Ubiquitination-independent degradation of proteins

Abstract

The present disclosure relates, at least in part, to methods and cells useful for promoting ubiquitination-independent protein degradation and for promoting protein stability using an engineered midnolin protein.

Description

UBIQUITINATION-INDEPENDENT DEGRADATION OF PROTEINSCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit under 35 U.S.C. § 119(e) of U.S. Provisional Application No. 63 / 515,548, filed July 25, 2023, entitled “UBIQUITINATION-INDEPENDENT DEGRADATION OF PROTEINS,” the entire disclosure of which is hereby incorporated by reference in its entirety.GOVERNMENT SUPPORT

[0002] This invention was made with government support under NS 115965 and AG011085 awarded by National Institutes of Health (NIH). The government has certain rights in the invention.REFERENCE TO AN ELECTRONIC SEQUENCE LISTING

[0003] The contents of the electronic sequence listing (H082470431WO00-SEQ-KVC.xml; Size: 46,566 bytes; and Date of Creation: July 25, 2024) is herein incorporated by reference in its entirety.BACKGROUND

[0004] Cells use ubiquitin to mark proteins for proteasomal degradation. While the proteasome also eliminates proteins that are not ubiquitinated, how this occurs mechanistically is unclear.SUMMARY

[0005] The present disclosure relates, at least in part, to the discovery that midnolin promotes the destruction of many nuclear proteins, including transcription factors encoded by the immediate-early-genes. Diverse stimuli induce midnolin and its overexpression is sufficient to cause the degradation of its targets by a mechanism that does not require ubiquitination. Instead, midnolin associates with the proteasome via an a helix, employs its Catch domain to bind a region within substrates that can form a P strand, and uses a ubiquitin-like domain to promote substrate destruction. Thus, midnolin contains three regions that function in concert to target a large set of nuclear proteins to the proteasome for degradation in a ubiquitination- independent manner.

[0006] Accordingly, aspects of the present disclosure relate to an engineered midnolin protein comprising: an aHelixC domain, a ubiquitin-like domain, and a Catch domain; wherein the Catch domain comprises Catch 1 subdomain and Catch 2 subdomain, wherein Catch 1 and Catch 2 subdomains are joined by a linker; and wherein the aHelixC domain, the ubiquitin-like domain, the Catch domain, Catch 1 subdomain, Catch 2 subdomain, and / or the linker is / are modified.

[0007] In some embodiments, the aHelixC domain further comprises a nuclear localization signal (NLS). In some embodiments, the NLS is modified.

[0008] In some embodiments, the engineered midnolin protein further comprises a localization signal.

[0009] In some embodiments, the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids longer than a wild type linker. In some embodiments, the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids shorter than a wild type linker.

[0010] In some embodiments, the engineered midnolin protein improves degradation of a target protein.

[0011] In some embodiments, the ubiquitin-like domain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type ubiquitin-like domain. In some embodiments, the amino acid substitution is L80E. In some embodiments, the amino acid substitution is G83E. In some embodiments, the ubiquitin-like domain is deleted from the engineered midnolin protein.

[0012] In some embodiments, the Catch 1 subdomain is deleted from the engineered midnolin protein. In some embodiments, the Catch 2 subdomain is deleted from the engineered midnolin protein. In some embodiments, the Catch 1 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 1 subdomain. In some embodiments, the amino acid substitution is V129D, L133D, S134, or L139D. In some embodiments, the Catch 2 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 2 subdomain. In some embodiments, the amino acid substitution is R301S, I306D, L310D, or L313D.

[0013] In some embodiments, the aHelixC domain is deleted from the engineered midnolin protein.

[0014] In some embodiments, the NLS is deleted from the engineered midnolin protein.

[0015] Aspects of the present disclosure relate to a method of promoting degradation of a target protein, the method comprising contacting the target protein with the engineered midnolin protein described herein. In some embodiments, the target protein comprises a P strand degron.In some embodiments, the Catch domain binds the P strand degron of the target protein. In some embodiments, the target protein is degraded by the proteasome after the Catch domain binds the P strand degron of the target protein. In some embodiments, degradation of the target protein is ubiquitination-independent.

[0016] In some embodiments, the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids longer than a wild type linker. In some embodiments, the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids shorter than a wild type linker. In some embodiments, modification to the linker affects binding specificity to the target protein.

[0017] Aspects of the present disclosure relate to a method of promoting stability of a target protein, the method comprising expressing in a cell a gene encoding the engineered midnolin protein described herein. In some embodiments, promoting stability of a target protein can be measured as increased half-life of the target protein. In some embodiments, promoting stability of a target protein can be measured as decreased degradation of the target protein.

[0018] In some embodiments, the ubiquitin-like domain of the engineered midnolin protein comprises an amino acid sequence comprising an amino acid substitution relative to a wild type ubiquitin-like domain. In some embodiments, the amino acid substitution is L80E. In some embodiments, the amino acid substitution is G83E. In some embodiments, the ubiquitin-like domain is deleted from the gene encoding the engineered midnolin protein.

[0019] In some embodiments, the Catch 1 subdomain is deleted from the gene encoding the engineered midnolin protein. In some embodiments, the Catch2 subdomain is deleted from the gene encoding the engineered midnolin protein. In some embodiments, the Catch 1 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 1 subdomain. In some embodiments, the amino acid substitution is V129D, L133D, S134, or L139D. In some embodiments, the Catch 2 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 2 subdomain. In some embodiments, the amino acid substitution is R301S, I306D, L310D, or L313D.

[0020] In some embodiments, the aHelixC domain is deleted from the gene encoding the engineered midnolin protein.

[0021] In some embodiments, the NLS is deleted from the gene encoding the engineered midnolin protein.

[0022] In some embodiments, the gene encoding midnolin (e.g., wild type or engineered midnolin) is expressed heterologously in the cell. In some embodiments, the gene encodingmidnolin (e.g., wild type or engineered midnolin) is expressed on a plasmid. In some embodiments, the gene is integrated into the genome of the cell.

[0023] Aspects of the present disclosure relate to a method of improving stability of a target protein, the method comprising contacting a cell with a small molecule that inhibits expression of midnolin in the cell.

[0024] Aspects of the present disclosure relate to a method of improving stability of a target protein, the method comprising deleting a gene encoding midnolin from the genome of a cell. In some embodiments, the step of deleting is accomplished using a genetic engineering technique. In some embodiments, the genetic engineering technique comprises a CRISPR / Cas9 system.

[0025] Aspects of the present disclosure relate to an engineered cell comprising the engineered midnolin protein described herein.

[0026] In some embodiments, the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids longer than a wild type linker. In some embodiments, the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids shorter than a wild type linker.

[0027] In some embodiments, the engineered midnolin protein improves degradation of a target protein.

[0028] In some embodiments, the ubiquitin-like domain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type ubiquitin-like domain. In some embodiments, the amino acid substitution is L80E. In some embodiments, the amino acid substitution is G83E. In some embodiments, the ubiquitin-like domain is deleted from the engineered midnolin protein.

[0029] In some embodiments, the Catch 1 subdomain is deleted from the engineered midnolin protein. In some embodiments, the Catch 2 subdomain is deleted from the engineered midnolin protein. In some embodiments, the Catch 1 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 1 subdomain. In some embodiments, the amino acid substitution is V129D, L133D, S134, or L139D. In some embodiments, the Catch 2 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 2 subdomain. In some embodiments, the amino acid substitution is R301S, I306D, L310D, or L313D.

[0030] In some embodiments, the aHelixC domain is deleted from the engineered midnolin protein.

[0031] In some embodiments, the NLS is deleted from the engineered midnolin protein.

[0032] In some embodiments, the target protein is relevant to a disease. In some embodiments, the disease is cancer, and immunological disease, or a brain disease. In some embodiments, the target protein is a transcription factor. In some embodiments, the target protein is IRF4 or an immediate early gene. In some embodiments, the immediate early gene is c-Fos, FosB, ERGl, or NR4A1.

[0033] Aspects of the present disclosure relate to a method of promoting degradation of a target protein, the method comprising increasing in a cell expression of a gene encoding midnolin.BRIEF DESCRIPTION OF THE DRAWINGS

[0034] FIGs. 1A-1E show genetic screens reveal midnolin as a regulator of IEG protein degradation. FIG. 1A shows a schematic of the FACS-based genome- wide CRISPR-Cas9 screens using the Global Protein Stability (GPS) reporter of IEG proteins in HEK-293T cells (created with BioRender.com). FIGs. 1B-1C show results of the genetic screens revealed MIDN as the top hit for negatively regulating the stability of both EGR1 and FosB. The proteasomal components showed a weaker enrichment likely due to their essentiality. The MAGeCK score represents the negative loglO of the “pos|score” value generated from MAGeCK (56). FIGs. 1D- 1E show losing midnolin stabilized, while overexpressing midnolin destabilized EGR1 and FosB. GPS EGR1 or FosB reporters were stably expressed in wild type and two independent MIDN KO HEK-293T single cell clones. Vectors expressing BFP control alone (no asterisk) or midnolin and BFP from a EFla promoter (labeled with asterisk) was transiently reintroduced by transfection before analyzing the GFP / DsRed ratio by flow cytometry.

[0035] FIGs. 2A-2F show Midnolin is induced and promotes the degradation of several IEG proteins in physiological settings. FIG. 2A shows the loss of midnolin increased the expression of IEG proteins in NIH / 3T3 cells. Immunoblotting was performed from NIH / 3T3 cells stably expressing Cas9 and control or MIDN targeting single guide RNAs. This population-level mutagenesis of MIDN may show lower penetrance relative to an isogenic mutant since the knockout efficiency is dependent on the efficacy of the single guide RNA. The cells were starved of serum overnight before serum restimulation for the indicated time points. Asterisks mark nonspecific cross-reactive proteins. FIG. 2B shows overexpressing midnolin decreased the expression of IEG proteins in NIH / 3T3 cells. Same assay as FIG. 2A but in NIH / 3T3 cells stably overexpressing an N-terminally 2xFLAG tagged human midnolin using a CMV promoter. FIG. 2C shows the loss of midnolin increased the expression of IEG proteins in primary cortical neurons. Neurons were isolated from E16.5 mouse brains and cultured in a dish. On day 3 postisolation, the neurons were infected with lentivirus encoding Cas9 with control or MIDNtargeting single guide RNAs. Immunoblotting was performed on day 11 post dissection from neurons that were silenced overnight with tetrodotoxin (TTX, a sodium channel blocker) and D- AP5 (a NMDA receptor antagonist) and stimulated with KC1 for the indicated time points to induce depolarization. FIG. 2D shows overexpressing midnolin decreased the expression of IEG proteins in primary cortical neurons. Similar assay as FIG. 2C but using lentivirus to overexpress a BFP control or human midnolin co-expressing BFP using an EFla promoter. FIGs. 2E-2F show qPCR analysis for mRNA levels of the indicated genes from FIG. 2E primary mouse cortical neurons that were KC1 stimulated or FIG. 2F from NIH / 3T3 cells that were serum restimulated for the indicated time points. Error bars represent the standard deviation from three biological replicates. Data were analyzed using an ordinary one-way ANOVA followed by Tukey’s multiple comparisons test where **** represents p < 0.0001.

[0036] FIGs. 3A-3G show midnolin can promote the degradation of numerous transcriptional regulators. FIG. 3A shows a schematic of midnolin GPS ORFeome screen. The GPS ORFeome library (-12,000 barcoded human ORFs tagged to GFP) was introduced into MIDN KO HEK- 293T and the library-expressing cells were transfected with BFP control or midnolin coexpressing BFP before FACS sorting the library into populations based on the GFP / DsRed ratio (created with BioRender.com). FIGs. 3B-3C show analysis of the GPS ORFeome screen showing the change in protein stability (APSI) between midnolin and BFP, which was calculated based on the change in read distribution of the barcoded ORFs. Approximately 5% of the library showed significant destabilization with APSI values less than -0.5. Several validated hits from the screen are shown in the boxed table. FIG. 3D shows gene set enrichment analysis (GSEA) based on the GPS ORFeome screen for molecular function. FIG. 3E shows validation of screen hits indicates their potent regulation by midnolin. GPS reporters for the indicated genes were stably expressed in MIDN KO HEK-293T cells and a control BFP or midnolin co-expressing BFP were transiently transfected before analyzing the GFP / DsRed ratio by flow cytometry. FIG. 3F shows Endogenous proteins of numerous screen hits are regulated by midnolin. Immunoblotting was performed from wild type, MIDN KO, and MIDN KO HEK-293T cells where midnolin expression was stably induced with doxycycline (100 ng / mL) for 2 days using a TRE promoter. Shown are putative midnolin targets based on the GPS ORFeome screen and negative controls (indicated by dagger symbol). FIG. 3G shows validation of midnolin-mediated degradation of endogenous IRF4 in Ramos B cells. Immunoblotting was performed from Ramos B cells expressing Cas9 and control or MIDN targeting single guide RNAs, or stably overexpressing midnolin using an EFla promoter.

[0037] FIGs. 4A-4E show midnolin associates with the proteasome to promote ubiquitination-independent degradation of bound substrates. FIG. 4A shows a 3xHA-tag wasintroduced at the N-terminus of the endogenous midnolin locus in HEK-293T cells using CRISPR-Cas9 initiated recombination. Cells were treated with MG 132 for 6 hours before immunoprecipitation of 3xHA-midnolin followed by mass spectrometry. The results revealed a large enrichment of the 26S proteasome and shown is a STRING analysis of the top coimmunoprecipitated proteins identified from the mass spectrometry. FIG. 4B shows midnolin co- immunoprecipitates the proteasome and IEG proteins endogenously. Immunoblotting was performed from anti-HA immunoprecipitants of endogenous 3xHA-midnolin from the knock-in HEK-293T cells treated with the indicated drugs for 6 hours. PM A was used to induce the transcription of lEGs. FIG. 4C shows endogenous midnolin protein levels are strongly increased by proteasomal inhibition but not by ubiquitin El inhibition. Immunoblotting was performed from wild type and 3xHA-midnolin knock-in HEK-293T cells treated with 10 pM MG132 or 500 nM TAK-243 for 6 hours. FIG. 4D shows lysine-dependent ubiquitination on substrates is not necessary for midnolin interaction. Immunoblotting was performed from anti-HA immunoprecipitants of HEK-293T cells that were transfected with the indicated constructs, either wild type or all lysine residues mutated to arginine residues (K to R). Cells were treated with 10 pM MG 132 for 6 hours. CBX8 serves as a negative control as it is not targeted by midnolin. FIG. 4E shows midnolin does not require lysine residues on substrates to promote degradation. Wild type and K to R mutant substrates were stably introduced into MIDN KO HEK-293T cells using a CMV promoter. Then, midnolin expression was induced using doxycycline (100 ng / mL) for 2 days using a TRE promoter before lysis and immunoblotting.

[0038] FIGs. 5A-5I show midnolin contains three regions that function in concert to promote proteasomal degradation of bound substrates. FIG. 5A shows midnolin structure prediction by AlphaFold (Q504T8-F1) reveals three regions with defined structure (26). FIG. 5B shows a schematic representation of mutations (57) or truncations introduced into the midnolin cDNA. See methods for the truncation boundaries and regions used for sufficiency experiments. FIG. 5C shows regions with defined structure are necessary for a functional midnolin. The GPS IRF4 reporter was stably expressed in MIDN KO HEK-293T cells and a control BFP or wild type and mutant versions of midnolin co-expressing BFP were transiently transfected before analyzing the GFP / DsRed ratio by flow cytometry. FIGs. 5D-5E show the midnolin Catch domain is necessary for binding substrates and the C-terminal a helix is necessary for proteasomal association. Immunoblotting was performed from anti-FLAG immunoprecipitants of HEK-293T cells stably expressing 2xFLAG-tagged midnolin using a CMV promoter. Cells were treated with 10 pM MG 132 for 6 hours. FIG. 5F shows the midnolin aHelix-C is sufficient to interact with the proteasome. Immunoblotting was performed from anti-FLAG immunoprecipitants of MIDN KO HEK-293T cells transfected with the indicated 2xFLAG-tagged proteins. FIG. 5G shows themidnolin Catch domain is sufficient to bind substrates. Immunoblotting was performed from anti-FLAG immunoprecipitants of MIDN KO HEK-293T cells transfected with the indicated 2xFLAG-tagged proteins. The 111 amino acid sequence between Catch 1 and Catch2 was shortened to 10 amino acids (ALoopl), 16 amino acids (ALoop2), or 28 amino acids (ALoop3). Cells were treated with 10 pM MG132 and 20 ng / pL of PMA for 6 hours. FIGs. 5H-5I show the Catchl and Catch2 regions of midnolin interact when expressed as independent proteins. Immunoblotting was performed from anti-HA immunoprecipitants of MIDN KO HEK-293T cells co-transfected with 2xHA-GFP-Catchl and 2xFLAG-MBP-Catch2 constructs, where “e” signifies empty 2xHA-GFP or 2xFLAG-MBP.

[0039] FIGs. 6A-6I show midnolin catches regions within its substrates that constitute a P strand degron. FIG. 6A shows an AlphaFold structure prediction of midnolin bound to its substrate IRF4 reveals an adopted P strand capture model. FIG. 6B shows that midnolin requires the predicted P strand within IRF4 to promote degradation. The GPS IRF4 reporters were stably expressed in MIDN KO HEK-293T cells and a control BFP or midnolin co-expressing BFP were transiently transfected before analyzing the GFP / DsRed ratio by flow cytometry. FIG. 6C shows the predicted P strands are necessary for interaction with midnolin. Immunoblotting was performed from anti-FLAG immunoprecipitants of 3xHA-midnolin knock-in HEK-293Ts transfected with 2xFLAG-tagged substrates. For FosB, the comparison is between the full-length protein and AFosB. Cells were treated with 10 pM MG132 for 6 hours. See methods for the truncation boundaries. FIG. 6D shows amino acid frequency of midnolin substrate P strands predicted by AlphaFold reveals a strong preference for hydrophobic residues. ‘Inward’ is defined by the residues buried within the Catch domain, while ‘outward’ is defined by the solvent- exposed residues. FIG. 6E shows the hydrophobicity of residues within the P strand was determined by a mean hydrophobicity index at pH 7 (58, 59) of residues immediately preceding, within, or immediately following the P strand. FIG. 6F shows an AlphaFold structure prediction of the midnolin Catch domain bound to IRF4. FIG. 6G shows hydrophobic P strand residues buried within the Catch domain are required for midnolin interaction. Similar assay as FIG. 6C from cells transfected with the 2xFLAG-tagged IRF4 constructs. FIG. 6H shows midnolin requires the hydrophobic P strand residues buried within the Catch domain to promote degradation. Similar assay as FIG. 6B, FIG. 61 shows regions encompassing predicted P strand(s) are sufficient for conferring an interaction with midnolin. Similar assay as FIG. 6C from cells transfected with the indicated 2xFLAG-GFP-peptide fusions.

[0040] FIG. 7 shows a model for how midnolin functions to promote ubiquitination- independent proteasomal degradation. Midnolin is induced by growth factors and neurological stimuli and its overexpression is sufficient to cause the degradation of its targets includingtranscription factors such as c-Fos, FosB, ERG1, NR4A1, IRF4, and potentially many other proteins within the nucleus, where midnolin primarily resides. The degradation of its substrates does not require ubiquitination. Instead, midnolin utilizes its Catch domain to bind unstructured hydrophobic regions within substrates that have the potential to form a P strand that functions as a midnolin degron. Midnolin associates with the proteasome using its long C-terminal a helix and promotes the destruction of Catch-bound substrates via its N-terminal ubiquitin-like domain. Structures of the midnolin domains are derived from AlphaFold predictions. How the C-terminal a helix of midnolin binds the proteasome, whether a conformational change occurs after substrate binding, and how the ubiquitin-like domain confers degradative activity require further investigation (created with BioRender.com).

[0041] FIGs. 8A-8E show validation of the CRISPR / Cas9 screens and mRNA expression changes. FIG. 8A shows GPS reporters for EGR1, NR4A1, and c-Fos were stably expressed in HEK-293T cells and BFP or midnolin co-expressing BFP were transiently transfected before analyzing the GFP / DsRed ratio by flow cytometry. FIG. 8B shows similar assay as FIG. 8A but gating the analysis for increasing levels of BFP (midnolin) for the GPS FosB reporter. FIG. 8C shows similar assay as FIG. 8A for GPS reporters of negative control nuclear proteins such as ATF2, CREB3, and CREB5. FIG. 8D shows midnolin overexpression does not grossly affect the mRNA levels of lEGs. qPCR analysis for mRNA levels of the indicated genes from primary mouse cortical neurons stably overexpressing a control BFP or midnolin co-expressing BFP that were KC1 stimulated for the indicated time points. Error bars represent the standard deviation from three biological replicates. Data were analyzed using a two-way ANOVA followed by Sidak’s multiple comparisons test where ns is not significant and *** represents a p < 0.001. FIG. 8E shows RiboTag immunoprecipitation followed by RNA seq data extracted from Mardinly et al., Nature 2016, highlighting MIDN and various lEGs are induced by light stimuli in the mouse visual cortex (11).

[0042] FIG. 9 shows validation of the midnolin GPS ORFeome screen. Validation of the GPS ORFeome screen showing the degradation of many putative midnolin substrates upon midnolin overexpression. A control BFP or midnolin co-expressing BFP were transfected into MIDN KO HEK-293T cells stably expressing the indicated GPS reporters before analyzing the GFP / DsRed ratio by flow cytometry. These proteins were extracted as possible midnolin substrates based on the GPS ORFeome screen.

[0043] FIGs. 10A-10B show midnolin does not require active ubiquitin El enzymes to function. FIG. 10A shows the BioPlex dataset reveals IEG proteins and the proteasomal component PSMD2 co-immunoprecipitate midnolin from HEK-293T cells (17). FIG. 10B shows an active ubiquitin El enzyme is not required for midnolin-dependent degradation. MIDN KOHEK-293T cells stably expressing the indicated GPS reporters for midnolin substrates were transfected with BFP or midnolin co-expressing BFP. The cells were then treated with DMSO, 10 pM MG132, or 500 nM TAK-243 for 6 hours before analyzing the GFP / DsRed ratio by flow cytometry. c-Myc serves as a negative control because it does not appear to be a midnolin substrate but remains targeted for ubiquitination-dependent degradation.

[0044] FIGs. 11A-11B show AlphaFold confidence and midnolin conservation. FIG. 11A shows an AlphaFold prediction of midnolin (Q504T8-F1) shaded by the predicted local distance difference test (pLDDT) as a measure of confidence or disorder (60). FIG. 11B shows an AlphaFold prediction of midnolin (Q504T8-F1) shaded by conservation using the ConSurf web server.

[0045] FIGs. 12A-12B show midnolin is localized primarily within the nucleus. FIG. 12A shows midnolin nuclear localization requires the C-terminal a helix. Anti-FLAG immunofluorescence was performed from HEK-293T cells stably expressing the indicated 2xFLAG-tagged midnolin constructs and treated with 10 pM MG132 for 6 hours. The Hoechst dye serves as a DNA / nucleus stain. FIG. 12B shows endogenous midnolin is localized primarily within the nucleus. Anti-HA immunofluorescence was performed from wild type unedited and knock-in 3xHA-midnolin HEK-293T cells treated with the indicated drugs for 6 hours.

[0046] FIGs. 13A-13C show further validation of the midnolin functional domains. FIG. 13A shows regions with defined secondary structure are crucial for a functional midnolin. The GPS FosB reporter was stably expressed in MIDN KO HEK-293T cells and a control BFP or wild type and mutant versions of midnolin co-expressing BFP were transiently transfected before analyzing the GFP / DsRed ratio by flow cytometry. FIG. 13B shows the interaction of midnolin with the proteasome is not affected by inhibitors of the proteasome or El ubiquitin-activating enzymes. Immunoblotting was performed from anti-FLAG immunoprecipitants of MIDN KO HEK-293T cells stably expressing 2xFLAG-tagged midnolin using a CMV promoter. Cells were treated with 10 pM MG132 or 500 nM TAK-243 for 6 hours. FIG. 13C shows the Catch domain is necessary for midnolin to interact with its substrates. Immunoblotting was performed from anti-FLAG immunoprecipitants of NIH / 3T3 cells stably expressing the indicated 2xFLAG- tagged midnolin constructs. The cells were serum starved overnight and serum restimulated in the presence and absence of 10 pM MG132 before lysis.

[0047] FIGs. 14A-14E show midnolin contains a Catch domain with a hydrophobic core that interacts with substrates. FIG. 14A shows the sequence between Catch 1 and Catch2 is not required to interact with substrates or the proteasome. Immunoblotting was performed from anti- FLAG immunoprecipitants of MIDN KO HEK-293T cells transfected with 2xFLAG-tagged midnolin constructs. See methods for truncation boundaries. Cells were treated with 10 pMMG 132 and 20 ng / pL PMA for 6 hours. FIG. 14B shows the sequence between Catch 1 and Catch2 is dispensable, while the hydrophobic residues that mediate the Catch 1 and Catch2 interaction are required for midnolin to promote substrate degradation. The GPS IRF4 or FosB reporter was stably expressed in MIDN KO HEK-293T cells and a control BFP and wild type or mutant versions of midnolin co-expressing BFP were transiently transfected before analyzing the GFP / DsRed ratio by flow cytometry. FIG. 14C shows an AlphaFold structure prediction and conservation of the Catch domain reveals a hydrophobic core. All residues within the core of the Catch domain are hydrophobic and the side chains are shown in the conservation figure generated using the ConSurf web server. FIG. 14D shows the hydrophobic residues that mediate the Catch 1 and Catch2 binding are required for the interaction of midnolin with its substrates. Similar assay as FIG. 14A. FIG. 14E shows the conserved hydrophobic core is necessary for the degradative function of midnolin. Similar assay as FIG. 14B. The immunoblot shows that all versions of midnolin express relatively evenly.

[0048] FIGs. 15A-15C show AlphaFold prediction of IRF4 bound to midnolin. FIG. 15A shows an AlphaFold structure prediction of IRF4 (AF-Q15306-F1), highlighting the unstructured region captured by the midnolin Catch domain with a dashed box. FIG. 15B shows an AlphaFold structure prediction of midnolin bound to IRF4, highlighting the same unstructured region that now forms a P strand conformation that functions as a midnolin degron. FIG. 15C shows the predicted local distance difference test (pLDDT) scores for residues that interact with the midnolin Catch domain were averaged and compared with the average of the rest of the substrate as an approximation of disorder in the original substrate not bound by midnolin (paired t-test, p = 2.72xl0'9).

[0049] FIGs. 16A-16H show AlphaFold-predicted P strand regions are necessary for capture by midnolin. Midnolin requires adopted P strands within substrates to promote degradation. GPS reporters for the indicated AlphaFold midnolin-substrate predictions were stably expressed in MIDN KO HEK-293T cells that were transfected with BFP or midnolin co-expressing BFP before analyzing the GFP / DsRed ratio by flow cytometry. FIGs. 16A-16B show the A P strand is defined by a deletion of residues 245-260 of FOXS1. FIGs. 16C-16D show the A P strand is defined by a deletion of residues 538-558 of CBX4. FIGs. 16E-16F show the A P strand is defined by a deletion of residues 277-288 of NeuroDl. FIGs. 16G-16H show the A P strand is defined by a deletion of residues 314-328 of SPINDOC (C11ORF84).

[0050] FIGs. 17A-17H show further validation of AlphaFold-predicted P strand regions captured by midnolin. FIG. 17A shows AFosB, a naturally expressed splice isoform, shows resistance to midnolin-promoted degradation. A BFP control or midnolin co-expressing BFP were transfected into HEK-293T cells stably expressing the indicated FosB GPS reporters beforeanalyzing the GFP / DsRed ratio by flow cytometry. See methods for more details regarding the truncation boundaries. FIG. 17B shows FosB contains multiple regions for midnolin capture. Similar assay as FIG. 17A, AFosB is defined by truncation of the last 101 amino acids, while AN P strand is defined by a deletion of residues 67-75 of FosB. FIG. 17C shows related to FIG. 6E. The hydrophobicity of residues comprising the P strand was determined by a two-sided paired t- test (p = 2.1xl0'69) between the average hydrophobicity index at pH 7 of residues within the P strand, with that of residues comprising the rest of the substrate. FIG. 17D shows similar analysis as FIG. 17C but comparing the P strand residues buried within the Catch domain and the residues pointing outward that are solvent exposed (p = 4.6xl0'28). FIG. 17E shows an AlphaFold structure prediction of the midnolin Catch domain bound to EGR1, highlighting the hydrophobic residues of EGR1 that are buried within the core of the Catch domain. FIG. 17F shows hydrophobic P strand residues buried within the Catch domain are required for midnolin capture. Immunoblotting was performed from anti-FLAG immunoprecipitants of 3xHA-midnolin knock-in HEK-293T cells transfected with 2xFLAG-tagged EGR1 constructs. Cells were treated with 10 pM MG132 for 6 hours. FIG. 17G and FIG. 17H show regions encompassing predicted P strand(s) are sufficient for midnolin to promote degradation. Similar assay as FIG. 17A.The IRF4 peptide is defined by the region spanning 192-248; the EGR1 peptide is defined by the region spanning 113-172; the FosB peptide is defined by the region spanning 238-338 amino acids.

[0051] FIGs. 18A-18B show the validation of midnolin clonal knockout cell lines and Proteintech antibody. FIG. 18A shows regions flanking the sgRNA cut site within the MIDN gene locus were PCR amplified, run on an agarose gel, and the PCR product was sequenced using Next-Generation DNA Sequencing to determine the allele identity. FIG. 18B shows an immunoblot from wild type and MIDN knockout clone #1. Cells were pre-treated for 6 hours with 10 pM MG132.

[0052] FIGs. 19A-19B show the Catch domain and substrate binding by the Catch domain. FIG. 19A shows the Catch 1 subdomain and Catch2 subdomain of the Catch domain. FIG. 19B shows the Catch domain binding IRF4, a midnolin substrate.

[0053] FIG. 20 shows an AlphaFold prediction of the Catch domain.

[0054] FIGs. 21A-21E show purification of midnolin-proteasome complexes. FIG. 21A shows that midnolin contains three domains necessary for its degradative function: a ubiquitin- like (Ubl) domain, a Catch domain, and a C-terminal helix (ocHelix-C) that bears a nuclear localization sequence. FIG. 21B shows a schematic representation of the process to affinity- purify the midnolin-proteasome complex by size-exclusion chromatography from HEK-293Tcells transiently overexpressing 2xFLAG-tagged midnolin variants. FIG. 21C shows a representative size-exclusion chromatography trace indicating the fractions collected corresponding to the midnolin-proteasome complex. FIG. 21D shows a representative Coomassie stain of the purified midnolin-proteasome complex showing the characteristic migration pattern of the proteasomal subunits after SDS-PAGE. FIG. 21E shows a representative negative stain electron micrograph of the midnolin-proteasome complex.

[0055] FIG. 22 shows cryogenic electron microscopy (cryo-EM) data processing for the midnolin-proteasome complex. Summary of cryo-EM data processing scheme for midnolin- proteasome complexes showing a representative micrograph, 2D classes, and 3D reconstructions. The four datasets derive from four midnolin-proteasome purifications as outlined in the methods below.

[0056] FIGs. 23A-23C show a cryo-EM map and model quality. FIG. 23A shows a courier shell correlation (FSC) vs. resolution (1 / A) curves for the indicated cryo-EM maps. FIG. 23B shows model vs. map FSC curves. FIG. 23C shows that the indicated cryo-EM map is shaded by local resolution.

[0057] FIG. 24 shows the structure of the midnolin-proteasome complex. Cryo-EM maps of the midnolin-proteasome complex in two conformations. The Ml state showed density for both the midnolin Ubl domain and ocHelix-C engaged with the 19S regulatory particle as well as low- resolution density for the Catch domain. The M2-A state contained density only for ocHelix-C.

[0058] FIG. 25 shows ATPase ring configurations of midnolin-proteasome complexes. The density of the ATPase ring observed from the 20S core particle. The central pore of the ATPase ring was closed in the Ml state and open in the M2 state with distinguishable density corresponding to a translocating substrate. A schematic that shows the configuration of the M2 ATPase subunits engaged with a substrate. The nucleotide status of each subunit is indicated by T = ATP, D = ADP, blank = apo.

[0059] FIGs. 26A-26J show a mechanism for assembling the midnolin-proteasome complex in the nucleus. FIG. 26A show an atomic model of midnolin ocHelix-C bound to PSMD2 / Rpnl. The ‘M site’ interacts directly with ocHelix-C and is distinct from the T1 and NT sites that bind ubiquitin. The midnolin nuclear localization sequence (NLS) mediates a part of the interaction. FIG. 26B shows immunoblotting of anti-FLAG immunoprecipitants of M1DN knockout HEK- 293T cells transiently overexpressing 2xFLAG-midnolin using a CMV promoter. FIG. 26C shows anti-FLAG immunofluorescence of M1DN knockout HEK-293T cells stably expressing 2xFLAG-midnolin. Cells were treated with 10 pM MG132 for 4 hours. FIG. 26D shows the same assay as FIG. 26C but the cells were transfected with control or different KPNB1 -targetingsiRNAs for 48 hours. FIG. 26E shows the same assay as FIG. 26B, but the cells were transfected with 2xFLAG-maltose-binding protein (MBP) fused to ocHelix-C. FIG. 26F shows the same assay as FIG. 26C using HEK-293T cells stably expressing 2xFLAG-MBP-ocHelix-C but were not pre-treated with MG 132. FIG. 26G shows in vitro co-immunoprecipitation followed by immunoblotting. Purified FLAG-MsyB -ocHelix-C were immobilized onto anti- FLAG beads and incubated with pure KPNB1, KPNA2 (FIG. 26H), human proteasomes (FIG. 261), or human proteasomes with either KPNB1 or KPNA2 to test for competition (FIG. 26 J).

[0060] FIGs. 27A-27B show the midnolin-PSMD2 interaction is evolutionarily conserved as predicted by AlphaFold-multimer. FIG. 27A shows AlphaFold-multimer predictions of full- length midnolin and PSMD2 / Rpnl using orthologous proteins. FIG. 27B shows the predicted aligned error (PAE) plots of the AlphaFold-multimer predictions suggest the interaction is high confidence. ocHelix-C interacts with the M site of PSMD2 via a hydrophobic interface on the N- terminal half of ocHelix-C and electrostatic interactions on the C-terminal half of the ocHelix-C analogously to what was observed by cryo-EM using the human proteins. The electrostatic interactions were primarily mediated by the nuclear localization sequence. This suggests that the mechanism to assemble the midnolin-proteasome complex primarily in the nucleus is conserved throughout evolution.

[0061] FIGs. 28A-28D show validation of the binding between ocHelix-C and PSMD2 or KPNB1. FIG. 28A shows AlphaFold-multimer prediction of full-length midnolin with PSMD2 / Rpnl. The M site represents PSMD2 residues that make direct contact with ocHelix-C. The M site does not overlap with the T1 and NT surfaces that bind ubiquitin but partially overlaps with the T2 site that binds USP14. The arginine residues within the midnolin nuclear localization sequence (NLS) mediate a part of the interaction. FIG. 28B shows that M1DN KO HEK-293T cells were first reconstituted with 2xFLAG-midnolin from a CMV promoter using lentivirus. These cells were then transfected with 2xHA-PSMD2. Shown is immunoblotting from anti-HA immunoprecipitates. FIG. 28C shows that 2xFLAG-midnolin was immunoprecipitated from the same, non-transfected cells as FIG. 28B, and the immunoprecipitates were subjected to liquid chromatography-mass spectrometry to identify midnolin interacting proteins. Shown are the top co-immunoprecipitated proteins, highlighting the proteasome, importin |3 (KPNB1), and the nuclear pore complex. FIG. 28D shows that immunoblotting was performed from anti-FLAG immunoprecipitants of M1DN KO HEK-293T cells that were transiently overexpressing 2xFLAG-tagged midnolin using a CMV promoter.

[0062] FIGs. 29A-29E show validations of siRNAs and protein purification. FIG. 29A shows anti-FLAG immunofluorescence of M1DN KO HEK-293T cells stably expressing 2xFLAG-tagged midnolin using a CMV promoter. Cells were transfected with control or different KPNA2-targeting siRNAs for 48 hours before treatment with 10 pM MG132 for 4 hours. FIG. 29B shows that HEK-293T cells were transfected with control, KPNB1-, or KPNA2-targeting siRNAs for 48 hours before assaying gene expression by immunoblotting. FIG. 29C shows Coomassie stains of the recombinant importin proteins, and FIG. 29D the recombinant FLAG- MsyB-ocHelix-C proteins that were purified by size-exclusion chromatography (SEC). The elutions represent the 6His-tagged proteins that were first affinity-purified with a Ni-NTA resin before fractionation by size-exclusion chromatography. The concentrated SEC Peak represents the purified protein used in the in vitro binding experiments. FIG. 29E shows Coomassie stain of purified human proteasomes showing the characteristic migration pattern of the proteasomal subunits after SDS-PAGE.

[0063] FIGs. 30A-30B show that the midnolin-PSMD14 interaction is evolutionarily conserved as predicted by AlphaFold-multimer. FIG. 30A shows AlphaFold-multimer predictions of full-length midnolin and PSMD14 / Rpnl l using orthologous proteins. FIG. 30B shows the predicted aligned error (PAE) plots of the AlphaFold-multimer predictions suggest the midnolin-PSMD14 interaction is high confidence. The hydrophobic linker residues between the Ubl and Catch are buried in PSMD14 analogously to what was observed by cryo-EM for the human proteins.

[0064] FIGs. 31A-31C show the midnolin ubiquitin-like domain positions the Catch domain above the substrate entry site. FIG. 31A shows an atomic model of the midnolin ubiquitin-like domain bound to the deubiquitinase PSMD14 / Rpnll. In comparison is the structure of ubiquitin bound to Rpnll during substrate translocation (PDB: 6EF3) (21). FIG. 31B shows M1DN knockout HEK-293T cells stably expressing a dual-fluorescence IRF4 stability reporter were transfected with control BFP or wild-type and Ubl-mutant midnolin co-expressing BFP using an EF-loc promoter. The cells were analyzed for the GFP / DsRed ratio two days post-transfection by flow cytometry. FIG. 31C shows AlphaFold3 prediction of D. gyrociliatus midnolin-PSMD2- PSMD14-ATPase (right) resembles the cryo-EM model of human Ml midnolin-PSMD2- PSMD14-ATPase complex (left).

[0065] FIG. 32 shows AlphaFold3 prediction of human midnolin bound to the proteasome. AlphaFold3 prediction of human midnolin-PSMD2-PSMD14-ATPase subunits (right) resembles the cryo-EM model of the human Ml state (left). The Catch domain is positioned right above the ATPase ring.

[0066] FIGs. 33A-33D show that Dimorphilus gyrociliatus evolved a miniaturized midnolin that functions in human cells. FIG. 33A shows AlphaFold prediction of a minimal midnolinfound in Dimorphilus gyrociliatus (segmented worm) that contains only the three functional domains: Ubl, Catch, and ocHelix-C. FIG. 33B shows anti-FLAG immunofluorescence of M1DN KO HEK-293T cells stably expressing 2xFLAG-tagged midnolin variants. Cells were pre-treated with 10 pM MG132 for 4 hours. FIG. 33C shows immunoblotting from anti-FLAG immunoprecipitations of M1DN KO HEK-293T cells transiently overexpressing 2xFLAG-tagged midnolin variants using a CMV promoter. FIG. 33D shows M1DN KO HEK-293T cells stably expressing a dual-fluorescence EGR1 stability reporter were transfected with control BFP or midnolin co-expressing BFP using an EF-loc promoter. The cells were analyzed for the GFP / DsRed ratio two days post-transfection by flow cytometry.

[0067] FIGs. 34A-34F show that the Catch domain can enhance the substrate interaction by forming an FG zipper. FIG. 34A shows a 2.5 A resolution crystal structure of an EGRl-Catch fusion protein. Beyond the hydrogen bonding between |3 strands, the interaction between EGR1 and the Catch domain occurs by alternating phenylalanine-glycine residues termed an FG zipper. FIG. 34B shows immunoblotting of anti-FLAG immunoprecipitants of HEK-293T cells expressing endogenous 3xHA-midnolin and transiently overexpressing EGRl-2xFLAG using a CMV promoter. Cells were treated with 10 pM MG132 for 4 hours. FIG. 34C shows AlphaFold- multimer prediction of IRF4 with midnolin reveals an incomplete FG zipper with valine 216 instead of a tyrosine. Same assay as FIG. 34B but from cells transfected with 2xFLAG-IRF4. FIG. 34D shows AlphaFold- multimer prediction of NeuroDl with midnolin reveals an incomplete FG zipper with isoleucine 279 and phenylalanine 285 that sterically clashes with phenylalanine 280 of midnolin. Same assay as FIG. 34B but from cells transfected with 2xFLAG- NeuroDl. FIG. 34E shows AlphaFold- multimer prediction of SPINDOC with midnolin reveals a lack of an FG zipper with glycine 319, leucine 321, and leucine 323 in place of tyrosine, glycine, and phenylalanine, respectively. Same assay as FIG. 34B but from cells transfected with 2xFLAG-SPINDOC. FIG. 34F shows the same assay as FIG. 34B. MIDN knockout HEK-293T cells were reconstituted with wild-type and zipper-swapped 2xHA- midnolin from a CMV promoter using lentivirus.

[0068] FIGs. 35A-35D show the purification and structure determination of the EGRl-Catch domain fusion. FIG. 35A shows Coomassie stain of the recombinant EGRl-Catch fusion protein, which was first affinity enriched and then purified by size-exclusion chromatography. FIG. 35B shows the EGRl-Catch domain crystals harvested for data acquisition. FIG. 35C shows summary of statistics related to the EGRl-Catch domain crystal structure. FIG. 35D shows the EGRl-Catch crystal structure showing density for the FG zipper residues.

[0069] FIG. 36 shows that the FG zipper residues of the Catch domain are evolutionarily conserved. The Catch domains of orthologous midnolin proteins were extracted from the AlphaFold web server. FG zipper residues and their location within the Catch domain are highly conserved during evolution.

[0070] FIG. 37 shows a model for how the midnolin-proteasome pathway degrades nuclear proteins independently of ubiquitination. When midnolin is synthesized, ocHelix-C is occupied with importin |3 (KPNB1), preventing proteasomal association in the cytosol and leading to nuclear transport. The high nuclear concentration of Ran-GTP liberates midnolin from KPNB 1 and midnolin can associate with the 19S regulatory particle through PSMD2 / Rpnl of nuclear proteasomes. The midnolin ubiquitin-like domain (Ubl) binds to the proteasomal component PSMD14 / Rpnl 1, which is normally required to cleave the ubiquitin chain from substrates but here serves as a receptor. The Ubl positions the Catch domain above the proteasomal entry site to initiate degradation (Ml state). After substrate translocation (M2 state), midnolin could then be reused to deliver other substrates. Alternatively, midnolin could be degraded due to proximity to the ATPase ring during the delivery process.DETAILED DESCRIPTION

[0071] In mammals, extracellular growth factors, cytokines, neurotrophic factors, and neurotransmitters bind their cognate receptors and activate rapid responses by inducing post- translational modifications of pre-existing proteins. More delayed responses are also induced by stimulating gene transcription within the nucleus (1, 2). This transcriptional response occurs in two steps. First, within minutes of the initial stimulus, a set of genes termed immediate-early- genes (lEGs) is activated. The lEGs encode transcription factors that then trigger a second wave of late-response-gene (LRG) expression that mediates the cellular response to the initial stimulus. The IEG response is activated in a stereotypical fashion in virtually every cell type in the body but induces programs of LRG expression that are cell-type specific (3). The plethora of cellular responses regulated by lEGs include the cell cycle re-entry of quiescent fibroblasts during wound healing, the activation of immune cells in response to cytokines, bacterial, and viral infections, and the adaptive responses of neurons to neurotransmitters during learning and memory (3).

[0072] The IEG family encodes transcription factors such as Fos, Egr, and Nr4a, that are rapidly and transiently induced in response to a wide range of extracellular stimuli (4). The IEG mRNAs accumulate to a high level, and once these mRNAs are translated, the IEG proteins undergo rapid decay (5). Thus, the protein stability of the IEG program is tightly controlled to allow for a relatively brief burst of protein expression that is crucial for appropriate cellularresponses to various stimuli. Mis-regulation of the signaling networks that control IEG expression can lead to cancer, immune deficiencies, and neurological disorders (3).

[0073] While the mechanisms that regulate IEG transcription are well characterized, it is unclear how IEG proteins are degraded. In many cases, conjugation of ubiquitin onto substrate proteins occurs as a prelude to their destruction by the proteasome. c-Fos and FosB have been reported to be targeted to the proteasome by both ubiquitination-dependent and -independent mechanisms, but the molecular events that orchestrate these processes are unknown (6, 7).

[0074] The present disclosure relates to the identification of a protein called midnolin that targets stimulus-induced transcription factors, such as c-Fos, FosB, EGR1, NR4A1, and IRF4, to the proteasome for degradation. Disrupting midnolin function in various cell types increases the peak abundance of these transcription factors and prolongs their expression. It is not uncommon for a protein to be targeted for proteasomal degradation by multiple mechanisms and these stimulus-responsive transcription factors may not be targeted for degradation solely by midnolin. Rather, it is possible that these proteins are also targeted by a ubiquitination-dependent pathway that functions in parallel with midnolin and could be the primary mechanism of degradation in some contexts. Indeed, it has been reported that Fos family members are targeted by both ubiquitination-dependent and -independent mechanisms (7). It was found that midnolin is induced by diverse stimuli, and its induction may act as a post-translational feedback circuit to limit the time course of expression of these stimulus-responsive transcription factors. Through a gain-of-function genetic screen, a large group of potential targets of midnolin was identified that is strongly enriched for nuclear proteins, especially transcriptional regulators, revealing that midnolin functions broadly to promote the degradation of proteins in the nucleus where midnolin itself is predominantly localized. In some embodiments, a midnolin can be a midnolin from different species, e.g., mammal (e.g., human, mouse, monkey), or segmented worms (e.g., Dimorphilus gyrociliatus).

[0075] The biological functions of midnolin are likely complex. Many of its substrates play central roles in the nervous and immune systems, and thus it will be important to establish the physiological function of midnolin in vivo. Midnolin was initially discovered due to its strong induction in the midbrain early during embryonic development (39). How midnolin expression is induced both during development and in cultured cells is currently unclear. Midnolin may have evolved to integrate various upstream stimuli and shape the proteome swiftly as a crucial response to the initial stimuli. Many midnolin substrates including IEG proteins and lineagespecific transcription factors undergo a transient burst of expression either in response to a particular stimulus or during a specific stage of development, and thus the prolonged expression of these transcription factors could be detrimental to organismal physiology. Indeed, previousstudies have shown that chronic and repeated exposure to drugs of abuse such as cocaine leads to week-long accumulation of AFosB, which plays a crucial role for addiction in the nucleus accumbens of the brain (35-37). The finding that AFosB is resistant to midnolin-dependent degradation provides a glimpse into the role that midnolin could play in brain function. Given that many IEG proteins are efficiently targeted for degradation by midnolin and that the precise expression of lEGs is critical for learning and memory (1, 3, 40), it is possible that disrupting or boosting midnolin function could impact the ability of animals to efficiently learn and store information in the brain.

[0076] IRF4, another midnolin substrate, is a lineage- specific transcription factor that is essential for the function and homeostasis of mature B and T cells and is an oncogenic driver of multiple myeloma (15, 16, 41). IRF4 protein expression is potently induced by diverse immunological stimuli including T cell receptor signaling, cytokines, and in fasting adipocytes (42-45). How the IRF4 protein returns to baseline after stimulation was unknown, and the finding that IRF4 is potently targeted for degradation by midnolin may provide insights into the function of midnolin in the immune system. Taken together, these findings suggest that midnolin may serve as a key regulator that determines the precise expression kinetics of stimulus-induced transcription factors by controlling their protein stability in various cell types or tissues. In principle, midnolin could be used to target different proteins under different circumstances. For example, in flies, it was reported that a midnolin ortholog, Stuxnet, is cell cycle regulated and promotes Pc protein degradation in mitosis during development (14).

[0077] Midnolin was found to promote the degradation of its targets in a proteasome- dependent, but ubiquitination-independent manner. This is supported by multiple lines of evidence. Removal of lysine residues from many midnolin targets did not abrogate the ability of midnolin to bind with and promote their destruction. A caveat to this interpretation is that ubiquitination can occur on the N-terminal amine of the first amino acid if it is not acetylated, or on other residues such as cysteine, serine, or threonine more rarely (24). However, in contrast to the effects of proteasomal inhibition, blocking the El ubiquitin-activating enzyme did not abrogate midnolin function while stabilizing canonical substrates of the ubiquitin-proteasome system. Together, these experiments allowed us to conclude that midnolin does not require ubiquitination for its degradative function. Instead, midnolin engages substrates using its Catch domain, which binds a hydrophobic region capable of P strand formation that functions as a degron. Midnolin associates with the proteasome using its long, C-terminal a helix and contains an N-terminal ubiquitin-like domain that is essential to promote the degradation of bound substrates. A model is hypothesized whereby the concerted action of these three regions ofmidnolin allows for the ubiquitination-independent proteasomal degradation of midnolin-bound proteins (FIG. 7).

[0078] Previous work demonstrated that ornithine decarboxylase (ODC) is targeted directly to the proteasome for degradation via a C-terminal unstructured sequence (46-49). However, the regions within substrates captured by midnolin are unlike any others previously described for E3 ubiquitin ligases. The midnolin degron appears to be generally unstructured but has potential to form a P strand upon binding midnolin, with aliphatic residues of the degron buried within the hydrophobic core of the Catch domain. The integration of the P strand degron into the Catch domain appears to complete a five-stranded anti-parallel P sheet structure. The somewhat degenerate nature of these hydrophobic degrons in various substrate proteins is likely explained by the fact that P sheets are stabilized not only by side chain interactions, but also by backbone hydrogen bonding (50), which may reduce the need for specific amino acid interactions. A region with the propensity to form a P strand that is also biochemically compatible with the hydrophobic core of the Catch domain may be present in many proteins, thereby explaining how midnolin recognizes a diverse set of proteins. Given the ubiquity of such P strands as structural elements of proteins, midnolin recognition of unfolded proteins could behave as a general quality control mechanism.

[0079] How midnolin initiates the degradation of bound substrates is not completely understood mechanistically. It is not known if midnolin interacts with its targets before binding to the proteasome, or whether midnolin associates with the proteasome constitutively and then recruits its substrates, thus defining a new subclass of proteosomes in the nucleus. A unique feature of midnolin may be that it interacts stably with the proteasome using a C-terminal a helix but not its Ubl domain. This is unlike the processivity factors Rad23 and Ubiquilin that bind ubiquitinated cargo and utilize their Ubl domain to interact with the proteasome (29, 30, 51-53). How the midnolin a helix binds the proteasome is currently unclear. However, the Ubl domain of midnolin is necessary for the degradation of bound substrates and it is noteworthy that midnolin itself is efficiently degraded by the proteasome in a ubiquitination-independent manner that requires the Ubl domain. Whether midnolin is reused or being degraded along with the substrate remains to be determined. It is hypothesized that when midnolin interacts with a substrate, a conformational change occurs that allows the proteasome to recognize the Ubl domain transiently to promote the degradation of the bound substrate, with or without midnolin also being degraded. Structural and biochemical analyses are required to deepen the understanding of this non- canonical docking and degradative mechanism.

[0080] The results presented herein suggest the midnolin-proteasome pathway may represent a general mechanism by which the proteasome bypasses the traditional ubiquitination system toachieve selective degradation of many nuclear proteins. It has been reported that in bacteria, which do not contain the ubiquitin-proteasome system, a hierarchy of adaptors mediate selective degradation of diverse proteins by the proteasome-equivalent ClpXP protease complex (54, 55). Thus, it will be of interest to determine whether, in the course of evolution, additional proteins have evolved to recruit proteins directly to the proteosome for degradation.

[0081] The present disclosure relates, in part, to the discovery that midnolin can promote the degradation of many transcription factors within the nucleus, where midnolin itself resides, via a non-canonical ubiquitination-independent mechanism. Midnolin interacts with its substrates using a structure termed the Catch domain, which binds to disordered regions within substrates that fold into a |3-strand, completing a |3-shcct tertiary structure (FIG. 19). Furthermore, midnolin stably associates with the proteasome and directly promotes the degradation of Catch bound substrates. This midnolin-proteasome pathway represents the first identified general mechanism whereby cells bypass the traditional ubiquitination system to achieve selective protein degradation.

[0082] The Catch domain is unique to midnolin as no other protein with a similar architecture was identified. With a hydrophobic core (FIG. 20), the midnolin Catch domain contains an internal hydrophobic pocket that might be used to tune the specificity of certain substrate(s). Indeed, what makes the Catch domain remarkable is that it can promiscuously bind many transcription factors (EGR1, c-Fos, FosB, IRF4), while retaining a degree of selectivity. For example, certain transcription factors bind the Catch domain with higher affinity compared to other transcription factors. The present disclosure shows that the Catch domain is not only necessary to interact with midnolin substrates, but it is also completely sufficient (FIG. 5G). In this experiment, evidence is provided that the Catch domain is engineerable because manipulating the length of the linker sequence connecting the Catch 1 and Catch2 subdomains tunes the substrate specificity.

[0083] Engineering the Catch domain, either pharmacologically or genetically, is an intriguing strategy to target traditionally undruggable transcription factors for degradation. The Catch domain has the potential to target many oncogenic transcription factors like IRF4 that drive diseases like multiple myeloma. Thus, the Catch domain could be engineered to increase the binding affinity towards a set of disease-relevant transcription factors. Conversely, it could also be beneficial to engineer a version of the Catch domain that has particularly low substrate specificity and is thus capable of binding many transcription factors promiscuously. One could use these engineered midnolin proteins as gene therapy to promote the degradation of many transcription factors simultaneously and sensitize cancer cells to death. Small molecule screenscould also be performed to identify chemicals that increase the affinity of the Catch domain to certain substrates, potentially through the hydrophobic pocket.

[0084] On the other hand, the immediate early gene encoded proteins (lEGs) such as c-Fos, FosB, EGR1, and NR4A1 are exceptional midnolin targets. The lEGs regulate proper learning and memory at an organismal level and high levels of c-Fos in hippocampal neurons promotes more accurate and reliable spatial memories. Loss of midnolin was found to lead to higher and more prolonged expression of multiple lEGs including c-Fos in primary cultured neurons. An inhibitor of midnolin that either targets the midnolin Catch domain or alpha-helix that binds the proteasome may lead to more efficient learning process, rapid memory formation, and thus improved cognitive functions. Therefore, small molecule screens could also be performed to identify chemicals that decrease the affinity of midnolin to its substrates or the proteasome for potential cognitive benefits.

[0085] Given pan proteasomal inhibitors like Velcade (bortezomib), Kyprolis (carfilzomib), and Ninlaro (ixazomib) are efficient treatments for multiple myeloma, targeting and engineering midnolin may represent a unique way to specifically target proteins of interest for potential therapeutic gain.

[0086] More broadly, the Catch domain could also be repurposed for other non-degradative functions. For example, the Catch domain could be fused to a signal peptide or another localization sequence to target a protein of interest to diverse cellular locations. One can use this strategy to retain transcriptional regulators in the cytosol to lower their activity, or to enable new functions of given proteins by introducing them to places in cells where they normally do not localize to. The Catch domain appears to interact with regions within its targets that are not functional domains, providing ground to this strategy.

[0087] Midnolin does not function as E3 ubiquitin ligases, which modify target proteins with ubiquitin to promote proteasomal degradation. Instead, midnolin directly shuttles proteins to the proteasome to achieve selective degradation, thus bypassing the traditional ubiquitination system. Coopting midnolin using small molecules or by engineering the Catch domain genetically provides means to control proteasomal degradation directly without the involvement of many other enzymes normally required in the ubiquitination cascade. The simplicity of the midnolin- proteasome pathway affords a new engineering potential of targeting proteins to the proteasome directly. This will have added benefits for diseases like cancer where resistance to current therapeutic avenues (PROTACs) might occur more commonly due to the complexity of the ubiquitination system.

[0088] Midnolin promotes proteasomal degradation of many nuclear proteins independently of ubiquitination. Accordingly, the present disclosure relates to an analysis of this process usingstructural and biochemical experiments. Midnolin contains an cx-helix that binds the nuclear import receptor (KPNB1) in the cytosol, preventing proteasomal association. Upon nuclear translocation, the midnolin cx-helix is released from KPNB 1 and associates with the proteasome through its subunit PSMD2 / Rpnl. The midnolin ubiquitin-like domain (Ubl) binds PSMD14 / Rpnl 1, which is normally required to cleave the ubiquitin chain from substrates, but here serves as a receptor. The Ubl positions the adjacent Catch domain, such that it can direct bound substrates into the proteasome. Substrates form a 0-strand with the Catch domain, an interaction that can be augmented by a phenylalanine-glycine zipper. Thus, the present disclosure relates, at least in part, to the illumination of the mechanism of the midnolin-proteasome pathway.

[0089] The 26S proteasome is a 2.5 megaDalton complex that degrades proteins in the nucleus and cytosol of eukaryotes (74). Two subcomplexes comprise the 26S proteasome: the 19S regulatory particle and the 20S core particle. The regulatory particle caps one or both ends of the barrel-shaped core particle that contains the proteolytic subunits to prevent nonselective degradation. Cells covalently attach ubiquitin chains onto substrates to selectively target proteins for degradation (75). Ubiquitinated proteins are recruited to the proteasome by the regulatory particle subunits PSMD2 / Rpnl, PSMD4 / RpnlO, or ADRM1 / Rpnl3 (76-79). PSMD2 / Rpnl also binds ubiquitin-like (Ubl) domains of the shuttling factor Rad23 and the deubiquitinase USP14 (79-82). A heterohexameric adenosine triphosphatase (ATPase) ring, which resides at the base of the regulatory particle, mechanically unfolds ubiquitinated proteins by translocation through a central pore and into the proteolytic chamber (83). Successful degradation requires PSMD14 / Rpnll, a zinc metalloprotease above the ATPase ring that removes ubiquitin chains en bloc from substrates to promote translocation (84-87).

[0090] In contrast, the midnolin-proteasome pathway does not require ubiquitination yet achieves selective degradation of many nuclear proteins (88). Midnolin loss is embryonically lethal in mice and flies, indicating that the midnolin-proteasome pathway is vital for organismal physiology (89, 90). Midnolin is a nuclear protein whose expression is induced by growth factors as well as neurological and immunological stimuli (88, 91, 92). In the nucleus, midnolin associates with the proteasome to promote the degradation of key transcriptional regulators, including EGR1, Fos, NR4A1, IRF4, NeuroDl, SPINDOC, and many other nuclear proteins (88). The Drosophila melanogaster ortholog of midnolin, Stuxnet, also associates with the proteasome and promotes the degradation of CBX4 independently of ubiquitination (89). The degradative function of midnolin requires its three conserved domains: an N-terminal ubiquitin- like (Ubl) domain, a “Catch” domain that binds substrate proteins, and a C-terminal helix(ocHelix-C) (FIG. 21A) (88). However, it is unknown how midnolin specifically assembles with the nuclear proteasome and bypasses ubiquitination to initiate the degradation of bound substrates. Here, structural and biochemical experiments were used to elucidate the mechanism of this non-canonical proteolysis pathway.

[0091] The data presented in the Examples of the present disclosure support a role for midnolin in both protein degradation and protein stability. Midnolin can support protein degradation by identifying a target protein for proteasomal degradation. Alternatively, midnolin can support or improve protein stability by identifying target proteins that degrade proteins of interest for proteasomal degradation. The role of midnolin in both protein degradation and protein stability suggests a role for midnolin as a therapeutic for a number of diseases. In some embodiments, midnolin is used in a gene therapy for a number of diseases.Engineered Midnolin Protein

[0092] Accordingly, the present disclosure relates, at least in part, to an engineered midnolin protein. In some embodiments, an engineered midnolin protein is used as a therapeutic for a number of diseases. In some embodiments, an engineered midnolin protein comprises one or more domains. In some embodiments, an engineered midnolin protein comprises an aHelixC domain. In some embodiments, an engineered midnolin protein comprises a ubiquitin-like domain. In some embodiments, an engineered midnolin protein comprises a Catch domain. In some embodiments, a Catch domain comprises a Catch 1 subdomain. In some embodiments, a Catch domain comprises a Catch2 domain. In some embodiments, an engineered midnolin protein comprises an aHelixC domain, a ubiquitin-like domain, and a Catch domain.

[0093] The present disclosure relates, at least in part, to the identification of midnolin, a protein that associates with target proteins and leads to proteasomal degradation of the target protein. A non-limiting example of an amino acid sequence for midnolin is provided by SEQ ID NO: 1:MEPQPGGARSCRRGAPGGACELGPAAEAAPMSLAIHSTTGTRYDLAVPPDETVEGLRKRLSQRLK VPKERLALLHKDTRLSSGKLQEFGVGDGSKLTLVPTVEAGLMSQASRPEQSVMQALESLTETQVS DFLSGRSPLTLALRVGDHMMFVQLQLAAQHAPLQHRHVLAAAAAAAAARGDPSIASPVSSPCRPV SSAARVPPVPTSPSPASPSPITAGSFRSHAASTTCPEQMDCSPTASSSASPGASTTSTPGASPAP RSRKPGAVIESFVNHAPGVFSGTFSGTLHPNCQDSSGRPRRDIGTILQILNDLLSATRHYQGMPP SLAQLRCHAQCSPASPAPDLAPRTTSCEKLTAAPSASLLQGQSQIRMCKPPGDRLRQTENRATRC KVERLQLLLQQKRLRRKARRDARGPYHWSPSRKAGRSDSSSSGGGGSPSEASGLGLDFEDSVWKP EANPDIKSEFWA (SEQ ID NO: 1).

[0094] A nucleotide sequence that encodes SEQ ID NO: 1 is provided as SEQ ID NO: 45:ATGGAGCCGCAGCCCGGCGGCGCCCGGAGCTGCCGGCGCGGGGCCCCCGGCGGCGCCTGCGAGCT GGGCCCGGCGGCTGAGGCGGCGCCCATGAGCCTCGCCATCCACAGCACCACGGGCACCCGCTACG ACCTGGCCGTGCCGCCCGACGAGACGGTGGAGGGGCTGCGCAAGCGGTTGTCCCAGCGCCTCAAA GTGCCCAAGGAGCGCCTGGCTCTTCTCCACAAAGACACCCGGCTCAGTTCGGGGAAGCTGCAGGA GTTCGGCGTGGGTGATGGCAGCAAGCTGACCTTGGTACCCACCGTGGAAGCGGGCCTCATGTCTC AGGCCTCAAGGCCGGAACAGTCCGTGATGCAAGCTCTCGAGAGTCTCACGGAGACGCAGGTCAGT GACTTCCTGTCGGGCCGTTCGCCACTGACACTGGCCTTGCGTGTGGGCGACCACATGATGTTCGT GCAGCTGCAGCTCGCGGCCCAGCACGCTCCACTGCAACACCGCCATGTGCTGGCCGCTGCGGCCG CCGCCGCTGCTGCGCGGGGGGACCCCAGCATAGCCTCCCCCGTGTCCTCGCCCTGCCGGCCGGTG TCCAGTGCCGCCCGAGTCCCCCCGGTGCCCACCAGCCCGTCCCCTGCATCTCCCTCGCCCATCAC AGCCGGCTCCTTCCGGTCCCACGCAGCCTCCACCACCTGCCCGGAGCAGATGGACTGCTCCCCCA CGGCCAGCAGCAGTGCCAGTCCTGGTGCCAGCACCACGTCTACCCCAGGGGCCAGCCCTGCCCCC CGCTCCCGAAAACCCGGCGCCGTCATCGAGAGCTTTGTGAATCACGCCCCGGGGGTCTTCTCAGG GACCTTCTCTGGCACGCTACACCCCAACTGCCAAGACAGCAGCGGGCGGCCGCGGCGTGACATCG GCACCATCCTGCAGATCCTGAACGACCTCCTGAGCGCCACCCGGCACTACCAGGGCATGCCCCCT TCGCTGGCCCAGCTCCGCTGCCACGCCCAGTGCTCCCCGGCCTCACCGGCCCCCGACCTGGCCCC CAGAACTACCTCCTGCGAGAAGCTCACGGCTGCCCCCTCAGCCTCCCTGCTGCAGGGCCAGAGCC AGATCCGCATGTGCAAGCCCCCGGGGGATCGGCTTCGGCAGACAGAAAACCGCGCCACGCGCTGC AAGGTGGAACGGCTGCAGCTGCTTCTGCAGCAGAAACGGCTCCGTAGAAAGGCCCGGCGGGACGC GCGGGGTCCGTACCACTGGTCACCCAGCCGCAAGGCCGGCCGCAGCGACAGCAGTAGCAGCGGGG GCGGCGGCAGCCCCAGCGAGGCCTCCGGCTTGGGCCTCGACTTCGAGGACTCCGTGTGGAAGCCA GAAGCCAACCCTGACATCAAGTCAGAGTTCGTGGTGGCTTAG (SEQ ID NO: 45)

[0095] A non-limiting example of an amino acid sequence for midnolin from D. gyrociliatus is provided by SEQ ID NO: 2:MNVNLI ICPSTGGEYRLSVDLGENVNDLRRSVASKFRLDAERIRLIYNDRNLTTGTLEENDVGEN SKI ILLPCLRSGLVPKSVESTSVQIAEALKTLSGKQIDDFLCGRKPLNLRLILGGHFVYVQLQLS STSKCPKQPSRIEDVRCHGSGVFSGTFSGTVRSPKSKNGASLVAGILNDILFSQRRRRDDQMTGE IRRAASRHSSPDKQRRVEDDQTRSKMAALKEALVERRRKKRKFSPYDRLLV (SEQ ID NO: 2).

[0096] In some embodiments, an engineered midnolin protein is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99% or at least 100% identical to SEQ ID NOs: 1 or 2. In some embodiments, an engineered midnolin protein is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99%, or at least 100% identical to a wildtype midnolin protein. In some embodiments, an engineered midnolin protein comprises one or more mutations relative to SEQ ID NOs: 1 or 2, or a wildtype midnolin protein. In some embodiments, a mutation is an amino acid substitution,deletion, or insertion. In some embodiments, an engineered midnolin protein is a truncated midnolin variant.

[0097] In some embodiments, an engineered midnolin protein comprises a Catch domain. A Catch domain is a domain within a midnolin protein that binds a substrate (e.g., a target protein) for proteasomal degradation. A non-limiting example of an amino acid sequence for a Catch domain is provided by SEQ ID NO: 3: MSQASRPEQSVMQALESLTETQVSDFLSGRSPLTLALRVGDHMMFVQLQLAAQHAPLQHRHVLAA AAAAAAARGDPSIASPVSSPCRPVSSAARVPPVPTSPSPASPSPITAGSFRSHAASTTCPEQMDC SPTASSSASPGASTTSTPGASPAPRSRKPGAVIESFVNHAPGVFSGTFSGTLHPNCQDSSGRPRR DIGTILQILNDLLSATRHYQGMPPSLAQLRCHAQCS (SEQ ID NO: 3).

[0098] In some embodiments, a Catch domain is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99%, or at least 100% identical to SEQ ID NO: 3. In some embodiments, a Catch domain is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99%, or at least 100% to a wildtype Catch domain. In some embodiments, a Catch domain comprises one or more mutations relative to SEQ ID NO: 3 or a wildtype Catch domain. In some embodiments, a mutation is an amino acid substitution, deletion, or insertion. In some embodiments, a Catch domain comprises one or more substitution mutations. In some embodiments, a Catch domain comprises an aspartate at a position corresponding to position 129 in SEQ ID NO: 1. In some embodiments, a Catch domain comprises an aspartate at a position corresponding to position 133 in SEQ ID NO: 1. In some embodiments, a Catch domain comprises a substitution mutation at a position corresponding to position 134 in SEQ ID NO: 1. In some embodiments, a Catch domain comprises an aspartate at a position corresponding to position 139 in SEQ ID NO: 1. In some embodiments, a Catch domain comprises a serine at a position corresponding to position 301 in SEQ ID NO: 1. In some embodiments, a Catch domain comprises an aspartate at a position corresponding to position 306 in SEQ ID NO: 1. In some embodiments, a Catch domain comprises an aspartate at a position corresponding to position 310 in SEQ ID NO: 1. In some embodiments, a Catch domain comprises an aspartate at a position corresponding to position 313 in SEQ ID NO: 1. In some embodiments, a Catch domain comprises one or more of the following amino acid substitution mutations relative to SEQ ID NO: 1: V129D, L133D, S134X, L139D, R301S, I306D, L310D, and / or L313D, wherein X represents any amino acid other than a serine. In some embodiments, a Catch domain is deleted from an engineered midnolin protein.

[0099] In some embodiments, a Catch domain comprises a Catch 1 subdomain. A nonlimiting example of an amino acid sequence for a Catch 2 domain is provided by SEQ ID NO: 42:MSQASRPEQSVMQALESLTETQVSDFLSGRSPLTLALRVGDHMMFVQLQLAAQHAPLQHR (SEQ ID NO: 42)

[0100] In some embodiments, a Catch 1 subdomain is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99% or at least 100% to a wildtype Catch 1 subdomain. In some embodiments, a Catch 1 subdomain comprises one or more mutations relative to a wildtype Catch 1 subdomain. In some embodiments, a mutation is an amino acid substitution, deletion, or insertion. In some embodiments, a Catch 1 subdomain comprises one or more substitution mutations. In some embodiments, a Catch 1 subdomain comprises an aspartate at a position corresponding to position 129 in SEQ ID NO: 1. In some embodiments, a Catch 1 subdomain comprises an aspartate at a position corresponding to position 133 in SEQ ID NO: 1. In some embodiments, a Catch 1 subdomain comprises a substitution mutation at a position corresponding to position 134 in SEQ ID NO: 1. In some embodiments, a Catch 1 subdomain comprises an aspartate at a position corresponding to position 139 in SEQ ID NO: 1. In some embodiments, a Catch 1 subdomain comprises one or more of the following amino acid substitution mutations relative to SEQ ID NO: 1: V129D, L133D, S134X, and / or L139D, wherein X represents any amino acid other than a serine. In some embodiments, a Catch 1 subdomain is deleted from an engineered midnolin protein.

[0101] In some embodiments, a Catch domain comprises a Catch 2 subdomain. A nonlimiting example of an amino acid sequence for a Catch 2 domain is provided by SEQ ID NO: 43:PGASPAPRSRKPGAVIESFVNHAPGVFSGTFSGTLHPNCQDSSGRPRRD IGTILQILNDLLSATR HYQGMPP SLAQLRCHAQCS (SEQ ID NO: 43)

[0102] In some embodiments, a Catch 2 subdomain is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99% or at least 100% to a wildtype Catch 2 subdomain. In some embodiments, a Catch 2 subdomain comprises one or more mutations relative to a wildtype Catch 2 subdomain. In some embodiments, a mutation is an amino acid substitution, deletion, or insertion. In some embodiments, a Catch 2 subdomain comprises one or more substitution mutations. In some embodiments, a Catch domain comprises a serine at a position corresponding to position 301 in SEQ ID NO: 1. In some embodiments, a Catch 2 subdomain comprises an aspartate at a position corresponding to position 306 in SEQ ID NO: 1. In some embodiments, a Catch 2 subdomain comprises an aspartate at a position corresponding to position 310 in SEQ ID NO: 1. In some embodiments, a Catch 2 subdomain comprises an aspartate at a position corresponding to position 313 in SEQ ID NO: 1. In some embodiments, a Catch 2 subdomain comprises one or more of the following amino acidsubstitution mutations relative to SEQ ID NO: 1: R301S, I306D, L310D, and / or L313D. In some embodiments, a Catch 2 subdomain is deleted from an engineered midnolin protein.

[0103] In some embodiments, an engineered midnolin protein comprises an aHelixC domain. An aHelixC domain is a C-terminal domain within a midnolin protein that is necessary for proteasomal association. A non-limiting example of an amino acid sequence for an aHelixC domain is provided by SEQ ID NO: 4: GDRLRQTENRATRCKVERLQLLLQQKRLRRKARRDARGP (SEQ ID NO: 4).

[0104] In some embodiments, an aHelixC domain is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99%, or at least 100% identical to SEQ ID NO: 4. In some embodiments, an aHelixC domain is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99%, or at least 100% to a wildtype aHelixC domain. In some embodiments, an aHelixC domain comprises one or more mutations relative to SEQ ID NO: 1, SEQ ID NO: 4, or a wildtype aHelixC domain. In some embodiments, a mutation is an amino acid substitution, deletion, or insertion. In some embodiments, an aHelixC domain comprises one or more substitution mutations. In some embodiments, an aHelixC domain comprises aspartate at a position corresponding to position 395 in SEQ ID NO: 1. In some embodiments, an aHelixC domain comprises aspartate at a position corresponding to position 405 in SEQ ID NO: 1. In some embodiments, an aHelixC domain comprises aspartate at a position corresponding to position 406 in SEQ ID NO: 1. In some embodiments, an aHelixC domain comprises aspartate at a position corresponding to position 409 in SEQ ID NO: 1. In some embodiments, an aHelixC domain comprises aspartate at a position corresponding to position 410 in SEQ ID NO: 1. In some embodiments, an aHelixC domain comprises aspartate at a position corresponding to position 413 in SEQ ID NO: 1. In some embodiments, an aHelixC domain comprises one or more of the following amino acid substitution mutations relative to SEQ ID NO: 1: L395D, R405D, R406D, R409D, R410D, and / or R413D. In some embodiments, an aHelixC domain is deleted from an engineered midnolin protein.

[0105] In some embodiments, an engineered midnolin protein comprises a ubiquitin-like domain. A non-limiting example of an amino acid sequence for a ubiquitin-like domain is provided by SEQ ID NO: 44: MSLAIHSTTGTRYDLAVPPDETVEGLRKRLSQRLKVPKERLALLHKDTRLSSGKLQEFGVGDGSK LTLVPTVEAG (SEQ ID NO: 44).

[0106] In some embodiments, a ubiquitin-like domain is at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99%, or at least 100%identical to a wildtype ubiquitin-like domain. In some embodiments, a ubiquitin-like domain comprises one or more mutations relative to a wildtype ubiquitin-like domain. In some embodiments, a mutation is an amino acid substitution, deletion, or insertion. In some embodiments, a ubiquitin- like domain comprises one or more substitution mutations. In some embodiments, a ubiquitin-like domain comprises a glutamate at a position corresponding to position 80 in SEQ ID NO: 1. In some embodiments, a ubiquitin-like domain comprises a glutamate at a position corresponding to position 83 in SEQ ID NO: 83. In some embodiments, a ubiquitin-like domain comprises an L80E and / or a G83E substitution mutation relative to SEQ ID NO: 1. In some embodiments, a ubiquitin-like domain is deleted from an engineered midnolin protein.

[0107] In some embodiments, a Catch 1 subdomain and a Catch 2 subdomain are joined by a linker. In some embodiments, a linker is one or more amino acids that connect two domains or subdomains. In some embodiments, a linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids in length. In some embodiments, a linker is a wildtype linker. In some embodiments, a linker is longer than a wildtype linker. In some embodiments, a linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids longer than a wildtype linker. In some embodiments, a linker is shorter than a wildtype linker. In some embodiments, a linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids shorter than a wildtype linker. In some embodiments, a linker affects the binding specificity of an engineered midnolin protein to a target protein.

[0108] In some embodiments, an engineered midnolin protein comprises a nuclear localization signal (NLS). In some embodiments, an aHelixC domain further comprises an NLS. In some embodiments, an NLS is a short amino acid sequence that signals for the transport of a protein into a cell nucleus.

[0109] As used herein, the term “identity” or “sequence identity” refers to the percentage of amino acid (or nucleic acid) residues of a candidate sequence that are identical to the amino acid (or nucleic acid) residues of a reference sequence, e.g., the wildtype sequences described herein, after aligning the sequences and introducing gaps, if necessary, to achieve the maximum percent identity (e.g., gaps can be introduced in one or both of the candidate and reference sequences for optimal alignment and non-homologous sequences can be disregarded for comparison purposes). Alteration of the amino acid sequence or nucleic acid coding sequences can be obtained by deletion, addition, or substitution of residues of the reference sequence. Alignment for purposesof determining percent identity can be achieved in various ways that are within the skill of one in the art, for instance, using publicly available computer software, such as BLAST, BLAST-2, BLAST-P, BLAST-N, BLAST-X, WU-BLAST-2, ALIGN, ALIGN-2, CLUSTAL, or Megalign (DNASTAR) software. Those skilled in the art can determine appropriate parameters for measuring alignment, including any algorithms needed to achieve maximal alignment over the full length of the sequences being compared. For instance, the percent amino acid (or nucleic acid) sequence identity of a given candidate sequence to, with, or against a given reference sequence (which can alternatively be phrased as a given candidate sequence that has or includes a certain percent amino acid (or nucleic acid) sequence identity to, with, or against a given reference sequence) is calculated as follows:100 x (fraction of A / B) where A is the number of amino acid (or nucleic acid) residues scored as identical in the alignment of the candidate sequence and the reference sequence, and where B is the total number of amino acid (or nucleic acid) residues in the reference sequence. In particular, a reference sequence aligned for comparison with a candidate sequence can show that the candidate sequence exhibits from, e.g., 50% to 100% identity across the full length of the candidate sequence or a selected portion of contiguous amino acid (or nucleic acid) residues of the candidate sequence. The length of the candidate sequence aligned for comparison purpose is at least 30%, e.g., at least 40%, e.g., at least 50%, 60%, 70%, 80%, 90%, 95%, 98%, 99%, or 100% of the length of the reference sequence. When a position in the candidate sequence is occupied by the same amino acid (or nucleic acid) residue as the corresponding position in the reference sequence (e.g., poly(A) sequence or any other sequences described herein), then the molecules are identical at that position.Methods of Use

[0110] Aspects of the present disclosure relate to the use of any engineered midnolin protein and / or nucleic acids encoding an engineered midnolin protein described herein to improve degradation of a target protein, promote stability of a target protein, or improve stability of a target protein. Midnolin is capable of ubiquitin-independent degradation of target proteins. In some embodiments, an engineered midnolin protein improves degradation of a target protein by associating the target protein with a proteasome. In some embodiments, degradation of a target protein is measured by determining the concentration of a target protein in a sample before and after contacting a cell with an engineered midnolin protein. In some embodiments, an engineered midnolin protein promotes or improves target protein stability by associating a protein that naturally degrades the target protein with a proteasome.

[0111] In some embodiments, an engineered midnolin protein is delivered to a cell to improve degradation of a target protein or improve or promote stability of a target protein. In some embodiments, degradation of a target protein comprises contacting a target protein with an engineered midnolin protein. In some embodiments, degradation of a protein that naturally degrades a target protein comprises contacting the protein with an engineered midnolin protein. In some embodiments, a target protein or a protein that naturally degrades a target protein comprises a P strand degron. A P strand degron is a protein structure to which midnolin binds. In some embodiments, the Catch domain of an engineered midnolin protein binds the P strand degron of a target protein or a protein that naturally degrades a target protein. In some embodiments, a target protein or a protein that naturally degrades a target protein is degraded by a proteasome after the Catch domain of an engineered mandolin protein binds the P strand degron of the target protein or protein that naturally degrades a target protein. In some embodiments, degradation by a proteasome is ubiquitination-independent. In some embodiments, a target protein is a transcription factor. In some embodiments, a target protein is IRF4 or an immediate early gene. In some embodiments, the immediate early gene is c-Fos, FosB, ERG1, or NR4A1.

[0112] In some embodiments, an engineered midnolin protein is delivered to a cell. Methods of delivering an engineered protein to a cell are known in the art. In some embodiments, a gene encoding an engineered midnolin protein is delivered to a cell. In some embodiments, a gene encoding an engineered midnolin protein is delivered to a cell on a plasmid. In some embodiments, a gene encoding an engineered midnolin protein is expressed in a cell heterologously. In some embodiments, a gene is heterologous if it does not naturally occur in that cell to which the gene is delivered. In some embodiments, a gene encoding an engineered midnolin protein is expressed in a cell exogenously. In some embodiments, a gene is exogenous if the gene is delivered to the cell, regardless of whether the gene is also found endogenously in the cell. In some embodiments, a gene encoding an engineered midnolin protein is integrated into the genome of a cell.

[0113] In some embodiments, midnolin activity is modulated by overexpressing or deleting a naturally occurring midnolin gene in a cell. In some embodiments, midnolin activity is modulated by contacting a cell with a small molecule that inhibits expression and / or activity of midnolin in the cell. In some embodiments, activity of midnolin is inhibited by a broad- spectrum small molecule protease inhibitor. In some embodiments, a small molecule protease inhibitor inhibits midnolin activity directly or indirectly. In some embodiments, a small molecule inhibitor that indirectly inhibits midnolin activity is MG- 132. In some embodiments, midnolin activity is modulated by contacting a cell that arguments midnolin activity. In some embodiments, activity of midnolin is augmented by a broad- spectrum small molecule protease inhibitor. In someembodiments, a small molecule protease inhibitor augments midnolin activity directly or indirectly. In some embodiments, a small molecule inhibitor that indirectly augments midnolin activity is TAK-243.

[0114] In some embodiments, midnolin activity is modulated by modifying, mutating, or deleting a gene encoding midnolin from the genome of a cell. In some embodiments, the step of deleting is accomplished by using any genetic engineering technique known in the art. For example, the genetic engineering technique can be a CRISPR / Cas9 system.

[0115] The methods described herein can be useful to treating one or more diseases. In some embodiments, a target protein described herein is relevant to a disease. In some embodiments, the disease is cancer, an immunological disease, or a neurological disease. In some embodiments, a cancer is blood or lymphoid cancer, a bone or soft tissue cancer, a brain or central nervous system cancer, breast cancer, a childhood cancer, a digestive system cancer, an eye cancer, a head or neck cancer, lung cancer, a pelvic area cancer, a skin cancer, a urinary cancer, or any other cancer known in the art. In some embodiments, an immunological disease is an autoimmune disease. In some embodiments a neurological disease is a neurodegenerative disease or a neurodevelopmental disease.EXAMPLESExample 1. Genetic screens reveal midnolin as a regulator of IEG protein degradation

[0116] To investigate the mechanism of IEG protein degradation, the Global Protein Stability (GPS) reporter system was first employed to assay IEG protein stability (8). GPS allows for the stable expression of DsRed as an internal control and a GFP-tagged protein from the same bicistronic mRNA. Thus, the ratio of GFP / DsRed analyzed by flow cytometry provides a measure of the relative stability of the GFP-fused protein. To identify regulators of IEG protein stability, HEK-293T cell lines stably expressing the GPS reporter for EGR1 or FosB were generated and genome- wide CRISPR-Cas9 screens were performed to hunt for genes whose disruption stabilized EGR1 or FosB (FIG. 1A). The comparison of EGR1 and FosB allowed us to investigate whether lEGs from different families are degraded by same or different mechanisms. The top hit from both screens was MIDN, a gene that in mammals encodes a largely uncharacterized protein named midnolin (FIGs. 1B-1C).

[0117] To validate the CRISPR-Cas9 screening results, MIDN clonal knockout (KO) HEK- 293T cells stably expressing the GPS reporter for EGR1 or FosB were generated, and assessed the effect of MIDN disruption and overexpression on the stability of EGR1 or FosB. Consistent with the screening results, loss of midnolin increased the stability of both EGR1 and FosB (FIGs. 1D-1E). Additional IEG proteins were tested, and midnolin overexpression was sufficient todecrease the levels of EGR1, FosB, c-Fos, and NR4A1 (FIGs. 1D-1E, FIGs. 8A-8B), but not several other transcription factors including ATF2, CREB3, and CREB5 (FIG. 8C). These findings raised an intriguing possibility that three distinct families of IEG proteins may be targeted for degradation by the same protein, midnolin.Example 2. Midnolin is induced and promotes the degradation of several IEG proteins in physiological settings

[0118] To investigate the requirement of midnolin for the degradation of IEG proteins in physiologically relevant settings, MIDN was mutated using CRISPR-Cas9 to generate a population-level KO of NIH / 3T3 fibroblasts, a well-characterized cell line for studying IEG inducibility during cell cycle re-entry. Serum deprivation synchronizes NIH / 3T3 cells in the GO phase of the cell cycle, and lEGs are rapidly and transiently transcribed within minutes after serum addition to these cells (9). This transient induction of IEG transcription is followed by an increase in the levels of IEG proteins that then quickly return to their basal levels. In MIDN KO cells, IEG protein levels remained high for several hours longer than in control cells, suggesting that the degradation of EGR1, NR4A1, FosB, and c-Fos was attenuated in the absence of midnolin (FIG. 2A). In contrast, stable overexpression of midnolin led to a decrease in the level of these IEG proteins (FIG. 2B).

[0119] In addition to their importance during cell cycle re-entry, lEGs mediate adaptive responses in neurons. In response to sensory stimuli, if enough glutamate is released at excitatory synapses in the brain to generate an action potential in the post-synaptic neuron, the depolarization of the post-synaptic neuron results in an influx of calcium that triggers the induction of lEGs (3). This membrane depolarization induction of lEGs also occurs in cultured mouse neurons within minutes of exposure to elevated levels of KC1 (10). To determine if midnolin regulates IEG protein levels in this paradigm, embryonic mouse cortical neurons were cultured and infected with lentivirus to generate population-level MIDN KO or midnolin- overexpressing cells. When exposed to elevated levels of KC1, IEG protein expression was increased in MIDN KO neurons but decreased in the midnolin-overexpressing neurons (FIGs. 2C-2D). In contrast, IEG mRNA expression was largely unchanged by midnolin overexpression (FIG. 8D), which is consistent with midnolin affecting IEG protein stability, but not their mRNA transcripts.

[0120] KC1 treatment of primary cortical neurons induced MIDN mRNA levels with kinetics similar to that of lEGs (FIG. 2E). In a previous RNAseq dataset, MIDN was found to be induced upon light stimulation of the visual cortex in vivo (FIG. 8E) (11). Midnolin was also induced upon serum restimulation of NIH / 3T3 cells (FIG. 2F) with kinetics similar to that of lEGs (12).These stimulus-dependent increases in midnolin expression may be involved in the rapid degradation of IEG proteins. Thus, midnolin is induced by various stimuli and promotes the degradation of IEG proteins in physiologically relevant settings, potentially through a feedback mechanism.Example 3. Midnolin can promote the degradation of numerous transcriptional regulators

[0121] To determine the extent to which midnolin regulates other cellular proteins beyond lEGs, a screen was performed to identify additional midnolin targets. The screen employed a previously described GPS ORFeome library, which contains -12,000 barcoded human open reading frames (ORFs) tagged with GFP in the GPS reporter system (13). The GPS ORFeome library was stably introduced into MIDN KO HEK-293T cells and plasmids expressing either a control blue fluorescent protein (BFP), or midnolin together with BFP, were transiently transfected into cells to yield two cell libraries, one lacking midnolin and the other overexpressing midnolin. Because the GPS system overexpresses proteins and endogenous midnolin levels are low, midnolin was overproduced to gain sensitivity in this setting. The cells in each library were then partitioned into six populations based on their GFP / DsRed ratios by fluorescence-activated cell sorting. The barcodes present in each population were then sequenced to determine the change in their distribution within the cell populations upon midnolin overexpression. If midnolin promoted the destruction of a given barcoded GFP-fusion protein, the distribution of the barcode would shift to a cell population with a lower GFP / DsRed ratio in the midnolin overexpressing library (FIG. 3A). This screen yielded previously characterized targets of midnolin including FosB and c-Fos, along with CBX4 (Pc in flies), which was previously shown to be regulated by the Drosophila melanogaster ortholog of midnolin, Stuxnet (14) (FIGs. 3B-3C). Midnolin overexpression also led to a robust reduction in the GFP / DsRed ratios of many proteins, consistent with a decrease in protein stability. Most of the proteins regulated by midnolin overexpression were nuclear proteins that regulate transcription (FIG. 3D), such as the lineage- specific transcription factors IRF4, NeuroDl, PAX8, and GATA1.

[0122] To validate the findings from the GPS ORFeome screen, midnolin was overexpressed in MIDN KO HEK-293T cells stably expressing individual GPS reporters for the identified proteins. Midnolin overexpression was sufficient to reduce the GFP / DsRed ratios of these proteins, consistent with a decrease in their stability (FIG. 3E and FIG. 9). The endogenous levels of various proteins identified in the screen were also substantially reduced in HEK-293T cells that expressed a doxycycline-induced midnolin (FIG. 3F). Among the lineage- specific transcription factors that are not expressed in HEK-293T cells, interferon regulatory factor 4 (IRF4) is essential for the function and homeostasis of mature B and T cell lymphocytes (15, 16).To test the ability of midnolin to promote the destruction of endogenous IRF4, Ramos B cell lines were generated with population-level KO or overexpression of midnolin. In these cell lines, the steady-state levels of IRF4 were markedly increased in the MIDN KO cells and decreased in the midnolin-overexpressed cells (FIG. 3G). Thus, through gain-of-function screening, many potential targets of midnolin were uncovered that have important tissue or cell-type specific functions in regulating gene expression.Example 4. Midnolin associates both with its substrates and the 26S proteasome

[0123] To begin to determine how midnolin promotes the degradation of a large diverse set of proteins, a HEK-293T cell line was generated in which endogenous midnolin was tagged at its N-terminus with 3xHA to facilitate immunoprecipitation. By mass spectrometry it was found that endogenous midnolin co-immunoprecipitated essentially all proteasomal subunits of the 19S regulatory particle and the 20S proteolytic core particle (FIG. 4A). These proteasomal subunits were the most abundant proteins detected by mass spectrometry, and there were no proteins besides the proteasomal proteins that were co-immunoprecipitated by midnolin and found to be shared with the genome-wide CRISPR-Cas9 screens (FIGs. 1A-1C). These results suggested that midnolin likely interacts directly with the proteasome. In addition, by mining the BioPlex protein-protein interaction dataset (17), several IEG proteins including EGR1, FosB, and NR4A1, as well as a proteasomal component, PSMD2, were found to co-immunoprecipitate endogenous midnolin in HEK-293T cells (FIG. 10A).

[0124] To confirm the mass spectrometry findings, HEK-293T cells expressing 3xHA- tagged midnolin were treated with MG132 to prevent substrate degradation or with phorbol 12- myristate 13-acetate (PMA), a Protein Kinase C (PKC) agonist to induce the transcription of lEGs (18). Endogenous midnolin interacted with c-Fos, FosB, EGR1, and NR4A1, as well as the proteasome, as indicated by PSMD2 and PSMA2, components of the 19S and 20S proteasome, respectively (FIG. 4B). Like serum- stimulated fibroblasts and KCl-treated neurons, exposure of HEK-293T cells to PMA led to an increase in the level of the midnolin protein (FIG. 4B). Thus, midnolin interacts with the proteasome to promote the degradation of midnolin-bound proteins.Example 5. Midnolin promotes ubiquitination-independent degradation of bound substrates

[0125] It was next investigated whether midnolin targets its substrates for destruction by a ubiquitination-and proteasome-dependent mechanism. Individual GPS reporters of midnolin substrates were stably expressed in MIDN KO HEK-293T cells, which were transfected with BFP control or midnolin co-expressing BFP. These reporter cells were then treated for 6 hourswith the proteasome inhibitor MG 132 or TAK-243, a potent inhibitor of the El ubiquitin- activating enzymes UBA6 and UAE that inhibits protein ubiquitination globally (19). While the proteasome inhibitor MG 132 strongly reduced the midnolin-mediated degradation of these midnolin substrates, the ubiquitin El inhibitor TAK-243 did not disrupt midnolin function (FIG. 10B). In contrast, both MG 132 and TAK-243 promoted stabilization of c-Myc, which does not appear to be a midnolin substrate (FIG. 10B). Furthermore, when HEK-293T cells expressing 3xHA-tagged midnolin were treated with MG 132 or TAK-243 for 6 hours, the proteasome inhibitor MG 132 was found to lead to a significant increase in the expression of the midnolin protein, but the El inhibitor TAK-243 caused a slight decrease in the level of midnolin. In contrast, the level of c-Myc and p27, two transcriptional regulators that are known to be targeted for proteasomal degradation in a ubiquitination-dependent manner increased upon exposure to TAK-243 or MG132 (FIG. 4C) (20-22). Thus, midnolin does not appear to require ubiquitination for its own turnover by the proteasome.

[0126] To further test the requirement of ubiquitination for midnolin-mediated degradation, potential ubiquitination sites were mutated in several midnolin substrates. Because canonical ubiquitination occurs on lysine residues, all lysine residues within these substrates were mutagenized to arginine to block lysine-dependent ubiquitination (23-25). Wild type and K to R mutant substrates, such as EGR1, FosB, c-Fos, NR4A1, NeuroDl, and IRF4, interacted with endogenous midnolin to a similar extent (FIG. 4D). When stably expressed in MIDN KO HEK- 293T cells, both wild type and K to R mutant substrates were efficiently degraded upon doxycycline-induced midnolin overexpression (FIG. 4E), indicating that lysine residues were not required for midnolin-dependent substrate destruction. Thus, it seems that midnolin directly associates with the proteasome and promotes the degradation of many transcriptional regulators without requiring their ubiquitination.Example 6. Midnolin contains three domains that function in concert to promote proteasomal degradation of bound substrates

[0127] To gain insight into how midnolin interacts with the proteasome and its numerous substrates, AlphaFold was used to obtain a predicted structure of midnolin, which revealed three confidently predicted and highly conserved regions with defined structure (FIG. 5A, FIGs. 1A- 11B) (26). Midnolin does not contain structural elements that are characteristic of RING- or HECT-type ubiquitin ligases (27) and does not contain a ubiquitin-binding domain characteristic of proteasomal processivity factors like Rad23 or Ubiquilin (28). Instead, midnolin contains a ubiquitin-like domain (Ubl) towards its N-terminus. Additionally, midnolin contains two discontinuous regions, each composed of two predicted anti-parallel P strands and two or three ahelices, that appear to fold together to form a domain with internal symmetry. For reasons discussed below, this region of midnolin was named the “Catch” domain. Finally, midnolin contains a long a helix towards its C-terminus, termed aHelix-C, which includes a predicted nuclear localization sequence (NLS). Indeed, endogenous midnolin was largely located within the nucleus and deletion of the predicted NLS, but not other regions of midnolin, resulted in its localization to the cytoplasm (FIGs. 12A-12B).

[0128] To examine whether these three regions are important for midnolin function, wild type and mutant versions of midnolin were transiently expressed in MIDN KO HEK-293T cells stably expressing the GPS IRF4 or FosB reporters (FIGs. 5A-5C). While wild type midnolin potently promoted IRF4 and FosB destruction, point mutations of the Ubl or deletions of the Ubl, Catch, aHelix-C, or NLS domains abrogated midnolin function (FIG. 5C, FIG. 13A).

[0129] Co-immunoprecipitation experiments were performed to identify midnolin domains that are required for its interaction with substrates and / or the proteasome. Point mutations or deletion of the Ubl domain did not disrupt the stable association of midnolin with EGR1 or the proteasome (FIGs. 5D-5E). However, mutagenesis of the Ubl domain potently increased midnolin levels and these Ubl mutants were only marginally sensitive to MG 132 (FIGs. 5D-5E). This suggests that the ubiquitination-independent degradation of midnolin that were observed previously (FIG. 4C) is dependent on its Ubl domain. In contrast, deletion of the NLS or entire C-terminal a helix revealed that this domain is necessary for midnolin to interact stably with the proteasome, but not its substrates (FIGs. 5D-5E). This association with the proteasome was not affected by inhibition of the proteasome or El ubiquitin-activating enzymes (FIG. 13B). When fused to maltose binding protein (MBP), the midnolin helix conferred the ability to interact with the proteasome (FIG. 5F). Thus, the C-terminal midnolin helix is both necessary and sufficient to bind the proteasome and midnolin, unlike the processivity factors Rad23 and Ubiquilin, engages the proteasome stably independent of a Ubl domain (FIGs. 5D-5E) (29, 30).

[0130] Deletion of the regions that fold together to form the Catch domain (the N-terminal Catch 1 and C-terminal Catch2 subdomains) abolished the interaction of midnolin with its substrates, without affecting its ability to bind the proteasome (FIGs. 5D-5E). In growth arrested NIH / 3T3 cells where the level of endogenous midnolin substrates was induced upon serum restimulation, the interaction of midnolin with its substrates also required the Catch domain (FIG. 13C). To determine if the Catch domain is both necessary and sufficient to engage midnolin substrates, transfected wild type, Catch domain-deleted midnolin, or the Catch domain alone were immunoprecipitated in HEK-293T cells and the interaction with various substrates and proteasomal components was assessed. Deletion of the Catch domain abolished the interaction of midnolin with its substrates while retaining proteasome binding, and the Catchdomain alone was sufficient to bind midnolin substrates (FIG. 5G). Catchl and Catch2 are separated by a long 111 amino acid unstructured sequence that is poorly conserved. This Catchl and Catch2 linker was shortened to 10 amino acids (ALoopl), 16 amino acids (ALoop2), or 28 amino acids (ALoop3). These midnolin variants still interacted with substrates and promoted their degradation, indicating that the long stretch connecting Catchl and Catch2 is largely dispensable (FIG. 5G, FIGs. 14A-14B). However, the length between Catchl and Catch2 could not be too short, because the shortest linker (ALoopl, 10 amino acids) did attenuate the interaction of midnolin with certain substrates (FIG. 5G and FIG. 14A). Thus, the linker between Catchl and Catch2 may still be important for proper flexibility or folding of the Catch domain.

[0131] Given the importance of the Catch domain for binding substrates, the AlphaFold prediction that Catchl and Catch2 fold together was further validated. This folding may be strongly driven by hydrophobicity, because the core of the Catch domain is comprised exclusively of highly conserved hydrophobic amino acids (FIG. 14C). To validate that Catchl and Catch2 fold back to bind each other in a process driven by hydrophobicity, each was expressed separately by tagging Catchl to 2xHA-GFP and Catch2 to 2xFLAG-MBP. A series of mutants for both Catchl and Catch2 were also generated where some hydrophobic residues that are predicted by AlphaFold to drive the interaction were changed to aspartic acid. As a control, a solvent-exposed residue on Catchl (SI 34) and Catch2 (R301) were mutagenized because these were not predicted to contribute to the binding between Catchl and Catch2. These epitope-tagged Catchl and Catch2 constructs were expressed in MIDN KO HEK-293T and 2xHA-GFP-Catchl was immunoprecipitated. Indeed, Catchl and Catch2 bound each other when expressed as independent proteins, and mutagenesis of the hydrophobic residues to aspartic acid abolished the interaction (FIGs. 5H-5I). This contrasts with mutagenesis of the solvent-exposed residues that did not alter the Catch 1-Catch2 interaction as anticipated. These same mutations were also introduced into full-length midnolin and only mutagenesis of the hydrophobic residues abolished the ability of midnolin to bind with and promote the degradation of its substrates (FIG. 14B, FIG. 14D). The hydrophobic core was important for the function of the Catch domain, because mutagenesis of additional hydrophobic residues to aspartic acid attenuated the ability of midnolin to promote the degradation of its substrates including IRF4 and FosB (FIG. 14E).

[0132] Thus, midnolin is primarily localized within the nucleus, associates with the proteasome using its long C-terminal a helix, binds substrates using its hydrophobic Catch domain, and contains a Ubl domain that is necessary to promote substrate degradation. Overall, these three regions of midnolin function in concert to promote ubiquitination-independent proteasomal degradation of bound substrates.Example 7. Midnolin “catches” regions within its substrates that constitute a p strand degron

[0133] Canonically, E3 ubiquitin ligases bind short linear motifs within substrates, termed degrons, in which the amino acid side chains play crucial roles in determining substrate specificity (31-34). To gain insights into how midnolin achieves substrate selection, AlphaFold was used to predict the structure of the midnolin- substrate bound complex (26). The AlphaFold predictions for midnolin bound to IRF4 (FIG. 6A) revealed that a predicted unstructured region of native IRF4 (FIG. 15A) formed a P strand upon binding to the midnolin Catch domain (FIG. 15B), thereby completing a five-stranded anti-parallel P sheet tertiary structure. Consistent with this AlphaFold predicted interaction, a small deletion within IRF4 that encompassed the predicted P strand abolished the ability of midnolin to promote IRF4 degradation (FIG. 6B).

[0134] To determine the generality of this P strand capture mechanism, AlphaFold structure predictions were performed for the 508 most destabilized proteins uncovered in the GPS ORFeome screen. The PDB files for these 508 predictions can be accessed using the following DOI (10.5061 / dryad.m905qfv6g). Approximately 40% (205 / 508) of the proteins have predictions that are consistent with them being captured by midnolin. The AlphaFold-predicted structure of midnolin substrates in the native and midnolin-bound states were systematically compared. The regions predicted to be captured by midnolin are generally more unstructured than the rest of the same protein in the native state but have the potential to form a P strand upon binding the Catch domain (FIG. 15C). These findings were validated by introducing into several representative midnolin substrates small deletions that encompassed the predicted P strand and tested the ability of midnolin to interact with and promote the degradation of these mutants. In each case, deletion of the predicted P strand abrogated the ability of midnolin to both interact with substrates (FIG. 6C) and promote their degradation (FIGs. 16A-16H).

[0135] The AlphaFold predictions yielded several different modes of midnolin-FosB interaction, raising the possibility that the N- and C-terminus of FosB have the potential to form multiple different P strands that could be captured by midnolin. Various truncated forms of FosB were generated to determine which regions were necessary for midnolin to promote degradation and found that the C-terminal 101 amino acids were required (FIG. 17A). Truncation of the last 101 amino acids of FosB results in a naturally existing splice isoform termed AFosB, which was shown to be significantly more stable than other Fos family members, although the molecular basis for this increased stability was unknown (35). Previous studies have shown that chronic and repeated exposure to drugs of abuse such as cocaine leads to week-long accumulation of AFosB protein expression in the nucleus accumbens, a brain region crucial for addiction (35, 36).Indeed, overexpression of AFosB in neurons of the nucleus accumbens sensitizes animals to the effects of cocaine and may thus contribute to addiction (37). In contrast to full-length FosB, midnolin weakly interacted with and less effectively promoted the degradation of AFosB (FIG. 6C and FIG. 17A). Thus, these findings provide a mechanistic explanation for the increased stability of AFosB. Nevertheless, AFosB is eventually degraded, and this may require the N- terminus of FosB, which is also predicted by AlphaFold to form a P strand that can be captured by midnolin. Deletion of the predicted N-terminal P strand region within AFosB largely abolished its ability to be targeted for decay by midnolin (FIG. 17B). Thus, a protein can contain more than one region that can be captured by midnolin.

[0136] Having identified sequences that mediate the midnolin-substrate interaction, it was asked whether specific amino acids were enriched or depleted over others within the predicted P strands captured by the midnolin Catch domain. Within these P strands, a marked depletion of charged amino acids including aspartic acid and glutamic acid were observed, as well as proline (FIG. 6D), which is known to disrupt P strands (38). Instead, there was a strong enrichment for hydrophobic amino acids within the midnolin induced P strands, which overall are significantly more hydrophobic than the average of all other regions within the same protein (FIG. 6E and FIG. 17C). Once bound by midnolin, these hydrophobic P strand residues were predicted to be buried within the core of the Catch domain (inward), while charged amino acids tended to point outward and were solvent exposed (FIG. 6D, FIG. 6F, and FIG. 17D). The enrichment for hydrophobicity within the P strands is consistent with the fact that the core of the Catch domain is also highly hydrophobic and required for catching hydrophobic regions within substrates. To validate these predictions, two residues (G218 and T219) at the center of the predicted IRF4 P strand were mutagenized to proline to potentially disrupt P strand formation, or the hydrophobic residues buried within the interior of the Catch domain (V216, and F220) were mutagenized to aspartic acid to disrupt potential hydrophobic interactions. Consistent with the AlphaFold predictions, both introduction of prolines and mutagenesis of the hydrophobic residues to aspartic acid within the P strand abolished the ability of midnolin to interact with (FIG. 6G) and promote the degradation of IRF4 (FIG. 6H). Similar results were obtained following mutagenesis of the corresponding residues within the EGR1 adopted P strand (FIGs. 17E-17F).

[0137] IT was next tested whether midnolin can interact with and promote the degradation of GFP fused to a short sequence containing the P strand degron(s) within substrates. Midnolin interacted with (FIG. 61) and promoted the degradation of (FIGs. 17G-17H) GFP fused to short stretches within EGR1, IRF4, and the C-terminal tail of FosB that was dependent on the hydrophobicity of the predicted P strand. Thus, the P strands predicted by AlphaFold are necessary and sufficient to interact with midnolin for proteasomal degradation.

[0138] It was concluded that midnolin achieves substrate selection via a general mechanism: captured regions within substrates possess the ability to form a P strand that is biochemically compatible with the hydrophobic core of the Catch domain.Example 8. Structure of the midnolin-proteasome complex

[0139] Midnolin-proteasome complexes were affinity-purified from human embryonic kidney (HEK-293T) cells overexpressing FLAG-tagged midnolin variants, including one N- terminally fused to an EGR1 substrate peptide (FIG. 21B and FIG. 21C). EGR1 was chosen because it co -immunoprecipitated the midnolin Catch domain most effectively among characterized substrates (88). The fused EGR1 peptide may occupy the Catch domain intramolecularly and potentially reduce the substrate heterogeneity. Purified samples contained primarily proteasomal complexes based on the characteristic Coomassie stain observed after sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (FIG. 21D) and negative stain electron microscopy (FIG. 21E). These purified samples were subjected to singleparticle cryogenic-electron microscopy (cryo-EM) (FIG. 22 and FIGs. 23A-23C, and Table 1). 3D classification revealed density associated with the 19S regulatory particle that could be assigned to midnolin. Two primary conformations of the midnolin-proteasome complex that, referred herein as Ml and M2 states, were resolved (FIG. 22 and FIGs. 23A-23C ).

[0140] The Ml state (3.9 A) contained the midnolin Ubl domain and ocHelix-C engaged with the 19S regulatory particle (FIG. 24). Above the ATPase ring was low-resolution density corresponding to the midnolin Catch domain. Associated with the ATPase ring was density corresponding to the proteasomal activator ZFAND5 (93). The Ml state resembled the previously reported Z+D proteasomal state (93), with PSMD2 positioned near the substrate entry site of the ATPase ring and PSMD14 / Rpnl 1 offset from the closed ATPase ring. The EGR1- midnolin fusion dataset lacked the Ml state but contained an Ml -like proteasome state without Ubl and Catch domain density (FIG. 22), possibly due to steric hindrance caused by the N- terminal EGR1 peptide fusion. The Ml -like state was unique to the EGR1 -midnolin fusion dataset and was abundant, consistent with the possibility that this state is stabilized when midnolin binds a substrate.

[0141] In comparison, two M2 states (A: 3.1 A and B: 3.4 A) were observed in all datasets. Both show only the midnolin ocHelix-C bound to the regulatory particle but display ATPase configurations in different stages of active translocation with distinguishable density corresponding to a substrate polypeptide (FIG. 25). The more abundant M2-A state resembled previously reported human ED2 or yeast 4D proteasomal states (94, 95).

[0142] Overall, the Ml state may represent the midnolin-proteasome complex that is in the process of delivering a substrate for unfolding. The M2 state may represent the complex after a substrate has been delivered for unfolding and subsequent degradation.Table 1. Cryo-EM data collection, refinement, and validation statisticsMl w / Ubl M2-A M2-A M2-B Ml w / o UblPDB 9BUI PDB 9BV1 (focused) PDB 9BV2 PDB 9BV3EMD-44910 EMD-44927 EMD-44928 EMD-44929 EMD-44930Data collection and processingMagnification 105,000 105,000 105,000 105,000Voltage (kV) 300 300 300 300Electron exposure (e- / A2) 50 to 54 50 to 54 50 to 54 54Defocus range (pm) -0.8 to -2.2 -0.8 to -2.2 -0.8 to -2.2 -0.8 to -1.8Pixel size (A) 0.825 0.825 0.825 0.825Symmetry imposed Cl Cl Cl ClInitial particle images (no.) 1,765,798 1,765,798 1,765,798 658,210Final particle images (no.) 35,838 186,489 87,326 113,946Map resolution (A) 3.9 3.1 3.0 3.4 2.9FSC threshold 0.143 0.143 0.143 0.143 0.143Map resolution range (A) 3.4 to 10 2.6 to 8.0 2.8 to 6.9 3.0 to 8.5 2.5 to 9.0RefinementInitial model used AlphaFold2, AlphaFold2, M2- A, Ml w / Ubl,ModelAngelo, ModelAngelo ModelAngelo ModelAngeloPDB 7QXWModel resolution (A) 4.4 3.4 3.8 3.3FSC threshold 0.5 0.5 0.5 0.5Model resolution range (A) 4.4 to 62.3 3.4 to 59.3 3.8 to 62.5 3.3 to 63.9Map sharpening B factor (A2) -23.7 -71.9 -87.9 -63.3 -56.9Model compositionNon-hydrogen atoms 90,266 90,308 88,606 88,575Protein residues 11,456 11,452 11,233 11,236Ligands 2 Mg2+1 Zn2+1 Zn2+5 Mg2+4 ATP 4 Mg2+4 ATP 4 ATP1 ADP 4 ATP 1 ADP 2 ADP2 ADPB factors (A2)Protein 208.50 70.89 114.57 76.24Ligand 215.03 50.77 97.23 45.60R.m.s. deviationsBond lengths (A) 0.005 0.005 0.005 0.004Bond angles (°) 0.989 0.981 0.959 0.976ValidationMolProbity score 2.45 2.11 2.11 2.08Clashscore 23.07 13.23 13.76 12.82Poor rotamers (%) 0.01 0.02 0.02 0.0Ramachandran plotFavored (%) 88.25 92.19 92.68 92.79Allowed (%) 11.66 7.78 7.30 7.18Disallowed (%) 0.09 0.03 0.02 0.04Example 9. Mechanisms for assembling the midnolin-proteasome complex in the nucleus

[0143] The Ml and M2 states showed a binding interface between midnolin ocHelix-C and a region of PSMD2 referred to herein as the ‘M site’ (FIG. 26A). The interaction between ocHelix-C and PSMD2 was conserved throughout evolution as predicted by AlphaFold-multimer (96) (FIGs. 27A-27B and FIG. 28A). The M site is distinct from the ubiquitin and Rad23 binding surfaces (T1 and NT), but partially overlaps with the USP14 binding site (T2) (79-82). cxHelix-C interacts with the M site of PSMD2 primarily through a hydrophobic interface on the N-terminal half of cxHelix-C and electrostatic interactions on the C-terminal half of cxHelix-C. The structural models were validated by introducing mutations to the M site of PSMD2 that were predicted to block the interaction and found that E319, L426, Y434, L465, and Y469 of PSMD2 were required to bind midnolin (FIG. 28B). E319 engages with the C-terminal half of cxHelix-C, while L426, Y434, L465, and Y469 engage with the N-terminal half. Thus, both halves of the interface are necessary for cxHelix-C to interact with PSMD2.

[0144] The arginine-rich, basic region of cxHelix-C (R405-R413), which serves as a nuclear localization sequence (NLS) (88), formed electrostatic interactions with acidic residues of PSMD2 (E315, D316, and E319) (FIG. 26A and FIG. 28A). The observation that the NLS may be critical for proteasome binding suggested a mechanism for assembling the midnolin- proteasome complex primarily in the nucleus. To investigate this mechanism, mutants were generated in cxHelix-C and their impact on proteasome binding and nuclear localization was tested. Mutation of leucine 395 (L395D) or five arginines comprising the NLS (5RD) abolished the midnolin-proteasome interaction (FIG. 26B). Although wild-type and L395D midnolin were localized primarily within the nucleus, 5RD midnolin was cytosolic (FIG. 26C). Thus, midnolin does not require proteasome binding to enter the nucleus.

[0145] Midnolin was immunoprecipitated and mass spectrometry was performed to identify potential machinery that regulates the nuclear transport of midnolin (FIG. 28C). Midnolin coimmunoprecipitated many nuclear pore proteins along with importin |3 (KPNB1), but none of the importin ex proteins (KPNA1-6) (FIG. 28C). Importin ex (KPNA1-6) normally binds to cargo containing a positively charged NLS (97, 98). The cargo-importin ex complex is then ferried to the nucleus by importin |3 (KPNB1) via its interaction with the nuclear pore complex (99-102). Although nuclear-localized wild-type and L395D midnolin co-immunoprecipitated importin |3 (KPNB1), 5RD midnolin did not (FIG. 28D), consistent with its cytosolic localization. Small interfering RNAs were then used to transiently reduce the expression of importin |3 (KPNB1) or the most abundant importin ex protein (KPNA2) (103). Depletion of importin |3 (KPNB1), but not importin ex (KPNA2), led to the accumulation of midnolin in the cytosol (FIG. 26D and FIGs. 29A-29B). Thus, importin |3 (KPNB1) is necessary for the nuclear transport of midnolin.

[0146] When fused to maltose-binding protein (MBP), wild-type and L395D cxHelix-C, but not 5RD cxHelix-C, co-immunoprecipitated importin |3 (KPNB1) (FIG. 26E). Additionally, wildtype cxHelix-C, but not L395D or 5RD cxHelix-C, interacted with the proteasome (FIG. 26E). Wild-type and L395D cxHelix-C also promoted the nuclear localization of MBP, but 5RD cxHelix-C remained cytosolic (FIG. 26F). Thus, cxHelix-C facilitates midnolin shuttling to the nucleus using importin |3 (KPNB1) and associates stably with the nuclear proteasome through PSMD2.

[0147] To investigate if cxHelix-C binds directly to importin |3 and the proteasome in a mutually exclusive manner, the interactions were reconstituted in vitro. KPNB1, KPNA2, and MsyB-tag fused to cxHelix-C were individually purified from bacteria (FIGs. 29C-29D), and the proteasome from human cells (FIG. 29E). cxHelix-C was then immunoprecipitated in the presence of importins or the proteasome. Wild-type and L395D cxHelix-C bound to importin |3 (KPNB1) while 5RD cxHelix-C failed to associate (FIG. 26G). No significant binding was observed between cxHelix-C and importin ex (KPNA2) (FIG. 26H), and only wild-type cxHelix-C associated with the proteasome (FIG. 261). cxHelix-C was immunoprecipitated in the presence of the proteasome and either KPNB1 or KPNA2. Importin |3 (KPNB1), but not importin ex (KPNA2), abrogated the proteasome- cxHelix-C interaction (FIG. 26J). Thus, importin |3 (KPNB1) and the proteasome compete for binding cxHelix-C in vitro.

[0148] It was hypothesized that if the midnolin-proteasome complex were to assemble in the cytosol, it might cause unproductive degradation of nuclear proteins before they can enter the nucleus to execute their functions, and of cytosolic proteins not meant to be degraded by the midnolin-proteasome pathway. The nuclear import receptor, KPNB1, may limit midnolin from associating with the cytosolic proteasome by sterically occluding cxHelix-C, allowing cells to assemble the midnolin-proteasome complex primarily in the nucleus.Example 10. The midnolin Ubl domain positions the Catch domain above the substrate entry site

[0149] The Ml state revealed that the midnolin Ubl domain bound the essential deubiquitinase PSMD14 / Rpnl l (FIG. 24). This interaction was conserved throughout evolution (FIGs. 30A-30B) and resembled how conjugated ubiquitin chains are normally cleaved from substrates by PSMD14 / Rpnl l during translocation (94) (FIG. 31A). The midnolin Ubl domain is followed by a short, unstructured linker that is positioned near the active site of PSMD14 as if it were ubiquitin ready to be cleaved.

[0150] It was previously showed that deletion of the Ubl domain prevented substrate degradation but did not disrupt the stable association of midnolin with the proteasome (88). The Global Protein Stability (GPS) reporter was employed to test the requirement of the Ubl- PSMD14 interaction for substrate degradation 104, 105). This dual fluorescent reporter expresses DsRed as an internal control along with green fluorescent protein (GFP) fused to the midnolin substrate IRF4 from the same bicistronic mRNA. The ratio of GFP / DsRed by flow cytometry provides a relative measure of the stability of the fusion protein. Wild-type and mutant versions of midnolin were overexpressed in M1DN knockout HEK-293T cells stably expressing the IRF4 reporter. Mutagenesis of the linker residues buried within PSMD14 (V99, V102, A104) abolished midnolin function, but disrupting the solvent-exposed linker residues (E103 and L106) had little or no effect (FIG. 31B). Thus, PSMD14 acts as a receptor for the midnolin Ubl domain.

[0151] The Catch domain is separated from the Ubl domain by a short linker sequence and is visible in the density map of the Ml proteasomal state. The Catch density is seen between the Ubl domain and ocHelix-C, right above the substrate entry site of the ATPase ring (FIG. 24). The Catch domain is low resolution likely because of its flexibility. An AlphaFold3 prediction ( 1 6) using the ATPase subunits, PSMD2, PSMD14, and midnolin confirmed the position of the Catch domain above the substrate entry site (FIG. 32). Interestingly, the midnolin ortholog of the segmented worm Dimorphilus gyrociliatus, which has a drastically compacted genome ( 107). contains only the Ubl domain, Catch domain, and ocHelix-C, with minimal unstructured regions in between (FIG. 33A). When expressed in human cells, D. gyrociliatus (worm) midnolin was nuclear, bound the proteasome, and promoted degradation of the EGR1 substrate (FIGs. 33B- 33D). AlphaFold3 predicted that D. gyrociliatus midnolin binds the proteasome identically to human midnolin (FIG. 31C). This minimal midnolin could only function with the Ml proteasomal state because the distance between PMSD2 (ochelix-C) and PMSD14 (Ubl domain) is too large in other states. These findings indicate that the Ml proteasomal state is crucial for the function of midnolin.Example 11. The Catch domain can enhance the substrate interaction by forming an FG zipper

[0152] The midnolin Catch domain selects substrates. AlphaFold suggests that the Catch domain contains two discontinuous, but semi-symmetrical regions each composed of two |3 strands and two or three ex helices {108). Using AlphaFold-multimer, it was previously found that an unstructured region within a midnolin substrate may adopt a |3 strand upon binding the Catch domain to complete an anti-parallel |3 sheet tertiary structure (88). The Catch domain could notbe observed with sufficient resolution to confirm the AlphaFold model by cryo-EM. Therefore, a 2.5 A resolution crystal structure of the Catch domain bound to the EGR1 substrate peptide was solved (FIG. 34A).

[0153] The substrate-free Catch domain, like full-length midnolin, was difficult to express and purify from bacteria. However, fusing the EGR1 peptide to the Catch domain drastically improved the expression in bacteria (FIG. 35A). This allowed us to purify protein that formed defined crystals (FIGs. 35B-35C). The crystal structure confirmed the |3 strand complementation mechanism. Among all proteins found to be midnolin substrates, EGR1 bound the Catch domain most effectively based on co-immunoprecipitation experiments (88). To understand the biochemical basis for this observation, the residues of EGR1 and midnolin buried within the Catch domain core were inspected. This led to the identification of alternating phenylalanine and glycine residues that were well-supported by the density and highly conserved throughout evolution (FIG. 35D and FIG. 36). This FG zipper could explain how EGR1 binds to midnolin so stably.

[0154] To test the requirement of the FG zipper, mutations were introduced within the EGR1 P strand and performed a co-immunoprecipitation experiment from cells endogenously expressing epitope-tagged midnolin. The midnolin-EGRl interaction was abrogated by substituting the FG zipper residues with alanine or changing the FG zipper from YxGxF to GxFxG (FIG. 34B).

[0155] Prior midnolin-substrate AlphaFold-multimer predictions were revisited (88). IRF4, NeuroDl, and SPINDOC represent midnolin substrates that bind the Catch domain less effectively than EGR1 and lack a complete FG zipper. It was tested if the midnolin-substrate interaction would improve with a complete FG zipper. IRF4 lacks one aromatic residue to complete the FG zipper. Replacement of valine 216 with tyrosine improved the midnolin-IRF4 binding (FIG. 34C). NeuroDl not only lacks one aromatic residue to complete the FG zipper but also contains phenylalanine 285 that may sterically clash with phenylalanine 280 of midnolin. Replacement of isoleucine 279 to tyrosine and phenylalanine 285 to leucine strongly improved the midnolin-NeuroDl binding (FIG. 34D). Finally, SPINDOC which lacks an FG zipper was tested. Replacement of glycine 319, leucine 321, and leucine 323 to tyrosine, glycine, and phenylalanine, respectively, drastically improved the midnolin-SPINDOC binding (FIG. 34E). Thus, the interaction between midnolin and its high-affinity substrates is driven not only by the complementation of a hydrophobic |3 strand, as previously reported, but also by specific residues that can form an FG zipper with the Catch domain.

[0156] To solidify the importance of the FG zipper, the zipper residues were swapped between midnolin and EGR1. Specifically, EGR1 was changed from YxGxF to GxFxG, while the Catch domain was changed from GxFxG to FxGxF. The wild-type EGR1 bound primarily to the wild-type midnolin, while the swapped EGR1 bound primarily to the swapped midnolin (FIG. 34F). This result suggests that the midnolin Catch domain can be engineered to bind other proteins of interest that do not bind to the wild-type Catch domain.Example 12. Model of the midnolin-proteasome pathway

[0157] Selective protein degradation is critical for cellular and organismal physiology. The ubiquitin-proteasome system (UPS) regulates this process through an elaborate network of more than 600 E3 ligases, which target proteins for ubiquitination and proteasomal degradation (709). In contrast, the midnolin-proteasome pathway represents a general mechanism that bypasses ubiquitin conjugation to degrade many nuclear proteins selectively. Here, structural and biochemical experiments were employed to elucidate the mechanism of the midnolin-proteasome pathway and propose the following model (FIG. 37).

[0158] After midnolin is synthesized, its ocHelix-C is bound by importin |3 (KPNB1), preventing midnolin association with the cytosolic proteasome and leading to nuclear transport. ocHelix-C is released from importin |3 (KPNB1) upon nuclear translocation and associates stably with the proteasome through PSMD2 / Rpnl. This mutually exclusive binding mechanism allows cells to assemble the midnolin-proteasome complex primarily in the nucleus. Substrates form a |3 strand to bind the Catch domain and the interaction can be augmented by a phenylalanine-glycine zipper. The midnolin Ubl domain binds PSMD14 / Rpnl 1, which is normally required to cleave the ubiquitin chain from substrates, but here serves as a receptor. The Ubl domain positions the adjacent Catch domain right above the ATPase ring to direct a bound substrate into the proteasome. The similarity of how the midnolin Ubl domain and ubiquitin engage PSMD14 / Rpnl l explains how midnolin initiates proteasomal degradation of bound substrates independently of ubiquitination. Whether PSMD14 / Rpnl 1 functions as a receptor to recruit other proteins that contain a Ubl domain is now possible.

[0159] It was previously observed that midnolin is very unstable and rapidly degraded by the proteasome (88). It remains unknown if midnolin acts catalytically or if midnolin is degraded stoichiometrically with a substrate. The latter could depend on the substrate binding affinity. For example, a tightly bound substrate may not dissociate from midnolin during translocation, dragging midnolin into the proteasome for degradation. It is also unclear if substrate binding affects the activity of midnolin or the proteasome. For example, substrate binding to the Catchdomain may allosterically impact midnolin orientation on the proteasome or facilitate proteasome activation. Establishing a reconstituted system to study the midnolin-proteasome pathway in vitro will elucidate these questions.

[0160] Overall, the molecular underpinnings of a non-canonical proteolysis pathway that is required for life and regulates development, sleep, and immune cell function is provided (89, 90, 110, 111). The presence of midnolin in the simple placozoan, Trichoplax sp H2, suggests that midnolin could have emerged in metazoans more than 800 million years ago concomitantly with multicellularity (112). The numerous and diverse substrates identified so far suggest the midnolin-proteasome pathway controls many physiological processes that await discovery.Materials and Methods related to Examples 1-7Cell culture

[0161] HEK-293T (ATCC, CRL-3216, RRID: CVCL_0063) and NIH / 3T3 cells (ATCC, CRL-1658, RRID: CVCL_0594) cells were cultured at 37°C and 5% CO2 in Dulbecco’s Modified Eagle’s Medium (DMEM) (Thermo Fisher Scientific, 11965118) supplemented with 100 units / mL of penicillin and 0.1 mg / mL of streptomycin (Thermo Fisher Scientific, 15070063) and 10% fetal bovine serum (Cytiva, SH30088.03). Ramos B cells (ATCC, CRL-1596, RRID: CVCL_0597) were cultured in RPMI 1640 medium (Thermo Fisher Scientific, A1049101) supplemented with 100 units / mL of penicillin and 0.1 mg / mL of streptomycin and 10% fetal bovine serum. NIH / 3T3 cells were starved overnight of serum and restimulated the following day with 20% serum for the indicated time points.

[0162] Animals were handled according to the protocol (IS00000074-6) approved by the Harvard University Office of the Institutional Animal Care and Use Committee, HMA Standing Committee on Animals and were in accordance with federal guidelines. Mouse cortical neurons were isolated and cultured as described previously (61). In short, embryonic cortices from wild type C57BL / 6NCrl mice (Charles River Laboratories, strain number: 027; 5-10 embryos: both males and females) were dissected at E16.5, and dissociated with papain (Sigma Aldrich, 10108014001). After terminating the papain digestion with ovomucoid (trypsin inhibitor from Worthington), cells were triturated gently through a P1000 pipette before passing through a 40- micron filter, then plated on cell culture dishes coated with poly-D-lysine (20 pg / mL) and laminin (4 pg / mL). The culture medium used for neurons was Neurobasal medium (GIBCO) that contains 2% B27 supplement, penicillin- streptomycin (50 U / mL penicillin and 50 U / mL streptomycin), and glutaMAX (1 mM). The neurons were cultured at 37°C and 5% CO2, treated with corresponding viruses at 3 days in vitro (DIV) while adding fresh culture medium at thesame time (35% of total volume), silenced on 10 DIV overnight by the addition of 1 |aM TTX (Abeam abl20055) and 100 |aM AP5 (Thermo Fisher 01-061-0), and harvested at 11 DIV after treatment with KC1 stimulation buffer comprised of: 52.8 mM KC1, 0.62 mM CaC12, 0.31 mM MgC12, 3.1 mM HEPES pH 7.4 for the indicated times.

[0163] Cells were treated with 20 ng / mL phorbol 12-myristate 13-acetate (PMA) (Thermo Fisher Scientific, J-63916-MA), 10 pM MG132 (Selleckchem, S2619), or 500 nM TAK-243 (Selleckchem, S8341) from lOOOx stock solution in DMSO for 6 hours unless stated otherwise.Plasmids and Cloning

[0164] The genome- wide CRISPR-Cas9 sgRNA Root library (5 sgRNAs / gene, 94,335 sgRNAs total) was used previously (62). The barcoded GPS ORFeome expression library was generated previously (13). The plasmids for human cDNAs containing a stop codon and conferring kanamycin resistance were obtained from the Ultimate ORF Clone collection (Thermo Fisher Scientific) in the form of Gateway entry clones: MIDN (IOH62653, BC094778.1), FOSB (IOH62162, NM_006732.2), FOS (IOH5624, NM_005252.3), ATF2 (IOH37849, NM_001880.2), CREB3 (IOH14437, BC009402.2), CREB5 (IOH53714, NM_001011666.1), ATF3 (IOH6465, NM_001674.2), BATF2 (IOH13295, BC012330.1), C0MMD9 (IOH12792, NM_014186.3), C11ORF31 (IOH58679, NM_170746.2), ZNF621 (IOH44483, NM_198484.3), SERTAD2 (IOH42292, NM_014755.2), LMX1A (IOH34878, NM_177398.2), LMX1B (IOH34707, NM_002316.2), HOXD3 (IOH5660, NM_006898.4), CHCHD2 (IOH3869, NM_016139.2), FNDC3B (IOH10620, BC012204.1), SOX12 (IOH40697, NM_006943.2), TAX1BP3 (IOH13074, NM_014604.2), PAX4 (IOH34754, BC074761.2), FOXA3 (IOH10014, NM_004497.2), STYX (IOH10157, NM_145251.3), ZNF764 (IOH6451, NM_033410.2), RelB (IOH11686, NM_006509.2), PRRX1 (IOH36664, NM_006902.3), IRF1 (IOH2022, NM_002198.2), IRF2 (IOH10126, NM_002199.3), IRF8 (IOH42114, NM_002163.2), IRF9 (IOH28745, NM_006084.4), MYC (IOH2954, P01106), SPINDOC (IOH28799, NM_138471.1), PPDPF (IOH4080, NM_024299.2), PAX8 (IOH3823, NM_003466.3), FOXS1 (IOH13387, NM_004118.3), NEURODI (IOH3394, NM_002500.2), MIER2 (IOH40210, NM_017550.1), IRF4 (IOH12141, NM_002460.2), GATA1 (IOH57792, NM_002049.3), CITED1 (IOH5542, BC004240.1), XRCC1 (IOH40644, NM_006297.2).

[0165] The plasmids for human cDNAs lacking a stop codon and conferring spectinomycin resistance were obtained from the Human ORFeome library V8.1 (Dana Farber Cancer Institute) in the form of Gateway entry clones: EGR1 (ORF_ID #14665, BC073983.1) and NR4A1 (ORF_ID #292, BC016147.1).

[0166] CBX4 and HA-tagged WT or K to R cDNA were generated by synthesis using IDT with attBl and attB2 overhangs for cloning into the pDONR221 (Thermo Fisher Scientific, 12536017) via a BP recombination reaction (Thermo Fisher Scientific, 11789020) to generate the entry clone. Similarly, the following attBl and attB2 overhangs were included in primers to generate fragments for sufficiency experiments by PCR for cloning into the pDONR221: attBl - GGGGACAAGTTTGTACAAAAAAGCAGGCTTAgccacc (SEQ ID NO: 5) attB2 - GGGGACCACTTTGTACAAGAAAGCTGGGTA (SEQ ID NO: 6)

[0167] Entry clones were mutagenized by PCR using the Q5 Site Directed Mutagenesis kit (NEB, E0554S) and the primers for mutagenesis were designed using the NEBaseChanger program.

[0168] Midnolin amino acid sequence used:MEPQPGGARSCRRGAPGGACELGPAAEAAPMSLAIHSTTGTRYDLAVPPDETVEGLRKR LSQRLKVPKERLALLHKDTRLSSGKLQEFGVGDGSKLTLVPTVEAGLMSQASRPEQSVM QALESLTETQVSDFLSGRSPLTLALRVGDHMMFVQLQLAAQHAPLQHRHVLAAAAAAA AARGDPSIASPVSSPCRPVSSAARVPPVPTSPSPASPSPITAGSFRSHAASTTCPEQMDCSP TASSSASPGASTTSTPGASPAPRSRKPGAVIESFVNHAPGVFSGTFSGTLHPNCQDSSGRPRRDIGTILQILNDLLSATRHYQGMPPSLAQLRCHAQCSPASPAPDLAPRTTSCEKLTAAPS ASLLQGQSQIRMCKPPGDRLRQTENRATRCKVERLQLLLQQKRLRRKARRDARGPYHW SPSRKAGRSDSSSSGGGGSPSEASGLGLDFEDSVWKPEANPDIKSEFVVA (SEQ ID NO: 1)

[0169] Midnolin regions for truncations or sufficiency experiments:AUbl - Residues 31-105ACatchl - Residues 112-156ACatch2 - Residues 266-332ANLS - Residues 402-413AaHelix-C - Residues 377-413 aHelix-C region fused to MBP - Residues 360-432Catch domain alone for sufficiency experiment - Residues 102-334ALoopl - Residues 160-260ALoop2 - Residues 163-257ALoop3 - Residues 169-251Catchl fused to 2xHA-GFP - Residues 107-166Catch2 fused to 2xFLAG-MBP - Residues 254-337FosB truncations:Deletion 1 - Residues 2-42Deletion2 - Residues 43-82Deletion3 - Residues 83-122Deletion4 - Residues 123-149Deletions - Residues 150-237AFosB - Residues 238-338Predicted P strand truncations:FOXS1 - Residues 245-260CBX4 - Residues 538-558NEURODI - Residues 277-288SPINDOC - Residues 314-328IRF4 - Residues 208-229FOSB (N-terminal P strand) - Residues 67-75EGR1 - Residues 128-145P strand sufficiency peptides:EGR1 - Residues 113-172IRF4 - Residues 192-248FosB - Residues 238-338

[0170] Entry clones were subcloned into the following lentiviral Gateway destination vectors using an LR recombination reaction (Thermo Fisher Scientific, 11791100): pHAGE-GPS 3.0 (13), pHAGE-GPS 3.2 (63), pHAGE-CMV-2xFLAG-N (this paper), pHAGE-CMV-2xHA-N (this paper), pHAGE-CMV Puromycin (64), pHAGE EFla BFP (this paper), CMV-C-2xFlag expression vector (Addgene, 118372), pHAGE TRE Blasticidin (this paper), pHAGE-CMV- 2xHA-GFP-N (this paper), pHAGE-CMV-2xFLAG-GFP-N (this paper), and pHAGE-CMV- 2xFLAG-MBP-N (this paper). The p!nducer20 mouseCD19 (64) plasmid was used to generate dox-on responsive HEK-293T cells.

[0171] Lentiviral CRISPR-Cas9 vectors containing the sgRNA of interest were cloned by first digesting the backbone lentiCRISPRv2 (Addgene, 52961) using BsmBI (NEB, R0739S). Then, sgRNA oligos containing CACC or AAAC overhangs were obtained from IDT,phosphorylated and annealed together, and ligated into the linear backbone by T4 (NEB, M0202S) ligation.Non-targeting sgControl - GTATTACTGATATTGGTGGG (SEQ ID NO: 7) Human sgMIDN #1 - GAAGCTGCAGGAGTTCGGCG (SEQ ID NO: 8) Human sgMIDN #2 - GCTGACCTTGGTACCCACCG (SEQ ID NO: 9) Mouse sgMIDN - GCGAGCTGAACACGGCCA (SEQ ID NO: 10)Lentivirus production

[0172] Lentivirus was generated by transfecting HEK-293T cells using PolyJet (SignaGen, SL100688) following the manufacturer’s instructions with plasmids encoding Tat, Rev, Gag-Pol, and VSV-G and lentiviral transfer vectors. Specifically, plasmid DNA was diluted into DMEM lacking supplements, and 3 pL of PolyJet reagent was used per 1 pg of plasmid DNA. One day post-transfection, the media was removed, and the cells were supplemented with fresh complete culture media. The lentiviral supernatant was collected 48- and 72-hours post- transfection, passed through a 0.45 pm filter, and either applied directly to cells or stored at -80°C for later use. For most experiments, lentivirus was packaged in 6-well plates. For library preparations, lentivirus was packaged in 8 15 cm plates using 13 pg of total DNA per plate for transfection, pooled, concentrated using the lenti-X concentrator (Takara, 631232), and aliquoted.Midnolin Overexpression and Flow cytometry

[0173] Cells were seeded in 6-well plates at 200,000 cells / well and were transfected two days later using Polyjet with EFla Midnolin co-expressing BFP, or BFP alone as a negative control. The media was changed one day post-transfection and cells were analyzed two days post-transfection.

[0174] Cells were prepared for flow cytometry by aspirating old media and rinsing cells once with PBS. The cells were detached using 0.05% trypsin at room temperature and the trypsin was neutralized using fresh complete media. The cells were then analyzed on a CytoFLEX S flow cytometer (Beckman Coulter, V2-B2-Y3-R2 version #C09762) and the CytExpert software (Beckman Coulter) was used to collect flow cytometry data. All flow cytometry data was analyzed using the FlowJo software. For fluorescence-activated cell sorting (FACS), a Sony MA900 was used for routine sorting of single clones, while a MoFlo Astrios (Beckman Coulter) instrument was used to collect cells for the CRISPR-Cas9 and GPS ORFeome screens.Generating doxycycline -inducible HEK-293T cells

[0175] Wild type and MIDN KO HEK-293T cells were infected with lentivirus encoding the p!nducer20 system expressing mouse CD19 as a doxycycline-inducible marker for cell-surface staining. After infection, the cells were treated with doxycycline (100 ng / mL) for 2 days and were stained on the cell surface using PE anti-mouse CD19 antibody (Biolegend, 152407, RRID: AB_2629816) for 30 minutes using 1 pL of antibody diluted in 100 pL of media containing 1 million cells. The cells were rinsed twice with PBS before sorting for the PE-positive population. The sorted cells were allowed to expand for more than 1 week before restaining the cells using PE anti-mouse CD 19 antibody in the absence of doxycycline induction. The PE- negative population was sorted, and the cells were expanded before another round of staining and sorting for the PE-negative population. This allowed for a population of cells that responded to doxycycline with minimal leakiness. Finally, lentivirus encoding the pHAGE TRE-Midnolin was infected into these MIDN KO HEK-293T cells and selected with blasticidin to allow for stable, doxycycline-inducible expression of midnolin.EGR1 and FosB CRISPR-Cas9 screens

[0176] Genome- wide CRISPR-Cas9 screens were performed to uncover regulators of EGR1 and FosB protein stability. Specifically, the plasmid library was packaged into lentivirus by transfecting HEK-293T cells using PolyJet as described earlier, and the lentivirus was titered to obtain a multiplicity of infection around 0.3. HEK-293T cells were generated to express the GPS 3.0 FosB or GPS 3.2 EGR1 reporters by selecting using hygromycin (200 pg / mL). These cells were then transduced with the titered CRISPR-Cas9 genome-wide Root library lentivirus at an MOI ~0.3 to maintain a 500x representation throughout. Cells were selected 48 hours posttransduction for 7 days using puromycin (2 pg / mL) to remove uninfected cells. On the ninth day of puromycin selection, the 95th percentile most stable cell population was collected based on the GFP / DsRed ratio by FACS using a MoFlo Astrios instrument (Beckman Coulter). Additionally, the unsorted input cells were collected based on the number of cells collected in the enriched population. Collected cells were rinsed once with PBS, pelleted, and stored in -80°C.Midnolin GPS ORFeome screen

[0177] The GPS ORFeome screen was performed as described previously (13) with some modifications. Sufficient cell numbers were used to maintain at least a 300-fold coverage of the library throughout. The library was packaged into lentivirus, which were used to transduce MIDN KO HEK-293T at a multiplicity of infection of 0.2. Two days post-transduction, theHEK-293T cells were treated with 2 pg / mE of puromycin for 6 days to remove uninfected cells, passaging the library once in between the selection period. The library-expressing cells were plated at 4 million cells / plate in a 15 cm dish and were transfected two days later using Polyjet with 8 pg DNA of EFla-Midnolin co-expressing BFP, or BFP alone as a negative control. The cells were harvested two days post-transfection and were sorted into six stability bins based on the GFP / DsRed ratio by FACS using a MoFlo Astrios instrument (Beckman Coulter). The sorting gates were established using the BFP control to ensure l / 6th of the population was collected per bin. Once the control populations were collected, the cells overexpressing midnolin were partitioned using the exact same sorting and gating settings as the control. The collected cells from each stability bin were rinsed once with PBS, pelleted, and frozen at -80°C for at least 12 hours.Deconvolution of the pooled screens

[0178] Cell pellets were thawed, and genomic DNA was harvested using a Centra Puregene Core Kit, Qiagen. The sgRNAs or barcodes were then amplified by PCR using all the genomic DNA as template (4 pg DNA per reaction) to include stagger sequences and Q5 Hot Start High- Fidelity DNA Polymerase from NEB. A second round of PCR was performed using the clean PCR1 product to add the Illumina P5 and P7 adaptor sequences. PCR2 samples were cleaned, pooled in the correct ratio, and sequenced on a NextSeq 500 instrument. The abundance of sgRNAs or barcodes were extracted from the raw sequencing data using Cutadapt (65) and were mapped onto the reference library using Bowtie2 (66).

[0179] MAGeCK was used to determine the enrichment of sgRNAs in the 95th percentile relative to the input population (56). The MAGeCK score plotted on the Y-axis represents the negative loglO of the “pos|score” value generated by MAGeCK.

[0180] For the GPS ORFeome analysis, the abundance of each ORF was corrected to account for sequencing depth and a protein stability index (PSI) score between 1 (most unstable) and 6 (most stable) was calculated using the following formula for each extracted ORF: PSI= _(i=l)A6 KRi*i2 , where i=the number of the stability bin denoted as an integer and Ri = the Illumina read proportion extracted from the bin i. The change in protein stability between midnolin and BFP is denoted as the difference in PSI (APSI).Gene set enrichment of GPS ORFeome hits

[0181] Gene set enrichment analyses (GSEA) was performed with GSEAPreranked (67) (v4.3.2) using rank weights derived from GPS ORFeome APSI values as input. These were tested for enrichment across the Human Molecular Signatures Database (68) (MSigDB v2022.1.Hs) C5 Gene Ontology collection (i.e. GO:BP, GO:MF, and GO:CC). The classic Kolmorogorov-Smirnov scoring scheme was used with 10,000 permutations and excluded gene sets with <10 or >1000 entries when intersected with the list of GPS ORFs.Generating MIDN knockout cells

[0182] To generate isogenic single clones lacking midnolin, HEK-293T cells were transfected with the lentiCRISPRv2 BFP plasmid encoding the given sgRNA of interest using Polyjet. Several days post-transfection, the BFP-positive cells were collected as single cells into 96- well plates using FACS. The single clones were allowed to expand for 2 weeks before screening for a knockout phenotype by immunoblotting as well as next generation sequencing of the genomic DNA locus encompassing the cut-site (FIG. 18A).

[0183] To generate a population-level depletion of midnolin in mouse NIH / 3T3 fibroblasts, cells were transduced with lentivirus encoding the given mouse lentiCRISPR v2 sgRNA coexpressing puromycin. After 3 days post transduction, cells were selected by 2 pg / mL puromycin for 5 days and expanded for further analysis.

[0184] To generate a population-level depletion of midnolin in primary cortical neurons, cells were transduced with lentivirus encoding the given mouse lentiCRISPRv2 sgRNA coexpressing puromycin.

[0185] To generate a population-level depletion of midnolin in Ramos B cells, the cells were transduced by spinf ection with lentivirus encoding the given lentiCRISPRv2 sgRNA# 1 coexpressing BFP. Specifically, the cells were incubated with lentivirus for 30 minutes with centrifugation at 2000 rpm at room temperature. The cells were allowed to expand for 6 days post transduction and the BFP-positive cells were collected by FACS.Generating 3xHA-MIDN knock-in HEK-293T cells

[0186] To generate an endogenous 3xHA-tagged midnolin cell line, the power of homology- directed repair (HDR) and CRISPR-Cas9 was employed. It was reasoned that an N-terminal epitope tag would be tolerable, as an N-terminal tagged midnolin transgene could interact with substrates and promote their degradation. To establish the knock-in line, the following custom Alt-R gRNA and ssDNA HDR template were generated by synthesis from IDT:gRNA: rCrCrGrGrGrCrUrGrCrGrGrCrUrCrCrArUrCrCrCrGrUrUrUrArArGrArGrCrUrArUrGrCrUrGr GrArArArCrArGrCrArUrArGrCrArArGrUrUrUrArArArUrArArGrGrCrUrArGrUrCrCrGrUrUr ArUrCrArArCrUrUrGrArArArArArGrUrGrGrCrArCrCrGrArGrUrCrGrGrUrGrCrUrUrUrUrUr UrU (SEQ ID NO: 11) ssDNA Template: CGGCGCCCGCCGCCCCCAGCCCCCCAGCGCGCGCCGGGGATGTATCCCTATGACGTG CCTGATTACGCCGGCGGAGGATCCTACCCCTATGATGTGCCTGACTACGCTGGCAGC GGAGGATACCCTTATGATGTGCCTGATTATGCTGGAGGTGGAGGTAGTGAGCCGCAG CCCGGCGGCGCCCGGAGCTGCCGGCGCGG (SEQ ID NO: 12)

[0187] spCas9, gRNA, and ssDNA template were introduced to cells by nucleofection. Specifically, per nucleofection of 100,000 cells in Lonza strip nucleofector system (V4XC-2032), mix 0.8 pL of 62.1 pM spCas9 (Aldevron, 9212-0.25MG), 0.8 pL of 100 pM gRNA (Alt-R from IDT), 0.25 pL of lOx NEB Buffer 3.1 (B6003S), and 0.65 pL of H2O to bring the final volume to 2.5 pL per nucleofection and incubate RNP at room temperature for 30 minutes. Making a master mix to troubleshoot the ssDNA template concentration is recommended to avoid small volumes. Then, add 2 pL of the RNP to 100,000 HEK-293T cells resuspended in 16.4 pL of NucleofectorTM solution plus 3.6 pL Supplement. Donor DNA was added directly to this solution at a final concentration of 500 nM or 2 pM. Cells were nucleofected using a 4D NucleofectorTM X Unit and a GFP positive control was included to ensure nucleofection worked properly. Once nucleofected, the cells were allowed to recover for 10 minutes at room temperature before adding the cells to 6-well plates containing warm media. The cells were allowed to expand for several days, and a cell lysate was collected to ensure the endogenous editing worked by immunoblotting. Then, single cells were partitioned into 96-well plates to obtain isogenic clones, which were validated by immunoblotting once expanded to ensure successful HA knock-in.Immunoprecipitation

[0188] Cells stably expressing the indicated epitope-tagged protein were cultured in 10 cm or 15 cm plates and allowed to reach 90% confluency before lysis. Alternatively, cells in 10 cm dishes were transiently transfected with 3 pg of the indicated plasmid DNA using Polyjet when 50% confluent. The media was changed one day post-transfection and the cells were lysed two days post-transfection with or without any necessary perturbations, such as proteasomalinhibition. For lysis, cells were rinsed once with ice-cold PBS by pouring and collected by scraping in 0.7 mL (10 cm plate) or 1 mL (15 cm plate) of lysis buffer containing 0.5% CHAPS, 40 mM HEPES pH 7.4, 100 mM NaCl, 4 mM EDTA, supplemented with lx protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific, 78441). Cell lysates were incubated with end-to-end rotation at 4°C for 30 minutes before clarification by centrifugation at 21,000xg for 15 minutes at 4°C. Anti-FLAG (Sigma, M8823, RRID: AB_2637089) or anti-HA (Thermo Fisher Scientific, 88836, RRID: AB_2749815) magnetic beads were rinsed three times in lysis buffer, using 15 pL of beads for every harvested plate. A 50 pL aliquot of the cell lysate was collected as input and the remaining supernatant was incubated with the beads for two hours at 4°C with end-to-end rotation. The immunoprecipitants were washed three times with the same lysis buffer and the cell lysates / immunoprecipitants were resuspended in Tris-Glycine SDS sample buffer (Thermo Fisher Scientific, LC2676) containing 10% 2-mercaptoethanol. Protein was eluted by heating at 95°C for 4 minutes before analysis of protein content by immunoblotting .Immunoblotting

[0189] For immunoprecipitation experiments, the collected samples were loaded directly for immunoblotting. For measuring steady-state abundance changes, cultured cells were lysed using lx RIPA buffer (Boston BioProducts, BP-115X) supplemented with lx protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific, 78441) for 15 minutes at 4°C. Samples were centrifuged at 21,000xg for 15 minutes at 4°C and protein concentration was normalized using a BCA assay (Thermo Fisher Scientific, 23225). Clarified supernatants were resuspended in Tris- Glycine SDS sample buffer (Thermo Fisher Scientific, LC2676) containing 10% 2- mercaptoethanol. Samples were loaded into 4-12% Tris-Glycine 15-well pre-cast gels (Thermo Fisher Scientific, XP04125BOX) and electrophoresis was run in lx Tris-Glycine SDS running buffer (Thermo Fisher Scientific, LC2675-4) at a constant 165-180 volts until the molecular weight ladder (Thermo Fisher Scientific, 26619) ran to the bottom. The protein within the gel was transferred to a 0.2 pm nitrocellulose membrane (BioRad, 170-4158) using the Trans-Blot Turbo Transfer System (BioRad). Nitrocellulose membranes were then blocked using 5% milk (LabScientific, M-0842) diluted in lx TBST (CST, 9997S) for at least 30 minutes at room temperature with gentle rocking. Primary antibodies were then diluted directly in the blocking solution at a 1:1000 dilution and incubated overnight at 4°C with gentle rocking. The following primary antibodies were used: rabbit anti-EGRl (CST, 4153, RRID: AB_2097038), rabbit anti- FosB (CST, 2251, RRID: AB_2106903), rabbit anti-c-Fos (in house) (69), rabbit anti-NR4Al (inhouse, warning has high background), rabbit anti-Midnolin (Proteintech, 18939-1-AP, RRID: AB_2878569), rabbit anti-PSMD2 (CST, 25430, RRID: AB_2798903), rabbit anti-PSMA2 (CST, 2455, RRID: AB_2171400), rabbit anti-HA (CST, 3724, RRID: AB_1549585), rabbit anti-FLAG (CST, 14793, RRID: AB_2572291), rabbit anti-mTOR (CST, 2983, RRID: AB_2105622), rabbit anti- Actin (CST, 4970, RRID: AB_2223172), rabbit anti-GAPDH (CST, 5174, RRID: AB_10622025), rabbit anti-CBX4 (CST, 30559, RRID: AB_2798991), rabbit anti- CBX8 (CST, 14696, RRID: AB_2687589), rabbit anti-ATF2 (CST, 35031, RRID: AB_2799069), rabbit anti-ATF3 (CST, 33593, RRID: AB_2799039), rabbit anti-IRFl (CST, 8478, RRID: AB_10949108), rabbit anti-RelB (CST, 4922, RRID: AB_2179173), rabbit anti- STAT3 (CST, 9139, RRID: AB_331757), rabbit anti-SPINDOC (Sigma, HPA040128, RRID: AB_10673027), rabbit anti-XRCCl (CST, 2735, RRID: AB_2218471), rabbit anti-CITEDl (Proteintech, 26999-1-AP, RRID: AB_2880718), rabbit anti-SOX12 (Proteintech, 23939-1-AP, RRID: AB_2879368), rabbit anti-FOXP3 (CST, 5298, RRID: AB_10839127), rabbit anti-c-Myc (CST, 5605, RRID: AB_1903938), rabbit anti-p27 (CST, 3686, RRID: AB_2077850), rabbit anti-Ubiquitin (CST, 43124, RRID: AB_2799235), and rabbit anti-IRF4 (CST, 4299, RRID: AB_10547141).

[0190] After overnight incubation, the blots were rinsed four times quickly and three additional times for longer, 10-minute incubations using lx TBST. After rinsing, the blots were incubated with 5% milk in lx TBST and the following secondary antibodies were applied directly at a 1:2000 dilution: anti-rabbit IgG, HRP-linked (CST, 7074, RRID: AB_2099233) or anti-mouse IgG, HRP-linked (CST, 7076, RRID: AB_330924). The blots were incubated in secondary antibody for 1 hour at room temperature with gentle rocking before rinsing as done for the primary antibody. The blots were exposed to either Pierce ECL western blotting substrate (Thermo Fisher Scientific, 32106) for strong antibodies or highly abundant protein, or Immobilon western chemiluminescent HRP substrate (Sigma, WBKLS0500) for weaker antibodies or less abundant proteins. All immunoblotting data was collected using high sensitivity autoradiography film (Denville Scientific, E3218).Mass spectrometry of endogenous midnolin immunoprecipitants

[0191] HEK-293T cells expressing endogenous 3xHA-tagged midnolin were cultured to 90% confluency in 5 15 cm plates per condition and unedited wild type HEK-293T cells were cultured in 5 15 cm plates. The knock-in cells were treated was DMSO or 10 pM MG 132 for 6 hours while the unedited wildtype HEK-293T served as background and were treated with 10 pM MG 132 for 6 hours. An anti-HA immunoprecipitation was performed using the same lysisconditions and protocol as described in the immunoprecipitation section of the methods. After the final wash, the beads were resuspended in 100 pL of 50 mM Tris pH 8.5 containing 5% SDS and the samples were heated at 95°C for 5 minutes to elute the proteins.

[0192] Eluted proteins were then digested using trypsin on S-Trap Micro columns (Protifi, C02-micro-10) following the manufacturer’s protocol. Specifically, proteins were first reduced using 5 mM TCEP for 15 minutes at 55°C and were subsequently alkylated with 20 mM iodoacetamide for 30 minutes in the dark at room temperature. After alkylation, the samples were acidified using phosphoric acid to a final concentration of 2.5% (v / v) and 10 volumes of 100 mM Tris, pH 7.55 in 90% methanol / 10% water were added to the samples to dilute the protein. This solution was then passed through S-Trap column by centrifuging for 30 seconds at 4,000xg. Multiple rounds of centrifugation were needed to load the entirety of one sample onto one column. Once the protein was trapped, the column was rinsed three times using 100 mM Tris, pH 7.55 in 90% methanol / 10% water, followed by a dry spin, before adding 2 pg of trypsin suspended in 20 pL 50 mM ammonium bicarbonate, pH 8. Columns were kept overnight at 37°C in a humid environment. After digestion, the peptides on the column were eluted by centrifuging three times for 1 minute at 4,000xg using three buffers applied sequentially: first 40 pL ammonium bicarbonate pH 8, second 40 pL 0.2% formic acid in water, and third 40 pL 50% acetonitrile in water. The pooled peptides were dried under reduced pressure using a SpeedVac and were resuspended in 30 pL 0.1% formic acid in water. LC-MS / MS data were acquired as reported previously (70) by injecting 10 pL of resuspended peptide sample.

[0193] A protein database consisting of the Human UniProt SwissProt proteome (downloaded on November 13th, 2022) was used to identify proteins that co-immunoprecipitated with endogenous 3xHA-midnolin. Specifically, the FragPipe graphical user interface (vl8.0) was used to search the data using the MSFragger search engine and to perform post-processing of the search results. The following parameters were used in the search. Tryptic peptides with a maximum of two missed cleavages were considered. Additionally, carbamidomethylation of cysteine was set as a fixed modification, and oxidation of methionine was allowed as a variable modification, with a maximum of four variable modifications per peptide. The allowed mass tolerances were 10 ppm for precursor ions and 0.04 Da for product ions. Peptide hits were filtered to a false discovery rate of 1% using PeptideProphet as implemented in FragPipe.Immunofluorescence

[0194] For experiments in FIGs. 12A-12B, 400,000 HEK-293T cells with indicated genetic background were plated on Poly-D-Eysine coated coverslips (TED PELLA, Inc.). On the following day, indicated treatments with DMSO, MG 132, or PMA were performed for 6 hoursbefore collection. Culture media was aspirated, and cells were washed with PBS once before fixation by 4% paraformaldehyde (PFA) in PBS for 15 minutes at room temperature. After three PBS washes, cells were permeabilized with 0.05% Triton in PBS for 5 minutes at room temperature. Cells were washed with PBS three times and placed in immunofluorescence blocking buffer (LI-COR, 927-70001) for 45 minutes at room temperature. Primary antibodies mouse anti-FLAG (Sigma, Fl 804, RRID: AB_262044) and rabbit anti-HA (CST, 3724, RRID: AB_1549585) were diluted 1: 400 in the blocking buffer and added on top of coverslips with cells for an overnight incubation at 4°C. After being rinsed three times with PBS, the cells were incubated in the dark at room temperature with secondary antibodies (1:500 dilution) and 8 pM Hoechst 33342 dye (Thermo Fisher Scientific, H3570), both diluted in blocking buffer, for 1 hour. Coverslips were rinsed thoroughly using PBS and were mounted on glass slides using ProLong gold antifade mountant (Thermo Fisher Scientific, P10144).

[0195] An Alexa 488-conjugated secondary antibody was utilized for the FLAG (Thermo Fisher Scientific, A-11001, RRID: AB_2534069) and HA (Thermo Fisher Scientific, A-11008, RRID: AB_143165) staining in FIGs. 12A-12B, the excitation wavelength was 488 nm. The excitation wavelength of the Hoechst nucleus-staining dye was 405 nm.

[0196] Image acquisition was done by a Zeiss AxioVert200M microscope with a 100X oil immersion objective as well as a Yokogawa CSU-22 spinning disk confocal head with a Borealis modification (Spectral Applied Research / Andor) along with a Hamamatsu ORCA-ER CCD camera. The image acquisition and hardware were controlled by the MetaMorph software package (Molecular Devices). The excitation lasers utilized to capture the images were 405 nm and 488 nm.RNA extraction and qPCR

[0197] Total RNA was extracted from cells with the RNeasy Plus Mini Kit (Qiagen, 74134) and cDNA was generated from freshly extracted RNA using the iScript cDNA Synthesis Kit (BioRad, 1708891) following the manufacturer instructions for both kits. Specifically, 250 ng of RNA was used for a 20 pL reaction to generate the cDNA. Platinum SYBR Green qPCR Supermix-UDG (Thermo Fisher Scientific, 11733038) and 2 pL of cDNA was used for qPCR reactions. Specifically, master mixes were prepared to contain 10 pL of SYBR, 7.5 pL of water, and 0.5 pL of 40x primers per 20 pL qPCR reaction. A Quantstudio 6 Pro (Thermo Fisher Scientific) was used to run the qPCR reactions. The following intercalating pre-mixed qPCR primers were obtained from IDT:Mouse MIDN (Mm.PT.58.10544931):GCGTCAACTTGCTCCCAT (SEQ ID NO: 13)AACGCCTCAAAGTACCCAAG (SEQ ID NO: 14)Mouse EGR1 (Mm.PT.58.29064929):GATAACTCGTCTCCACCATCG (SEQ ID NO: 15)AGCGCCTTCAATCCTCAAG (SEQ ID NO: 16)Mouse c-Fos (Mm.PT.58.29977214):GGCACTAGAGACGGACAGAT (SEQ ID NO: 17)ACAGCCTTTCCTACTACCATTC (SEQ ID NO: 18)Mouse FosB (Mm.PT.58.10990878):AGAGACACTTACCCCAGAAGA (SEQ ID NO: 19)GCTCTGCCTTTTCCTCTTCA (SEQ ID NO: 20)Mouse Actin (Mm.PT.39a.22214843.g):GACTCATCGTACTCCTGCTTG (SEQ ID NO: 21)GATTACTGCTCTGGCTCCTAG (SEQ ID NO: 22)Mouse mTOR (Mm.PT.58.28403918):TGCATCACTCGTTCATCCTG (SEQ ID NO: 23)AAGTCATCACATCCAAGCAGA (SEQ ID NO: 24)

[0198] Changes in mRNA levels were determined by subtracting the Cq values generated during the qPCR between the gene of interest and the control to yield a ACq value. Data were then normalized to the indicated control condition by substracting the ACq values by the average ACq of the indicated control condition to generate the AACq. Plotted in graphs are 2A(-AACq) from three biological replicates and the following statistical tests were used.

[0199] FIGs. 2E-2F: Data were analyzed using an ordinary one-way ANOVA followed by Tukey’s multiple comparisons test where **** represents p < 0.0001.

[0200] FIG. 8D: Data were analyzed using a two-way ANOVA followed by Sidak’s multiple comparisons test where ns is not significant and *** represents a p < 0.001.

[0201] Computational identification of substrate P strandsAlphaFold, multimer predictions

[0202] To identify P strands within hits identified in the ORFeome GPS screen, genes with APSI < -0.5 were taken (n=508) and the longest sequence across corresponding protein accession IDs (either NCBI Reference Sequence or Ensembl ID) was used as the input sequence for downstream steps (as barcodes from the screen were grouped at the gene level but could represent multiple isoforms). These sequences were individually paired with the MIDN sequence (UniProtKB: Q504T8) as a two-sequence FASTA file input into AlphaFold (v2.2.0) for multimer prediction with default reference databases specified as in (26) and max_template_date=2022-01- 01. Any selenocysteines were recoded as cysteines and three substrates (ACSBG2, ACSS2, and RIMBP3) that failed MSA using the default settings were rerun successfully by replacing the UniClust30_2018_08 database with UniRef30_2022_02.Identification of substrate [> strands within Midnolin [> sheet

[0203] The 25 ranked PDB models from each AlphaFold run with MIDN and one of the substrates were then processed by a custom Python script to identify PDB models that folded a linear stretch of the substrate into P strand conformation placed between P strands of the corresponding MIDN domain. In more detail, a pairwise distance matrix was first computed between each a-carbon atom in MIDN and each cr-carbon atom in the substrate as

[0204] where xi,yi, and zi are the coordinates of the ith substrate a-carbon atom and xj, yj, and zj are the coordinates of the jth MIDN a-carbon atom. As most P sheets have inter-strand distances < 5 A (71), the distance matrix was scanned to identify sequential substrate residues < 5.5 A from corresponding linear stretches within each adjacent MIDN P strand (ie. Di,j < 5.5 for both some sequential set of i with some sequential set of j, where 148 < j <157, as well as the same set of i with another sequential set of j, where 279 < j < 286).

[0205] Secondary structure assignment for the PDB model was done with the DSSP algorithm (72). Substrate residues satisfying the distance requirements specified above were then retained if they were assigned the extended P strand secondary structure (i.e., “E” coding). As DSSP relies on flanking residues to call secondary structure, the most N- and C-terminal residues are not assigned secondary structure. To avoid excluding them from P strand assignments, theywere assigned “E” coding if the adjacent residue had been assigned “E” coding. To catch residues that are part of a P strand, but slightly further from one or both of the MIDN P strands, this set of residues was then expanded by 7 residues in each direction and again only those with extended P strand secondary structure were kept. Finally, the longest contiguous stretch of P strand secondary structure was kept (if any) for final reporting.Relative disorder comparison

[0206] ORFs used in AlphaFold multimer folding alongside midnolin were matched with UniProtKB accessions using UniParc to find identical proteins that had already been folded as monomers in the AlphaFold Protein Structure Database (alphafold.ebi.ac.uk). This yielded existing models for 126 / 205 substrates predicted to interact with the midnolin Catch domain, for which the predicted local distance difference test (pLDDT) scores were extracted from corresponding PDB files. The pLDDT scores for these residues interacting with the midnolin Catch domain were averaged and compared with the average of the rest of the substrate (paired t- test, p = 2.72x10-9) as an approximation of disorder in the original substrate (60), where lower pLDDT scores correspond to increased disorder. In conclusion, the regions that are predicted to interact with the Catch domain are predicted to be unstructured by AlphaFold in the native, midnolin-free state.Amino acid enrichment

[0207] For each residue, n, going into the P strand from either the N- or C-terminal side, the overall frequency of each amino acid for P strands of length > 2n (to avoid double counting) was normalized by the background amino acid frequency across substrate sequences. Frequencies were then computed for amino acids preceding and following all P strands, provided the P strand was not the N- or C-terminus of the protein respectively.Hydrophobicity

[0208] The relative hydrophobicity of residues comprising the MIDN-interacting P strand was assessed by a two-sided paired t-test between the mean hydrophobicity index (58, 59) (at pH 7) of residues in the P strand with that of residues comprising the rest of the substrate.Identification ofMIDN-facing fi strand side chains

[0209] As P strands within a P sheet make contacts with each other through backbone interactions, where side chains altematingly project above and below the plane of the P sheet,residues composing the MIDN-interacting P strand can be parsed into those facing towards or away from the MIDN Catch domain. Side chains of the substrate P strand facing MIDN Catch domain were identified by first generating another distance matrix as before, but between each a- carbon atom in MIDN and each ^-carbon atom in the substrate. For each residue in the identified substrate P strand, those with ^-carbon distance to MIDN Catch domain (approximated by a- carbon position for MIDN isoleucine residue 309) less than their cr-carbon distance were annotated as MIDN-facing. Because glycine lacks a ^-carbon and residues at the ends of P strands may have more rotational variability, but side chain orientations along the P strand should alternate between facing towards or away from the MIDN Catch domain, a filter was then applied to ask whether the identified MIDN-facing side chains better matched either the set of odd or set of even residues and MIDN-facing side chains were then annotated as that set. Properties such as relative frequency and hydrophobicity were then computed for inward- and outward-facing side chains.Materials and Methods related to Examples 8-12Cell Culture

[0210] Human embryonic kidney (HEK)-293T (ATCC, CRL-3216, RRID: CVCL_0063) were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (Thermo Fisher Scientific, 11965118) containing 100 units / mL of penicillin and 0.1 mg / mL of streptomycin (Thermo Fisher Scientific, 15070063) and 10% fetal bovine serum (Cytiva, SH30088.03). Cells were kept at 37°C and 5% CO2. In the indicated experiments, cells were treated with 10 pM MG132 for 4 hours from lOOOx stock solution in dimethyl sulfoxide (DMSO).Plasmids

[0211] As previously described (88), plasmid cDNAs containing a stop codon and conferring kanamycin resistance were obtained from the human Ultimate ORF collection (Thermo Fisher Scientific) in the form of a Gateway entry clone: MIDN (IOH62653, BC094778.1), PSMD2 (IOH27806, BC007897.2), IRF4 (IOH12141, NM_002460.2), NeuroDl (IOH3394, NM_002500.2), and SPINDOC (IOH28799, NM_138471.1). The EGR1 Gateway entry clone (ORF_ID no. 14665, BC073983.1) lacking a stop codon and conferring spectinomycin resistance was obtained from the Human ORFeome library V8.1 (Dana Farber Cancer Institute). The entry clone for D. gyrociliatus midnolin was generated by synthesis.

[0212] The entry clone for the ocHelixC (residues 360 to 432) used in sufficiency experiments was generated by polymerase chain reaction (PCR) using primers containing thefollowing overhangs for cloning into pDONR221 via a BP reaction (Thermo Fisher Scientific, 11789020): attBl, GGGGACAAGTTTGTACAAAAAAGCAGGCTTAgccacc (SEQ ID NO: 5); attB2, GGGGACCACTTTGTACAAGAAAGCTGGGTA (SEQ ID NO: 6). EGR1 residues 113 to 172 were N-terminally appended to midnolin by PCR to generate an entry clone.

[0213] Mutations to entry clones were introduced by PCR using the Q5 Site-Directed Mutagenesis kit (NEB, E0554S). The PCR primers were designed using the NEBaseChanger online tool. Wild-type and mutant versions of entry clones were subcloned into the following destination vectors via an LR reaction (Thermo Fisher Scientific, 11791100): pHAGE CMV 2xFLAG-destination vector for N-terminally tagging MIDN, IRF4, NeuroDl, and SPINDOC, a pHAGE CMV 2xHA-destination vector for N-terminally tagging PSMD2, a CMV-C-2xFLAG destination vector (Addgene, 118372) for C-terminally tagging EGR1, a pHAGE CMV 2xFLAG-MBP-destination vector for N-terminally tagging the midnolin oc-HelixC, a GPS 3.0 destination vector for GFP-IRF4, a GPS 3.2 destination vector for EGR1-GFP, or a pHAGE EF lex-destination vector (blue fluorescent protein, BFP) for expressing untagged midnolin for flow cytometry experiments.

[0214] FLAG-MsyB-cxHelixC (WT, L395D, and R405 / 406 / 409 / 410 / 413D), KPNB 1,KPNA2, and the EGR1 -Catch domain oligonucleotides codon-optimized for expression in bacteria were generated by synthesis from Integrated DNA Technologies (IDT) to contain overhangs for Gibson assembly (NEB, E2611S). A pET 6xHis-3C bacterial expression vector obtained from the Dana-Farber Cancer Institute Crystallography Core Facility was digested with BamHI and Notl for direct Gibson assembly of the synthetic oligonucleotides.2xFLAG-EGRl -midnolin (for cryo-EM):MDYKDDDDKDYKDDDDKGGGGSGGGGFETSLYKKAGLATMEKVLVETSYPSQTTRLP PITYTGRFSLEPAPNSGNTLWPEPLFSLVSGLVSMTNPPASSSEPQPGGARSCRRGAPGGA CELGPAAEAAPMSLAIHSTTGTRYDLAVPPDETVEGLRKRLSQRLKVPKERLALLHKDT RLSSGKLQEFGVGDGSKLTLVPTVEAGLMSQASRPEQSVMQALESLTETQVSDFLSGRS PLTLALRVGDHMMFVQLQLAAQHAPLQHRHVLAAAAAAAAARGDPSIASPVSSPCRPV SSAARVPPVPTSPSPASPSPITAGSFRSHAASTTCPEQMDCSPTASSSASPGASTTSTPGASP APRSRKPGAVIESFVNHAPGVFSGTFSGTLHPNCQDSSGRPRRDIGTILQILNDLLSATRH YQGMPPSLAQLRCHAQCSPASPAPDLAPRTTSCEKLTAAPSASLLQGQSQIRMCKPPGDR LRQTENRATRCKVERLQLLLQQKRLRRKARRDARGPYHWSPSRKAGRSDSSSSGGGGS PSEASGLGLDFEDSVWKPEANPDIKSEFVVA (SEQ ID NO: 25)D. gyrociliatus midnolin:MNVNLIICPSTGGEYRLSVDLGENVNDLRRSVASKFRLDAERIRLIYNDRNLTTGTLEENDVGENSKIILLPCLRSGLVPKSVESTSVQIAEALKTLSGKQIDDFLCGRKPLNLRLILGGHF VYVQLQLSSTSKCPKQPSRIEDVRCHGSGVFSGTFSGTVRSPKSKNGASLVAGILNDILFS QRRRRDDQMTGEIRRAASRHSSPDKQRRVEDDQTRSKMAALKEALVERRRKKRKFSPY DRLLV (SEQ ID NO: 2)

[0215] The following proteins were expressed in bacteria using a pET plasmid:FLAG-MsyB-ocHelixC (wild type):MQLSHHHHHHSSGLEVLFQGPGGGSDYKDDDDKGGGGSMYATLEEAIDAAREEFLADNPGIDAEDANVQQFNAQKYVLQDGDIMWQVEFFADEGEEGECLPMLSGEAAQSVFDG DYDEIEIRQEWQEENTLHEWDEGEFQLEPPLDTEEGRAAADEWDERETSLYKKAGLATSASLLQGQSQIRMCKPPGDRLRQTENRATRCKVERLQLLLQQKRLRRKARRDARGPYHW SPSRKAGRSDSSSS (SEQ ID NO: 26)FLAG-MsyB-ocHelixC (L395D):MQLSHHHHHHSSGLEVLFQGPGGGSDYKDDDDKGGGGSMYATLEEAIDAAREEFLADNPGIDAEDANVQQFNAQKYVLQDGDIMWQVEFFADEGEEGECLPMLSGEAAQSVFDG DYDEIEIRQEWQEENTLHEWDEGEFQLEPPLDTEEGRAAADEWDERETSLYKKAGLATSASLLQGQSQIRMCKPPGDRLRQTENRATRCKVERDQLLLQQKRLRRKARRDARGPYHW SPSRKAGRSDSSSS (SEQ ID NO: 27)FLAG-MsyB-ocHelixC (5RD):MQLSHHHHHHSSGLEVLFQGPGGGSDYKDDDDKGGGGSMYATLEEAIDAAREEFLADNPGIDAEDANVQQFNAQKYVLQDGDIMWQVEFFADEGEEGECLPMLSGEAAQSVFDG DYDEIEIRQEWQEENTLHEWDEGEFQLEPPLDTEEGRAAADEWDERETSLYKKAGLATS ASLLQGQSQIRMCKPPGDRLRQTENRATRCKVERLQLLLQQKRLDDKADDDADGPYH WSPSRKAGRSDSSSS (SEQ ID NO: 28)KPNB1:MQLSHHHHHHSSGLEVLFQGPGGGGLNDIFEAQKIEWHESSGLEVLFQGPGSELITILEKTVSPDRLELEAAQKFLERAAVENLPTFLVELSRVLANPGNSQVARVAAGLQIKNSLTSK DPDIKAQYQQRWLAIDANARREVKNYVLQTLGTETYRPSSASQCVAGIACAEIPVNQWPELIPQLVANVTNPNSTEHMKESTLEAIGYICQDIDPEQLQDKSNEILTAIIQGMRKEEPSNNVKLAATNALLNSLEFTKANFDKESERHFIMQVVCEATQCPDTRVRVAALQNLVKIMSLY YQYMETYMGPALFAITIEAMKSDIDEVALQGIEFWSNVCDEEMDLAIEASEAAEQGRPP EHTSKFYAKGALQYLVPILTQTLTKQDENDDDDDWNPCKAAGVCLMLLATCCEDDIVP HVLPFIKEHIKNPDWRYRDAAVMAFGCILEGPEPSQLKPLVIQAMPTLIELMKDPSVVVR DTAAWTVGRICELLPEAAINDVYLAPLLQCLIEGLSAEPRVASNVCWAFSSLAEAAYEA ADVADDQEEPATYCLSSSFELIVQKLLETTDRPDGHQNNLRSSAYESLMEIVKNSAKDC YPAVQKTTLVIMERLQQVLQMESHIQSTSDRIQFNDLQSLLCATLQNVLRKVQHQDALQ ISDVVMASLLRMFQSTAGSGGVQEDALMAVSTLVEVLGGEFLKYMEAFKPFLGIGLKN YAEYQVCLAAVGLVGDLCRALQSNIIPFCDEVMQLLLENLGNENVHRSVKPQILSVFGDI ALAIGGEFKKYLEVVLNTLQQASQAQVDKSDYDMVDYLNELRESCLEAYTGIVQGLKG DQENVHPDVMLVQPRVEFILSFIDHIAGDEDHTDGVVACAAGLIGDLCTAFGKDVLKLV EARPMIHELLTEGRRSKTNKAKTLATWATKELRKLKNQA (SEQ ID NO: 29)KPNA2:MQLSHHHHHHSSGLEVLFQGPGGGGLNDIFEAQKIEWHESSGLEVLFQGPGSSTNENAN TPAARLHRFKNKGKDSTEMRRRRIEVNVELRKAKKDDQMLKRRNVSSFPDDATSPLQE NRNNQGTVNWSVDDIVKGINSSNVENQLQATQAARKLLSREKQPPIDNIIRAGLIPKFVS FLGRTDCSPIQFESAWALTNIASGTSEQTKAVVDGGAIPAFISLLASPHAHISEQAVWALG NIAGDGSVFRDLVIKYGAVDPLLALLAVPDMSSLACGYLRNLTWTLSNLCRNKNPAPPI DAVEQILPTLVRLLHHDDPEVLADTCWAISYLTDGPNERIGMVVKTGVVPQLVKLLGAS ELPIVTPALRAIGNIVTGTDEQTQVVIDAGALAVFPSLLTNPKTNIQKEATWTMSNITAGR QDQIQQVVNHGLVPFLVSVLSKADFKTQKEAVWAVTNYTSGGTVEQIVYLVHCGIIEPL MNLLTAKDTKIILVILDAISNIFQAAEKLGETEKLSIMIEECGGLDKIEALQNHENESVYK ASLSLIEKYFSVEEEEDQNVVPETTSEGYTFQVQDGAPGTFNF (SEQ ID NO: 30)EGRl-Catch domain (for crystallography):MQLSHHHHHHSSGLEVLFQGPGGSPPITYTGRFSLEPAPNSGNTLSRPEQSVMQALESLT ETQVSDFLSGRSPLTLALRVGDHMMFVQLQLAAQHAPPAPRSRKPGAVIESFVNHAPGV FSGTFSGTLHPNCQDSSGRPRRDIGTILQILNDLLSATRHYQGMPPSLAQLRCHA (SEQ ID NO: 31)Purification of the Midnolin-Proteasome Complex

[0216] Four purifications were performed for cryo-EM analysis. (1) The lysis buffer lacked ATP and magnesium and the bait was wild-type 2xFLAG-midnolin. (2) The lysis buffer contained both ATP and magnesium and the bait was wild-type 2xFLAG-midnolin. (3) The lysisbuffer contained both ATP and magnesium and the bait was G83E 2xFLAG-midnolin. (4) The lysis buffer contained both ATP and magnesium and the bait was 2xFLAG-EGRl-midnolin.

[0217] 3.5 million HEK-293T cells were seeded into 50 15 cm plates per prep. Two days after seeding, the cells were transfected using PolyJet (SignaGen, SL100688) with 10 pg DNA per plate of 2xFLAG-midnolin under the control of a CMV promoter (500 pL DMEM + DNA and 500 pL DMEM + 30 pL Poly Jet per plate). The media was changed one day after the transfection. Two days post-transfection, the cells were rinsed with ice-cold PBS and harvested by scraping in 1 mL of lysis buffer containing 40 mM HEPES, 100 mM NaCl, 2 mM MgCh, 5 mM ATP, lx HALT protease, 0.5% CHAPS per 15 cm plate. The lysate was incubated end-to- end for 20 minutes at 4°C before centrifuging at 21,000 x g. The clarified lysate was poured into two 50 mL falcon tubes. Using the lysis buffer, 1.8 mL of packed M2 FLAG agarose resin (Millipore, A2220) was rinsed four times in 2 mL Eppendorf tubes. Note that the FLAG M2 resin may change to a yellow / light brown color because of the HALT cocktail. The washed resin was added to the 50 mL falcons containing the clarified lysate and rotated end-to-end for 1.5 hours at 4°C. The falcon lids were sealed with parafilm to ensure there was no leaking. After incubating, the resin-containing cell lysate was flowed through a column (BioRad, 7321010) at 4°C by gravity. The two 50 mL falcons were rinsed with 15 mL lysis buffer for each falcon to wash off residual FLAG resin and the solution was flowed through the column. Finally, the resin on the column was rinsed once with 20 mL of lysis buffer, followed by 20 mL of size-exclusion chromatography (SEC) buffer containing 40 mM HEPES, 100 mM NaCl, 2mM MgCh, 5mM ATP, 0.05% CHAPS. The proteins were eluted from the resin using SEC buffer supplemented with 0.5 mg / mL of 3xFLAG peptide (APExBIO, A6001). Specifically, the resin on the column was incubated with 1.8 mL of elution buffer for 20 minutes at room temperature and the eluate was collected in a fresh tube. This process was repeated for a total of four elutions. To determine if protein was present in the elutions, an absorbance measurement at 280 nm was recorded using the elution buffer as a blank to determine if protein was present in the elutions.

[0218] The elutions containing proteins were pooled and concentrated (3,500 x g) using an Amicon Ultra-15, 10 kDa MWCO concentrator until -500 pL. It was important to pre-wet the filter using the SEC buffer before adding protein samples and to pipet up and down gently between centrifuging rounds to prevent unnecessary crashing. The concentration before SEC was around 4-5 mg / mL. The protein was centrifuged at 21,000 x g for 1 minute at 4°C to remove any debris. A Superose6 column was equilibrated using the SEC buffer and a 0.5 mL injection loop was washed thoroughly with the SEC buffer. The loop was then manually injected with 0.5 mL of concentrated eluted protein. The 0.5 mL fractions corresponding to the protein of interest were collected, pooled, and concentrated using Amicon Ultra-15, 10 kDa MWCO (4,800 x g, 6minutes intervals) to a final volume of around 300 |aL at 3-4 mg / mL of pure protein. The purified proteins were used directly for negative stain electron microscopy and cryo-EM sample preparation to prevent freezing and thawing of the sample.Negative Stain Electron Microscopy

[0219] 3.5 pl of midnolin-proteasome complexes purified as above was applied to glow- discharged (30 seconds, 30 mA) 400-mesh copper grids (EMS400-Cu) coated with a continuous layer of approximately 10 nm carbon. After 1 minute, the grid was blotted using filter paper (VMR, 28310-081), immediately stained with 1.5 % uranyl formate solution (Electron Microscopy Science, 22451), and blotted again. The staining procedure was repeated twice with a 1 -minute incubation with uranyl formate before the final blotting step. The grid was then airdried at room temperature before imaging. Images were collected using a FEI Tecnai T12 transmission electron microscope operating at 120 kV and equipped with a Gatan UltraScan 895 (4k x 4k) CCD camera at a nominal magnification of 67,000x corresponding to a calibrated pixel size of 1.68 A and a defocus value of -3.0 pm.Cryo-EM Sample Preparation and. Data Collection

[0220] 3.5 pl of 4 mg / mL midnolin-proteasome complexes purified as above was applied to glow-discharged R0.6 / 1 UltrAuFoil 300 mesh grids (Quantifoil) and frozen in liquid ethane using a Vitrobot Mark IV (Thermo Fisher Scientific) set at 4°C and 100% humidity with a 10 second wait time, 3 second blot time, and +8 blot force. All datasets were collected using a Titan Krios (Thermo Fisher Scientific) operating at 300 kV and equipped with a BioQuantum K3 imaging filter with a 20-eV slit width and a K3 summit direct electron detector (Gatan) in counting mode at a nominal magnification of 105,000x corresponding to a calibrated pixel size of 0.825 A. Semi-automated data collection was performed with SerialEM v4.0.5. For the WT MIDN -ATP dataset, a 2.597-second exposure was fractionated into 50 frames, resulting in a total exposure of 52.98 e- / A2. The defocus targets were -1.0 to -2.2 pm. For the WT MIDN +ATP dataset, a 2.652-second exposure was fractionated into 48 frames, resulting in a total exposure of 50.54 e- / A2. The defocus targets were -1.0 to -2.2 pm. For the G83E MIDN +ATP dataset, a 2.65- second exposure was fractionated into 50 frames, resulting in a total exposure of 53 e- / A2. The defocus targets were -1.0 to -2.2 pm. For the EGR1-MIDN +ATP dataset, a 3.02-second exposure was fractionated into 54 frames, resulting in a total exposure of 53.79 e- / A2. The defocus targets were -0.8 to -1.8 pm.Image Processing and Model Building

[0221] Data processing for all 4 datasets was performed in cryoSPARC v4.3.1 (FIG. 22). After patch-based motion correction and CTF estimation, micrographs with severe contamination or poor CTF fits were removed. 3,914 (prep 1, WT MIDN -ATP), 3,396 (prep 2, WT MIDN +ATP), 4,095 (prep 3, G83E MIDN +ATP) or 8,035 (prep 4, EGR1-MIDN +ATP) micrographs were subjected to automated particle picking using templates generated from blob-based picking. The particles were extracted with a box size of 800 and downsampled to a box size of 200 for the initial classification steps. After 2D classification, 380,637 (prep 1), 380,842 (prep 2), 285,529 (prep 3), and 658,210 (prep 4) particles were selected for heterogeneous refinement using multiple reference volumes generated by ab initio reconstruction. Particles in the best classes were subjected to homogeneous refinement with C2 symmetry, followed by symmetry expansion. Afterward, all particles were re-centered on the 19S particle, extracted with a box size of 440, and downsampled to a box size of 220. Good 19S particles after 2D classification were then subjected to multiple rounds of heterogeneous refinements.

[0222] 35,838 particles from prep 1, 2, and 3 datasets in classes with density corresponding to the Ubl domain were unbinned, local motion corrected, combined, and subjected to non- uniform and local refinement. 340,023 particles from the EGR1-MIDN +ATP dataset with a similar AAA+ ATPase state but no obvious Ubl density were subjected to 3D classification with a mask focused on PSMD2. 113,946 particles in the class with the clearest PSMD2 density were unbinned, local motion corrected, and subjected to non-uniform and local refinement. 228,049 particles from the EGR1-MIDN +ATP dataset with clear PSMD2 were subjected to an additional round of heterogeneous refinement to reveal two classes showing the midnolin ocHelix-C bound to PSMD2 and different AAA+ ATPase states. The particles from these classes were combined with particles in corresponding classes in prep 1, prep 2, and prep 4 datasets after unbinning, local motion correction, and additional heterogeneous refinement with a mask focused on PSMD2. 186,489 particles in the M2-A state and 87,326 particles in the M2-B state containing clear PSMD2 density were used for final non-uniform and local refinements.

[0223] Unsharpened and post-processed maps from cryoSPARC and deepEMhancer were used for interpretation. Local resolution estimation (FIGs. 23A-23C) and 3D variability analysis were performed in cryoSPARC. An initial model for the M2-A conformation was generated primarily by rigid body fitting individual AlphaFold models of each proteasomal subunit, the AlphaFold multimer model of PSMD2 with the midnolin ochelix-C, cross-referencing the output from ModelAngelo (773), particularly for the AAA+ ATPase subunits. The model was manually adjusted in Coot v.0.9.8 in between iterative rounds of real space refinement in Phenix v.1.19.2. The M2-A model was then used as an initial model for the M2-B conformation, which wasrefined the same way after adjustment of the AAA+ ATPase subunit conformations in Coot guided by the output from ModelAngelo. The Ml conformations were modeled in the same way, starting with rigid body fitting of individual AlphaFold models, including the AlphaFold- multimer model of PSMD14 with the midnolin Ubl domain, and cross-referencing with the output from ModelAngelo. Identification and placement of ZFAND5 in the Ml w / Ubl conformation was guided by PDB 7QXW (93). Figures were made using ChimeraX v.1.5 - v.1.7.Crystallography

[0224] Protein expression and purification: The chimeric EGR1 -Catch domain construct was overexpressed in E. coli BL21 (DE3) and purified using affinity chromatography and sizeexclusion chromatography. Briefly, cells were grown at 37°C in TB medium in the presence of 50 pg / mL of kanamycin to an OD of 0.8, cooled to 17°C, induced with 400 pM isopropyl- 1-thio- D-galactopyranoside (IPTG), incubated overnight for 20 hours at 17°C, collected by centrifugation, and stored at -80°C. Cell pellets were lysed in buffer A (25 mM HEPES, pH 7.5, 500 mM NaCl, and 1 mM TCEP) using a Microfluidizer (Microfluidics), and the resulting lysate was centrifuged at 30,000 x g for 40 minutes. Ni-NTA beads (Qiagen) were mixed with cleared lysate for 45 minutes and washed with buffer A. Beads were transferred to an FPLC-compatible column, and the bound protein was washed further with buffer A for 10 column volumes and eluted with buffer B (25 mM HEPES, pH 7.5, 500 mM NaCl, 10% glycerol, 0.5 mM TCEP, and 400 mM imidazole). The eluted sample was concentrated and purified further using a Superdex 200 16 / 600 column (Cytiva) in buffer C containing 20 mM HEPES, pH 7.5, 300 mM NaCl, and 0.5 mM TCEP. 3C protease was added to SEC fractions containing the EGRl-Catch domain, and His-tag was removed by applying it to a second Ni-NTA column. The eluant containing cleaved EGRl-Catch domain protein was concentrated to ~15 mg / mL and stored at -80°C.Crystallization: A solution of 15 mg / mL EGRl-Catch domain was crystallized in 1.4 M ammonium sulfate and 100 mM HEPES, pH 7.0 by sitting-drop vapor diffusion at 27°C. Crystals were transferred briefly into a crystallization buffer containing 25% glycerol before flashfreezing in liquid nitrogen.Determination: Diffraction data were collected at beamline 17-ID-2 of the National Synchrotron Light Source II (Brookhaven National Laboratory). Datasets were integrated and scaled using X- ray Detector Software (XDS) (114). Structures were solved by molecular replacement using the program Phaser (115) and the AlphaFold-predicted model as a search model. Iterative manual model building and refinement using Phenix (116) and Coot (117) led to a model with excellent statistics.Purification of Importins and MsyB-tagged ahelix-C

[0225] 20 L of chloramphenicol-resistant Rosetta (DE3) cells (Millipore, 71397-3) were transformed with pET plasmid conferring kanamycin resistance and IPTG-inducible expression of 6xHis-FLAG-MsyB-cxhelix-C, -KPNB1, or -KPNA2. Transformed cells were plated onto chloramphenicol and kanamycin and grown at 37°C overnight. The next day, single colonies were picked into 50 mL LB starter cultures containing chloramphenicol and kanamycin, and the bacteria were grown overnight at 37°C with 200 rpm shaking. The next day, the starter cultures were transferred to 1 L of fresh LB containing chloramphenicol and kanamycin, and the bacteria were grown at 37°C to an O.D. = 0.6. Once reaching this density, the bacteria were cooled at 4°C for 10 minutes, and 0.5 mM isopropyl |3- d-1 -thiogalactopyranoside IPTG (Sigma, 15502- 10G) was added directly to the liquid cultures. The bacteria were then grown at 16°C for 16-18 hours to induce protein expression.

[0226] Post induction, the bacteria were transferred to 1 L centrifuge bottles and pelleted at 4,000 x g for 15 minutes at 4°C. The pellet was resuspended in 35 mL of buffer containing 50 mM HEPES pH 7.4, 150 mM NaCl, and 10 mM imidazole (Sigma, 68268-500ML-F). The resuspended bacteria were again pelleted at 4,000 x g for 15 minutes at 4°C. The supernatant was removed, and the cells were resuspended in 35 mL of lysis buffer containing 50 mM HEPES pH 7.4, 150 mM NaCl, 10 mM imidazole, 1 mM dithiothreitol DTT (Thermo Fisher Scientific, R0862), benzonase (Millipore, 71206-3), and protease inhibitor tablets (Sigma, 11873580001). The cells were lysed using a microfluidizer (700 bar, 4 cycles, 20 mL / minute flow rate). Lysate was clarified by centrifuging at 21,000 x g for 20 minutes at 4°C. While centrifuging, 0.8 mL column volume (1.6 mL total volume per protein prep) of Ni-NTA agarose (Qiagen, 30210) was washed three times with lysis buffer. The supernatant was then incubated with the washed Ni- NTA agarose in a 50 mL conical tube for 30 minutes with end-to-end rotation at 4°C. After incubation, the mixture was subjected to gravity flow and both the conical tube, and the columns were rinsed with 50-60 mL total buffer at 4°C. The protein was eluted 4 times from the resin by sequentially incubating for 5 minutes with 500 pL of buffer containing 50 mM HEPES, 150 mM NaCl, and 200 mM imidazole. The absorbance A280 of each elution was measured using a nanodrop.

[0227] The elutions containing the highest amount of protein, usually elutions 1, 2, and 3, were pooled and concentrated to 500 pL using Amicon Ultra- 15, 3 kDa MWCO (Millipore, UFC9003). The protein was centrifuged at 21,000 x g for 1 minute to remove any debris. Superdex 200 size-exclusion chromatography was equilibrated using a buffer containing 50 mM HEPES, 150 mM NaCl, and 1 mM DTT. A 0.5 mL injection loop was washed thoroughly withthis same buffer and the loop was manually injected with 0.5 mL of concentrated eluted protein. The 0.5 mL fractions corresponding to the protein of interest were collected, pooled, and concentrated using Amicon Ultra-15, 3 kDa MWCO to a final volume of around 300-400 pL at 3-8 mg / mL of pure protein.Purification of Human Proteasome

[0228] HBTH-Rpnl 1 suspension HEK-293 cells 118 were grown in 600 mL of Expi293 expression media (Thermo Fischer Scientific, A1435101) to a final concentration of 5 million cells / mL. Cells were harvested by spinning at 1000 x g for 5 minutes. The media was removed, and the cell pellet was rinsed once with 100 mL of PBS before centrifuging again at 1000 x g for 5 minutes. The PBS was removed, and the cells were resuspended in 70 mL of lysis buffer containing 40 mM HEPES pH 7.4, 100 mM NaCl, 0.5 % CHAPS, 1 mM MgCl2, 1 mM DTT, 5 mM ATP, and lx HALT protease and phosphatase inhibitor. Cells were Dounce homogenized gently for 12 stokes / 2 mL of lysis buffer. The homogenate was subsequently incubated at 4°C for 20 minutes with end-to-end rotation. The lysate was then centrifuged at 21,000 x g for 25 minutes at 4°C to remove cell debris. During the spin, prepare ~2.5 mL slurry of NeutrAvidin Agarose (Thermo Fischer Scientific, 29200) and wash 4 times with lysis buffer to remove the storage buffer. Incubate the resin with the cleared supernatant, rotating end-to-end at 4°C for 2 hours. The beads were rinsed with lysis buffer 3 times, 20 mL each time, with gentle spinning at 2000 x g for 3 minutes at 4°C to precipitate the beads after each wash. The resin was then rinsed twice with 25 mL of TEB buffer containing 50 mM Tris pH 7.5, 50 mM NaCl, 2.5 mM ATP, 1 mM MgC12, and 1 mM DTT. The beads were then dried with a needle and vacuum before adding 4 mL of TEB containing 18 pL of HisTEV (Thermo Fischer Scientific, 12-575-015). Cleavage was allowed to occur overnight at 4°C with gentle shaking. The eluate (-4 mL) was collected. The beads were washed with an additional 2.5 mL of TEB buffer. The 6.5 mL pool containing the two elutes was concentrated using an Amicon Ultra-15, 10 kDa MWCO concentrator (Millipore, UFC9010) to a final volume of -500 pL containing about 0.6 mg / mL of human proteasomes. The solution was aliquoted and snap-frozen in liquid nitrogen for storage at -80°C.Coomassie Staining

[0229] Tris-Glycine gels soaked in Coomassie Brilliant Blue R-250 staining solution (BioRad, 1610436) were microwaved for 15 seconds and incubated at room temperature with gentle rocking for 30 minutes. Gels were then destained using 35 % methanol and 10 % acetic acid in water in the presence of a Kimwipe with gentle rocking overnight.Lentivirus Production

[0230] HEK-293T cells were seeded into 6-well culture plates at 300,000 cells / well. The cells were allowed to grow for 2 days before they were transfected using PolyJet (SignaGen, SL100688) following the manufacturer’s instructions. Briefly, plasmids encoding Tat, Rev, Gag- Pol, and VSV-G were mixed equimolar with lentiviral transfer vectors. Specifically, 1 pg of total plasmid DNA was diluted into DMEM lacking supplements. Separately, 3 pL of PolyJet reagent was diluted into DMEM lacking supplements. The Poly Jet solution was added to the DNA solution. The mixture was incubated for 15 minutes at room temperature and added to cells dropwise. The media was replaced with fresh media one day post-transfection. The lentiviral supernatant was collected at 48 hours post-transfection, passed through a 0.45 pm filter, and applied directly onto cells.Immunoprecipitation

[0231] HEK-293T cells were seeded at 1 million cells per 10 cm culture dish. Two days post-seeding, the cells were transfected with 3 pg of the indicated plasmid using PolyJet. The media was replaced with fresh media one day post-transfection. Two days post-transfection, the cells were rinsed once with ice-cold PBS and 700 pL of lysis buffer containing 0.5 % CHAPS, 100 mM NaCl, 40 mM HEPES pH 7.4, lx protease-phosphatase inhibitor cocktail (Thermo Fisher Scientific, 78441) was applied directly to the dish. The cells were harvested into Eppendorf tubes by scraping and lysates were incubated with end-to-end rotation for 15-20 minutes at 4°C. The tubes were then centrifuged at 21,000 x g for 15 minutes at 4°C. During the centrifuge step, 15 pL / plate of anti-FLAG (Sigma, M8823, RRID: AB_2637089) or anti-HA (Thermo Fisher Scientific, 88836, RRID: AB_2749815) magnetic beads were rinsed 3 times using 700 pL of lysis buffer. After centrifuging, a 30 pL aliquot of whole-cell lysate was collected into 200 pL of Tris-Glycine SDS sample buffer (Thermo Fisher Scientific, LC2676) containing 10% 2-mercaptoethanol. The remainder of the supernatant was incubated with the magnetic beads end-to-end for 1 hour at 4°C. The beads were then rinsed 3 times using 700 pL of lysis buffer and resuspended in 50 pL of Tris-Glycine SDS sample buffer. The protein samples were then denatured for 3 minutes at 95°C. The IP samples were diluted 1:20 in Tris-Glycine SDS sample buffer to blot for the bait protein. 15 pL of whole-cell lysate or IP sample was loaded to blot for co-immunoprecipitated proteins.Immunoblotting

[0232] Protein samples were resuspended in Tris-Glycine SDS sample buffer (Thermo Fisher Scientific, LC2676) containing 10% 2-mercaptoethanol and denatured for 3 minutes at95°C. The samples were then loaded into 4-12% Tris-Glycine 15-well pre-cast gels (Thermo Fisher Scientific, XP04125BOX) and electrophoresis was carried out at a constant 165 volts in lx Tris-Glycine SDS running buffer (Thermo Fisher Scientific, LC2675-4) until the molecular weight ladder (Thermo Fisher Scientific, 26619) reached the bottom of the gel. Proteins in the gel were transferred to a 0.2 pm nitrocellulose membrane (BioRad, 170-4158) using the Trans-Blot Turbo Transfer System (BioRad). The protein-containing membranes were blocked using 5% milk (LabScientific, M-0842) in lx TBST (CST, 9997S) for 15-30 minutes with gentle rocking at room temperature. All primary antibodies were diluted directly in the 5% milk blocking solution at a 1:1000 dilution by volume and incubated overnight with gentle rocking at 4°C.

[0233] The following primary antibodies were used: rabbit anti-PSMD2 (CST, 25430, RRID: AB_2798903), rabbit anti-PSMA2 (CST, 2455, RRID: AB_2171400), rabbit anti-HA (CST, 3724, RRID: AB_1549585), rabbit anti-FLAG (CST, 14793, RRID: AB_2572291), rabbit anti-mTOR (CST, 2983, RRID: AB_2105622), rabbit anti-KPNBl (CST, 51186, RRID: AB_2799386), rabbit anti-KPNA2 (CST, 14372, RRID: AB_2798467).

[0234] After incubating overnight in primary antibody, the membranes were rinsed 3 times using lx TBST as quick washes. The membranes were also incubated in lx TBST for three 10- minute washes with gentle rocking. After these rinses, the membranes were incubated with 5% milk in lx TBST containing a 1:2000 dilution by volume of anti-rabbit IgG, HRP-linked (CST, 7074, RRID: AB_2099233) secondary antibody. The membranes were incubated in secondary antibody for 1 hour at room temperature with gentle rocking before rinsing as done after the primary antibody incubation. After rinsing, the membranes were exposed to Pierce ECL western blotting substrate (Thermo Fisher Scientific, 32106) for 1-2 minutes at room temperature. High- sensitivity autoradiography film (Denville Scientific, E3218) was used to collect all immunoblotting data.Immunofluorescence

[0235] HEK-293T cells were seeded in 6-well culture plates containing Poly-D-Lysine coated coverslips (TED PELLA, Inc.) at 300,000 cells per well. On the following day, the culture media was aspirated, and the cells were rinsed with cold phosphate-buffered saline (PBS) once before fixation using 4% paraformaldehyde (PFA) diluted in PBS for 15 minutes at room temperature. The cells were then rinsed 3 times using PBS and the cells were permeabilized with 0.05% Triton diluted in PBS for 5 minutes at room temperature. The cells were rinsed 3 times with PBS and incubated in immunofluorescence blocking buffer (LLCOR, 927-70001) for 45 minutes at room temperature. The coverslips were then transferred faceup from the 6-well culture plate to a parafilm-coated 15 cm culture dish. Mouse anti-FLAG (Sigma, F1804, RRID:AB_262044) primary antibody was diluted 1:400 in the blocking buffer. The cells facing upward were incubated in 100 pL of the primary antibody solution by applying this solution directly on top of the cells. The cells were incubated in primary antibody overnight at 4°C in the 15 cm culture dish that was kept hydrated with a wet Kimwipe. The following day, the cells were rinsed by gently dipping the coverslips 50 times in and out of a 50 mL conical tube containing PBS. After rinsing, the coverslips were again placed faceup on a parafilm-coated dish. A 100 pL blocking solution containing a 1:500 dilution of anti-mouse Alexa 488-conjugated secondary antibody (Thermo Fisher Scientific, A-11001, RRID: AB_2534069) and 8 pM Hoechst 33342 dye (Thermo Fisher Scientific, H3570) was applied directly on top of the cells. The cells were incubated in this solution for 1 hour at room temperature in the dark. After incubation, the coverslips were again rinsed by gently dipping the coverslips 50 times in and out of a 50 mL conical tube containing PBS. The coverslips were mounted on glass slides using ProLong Glass antifade mountant (Thermo Fisher Scientific, P36982).

[0236] Images were acquired using a W1 Yokogawa Spinning disk Nikon Ti inverted confocal microscope with a 50 pm pinhole disk. All images were acquired using a Plan Apo 60X / 1.4 oil immersion objective and an Andor Zyla 4.2 Plus sCMOS monochrome camera. Nikon Elements Acquisition Software AR 5.02 was used to collect data. Excitation laser lines used were 405 nm and 488 nm.In Vitro Binding Assays

[0237] 10 pL of anti-FLAG beads (Sigma, M8823, RRID: AB_2637089) per reaction were rinsed 3 times using 700 pL of CHAPS buffer containing 100 mM NaCl, 40 mM HEPES pH 7.4, 0.5 % CHAPS, and lx HALT protease / phosphatase inhibitor. After rinsing, the beads were resuspended in 700 pL of CHAPS buffer containing 50 mg / mL of milk (LabScientific, M-0842). The beads were blocked for 1 hour at 4°C with end-to-end rotation. After blocking, the beads were rinsed 3 times with 700 pL of CHAPS buffer. The beads were transferred to low-binding, pre-lubricated Eppendorf tubes (Millipore, CLS3207) and resuspended in 500 pL of CHAPS buffer. 5 pL (19.5 pg per condition) of the indicated FLAG-MsyB-ocHelix-C pure protein was added to each tube. The beads were incubated for 1.5 hours at 4°C with end-to-end rotation to immobilize the bait. After incubation, the tubes were rinsed 3 times with 500 pL of CHAPS buffer and resuspended in 500 pL of CHAPS buffer containing the following proteins: KPNB1 (103 kDa) and KPNA2 (63 kDa) were kept equimolar.FIG. 26G: 7.9 pg KPNB1FIG. 26H: 4.9 pg KPNA2FIG. 261: 12 pg proteasomeFIG. 26J: 12 pg proteasome and / or 435 pg KPNB1 and / or 266 pg KPNA2

[0238] After adding the pure proteins, the beads were incubated end-to-end rotation for 1 hour at 4°C. After incubation, the beads were rinsed three times using 500 pL of CHAPS buffer. The beads were resuspended with 25 pL Tris-Glycine SDS sample buffer and boiled at 95°C for 3 minutes. The purified proteins were collected as input in the same sample buffer. The bait was diluted 1:20 in the same sample buffer for the anti-FLAG blot. 8 pL of each sample was loaded into gels for immunoblotting.Mass Spectrometry

[0239] Using lentivirus, MIDN mutant HEK-293T cells were reconstituted with 2xFLAG- tagged wild-type midnolin. MIDN mutant cells alone and mutant cells stably overexpressing 2xFLAG-tagged midnolin were each grown to >90% confluency in 5 15 cm plates. An anti- FLAG immunoprecipitation was performed as described in the immunoprecipitation section of the materials and methods, using 1 mL of lysis buffer per plate. After the final rinse, the anti- FLAG beads were resuspended with 100 pL of 50 mM Tris pH 8.5 containing 5% SDS, and the protein was eluted off the beads by heating at 95 °C for 4 minutes.

[0240] The eluted proteins were digested with trypsin on S-Trap Micro columns (Protifi, C02- micro- 10). Proteins were first reduced with 5 mM TCEP at 55 °C for 15 minutes and alkylated at room temperature with 20 mM iodoacetamide for 30 minutes in the dark. The proteins were then acidified using phosphoric acid to a final concentration of 2.5% (v / v). The protein was then diluted 10-fold with 100 mM Tris, pH 7.55 in 90% methanol / 10% water. The solution was passed through S-Trap columns by multiple rounds of centrifugation for 30 seconds at 4000 x g. The trapped protein was then rinsed 3 times with 150 pL of 100 mM Tris, pH 7.55 in 90% methanol / 10% water, ending with a dry spin at 4000 x g for 1 minute. A 20 pL solution of 50 mM ammonium bicarbonate, pH 8 containing 2 pg of trypsin was added directly to each column. Digestion was carried out at 37°C in a humid environment overnight. The peptides were eluted from the column by 3 sequential centrifugations for 1 minute at 4,000 x g using first 40 pL ammonium bicarbonate pH 8, second 40 pL 0.2% formic acid in water, and third 40 pL 50% acetonitrile in water. The pooled eluate was dried using a SpeedVac under reduced pressure and the peptides were resuspended in 30 pL 0.1% formic acid in water before data acquisition by LC- MS / MS as previously described 119).

[0241] A Human UniProt SwissProt proteome was downloaded on April 24th, 2022, and was used as the reference database to search for co-immunoprecipitated proteins. Briefly, FragPipe(vl8.0) was used to search the data using the MSFragger search engine. Tryptic peptides with at most two missed cleavages were considered. Fixed peptide modification included carbamidomethylation of cysteine, while oxidation of methionine was a variable modification with at most four variable modifications per peptide. The mass tolerances were 10 ppm for precursor ions and 0.04 Da for product ions. PeptideProphet was used to filter peptide hits to a false discovery rate of 1%. siRNA Transfection

[0242] HEK-293T were seeded in a 6-well plate at 250,000 cells. One day after seeding, the media was replaced with 1.5 mL of fresh culture media containing FBS / antibiotics. The following siRNAs from Dharmacon were resuspended in lx siRNA buffer (Dharmacon, B- 002000-UB-100) to obtain a 20 pM stock solution:Non-targeting Control (Dharmacon, D-001810-01-05) Non-targeting Control (Dharmacon, D-001810-02-05) Non-targeting Control (Dharmacon, D-001810-03-05) KPNB1 #1 (Dharmacon, J-017523-06-0005) KPNB1 #2 (Dharmacon, J-017523-07-0005) KPNB1 #3 (Dharmacon, J-017523-08-0005) KPNB1 #4 (Dharmacon, J-017523-09-0005) KPNA2 #1 (Dharmacon, J-004702-07-0005) KPNA2 #2 (Dharmacon, J-004702-08-0005) KPNA2 #3 (Dharmacon, J-004702-09-0005) KPNA2 #4 (Dharmacon, J-004702- 10-0005)

[0243] 6 pL of siRNA (20 pM) was diluted in 250 pL of Opti-MEM media (Thermo FisherScientific, 31985070) lacking FBS / antibiotics and mixed gently. Then, 5 pL of Lipofectamine RNAiMAX (Thermo Fisher Scientific, 13778075) was diluted in a separate 250 pL of Opti- MEM media lacking FBS / antibiotics. The RNAiMAX solution was immediately added to the siRNA-containing solution to obtain a 500 pL volume. This solution was allowed to incubate for 15-18 minutes at room temperature before adding the solution to cells, obtaining a 60 nM siRNA final concentration. One day post-transfection, the media was changed. The cells were harvested for downstream assays, such as immunoblotting and immunofluorescence 48 hours posttransfection.Midnolin Overexpression and Flow Cytometry

[0244] MIDN mutant HEK-293T cells, generated previously (SS), were infected with lentivirus to stably overexpress the GPS IRF4 reporter. Once this reporter cell line wasestablished, the cells were seeded in 6-well culture plates at 200,000 cells / well. After growing out the cells for 2 days, the cells were transfected with 1 pg of DNA using PolyJet to overexpress a control blue-fluorescent protein (BFP), or midnolin (wild type or mutant versions) together with BFP under an EFl ex promoter. The media was replaced with fresh media one day posttransfection. Two days post-transfection, the cells were detached from the culture plate using 0.05% trypsin and neutralized with fresh media. The cells were subsequently analyzed for BFP+, GFP+, and DsRed+ cells on a CytoFLEX S flow cytometer (Beckman Coulter, V2-B2-Y3-R2 version #C09762). The CytExpert software (Beckman Coulter) was used to collect flow cytometry data. Figures were generated using the FlowJo software.References1. K. Minatohara, M. Akiyoshi, H. Okuno, Role of Immediate-Early Genes in Synaptic Plasticity and Neuronal Ensembles Underlying the Memory Trace. Front Mol Neurosci 8, 78 (2015).2. S. B. McMahon, J. G. Monroe, The role of early growth response gene 1 (egr-1) in regulation of the immune response. J Leukoc Biol 60, 159-166 (1996).3. E. L. Yap, M. E. Greenberg, Activity-Regulated Transcription: Bridging the Gap between Neural Activity and Behavior. Neuron 100, 330-348 (2018).4. E. Benito, A. Barco, The neuronal activity-driven transcriptome. Mol Neurobiol 51, 1071-1088 (2015).5. S. Bahrami, F. Drablos, Gene regulation in the immediate-early response process. Adv Biol Regul 62, 37-49 (2016).6. G. Bossis, P. Ferrara, C. Acquaviva, I. Jariel-Encontre, M. Piechaczyk, c-Fos protooncoprotein is degraded by the proteasome independently of its own ubiquitinylation in vivo. Mol Cell Biol 23, 7425-7436 (2003).7. T. Gomard et al., Fos family protein degradation by the proteasome. Biochem Soc Trans 36, 858-863 (2008).8. H. C. Yen, Q. Xu, D. M. Chou, Z. Zhao, S. J. Elledge, Global protein stability profiling in mammalian cells. Science 322, 918-923 (2008).9. M. E. Greenberg, E. B. Ziff, Stimulation of 3T3 cells induces transcription of the c-fos proto-oncogene. Nature 311, 433-438 (1984).10. J. Y. Joo, K. Schaukowitch, L. Farbiak, G. Kilaru, T. K. Kim, Stimulus-specific combinatorial functionality of neuronal c-fos enhancers. Nat Neurosci 19, 75-83 (2016).11. A. R. Mardinly et al., Sensory experience regulates cortical inhibition by inducing IGF1 in VIP neurons. Nature 531, 371-375 (2016).12. N. Sagehashi et al., Insulin Enhances Gene Expression of Midnolin, a Novel Genetic Risk Factor for Parkinson's Disease, via Extracellular Signal-Regulated Kinase, Phosphoinositide 3- Kinase and Multiple Transcription Factors in SH-SY5Y Cells. J Pharmacol Exp Ther 381, 68-78 (2022).13. I. Koren et al., The Eukaryotic Proteome Is Shaped by E3 Ubiquitin Ligases Targeting C- Terminal Degrons. Cell 173, 1622-1635 el614 (2018).14. J. Du et al., Stuxnet Facilitates the Degradation of Polycomb Protein during Development. Dev Cell 37, 507-519 (2016).15. H. W. Mittrucker et al., Requirement for the transcription factor LSIRF / IRF4 for mature B and T lymphocyte function. Science 275, 540-543 (1997).16. U. Klein et al., Transcription factor IRF4 controls plasma cell differentiation and classswitch recombination. Nat Immunol 7, 773-782 (2006).17. E. L. Huttlin et al., Dual proteome- sc ale networks reveal cell-specific remodeling of the human interactome. Cell 184, 3022-3040 e3028 (2021).18. T. H. Wu, L. Shi, A. W. Lowe, M. R. Nicolls, P. N. Kao, Inducible expression of immediate early genes is regulated through dynamic chromatin association by NF45 / ILF2 and NF90 / NF110 / ILF3. PLoS One 14, e0216042 (2019).19. M. L. Hyer et al., A small-molecule inhibitor of the ubiquitin activating enzyme for cancer treatment. Nat Med 24, 186-193 (2018).20. A. Montagnoli et al., Ubiquitination of p27 is regulated by Cdk-dependent phosphorylation and trimeric complex formation. Genes Dev 13, 1181-1189 (1999).21. L. M. Tsvetkov, K. H. Yeh, S. J. Lee, H. Sun, H. Zhang, p27(Kipl) ubiquitination and degradation is regulated by the SCF(Skp2) complex through phosphorylated Thrl87 in p27. Curr Biol 9, 661-664 (1999).22. M. Weicker et al., Two diphosphorylated degrons control c-Myc degradation by the Fbw7 tumor suppressor. Sci Adv 8, eabl7872 (2022).23. M. M. Pena, Y. Y. Xing, S. Koli, F. G. Berger, Role of N-terminal residues in the ubiquitin-independent degradation of human thymidylate synthase. Biochem J 394, 355-363 (2006).24. K. N. Swatek, D. Komander, Ubiquitin modifications. Cell Res 26, 399-422 (2016).25. V. Chau et al., A multiubiquitin chain is confined to specific lysine in a targeted shortlived protein. Science 243, 1576-1583 (1989).26. J. Jumper et al., Highly accurate protein structure prediction with AlphaFold. Nature 596, 583-589 (2021).27. N. Zheng, N. Shabek, Ubiquitin Ligases: Structure, Function, and Regulation. Annu Rev Biochem 86, 129-157 (2017).28. I. Dikic, S. Wakatsuki, K. J. Walters, Ubiquitin-binding domains - from structures to functions. Nat Rev Mol Cell Biol 10, 659-671 (2009).29. C. Schauber et al., Rad23 links DNA repair to the ubiquitin / proteasome pathway. Nature 391, 715-718 (1998).30. S. Elsasser et al., Proteasome subunit Rpnl binds ubiquitin-like protein domains. Nat Cell Biol 4, 725-730 (2002).31. A. Bachmair, D. Finley, A. Varshavsky, In vivo half-life of a protein is a function of its amino-terminal residue. Science 234, 179-186 (1986).32. A. Varshavsky, Naming a targeting signal. Cell 64, 13-15 (1991).33. R. T. Timms, I. Koren, Tying up loose ends: the N-degron and C-degron pathways of protein degradation. Biochem Soc Trans 48, 1557-1567 (2020).34. A. Varshavsky, N-degron and C-degron pathways of protein degradation. Proc Natl Acad Sci U S A 116, 358-366 (2019).35. E. J. Nestler, M. Barrot, D. W. Self, DeltaFosB: a sustained molecular switch for addiction. Proc Natl Acad Sci U S A 98, 11042-11046 (2001).36. B. T. Hope et al., Induction of a long-lasting AP-1 complex composed of altered Fos-like proteins in brain by chronic cocaine and other chronic treatments. Neuron 13, 1235-1244 (1994).37. M. B. Kelz et al., Expression of the transcription factor deltaFosB in the brain controls sensitivity to cocaine. Nature 401, 272-276 (1999).38. S. C. Li, N. K. Goto, K. A. Williams, C. M. Deber, Alpha-helical, but not beta-sheet, propensity of proline is determined by peptide environment. Proc Natl Acad Sci U S A 93, 6676- 6681 (1996).39. M. Tsukahara, H. Suemori, S. Noguchi, Z. S. Ji, H. Tsunoo, Novel nucleolar protein, midnolin, is expressed in the mesencephalon during mouse development. Gene 254, 45-55 (2000).40. F. T. Gallo, C. Katche, J. F. Morici, J. H. Medina, N. V. Weisstaub, Immediate Early Genes, Memory and Psychiatric Disorders: Focus on c-Fos, Egrl and Arc. Front Behav Neurosci 12, 79 (2018).41. A. L. Shaffer et al., IRF4 addiction in multiple myeloma. Nature 454, 226-231 (2008).42. P. S. Biswas, G. Bhagat, A. B. Pernis, IRF4 and its regulators: evolving insights into the pathogenesis of inflammatory arthritis? Immunol Rev 233, 79-96 (2010).43. R. Nayar et al., TCR signaling via Tec kinase ITK and interferon regulatory factor 4 (IRF4) regulates CD8+ T-cell differentiation. Proc Natl Acad Sci U S A 109, E2794-2802 (2012).44. J. W. Williams et al., Transcription factor IRF4 drives dendritic cells to promote Th2 differentiation. Nat Commun 4, 2990 (2013).45. J. Eguchi et al., Transcriptional control of adipose lipid handling by IRF4. Cell Metab 13, 249-259 (2011).46. Y. Murakami, S. Matsufuji, S. Hayashi, N. Tanahashi, K. Tanaka, Degradation of ornithine decarboxylase by the 26S proteasome. Biochem Biophys Res Commun 267, 1-6 (2000).47. L. Ghoda, T. van Daalen Wetters, M. Macrae, D. Ascherman, P. Coffino, Prevention of rapid intracellular degradation of ODC by a carboxyl-terminal truncation. Science 243, 1493- 1495 (1989).48. Y. Murakami et al., Ornithine decarboxylase is degraded by the 26S proteasome without ubiquitination. Nature 360, 597-599 (1992).49. M. A. Hoyt, M. Zhang, P. Coffino, Ubiquitin-independent mechanisms of mouse ornithine decarboxylase degradation are conserved between mammalian and fungal cells. J Biol Chem 278, 12135-12143 (2003).50. J. S. Merkel, J. M. Sturtevant, L. Regan, Sidechain interactions in parallel beta sheets: the energetics of cross-strand pairings. Structure 7, 1333-1343 (1999).51. I. Marin, The ubiquilin gene family: evolutionary patterns and functional insights. BMC Evol Biol 14, 63 (2014).52. S. Raasi, C. M. Pickart, Rad23 ubiquitin-associated domains (UBA) inhibit 26 S proteasome-catalyzed proteolysis by sequestering lysine 48-linked polyubiquitin chains. J Biol Chem 278, 8951-8959 (2003).53. L. Chen, U. Shinde, T. G. Ortolan, K. Madura, Ubiquitin-associated (UBA) domains in Rad23 bind ubiquitin and promote inhibition of multi-ubiquitin chain assembly. EMBO Rep 2, 933-938 (2001).54. K. K. Joshi, M. Berge, S. K. Radhakrishnan, P. H. Viollier, P. Chien, An Adaptor Hierarchy Regulates Proteolysis during a Bacterial Cell Cycle. Cell 163, 419-431 (2015).55. N. J. Kuhlmann, D. Doxsey, P. Chien, Cargo competition for a dimerization interface restricts and stabilizes a bacterial protease adaptor. Proc Natl Acad Sci U S A 118, (2021).56. W. Li et al., MAGeCK enables robust identification of essential genes from genome-scale CRISPR / Cas9 knockout screens. Genome Biol 15, 554 (2014).57. G. A. Collins, A. L. Goldberg, Proteins containing ubiquitin-like (Ubl) domains not only bind to 26S proteasomes but also induce their activation. Proc Natl Acad Sci U S A 117, 4664- 4674 (2020).58. O. D. Monera, T. J. Sereda, N. E. Zhou, C. M. Kay, R. S. Hodges, Relationship of sidechain hydrophobicity and alpha-helical propensity on the stability of the single- stranded amphipathic alpha-helix. J Pept Sci 1, 319-329 (1995).59. T. J. Sereda, C. T. Mant, F. D. Sonnichsen, R. S. Hodges, Reversed-phase chromatography of synthetic amphipathic alpha-helical peptides as a model for ligand / receptor interactions. Effect of changing hydrophobic environment on the relative hydrophilicity / hydrophobicity of amino acid side-chains. J Chromatogr A 676, 139-153 (1994).60. K. Tunyasuvunakool et al., Highly accurate protein structure prediction for the human proteome. Nature 596, 590-596 (2021).61. N. Sharma et al., ARNT2 Tunes Activity-Dependent Gene Expression through NCoR2- Mediated Repression and NPAS4-Mediated Activation. Neuron 102, 390-406 e399 (2019).62. J. G. Doench et al., Optimized sgRNA design to maximize activity and minimize off- target effects of CRISPR-Cas9. Nat Biotechnol 34, 184-191 (2016).63. R. T. Timms et al., A glycine- specific N-degron pathway mediates the quality control of protein N-myristoylation. Science 365, (2019).64. T. D. Martin et al., The adaptive immune system is a major driver of selection for tumor suppressor gene inactivation. Science 373, 1327-1335 (2021).65. M. Martin, Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet.joumal 17, 10-12 (2011).66. B. Langmead, S. L. Salzberg, Fast gapped-read alignment with Bowtie 2. Nat Methods 9, 357-359 (2012).67. A. Subramanian et al., Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 102, 15545-15550 (2005).68. A. Liberzon et al., Molecular signatures database (MSigDB) 3.0. Bioinformatics 27, 1739-1740 (2011).69. Y. Lin et al., Activity-dependent regulation of inhibitory synapse development by Npas4. Nature 455, 1198-1204 (2008).70. C. Nardone et al., A central role for regulated protein stability in the control of TFE3 and MITF by nutrients. Mol Cell 83, 57-73 e59 (2023).71. J. Martin et al., Protein secondary structure assignment revisited: a detailed analysis of different assignment methods. BMC Struct Biol 5, 17 (2005).72. W. Kabsch, C. Sander, Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22, 2577-2637 (1983).73. X. Gu, C. Nardone et al., The midnolin-proteasome pathway catches proteins for ubiquitination-independent degradation. Dryad, Dataset, DOI: 10.506 l / dryad.m905qfv6g. (2023).74. J. A. M. Bard et al., Structure and Function of the 26S Proteasome. Annu Rev Biochem 87, 697-724 (2018).75. A. Hershko, A. Ciechanover, The ubiquitin system. Annu Rev Biochem 67, 425-479 (1998).76. S. Elsasser et al., Proteasome subunit Rpnl binds ubiquitin-like protein domains. Nat Cell Biol 4, 725-730 (2002).77. S. Elsasser, D. Chandler-Militello, B. Muller, J. Hanna, D. Finley, Rad23 and RpnlO serve as alternative ubiquitin receptors for the proteasome. J Biol Chem 279, 26817-26822 (2004).78. K. Husnjak et al., Proteasome subunit Rpnl3 is a novel ubiquitin receptor. Nature 453, 481-488 (2008).79. Y. Shi et al., Rpnl provides adjacent receptor sites for substrate binding and deubiquitination by the proteasome. Science 351, (2016).80. A. J. Boughton, L. Liu, T. Lavy, O. Kleifeld, D. Fushman, A novel recognition site for polyubiquitin and ubiquitin-like signals in an unexpected region of proteasomal subunit Rpnl. J Biol Chem 297, 101052 (2021).81. X. Chen et al., Structures of Rpnl Tl:Rad23 and hRpnl3:hPLIC2 Reveal Distinct Binding Mechanisms between Substrate Receptors and Shuttle Factors of the Proteasome. Structure 24, 1257-1270 (2016).82. S. Zhang et al., USP14-regulated allostery of the human proteasome by time-resolved cryo-EM. Nature 605, 567-574 (2022).83. J. Rabi et al., Mechanism of gate opening in the 20S proteasome by the proteasomal ATPases. Mol Cell 30, 360-368 (2008).84. R. Verma et al., Role of Rpnl 1 metalloprotease in deubiquitination and degradation by the 26S proteasome. Science 298, 611-615 (2002).85. T. Yao, R. E. Cohen, A cryptic protease couples deubiquitination and degradation by the proteasome. Nature 419, 403-407 (2002).86. E. J. Worden, C. Padovani, A. Martin, Structure of the Rpnl 1-Rpn8 dimer reveals mechanisms of substrate deubiquitination during proteasomal degradation. Nat Struct Mol Biol 21, 220-227 (2014).87. M. E. Matyskiela, G. C. Lander, A. Martin, Conformational switching of the 26S proteasome enables substrate degradation. Nat Struct Mol Biol 20, 781-788 (2013).88. X. Gu et al., The midnolin-proteasome pathway catches proteins for ubiquitination- independent degradation. Science 381, eadh5021 (2023).89. J. Du et al., Stuxnet Facilitates the Degradation of Polycomb Protein during Development. Dev Cell 37, 507-519 (2016).90. S. M. Kweon et al., Heterozygous midnolin knockout attenuates severity of nonalcoholic fatty liver disease in mice fed a Western-style diet high in fat, cholesterol, and fructose. Am J Physiol Gastrointest Liver Physiol 325, G147-G157 (2023).91. M. Tsukahara, H. Suemori, S. Noguchi, Z. S. Ji, H. Tsunoo, Novel nucleolar protein, midnolin, is expressed in the mesencephalon during mouse development. Gene 254, 45-55 (2000).92. N. Sagehashi et al., Insulin Enhances Gene Expression of Midnolin, a Novel Genetic Risk Factor for Parkinson's Disease, via Extracellular Signal-Regulated Kinase, Phosphoinositide 3- Kinase and Multiple Transcription Factors in SH-SY5Y Cells. J Pharmacol Exp Ther 381, 68-78 (2022).93. D. Lee et al., Molecular mechanism for activation of the 26S proteasome by ZFAND5. Mol Cell 83, 2959-2975 e2957 (2023).94. A. H. de la Pena, E. A. Goodall, S. N. Gates, G. C. Lander, A. Martin, Substrate-engaged 26S proteasome structures reveal mechanisms for ATP -hydrolysis-driven translocation. Science 362, (2018).95. Y. Dong et al., Cryo-EM structures and dynamics of substrate-engaged human 26S proteasome. Nature 565, 49-55 (2019).96. R. Evans et al., Protein complex prediction with AlphaFold-Multimer. bioRxiv, 2021.2010.2004.463034 (2022).97. A. Lange et al., Classical nuclear localization signals: definition, function, and interaction with importin alpha. J Biol Chem 282, 5101-5105 (2007).98. D. S. Goldfarb, A. H. Corbett, D. A. Mason, M. T. Harreman, S. A. Adam, Importin alpha: a multipurpose nuclear-transport receptor. Trends Cell Biol 14, 505-514 (2004).99. G. Cingolani, C. Petosa, K. Weis, C. W. Muller, Structure of importin-beta bound to the IBB domain of importin-alpha. Nature 399, 221-229 (1999).100. M. K. Iovine, J. L. Watkins, S. R. Wente, The GLFG repetitive region of the nucleoporin Nupl l6p interacts with Kap95p, an essential yeast nuclear import factor. J Cell Biol 131, 1699- 1713 (1995).101. A. Radu, M. S. Moore, G. Blobel, The peptide repeat domain of nucleoporin Nup98 functions as a docking site in transport across the nuclear pore complex. Cell 81, 215-222 (1995).102. R. Bayliss, T. Littlewood, L. A. Strawn, S. R. Wente, M. Stewart, GLFG and FxFG nucleoporins bind to overlapping sites on importin-beta. J Biol Chem 277, 50597-50606 (2002).103. N. H. Cho et al., OpenCell: Endogenous tagging for the cartography of human cellular organization. Science 375, eabi6983 (2022).104. H. C. Yen, Q. Xu, D. M. Chou, Z. Zhao, S. J. Elledge, Global protein stability profiling in mammalian cells. Science 322, 918-923 (2008).105. H. C. Yen, S. J. Elledge, Identification of SCF ubiquitin ligase substrates by global protein stability profiling. Science 322, 923-929 (2008).106. J. Abramson et al., Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature 630, 493-500 (2024).107. J. M. Martin-Duran et al., Conservative route to genome compaction in a miniature annelid. Nat Ecol Evol 5, 231-242 (2021).108. J. Jumper et al., Highly accurate protein structure prediction with AlphaFold. Nature 596, 583-589 (2021).109. D. Popovic, D. Vucic, I. Dikic, Ubiquitination in disease pathogenesis and treatment. Nat Med 20, 1242-1253 (2014).110. Z. Zhao et al., Epigenetic regulator Stuxnet modulates octopamine effect on sleep through a Stuxnet-Polycomb-Octbeta2R cascade. EMBO Rep 22, e47910 (2021).111. X. Zhong et al., Viable mutations of mouse midnolin suppress B cell malignancies. J Exp Med 221, (2024).112. S. R. Najle et al., Stepwise emergence of the neuronal gene expression program in early animal evolution. Cell 186, 4676-4693 e4629 (2023).113. K. Jamali et al., Automated model building and protein identification in cryo-EM maps.Nature, (2024).114. W. Kabsch, Xds. Acta Crystallogr D Biol Crystallogr 66, 125-132 (2010).115. A. J. McCoy et al., Phaser crystallographic software. J Appl Crystallogr 40, 658-674 (2007).116. D. Liebschner et al., Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr D Struct Biol 75, 861-877 (2019).117. P. Emsley, B. Lohkamp, W. G. Scott, K. Cowtan, Features and development of Coot.Acta Crystallogr D Biol Crystallogr 66, 486-501 (2010).118. X. Wang et al., Mass spectrometric characterization of the affinity-purified human 26S proteasome complex. Biochemistry 46, 3553-3565 (2007).119. C. Nardone et al., A central role for regulated protein stability in the control of TFE3 and MITF by nutrients. Mol Cell 83, 57-73 e59 (2023).EQUIVALENTS AND SCOPE

[0245] In the claims articles such as “a,” “an,” and “the” may mean one or more than one unless indicated to the contrary or otherwise evident from the context. Claims or descriptions that include “or” between one or more members of a group are considered satisfied if one, more than one, or all of the group members are present in, employed in, or otherwise relevant to a given product or process unless indicated to the contrary or otherwise evident from the context. The disclosure includes embodiments in which exactly one member of the group is present in, employed in, or otherwise relevant to a given product or process. The disclosure includes embodiments in which more than one, or all of the group members are present in, employed in, or otherwise relevant to a given product or process.

[0246] Furthermore, the disclosure encompasses all variations, combinations, and permutations in which one or more limitations, elements, clauses, and descriptive terms from one or more of the listed claims is introduced into another claim. For example, any claim that is dependent on another claim can be modified to include one or more limitations found in any other claim that is dependent on the same base claim. Where elements are presented as lists, e.g., in Markush group format, each subgroup of the elements is also disclosed, and any element(s) can be removed from the group. It should it be understood that, in general, where the disclosure, or aspects described herein, is / are referred to as comprising particular elements and / or features, certain embodiments described herein or aspects described herein consist, or consist essentially of, such elements and / or features. For purposes of simplicity, those embodiments have not been specifically set forth in haec verba herein. It is also noted that the terms “comprising” and “containing” are intended to be open and permits the inclusion of additional elements or steps. Where ranges are given, endpoints are included. Furthermore, unless otherwise indicated or otherwise evident from the context and understanding of one of ordinary skill in the art, values that are expressed as ranges can assume any specific value or subrange within the stated ranges in different embodiments described herein, to the tenth of the unit of the lower limit of the range, unless the context clearly dictates otherwise.

[0247] This application refers to various issued patents, published patent applications, journal articles, and other publications, all of which are incorporated herein by reference. If there is a conflict between any of the incorporated references and the instant specification, the specification shall control. In addition, any particular embodiment of the present disclosure that falls within the prior art may be explicitly excluded from any one or more of the claims. Becausesuch embodiments are deemed to be known to one of ordinary skill in the art, they may be excluded even if the exclusion is not set forth explicitly herein. Any particular embodiment described herein can be excluded from any claim, for any reason, whether or not related to the existence of prior art.

[0248] Those skilled in the art will recognize or be able to ascertain using no more than routine experimentation many equivalents to the specific embodiments described herein. The scope of the present embodiments described herein is not intended to be limited to the above Description, but rather is as set forth in the appended claims. Those of ordinary skill in the art will appreciate that various changes and modifications to this description may be made without departing from the spirit or scope of the present disclosure, as defined in the following claims.

Claims

CLAIMSWhat is claimed is:

1. An engineered midnolin protein comprising: an aHelixC domain, a ubiquitin-like domain, and a Catch domain; wherein the Catch domain comprises Catch 1 subdomain and Catch 2 subdomain, wherein Catch 1 and Catch 2 subdomains are joined by a linker; and wherein the aHelixC domain, the ubiquitin-like domain, the Catch domain, Catch 1 subdomain, Catch 2 subdomain, and / or the linker is / are modified.

2. The engineered midnolin protein of claim 1, wherein the aHelixC domain further comprises a nuclear localization signal (NLS).

3. The engineered midnolin protein of claim 1 or 2, wherein the NLS is modified.

4. The engineered midnolin protein of any one of claims 1-3, further comprising a localization signal.

5. The engineered midnolin protein of any one of claims 1-4, wherein the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids longer than a wildtype linker.

6. The engineered midnolin protein of any one of claims 1-4, wherein the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids shorter than a wildtype linker.

7. The engineered midnolin protein of any one of claims 1-6, wherein the engineered midnolin protein improves degradation of a target protein.

8. The engineered midnolin protein of any one of claims 1-7, wherein the ubiquitin-like domain comprises an amino acid sequence comprising an amino acid substitution relative to a wildtype ubiquitin-like domain.

9. The engineered midnolin protein of claim 8, wherein the amino acid substitution is L80E.

10. The engineered midnolin protein of claim 8, wherein the amino acid substitution is G83E.

11. The engineered midnolin protein of any one of claims 1-10, wherein the ubiquitin-like domain is deleted from the engineered midnolin protein.

12. The engineered midnolin protein of any one of claims 1-11, wherein the Catch 1 subdomain is deleted from the engineered midnolin protein.

13. The engineered midnolin protein of any one of claims 1-12, wherein the Catch 2 subdomain is deleted from the engineered midnolin protein.

14. The engineered midnolin protein of any one of claims 1-11, wherein the Catch 1 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 1 subdomain.

15. The engineered midnolin protein of claim 14, wherein the amino acid substitution is V129D, L133D, S134, or L139D.

16. The engineered midnolin protein of any one of claims 1-11, wherein the Catch 2 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 2 subdomain.

17. The engineered midnolin protein of claim 16, wherein the amino acid substitution is R301S, I306D, L310D, or L313D.

18. The engineered midnolin protein of any one of claims 1-17, wherein the aHelixC domain is deleted from the engineered midnolin protein.

19. The engineered midnolin protein of any one of claims 2-18, wherein the NLS is deleted from the engineered midnolin protein.

20. A method of promoting degradation of a target protein, the method comprising contacting the target protein with the engineered midnolin protein according to any one of claims 1-4.

21. The method of claim 20, wherein the target protein comprises a P strand degron.

22. The method of claim 20 or 21, wherein the Catch domain binds the P strand degron of the target protein.

23. The method of any one of claims 20-22, wherein the target protein is degraded by the proteasome after the Catch domain binds the P strand degron of the target protein.

24. The method of any one of claims 20-23, wherein degradation of the target protein is ubiquitination-independent.

25. The method of any one of claims 20-24, wherein the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids longer than a wildtype linker.

26. The method of any one of claims 20-24, wherein the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids shorter than a wildtype linker.

27. The method of any one of claims 20-26, wherein modification to the linker affects binding specificity to the target protein.

28. A method of promoting stability of a target protein, the method comprising expressing in a cell a gene encoding the engineered midnolin protein according to any one of claims 1-19.

29. The method of claim 28, wherein the ubiquitin-like domain of the engineered midnolin protein comprises an amino acid sequence comprising an amino acid substitution relative to a wild type ubiquitin-like domain.

30. The method of claim 29, wherein the amino acid substitution is L80E.

31. The method of claim 29, wherein the amino acid substitution is G83E.

32. The method of any one of claims 28-31, wherein the ubiquitin-like domain is deleted from the gene encoding the engineered midnolin protein.

33. The method of any one of claims 28-32, wherein the Catch 1 subdomain is deleted from the gene encoding the engineered midnolin protein.

34. The method of any one of claims 28-33, wherein the Catch 2 subdomain is deleted from the gene encoding the engineered midnolin protein.

35. The method of any one of claims 28-32, wherein the Catch 1 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 1 subdomain.

36. The method of claim 35, wherein the amino acid substitution is V129D, L133D, S134, or L139D.

37. The method of any one of claims 28-32, wherein the Catch 2 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 2 subdomain.

38. The method of claim 37, wherein the amino acid substitution is R301S, I306D, L310D, or L313D.

39. The method of any one of claims 28-38, wherein the aHelixC domain is deleted from the gene encoding the engineered midnolin protein.

40. The method of any one of claims 28-39, wherein the NLS is deleted from the gene encoding the engineered midnolin protein.

41. The method of any one of claims 28-40, wherein the gene is expressed heterologously in the cell.

42. The method of any one of claims 28-41, wherein the gene is expressed on a plasmid.

43. The method of any one of claims 28-42, wherein the gene is integrated into the genome of the cell.

44. A method of promoting stability of a target protein, the method comprising contacting a cell with a small molecule that inhibits expression of midnolin in the cell.

45. A method of promoting stability of a target protein, the method comprising deleting a gene encoding midnolin from the genome of a cell.

46. The method of claim 45, wherein the step of deleting is accomplished using a genetic engineering technique.

47. The method of claim 45 or 46, wherein the genetic engineering technique comprises a CRISPR / Cas9 system.

48. An engineered cell comprising the engineered midnolin protein according to any one of claims 1-4.

49. The engineered cell of claim 48, wherein the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids longer than a wildtype linker.

50. The engineered cell of claim 49, wherein the linker is at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 50, or at least 100 amino acids shorter than a wildtype linker.

51. The engineered cell of claim 49 or 50, wherein the engineered midnolin protein improves degradation of a target protein.

52. The engineered cell of claim 48, wherein the ubiquitin-like domain comprises an amino acid sequence comprising an amino acid substitution relative to a wildtype ubiquitin-like domain.

53. The engineered cell of claim 52, wherein the amino acid substitution is L80E.

54. The engineered cell of claim 52, wherein the amino acid substitution is G83E.

55. The engineered cell of any one of claims 48 or 52-54, wherein the ubiquitin-like domain is deleted from the engineered midnolin protein.

56. The engineered cell of any one of claims 48 or 52-55, wherein the Catch 1 subdomain is deleted from the engineered midnolin protein.

57. The engineered cell of any one of claims 48 or 52-56, wherein the Catch 2 subdomain is deleted from the engineered midnolin protein.

58. The engineered cell of any one of claims 48-55, wherein the Catch 1 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 1 subdomain.

59. The engineered cell of claim 58, wherein the amino acid substitution is V129D, L133D, S134, or L139D.

60. The engineered cell of any one of claims 48-55, wherein the Catch 2 subdomain comprises an amino acid sequence comprising an amino acid substitution relative to a wild type Catch 2 subdomain.

61. The engineered cell of claim 60, wherein the amino acid substitution is R301S, I306D, L310D, or L313D.

62. The engineered cell of any one of claims 48 or 52-61, wherein the aHelixC domain is deleted from the engineered midnolin protein.

63. The engineered cell of any one of claims 48 or 52-62, wherein the NLS is deleted from the engineered midnolin protein.

64. The method of any one of claims 20-47, wherein the target protein is relevant to a disease.

65. The method of claim 64, wherein the disease is cancer, an immunological disease, or a brain disease.

66. The method of claim 64 or 65, wherein the target protein is a transcription factor.

67. The method of any one of claims 64-66, wherein the target protein is IRF4 or an immediate early gene.

68. The method of claim 67, wherein the immediate early gene is c-Fos, FosB, ERG1, or NR4A1.

69. The engineered cell of any one of claims 48-63, wherein the target protein is relevant to a disease.

70. The engineered cell of claim 69, wherein the disease is cancer, an immunological disease, or a brain disease.

71. The engineered cell of claim 69 or 70, wherein the target protein is a transcription factor.

72. The engineered cell of any one of claims 69-71, wherein the target protein is IRF4 or an immediate early gene.

73. The engineered cell of claim 72, wherein the immediate early gene is c-Fos, FosB, ERG1, or NR4A1.

74. A method of promoting degradation of a target protein, the method comprising increasing in a cell expression of a gene encoding midnolin.

Images