Abstract
The Pro/N-degron recognizing C-terminal to LisH (CTLH) complex is an E3 ligase of emerging interest in the developmental biology field and for targeted protein degradation (TPD) modalities. The human CTLH complex forms distinct supramolecular ring-shaped structures dependent on the multimerization of WDR26 or muskelin β-propeller proteins. Here, we find that, in HeLa cells, CTLH complex E3 ligase activity is dictated by an interplay between WDR26 and muskelin in tandem with muskelin autoregulation. Proteomic experiments revealed that complex-associated muskelin protein turnover is a major ubiquitin-mediated degradation event dependent on the CTLH complex in unstimulated HeLa cells. We observed that muskelin and WDR26 binding to the scaffold of the complex is interchangeable, indicative of the formation of separate WDR26 and muskelin complexes, which correlated with distinct proteomes in WDR26 and muskelin knockout cells. We found that mTOR inhibition-induced degradation of Pro/N-degron containing protein HMGCS1 is distinctly regulated by a muskelin-specific CTLH complex. Finally, we found that mTOR inhibition also activated muskelin degradation, likely as an autoregulatory feedback mechanism to regulate CTLH complex activity. Thus, rather than swapping substrate receptors, the CTLH E3 ligase complex controls substrate selectivity through the differential association of its β-propeller oligomeric subunits WDR26 and muskelin.
Similar content being viewed by others
Introduction
Ubiquitination involves the covalent attachment of one or more ubiquitin (Ub) proteins to mainly lysine residues of substrate proteins. Depending on the type of ubiquitin linkage and topology of the poly-Ub chain, various effects on the targeted substrate can occur such as proteasomal degradation, changes in interaction partners, or alterations of subcellular localization1,2. These regulatory changes are necessary for the proper development of multicellular organisms. Hence, dysfunctional ubiquitination cascades can lead to aberrant cellular differentiation and developmental abnormalities, among other pathologies3,4.
The transfer of ubiquitin to a substrate occurs through the coordinated action of E1 Ub-activating enzymes, E2 Ub-conjugating enzymes, and E3 Ub-protein ligases, with the latter conferring substrate specificity5. Most E3 ligases contain RING (Really Interesting New Gene) domains that facilitate the interaction between the E2~Ub conjugate and a specific substrate. Over 600 RING E3 ligases are predicted to be expressed in human cells to regulate various molecular pathways. Many of these exist in multi-subunit complexes that include interchangeable substrate receptors, adaptors, scaffolds, and RING catalytic modules - thereby increasing the theoretical number of proteins that could be regulated through ubiquitination6,7.
The C-terminal to LisH (CTLH) complex – also known as the Glucose-Induced Degradation Deficient (GID) complex – is a multi-subunit RING E3 ligase that has been implicated in a growing list of biological processes such as cell proliferation, haemopoietic differentiation, glycolysis, autophagy, and maternal to zygotic transition8,9,10,11,12,13,14,15,16,17. Recent studies have elucidated the composition and general structure of the CTLH complex in vitro using electron cryomicroscopy and in vitro binding assays18,19,20,21. Briefly, E3 ligase activity is mediated by the RING catalytic module formed between a MAEA/RMND5A heterodimer where each monomer binds to a separate GID8 (TWA1) protein. The general scaffold is comprised of GID8, RanBP9 (RanBPM), and ARMC8 isoforms α and β. The Pro/N-degron recognizing substrate receptor GID4 then binds to ARMC8α, resulting in a horseshoe-shaped structure with the RMND5A-MAEA heterodimer linking two separate CTLH scaffold + GID4 moieties, with one RanBP9 protein at each free end. Addition of WDR26 promotes the formation of a 1.5 MDa supramolecular donut-shaped complex by forming 2 homodimers and connecting the 2 horseshoe-shaped structures through RanBP9 on opposite ends of the complex. This supramolecular complex can accommodate substrates in its internal cavity, leading to their ubiquitination19,20.
Interestingly, certain species, including mammals, have another CTLH complex member called muskelin8,22. Similar to WDR26, this protein contains a β-propeller-like kelch domain at its C-terminus. In vitro, muskelin was also shown to mediate the oligomerization of the CTLH complex through a homotetramer instead of a homodimer, potentially creating a structurally distinct central pocket for oligomeric substrates compared to WDR2618,20.
Although muskelin and WDR26 can both mediate the formation of CTLH complex oligomerization in vitro, only the loss of WDR26, and not muskelin, has a profound effect on CTLH complex formation in cells20,23. An additional distinction is that muskelin protein levels, but not WDR26, are regulated by ubiquitination in a MAEA/RMND5A-dependent manner22,24. Despite differences between these two structurally similar subunits, it remains unclear whether they translate into differences in CTLH complex-mediated regulation. In this study, using proteomics and complex formation assays in human cells, we found that WDR26 and muskelin can mediate the formation of two distinct supramolecular CTLH complexes that have distinct functions. We also discovered that inhibition of mTOR signalling regulates muskelin-specific CTLH complex activity in human cells and leads to muskelin autoregulation.
Results
Muskelin autoregulation is a dominant activity of the CTLH complex
To examine possible ubiquitination targets of the CTLH complex, we conducted global label-free quantitative proteomics in RMND5A (n = 5) or MAEA (n = 4) HeLa knockout (KO) cells. We reasoned that degradation targets of the CTLH complex would be increased in RMND5A or MAEA KO cells relative to control cells. Principal component analysis and hierarchical clustering revealed that knockouts clustered distinctly from control samples (Supplemental Fig. 1A). Out of 4769 quantified proteins, 202 were significantly increased in RMND5A KO and 181 proteins were increased in MAEA KO samples compared to controls (adjusted p < 0.05, KO/WT fold change > 1.5) (Fig. 1A, B, Supplemental data 1). Overall, there was substantial agreement between changed proteins in both knockouts, with 108 significantly changed proteins identified in common (Fig. 1C).
A Heatmap of differentially abundant proteins (adjusted p-value < 0.05, fold change > 1.5, statistics derived from limma, see methods) in whole-cell extracts of RMND5A (n = 5) or MAEA KO (n = 4) HeLa cells. Each row represents one protein and samples are represented in columns. Protein abundance is shown as z-scores quantified across rows. Hierarchical clustering of samples and proteins is shown. B Volcano plot of all proteins with log2 fold change shown on the x-axis and -log10 adjusted p values shown on the y axis. RMND5A KO vs respective controls on the left, MAEA KO vs respective controls on the right. Muskelin (MKLN1) is indicated in red. C Overlap of deferentially abundant proteins in RMND5A or MAEA KO cells. D Median centered intensity of MKLN1 in the proteome samples. Boxplot midline indicates median values, bounds of the box indicate 25th and 75th percentiles, and maxima and minima indicate the largest point above or below 1.5 times the interquartile range. E Median centered intensities for specific proteins containing Pro/N-degrons known to be recognized by GID4. Evolutionary conservation of first ten amino acids of N-termini are indicated. Boxplot midline indicates median values, bounds of the box indicate 25th and 75th percentiles, and maxima and minima indicate the largest point above or below 1.5 times the interquartile range. F Comparison of the diGLY-enriched proteome vs total proteome in RMND5A KO HeLa cells. MG132-treated control or RMND5A KO HeLa cells were lysed, digested, and peptides containing the diGLY motif originating from ubiquitin were enriched and analyzed by LC-MS/MS. Shown are diGLY peptides plotted with their log2 fold change on the y-axis against the log2 fold change of their matched total protein levels. diGLY peptides with log2FC> 1 are labelled with protein name and modified Lysine residue position. Muskelin K260 is indicated in red. G Restoration of muskelin levels in RMND5A KO cells by exogenous WT RMND5A or I338A/L399A E2 binding deficient mutant. RMND5A KO HeLa cells were transfected with carrier plasmid (mock), HA-WT-RMND5A, or HA-I338A/L339A-RMND5A. Whole cell lysates were prepared 24 h later and analyzed by western blot with the indicated antibodies. Boxplot midline indicates median values of muskelin intensity (normalized to tubulin, relative to mock for each replicate), bounds of the box indicate 25th and 75th percentiles, and maxima and minima indicate the largest point above or below 1.5 times the interquartile range (n = 4). Welch two sample t-test, t = 11.611, df ≈ 6.
We previously discovered that protein levels of CTLH complex subunit muskelin (MKLN1) are increased in RMND5A and MAEA KO HeLa and HEK293 cell lines22. Interestingly, muskelin was one of the most highly increased proteins of the quantifiable proteomes of both RMND5A and MAEA KO cells (Fig. 1B, D). Additionally, alanyl aminopeptidase (ANPEP), BAR/IMD domain containing adaptor protein 2 (BAIAP2), and LIMA1 (LIM domain and actin binding 1), all previously identified as interactors of CTLH complex subunits23,25, were significantly increased in one or both knockout cell lines (Supplementary Fig. 1B). We also noted that some proteins increased in RMND5A and/or MAEA knockouts contain an N-terminal sequence that matches the Pro/N-degron recognized by CTLH complex substrate receptor GID4 (Fig. 1E)26,27,28. This included mevalonate pathway enzyme HMG-CoA synthase 1 (HMGCS1), which was recently identified as a GID4 substrate29, and two other proteins not previously linked to GID4 or the CTLH complex: CSNK2A2, which encodes the α’ subunit of casein kinase II (CK2), and cytoskeleton-associated protein 4 (CKAP4), which is associated with cell migration and metastasis30,31.
A comparison of the RMND5A-dependent proteome and ubiquitinome showed few changes consistent with degradative ubiquitination (Fig. 1F, Supplemental data 2, Supplemental Fig. 1C, D), which is in line with our previous observations for CTLH complex scaffolding subunit RanBP910. In fact, muskelin was the top increased protein with a more than 2-fold decreased ubiquitin site (Fig. 1F). This occurred on lysine 260 of muskelin, a conserved and – based on the Alphafold predicted structure – surface exposed residue located in the connecting region between the CTLH motif and the kelch repeats (Supplementary Fig. 1E, F). This suggests that K260 is at least one ubiquitination site of muskelin that is RMDN5A-dependent.
Next, we used RMND5A mutants to establish that the muskelin regulation is mediated by the E3 ligase activity of the complex. We previously demonstrated that exogenous expression of RMND5A cDNA restores muskelin levels back to control levels in RMND5A KO cells22. Using the same approach, we compared the effect of transfecting wildtype RMND5A with a I338A/L339A mutant previously shown to be deficient in E2 binding in vitro20. While exogenous WT RMND5A reduced muskelin protein levels, the I338A/L339A mutant had no effect, thus establishing that the regulation of muskelin is due to the ubiquitin ligase activity of the CTLH complex (Fig. 1G). Together with our previous studies on global CTLH functions10,23,29, these data suggest that most of the CTLH-dependent ubiquitination is non-degradative in unstimulated HeLa cells, but its own subunit muskelin is one of its primary degradative ubiquitination targets.
Muskelin and WDR26 both engage in similar interactions with RanBP9
Intrigued by why the CTLH complex E3 activity would be mainly dedicated to regulating muskelin levels in unstimulated conditions, but not those of WDR26, we sought to compare functions and interactions of WDR26 and muskelin. Studies using in vitro reconstituted systems have determined that, in the presence of muskelin or WDR26, both of which bind RanBP9, a supramolecular CTLH complex is formed18,20,21. To establish the requirements for this in a cellular model, we evaluated the RanBP9, muskelin, and WDR26 domains required to mediate interactions with the CTLH complex. WDR26 and muskelin have a similar domain architecture, namely α-helical LisH, CTLH, and CRA motifs and a C-terminal β-propeller domain – through WD40 repeats in WDR26 or kelch repeats in muskelin (Supplementary Fig. 2A). Unique to muskelin, however, is a discoidin domain near the N-terminus. RanBP9 also has LisH, CTLH, and CRA motifs. We performed co-immunoprecipitation experiments with transfected RanBP9, muskelin, or WDR26 domain deletion constructs as baits (Supplementary Fig. 2B). Though not required for MAEA association, the CTLH motif of RanBP9 was necessary for its interaction with both muskelin and WDR26 (Fig. 2A). Deletion of the LisH motif of RanBP9, meanwhile, had no effect on muskelin and WDR26 association (Fig. 2A).
A The indicated HA-RanBP9 domain deletion constructs (see supplemental Fig. 2B) were transfected in shRanBP9 HCT116 cells. After 24 h, cells were lysed in whole-cell extract buffer and immunoprecipitated with an anti-HA antibody. Elutions were run on a western blot and analyzed for the presence of the indicated CTLH complex subunits. Experiments were performed in triplicate (n = 3). B The indicated FLAG-muskelin domain deletion constructs (containing a streptavidin binding peptide, see supplementary Fig. 2B) were transfected in muskelin KO HeLa cells. After 24 h, cells were lysed in whole-cell extract buffer and muskelin complexes were isolated with streptavidin beads. Elutions were run on a western blot and analyzed for the presence of the indicated CTLH complex subunits. Experiments were performed in triplicate (n = 3). C The indicated FLAG-WDR26 domain deletion constructs (containing a streptavidin binding peptide, see supplementary Fig. 2B) were transfected in WDR26 KO HeLa cells. After 24 h, cells were lysed in whole-cell extract buffer and WDR26 complexes were isolated with streptavidin beads. Elutions were run on a western blot and analyzed for the presence of the indicated CTLH complex subunits. Experiments were performed in triplicate (n = 3). D Immunoprecipitation of endogenous RanBP9. WT or indicated KO HeLa cells were lysed in whole-cell lysis buffer and subjected to immunoprecipitation with a RanBP9 antibody that was crosslinked to beads. Elutions were run on a Western blot and analyzed for the presence of the indicated CTLH complex subunits. Experiments were performed in triplicate (n = 3).
For muskelin, deletion of CTLH and CRA motifs prevented binding to RanBP9 whereas deletion of the LisH, discoidin or kelch repeats did not (Fig. 2B). Similarly, the CTLH and CRA motifs of WDR26 were required for RanBP9 binding but the LisH motif and WD40 domain were not (Fig. 2C). Next, we compared co-immunoprecipitations of endogenous RanBP9 across WT and KO cell lines of CTLH complex subunits. As expected, RanBP9 was still able to associate with CTLH complex subunits in either muskelin or WDR26 KO cells, and both muskelin and WDR26 still interacted with RanBP9 in any KO cell line other than their own (Fig. 2D). These results are supported by similar co-IP experiments conducted for the S. cerevisiae complex as well as previously determined cryo-EM structures and in vitro binding assays for the human complex18,20,21,32. Overall, it is clear muskelin and WDR26 associate similarly with the CTLH complex via interactions between their CTLH-CRA motifs and the CTLH-CRA motifs of RanBP9.
Muskelin can substitute for WDR26 as the β-propeller link for supramolecular complex formation
It has yet to be determined in a cellular model whether WDR26 and muskelin co-exist as part of a supramolecular complex. Since the supramolecular complex contains two regions where WDR26 dimers or muskelin tetramers could bind to RanBP9, a “hybrid” supramolecular complex could in theory be formed with both moieties present together. To address this possibility, transiently-transfected FLAG-muskelin was immunoprecipitated from either DMSO or MG132-treated WT HeLa cells and co-immunoprecipitated WDR26 and RanBP9 levels were analyzed. Because of the muskelin autoregulation reported above, the co-immunoprecipitation was performed with lysates derived from MG132-treated cells to address the possibility that hybrid complexes are being degraded through the proteasome in nascent conditions. Although RanBP9 readily co-immunoprecipitated with overexpressed muskelin in both conditions, no detectable WDR26 was co-immunoprecipitated in either condition (Fig. 3A, Supplementary Fig. 3A). Similarly, a negligible amount of endogenous muskelin was co-immunoprecipitated with exogenous WDR26 in comparison to endogenous RanBP9 (Fig. 3B, Supplementary Fig. 3A), suggesting that the cell contains largely WDR26- or muskelin-specific supramolecular complexes.
A FLAG-muskelin or pcDNA3.1 empty vector was transfected in DMSO- or MG132-treated muskelin KO HeLa cells. After 24 hours, cells were lysed in whole cell extract buffer and FLAG antibody was used for the co-immunoprecipitation assay. Elutions were run on a western blot and analyzed using the indicated antibodies. Experiments were performed in triplicate (n = 3). B FLAG-WDR26 or pcDNA3.1 empty vector was transfected in WDR26 KO HeLa cells. After 24 h, cells were lysed in whole cell extract buffer and FLAG or IgG antibody was used for the co-immunoprecipitation assay. Elutions were run on a western blot and analyzed using the indicated antibodies. Experiments were performed in triplicate (n = 3). C Whole cell extracts were prepared from WDR26 KO and WDR26/muskelin double knockout (DKO) HeLa cells treated with DMSO or MG132 and separated by a 5–40% sucrose gradient. The resulting fractions were loaded on an SDS-PAGE gel, prepared for western blotting and analyzed with the indicated antibodies. Experiments were performed in triplicate (n = 3). D Schematic summarizing the results that a muskelin exclusive complex exists but a hybrid complex does not. Created in BioRender. Maitland, M. (2025) https://BioRender.com/b17n399.
WDR26 is required in vivo for nuclear formation of the high order molecular complexes; however, we previously found that in WDR26 KO cells, proteasome inhibition by MG132 restored higher order nuclear complex formation23. This indicated that if the CTLH subunits are not degraded in WDR26 KO cells, the supramolecular structure can still form even in the absence of WDR26. To test if muskelin is compensating for the lack of WDR26, we created a double muskelin/WDR26 KO cell line and performed sucrose gradients in the presence or absence of MG132. Sucrose gradients are comparable between WDR26 KO and muskelin/WDR26 KO cells in DMSO vehicle conditions, as has been observed previously20; however, we found that while proteasome inhibition restored CTLH complex formation in WDR26 KO cells, it could not if both muskelin and WDR26 were absent (Fig. 3C). We also tested whether muskelin or WDR26 were able to displace each other using sucrose gradients from WT cells transfected with either muskelin or WDR26. In untransfected cells, endogenous muskelin and WDR26 migrated mostly to fractions corresponding to the supramolecular complex (fractions 5 and 6); however, upon transfection of exogenous muskelin, a fraction of endogenous WDR26 migrated to lower molecular mass fraction 3 and reciprocally, WDR26 transfection triggered relocalization of muskelin into that same fraction (Supplementary Fig. 3B). In addition, co-immunoprecipitation experiments showed overexpressed muskelin or WDR26 competed with one another for interaction with RanBP9 (Supplementary Fig. 3C). To summarize, muskelin and WDR26 can both serve as the β-propeller link for the supramolecular CTLH complex, but a hybrid complex is not observed (Fig. 3D). In addition, their association can be modulated by varying protein levels of either complex member (Supplementary Fig. 3D).
Muskelin instability requires both binding to RanBP9 and oligomerization
We next determined the requirements for muskelin to be regulated by the CTLH complex. We conducted cycloheximide (CHX) chase assays in WT and RMND5A KO cells transfected with 3 forms of muskelin. First, wild-type muskelin that can interact with RanBP9 and form a tetramer (WT); second, a mutant muskelin lacking the CRA motif that cannot interact with RanBP9 but can form a tetramer (∆CRA); and finally, a mutant lacking the LisH motif that can interact with RanBP9 but cannot form a tetramer (∆LisH)18,33. WT-muskelin was more unstable compared to either muskelin deletion mutant (Fig. 4), demonstrating that muskelin that can form a supramolecular complex is degraded. This data supports the hypothesis that muskelin-containing supramolecular CTLH complexes are regulating complex-bound muskelin protein levels in an autoregulatory fashion.
Western blot quantification of muskelin protein levels in a cycloheximide (CHX) pulse-chase in muskelin KO HeLa cells. Representative western blots are shown (left) and quantification from n = 3 independent experiments (right). A Two-Way ANOVA revealed a significant interaction between 50 µg/µL CHX treatment time and muskelin constructs (F(2,28) = 9.318, p = 0.000791). Post-hoc comparisons using Tukey’s HSD test indicated that ΔCRA and ΔLisH significantly increased muskelin protein level at 18 and/or 24 h CHX compared to WT (adjusted p-value indicated on plot). Linear models were fit to the data using least squares method with each solid line indicating the best fit to the data and 95% confidence intervals indicated by the surrounding shaded area.
Distinct proteome regulation by WDR26 and muskelin
Considering the observed interplay between WDR26 and muskelin in binding to the CTLH complex, we thought that the differential association of these two subunits could result in CTLH complexes with distinct functions. Consequently, we performed a comprehensive proteomic analysis of their respective knockout cell lines to investigate potential functional differences. A principal component analysis of muskelin (n = 5) and WDR26 (n = 6) knockout proteomes revealed two discrete clusters delineated by their knockout status, wherein KO and control samples were segregated along the first principal component (PC1), accounting for approximately 70% of the observed variance in each case (Supplementary Fig. 4A). Overall, MKLN1 KO produced a less pronounced effect on global proteome regulation than WDR26 KO, since there were 132 significantly changed proteins out of 4218 in WDR26 KO, compared to only 39 changed proteins after loss of MKLN1 (adjusted p < 0.05, fold change cut-off of 1.5, Fig. 5A, B, Supplementary Data 3). Since muskelin protein levels are decreased in WDR26 KO cells23, it is not surprising that there were 16 significantly changed proteins in common between both proteomes, corresponding to 41% of all changed proteins in MKLN1 KO cells (Fig. 5C). Shared changed proteins included LIMA1 and ANPEP which both increased (Fig. 5D), agreeing with their changes after either RMND5A or MAEA knockout (Supplementary Fig. 1B). Additionally, 6 proteins were changed in all four proteomes, 20 were changed in MAEA, RMND5A, and WDR26 KO proteomes, and 2 were changed in MKLN1, MAEA, and RMND5A KO proteomes (Supplementary Fig. 4B). Of the increased proteins in any KO cell line proteome, 13 contained a Pro/N-degron sequence, the recognition element for the CTLH complex substrate receptor GID4 (Supplementary Fig. 4C).
A Heatmap of differentially abundant proteins (adjusted p-value < 0.05, fold change > 1.5, statistics derived from limma, see methods) in whole-cell extracts of MKLN1 (n = 5) or WDR26 KO (n = 6) HeLa cells. Each row represents one protein and samples are represented in columns. Protein abundance is shown as z-scores quantified across rows. Hierarchical clustering of samples and proteins is shown. B Volcano plot of all proteins with log2 fold change shown on the x-axis and -log10 adjusted p values shown on the y axis. MKLN1 KO vs respective controls on the left, WDR26 KO vs respective controls on the right. HMGCS1 is indicated in red. C Overlap of proteins that had significant differential abundance in MKLN1 or WDR26 KO cells. D Median centered intensity of proteins increased in both WDR26 and muskelin KO proteome samples, and in one (LIMA1) or both (ANPEP) of RMND5A and MAEA KO proteomes (see Supplemental Fig. 1B). Boxplot midline indicates median values, bounds of the box indicate 25th and 75th percentiles, and maxima and minima indicate the largest point above or below 1.5 times the interquartile range. E Median centered intensities for Pro/N-degron containing proteins, increased in RMND5A KO and/or MAEA KO (Fig. 1) proteomes that are increased in either MKLN1 (HMGCS1) or WDR26 (CKAP4) KO proteome samples. Boxplot midline indicates median values, bounds of the box indicate 25th and 75th percentiles, and maxima and minima indicate the largest point above or below 1.5 times the interquartile range. F Top 10 GO terms functional annotation of proteins significantly increased in MKLN1 KO or WDR26 KO proteome samples. Gene ratio represents the fraction of genes in each GO category that were identified. P-values were adjusted using the Benjamini-Hochberg method.
In contrast to these similarities, many proteins showed specific regulation by either MKLN1 or WDR26. Two noteworthy examples are Pro/N-degron containing proteins CKAP4 and HMGCS1, which we previously found to be increased in RMND5A or MAEA knockouts (Fig. 5E, Fig. 1E). CKAP4 was not changed after MKLN1 knockout but increased in cells lacking WDR26 (Fig. 5E). HMGCS1 levels, on the other hand, increased after MKLN1 knockout but were not changed in response to WDR26 knockout (Fig. 5B, E).
To further investigate the specific functional differences in proteome regulation by muskelin and WDR26, we performed GO terms analysis on the list of increased proteins in each knockout. The top enriched GO terms associated with WDR26-increased proteins were associated with embryonic morphogenesis (GO:0048598). In contrast, the top ten significantly enriched GO terms (ranked by gene ratio) associated with MKLN1 increased proteins were largely associated with cell migration (GO:0016477) and cell motility (GO:0048870) (Fig. 5F). Taken together, these findings are indicative of the distinct proteome regulation functions of MKLN1 and WDR26 acting in concert with the CTLH complex, mediated through mutually exclusive binding to the CTLH complex via RanBP9.
The muskelin-containing supramolecular CTLH complex regulates HMGCS1 protein levels in an mTOR- and GID4-dependent manner
The proteome data determined that HMGCS1 was significantly increased in RMND5A KO cells and muskelin KO cells, but not WDR26 KO cells (Fig. 1D, Fig. 5B, E). HMGCS1 contains a Pro/N-degron and we previously demonstrated it is regulated by CTLH complex substrate receptor GID429. It therefore represents a potential CTLH target that may be distinctly regulated by muskelin and WDR26.
HMGCS1 was previously shown to be ubiquitinated and undergo rapid proteasomal degradation upon treatment with the mTOR inhibitor Torin-1, though an E3 ligase responsible for this was not identified34. By conducting a similar experiment with Torin-1 treatment in HeLa cells, we also observed rapid degradation of HMGCS1 (Fig. 6A). Remarkably, this degradation did not occur in RMND5A KO HeLa cells (Fig. 6A). Degradation was also impeded if HeLa cells were pre-treated with the GID4 chemical probe PFI-7 that occludes the GID4 binding pocket, but not with the negative control compound PFI-7N that cannot bind GID4 (Fig. 6B)29. As HMGCS1 was determined to be uniquely dependent on muskelin in the proteome datasets derived from unstimulated cells, we also compared the effect on HMGCS1 by Torin-1 treatment in WDR26 or muskelin KO HeLa cells. Expectedly, while Torin-1-induced HMGCS1 degradation in WDR26 KO cells was indistinguishable from WT cells, no degradation occurred in muskelin KO cells (Fig. 6A). Together, this implicates the muskelin exclusive CTLH complex as the E3 responsible, via GID4 Pro/N-degron binding, for Torin-1 induced HMGCS1 ubiquitination and degradation.
A Whole-cell lysates were prepared from WT (blue, circle), MKLN1 KO (burgundy, square), WDR26 KO (gold, triangle), or RMND5A KO (green, diamond) HeLa cells treated with 250 nM Torin-1 for the indicated time points and analyzed by western blot with the indicated antibodies. Quantification from n = 3 independent experiments on the right. A Two-Way ANOVA revealed a significant interaction between treatment time and HMGCS1 levels (F(6,24) = 3.902, p = 0.00734). Post-hoc comparisons using Tukey’s HSD test indicated that muskelin and RMND5A KO cells had significantly increased HMGCS1 protein level at 4 h compared to WT (adjusted p-values indicated on plot). Linear models were fit to the data using least squares method with each solid line indicating the best fit to the data and 95% confidence intervals indicated by the surrounding shaded area. B Whole-cell lysates were prepared from WT HeLa cells pretreated for 24 h with 10 μM negative control compound PFI-7N (blue, circle) or PFI-7 (red, triangle). Cells were then treated with 50 μg/mL cycloheximide (CHX) or CHX and 250 nM Torin-1 for the indicated time points and analyzed by western blot with the indicated antibodies. Quantification from n = 3 independent experiments on the right. A Two-Way ANOVA revealed a significant interaction between treatment time and HMGCS1 levels (F(2,12) = 5.401, p = 0.021247). Post-hoc comparisons using Tukey’s HSD test indicated that PFI-7 treatment had significantly increased HMGCS1 protein level at 2 h compared to PFI-7N (adjusted p-values indicated on plot). Linear models were fit to the data using least squares method with each solid line indicating the best fit to the data and 95% confidence intervals indicated by the surrounding shaded area. C The same lysates from (A) were run on western blot to assess muskelin levels. Quantification from n = 3 independent experiments on the right. A Two-Way ANOVA revealed a significant interaction between treatment time and HMGCS1 levels (F(4,18) = 6.302, p = 0.00236). Post-hoc comparisons using Tukey’s HSD test indicated that RMND5A KO cells had significantly increased HMGCS1 protein level at 2 and 4 h compared to WDR26 KO (adjusted p-values indicated on plot). Linear models were fit to the data using least squares method with each solid line indicating the best fit to the data and 95% confidence intervals indicated by the surrounding shaded area. D Schematic depicting the results presented here. The muskelin-CTLH complex is activated upon Torin-1 treatment and degrades HMGCS1 via GID4 binding. A negative feedback loop is present wherein muskelin is also degraded. Created in BioRender. Maitland, M. (2025) https://BioRender.com/b17n399.
Curiously, Torin-1 treatment resulted in decreased muskelin levels in WT and WDR26 KO cells, but not in RMND5A KO cells (Fig. 6C). These results were replicated in HEK293 cells, demonstrating that mTOR-mediated muskelin autoregulation is a general mechanism (Supplementary Fig. 5A). To assess if Torin-1 changes the formation of CTLH-muskelin complexes, we immunoprecipitated RanBP9 in cells treated with MG132 and with or without Torin-1. Surprisingly, muskelin and GID4 association with RanBP9 was unaffected by Torin-1 treatment (Supplementary Fig. 5B), indicating that increased HMGCS1 degradation upon mTOR inhibition is not through increased formation of CTLH-muskelin complexes. However, we could not detect HMGCS1 in the immunoprecipitates, likely due to the transient nature of its association with the complex. Altogether, these findings suggest that Torin-1 activates the muskelin-CTLH complex to degrade HMGCS1 using GID4 as the substrate receptor, but only briefly as muskelin autoregulation soon follows (Fig. 6D).
Discussion
The CTLH complex is a multi-subunit E3 ligase with recently reported chemical ligands to the substrate receptor GID4, generating interest for potential use of this E3 in targeted protein degradation (TPD) modalities29,35,36,37. Prior to its implementation in TPD, however, the function and regulation of the CTLH complex must be better understood. Here, we identify a functional dichotomy between the β-propeller subunits muskelin and WDR26 that translates into the formation of distinct CTLH complexes and correlates with the targeting of different substrates. Additionally, we provide the first evidence of a pathway that regulates the human CTLH complex activity. We demonstrate that the CTLH complex comprising muskelin, but not WDR26, rapidly degrades HMGCS1 upon mTOR inhibition, highlighting a biological regulation/pathway that affects the activity of a specific CTLH sub-complex. Muskelin degradation is also triggered, providing a mechanism for this complex to autoregulate its activity.
A recent report focused on GID4 in the presence or absence of its selective antagonist (PFI-7) identified HMGCS1 as the first human GID4 target29. Here we show that HMGCS1 - a protein destabilized by the mTOR pathway34 - is degraded in a muskelin- and GID4-dependent manner when mTOR is inhibited. Since muskelin is also required for HMGCS1 degradation, its association with the core complex potentially creates the correct structure within the cavity of the supramolecular complex that facilitates GID4 anchoring of the HMGCS1 degron and efficient polyubiquitination of HMGCS1 by the CTLH catalytic subunits. This mechanism is likely similar to how Gid7 mediates Fbp1 degradation in yeast20,38.
Indeed, while this manuscript was under review, another study similarly reported that the Torin-1 induced degradation of HMGCS1 occurs through GID4 binding and a muskelin-CTLH complex39. In particular, they also validated HMGCS1 ubiquitination in vitro, established the requirement of the HMGCS1 N-terminal proline for degradation, and demonstrated HMGCS1 degradation is important for metabolic flux and cell proliferation39.
The S. cerevisiae Gid complex, homologous to the CTLH complex, achieves substrate diversity and regulation through distinct substrate receptors8. For example, the Gid complex can function through the substrate receptors Gid4, Gid10, and Gid11 that differ in their degron preferences, are activated by different stimuli, and undergo autoregulation once ubiquitination of targets is completed19,40,41,42.
Thus far, only GID4 substrate targeting has been defined for the human complex. Our data indicates that in order to regulate substrate diversity with only one substrate receptor, the CTLH complexes in multicellular organisms evolved to utilize and regulate the structurally similar but distinct β-propeller subunits WDR26 and muskelin. By way of their differing oligomeric states, the control of WDR26 versus muskelin complex formation dictates CTLH complex substrate preferences. This could be the reason why autoregulation of human CTLH complex activity occurs via its ability to degrade muskelin, allowing a quick shift between WDR26 and muskelin within the complex while maintaining availability of the general scaffold. Curiously, only muskelin – and not WDR26 – is regulated through autoubiquitination by the CTLH complex22,24. One potential explanation may involve the structural differences between muskelin- and WDR26-containing CTLH complexes. For instance, muskelin has been reported to form a tetramer that can give rise to a sheet-like CTLH complex appearance while WDR26 dimerizes and mediates mainly donut-shaped supramolecular complexes18. These structural differences potentially allow RING subunits MAEA and RMND5A to be in the proper position for muskelin ubiquitination, but not WDR2618,20. Previous studies have shown that WDR26 protein levels may be regulated by specific signalling pathways at the level of gene expression, since WDR26 protein levels correlate with changes in its gene expression during erythroid maturation13,43,44.
Our proteomic data in unstimulated HeLa cells highlighted HMGCS1 as a clear muskelin-specific target. We then demonstrated that HMGCS1 regulation by the CTLH complex was accelerated when mTOR was inhibited after finding evidence that mTOR regulates HMGCS1 protein levels34. The mechanism through which mTOR inhibition promotes degradation of HMGCS1 through the muskelin-CTLH complex still remains to be elucidated. We found it did not affect muskelin or GID4 association with the complex, suggesting the recruitment of HMGCS1 could involve post-translational modification of CTLH complex member(s) and/or the involvement of a substrate adaptor or a regulatory factor, similar to YPEL5 regulation of substrate binding to WDR2645.
One potential WDR26-specific target highlighted by our proteomic screens was CKAP4. CKAP4 is a Pro/N-degron containing protein that was increased in RMND5A, MAEA, and WDR26 KO cells, but not muskelin KO cells. In the future, other targets of such nature could be uncovered by conducting the proteomics in additional cell types and with a stimulus. However, we acknowledge that an increased protein in the proteome datasets may not necessarily mean it is a degradation target because (1) long-term KO cell lines can produce adaptive responses and (2) protein abundance changes can occur through multiple mechanisms such as at the RNA level. Additionally, methods other than global proteomics would be valuable to explore non-degradative ubiquitination by the CTLH complex during mTOR inhibition and/or in the context of the interplay between muskelin and WDR26.
Overall, we have integrated and extended the in vitro findings of yeast Gid7, its mammalian homologue WDR26, and the structurally similar muskelin in mediating supramolecular complex formation19,20,21,22 into biological relevance using a human cell model. The biological importance of the interplay we uncovered between WDR26 and muskelin and how it affects CTLH complex function was evident in the case of Torin-1 induced degradation of HMGCS1. The increased activity of muskelin-containing CTLH complexes during mTOR inhibition provides clues into which biological processes require proper activation for normal function. The mTOR-AMPK axis regulates many processes including cellular proliferation, cilia development, glucose metabolism, and ribosome biogenesis, all of which have been proposed as CTLH complex-regulated processes in previous studies9,10,12,23,46,47. This interplay may also be relevant during human cytomegalovirus virus infection in which a WDR26-independent shift in complex composition was observed48 and for the drosophila CTLH complex in which the muskelin-dependent CTLH complex is responsible for degrading proteins as part of the maternal to zygotic transition (MZT), followed by muskelin ubiquitination and degradation14,15. How this WDR26 and muskelin interplay and autoregulation may extend to other pathways and biological processes will be an exciting area of future study, and a necessary consideration for future development of CTLH complex-targeting TPD compounds.
Materials and methods
Cell culture
HeLa and HEK293 cells were obtained from the American Type Culture Collection (ATCC). Generation of KO HeLa and HEK293 cell lines have been described previously22,23. All cells were cultured in high glucose (4.5 g/L) Dulbecco’s modified Eagle’s medium (Wisent Bioproducts, St. Bruno, Quebec, Canada) supplemented with 10% fetal bovine serum, 1% sodium pyruvate, and 1% L-glutamine (609-065-EL, Wisent Bioproducts) at 37 °C and 5% CO2. Cells were treated with DMSO (DMS666, Bioshop), 50 µM cycloheximide (Cyc003.1, Bioshop), 10 µM MG132 (474790, Cedarlane), 250 nM Torin1 (475991, Millipore Sigma), and 10 µM PFI-7N or PFI-7 (Structural Genomics Consortium) for the indicated times. Plasmid transfections were carried out with jetPRIME (CA89129-924, Polypus Transfection) on 80% confluent cells seeded the day before according to the manufacturer’s protocol. Cells were regularly tested to ensure absence of mycoplasma contamination.
Cloning
All template DNA used in this study was described previously22, see Supplemental Table 1 below for a list of all cloning reagents. RanBP9 domain deletions have also been described previously49,50. Point mutations were generated via site directed mutagenesis with KOD polymerase (71086-3, Novagen) followed by Dpn1 digest (R0176, New England Biotechnology). Domain deletions were generated using tail-to-tail PCR with KOD polymerase and primers complementary to the regions immediately before and after the DNA to be deleted. All primers listed below were ordered from Integrated DNA Technologies (IDT). The entire cDNA sequence was validated by Sanger sequencing.
Cell extracts
Whole cell extracts for immunoprecipitation and western blot were prepared by lysing cells in whole cell extract buffer containing fresh inhibitors (50 mM HEPES pH7.4, 150 mM NaCl, 10 mM EDTA, 10% glycerol, 0.5% NP40, 10 mM DTT, 1 mM Na3VO4, 10 mM NaF, 1 mM phenylmethylsulphonyl fluoride (PMSF), 1 μg/ml of aprotinin, 10 μg/ml of pepstatin, and 1 μg/ml of leupeptin) on ice for 20 min, then centrifuging for 20 minutes 13,000 rpm at 4 ˚C, as previously described in ref. 22. Whole cell extracts prepared for sucrose gradients underwent the same lysis but with lysis buffer containing no glycerol. For the PFI-7 experiment, cells were lysed in lysis buffer (20 mM Tris–HCl pH 8, 150 mM NaCl, 1 mM EDTA, 10 mM MgCl2, 0.5% Triton X-100, 12.5 U mL−1 benzonase). After 2 min incubation at RT, SDS was added to final 1% concentration. Preparation of nuclear extracts was also previously described in ref. 23; briefly, scraped cells were incubated in cytoplasmic lysis buffer (50 mM HEPES pH 7.4, 150 mM NaCl) supplemented with 10 µg/mL aprotinin, 2 µg/mL leupeptin, 2.5 µg/mL pepstatin, 1 mM DTT, 2 mM NaF, 2 mM Na3VO4, 0.1 mM PMSF, and 75 µg/mL digitonin (D141, Millipore Sigma). After a 10 min incubation on ice, cells were centrifuged at 500 x g for 1 min and the supernatant was discarded. The resulting pellet was resuspended with nuclear lysis buffer (10 mM Tris-HCl pH 8.0, 100 mM NaCl, 0.1% sodium deoxycholate, 1 mM EDTA, 25% glycerol) supplemented with 10 µg/mL aprotinin, 2 µg/mL leupeptin, 2.5 µg/mL pepstatin, 1 mM DTT, 2 mM NaF, 2 mM Na3VO4, 0.1 mM PMSF, and 100 units/mL of Benzonase nuclease (E1014, Sigma Aldrich) and incubated on ice for 30 min. The samples were then centrifuged at 17,000 x g for 20 min and supernatants were isolated from cell debris. Protein concentration was estimated by Pierce 660 nm protein assay (22660, ThermoFisher).
Immunoprecipitation
For RanBP9 IPs, preconjugation of antibody and immunoprecipitation protocol has been described elsewhere with slight modifications10. Four µg of RanBP9 antibody (F-1, sc-271727, Santa Cruz Biotechnology) and mouse IgG (sc-2025, Santa Cruz Biotechnology) were conjugated to 15 µL Dynabeads Protein G (10004D, Invitrogen) for 1 h at 4 °C with end-over-end rotation. Afterwards, antibody-bead conjugate was washed three times in 100 mM sodium Borate, pH 9 (M5162, Sigma-Aldrich), then resuspended in 20 mM DMP (Dimethyl pimelimidate dihydrochloride, D8388, Sigma-Aldrich) in 100 mM sodium Borate, pH 9 and rotated at room temperature for 30 min. Crosslinking was quenched by washing and incubation in 200 mM ethanolamine, pH 8 (S9640, Sigma-Aldrich) for 2 h at 4 °C with end-over-end rotation. After blocking, beads were washed three times in 1 mL whole cell extract buffer and kept on ice. 1 mg of whole cell extracts, prepared as described above, were then adjusted to 0.25% NP-40 and precleared using 5 µL Dynabeads Protein G. The preconjugated beads were then incubated with the precleared whole cell extracts overnight at 4 °C while rotating. For FLAG and HA IPs, extracts were adjusted to 0.25% NP-40, pre-cleared and rotated overnight at 4 °C with FLAG (M2, F1804, Sigma-Aldrich) or HA (HA-7, H9658, Sigma-Aldrich) then incubated for 1 h at 4 °C with Dynabeads protein G. In all IP types, beads were washed three times in wash IP buffer (25 mM HEPES pH 7.9, 60 mM KCl, 0.5 mM EDTA, 0.25% NP-40, 12% glycerol), resuspended in SDS loading buffer, and boiled at 95 °C for 10 min to elute proteins prior to SDS-PAGE.
Sucrose Gradient
Preparation of sucrose gradient fractions was described in a previous publication23. Three hundred µg of either whole cell extract or nuclear extract, prepared as described above was loaded onto a 5–40% sucrose gradient (w/v) made with whole cell or nuclear lysis buffer the night before. Gradients were centrifuged in a SW41 rotor at 30 400 RPM for 16 h at 4 °C. Fractions were collected and 10% TCA and acetone were used to precipitate the proteins from each fraction. The protein pellets were resuspended in SDS loading buffer and boiled at 95 °C for 10 min prior to SDS-PAGE.
Western blot
For western blot using chemiluminescent imaging, protein was loaded onto a 10% or 8% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) gel, transferred to Polyvinylidene difluoride (PVDF) membrane, hybridized, and assessed as described previously22. For fluorescence-based imaging, protein was loaded onto a 10% or 8% SDS-PAGE gel and was transferred to Immobilon-FL PVDF membrane (IPFL00005, Millipore). The blot was blocked for one hour and hybridized with primary antibody overnight in 0.5% fish gelatin (G7041, Sigma) dissolved in TBST (0.05% tween). The next day, the blot was washed 3 times in TBST for 10 minutes at room temperature and incubated in one of four secondary antibodies (IRDye 680RD Goat anti-Rabbit IgG, 926-68071, LI-COR Biotechnology; IRDye 680RD Donkey anti-Mouse IgG, 926-68072, LI-COR Biotechnology; IRDye 800CW Donkey anti-Rabbit IgG, 926-32213, LI-COR Biotechnology; IRDye 800CW Donkey anti-Mouse IgG, 926-32212, LI-COR Biotechnology) diluted at 1:10000 in 0.5% fish gelatin dissolved in TBST. After 3 washes in TBST for 10 min in the dark at room temperature, the blots were imaged using ChemiDoc MP (BioRad). Images were prepared and band intensities were quantified using ImageLab (BioRad). For immunoblotting of the PFI-7 experiment, samples were boiled in SDS loading buffer before western blotting using the NuPAGE electrophoresis and transfer system (Invitrogen). Immunoblots were imaged on a Li-Cor Odyssey CLx and quantified in Image Studio Lite v5.2.5 (Li-Cor Biosciences). Primary antibodies used in any experiments of this study are: Vinculin (E1E9V, Cell Signaling Technology); Muskelin (C-12, sc-398956, Santa Cruz Biotechnology); HA (HA-7, H3663 Sigma-Aldrich); FLAG (M2, F1804, Sigma-Aldrich); WDR26 (ab85962, Abcam); RanBP9 (5M, 71-001, Bio academia); GID4 (NBP1-53185, Novus), ARMC8 (E-1, sc-365307, Santa Cruz Biotechnology); Tubulin (T5168, Sigma-Aldrich); RMND5A (custom, Yenzyme Antibodies); p-S6 (2215S, Cell Signaling Techonologies); S6 (2317S, Cell Signaling Technologies); Actin (A5441, Sigma-Aldrich); and HMGCS1 (A-6, sc-166763, Santa Cruz Biotechnology).
MS sample preparation for global proteomics
HeLa cells at 75–80% confluency were trypsinized, cell pellet was collected by centrifugation, washed, and then frozen at −80 °C. Cells were lysed by resuspension in 8 M urea, 50 mM ammonium bicarbonate (ABC), 10 mM Dithiothreitol (DTT), 2% Sodium dodecyl sulfate (SDS) and then sonicated with a probe sonicator (20 × 0.5 s pulses; Level 1). Twenty-five µg of protein lysate, as quantified by Pierce™ 660 nm Protein Assay (22660, 22663, ThermoFisher Scientific), was reduced in 10 mM DTT for 25 min, alkylated in 100 mM iodoacetamide for 25 min in the dark, followed by methanol precipitation as previously described in ref. 51. The protein pellet was resuspended in 200 µL of 50 mM ABC and subjected to a sequential digest first with 250 ng of LysC (125-05061, Wako Chemicals, USA) for 4 h, then 500 ng of Trypsin/LysC (V5071, Promega) for 16 h, followed by 500 ng of Trypsin (V5111, Promega) for an additional 4 h. Digestions were incubated at 37 ˚C at 600 rpm with interval mixing (30 s mix, 2 min pause) on a Thermomixer C (2231000667, Eppendorf). After the last digestion, samples were acidified with 10% formic acid (FA) to pH 3–4 and centrifuged at 14,000 x g to pellet insoluble material.
diGLY enrichment
HeLa cells at 75–80% confluency were treated with 10 μM MG132 for 4 h and processed exactly as described for global proteomics. After methanol precipitation, 1 mg protein was digested sequentially as described for global proteomics but with 6.66 μg of Lys-C, 20 μg of Trypsin/Lys-C, and 20 μg of Trypsin. Peptides were then dried using a SpeedVac (Thermo Scientific). PTMScan® Ubiquitin Remnant Motif (K-ε-GG) Antibody Bead Conjugate (5562, Cell Signaling Technology) (25 µL per sample) was crosslinked and used as previously described52. Briefly, antibody-bead conjugate was washed three times in 100 mM sodium borate, pH 9 (M5162, Sigma-Aldrich), then resuspended in 20 mM DMP (Dimethyl pimelimidate dihydrochloride, D8388, Sigma-Aldrich) in 100 mM sodium borate, pH 9 and rotated at room temperature for 30 min. Crosslinking was quenched by washing and incubation in 200 mM ethanolamine, pH 8 (S9640, Sigma-Aldrich) for 2 h at 4 °C with end-over-end rotation. After blocking, beads were washed three times in 1 mL of 1x IAP buffer (provided by the kit) and kept on ice. Dried peptides were resuspended in 1 mL of IAP buffer, centrifuged, and the supernatant was added to the crosslinked antibody-beads and incubated with rotation for 1 h at 4 °C. After enrichment, beads were washed twice with IAP buffer, three times with PBS, and then peptides were eluted with two rounds of 5 min incubations in 0.15% Trifluoroacetic acid (TFA). The eluted peptides were dried in a SpeedVac and reconstituted in 0.1% FA.
Liquid Chromatography Tandem Mass-Spectrometry (LC-MS/MS) for Global Proteome and RMND5A diGly enrichment
Approximately 1 µg of peptide sample (as determined by Pierce BCA assay) was injected onto a Waters M-Class nanoAcquity HPLC system (Waters) coupled to an ESI Orbitrap mass spectrometer (Q Exactive plus, ThermoFisher Scientific) operating in positive mode. Buffer A consisted of mass spectrometry grade water with 0.1% FA and buffer B consisted of acetonitrile with 0.1% FA (ThermoFisher Scientific). All samples were trapped for 5 min at a flow rate of 5 µL/min using 99% buffer A and 1% buffer B on a Symmetry BEH C18 Trapping Column (5 mm, 180 mm x 20 mm, Waters). Peptides were separated using a Peptide BEH C18 Column (130 Å, 1.7 mm, 75 mm x 250 mm) operating at a flow rate of 300 nL/min at 35 ˚C (Waters). Proteome Samples were separated using a non-linear gradient consisting of 1%–7% buffer B over 1 min, 7%–23% buffer B over 179 min and 23%–35% buffer B over 60 min, before increasing to 98% buffer B and washing. RMND5A diGLY enriched samples were trapped for 5 min then separated using a non-linear gradient consisting of 1%–7.5% buffer B over 1 min, 7.5%–25% buffer B over 179 min, 25%–32.5% buffer B over 40 min and 32.5%–40% over 20 min before increasing to 98% buffer B and washing. MS acquisition settings are provided in Supplementary Table 2.
Proteomic data analysis
The proteomic samples were processed using a label-free data-dependent acquisition (DDA) approach. Raw mass spectrometry files were converted to mzml format using msconvert within proteowizard (version 3.0.22167)53. Database searches were performed using MSFragger (version 3.7)54,55 within FragPipe (version 19.0). For all database searches, MSFragger was used to search against the reviewed human proteome retrieved from Uniprot (2023-04-19 version) containing 40910 entries with reverse decoys and common contaminants added. Datasets were analysed in batches according to their acquisition time and analysis requirements. MKLN1 and WDR26 proteomes made up one analysis batch while the second batch consisted of RMND5A and MAEA proteomes. Ubiquitin remnant proteomic samples were processed in a separate run. Precursor and fragment mass tolerances were set to 20 ppm for total proteome. Enzyme specificity was set to trypsin, with up to two missed cleavages allowed for whole proteomes and three for ubiquitin remnant. Carbamidomethylation of cysteine (+57.02146 Da) was set as a fixed modification and oxidation of methionine (+15.9949 Da) and N-terminal acetylation (+42.0106 Da) were set as variable modifications. For ubiquitin remnant proteomics, GlyGly (+114.04293) was set as an additional variable modification. Filtering of peptide-spectrum matches (PSM) was performed using Philosopher (version 4.8.1)56 with false-discovery rate set to 1% at the PSM, peptide, and protein level. MSBooster was turned on (version 1.1.11)57 and for ubiquitin remnant site localization, PTMProphet was used58 with a minimum site threshold of 0.5. Label-Free Quantification was performed by IonQuant (version 1.8.10)59 using the MaxLFQ algorithm with match-between-runs enabled between all runs within each batch with a false-discovery rate threshold of 1%.
Subsequent data analysis was performed in R (version 4.1.2) with fully annotated analysis code available at https://github.com/d0minicO/CTLH_proteomes_analysis. DEP (version 1.16.0)60 was used for data normalization and imputation. Protein intensities for whole proteome data or peptide intensities for ubiquitin remnant data were filtered to retain rows missing in no more than two replicates of each condition. For whole proteome data, proteins with two or more peptides and at least one unique peptide was retained. MS1 intensities were log2-transformed and variance stabilizing normalization (vsn) was applied. Mixed imputation was performed to impute missing values with K-nearest neighbour (KNN) used for missing at random (MAR) and Quantile Regression Imputation of Left Censored Data (QRILC) used for missing not at random (MNAR). Limma (version 3.50.3)61 was used to detect significant protein abundance changes with separate regression models for each batch. P-values were adjusted using Benjamini-Hochberg procedure with a significance threshold of 0.05 and fold-change threshold of 1.5. To visualise relevant biological variability across each dataset, dimensionality reduction was performed using principal components analysis (PCA) of a matrix containing all samples in each batch with all ubiquitin remnant-containing peptides or the subset of significantly changed proteins. Pathways analysis on sets of changed proteins was performed using clusterProfiler (version 4.2.2)62 with all quantified proteins in each batch as background. Hierarchical clustering on heatmaps was done using the complete method within pheatmap (version 1.0.12) (https://github.com/raivokolde/pheatmap). To aid visualization of protein abundances in boxplots, median centering was applied.
Statistics and reproducibility
Statistical analyses were performed using R (4.3.1). Statistical test used and associated metrics are defined in each figure legend. Results were considered significant when P < 0.05. Experiments were conducted a minimum of 3 times (or as indicated) with separate cell culture preparations or transfections.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
Mass spectrometry raw files have been deposited in ProteomeXchange through partner Mass spectrometry Interactive Virtual Environment (MassIVE) with accession code MSV000094320. Uncropped/unedited blots are provided in Supplemental Fig. 6. Source data underlying graphs and charts can be obtained in Supplementary Data 4.
References
Komander, D. & Rape, M. The Ubiquitin Code. Annu. Rev. Biochem. 81, 203–229 (2012).
Dikic, I. & Schulman, B. A. An expanded lexicon for the ubiquitin code. Nat. Rev. Mol. Cell Biol. 24, 273–287 (2023).
Rape, M. Ubiquitylation at the crossroads of development and disease. Nat. Rev. Mol. Cell Biol. 19, 59–70 (2017).
Popovic, D., Vucic, D. & Dikic, I. Ubiquitination in disease pathogenesis and treatment. Nat. Med. 20, 1242–1253 (2014).
Pickart, C. M. Mechanisms Underlying Ubiquitination. Annu Rev. Biochem. 70, 503–533 (2001).
Deshaies, R. J. & Joazeiro, C. A. P. RING Domain E3 Ubiquitin Ligases. Annu Rev. Biochem. 78, 399–434 (2009).
Vittal, V., Stewart, M. D., Brzovic, P. S. & Klevit, R. E. Regulating the Regulators: Recent Revelations in the Control of E3 Ubiquitin Ligases. J. Biol. Chem. 290, 21244–21251 (2015).
Maitland, M. E. R., Lajoie, G. A., Shaw, G. S. & Schild-Poulter, C. Structural and Functional Insights into GID/CTLH E3 Ligase Complexes. Int J. Mol. Sci. 23, 5863 (2022).
McTavish, C. et al. Regulation of c-Raf Stability through the CTLH Complex. Int J. Mol. Sci. 20, 934 (2019).
Maitland, M. E. R., Kuljanin, M., Wang, X., Lajoie, G. A. & Schild-Poulter, C. Proteomic analysis of ubiquitination substrates reveals a CTLH E3 ligase complex-dependent regulation of glycolysis. FASEB J. 35, e21825 (2021).
Lampert, F. et al. The multi-subunit GID/CTLH E3 ubiquitin ligase promotes cell proliferation and targets the transcription factor Hbp1 for degradation. Elife 7, e35528 (2018).
Liu, H. et al. The GID ubiquitin ligase complex is a regulator of AMPK activity and organismal lifespan. Autophagy 16, 1618–1634 (2020).
Sherpa, D. et al. Modular UBE2H-CTLH E2-E3 complexes regulate erythroid maturation. Elife 11, e77937 (2022).
Zavortink, M. et al. The e2 marie kondo and the ctlh e3 ligase clear deposited rna binding proteins during the maternal-to-zygotic transition. Elife 9, 1–48 (2020).
Cao, W. X. et al. Precise temporal regulation of post-transcriptional repressors is required for an orderly drosophila maternal-to-Zygotic transition. Cell Rep. 31, 107783 (2020).
Pfirrmann, T. et al. RMND5 from Xenopus laevis Is an E3 Ubiquitin-Ligase and Functions in Early Embryonic Forebrain Development. PLoS One 10, e0120342 (2015).
Santt, O. et al. The Yeast GID Complex, a Novel Ubiquitin Ligase (E3) Involved in the Regulation of Carbohydrate Metabolism. Mol. Biol. Cell 19, 3323–3333 (2008).
van gen Hassend, P. M. et al. RanBP9 controls the oligomeric state of CTLH complex assemblies. J. Biol. Chem. 299, 102869 (2023).
Qiao, S. et al. Interconversion between Anticipatory and Active GID E3 Ubiquitin Ligase Conformations via Metabolically Driven Substrate Receptor Assembly. Mol. Cell 77, 150–163.e9 (2020).
Sherpa, D. et al. GID E3 ligase supramolecular chelate assembly configures multipronged ubiquitin targeting of an oligomeric metabolic enzyme. Mol. Cell 81, 2445–2459.e13 (2021).
Mohamed, W. I. et al. The human GID complex engages two independent modules for substrate recruitment. EMBO Rep. 22, e52981 (2021).
Maitland, M. E. R. et al. The mammalian CTLH complex is an E3 ubiquitin ligase that targets its subunit muskelin for degradation. Sci. Rep. 9, 1–14 (2019).
Onea, G., Maitland, M. E. R., Wang, X., Lajoie, G. A. & Schild-Poulter, C. Distinct nuclear and cytoplasmic assemblies and interactomes of the mammalian CTLH E3 ligase complex. J. Cell Sci. 135, jcs259638 (2022).
Jordan, V. N., Ordureau, A. & An, H. Identifying E3 Ligase Substrates With Quantitative Degradation Proteomics. ChemBioChem 24, 1–9 (2023).
Huttlin, E. L. et al. Dual proteome-scale networks reveal cell-specific remodeling of the human interactome. Cell 184, e28 (2021).
Dong, C. et al. Molecular basis of GID4-mediated recognition of degrons for the Pro/N-end rule pathway. Nat. Chem. Biol. 14, 466–473 (2018).
Chrustowicz, J. et al. Multifaceted N-Degron Recognition and Ubiquitylation by GID/CTLH E3 Ligases. J. Mol. Biol. 434, 167347 (2022).
Chen, S.-J., Wu, X., Wadas, B., Oh, J.-H. & Varshavsky, A. An N-end rule pathway that recognizes proline and destroys gluconeogenic enzymes. Science (1979) 355, eaal3655 (2017).
Owens, D. D. G. et al. A chemical probe to modulate human GID4 Pro/N-degron interactions. Nat. Chem. Biol. 20, 1164–175 (2024).
Sun, X. et al. Elucidation of CKAP4-remodeled cell mechanics in driving metastasis of bladder cancer through aptamer-based target discovery. Proc. Natl. Acad. Sci. USA 119, e2110500119 (2022).
Osugi, Y., Fumoto, K. & Kikuchi, A. CKAP4 Regulates Cell Migration via the Interaction with and Recycling of Integrin. Mol. Cell Biol. 39, e00073–19 (2019).
Menssen, R. et al. Exploring the topology of the gid complex, the E3 ubiquitin ligase involved in catabolite-induced degradation of gluconeogenic enzymes. J. Biol. Chem. 287, 25602–25614 (2012).
Delto, C. F. et al. The LisH Motif of Muskelin Is Crucial for Oligomerization and Governs Intracellular Localization. Structure 23, 364–373 (2015).
Zhao, J., Zhai, B., Gygi, S. P. & Goldberg, A. L. MTOR inhibition activates overall protein degradation by the ubiquitin proteasome system as well as by autophagy. Proc. Natl Acad. Sci. USA 112, 15790–15797 (2015).
Chana, C. K. et al. Discovery and Structural Characterization of Small Molecule Binders of the Human CTLH E3 Ligase Subunit GID4. J. Med Chem. 65, 12725–12746 (2022).
Schapira, M., Calabrese, M. F., Bullock, A. N. & Crews, C. M. Targeted protein degradation: expanding the toolbox. Nat. Rev. Drug Discov. 18, 949–963 (2019). 2019 18:12.
Serebrenik, Y. V., Mani, D., Maujean, T., Burslem, G. M. & Shalem, O. Pooled endogenous protein tagging and recruitment for systematic profiling of protein function. Cell Genomics 4, 100651 (2024).
Chrustowicz, J. et al. Multisite phosphorylation dictates selective E2-E3 pairing as revealed by Ubc8/UBE2H-GID/CTLH assemblies. Mol Cell 84, 293–308.e141–16 (2024).
Yi, S. A., Sepic, S., Schulman, B. A., Ordureau, A. & An, H. mTORC1-CTLH E3 ligase regulates the degradation of HMG-CoA synthase 1 through the Pro/N-degron pathway. Mol. Cell 84, 2166–2184.e9 (2024).
Menssen, R., Bui, K. & Wolf, D. H. Regulation of the Gid ubiquitin ligase recognition subunit Gid4. FEBS Lett. 592, 3286–3294 (2018).
Langlois, C. R. et al. A GID E3 ligase assembly ubiquitinates an Rsp5 E3 adaptor and regulates plasma membrane transporters. EMBO Rep. 23, 1–17 (2022).
Kong, K.-Y. E. et al. Timer-based proteomic profiling of the ubiquitin-proteasome system reveals a substrate receptor of the GID ubiquitin ligase. Mol. Cell 81, 2460–2476.e11 (2021).
Zhen, R. et al. Wdr26 regulates nuclear condensation in developing erythroblasts. Blood 135, 208–219 (2020).
An, X. et al. Global transcriptome analyses of human and murine terminal erythroid differentiation. Blood 123, 3466–3477 (2014).
Gottemukkala, K. V. et al. Non-canonical substrate recognition by the human WDR26-CTLH E3 ligase regulates prodrug metabolism. Mol. Cell 84, 1948–1963.e11 (2024).
Ling, N. X. Y. et al. mTORC1 directly inhibits AMPK to promote cell proliferation under nutrient stress. Nat. Metab. 2, 41–49 (2020). 2020 2:1.
Hantel, F. et al. Cilia-localized GID/CTLH ubiquitin ligase complex regulates protein homeostasis of sonic hedgehog signaling components. J. Cell Sci. 135, 260203 (2022).
Hashimoto, Y., Sheng, X., Murray-Nerger, L. A. & Cristea, I. M. Temporal dynamics of protein complex formation and dissociation during human cytomegalovirus infection. Nat. Commun. 11, 1–20 (2020).
Salemi, L. M., Loureiro, S. O. & Schild-Poulter, C. Characterization of RanBPM Molecular Determinants that Control Its Subcellular Localization. PLoS One 10, e0117655 (2015).
Salemi, L. M., Almawi, A. W., Lefebvre, K. J. & Schild-Poulter, C. Aggresome formation is regulated by RanBPM through an interaction with HDAC6. Biol. Open 3, 418–430 (2014).
Kuljanin, M., Dieters-Castator, D. Z., Hess, D. A., Postovit, L. M. & Lajoie, G. A. Comparison of sample preparation techniques for large-scale proteomics. Proteomics 17, 1600337 (2017).
Udeshi, N. D., Mertins, P., Svinkina, T. & Carr, S. A. Large-scale identification of ubiquitination sites by mass spectrometry. Nat. Protocols. 8, 1950–1960 (2013).
Chambers, M. C. et al. A cross-platform toolkit for mass spectrometry and proteomics. Nat. Biotechnol. 30, 918–920 (2012).
Teo, G. C., Polasky, D. A., Yu, F. & Nesvizhskii, A. I. Fast Deisotoping Algorithm and Its Implementation in the MSFragger Search Engine. J. Proteome Res 20, 498–505 (2021).
Kong, A. T., Leprevost, F. V., Avtonomov, D. M., Mellacheruvu, D. & Nesvizhskii, A. I. MSFragger: ultrafast and comprehensive peptide identification in mass spectrometry–based proteomics. Nat. Methods 14, 513–520 (2017).
da Veiga Leprevost, F. et al. Philosopher: a versatile toolkit for shotgun proteomics data analysis. Nat. Methods 17, 869–870 (2020).
Yang, K. L. et al. MSBooster: improving peptide identification rates using deep learning-based features. Nat. Commun. 14, 1–14 (2023).
Shteynberg, D. D. et al. PTMProphet: Fast and Accurate Mass Modification Localization for the Trans-Proteomic Pipeline. J. Proteome Res. 18, 4262–4272 (2019).
Yu, F., Haynes, S. E. & Nesvizhskii, A. I. IonQuant Enables Accurate and Sensitive Label-Free Quantification With FDR-Controlled Match-Between-Runs. Mol. Cell. Proteom. 20, 100077 (2021).
Zhang, X. et al. Proteome-wide identification of ubiquitin interactions using UbIA-MS. Nat. Protoc. 13, 530–550 (2018).
Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47–e47 (2015).
Wu, T. et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innovation 2, 100141 (2021).
Acknowledgements
This work was supported by the Canadian Institutes of Health Research (MOP-142414 and PJT-169101 to C.S.P.; FDN154328 to C.H.A.); Mass spectrometry analyses were performed on equipment funded by a grant from the Canada Foundation for Innovation to G.A.L.; G.O., M.E.R.M. and B.G.C. were supported by a Postgraduate Doctoral Scholarship from the Natural Sciences and Engineering Research Council of Canada. The Structural Genomics Consortium is a registered charity (no: 1097737) that receives funds from Bayer AG, Boehringer Ingelheim, Bristol Myers Squibb, Genentech, Genome Canada through Ontario Genomics Institute (OGI-196), EU/EFPIA/OICR/McGill/KTH/Diamond Innovative Medicines Initiative 2 Joint Undertaking (EUbOPEN grant 875510), Janssen, Merck KGaA (aka EMD in Canada and US), Pfizer and Takeda.
Author information
Authors and Affiliations
Contributions
C.S.P., M.E.R.M., and G.O. designed project and experiments. M.E.R.M. and G.O. conducted most experiments, data analysis and generated most figures. B.C.G.C. and X.W. performed experiments. M.E.R.M. and D.D.G.O performed bioinformatic analyses and resulting figures. M.E.R.M. and G.O. wrote the manuscript with input from C.S.P. G.A.L., D.B.L., C.H.A. and C.S.P. provided supervision and funding.
Corresponding author
Ethics declarations
Competing interests
D.D.G.O. is an employee of Amphista Therapeutics, a company that is developing targeted protein degradation therapeutic platforms. The remaining authors declare no competing interests.
Peer review
Peer review information
Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editors: Albert Lee and Ophelia Bu. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Maitland, M.E.R., Onea, G., Owens, D.D.G. et al. Interplay between β-propeller subunits WDR26 and muskelin regulates the CTLH E3 ligase supramolecular complex. Commun Biol 7, 1668 (2024). https://doi.org/10.1038/s42003-024-07371-3
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s42003-024-07371-3
