EEL-1/HUWE1 promotes degradation of ubiquitin-fused GFP

Reporter proteins fused at their N-terminus to the G76V mutant form of Ub are subject to rapid degradation via the Ub fusion degradation (UFD) pathway and provide a powerful tool to study proteasome-mediated degradation in vivo [53–57]. UFD depends on multiple Ub ligases that act in concert to extend Ub chains on the fused ubiquitin moiety [56,58–63]. UFD is remarkably robust to inhibition of the proteolytic capacity of the proteasome; there is no detectable accumulation of Ub[G76V]-fused GFP in cancer cells unless >80% of proteasome activity is eliminated [53]. These observations suggest that the UPS is highly robust to proteolytic challenges. However, the mechanistic basis and physiologic significance of robust UPS function in this context is not clear.

We previously generated a transgene that ubiquitously expresses Ub[G76V]::GFP (hereafter UbV-GFP) in C. elegans [14,16]. UbV-GFP fluorescence is not detectable in wild type animals but accumulates to low but detectible levels in intestinal cells of skn-1a and png-1 mutant animals, reflecting a requirement for SKN-1A/Nrf1 for adequate proteasome levels [14]. We hypothesized that UbV-GFP accumulation in SKN-1A pathway mutants would be exacerbated if other components that promote robust UPS function are inactivated. To uncover these factors, we carried out a forward genetic screen for increased UbV-GFP fluorescence in animals lacking either skn-1a or png-1. This screen identified several alleles affecting the eel-1 gene (Fig 1A). eel-1 encodes a highly conserved ~465 kDa HECT-domain E3 ubiquitin ligase orthologous to human HUWE1 and S. cerevisiae Tom1. HUWE1 knockdown in human cancer cells causes increased accumulation of UbV-GFP [58] and inactivation of Tom1 in yeast causes partial stabilization of UFD substrates [60]. RNAi-mediated knockdown of eel-1 leads to increased UbV-GFP accumulation in C. elegans using an independently generated UbV-GFP transgene [55]. Thus, identification of eel-1 in our screen supports that this ubiquitin ligase plays a conserved role in degradation of Ub fusion proteins.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. EEL-1/HUWE1 promotes turnover of UbV-GFP.

A) Schematic of the eel-1 gene. Locations and effects of mutations identified in a mutagenesis screen for increased accumulation of UbV-GFP are labelled in black. Locations and effects of CRISPR/Cas9-engineered alleles are labelled in red. B) Fluorescence micrographs showing that simultaneous inactivation of skn-1a and eel-1 causes dramatic accumulation of UbV-GFP. All images are of L4 stage animals. Scale bar 100 μM. C) Quantification of UbV-GFP levels shown in (B). UbV-GFP accumulation is not increased in eel-1 single mutants and is moderately increased in skn-1a single mutants. UbV-GFP accumulation is dramatically increased in eel-1 skn-1a double mutants. Error bars show mean ± SD. n = 45 animals per genotype. Ns p > 0.05, **** p < 0.0001, ordinary one-way ANOVA with Šídák’s multiple comparisons test. D) Fluorescence micrographs showing that accumulation of UbV-GFP following proteasome inhibition is increased in eel-1 mutants. L4 animals were exposed to the indicated BTZ concentration (or DMSO control) for 24 hours prior to imaging. Imaged animals are day 1 adults. Scale bar 100 μM. E) Quantification of UbV-GFP levels following BTZ exposure as described in panel (D). Exposure of eel-1 mutant animals to 0.5-1 μM BTZ leads to increased accumulation of UbV-GFP but has no effect on the wild type. Error bars show mean ± SD. N = 30 animals per genotype and condition (BTZ concentrations tested were: DMSO control, 0.1 μM, 0.5 μM, 1 μM). P-values indicate comparison between eel-1 mutant and wild type for each drug concentration. Ns p > 0.05, **** p < 0.0001, ordinary two-way ANOVA with Šídák’s multiple comparisons test.

https://doi.org/10.1371/journal.pgen.1012273.g001

To validate our screening results, we used CRISPR/Cas9 to introduce a premature termination codon/frameshift cassette at two positions in the eel-1 coding sequence (G173STOP, D2410STOP; Fig 1A). Both alleles are expected to act as nulls, given that they were engineered to include a frameshift downstream of the early stop codon and nonsense mediated decay is likely to prevent expression of the truncated protein. In addition, both alleles lack all regions encoding known domains important for EEL-1 function, including the HECT domain. Strikingly, in a wild type genetic background, neither allele causes detectable accumulation of UbV-GFP fluorescence. However, in a skn-1a or png-1 mutant background, eel-1 inactivation causes very high levels of UbV-GFP, far exceeding levels observed in single mutants (Figs 1B-1C; S1). This suggests that EEL-1 is not essential for proteasomal degradation of UbV-GFP but becomes salient when SKN-1A-mediated control of proteasome levels is lost. Since both premature termination codon mutations have the same effect on UbV-GFP, we used the G173STOP allele for all subsequent experiments.

The genetic interaction between eel-1 and skn-1a suggested that EEL-1 could play an important role to ensure turnover of ubiquitin fusion proteins when proteasome levels or function is reduced. The proteasome inhibitor bortezomib (BTZ) impairs proteasome function by blocking the β5/chymotrypsin-like active site, which is rate-limiting for protein turnover [64,65]. In WT animals, exposure to BTZ at high concentrations causes accumulation of UbV-GFP [16]. We found that in animals lacking EEL-1, the concentration of BTZ required to cause accumulation of UbV-GFP, as measured by live imaging, is significantly lower compared to the wild type (Fig 1D-1E). We conclude that EEL-1 is important to ensure turnover of UbV-GFP in contexts where the capacity for proteasome-mediated protein turnover is impaired.

As already mentioned, UFD depends on multiple ubiquitin ligases, including EEL-1/HUWE1/Tom1, which act in concert to poly-ubiquitinate the fused ubiquitin [56,58–63]. By western blot, we did not detect any reduction in ubiquitination of the UbV-GFP species that accumulate in eel-1 mutants following BTZ challenge or in the skn-1a mutant background (S2 Fig). In addition, we found that in the absence of any proteasomal insult, eel-1 mutants accumulate a small amount of UbV-GFP (S2 Fig), a difference undetectable by live imaging (see Fig 1B,1C). Together, these results indicate that the EEL-1 ubiquitin ligase, although not essential for ubiquitination of UbV-GFP, promotes its proteasomal degradation. The striking accumulation of ubiquitinated forms of UbV-GFP in eel-1 skn-1a double mutants indicates that in the context of proteasome insufficiency, EEL-1 is needed to ensure turnover even though UbV-GFP is still subject to ubiquitination by other ligases.

Simultaneous inactivation of EEL-1 and SKN-1A causes severe defects in development and physiology

Inactivation of EEL-1 causes a range of developmental and physiological defects [39–42]. During our analysis, we noticed that adverse consequences of eel-1 inactivation are severely exacerbated by loss of the SKN-1A/Nrf1 pathway. In eel-1 skn-1a or png-1; eel-1 double mutants, we observed frequent vulval rupture during the first two days of adulthood (Fig 2A, 2B). Additional rupture events occur later and almost all double mutant animals rupture by day 4 (Fig 2C). In contrast, single mutants lacking eel-1 rarely rupture within the first two days of adulthood, but do show rupture later in adulthood, affecting ~30% of animals by day 4 (Fig 2A-2C). skn-1a and png-1 mutants show elevated vulval rupture later in life (by day 7) [16], but no rupture during the first 4 days of adulthood (Fig 2A-2C). Thus, EEL-1 and SKN-1A each contribute to maintenance of vulval integrity during adulthood and simultaneous inactivation of both is synergistically detrimental. In most png-1; eel-1 mutant animals, the vulva ruptures prior to significant progeny production and the progeny produced rarely develop into viable adults. As a result, we could not maintain a double mutant population for more than one generation and have not further characterized genetic interactions between png-1 and eel-1.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Simultaneous inactivation of EEL-1/HUWE1 and the SKN-1A/Nrf1 pathway causes severe defects in development and physiology that are not caused by reduced proteasome levels.

A) Micrographs showing day 1 adult animals. Red arrows indicate vulval rupture. Vulval rupture occurs frequently in eel-1 skn-1a and png-1; eel-1 double mutants but is not observed in wild type animals or in single mutants. Scale bar 100 μM. B) Quantification of vulval rupture in day 2 adults. Results are shown for n = 4 replicate experiments, > 25 animals were assayed in each replicate experiment. Error bars show mean ± SD. Ns p > 0.05, **** p < 0.0001, indicates comparison to the WT control, ordinary one-way ANOVA with Šídák’s multiple comparisons test. C) Quantification of vulval rupture in day 4 adults. Results are shown for n = 4 replicate experiments, > 25 animals were assayed in each replicate experiment. Error bars show mean ± SD. Ns p > 0.05, **** p < 0.0001, indicates comparison to the WT control, ordinary one-way ANOVA with Šídák’s multiple comparisons test. D) Fluorescence micrographs showing expression of the pbs-5_p::gfp transcriptional reporter of proteasome subunit gene expression. Expression of the reporter is increased in eel-1 mutant animals and reduced in skn-1a mutant animals. Expression of the reported is not further reduced in eel-1 skn-1a double mutants. Images show L4 stage animals. Scale bar 100 μM. E) Quantification of pbs-5_p::gfp reporter expression shown in (D). Expression of the reporter is increased in eel-1 mutant animals compared to the wild type. Expression is reduced in skn-1a mutants. Although expression in eel-1 skn-1a double mutants is reduced compared to wild type, it is slightly increased compared to skn-1a single mutants. Results are shown for n = 15 animals. Error bars show mean ± SD. * p < 0.05, **** p < 0.0001, ordinary one-way ANOVA with Tukey’s multiple comparisons test. F) Fluorescence micrographs showing expression of the pbs-5_p::gfp transcriptional reporter of proteasome subunit gene expression following exposure to BTZ. L4 animals were exposed to 500 nM BTZ (or DMSO control) for 24 hours prior to imaging. Imaged animals are young adults. Expression of the pbs-5_p::gfp reporter is increased in eel-1 mutants as compared to the wild type under both control (DMSO) conditions and following BTZ exposure. Scale bar 100 μM. G) Quantification of pbs-5_p::gfp reporter expression shown in (F). Expression of the pbs-5_p::gfp GFP reporter is higher in eel-1 mutants in control conditions and following BTZ exposure. Results are shown for n = 30 animals for each genotype/condition. Error bars show mean ± SD. ** p < 0.01, **** p < 0.0001, ordinary two-way ANOVA, uncorrected Fisher’s LSD. H) Fluorescence micrographs showing levels of an endogenously tagged proteasome subunit, RPN-9, tagged with mScarlet. Proteasome levels are increased in eel-1 mutants and reduced in skn-1a mutants but are not further reduced in eel-1 skn-1a double mutants. Images show L4 stage animals. Scale bar 100 μM. I) Quantification of levels of the RPN-9 proteasome subunit shown in (H). RPN-9 levels are increased in eel-1 mutants and reduced in skn-1a mutants. RPN-9 levels are unchanged compared to the wild type in eel-1 skn-1a double mutants. Results are shown for n = 15 animals per genotype. Error bars show mean ± SD. Ns p > 0.05, *** p < 0.001, **** p < 0.0001, ordinary one-way ANOVA with Šídák’s multiple comparisons test. J) Fluorescence micrographs showing auxin-induced degradation of degron-GFP. Images show L4 stage animals either before, or 60 minutes after, transfer to plates supplemented with 50 mM auxin. Auxin-induced degradation of degron-GFP is reduced in skn-1a mutants but not in eel-1 mutants. K) Quantification of degron-GFP fluorescence following auxin exposure under the same conditions shown in (J); animals were imaged over a 4-hour period following auxin exposure. Destruction of degron-GFP is delayed in skn-1a mutants but is not affected by eel-1. For each genotype, fluorescence intensity is normalized to the initial (pre-auxin-exposure) value to calculate the percent of fluorescence remaining at each time point. N = 40-45 animals per genotype at each time-point. Error bars show mean ± SD. Trend line shows non-linear regression (one phase decay). degron-GFP half-life in each genotype (95% confidence interval, mins): WT: 28-38; skn-1a: 101-121; eel-1: 28-44; eel-1 skn-1a: 66-88. In all panels of this figure, eel-1 denotes the eel-1[G173STOP] mutant.

https://doi.org/10.1371/journal.pgen.1012273.g002

We also noticed that eel-1 mutants show a marked delay in larval growth and development, a defect that is particularly severe at elevated temperatures (S3 Fig). skn-1a single mutants develop at the same rate as the wild type, but the growth delay caused by eel-1 inactivation is strongly enhanced in double mutants (S3 Fig). These data indicate that eel-1 is required for timely developmental progression, and skn-1a becomes necessary to support development when eel-1 is inactive. We conclude that abnormalities in development and physiology caused by inactivation of the EEL-1 ubiquitin ligase are exacerbated by loss of proteasome regulation by SKN-1A. Given the synergistic impact of eel-1 and skn-1a mutations on UbV-GFP levels, these data suggest that simultaneous loss of EEL-1/HUWE1 and SKN-1A/Nrf1 severely impairs the UPS, causing major developmental and physiological defects.

EEL-1 is not required to maintain adequate proteasome levels

The strong genetic interaction between eel-1 and skn-1a suggests that SKN-1A and EEL-1 could act in separate pathways that independently enhance proteasome levels or activity. To examine whether EEL-1 impacts transcriptional control of proteasome levels, we generated a reporter in which the promoter of the pbs-5 gene drives GFP expression (pbs-5_p::gfp). The pbs-5 gene encodes the β5 subunit of the proteasome. PBS-5/β5 is essential for proteasome assembly and function, is regulated by SKN-1A, and is likely rate-limiting for proteasome function in vivo [14,17,66–69]. We find that loss of eel-1 does not reduce, but rather increases, pbs-5_p::gfp expression and does not further decrease pbs-5 _p::gfp expression in skn-1a mutants (Fig 2D-2E). Exposure to BTZ leads to increased pbs-5_p::gfp expression in wild type animals, as expected, and the extent of activation is higher in eel-1 mutants (Fig 2F-2G). Therefore, the dramatic UbV-GFP accumulation in eel-1 skn-1a double mutants is not explained by reduced expression of proteasome subunit genes.

EEL-1 could impact proteasome levels via a post-transcriptional effect not captured by the pbs-5_p::gfp reporter. We therefore examined animals in which an endogenous proteasome subunit is tagged with mScarlet (rpn-9::mScarlet) [70]. Supporting observations made with the pbs-5_p::gfp transcriptional reporter for proteasome subunit gene expression, mScarlet-tagged RPN-9 protein levels are increased in eel-1 mutant animals compared to the wild type (Fig 2H, 2I). As expected, skn-1a mutants show reduced RPN-9 levels [14], but levels are not further reduced in the double mutants lacking both EEL-1 and SKN-1A (Fig 2H, 2I). Thus, the strong genetic interaction between eel-1 and skn-1a is not explained by a change in proteasome abundance.

EEL-1 is not generally required for protein degradation by the proteasome

We next sought to address the possibility that proteasome function is generally impaired in animals lacking EEL-1. In western blots for total ubiquitin conjugates, we observed an increase in accumulation of ubiquitinated proteins in skn-1a mutant animals, as expected given their reduced proteasome levels. In contrast, we did not observe any increase caused by eel-1 inactivation, either alone or in combination with skn-1a (S4 Fig). Thus, these data suggest that EEL-1 loss does not cause global impairment in the clearance of ubiquitinated proteins by the proteasome.

As an additional test, we examined auxin-induced degradation (AID) of GFP. AID is a heterologous protein degradation system derived from plant hormone signaling and has been used to drive conditional proteasome-dependent protein depletion [71,72]. GFP fused to an auxin-sensitive degron (degron-GFP; expressed in all tissues) is degraded by the proteasome following addition of auxin, triggering interaction with the ubiquitin ligase TIR1. In this system, exposure to 50 mM auxin leads to rapid destruction of degron-GFP, with diminished fluorescence readily detectable after 1 hour [73]. Consistent with reduced proteasome levels causing a general impairment of protein turnover, clearance of degron-GFP is delayed in skn-1a mutant animals, with a roughly 3-fold increase in degron-GFP half-life (Fig 2J, 2K). In contrast, auxin-induced degradation is unchanged in eel-1 mutants. Additionally, there is no further delay in degron-GFP turnover in eel-1 skn-1a double mutants compared to skn-1a (Fig 2J, 2K). We conclude that EEL-1 is not required for AID of GFP. Since AID is proteasome-dependent, this argues that the strong genetic interaction between eel-1 and skn-1a is not explained by a general reduction in all proteasome functions.

EEL-1 ubiquitin ligase activity promotes turnover of UbV-GFP and vulval integrity

We sought to understand the mechanism by which EEL-1 promotes degradation of UbV-GFP. EEL-1/HUWE1 contains a conserved C-terminal HECT domain responsible for transfer of ubiquitin to substrates. The ubiquitination reaction requires a conserved cysteine residue found within the HECT domain [74]. We engineered inactivating mutations in the EEL-1 HECT domain by CRISPR/Cas9 (C4144A and C4144S) (Fig 3A). These mutations dramatically enhance UbV-GFP accumulation in skn-1a mutants (Fig 3B, 3C). We therefore conclude that the E3 ubiquitin ligase activity of EEL-1 is required to promote UbV-GFP turnover, consistent with a conserved role of EEL-1/HUWE1/Tom1 as a ubiquitin ligase in the UFD pathway [58,60].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. EEL-1/HUWE1 ubiquitin ligase activity is required to promote turnover of UbV-GFP and vulval integrity.

A) Schematic of the EEL-1 protein showing domain architecture and locations of mutations studied in this figure. B) Fluorescence micrographs showing that mutation of the EEL-1 active site causes accumulation of UbV-GFP in a skn-1a mutant background. Images show L4 stage animals. Scale bar 100 μM. C) Quantification of UbV-GFP levels shown in (B). All animals are skn-1a mutant; the x-axis label indicates the eel-1 genotype. Results are shown for n ≥ 15 animals per genotype. Error bars show mean ± SD. **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Dunnett’s multiple comparisons test. D) Fluorescence micrographs showing that deletion of the EEL-1 3xUBM module causes accumulation of UbV-GFP in a skn-1a mutant background. Scale bar 100 μM. E) Quantification of UbV-GFP levels shown in (D). All animals are skn-1a mutant; the x-axis label indicates the eel-1 genotype. Results are shown for n = 45 animals per genotype. Bars show mean ± SD. **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, Welch’s two-tailed t test. F) Fluorescence micrographs showing the effects of mutation of the EEL-1 UBA domain on UbV-GFP accumulation in a skn-1a mutant background. The L1446M mutation is a conservative substitution that humanizes EEL-1 at this residue. The L1446V mutation is equivalent to a patient allele, and the L1446A mutation is expected to disrupt ubiquitin binding similarly to L1446V. UbV-GFP levels are slightly increased in L1446V and L1446A mutant animals. Scale bar 100 μM. G) Quantification of UbV-GFP levels shown in (F). All animals are skn-1a mutant; the x-axis label indicates the eel-1 genotype. Results are shown for n = 45 animals per genotype. Error bars show mean ± SD. Ns p > 0.5, ** p < 0.01, **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Dunnett’s multiple comparisons test. H) Quantification of vulval rupture phenotypes caused by skn-1a and eel-1 null mutations in animals carrying the UbV-GFP transgene. The eel-1 mutation tested is eel-1[G173STOP]. Inactivation eel-1 and skn-1a has the same effect on vulval integrity in the UbV-GFP transgenic background as in non-transgenic animals (compare to Fig 2B). Results are shown for n = 3-7 replicate experiments; 25-35 animals were assayed in each replicate experiment. Error bars show mean ± SD. Ns p > 0.05, **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Šídák’s multiple comparisons test. I) Quantification of vulval rupture phenotypes caused by various eel-1 mutations. All animals are skn-1a mutant and carry the UbV-GFP transgene, the x-axis label indicates the eel-1 genotype. Results are shown for n = 3-4 replicate experiments; 25-35 animals were assayed in each replicate experiment. Error bars show mean ± SD. Ns p > 0.05, *** p < 0.001, **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Šídák’s multiple comparisons test. Additional statistical analysis of these data is provided in S1 Table.

https://doi.org/10.1371/journal.pgen.1012273.g003

EEL-1/HUWE1/Tom1 contains conserved ubiquitin-binding domains including a ubiquitin associated (UBA) domain and three repeated ubiquitin binding motifs (UBMs) (Fig 3A). Our screen isolated two alleles (E3064K, E3119K) affecting the UBMs. We noted that the E3064K mutation alters a conserved residue within EEL-1’s second UBM (S5 Fig), near an LP dipeptide required for Ub binding [75,76]. To directly test whether ubiquitin-binding is important for EEL-1 function, we first generated a CRISPR/Cas9 allele, P3059A, which is predicted to disrupt Ub-binding [75]. However, P3059A had no effect on UbV-GFP turnover in skn-1a mutant animals (S6 Fig). This suggested that the 3xUBM module may not be required for EEL-1 function in UbV-GFP turnover, or the mutation does not fully disrupt its function. We therefore generated an in-frame deletion allele that removes all three UBMs (Fig 3A). This eel-1[∆2974–3120] mutation strongly enhances accumulation of UbV-GFP (Fig 3D, 3E), suggesting that robust UbV-GFP turnover requires the 3xUBM Ub-binding domain of EEL-1. Although a similar deletion has been shown to have minimal effects on activity of a fungal ortholog in vitro [46], we cannot exclude the possibility that the 3xUBM deletion may alter the overall structure or activity of EEL-1.

A mutation that disrupts the UBA domain of HUWE1, M1328V, causes a neurodevelopmental syndrome [51,77]. The equivalent residue in C. elegans EEL-1 is L1446; M1328/L1446 lies within the conserved binding interface through which various UBA domains interact with ubiquitin [78–80]. A conserved mode of Ub-binding involving M1328/L1146 is supported by sequence alignment and Alpha Fold structural prediction (S7 Fig). We generated animals harboring the patient-equivalent mutation L1446V and another substitution L1146A, both of which are expected to disrupt Ub binding. We additionally generated a strain in which the UBA domain of C. elegans EEL-1 is ‘humanized’ by a L1446M substitution, which is not expected to disrupt Ub binding. The EEL-1[L1446A] and EEL-1[L1446V] alleles both cause a mild but significant increase in accumulation of UbV-GFP in skn-1a mutant animals, whereas there is no effect of the EEL-1[L1446M] substitution (Fig 3F, 3G). Thus, the UBA domain of EEL-1 contributes to robust UbV-GFP turnover, though the effect is modest compared to deletion of the 3xUBM domain. These data suggest that Ub-binding domains of EEL-1 are involved in ensuring optimal degradation of UbV-GFP, consistent with a model in which EEL-1 binds UbV-GFP via Ub-binding domains and promotes UbV-GFP degradation via ubiquitin-dependent ubiquitin ligase activity [58,60].

We next asked whether the eel-1 alleles affecting the ubiquitin ligase or Ub-binding domains have effects on vulval rupture. We carried out these experiments using animals harboring the UbV-GFP transgene, which we confirmed does not alter the rupture phenotypes caused by eel-1 and/or skn-1a inactivation (Fig 3H; compare to Fig 2B). In a skn-1a mutant background, the C4144A HECT domain mutation caused vulval rupture to a similar extent as the eel-1 null, while mutations affecting ubiquitin binding domains corresponded to their effects on UbV-GFP: only eel-1[∆2974–3120] showed significant vulval rupture (Fig 3I). Thus, EEL-1 ensures turnover of UbV-GFP and promotes vulval integrity via similar mechanisms, suggesting that ubiquitin ligase activity of EEL-1/HUWE1 plays important roles in animal physiology by promoting optimal protein turnover.

The Acid-IDR of EEL-1 is not required for degradation of UbV-GFP or vulval integrity

Our screen isolated a cluster of amino acid substitutions that do not affect ubiquitin binding domains or the HECT domain (E2331K, E2394K, E2412K, E2498K; Fig 4A). These substitutions cluster within a region of the protein that is highly enriched in acidic residues (~50% glutamate or aspartate) and is intrinsically disordered (Fig 4A-4C). We therefore term this region the Acid-IDR domain. The Acid-IDR domain is involved in recognition of orphaned ribosomal subunits and other basic proteins for ubiquitination and degradation [24,46]. The alleles isolated in our screen each introduce a basic amino acid, lysine, into this domain, which could disrupt electrostatic interactions with substrates or regulators. To test the role of the Acid-IDR, we generated in-frame deletion alleles that remove all or part of the domain (Fig 4A). We observed no UbV-GFP accumulation following deletion of the Acid-IDR in a skn-1a mutant background (Fig 4D, 4E). Further, we measured vulval rupture and found no increase compared to control animals (Fig 4F). We conclude that the Acid-IDR of EEL-1 is not required for degradation of UbV-GFP or vulval integrity in skn-1a mutant animals.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. The Acid-IDR domain of EEL-1/HUWE1 is not required for UbV-GFP turnover or vulval integrity.

A) Schematic of the EEL-1 protein showing domain architecture and locations of in-frame deletion mutations studied in this figure. Black arrows indicate locations of amino acid substitutions identified in our screen for increased UbV-GFP accumulation (E2331K, E2394K, E2412K, E2498K). B) The EEL-1 Acid-IDR corresponds to a region of predicted disorder as calculated by AIUPred. C) Amino acid composition of EEL-1, showing enrichment of acidic residues in the Acid-IDR. The percent acidic amino acids (glutamate or aspartate) in a 100 amino acid window around each position is plotted. D) Fluorescence micrographs showing that deletion of the Acid-IDR of EEL-1 does not lead to increased accumulation of UbV-GFP in skn-1a mutants. The G173STOP mutant is included as a positive control. Scale bar 100 μM. E) Quantification of UbV-GFP levels shown in (D). All animals are skn-1a mutants; the x-axis label indicates the eel-1 genotype. Results are shown for n = 45 animals per genotype. Error bars show mean ± SD. Ns p > 0.5, *** p < 0.001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Dunnett’s multiple comparisons test. F) Quantification of vulval rupture phenotypes caused by various eel-1 mutations affecting the Acid-IDR domain. All animals are skn-1a mutants and carry the UbV-GFP transgene; the x-axis label indicates the eel-1 genotype. Results with the eel-1[G173STOP] allele are included for comparison. Results are shown for n = 3-7 replicate experiments; 25-35 animals were assayed in each replicate experiment. Error bars show mean ± SD. Ns p > 0.05, **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Šídák’s multiple comparisons test. Additional statistical analysis of these data is provided in S1 Table.

https://doi.org/10.1371/journal.pgen.1012273.g004

Lysine desert mutations in eel-1 disrupt turnover of UbV-GFP

Interestingly, our unbiased mutagenesis screen identified several amino acid substitutions that introduce lysine (K) residues in two distinct clusters (Fig 5A). The clusters are in or nearby to the Acid-IDR (E2331K, E2394K, E2412K, E2598K) and 3xUBM domains (E3064K, E3119K, R3172K). Comparing the sequence composition and domain organization of EEL-1/HUWE1/Tom1 orthologs across diverse eukaryotes, these domains are positioned within deeply conserved lysine-deficient regions (termed lysine deserts; Figs 5A, S8, S9). Lysine deserts are frequently found in UPS components and prevent their inactivation via accidental ubiquitination [81–86]. We therefore hypothesized that the absence of lysine from these domains is required for EEL-1 function.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Lysine deficiency within EEL-1/HUWE1 lysine deserts is required for degradation of UbV-GFP and its disruption has varied effects on vulval integrity.

A) Schematic showing the EEL-1 protein, showing locations of all lysine residues. The Acid-IDR and 3xUBM domains are located within extensive lysine-free regions. Locations of amino acid substitution mutations affecting these domains are indicated. Black arrows indicate mutations isolated via mutagenesis screening; red arrows indicate mutations generated by CRISPR/Cas9 gene editing. B) Fluorescence micrographs showing that introducing a lysine (K) at position E2412 or E3064 causes increased accumulation of UbV-GFP in the skn-1a mutant background, but that introducing alanine (A) or arginine (R) does not. Scale bar 100 μM. C) Quantification of UbV-GFP levels shown in (B). All animals are skn-1a mutant; the x-axis label indicates the eel-1 genotype. Results are shown for n = 45 animals. Error bars show mean ± SD. Ns p > 0.5, ** p < 0.01, **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Dunnett’s multiple comparisons test. D) Fluorescence micrographs showing that introducing a lysine (K) at position R2567 causes increased accumulation of UbV-GFP in the skn-1a mutant background, but that introducing alanine (A) does not. Scale bar 100 μM. E) Quantification of UbV-GFP levels shown in (D). All animals are skn-1a mutant; the x-axis label indicates the eel-1 genotype. Results are shown for n = 45 animals per genotype. Error bars show mean ± SD. Ns p > 0.5, **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Dunnett’s multiple comparisons test. F) Fluorescence micrographs showing that insertion of an arbitrary lysine-free sequence, GSSRPSTR, at position 2410 does not cause increased accumulation of UbV-GFP in the skn-1a mutant background, but that introducing a similar lysine-containing sequence, GSSRPSTK, does. Scale bar 100 μM. G) Quantification of UbV-GFP levels shown in (F). All animals are skn-1a mutant; the x-axis label indicates the eel-1 genotype. Results are shown for n = 45 animals per genotype. Error bars show mean ± SD. Ns p > 0.5, **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Dunnett’s multiple comparisons test. H) Quantification of vulval rupture phenotypes caused by various eel-1 mutations affecting lysine deserts. All animals are skn-1a mutants and carry the UbV-GFP transgene; the x-axis label indicates the eel-1 genotype. Results with the eel-1[G173STOP] allele are included for comparison. Results are shown for n = 3-7 replicate experiments; 25-35 animals were assayed in each replicate experiment. Error bars show mean ± SD. Ns p > 0.05, **** p < 0.0001, indicates P-value for comparison to the eel-1[WT] control, ordinary one-way ANOVA with Šídák’s multiple comparisons test. Additional statistical analysis of these data is provided in S1 Table.

https://doi.org/10.1371/journal.pgen.1012273.g005

We used CRISPR/Cas9 to recreate two of the mutations isolated in our screen; E2412K (a substitution within the Acid-IDR) and E3064K (within the 3xUBMs) (Fig 4A). As expected, both cause increased accumulation of UbV-GFP in the skn-1a mutant background (Fig 5B, 5C). Importantly, substitution of alanine (A) or arginine (R) at these positions has no effect (Fig 5B, 5C). These data suggest that disruption of EEL-1/HUWE1 function by these mutations is a specific consequence of introducing lysine to EEL-1’s lysine-free regions.

The lysine deficient regions of EEL-1 span hundreds of amino acids. We wondered whether our screen had pinpointed specific locations within the desert that are particularly sensitive to introduction of a lysine. To test this idea, we arbitrarily selected a basic residue within this region, R2567, which is at the periphery of the Acid-IDR domain but near the center of a lysine desert (Fig 5A). Strikingly, in a skn-1a mutant background, R2567K, but not R2567A, causes a strong enhancement of UbV-GFP accumulation (Fig 5D, 5E). The strong effect of introducing a lysine at a randomly selected location suggests that this lysine-deficient region of EEL-1/HUWE1 is intolerant of lysine at many or all positions, as implied by its deep evolutionary conservation.

Our data indicate that ectopic lysine residues in the Acid-IDR of EEL-1 are not tolerated, even though this domain can be deleted without disrupting function. This suggests that the ectopic lysine residues do not simply disrupt an Acid-IDR-dependent function. Rather, EEL-1 function in general requires that this domain is lysine-free. To test this, we inserted a short lysine-free arbitrary sequence, GSSRPSTR, into the Acid-IDR after the 2410^th residue (eel-1[PSTR²⁴¹⁰]), and a similar lysine-bearing variant insertion, GSSRPSTK (eel-1[PSTK²⁴¹⁰]). Strikingly, only the lysine-bearing insertion causes a UbV-GFP accumulation in a skn-1a mutant background (Fig 5F, 5G). Thus, lysine residues, even if engineered to alter a novel sequence not normally present in the protein, still potently disrupt function. This confirms that the need for these regions to be lysine-free includes residues that have no direct relevance to the surrounding domain’s function(s). This suggests that conservation of the lysine-deficient domains is driven by strict selection to be lysine-free, which creates an additional sequence constraint on top of those imposed by the domains’ functions.

Lysine desert mutations in eel-1 variably disrupt vulval integrity

To gauge the physiological impact of manipulations to the lysine desert of EEL-1, we measured their effect on vulval integrity in day 2 adults (Fig 5H). In the skn-1a genetic background, none of the non-lysine substitutions or insertions caused any increase in vulval rupture, consistent with little or no effect on eel-1 function. In contrast, mutations that introduce an ectopic lysine cause vulval rupture in the skn-1a mutant background, albeit to very varying extents. The variant we engineered at a randomly selected position, R2567K, caused the strongest phenotype, equivalent to an eel-1 null allele, but the amino acid substitutions isolated in our mutagenesis screen had less dramatic effects (Fig 5H, S1 Table). The E2412K variant causes an intermediate rupture phenotype, whereas we did not observe elevated vulval rupture in E3064K mutants. However, eel-1[E3064K] skn-1a double mutants did display markedly elevated vulval rupture by day 4 of adulthood (S10 Fig). When analyzed as single mutants, neither eel-1[E2412K] or eel-1[E3064K] caused vulval rupture at day 4 of adulthood, a phenotype we had observed in the eel-1 null, suggesting that these mutations do not fully inactivate EEL-1 (Figs 2C, S10). Thus, the physiological impacts of lysine desert mutations are variable and dependent on the specific location at which ectopic lysine is introduced.

Mutant phenotypes caused by an ectopic lysine residue are rescued by an engineered protein-protein interaction

Lysine deserts are often found in intrinsically disordered domains that are likely to be dynamically exposed to solvent and/or cellular interactors, properties that might predispose adventitious ubiquitination [81]. To investigate the relationship between disorder, solvent exposure, and lysine desert mutations’ functional impact, we inserted a proline-flanked ALFA tag (PSRLEEELRRRLTEP, a lysine-free engineered sequence with high helical propensity [87]) after the 2410^th residue of EEL-1 (eel-1[ALFA²⁴¹⁰]). We generated a second edit in which the same insertion is preceded by a lysine residue (KPSRLEEELRRRLTEP; eel-1[K-ALFA²⁴¹⁰]). We find that insertion of ALFA into the disordered Acid-IDR does not cause UbV-GFP accumulation in the skn-1a mutant background, unless accompanied by insertion of a lysine residue (Fig 6A, 6B). Consistently, in the skn-1a mutant background, eel-1[K-ALFA²⁴¹⁰], but not eel-1[ALFA²⁴¹⁰], causes vulval rupture (Fig 6C). In fact, the extent of vulval rupture caused by the K-ALFA insertion is higher than that caused by the non-helical GSSRPSTK insertion at the same location. Thus, a short helical insertion in the Acid-IDR does not impair function unless accompanied by an ectopic lysine residue. Moreover, the detrimental effect of an ectopic lysine at this location is not blunted, but rather exacerbated, if adjacent to an engineered helical insertion in the disordered Acid-IDR.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Engineered changes to the lysine desert of EEL-1/HUWE1 modulate UbV-GFP turnover and vulval integrity.

A) Fluorescence micrographs showing that insertion of a proline-flanked ALFA tag (“ALFA”’; PSRLEEELRRRLTEP) at position 2410 does not cause increased accumulation of UbV-GFP in the skn-1a mutant background, but the same insertion accompanied by a flanking lysine residue (“K-ALFA”; KPSRLEEELRRRLTEP), does. The UbV-GFP accumulation caused by the K-ALFA insertion is substantially reduced by expression of NbALFA::tagRFP. Scale bar 100 μM. B) Quantification of UbV-GFP levels shown in (A). All animals are skn-1a mutant; the x-axis label indicates the eel-1 genotype and the presence of NbALFA::tagRFP. Results are shown for n = 45 animals per genotype. Error bars show mean ± SD. Ns p > 0.5, **** p < 0.0001, ordinary one-way ANOVA with Tukey’s multiple comparisons test. C) Quantification of vulval rupture phenotypes caused by engineered insertion of an ALFA tag in the EEL-1 lysine desert. All animals are skn-1a mutants and carry the UbV-GFP transgene; the x-axis label indicates the eel-1 genotype and presence of NbALFA::tagRFP. Results are shown for n = 3 replicate experiments; 25-35 animals were assayed in each replicate experiment. Error bars show mean ± SD. Ns p > 0.05, **** p < 0.0001, indicates comparison to the eel-1[WT] control, ordinary one-way ANOVA with Šídák’s multiple comparisons test. Additional statistical analysis of these data is provided in S1 Table. D) Fluorescence micrographs showing that engineered insertion of an ALFA tag at position 2410 of eel-1 leads to degradation of NbALFA::GFP, and that the extent of degradation is limited if the engineered insertion includes an adjacent lysine residue. Scale bar 100 μM. E) Quantification of NbALFA::GFP levels shown in (D). The x-axis label indicates the eel-1 genotype. Results are shown for n = 15 animals per genotype. Error bars show mean ± SD. **** p < 0.0001, ordinary one-way ANOVA with Tukey’s multiple comparisons test.

https://doi.org/10.1371/journal.pgen.1012273.g006

NbALFA, a single chain antibody, binds to the ALFA tag with very high affinity [87]. We wondered whether NbALFA binding would have any effect on the function of EEL-1[ALFA²⁴¹⁰] or EEL-1[K-ALFA²⁴¹⁰]. We hypothesized that tightly associating the Acid-IDR, which is disordered and lysine deficient, to a non-disordered and lysine-containing protein, could interfere with EEL-1[ALFA²⁴¹⁰] function. Alternatively, for EEL-1[K-ALFA²⁴¹⁰], placing a bulky binding partner adjacent to the ectopic lysine could modulate detrimental effects. We engineered animals to express NbALFA fused to tagRFP under the strong rpl-28 promoter from a single-copy genomic insertion. Interestingly, NbALFA::tagRFP has no apparent effect on EEL-1[ALFA²⁴¹⁰] function, as measured by accumulation of UbV-GFP or vulval rupture in a skn-1a mutant background (Fig 6A-6C). In contrast, the defect in UbV-GFP turnover and the vulval rupture of eel-1[K-ALFA²⁴¹⁰] animals is strongly suppressed by NbALFA::tagRFP (Fig 6A-6C). This suggests that the detrimental effects of the ectopic lysine residue in EEL-1[K-ALFA²⁴¹⁰] is dampened by NbALFA binding, likely by restricting the exposure of the ectopic lysine to the cellular milieu.

Ectopic lysine impairs degradation of an engineered EEL-1 substrate

In EEL-1[ALFA²⁴¹⁰]-expressing animals, binding of NbALFA fusion proteins to the ALFA tag would recruit them to the substrate recognition arena of EEL-1 and would likely cause their degradation. We examined the levels of an NbALFA::GFP fusion in eel-1[ALFA²⁴¹⁰] animals and observed a significant reduction, suggesting that the engineered interaction between EEL-1 and NbALFA::GFP causes its ubiquitination and destruction. Interestingly, destabilization of NbALFA::GFP by EEL-1[K-ALFA²⁴¹⁰] is diminished compared to EEL-1[ALFA²⁴¹⁰], although both contain an intact ALFA sequence (Fig 6D, 6E). Thus, introduction of a lysine residue within the Acid-IDR domain interferes with degradation of an engineered substrate that binds adjacent to the ectopic lysine residue. These data suggest that the ectopic lysine residue reduces the ability of EEL-1 to ubiquitinate bound substrates and/or trigger their subsequent destruction.

Lysine desert mutations destabilize EEL-1 in a ligase activity-dependent manner

Lysine deserts can safeguard against accidental ubiquitination leading to self-destruction of UPS components [81–86]. To address whether lysine amino acid substitutions affect the stability of EEL-1, we engineered E2412K and E3064K substitutions into a strain in which endogenous EEL-1 is tagged with HA (HA::EEL-1; the tag does not disrupt function, S11 Fig). Western blots revealed that HA::EEL-1 levels are significantly reduced in E3064K mutants and undetectable in E2412K mutants, demonstrating that these lysine substitutions destabilize EEL-1 (Fig 7A). Το test whether this destabilization is due to autoubiquitination, we tested the effect of an active site mutation (C4144A) on HA::EEL-1[E2412K] stability. We found that the levels of HA::EEL-1[E2412K, C4144A] increased dramatically compared to HA::EEL-1[E2412K] (Fig 7B), suggesting that lysine-introducing mutations destabilize EEL-1 in a manner dependent on its ubiquitin ligase activity. While performing these experiments we found that the active site mutation causes a massive increase in levels of HA::EEL-1 regardless of the presence of a lysine-introducing mutation (Fig 7B). This suggests that catalytically inactivating EEL-1 increases its stability even in the absence of a lysine-introducing mutation. Taken together, these data suggest that autoubiquitination of EEL-1/HUWE1 normally functions to control its protein levels, but excessive autoubiquitination in lysine desert mutants leads to catastrophic loss of EEL-1/HUWE1 stability and function.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Ectopic lysine residues destabilize EEL-1/HUWE1 via autoubiquitination.

A) Western blot showing the effect of lysine-introducing mutations on levels of HA-tagged endogenous EEL-1. The E2412K mutation reduces HA::EEL-1 to undetectable levels, and the E3064K mutation reduces HA::EEL-1 to barely detectable levels. B) Western blot showing the effect of mutation of the HECT-domain active site (AS; C4144A) on the stability of HA-tagged endogenous EEL-1. The AS mutation causes a massive increase in the levels of wild type HA::EEL-1 and of the E2412K mutant.

https://doi.org/10.1371/journal.pgen.1012273.g007

C. elegans strains and maintenance

C. elegans were maintained on standard nematode growth media (NGM) at 20°C and fed E. coli OP50 unless otherwise noted. A list of strains used in this study is provided in S2 Table. Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). Ethane-methyl-sulfonate (EMS) mutagenesis was carried out by treating ~10,000 L4 animals with 47 mM EMS for 4 hours at 20°C. After mutagenesis, large populations of synchronized F2 animals were screened and individuals with increased accumulation of UbV-GFP were isolated. This screen was carried out using skn-1a(mg570); mgIs77[UbV-GFP] and in png-1(ok1654); mgIs77[UbV-GFP] animals.

Whole genome sequencing and analysis

Genomic DNA was prepared using the Gentra Purgene Tissue kit (Qiagen, #158689). Genomic DNA libraries were prepared using the NEBNext genomic DNA library construction kit (New England Biolabs, #E6040). Libraries were sequenced using an Illumina HiSeq instrument. Deep sequencing reads were analyzed using a custom Galaxy workflow adapted from CloudMap [113]. Candidate causative alleles were identified by the isolation of multiple independent mutations affecting the same gene.

Growth rate assay

Embryos were isolated by bleaching of gravid adult animals and then hatched overnight in M9 at 20°C while rotating. The resulting population of synchronized L1 larvae were plated to fresh NGM plates seeded with OP50 and incubated at the desired temperature. After 48–72 hours, plates were imaged using a Leica M165FC equipped with a 910 Leica K5 sCMOS camera and using LAS X software. Animal length was measured using the segmented line tool in ImageJ.

Vulval rupture assay

To measure adult rupture, 30–40 L4 animals were selected at random from mixed stage cultures and transferred to a fresh plate. The animals were monitored for rupture (indicated by intestinal or germline material outside the vulva) after 48 hours (day 2 adults), and again after a further 48 hours (day 4 adults). Throughout the rupture assay, adults were transferred to fresh plates every 2–3 days to separate them from progeny. To image vulval rupture in day 1 adults, L4 animals were selected at random from mixed stage cultures and transferred to a fresh plate. After 24 hours, the animals were imaged using a Leica M165FC equipped with a 910 Leica K5 sCMOS camera and using LAS X software.

Plasmid constructs and transgenesis

All cloning was carried out using the NEBuilder HiFi DNA Assembly kit (New England Biolabs #E2621). The sequence encoding NbALFA was codon optimized for expression in C. elegans using the transgenebuilder tool (https://www.wormbuilder.org/transgenebuilder/) and synthesized to include a single intron (from the rps-0 gene) and a C-terminal in-frame linker sequence (STSGGSGGTGGSS). Gene synthesis was carried out by Twist Biosciences. Gibson assembly was used to generate plasmids in which NbALFA fusion proteins are expressed under control of the ubiquitous rpl-28 promoter (605 bp immediately upstream of the rpl-28 ORF) and are followed by the tbb-2 3’UTR (376 bp immediately downstream of the tbb-2 stop codon). These constructs were inserted to the pNL43 backbone, to allow generation of transgenics by miniMos [15,114]. pNL559 (Rpl-28p::NbALFA::tagRFP::tbb-2) and pNL549 (Rpl-28p::NbALFA::GFP::tbb-2), were used to generate transgenic (miniMos insertion) animals harboring single-copy transgenes driving ubiquitous expression of NbALFA::tagRFP and NbALFA:tagRFP respectively [114].

The pbs-5_p::gfp reporter was generated by placing the promoter of CEOP1752 (850 bp upstream of the first open reading frame of the operon, K05C4.2 (pbs-5 is the second gene in the operon), upstream of a fragment containing the GFP coding sequence and tbb-2 3’UTR (376 bp immediately downstream of the tbb-2 stop codon), and inserted to the pBlueScript backbone, to generate pNL136 (pbs-5_p::GFP). High-copy-number arrays containing pNL136 were integrated to chromosome II via CRISPR/Cas9 [115,116].

Genome Modification by CRISPR/Cas9

CRISPR/Cas9 genome editing was carried out through microinjection of guide RNA/Cas9 ribonucleoprotein (RNP) complexes (IDT #1081058, #1072532, and custom crRNA) along with single-stranded oligonucleotides (IDT) as homology-directed repair (HDR) templates [117,118]. Sequences of all guide RNAs and HDR oligos are provided in S3 Table. All CRISPR edits were identified by diagnostic PCR and confirmed by sequencing.

Microscopy

Brightfield and fluorescence images of whole animals (UbV-GFP, pbs-5_p::gfp, rpn-9::mScarlet, degron-GFP, NbALFA::GFP) were collected on a Leica M165FC equipped with a 910 Leica K5 sCMOS camera and using LAS X software. For fluorescence imaging, worms were immobilized using sodium azide and mounted on 2% agarose pads. All images were processed and analyzed using ImageJ/Fiji software. Images shown within the same figure panel were collected using the same exposure time and were then processed identically. To quantify accumulation of UbV-GFP, the mean pixel intensity in the anterior intestinal cells was measured. To quantify pbs-5_p::GFP reporter expression, rpn-9::mScarlet levels, degron-GFP levels, and NbALFA::GFP levels, the mean pixel intensity was measured using a polygon selection around each animal.

Bortezomib treatment

NGM plates containing the desired concentration of bortezomib (LC Laboratories, #B1408) were prepared by directly applying bortezomib solution to NGM plates seeded with OP50 and leaving the plates to dry for 1–2 hours. Control plates supplemented with DMSO were prepared in parallel. L4 stage animals to were shifted to bortezomib or control DMSO-supplemented plates and imaged after 24 hours.

Western blot

For UbV-GFP western blots: approximately 60 L4 stage animals were placed on regular NGM plates, plates supplemented with 1 μM bortezomib, or control DMSO plates, and incubated overnight at 20°C. The following day, 50 day 1 adults were transferred into microcentrifuge tubes containing 30 μL M9. Then, 15 μL of Bolt LDS sample buffer (Invitrogen; #B0008) was added to each tube. Samples were frozen at -80°C for 15 minutes, followed by incubation at 70°C for 10 minutes. Samples were then vortexed at medium speed for 5 minutes. Then, 4.5 μL of Bond- Breaker TCEP solution (Thermo Scientific; #77720) was added, and samples were incubated at room temperature for 5 minutes. 20–30 μL of each sample was loaded onto protein gels. For HA-tagged EEL-1 and Ub western blots: 50 day 1 adults were transferred into microcentrifuge tubes containing 20 μL M9. Then, 15 μL of NuPAGE LDS sample buffer (Invitrogen; #NP0007) was added to each tube. Samples were frozen at -80°C for 15 minutes, followed by incubation at 70°C for 10 minutes. Samples were then vortexed at medium speed for 5 minutes. Then, 3.5 μL of Bond-Breaker TCEP solution (Thermo Scientific; #77720) was added, and samples were incubated at room temperature for 5 minutes. 25 μL of each sample was loaded onto protein gels.

SDS-PAGE was performed using Bolt 4–12% Bis–Tris Plus gels (Invitrogen, #NW04122BOX) and MOPS SDS running buffer (Invitrogen; #B0001; UbV-GFP, Ub) or NuPAGETM 3–8% Tris-Acetate gels (Invitrogen; #EA0375BOX) and Tris-Acetate SDS running buffer (Invitrogen, #LA0041; HA-tagged EEL-1). Western blotting was performed using iBlot 2 nitrocellulose mini stacks (Invitrogen; #IB23002) or iBlot 2NC regular stacks (Invitrogen; IB23001), according to the manufacturer’s instructions. Membranes were blocked using 5% milk or 4% bovine serum albumin (BSA; Gold-Bio; #A-420–50). The following antibodies were used: anti-GFP mouse monoclonal antibody (SIGMA; 11814460001; 1:1,000 dilution), anti-HA tag (F7) mouse monoclonal antibody (Santa Cruz; #sc-7392; 1:1,000 dilution), anti-ubiquitin mouse monoclonal antibody (Santa Cruz #sc-8017), anti-α-tubulin mouse monoclonal antibody (SIGMA, #T6074; 1:10,000 dilution), and goat anti-mouse IgG antibody, HRP conjugate (Millipore, #12–349; 1:2,500 dilution). The SuperSignal West Pico PLUS chemiluminescent substrate (Thermo Scientific, #34577) or SuperSignal West Dura extended duration substrate (Thermo; #34075) were used for detection of the HRP signal.

Auxin-induced degradation

NGM plates containing 50 mM auxin (α-Napthaleneacetic acid, Phytotech Labs #N610) were prepared by directly adding auxin solution to NGM plates seeded with OP50 and leaving plates to dry for 1–2 hours. Control (non-auxin-exposed) L4 animals were imaged immediately prior to beginning the auxin treatment. L4 stage animals were shifted to auxin-supplemented plates and imaged after 1, 2 and 4 hours.

Protein sequence analysis

EEL-1/HUWE1/Tom1 orthologs were identified in the PANTHER database subfamily: E3 UBIQUITIN-PROTEIN LIGASE HUWE1 (PTHR11254:SF67) [119]. Protein sequences of EEL-1 and orthologs were retrieved from Uniprot [120]. Multiple sequence alignment was carried out using Clustal Omega with default settings [121]. AIUPred was used to predict intrinsically disordered regions [122]. Sequence composition was analyzed in excel. Structure prediction was carried out using Alpha Fold 3 and structure predictions were visualized using UCSF ChimeraX [123,124].

Statistical analysis

Statistical tests were carried out using GraphPad Prism. All biological replicates were performed using independent populations of animals. All numerical data used in graphs and statistical tests are provided in S1 Data File.