IP6-assisted CSN-COP1 competition regulates a CRL4-ETV5 proteolytic checkpoint to safeguard glucose-induced insulin secretion

Lin, Hong; Yan, Yuan; Luo, Yifan; So, Wing Yan; Wei, Xiayun; Zhang, Xiaozhe; Yang, Xiaoli; Zhang, Jun; Su, Yang; Yang, Xiuyan; Zhang, Bobo; Zhang, Kangjun; Jiang, Nan; Chow, Billy Kwok Chong; Han, Weiping; Wang, Fengchao; Rao, Feng

doi:10.1038/s41467-021-22941-3

Download PDF

Article
Open access
Published: 28 April 2021

IP₆-assisted CSN-COP1 competition regulates a CRL4-ETV5 proteolytic checkpoint to safeguard glucose-induced insulin secretion

Hong Lin¹^na1,
Yuan Yan¹^na1,
Yifan Luo^1,2^na1,
Wing Yan So³^na1,
Xiayun Wei¹,
Xiaozhe Zhang¹,
Xiaoli Yang¹,
Jun Zhang¹,
Yang Su¹,
Xiuyan Yang¹,
Bobo Zhang¹,
Kangjun Zhang⁴,
Nan Jiang⁴,
Billy Kwok Chong Chow ORCID: orcid.org/0000-0003-3390-0307²,
Weiping Han³,
Fengchao Wang ORCID: orcid.org/0000-0002-3595-2859⁵ &
…
Feng Rao ORCID: orcid.org/0000-0002-6038-6406¹

Nature Communications volume 12, Article number: 2461 (2021) Cite this article

4193 Accesses
11 Citations
3 Altmetric
Metrics details

Subjects

Abstract

COP1 and COP9 signalosome (CSN) are the substrate receptor and deneddylase of CRL4 E3 ligase, respectively. How they functionally interact remains unclear. Here, we uncover COP1–CSN antagonism during glucose-induced insulin secretion. Heterozygous Csn2^WT/K70E mice with partially disrupted binding of IP₆, a CSN cofactor, display congenital hyperinsulinism and insulin resistance. This is due to increased Cul4 neddylation, CRL4^COP1 E3 assembly, and ubiquitylation of ETV5, an obesity-associated transcriptional suppressor of insulin secretion. Hyperglycemia reciprocally regulates CRL4-CSN versus CRL4^COP1 assembly to promote ETV5 degradation. Excessive ETV5 degradation is a hallmark of Csn2^WT/K70E, high-fat diet-treated, and ob/ob mice. The CRL neddylation inhibitor Pevonedistat/MLN4924 stabilizes ETV5 and remediates the hyperinsulinemia and obesity/diabetes phenotypes of these mice. These observations were extended to human islets and EndoC-βH1 cells. Thus, a CRL4^COP1-ETV5 proteolytic checkpoint licensing GSIS is safeguarded by IP₆-assisted CSN-COP1 competition. Deregulation of the IP₆-CSN-CRL4^COP1-ETV5 axis underlies hyperinsulinemia and can be intervened to reduce obesity and diabetic risk.

Lysosomal degradation of newly formed insulin granules contributes to β cell failure in diabetes

Article Open access 25 July 2019

Adrien Pasquier, Kevin Vivot, … Romeo Ricci

TRIM24 is an insulin-responsive regulator of P-bodies

Article Open access 08 July 2022

Wen Wei, Qiaoli Chen, … Shuai Chen

UDP-glucose, cereblon-dependent proinsulin degrader

Article Open access 26 August 2022

Jaeyong Cho, Atsushi Miyagawa, … Takeshi Imai

Introduction

The cullin-rING ubiquitin ligases (CRLs) are the largest family of E3 enzymes that work together with ubiquitin-activating E1 and ubiquitin-conjugating E2 enzymes to ubiquitylate protein substrates for proteasomal degradation or signal transduction. CRLs are composed of a scaffold Cullin protein (Cul1, 2,-3, 4A/B, 5, or 7), an E2-recruiting RING protein Rbx1/2 (also called Roc1/2), and Cullin-specific adaptors (e.g., DDB1 for CRL4) and/or substrate receptors (e.g., DDB1–CUL4-associated factors (DCAFs) for CRL4)^1,2. While CRLs ubiquitylate numerous proteins involved in diverse biological processes and are emerging therapeutic targets and tools in cancer research^3,4, their roles in glucose homeostasis and metabolic diseases are less understood⁵.

Glucose homeostasis is controlled by circulating insulin levels and insulin signaling. Insulin levels, in particular, is under tight regulation, with hypersecretion/hyposecretion and dysregulated clearance all posing diabetes risk⁵. While several mediators of insulin signaling are CRL substrates^6,7, direct involvement of CRLs in insulin secretion has not been reported. The COP1 protein, first identified in Arabidopsis based on the constitutive photomorphogenesis (COP) phenotype of its mutants, is a candidate CRL4 substrate receptor that promotes murine insulin secretion by ubiquitylating the ETV family of transcription factors^8,9, among which, ETV5 is strongly associated with obesity and T2D according to genome-wide association studies (GWAS)^10,11,12. Such GWAS link often relates to islet function¹³. However, it remains unknown if any physiologic stimuli regulate insulin release via COP1-mediated ETV degradation. Moreover, since COP1 contains a RING domain and can serve as an E3 ligase independent of CRL4, whether CRL4 is required for COP1’s function in insulin secretion remains to be validated.

The E3 ligase activity of CRL is regulated by cycles of neddylation, an ubiquitin-like Nedd8 modification required for CRL activity^14,15,16. The eight-subunit COP9 Signalosome (CSN) complex, another mediator of Arabidopsis photomorphogenesis¹⁷, is a deneddylase that catalyzes Cullin deneddylation while also sterically inhibiting substrate access^18,19,20,21. We have identified inositol hexakisphosphate (IP₆), an abundant cellular metabolite²², as a “glue”-like molecule facilitating CRL4–CSN complex assembly²³, and also elucidated the structural mechanisms explaining the biochemical role of IP₆ at CRL–CSN interface^24,25. IP₆ primarily binds to a basic pocket in CSN2 centered on Lys70, but also makes direct contact with CRL to bridge CRL–CSN interactions²⁵. Based on these mechanistic insights, we generated IP₆-binding-deficient CSN2-K70E mutant mice. The early embryonic lethality of homozygous Csn2^K70E/K70E mutants underscores the essentiality of IP₆ as a CSN cofactor, but also hampered the elucidation of physiological contexts, whereby CRL is regulated by the IP₆ metabolite.

Here, by searching for haploinsufficient phenotypes in the viable heterozygous Csn2^WT/K70E mice, we uncovered a glucose-responsive IP₆–CSN–CRL4^COP1–ETV5 axis in insulin secretion control. The anti-obese/diabetic implications of this regulatory axis are further validated by preclinical testing of the CRL neddylation inhibitor MLN4924 (also known as pevonedistat²⁶) in high-fat diet (HFD) and ob/ob mouse disease models. Mechanistically, COP1 is an unusual CRL substrate receptor that directly competes with CSN. IP₆ binding to CSN modulates mutually exclusive assembly of the active CRL4^COP1 E3 ligase and the inert CRL4–CSN complex, with such CRL4 toggling regulated by glycemic status. Our study therefore extends the evolutionarily conserved interaction between COP1 and COP9 from photomorphogenesis in plants to glucose homeostasis in mammals, two fundamental energy coping processes.

Results

Heterozygous Csn2 ^WT/K70E mice with partial abolishment of the IP₆ cofactor display congenital hyperinsulinism, and uncoupling between glucose and insulin tolerance

Using a combination of structural and biochemical approaches, we recently showed that IP₆ promotes CRL deneddylation by acting as a CSN cofactor via direct interaction with a surface patch on CSN2 centered on Lys70 (ref. ²⁵). To understand the mammalian physiology of IP₆-mediated CRL–CSN complex formation, we attempted to generate a mouse line carrying the CSN2-K70E mutation. The K70E mutation selectively abolishes IP₆-augmented CSN–CRL interaction without disrupting CSN2 overall structure or its integration into CSN holo-complex²⁵, ensuring that potential phenotypes originate specifically from loss of the CSN cofactor IP₆. Homozygous Csn2^K70E/K70E embryos die at the stage of peri-implantation (Table 1), similar to the Csn2 null mice²⁷. We reasoned that partial CSN dysfunction better mimics tolerable CSN inhibition²⁸, and therefore searched for haploinsufficient phenotypes in the viable Csn2^WT/K70E heterozygous mice.

Table 1 Genotypes of offspring from Csn2^WT/K70E intercrosses.

Full size table

The Csn2^WT/K70E heterozygous mice are born in Mendelian ratio from WT-Het intercross and weigh similarly to wild-type littermates before adulthood, but are significantly heavier afterward (Fig. 1a), despite less food intake (Fig. S1a). This leads us to examine their metabolic parameters. Compared to 2.5-month wild-type littermates, the Csn2^WT/K70E mice have normal levels of fed blood glucose, free fatty acid, and triglyceride (Fig. 1b and Fig. S1b, c), but display fasting hypoglycemia (Fig. 1b), hyperinsulinemia (Fig. 1c), and elevated levels of the degradation-resistant proinsulin cleavage product: C-peptide (Fig. 1d), indicating enhanced insulin secretion. When challenged with tolerance tests, the Csn2^WT/K70E mice are more glucose tolerant (Fig. 1e), but are mildly resistant to insulin (Fig. 1f), with more pronounced insulin resistance at 4-month age (Fig. S1d). This rare uncoupling between glucose tolerance and insulin sensitivity further suggests that hyperinsulinemia is not driven primarily by insulin resistance, which would also cause impaired glucose tolerance and hyperglycemia⁵, phenotypes not found in Csn2^WT/K70E mice. Treatment with the insulin sensitizer metformin, following a previous protocol²⁹, significantly improves insulin sensitivity of Csn2^WT/K70E mice (Fig. S1e), but hyperinsulinemia remains (Fig. S1f), again suggesting that hyperinsulinemia is unlikely to be downstream of insulin resistance in Csn2^WT/K70E mice.

**Fig. 1: Enhanced insulin secretion and congenital hyperinsulinemia in *Csn2*^WT/K70E (Het) mice.**

Chronic exposure to excess insulin is known to induce insulin resistance^{30,31,32,33,34}, we therefore investigated whether insulin oversecretion underlies the observed metabolic derangements in Csn2^WT/K70E mice. Decreasing insulin secretion by the β-cell toxin streptozocin eliminates the differences in insulin sensitivity (Fig. 1g), serum insulin levels (Fig. 1h), and body weight (Fig. S1g) between wild-type and Csn2^WT/K70E mice, suggesting that these differences originate from pancreatic β cells. Importantly, both Csn2^WT/K70E mice (Fig. 1i) and their pancreatic islets (Fig. 1j) secrete more insulin under basal conditions, or when stimulated with glucose or another insulin secretagogues (l-arginine; Fig. 1k). The first and second phases of glucose-stimulated islet insulin secretion are both augmented (Fig. 1j). Together, these data suggest that, due at least in part to enhanced insulin secretion, adult Csn2^WT/K70E mice develop early signs of type II diabetes, including congenital hyperinsulinism, insulin resistance, and obesity, though not yet diabetic.

Insulin production and vesicle release are increased in Csn2 ^WT/K70E mice

We next examined which molecular events in the islet insulin production and secretion pathway are altered in Csn2^WT/K70E mice. Histological, immunohistochemical, and cytoplasmic Ca²⁺ recording experiments showed that islet number (Fig. 2a), size (Fig. 2b), α- and β-cell percentages (Fig. 2c), and glucose- or KCl-induced Ca²⁺ influx (Fig. 2d, e) are comparable between wild-type and Csn2^WT/K70E mice. In contrast, total islet insulin content, assayed by immunoflurescence (Fig. 2c) and ELISA (Fig. 2f), are higher in Csn2^WT/K70E islets, which is further supported by increased levels of Ins1 and Ins2 mRNA (Fig. 2g). Electron microscopic analysis reveal significantly more docked insulin granules in Csn2^WT/K70E islets (Fig. 2h), suggesting augmented vesicle release. Consistently, insulin secretion measured as percentage of total insulin content is also higher in Csn2^WT/K70E islets (Fig. 2i). Together, these data suggest that both insulin biosynthesis and insulin vesicle release are augmented in Csn2^WT/K70E mice, resulting in hyperinsulinemia.

**Fig. 2: Increased insulin production and vesicle release contribute to hyperinsulinemia in *Csn2*^WT/K70E (Het) mice.**

CRL4^COP1-mediated ubiquitylation and degradation relieves the ETV5 transcriptional “brake” to enhance insulin secretion in Csn2 ^WT/K70E mice

CSN deneddylates and inactivates CRL in an IP₆-dependent manner^23,25. To elucidate the molecular mechanism underlying enhanced insulin secretion in Csn2^WT/K70E mice, we first verified that this mutation indeed disrupts CRL–CSN interaction. Consistent with such an expectation, CSN3 immunoprecipitation of Cul4 is significantly diminished in Csn2^WT/K70E pancreas (Fig. S2a). Furthermore, IP6 addition increases CSN3-Cul4 co-immunoprecipitation in both wild-type and CSN2^WT/K70E mice, suggesting that the remaining wild-type allele still responds to the bridging effect of IP₆. These data support our proposition that the Csn2^WT/K70E mice display partial loss of the IP₆ cofactor and consequent CRL regulation.

To validate that enhanced insulin secretion in Csn2^WT/K70E mice is due to diminished CRL regulation, we examined the effects of CRL re-inhibition by using the well-characterized CRL neddylation inhibitor: pevonedistate/MLN4924 (ref. ^26,35). While MLN4924 treatment does not affect depolarization-induced Ca²⁺ influx (Fig. S2b), it markedly suppresses basal and glucose-stimulated islet insulin secretion, with more pronounced inhibition under high glucose (Fig. 3a). Weak but significant GSIS is still observed in the presence of MLN4924, suggesting that the islets retain glucose responsiveness to some extent (Fig. 3a). However, MLN4924 eliminates the difference between wild-type and Csn2^WT/K70E islets (Fig. 3a), suggesting that CRL disinhibition differentiates wild-type and Csn2^WT/K70E islets. When applied in mice, MLN4924 restores serum insulin levels (Fig. 3b) and body weight (Fig. 3c) of the Csn2^WT/K70E mice to normal levels over time, without affecting the insulin levels and body weight of wild-type littermates (Fig. 3b, c), nor their food intake (Fig. S2c). Thus, CRL re-inhibition reverses congenital hyperinsulinism and obesity of Csn2^WT/K70E mice.

**Fig. 3: Activation of the CRL4^COP1–ETV5 degradation axis underlies glucose-induced insulin secretion and is constitutively augmented in *Csn2*^WT/K70E (Het) mice.**

To identify specific CRL(s) that is dysregulated in Csn2^WT/K70E mice, we then examined the neddylation status of Cullins. Among all the canonical Cullins screened (Cul1–5), only Cul4A/B display consistently increased neddylation ratio in Csn2^WT/K70E pancreas (Fig. 3d). This data, together with earlier findings that Cul4 immunoprecipitates are particularly enriched for CSN subunits^36,37, suggests that partial loss of the IP₆ “glue” primarily releases CRL4 from CSN sequestration.

To identify downstream CRL4 modules, we next employed an in-islet viral knockdown assay to screen a collection of putative and verified CRL4 substrate receptors for insulin secretion promoter(s). This screen reveals COP1 as a prominent positive regulator of insulin secretion (Fig. 3e). Consistently, COP1 knockdown markedly diminishes islet insulin secretion and eliminates the differences between wild-type and Csn2^WT/K70E islets (basal or glucose-stimulated; Fig. 3f). COP1 can function as an independent E3 ligase, or as a CRL4 substrate receptor by working together with DET1 (refs. ^34,35). In the pancreas, COP1 is islet-enriched and promotes insulin secretion by degrading transcription factors ETV1/4/5 (ref. ⁸), a pathway compromised in low-exocytotic β cells from T2D patients³⁸. Depleting COP1 or DET1 leads to the accumulation of ETV5 and, to a mild extent, ETV4 in MIN6 insulinoma cells (Fig. 3g), whereas co-expressing COP1 and DET1 diminishes ETV5 levels (Fig. 3h), suggesting that CRL4^COP1/DET1 mediates ETV5 degradation, which is further supported by ETV5 stabilization upon MLN4924 treatment (Fig. 3i). Levels of ETV5, but not ETV1, ETV4, or other COP1-indpendent CRL4 substrates (CSB, Cdt1), is significant downregulated in Csn2^WT/K70E pancreas (Fig. 3d) and in isolated Csn2^WT/K70E islets (Fig. 3i), but can be restored to the same level as in wild-type islets upon MLN4924 treatment (Fig. 3i). The diminished ETV5 levels in Csn2^WT/K70E mice correlates with enhanced ETV5 ubiquitylation (Fig. 3j). CSN2 knockdown similarly enhances ETV5 ubiquitylation (Fig. 3k). These data suggest that partial abolishment of CSN sequestration activates a CRL4^COP1–ETV5 degradation axis in Csn2^WT/K70E pancreas.

ETV5 is linked to body mass index and other obesity-related traits in GWAS^10,11,12. Given that obesity epidemiologically correlates with insulin levels^39,40,41, yet both positive and negative regulation of insulin secretion has been proposed for ETV5 (refs. ^8,42), we reinvestigated its precise role. Depleting Etv5 enhances basal and glucose-stimulated insulin secretion in islets (Fig. 3l), whereas overexpressing Etv5 represses basal and glucose-stimulated insulin secretion (Fig. 3m). Mechanistically, we examined two previously identified ETV transcriptional targets that are exocytosis regulators: synaptotagmin-like 3 (Sytl3) and exocyst 6 (Exoc6)⁸. Depleting Etv5 decreases Sytl3 and Exoc6 transcript levels in islets, while increasing that of Ins1 (Fig. 3n). Conversely, overexpressing Etv5 upregulates the mRNA levels of Sytl3 and Exoc6, but not Ins1 (Fig. 3o). Both Sytl3 and Exoc6 exert anti-exocytotic effects, since their overexpression significantly inhibits insulin secretion from INS1 cells, with their combined expression having the most pronounced inhibitory effects (Fig. 3p). Further chromatin immunoprecipitation (ChIP) experiments validated ETV5 binding to Sytl3 and Exoc6 promoter regions (Fig. S2d). These data, together with diminished mRNA (Fig. 2g) and protein (Fig. 3d) levels of Sytl3 and Exoc6 in Csn2^WT/K70E pancreas, supports an inhibitory ETV5-Sytl3/Exoc6 transcriptional axis in controlling insulin secretion⁸, which is compromised in Csn2^WT/K70E mice due to CRL4^COP1-mediated ETV5 degradation.

COP1 outcompetes CSN to assemble the CRL4^COP1 E3 ligase in Csn2 ^WT/K70E mice

The above data suggest that CSN is a haploinsufficient regulator of CRL4^COP1. However, it remains unclear why CRL4^COP1 seems particularly sensitive to CSN regulation. CSN can bind CRL4 loaded with classical WD40 domain-containing substrate receptors known as DCAFs (e.g., DDB2)^19,43, but ubiquitylation substrates sterically compete with CSN¹⁹. COP1 is an unorthodox substrate receptor whose binding to CRL4 is indirectly mediated by the DCAF DET1 (refs. ^44,45). Consistently, DET1 knockdown abolishes COP1 interaction with Cul4 (Fig. S3a). Using truncation and deletion variant mapping, we identified the N-terminal (aa 1–110) and C-terminal (aa 287–450) regions of DET1 mediates binding to the CRL4 adaptor DDB1 (Fig. S3c) and COP1 (Fig. S3d, e), respectively. Furthermore, exon 7 (aa 277–296) of COP1 mediates its binding to DET1 and indirectly DDB1 (Fig. S3b), but not the ubiquitylation substrate ETV5, consistent with a prior report⁴⁵; When these information were integrated to build a structural model of CRL4^COP1, potential steric overlap between COP1 and CSN in the cryo-EM structure of CRL4–CSN^19,43 was noted (Fig. 4a). We therefore hypothesized that, similar to ubiquitylation substrates, COP1 competes with CSN for mutually exclusive CRL4 binding. Consistent with this notion, CSN can pull-down DET1, but not COP1 (Fig. 4b). Rather, co-expressing COP1 with DET1 significantly diminishes DET1 and Cul4 co-immunoprecipitation with CSN, suggesting that CRL4^COP1 and CRL4–CSN are mutually exclusive complexes because of steric incompatibility between COP1 and CSN.

**Fig. 4: COP1 outcompetes CSN to assemble the CRL4^COP1 E3 ligase for ETV5 degradation in *Csn2*^WT/K70E (Het) mice and in response to nutrients.**

We then reasoned that, in heterozygous Csn2^WT/K70E mice with weakened CRL4–CSN interaction, the COP1–CSN competition is likely altered to favor CRL4^COP1 complex assembly. Consistent with this notion, COP1 co-immunoprecipitation with Cul4A and Cul4B is greatly increased in Csn2^WT/K70E pancreas (Fig. 4c, d). Importantly, addition of the metabolite IP₆ into lysates dramatically diminishes COP1 immunoprecipitation with Cul4 (Fig. 4e), suggesting that IP₆ delimits CRL4–COP1 complex formation. Importantly, we found that nutrient availability reciprocally regulates the formation of CLR4^COP1 vs CRL4–CSN complexes. Thus, glucose withdrawal increases CRL4–CSN complex formation and decreases CRL4^COP1 assembly, both effects reversed by refeeding glucose containing medium (Fig. 4f, g). Together, these results suggest that direct CSN competition with COP1 for CRL4 requires the IP₆ cofactor, whose partial absence in Csn2^WT/K70E mice leads to CRL4^COP1 assembly.

The CSN–CRL4^COP1–ETV5 axis is a glucose-regulated checkpoint for nutrient-induced insulin secretion

Given the pleiotropic control of insulin homeostasis by ETV5, we investigated its physiological regulation during nutrient-stimulated insulin secretion. Glucose addition to medium promotes CRL4^COP1 complex assembly in β cells (Fig. 4f), whereas refeeding fasted mice promotes CRL4^COP1 complex assembly in pancreas (Fig. 5a), suggesting nutrient regulation of a CRL4^COP1-mediated degradation pathway. Interestingly, pancreatic levels of ETV5, but not ETV4, time-dependently accumulates upon fasting (Fig. 5b), and diminishes upon refeeding (Fig. S4a), with near-complete ETV5 degradation after 1-h refeeding in whole pancreas and in islets, evidenced by western blotting (Fig. 5c) and immunostaining (Fig. 5d), respectively. This feeding-induced ETV5 degradation is not observed in other metabolic organs, such as liver or the white adipose tissue (Fig. S4b, c). Similar to fasting in mice, glucose withdrawal stabilizes islet ETV5 (Fig. S4d). In contrast, direct glucose application depletes ETV5 in mice (Fig. 5d) and in isolated pancreatic islets in a time-dependent manner (Fig. 5e and Fig. S4e), which is accompanied by decreased transcription of Sytl3 and Exoc6 (Fig. S4f). These data suggest that ETV5 is degraded in a glucose-dependent manner to relieve the inhibition on insulin secretion.

**Fig. 5: ETV5 as a CRL4^COP1 regulated physiologic checkpoint for nutrient/glucose-induced insulin secretion.**

We then investigated whether glucose-induced ETV5 degradation is mediated by CRL4. The CRL inhibitor MLN4924 completely reverses feeding-induced ETV5 depletion, without altering fasting ETV5 levels, suggesting involvement of CRL4 (Fig. 5g). Consistently, elevating glucose concentrations increases Cul4 neddylation in islets (Fig. 5h). Such glucose-dependent changes in Cul4 neddylation could be attributed to diminished CSN-mediated deneddylation, since CRL–CSN interaction is stronger under fasting and is diminished upon glucose supplementation (Fig. 4f), although total levels of IP₆ remain unchanged between fasted and fed mice (Fig. S4g). Together, these data together suggest that glucose-induced assembly of CRL4^COP1 destabilizes ETV5.

In summary, the data presented so far suggest that CRL4^COP1 liberation from IP₆-dependent CSN sequestration and consequent ETV5 degradation relieves the checkpoint for nutrient (glucose)-induced insulin secretion, a process constitutively augmented in Csn2^WT/K70E mice.

The CRL4^COP1–ETV5 axis in human islets and β cells

Given the GWAS-identified ETV5 association with diabetes traits^10,11,12, and the compromised COP1 activity in β cells from T2D patients³⁸, we also examined the relevance of the CRL4^COP1–ETV5 pathway in human using isolated islets and the EndoC-βH1 human β cells⁴⁶. CRL inhibition and COP1 knockdown both decrease insulin secretion from isolated islets and EndoC-βH1 cells, more prominently under high glucose conditions, when CRL4^COP1 is assembled and active (Fig. 6a, b, e, f). By contrast, ETV5 knockdown enhances insulin secretion (Fig. 6c, g). Consistently, GSIS is accompanied by the depletion of ETV5 levels from isolated islets and EndoC-βH1 cells (Fig. 6d,h). These data suggest that the CRL4^COP1–ETV5 axis also operates during insulin secretion in human.

**Fig. 6: The CRL4^COP1–ETV5 axis operates in human islets and β cells.**

CRL neddylation inhibition prevents ETV5 degradation and hyperinsulinemia in diet- and leptin deficiency-induced obesity/diabetes

One emerging approach to treat obesity and insulin resistance is mild suppression of hyperinsulinemia⁴⁷. The CRL inhibitor MLN4924 specifically inhibits hyperinsulinemia in Csn2^WT/K70E mice, without significantly affecting serum insulin levels in wild-type littermates (Fig. 3b). Given that the pancreatic CRL4^COP1–ETV5 axis is hyperactive not only in Csn2^WT/K70E mice (Fig. 3), but also in HFD-fed mice (Fig. 7a), we examined whether CRL inhibition could rescue disease progression in this obesity/diabetes model. Indeed, MLN4924 attenuates HFD-induced ETV5 degradation (Fig. 7a), obesity (Fig. 7b), hyperinsulinemia (Fig. 7c), hyperglycemia (Fig. 7d), glucose intolerance (Fig. S5b), and insulin resistance (Fig. S5c). In comparison, in mice fed with normal chow diet (NCD), levels of ETV5 (Fig. S5a), blood glucose (Fig. 7c), serum insulin (Fig. 7d), glucose clearance (Fig. S5b), and insulin sensitivity (Fig. S5c) are not significantly affected by the same regime of MLN4924 treatment. While the effect of MLN4924 maybe global⁴⁸, mimicking hyperinsulinemia via chronic administration of long-acting insulin (Glargine®) reverses the inhibitory effect of MLN4924 on HFD-induced weight gain (Fig. S5d), suggesting that MLN4924 inhibits diet-induced obesity at least in part via inhibiting insulin hypersecretion.

**Fig. 7: The CRL inhibitor pevonedistat/MLN4924 stabilizes ETV5 and improves diet-induced or leptin deficiency-induced obesity/diabetes.**

To exclude possible HFD-specific effects of MLN4924, we examined the leptin-deficient ob/ob mice, whose uncontrolled food intake elicits hyperinsulinemia and low ETV5 levels under normal diet (Fig. 7e). MLN4924 also attenuates excessive ETV5 degradation (Fig. 7e), obesity, and diabetes development in ob/ob mice (Fig. 7f), improving their hyperglycemia (Fig. 7g), hyperinsulinemia (Fig. 7h), glucose tolerance, and insulin sensitivity (Fig. S5e, f). Together, these data suggest that CRL inhibition is a plausible treatment option for hyperinsulinemia, diet/overeating-induced obesity, and obesity-associated diabetes (Fig. 8).

Fig. 8: Scheme depicting the IP₆–CSN–CRL4^COP1–ETV5 pathway as a checkpoint controlling insulin secretion and its potential role in the development of diet/overeating-induced obesity and diabetes.

Discussion

Compared to mechanisms facilitating glucose-induced insulin release, less is known about β-cell autonomous regulators that restrict unwanted or untimely insulin production and secretion. In dissecting the physiological implications of metabolite (IP₆)-mediated CSN–CRL interaction, we demonstrate a hitherto unknown role for mammalian CSN in metabolic control. CSN regulates pancreatic insulin secretion via sequestrating CRL4 from CRL4^COP1 E3 ligase assembly, leading to the build-up ETV5, a transcriptional “brake” that restricts insulin secretion and is relieved upon glycemic rise (Fig. 8). CRL4^COP1 also plays other roles in development and differentiation^9,49, whether glucose regulates CRL4^COP1 in these contexts remain to be investigated. In plants, COP1 and CSN both mediate light-controlled growth via interacting with the CRL4 E3 ligase^50,51. While it would be interesting to examine whether COP1–CSN competition also determines the assembly and E3 ligase activity of CRL4^COP1 in plants, fundamental interaction between COP1 and CSN in growth and metabolism seems preserved across evolution.

The CSN cofactor IP₆ directly interacts with the Rbx1 subunit of CRL, not Cullins, thus potentially influencing all Cullin members that heterodimerizes with Rbx1/2 (refs. ²⁵). Such broad impact of IP₆ could possibly account for embryonic lethality in Csn2^K70E/K70E homozygous mice²⁵. In the Csn2^WT/K70E heterozygous mice, however, partial abolishment of the IP₆ cofactor has rather specific influences on the pancreatic CRL4^COP1–ETV5 axis at multiple levels. First, only Cul4A/B display significantly enhanced neddylation, which is likely due to larger percentage of Cul4s being CSN-bound during their catalytic cycle^29,30, rendering them sensitive to even partial CSN dysfunction. Given that Cul4A and Cul4B are both hyper-neddylated, and that assembly of CRL4A^COP1 and CRL4B^COP1 E3 ligases are both increased in Csn2^WT/K70E mice, it remains to be determined if CRL4A^COP1 and CRL4B^COP1 are equally involved in ETV5 ubiquitylation. Second, among the CRL4s, the assembly of CRL4^COP1 is prominently upregulated in Csn2^WT/K70E pancreas. This is because COP1 is islet-enriched⁸, where it partners with DET1 to compete with CSN for free CRL4, unlike canonical substrate receptors (e.g., DDB2, CSA) that coexist with CSN-bound CRL4 (ref. ¹⁹). Finally, ETV5, but not other CRL4^COP1 substrates examined, is degraded in the Csn2^WT/K70E mutant mice and, more physiologically, in fasted wild-type mice upon refeeding. The same treatment does not deplete ETV5 levels in liver or adipose tissue, pointing toward a highly dedicated IP₆–CSN–CRL4^COP1–ETV5 pathway in pancreatic nutrient response, whereby glucose-induced CRL4^COP1 assembly and yet-unknown modifications on ETV5 may serve as coincidental triggers for its recognition and degradation. Given that ETV5 degradation is nutrient-regulated in β cells and in whole pancreas, its role in other cell types worth further investigation.

The clinical relevance of ETV5 as a CRL4^COP1-degradable checkpoint in feeding-induced insulin secretion is underscored by GWAS identification of ETV5 as a diabetes/obesity-related gene^10,11,12. GWAS-identified gene associated with diabetes are increasingly appreciated to relate to islet function¹³. Given the pronounced assembly of CRL4^COP1 E3 ligase upon feeding, other COP1 targets may also be ubiquitylated in a stimuli-dependent context, thereby contributing to glucose homeostasis regulation. In this regard, the COP1 degradation axis has been recently shown to malfunction in T2D patient β cells with low-exocytotic capacity, whereby upregulated ETV1 suppresses insulin secretion via an inflammatory pathway³⁸, suggesting a COP1-ETV1 axis preventing β cell failure at later stages of T2D. These insights provide a new window to explore the etiology of potential disease-associated variants of COP1 and ETVs.

Two tightly linked and co-evolving hallmarks of diet-induced obesity and diabetes are insulin resistance and hyperinsulinemia⁵², the cause and consequence between them remain a subject of discussion^5,30,47. Mice with reduced insulin gene dosage are resistant to diet-induced obesity and insulin resistance⁵³, while mice carrying extra copies of insulin gene develop insulin resistance as a consequence of hyperinsulinemia⁵⁴. The Csn2^WT/K70E mice we generated represent a rare nontransgenic model of “congenital hyperinsulinism”, whereby insulin oversecretion drives resistance. These observations, together with the notable correlation between insulin levels and obesity/diabetes in humans^39,40,41, indicate that mild suppression of hyperinsulinemia can be therapeutic. Indeed, lowering insulin levels with β-cell K_ATP channel openers has proved beneficial in rodents and humans⁵⁵. However, the challenge remains to identify agents that selectively inhibits hyperinsulinemia without affecting normal insulin secretion⁴⁷. Given that the general CRL inhibitor MLN4924 efficiently inhibits hyperinsulinemia, but not normal insulin levels or sensitivity, and that CRL4^COP1 is particularly sensitive to partial disruption of CSN regulation, more specific modulation of insulin secretion via targeting Cul4 regulators that determine CRL4^COP1 assembly, including the inositol polyphosphate pathway, may benefit obese and early T2D patients.

Methods

Materials

HEK 293, 293T, MIN6, and INS1 cells were purchased from ATCC and cultured in DMEM/10% FBS/1% glutamine/1% Penstrep at 5% CO₂ (vol/vol). The primary antibodies used were: Cul1 (1:1000, CST, 4995), Cul3 (1:1000, CST, 2759), CSN5 (1:1000, CST, 6895), CDT1 (1:1000, Proteintech, 14382-1-AP); ETV1 (1:1000, Abcam, ab184120), ETV5 (1:1000, Abcam, ab102010, CHIP and WB), Cul4A (1:1000, Abcam, ab72548, WB and IP), CSN3 (1:1000, Abcam, ab79698), Cul2 (1:1000, Bethyl, A302–476A), Cul5 (1:1000, Bethyl, A302–173A), COP1 (1:1000, Bethyl, A302–173A), CSB (1:1000, Genetex, GTX104589), GAPDH (1:3000, Proteintech, 10494-1-AP), Exoc6 (1:1000, Proteintech, 12723-1-AP), Sytl3 (1:1000, Proteintech, 22076-1-AP), CSN2 (1:3000, Proteintech, 10969-2-AP), ETV4 (1:1000, Thermo, MA5-15424), Cul4B (1:1000, Sigma, C9995, WB and IP), GST (1:3000, Sigma, A7340), Det1 (1:1000, Santa Cruz, sc-514348), insulin (1:100, R&D Systems, I2018, for IF), and glucagon (1:100, Abcam, ab10988, for IF). Trypsin, shRNA for CSN2 (id: 1773s21c1), and scrambled shRNA were purchased from Sigma. The other shRNAs were generated by cloning into pLKO.1 vector. The corresponding primers are synthesized by Sangon Biotech and are listed in the Supplementary Table 1, which also contain other primers used in this study. IgG beads were purchased from GE Healthcare. GST-TUBE was from LifeSensors. Premade insulin ELISA kit was from Crystal Chem. Glucose-free DMEM and RPMI are from Gibco. All other reagents were purchased from Sigma unless otherwise specified.

GST pulldown and co-immunoprecipitation

Cells were lysed in lysis buffer (50 mM Tris-HCl, pH 7.5, 125 mM NaCl, l% Triton X-100, 1× phosphatase inhibitor, and 1× protease inhibitor). For GST pulldown experiments, cell lysate was incubated with glutathione beads at 4 °C for 2 h. For immunoprecipitation experiments, cell lysate was first incubated with desired antibody at 4 °C overnight, followed by incubation with protein A/G plus beads (Santa Cruz, Inc) at 4 °C for another 1 h. Beads were then washed three times with lysis buffer and boiled in SDS sample buffer. Samples were loaded to SDS–PAGE gel for western blotting of the indicated proteins. Images were acquired using ChampChemi 610 Chemiluminescence Analysis System (SageCreation).

IP₆ detection

The purification and visualization of IP₆ from pancreas were performed according to previously established methods^25,56,57. Tissues were hmogenized and extracted in 1 M perchloric acid (PA) containing 5 mM EDTA. After votexing for 10 min, samples were centrifuged at 18,000 × g for 5 min at 4 °C. The titanium dioxide (TiO₂) beads (Titansphere TiO 5 µm;GL Sciences) were prepared by washing in water and then in 1 M PA containing 5 mM EDTA (4–5 mg for one sample). The supernatants from tissue lysate were mixed with TiO₂ beads added by rotation for 15 min at 4 °C. After centrifuging at 3500 × g for 1 min, beads were washed twice in PA. Bound inositol phosphates were then eluted with 200 µl 5% ammonium hydroxide, which were then vacuum evaporated to 50 µl and subjected to 35.5% PAGE. Phosphate-rich metabolites were imaged with toluidine blue staining, using commercial IP₆ as a control.

Chromatin immunoprecipitation

ETV5 ChIP was performed as previously described⁵⁸. Briefly, INS1 cells were crosslinked with formaldehyde, and lysed in lysis buffer [10 mM Tris, 10 mM NaCl, 0.2% NP40 (Ph8.0), and protease inhibitors]. Nucleus were extracted in 1% SDS and sheared to an average length of 300–500 bp by sonication (Qsonica Q700, 6 pulses, 15 s on, 45 s off at 5 W), followed by dilution in i.p. buffer and overnight incubation with ETV5 antibody (ab102010, 1:100). Upon protein A/G-agarose beads precipitation and wash, immunoprecipitated DNA were de-crosslinked by heating (95 °C, 15 min). Q-PCR was performed using SYBR ® Green Real-time PCR Master Mix, with the following primers: Sytl3-F:5′-ACTCAAACTGGGCTCCCAAG-3′, Sytl3-R:5′-AATGAGAATCTGTACCTAGA-3′; Exoc6-F:5′-AATATCATCACCCAGGATTC-3′, Exoc6-R:5′-GTCAGGCAGGCTTTGCACTA -3′.

Mouse genetics

All mice were maintained in specific pathogen-free facilities in the Laboratory Animal Research Center of Southern University of Science and Technology. Mice were housed in colony cages at 25 °C with 12-h light/12-h dark cycles, with 20–60% humidity. The dark cycle began at 7 p.m. Experiments were conducted following the SUSTech Guide for the Care and Use of Laboratory Animals, and approved by the Animal Ethics Committee (Approval code: SUSTC-2017-076). The knockin mouse model was generated in a C57BL/6J background by targeting the mouse Csn2 gene and introducing the mutation into the equivalent position (K70E) in exon 3 by CRISPR/Cas9. Genotyping was performed using specific primers against mutant allele and WT allele, respectively, followed by restriction digestion and/or DNA sequencing. Heterozygous mutant mice were backcrossed to C57BL/6J wild-type mice. Experiments were conducted exclusively on littermates and, unless specified otherwise, male mice were used for analysis, although female Het mice also display heavier weight and hyperinsulinemia. Ob/ob mice were purchased from Model Animal Research Center, Nanjing University.

HFD treatment

At the age of 8 weeks, mice were fed with HFD (D12492, Research Diets), which consists of Rodent Diet with 60% kcal fat. HFD treatment lasted for 9 weeks, with mice’s body weights, blood glucose levels monitored weekly. MLN4924 (ApexBio) was dissolved in (2-hydroxypropyl)-β-cyclodextrin (Vehicle) and injected intraperitoneally.

Measurement of glucose, FFA, TG, and insulin levels

Blood samples are taken by tail vein nick, the glucose levels are measured by glucometer (Contour TS, Bayer), and levels of insulin, nonesterified free fatty acids, and triglycerides are detected with serum samples, using commercially available kit.

Glucose tolerance test

Mice were fasted for 16 h, and injected intraperitoneally with glucose (1.5 g/kg body weight). Blood was longitudinally collected at 0, 5, 10, 15, 30, 45, 60, 90, and 120 min from the same mice to measure glucose levels with a glucometer.

Insulin tolerance test

Mice were starved for 4–6 h prior to the ITT, Recombinant Human Insulin (ProSpec) was intraperitoneally injected (0.75 U/kg body weight). Blood was then longitudinally collected at 0, 15, 30, 45, 60, 90, and 120 min from the same mice to measure glucose levels.

Streptozocin treatment

Ten-week-old mice were intraperitoneally injected with 50 mg/kg Streptozocin [Sigma, 10 mg/ml in 0.1 M Sodium Citrate Buffer (pH 5.5)] for four consecutive days. Mice’s body weight and blood glucose levels were then monitored weekly to ensure successful destruction of β cells, ITT was performed 6 weeks after streptozocin treatment.

Glucose-stimulated insulin secretion assay

Mice were fasted for at least 16 h prior to the GSIS assay, followed by glucose (1.5 g/kg body weight) injection described for the GTT. Blood samples (>20 µl) are longitudinally collected from the same mice at 5, 10, 20, and 30 min after injection. Blood was allowed to clot at room temperature for 15–30 min, spun at 2000 × g in a tabletop centrifuge for 20 min at 4 °C, and the supernatant serum was collected for Insulin measurement using Insulin ELISA Kit (Crystal Chem).

Free cytoplasmic Ca²⁺ recording

Ca²⁺ recording and analysis followed published methods⁸. Briefly, islets cultured overnight on ploy-l-lysine-coated glass dishes were transferred to in REC buffer (119 mM NaCl, 2.5 mM CaCl₂, 4.7 mM KCl, 10 mM Hepes,1.2 mM MgSO₄, 1.2 mM KH2PO₄, and 2.8 mM glucose) with 5 µM FURA2 (invitrogen) and for 30 min. They were then put into fresh REC buffer to record basal calcium. At 3 min, 20 mM glucose or 25 mM KCl was injected to stimulate calcium influx. The A340 nm/A380 nm ratio was recorded every 2 s using the Metafluor system (Molecular Device).

Isolated islets insulin secretion assay

The protocol described by Suriben et al. was adapted for measuring insulin secretion from isolated islets⁸. Briefly, pancreas was frist distended and then digested in Collagenase XI/HBSS solution at 37.54 °C for 15 min, with shaking every 2–3 min. Upon digestion termination by CaCl2/HBSS solution, cells were pelleted, resuspended, and the islets were captured by filtration through a 70 μm cell strainer. After overnight recovery in RPMI medium containing 11.8 mM glucose, 10% FBS, 50 U/ml penicillin, and 50 μg/ml streptomycin, ten islets per mouse were transferred into Kreb’s buffer containing 2.8 mM glucose for 1 h to measure basal insulin secretion. After supernatant removal, secretogogues (16.8 mM glucose or 20 mM arginine in 2.8 mM glucose) in Kreb’s buffer were added to the islets for 1 h. Insulin in the supernatants was measured as described above. Results are presented as total insulin secreted after normalization to basal secretion in 2.8 mM glucose.

For time-split GSIS measurement, insulin concentrations detected from each time points were normalized to a concentration with total volume being 1 ml, and prior insulin levels were subtracted.

For shRNA-based screening of CRL4 adaptors regulating insulin secretion, the primer sequences are described in Supplementary Table 1. Validated shRNAs in pLKO.1 vectors were used to package lentiviral particles in HEK293T cells cultured in 10-cm plates as previously described²⁵. Lentiviruses in the medium were pelleted by centrifugation and resupended in 200 μl islet culture medium, before being added to islets isolated on the same day. After infection for 24 h, viruses were removed and islets were cultured for another 24 h. Afterward, GSIS was performed as described above.

Islet electron microscopy

Islets were harvested from 3-month-old mice by collagenase digestion and hand picked. Samples were fixed in the 1/2 Karnovsky’s Fixative (2% paraformaldehyde and 2.5% glutaraldehyde in PBS) overnight, post fixed in 1% aqueous osmium tetroxide in PBS for 2 h at 4 °C, and dehydrated through a series of ethanol washes followed by two propylene oxide washes. Samples were embedded in Eponate 12 (TEPELLA-18010) and cured at 60 °C overnight. Ultrathin (80 nm) sections were obtained with an Ultracut microtome (Leica). The ultrathin sections were stained with 1% uranyl acetate in methanol for 5 min, and washed with ddH₂O three times. After drying with filter paper, samples were stained with 0.2% lead citrate for 5 min, then washed with ddH₂O three times, and dried in the dish overnight. Sections were examined in a transmission electron microscope (Hitachi TEM-HT7700) at 100 kV.

Pancreas morphology analysis

Pancreas was dissected rapidly when the mouse was sacrificed, and fixed overnight in 4% paraformaldehyde. Paraffin sections were blocked in 10% horse serum, and stained with different methods including H&E staining, Aldehyde fuchsin staining (Huayueyang Biotechnology), Immunohistochemsitry (ETV5) and immunofluorescence staining. In IF staining, islets were stained for insulin (R&D Systems) and glucagon (Abcam). And the stained area was quantified using Image-Pro Plus software (Media Cybernetics).

Human islets experiments

The study was approved by the Ethics Committee of The Third People’s Hospital of Shenzhen. Pancreatic tissues from metabolically healthy donors were donated via a dedicated Organ Procurment Organization in a double-blinded manner. Informed consents were obtained from donor’s next of kin authorized to give consent. The islet isolation procedures were similarly as for as mouse pancreas. Human islets were cultured in RPMI 1640 media (5.5 mM glucose and 10% FBS) and processed for GSIS with/without MLN4924 treatment or COP1 and ETV5 knockdown, with five islets per sample. To determine the effect of glucose on islet ETV5 levels, 50 islets each were cultured in Kreb’s buffer containing either 2.8 or 16.8 mM glucose for 1 h, before being assayed for ETV5 by western blotting.

EndoC-βH1 cell experiments

EndoC-βH1 cells (EndoCells, Paris, France) were cultured as previously described⁴⁶. Briefly, cells were cultured in DMEM containing 5.6 mM glucose, 2% BSA fraction V (Roche), 10 mM nicotinamide (Merck Millipore), 50 μM 2-mercaptoethanol, 5.5 μg/ml transferrin, 6.7 ng/ml sodium selenite (Sigma-Aldrich), and 1% (vol/vol) penicillin and streptomycin (Thermo Fisher Scientific), and were maintained at 37 °C in air with 5% CO₂.

For insulin secretion assay, cells were preincubated in Krebs-Ringer bicarbonate buffer (KRBB; supplemented with 10 mmol/l HEPES and 2 mg/ml BSA) with 2.8 mmol/l glucose for 2 h. After that, cells were treated with MLN4924 (0, 1, and 5 μM) for 2 h followed by incubation in KRBB containing 2.8 mM or 16.8 M glucose for an additional 1 h. Insulin release into KRBB was measured by ELISA kits. For COP1 and ETV5 knockdown experiments, EndoC-βH1 cells were infected with the respective lentivirus and subjected to puromycin (1 μg/ml) selection for stable knockdown cell lines, which were then used for insulin secretion assays.

Modeling

Structural models of DET1 (Uniprot id: Q7L5Y6) and COP1 (Uniprot id: Q8NHY2) were generated by homology modeling using I-TASSER (version 5.1)⁵⁹. After that, a model of the CRL4^DET1/COP1 complex was manually built based on experiments that determined the binding interface between DET1 and DDB1, and between DET1 and COP1 (Fig. S3). Thus, DET1 was positioned into the CRL4–CSN complex by replacing DDB2 in the cryo-EM structure of CRL4^DDB2–CSN¹⁹, with the helical motif of DET1 (aa 14–26) aligning to DDB1-binding helical motif of DDB2 (ref. ⁶⁰). COP1 was positioned based on prior findings that its coiled-coil domain-containing exon 7 (aa 277–296) mediates binding to DET1 (ref. ⁴⁵). Images were preparaed in Pymol (version 2.2.0), as previously described²⁵.

Statistics and reproducibility

All results are presented as the mean with standard error (SEM) or standard deviation (SD) of at least triplicates. Bar graph figures were generated using Graphpad Prism (version 8.0.1). The sample size (n) defines the number of mice for animal experiments and the number of replicates for cell studies. All datapoints are displayed in the plots when possible. Western blot experiments were indpendently repeated at least twice with similar results.

When comparing data with one variable factor, statistical significance was calculated by two-tailed Student’s tmtest using EXCEL 2010, the exact p values are labaled in the figure. For data with two variable factors, statistical significance was determined by two-way ANOVA with Bonferoni posttest using SigmaPlot (SPSS version 12.3; *p < 0.05, **p < 0.01, and ***p < 0.001). Factor A is time, and factor B is genotype or drug treatment. For data involving time as the third variant, statistical significance was determined by two-way repeated measures ANOVA with Hold-Sidak posttest using SigmaPlot (*p < 0.05, **p < 0.01, and ***p < 0.001). The genotype × timepoint interactions or treatment × timepoint interactions were tested and reported. Other statistical details of the experiments, such as the exact number of mice, can be found in the figure legends and in Supplementary Information files.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The structural models of COP1 and DET1 generated using I-Tasser are available upon request. All other data generated or analyzed during this study are included in this published article (and its Supplementary Information files). Data and reagents requests should be addressed to F.R. (raof@sustech.edu.cn). Source data are provided with this paper.

References

Petroski, M. D. & Deshaies, R. J. Function and regulation of cullin-RING ubiquitin ligases. Nat. Rev. Mol. Cell Biol. 6, 9–20 (2005).
Article CAS PubMed Google Scholar
Lydeard, J. R., Schulman, B. A. & Harper, J. W. Building and remodelling Cullin-RING E3 ubiquitin ligases. EMBO Rep. 14, 1050–1061 (2013).
Article CAS PubMed PubMed Central Google Scholar
Skaar, J. R., Pagan, J. K. & Pagano, M. SCF ubiquitin ligase-targeted therapies. Nat. Rev. Drug Discov. 13, 889–903 (2014).
Article CAS PubMed PubMed Central Google Scholar
Bondeson, D. P. & Crews, C. M. Targeted protein degradation by small molecules. Annu. Rev. Pharm. Toxicol. 57, 107–123 (2017).
Article CAS Google Scholar
Czech, M. P. Insulin action and resistance in obesity and type 2 diabetes. Nat. Med. 23, 804–814 (2017).
Article CAS PubMed PubMed Central Google Scholar
Liu, B. et al. Aberrant expression of FBXO2 disrupts glucose homeostasis through ubiquitin-mediated degradation of insulin receptor in obese mice. Diabetes 66, 689–698 (2017).
Article CAS PubMed Google Scholar
Zhang, Z. et al. Insulin resistance and diabetes caused by genetic or diet-induced KBTBD2 deficiency in mice. Proc. Natl Acad. Sci. USA 113, E6418–E6426 (2016).
Article CAS PubMed PubMed Central Google Scholar
Suriben, R. et al. beta-Cell insulin secretion requires the ubiquitin ligase COP1. Cell 163, 1457–1467 (2015).
Article CAS PubMed Google Scholar
Zhang, Z. et al. Transcription factor Etv5 is essential for the maintenance of alveolar type II cells. Proc. Natl Acad. Sci. USA 114, 3903–3908 (2017).
Article CAS PubMed PubMed Central Google Scholar
Locke, A. E. et al. Genetic studies of body mass index yield new insights for obesity biology. Nature 518, 197–206 (2015).
Article CAS PubMed PubMed Central Google Scholar
Berndt, S. I. et al. Genome-wide meta-analysis identifies 11 new loci for anthropometric traits and provides insights into genetic architecture. Nat. Genet. 45, 501–512 (2013).
Article CAS PubMed PubMed Central Google Scholar
Thorleifsson, G. et al. Genome-wide association yields new sequence variants at seven loci that associate with measures of obesity. Nat. Genet. 41, 18–24 (2009).
Article CAS PubMed Google Scholar
Mattis, K. K. & Gloyn, A. L. From genetic association to molecular mechanisms for islet-cell dysfunction in type 2 diabetes. J. Mol. Biol. 432, 1551–1578 (2020).
Article CAS PubMed Google Scholar
Cope, G. A. et al. Role of predicted metalloprotease motif of Jab1/Csn5 in cleavage of Nedd8 from Cul1. Science 298, 608–611 (2002).
Article ADS CAS PubMed Google Scholar
Duda, D. M. et al. Structural insights into NEDD8 activation of cullin-RING ligases: conformational control of conjugation. Cell 134, 995–1006 (2008).
Article CAS PubMed PubMed Central Google Scholar
Baek, K. et al. NEDD8 nucleates a multivalent cullin-RING-UBE2D ubiquitin ligation assembly. Nature 578, 461–466 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Chamovitz, D. A. et al. The COP9 complex, a novel multisubunit nuclear regulator involved in light control of a plant developmental switch. Cell 86, 115–121 (1996).
Article CAS PubMed Google Scholar
Lyapina, S. et al. Promotion of NEDD-CUL1 conjugate cleavage by COP9 signalosome. Science 292, 1382–1385 (2001).
Article ADS CAS PubMed Google Scholar
Cavadini, S. et al. Cullin-RING ubiquitin E3 ligase regulation by the COP9 signalosome. Nature 531, 598–603 (2016).
Article ADS CAS PubMed Google Scholar
Enchev, R. I. et al. Structural basis for a reciprocal regulation between SCF and CSN. Cell Rep. 2, 616–627 (2012).
Article CAS PubMed PubMed Central Google Scholar
Lingaraju, G. M. et al. Crystal structure of the human COP9 signalosome. Nature 512, 161–165 (2014).
Article ADS CAS PubMed Google Scholar
Hatch, A. J. & York, J. D. SnapShot: inositol phosphates. Cell 143, 1030–1030 e1 (2010).
Article CAS PubMed PubMed Central Google Scholar
Scherer, P. C. et al. Inositol hexakisphosphate (IP6) generated by IP5K mediates cullin-COP9 signalosome interactions and CRL function. Proc. Natl Acad. Sci. USA 113, 3503–3508 (2016).
Article ADS CAS PubMed PubMed Central Google Scholar
Rao, F., Lin, H. & Su, Y. Cullin-RING ligase regulation by the COP9 signalosome: structural mechanisms and new physiologic players. Adv. Exp. Med. Biol. 1217, 47–60 (2020).
Article CAS PubMed Google Scholar
Lin, H. et al. Basis for metabolite-dependent Cullin-RING ligase deneddylation by the COP9 signalosome. Proc. Natl Acad. Sci. USA 117, 4117–4124 (2020).
Article CAS PubMed PubMed Central Google Scholar
Soucy, T. A. et al. An inhibitor of NEDD8-activating enzyme as a new approach to treat cancer. Nature 458, 732–736 (2009).
Article ADS CAS PubMed Google Scholar
Lykke-Andersen, K. et al. Disruption of the COP9 signalosome Csn2 subunit in mice causes deficient cell proliferation, accumulation of p53 and cyclin E, and early embryonic death. Mol. Cell Biol. 23, 6790–6797 (2003).
Article CAS PubMed PubMed Central Google Scholar
Schlierf, A. et al. Targeted inhibition of the COP9 signalosome for treatment of cancer. Nat. Commun. 7, 13166 (2016).
Article ADS CAS PubMed PubMed Central Google Scholar
Gray, S. L., Donald, C., Jetha, A., Covey, S. D. & Kieffer, T. J. Hyperinsulinemia precedes insulin resistance in mice lacking pancreatic beta-cell leptin signaling. Endocrinology 151, 4178–4186 (2010).
Article CAS PubMed Google Scholar
Shanik, M. H. et al. Insulin resistance and hyperinsulinemia Is hyperinsulinemia the cart or the horse? Diabetes Care 31, S262–S268 (2008).
Article CAS PubMed Google Scholar
Ning, J. et al. Insulin and insulin signaling play a critical role in fat induction of insulin resistance in mouse. Am. J. Physiol. Endocrinol. Metab. 301, E391–E401 (2011).
Article CAS PubMed PubMed Central Google Scholar
Flores-Riveros, J. R., McLenithan, J. C., Ezaki, O. & Lane, M. D. Insulin down-regulates expression of the insulin-responsive glucose transporter (GLUT4) gene: effects on transcription and mRNA turnover. Proc. Natl Acad. Sci. USA 90, 512–516 (1993).
Article ADS CAS PubMed PubMed Central Google Scholar
Cen, H., Botezelli, J. D., Wang, S. & Johnson, J. D. Modulation of muscle cell Insr and insulin receptor signaling by hyperinsulinemia in vitro and in vivo. Preprint at bioRxiv https://doi.org/10.1101/556571 (2019).
Tang, J., Neidigh, J. L., Cooksey, R. C. & McClain, D. A. Transgenic mice with increased hexosamine flux specifically targeted to beta-cells exhibit hyperinsulinemia and peripheral insulin resistance. Diabetes 49, 1492–1499 (2000).
Article CAS PubMed Google Scholar
Sekeres, M. A., Fram, R. J., Hua, Z. & Ades, L. Phase 3 study of first line pevonedistat (PEV) + azacitidine (AZA) versus single-agent AZA in patients with higher-risk myelodysplastic syndromes (HR MDS), chronic myelomonocytic leukemia (CMML) or low-blast acute myelogenous leukemia (AML). J. Clin. Oncol. 36, TPS7077 (2018).
Bennett, E. J., Rush, J., Gygi, S. P. & Harper, J. W. Dynamics of cullin-RING ubiquitin ligase network revealed by systematic quantitative proteomics. Cell 143, 951–965 (2010).
Article CAS PubMed PubMed Central Google Scholar
Reichermeier, K. M. et al. PIKES analysis reveals response to degraders and key regulatory mechanisms of the CRL4 network. Mol. Cell 77, 1092–1106 (2020).
Camunas-Soler, J. et al. Patch-seq links single-cell transcriptomes to human islet dysfunction in diabetes. Cell Metab. 31, 1017–1031 e4 (2020).
Article CAS PubMed PubMed Central Google Scholar
Corkey, B. E. Banting lecture 2011: hyperinsulinemia: cause or consequence? Diabetes 61, 4–13 (2012).
Article CAS PubMed Google Scholar
Kim, M. K., Reaven, G. M. & Kim, S. H. Dissecting the relationship between obesity and hyperinsulinemia: Role of insulin secretion and insulin clearance. Obesity 25, 378–383 (2017).
Article CAS PubMed Google Scholar
Templeman, N. M., Skovso, S., Page, M. M., Lim, G. E. & Johnson, J. D. A causal role for hyperinsulinemia in obesity. J. Endocrinol. 232, R173–R183 (2017).
Article CAS PubMed Google Scholar
Gutierrez-Aguilar, R. et al. The role of the transcription factor ETV5 in insulin exocytosis. Diabetologia 57, 383–391 (2014).
Article CAS PubMed Google Scholar
Jin, J., Arias, E. E., Chen, J., Harper, J. W. & Walter, J. C. A family of diverse Cul4-Ddb1-interacting proteins includes Cdt2, which is required for S phase destruction of the replication factor Cdt1. Mol. Cell 23, 709–721 (2006).
Article CAS PubMed Google Scholar
Rizzini, L. et al. Cryptochromes-mediated inhibition of the CRL4(Cop1)-complex assembly defines an evolutionary conserved signaling mechanism. Curr. Biol. 29, 1954–1962 e4 (2019).
Article CAS PubMed PubMed Central Google Scholar
Wertz, I. E. et al. Human De-etiolated-1 regulates c-Jun by assembling a CUL4A ubiquitin ligase. Science 303, 1371–1374 (2004).
Article ADS CAS PubMed Google Scholar
Ravassard, P. et al. A genetically engineered human pancreatic beta cell line exhibiting glucose-inducible insulin secretion. J. Clin. Investig. 121, 3589–3597 (2011).
Article CAS PubMed PubMed Central Google Scholar
Page, M. M. & Johnson, J. D. Mild suppression of hyperinsulinemia to treat obesity and insulin resistance. Trends Endocrinol. Metab. 29, 389–399 (2018).
Article CAS PubMed Google Scholar
Park, H. S. et al. PPARgamma neddylation essential for adipogenesis is a potential target for treating obesity. Cell Death Differ. 23, 1296–1311 (2016).
Article CAS PubMed PubMed Central Google Scholar
Cui, C. P. et al. Author correction: dynamic ubiquitylation of Sox2 regulates proteostasis and governs neural progenitor cell differentiation. Nat. Commun. 10, 173 (2019).
Article ADS CAS PubMed PubMed Central Google Scholar
Wei, N. & Deng, X. W. The COP9 signalosome. Annu. Rev. Cell Dev. Biol. 19, 261–286 (2003).
Article CAS PubMed Google Scholar
Sanchez-Barcelo, E. J., Mediavilla, M. D., Vriend, J. & Reiter, R. J. Constitutive photomorphogenesis protein 1 (COP1) and COP9 signalosome, evolutionarily conserved photomorphogenic proteins as possible targets of melatonin. J. Pineal Res. 61, 41–51 (2016).
Article CAS PubMed Google Scholar
Kim, S. H. & Reaven, G. M. Insulin resistance and hyperinsulinemia: you can’t have one without the other. Diabetes Care 31, 1433–1438 (2008).
Article CAS PubMed PubMed Central Google Scholar
Mehran, A. E. et al. Hyperinsulinemia drives diet-induced obesity independently of brain insulin production. Cell Metab. 16, 723–737 (2012).
Article CAS PubMed Google Scholar
Marban, S. L. & Roth, J. Transgenic hyperinsulinemia: a mouse model of insulin resistance and glucose intolerance without obesity. in Lessons from Animal Diabetes, 201–224 (Birkhauser, 1996).
Erion, K. & Corkey, B. E. beta-Cell failure or beta-cell abuse? Front. Endocrinol. 9, 532 (2018).
Article Google Scholar
Zhang, X. et al. Suramin and NF449 are IP5K inhibitors that disrupt inositol hexakisphosphate-mediated regulation of cullin-RING ligase and sensitize cancer cells to MLN4924/pevonedistat. J. Biol. Chem. 295, 10281–10292 (2020).
Article CAS PubMed PubMed Central Google Scholar
Wilson, M. S. & Saiardi, A. Inositol phosphates purification using titanium dioxide beads. Bio Protoc 8, e2959 (2018).
Lopez-Delisle, L. et al. Activated ALK signals through the ERK-ETV5-RET pathway to drive neuroblastoma oncogenesis. Oncogene 37, 1417–1429 (2018).
Article CAS PubMed PubMed Central Google Scholar
Yang, J. & Zhang, Y. I-TASSER server: new development for protein structure and function predictions. Nucleic Acids Res. 43, W174–W181 (2015).
Article CAS PubMed PubMed Central Google Scholar
Li, T., Robert, E. I., van Breugel, P. C., Strubin, M. & Zheng, N. A promiscuous alpha-helical motif anchors viral hijackers and substrate receptors to the CUL4-DDB1 ubiquitin ligase machinery. Nat. Struct. Mol. Biol. 17, 105–111 (2010).
Article PubMed CAS Google Scholar

Download references

Acknowledgements

This work was supported by grants from the National Science Foundation of China (31872798 and 91853129 to F.R.), the Shenzhen Municipal Government (KQJSCX20180322152418316, JCYJ20170412153517422, and JCYJ20170817104311912 to F.R.), the Department of Science and Technology of Guangdong Province (2018A030313207 to F.R.), and the Hong Kong Government GRF (17127718 and 17127215 to B.K.C.C).

Author information

These authors contributed equally: Hong Lin, Yuan Yan, Yifan Luo, Wing Yan So.

Authors and Affiliations

School of Life Sciences, Department of Biology, Southern University of Science and Technology, Shenzhen, Guangdong, China
Hong Lin, Yuan Yan, Yifan Luo, Xiayun Wei, Xiaozhe Zhang, Xiaoli Yang, Jun Zhang, Yang Su, Xiuyan Yang, Bobo Zhang & Feng Rao
School of Biological Sciences, The University of Hong Kong, Pokfulam, Hong Kong
Yifan Luo & Billy Kwok Chong Chow
Singapore Bioimaging Consortium, Agency for Science, Technology, and Research, Singapore, Singapore
Wing Yan So & Weiping Han
Department of Hepatic Surgery, the Third People’s Hospital of Shenzhen and the Second Affiliated Hospital of Southern University of Science and Technology, Shenzhen, Guangdong, China
Kangjun Zhang & Nan Jiang
National Institute of Biological Sciences, Beijing, China
Fengchao Wang

Authors

Hong Lin
View author publications
You can also search for this author in PubMed Google Scholar
Yuan Yan
View author publications
You can also search for this author in PubMed Google Scholar
Yifan Luo
View author publications
You can also search for this author in PubMed Google Scholar
Wing Yan So
View author publications
You can also search for this author in PubMed Google Scholar
Xiayun Wei
View author publications
You can also search for this author in PubMed Google Scholar
Xiaozhe Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Xiaoli Yang
View author publications
You can also search for this author in PubMed Google Scholar
Jun Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Yang Su
View author publications
You can also search for this author in PubMed Google Scholar
Xiuyan Yang
View author publications
You can also search for this author in PubMed Google Scholar
Bobo Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Kangjun Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Nan Jiang
View author publications
You can also search for this author in PubMed Google Scholar
Billy Kwok Chong Chow
View author publications
You can also search for this author in PubMed Google Scholar
Weiping Han
View author publications
You can also search for this author in PubMed Google Scholar
Fengchao Wang
View author publications
You can also search for this author in PubMed Google Scholar
Feng Rao
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

Y.Y., Y.L., B.K.C.C., and F.R. designed research; H.L., Y.Y., Y.L., W.Y.S., X.Z., Xiaoli Yang, J.Z., X.W., Y.S., B.Z., and Xiuyan Yang performed research; B.K.C.C., W.H., and F.R. supervised research. N.J., K.Z., and F.W. contributed new reagents/analytic tools; H.L., Y.Y., X.Z., and F.R. analyzed data; and F.R. wrote the paper. All authors reviewed and edited the manuscript.

Corresponding author

Correspondence to Feng Rao.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks Wolfgang Dubiel and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, 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 license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license 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 license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Lin, H., Yan, Y., Luo, Y. et al. IP₆-assisted CSN-COP1 competition regulates a CRL4-ETV5 proteolytic checkpoint to safeguard glucose-induced insulin secretion. Nat Commun 12, 2461 (2021). https://doi.org/10.1038/s41467-021-22941-3

Download citation

Received: 06 April 2020
Accepted: 08 April 2021
Published: 28 April 2021
DOI: https://doi.org/10.1038/s41467-021-22941-3

This article is cited by

E26 transformation-specific transcription variant 5 in development and cancer: modification, regulation and function
- Yi Wei
- Shenqi Han
- Bixiang Zhang
Journal of Biomedical Science (2023)
Constitutive photomorphogenic protein 1 ubiquitinates interleukin-1 receptor accessory protein in human liver cancer
- Kuan Cao
- Zhiyi Liu
- Bin Zhang
Journal of Cancer Research and Clinical Oncology (2023)
Chromatin accessibility analysis identifies the transcription factor ETV5 as a suppressor of adipose tissue macrophage activation in obesity
- Ren-Dong Hu
- Wen Zhang
- Kun Qu
Cell Death & Disease (2021)
Role of NEDD8 and neddylation dynamics in DNA damage response
- Yifan Luo
- Yang Su
- Feng Rao
Genome Instability & Disease (2021)
5-IP7 is a GPCR messenger mediating neural control of synaptotagmin-dependent insulin exocytosis and glucose homeostasis
- Xiaozhe Zhang
- Na Li
- Feng Rao
Nature Metabolism (2021)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.