SEL1L-HRD1 ER-associated degradation regulates leptin receptor maturation and signaling in POMC neurons in diet-induced obesity

Endoplasmic reticulum (ER) homeostasis in the hypothalamus has been implicated in the pathogenesis of certain patho-physiological conditions such as diet-induced obesity (DIO) and type 2 diabetes; however, the significance of ER quality control mechanism(s) and its underlying mechanism remain largely unclear and highly controversial in some cases. Moreover, how the biogenesis of nascent leptin receptor in the ER is regulated remains largely unexplored. Here we report that the SEL1L-HRD1 protein complex of the highly conserved ER-associated protein degradation (ERAD) machinery in POMC neurons is indispensable for leptin signaling in diet-induced obesity. SEL1L-HRD1 ERAD is constitutively expressed in hypothalamic POMC neurons. Loss of SEL1L in POMC neurons attenuates leptin signaling and predisposes mice to HFD-associated pathologies including leptin resistance. Mechanistically, newly synthesized leptin receptors, both wildtype and disease-associated human mutant Cys604Ser (Cys602Ser in mice), are misfolding prone and bona fide substrates of SEL1L-HRD1 ERAD. Indeed, defects in SEL1L-HRD1 ERAD markedly impair the maturation of these receptors and causes their ER retention. This study not only uncovers a new role of SEL1L-HRD1 ERAD in the pathogenesis of diet-induced obesity and central leptin resistance, but a new regulatory mechanism for leptin signaling.

Among over a dozen of putative ERAD complexes, the SEL1L-HRD1 protein complex represents the most evolutionarily conserved ERAD branch where SEL1L/Hrd3p is an obligatory cofactor for the E3 ligase HRD1 [28][29][30][32][33][34] . Recentstudies using cell type-specific SEL1L or HRD1 knockout mouse models have revealed the patho-physiological importance of SEL1L-HRD1 ERAD in a substrate-specific manner [35][36][37][38][39][40][41][42] .Particularly relevant to this study, SEL1L-HRD1 ERAD has been reported as indispensable for AVP and POMC neurons to control water balance and food intake via the maturation of prohormones, proAVP and POMC, respectively 39,40 .POMC neuron-specific Sel1L deletion leads to hyperphagia and age-associated obesity starting around 13 weeks of age when fed a low-fat chow diet 39 .Given the importance of POMC neurons in maintaining energy homeostasis under various nutritional status, one outstanding question is the relevance and significance of SEL1L-HRD1 ERAD in POMC neurons under pathophysiological conditions, including DIO.
Here, we show that SEL1L-HRD1 ERAD in POMC neurons at the arcuate nucleus (ARC) of the hypothalamus, a key group of metabolic neurons that control food intake and energy expenditure 43 , controls DIO pathogenesis and leptin sensitivity via the regulation of leptin receptor biogenesis and signaling.Soon after weaning, POMC-specific Sel1L deficient (Sel1L POMC ) mice are hypersensitive to DIO.Much to our surprise, SEL1L-HRD1 ERAD is indispensable for the maturation of nascent leptin receptor to reach to the cell surface.Hence, SEL1L-HRD1 ERAD is a critical regulator of the maturation of leptin receptor in the ER and thereby leptin signaling in POMC neurons.

Transient upregulation of SEL1L-HRD1 ERAD expression in the hypothalamus in
response to high fat diet (HFD) feeding.
We previously showed that the SEL1L-HRD1 protein complex is constitutively expressed in the ARC of the hypothalamus 39 .Here we first asked whether its expression in the ARC region is regulated in response to overnutrition by placing the mice on 60% HFD (60% calories derived from fat) for 1 or 8 weeks.HFD feeding expectedly reduced the expression of Pomc, Npy and Agrp (Supplementary Fig. 1A), while enhance the protein levels of POMC derivatives β-Endorphin and α-MSH (Supplementary Fig. 1B-E).Moreover, HFD feeding enhanced neuronal activity in PVH region as measured by nuclear c-FOS following both 1-and 8-week HFD (Supplementary Fig. 1D, E).One-week HFD significantly induced Hrd1 mRNA level, but not Sel1L mRNA level, while 8-week HFD feeding had no such effect (Fig. 1A).At the protein levels, both SEL1L and HRD1 proteins, were elevated at 1-week HFD, and returned to the basal levels after 8-week HFD (Fig. 1B, C), pointing to the transient response to SEL1L-HRD1 expression in the hypothalamus in response to HFD challenge.We next performed confocal microscopy to visualize SEL1L-HRD1 expression in the ARC regions in response to HFD.To visualize POMC neurons, we used POMC-eGFP transgenic mice where eGFP is under the control of POMC promoter 39,44 .SEL1L protein level was increased specifically in POMC neurons upon 1-week HFD, and returned to the basal level with prolonged HFD feeding (Fig. 1D, E).Similar observation was obtained for HRD1 protein levels in POMC neurons, but unlike SEL1L, HRD1 protein level was transiently upregulated in non POMC neurons as well (Fig. 1F, G).Hence, SEL1L-HRD1 expression in POMC neurons are responsive to acute, but not chronic, nutrient overload.

Hypothalamic POMC-specific ERAD deficiency leads to early-set DIO and its pathologies.
To delineate the significance of hypothalamic ERAD in DIO, we next characterized the phenotypes of Sel1L POMC mice, generated by crossing Sel1L f/f with the Pomc-Cre mouse line 39 , following 8-week HFD feeding from 5 weeks of age.While, in line with our previous report, Sel1L POMC mice appeared comparably to WT littermates in terms of body weight on chow diet for the first 13 weeks of age 39 (Fig. 2A), Sel1L POMC mice, both sexes, gained significantly more body weight soon after HFD feeding (Fig. 2A).Body composition analysis showed that fat content was significantly increased in Sel1L POMC mice, reaching over 50% of body mass after 8week HFD (Fig. 2B and Supplementary Fig. 2A) with more lipid deposition in the livers, as well as both white and brown adipose tissues (WAT and BAT) (Fig. 2C).Sel1L POMC mice became highly glucose intolerant and insulin resistant following 8-week HFD (Fig. 2D, E), with elevated ad libitum and fasting blood glucose (Fig. 2F) and ad libitum insulin levels (Fig. 2G).In addition, glucagon and corticosterone levels were elevated in Sel1L POMC mice (Supplementary Fig. 2B,     C), while rectal temperature in Sel1L POMC mice was decreased by 2 degrees compared to that of WT littermates (Supplementary Fig. 2D).Hence, we concluded that mice with POMC-specific ERAD defects exhibit early onset DIO and its pathologies including glucose and insulin resistance.

Hypothalamic ERAD deficiency triggers hyperphagia and leptin resistance.
We next explored the possible mechanism underlying the susceptibility to DIO in Sel1L POMC mice.Sel1L POMC mice consumed ~ 40% more food daily, i.e., hyperphagia, upon both 1-and 8week HFD feeding (Fig. 3A).To directly demonstrate the direct causal link between food intake and weight gain, we performed pair feeding (giving the same amount of the food as WT littermates consume) following 8-week ad libitum HFD feeding.Sel1L POMC mice gained weight quite rapidly under ad libitum feeding of HFD; however, their weight gain was significantly slowed down following pair-feeding and recovered when placed on ad libitum HFD feeding again (Fig. 3B).Indeed, weight gain of Sel1L POMC mice was comparable to that of WT littermates if pair-feeding was performed at the beginning of HFD feeding (Fig. 3C).We then tested whether hyperphagia of Sel1L POMC mice is caused by leptin resistance by leptin injection (Fig. 3D).Leptin injection was expected to induce body weight loss in WT mice, but not Sel1L POMC mice.Indeed, unlike WT mice, Sel1L POMC mice continued to gain body weight following leptin injection (Fig. 3D, E).This difference in body weight gain was likely due to the differences in food intake in response to leptin injection (Fig. 3F), pointing to a significant leptin resistance in Sel1L POMC mice.Sel1L POMC mice exhibited progressively marked hyperleptinemia with HFD feeding (Fig. 3G).Hence, we concluded that hypothalamic POMC neurons-specific ERAD deficiency triggers hyperphagia and leptin resistance.

The effect of hypothalamic SEL1L-HRD1 ERAD in DIO is mediated by leptin resistance.
To further establish the effect of leptin resistance in ERAD deficiency-associated DIO, we next performed parabiosis where two littermates were surgically stitched together to allow the sharing of the circulation (Fig. 4A).Following two weeks of recovery on chow diet, the parabionts WT: Sel1L POMC (Group III) were placed on HFD for 8 weeks (Fig. 4A).Two control parabionts, WT: WT (Group I) and Sel1L POMC :Sel1L POMC (Group II), gained weight as expected with the latter pair becoming obese (Fig. 4B).However, in WT: Sel1L POMC (Group III) parabionts, body weight gain for WT mice was attenuated compared to WT mice in WT: WT (Group I) control parabionts (P=0.08), while body weight gain for Sel1L POMC mice was comparable to that of Sel1L POMC :Sel1L POMC parabionts (Group II) (Fig. 4B).Body compositions (i.e., lean vs. fat) in parabionts were not affected by the partner (Fig. 4C).Moreover, serum leptin and insulin levels were highly elevated in the Sel1L POMC mice, but unaltered in WT mice regardless of the partners (Fig. 4D, E).Hence, these data suggested that hypothalamic SEL1L-HRD1 ERAD controls DIO pathogenesis via hyperleptinemia.

Hypothalamic SEL1L-HRD1 ERAD deficiency impairs leptin-pSTAT3 signaling.
We next asked how POMC-specific SEL1L-HRD1 ERAD regulates leptin sensitivity.As leptin signaling induces phosphorylation of STAT3 (pSTAT3), we next examined the levels of pSTAT3 in POMC neurons following leptin challenge.To visualize POMC neurons, we generated Sel1L POMC mice on the POMC-eGFP background (Sel1L POMC ;POMC-eGFP) 39,44 .HFD feeding progressively blunted leptin-induced pSTAT3 in the POMC neurons of the ARC region of WT mice, but to a much greater extent, in Sel1L POMC mice (Fig. 5A-D and Supplementary Fig. 3).In keeping with the notion that pSTAT3 a critical transcription factor for the Pomc gene 45 , hypothalamic Pomc mRNA expression was markedly decreased in Sel1L POMC mice with HFD (Fig. 5E).Moreover, Western blot analysis of pSTAT3 of the ARC region also showed a greater reduction of the percent of STAT3 being phosphorylated following HFD feeding (Fig. 5F, G).Thus, our data suggested that SEL1L-HRD1 ERAD in POMC neurons is vital for maintaining central leptin sensitivity during DIO pathogenesis.

The effect of POMC-specific ERAD in DIO is likely uncoupled from UPR or inflammation.
As ERAD deficiency expectedly causes the accumulation of unfolded/misfolded proteins in the ER that can potentially trigger UPR and given the reported role of UPR in DIO pathogenesis, we next tested whether ERAD deficiency activates UPR and if so, to what extent.There was no detectable activation of the PERK pathway as measured by phosphorylation of PERK and its downstream phosphorylation of eIF2α (Fig. 6A and Supplementary Fig. 4A).Phosphorylation of IRE1α, on the other hand, was moderately elevated in the ARC of Sel1L POMC mice, so was the splicing of Xbp1 mRNA (a downstream effector of IRE1α) (Fig. 6B, C and Supplementary Fig. 4B, C).Consistently, ER chaperons BiP (an XBP1 target) was mildly elevated in the ARC of Sel1L POMC mice (Fig. 6A and Supplementary Fig. 4A, D).In vitro, treatment with an ER stress inducer thapsigargin (Tg) induced strong ER stress, but failed to affect leptin signaling in WT HEK293T cells transfected with long isoform of mouse Leptin receptors (mLepRb) (Fig. 6D and Supplementary Fig. 4E), indicating that UPR is not sufficient to induce leptin resistance.
Importantly, we found no significant POMC neuronal loss in the ARC of Sel1L POMC ;POMC-eGFP mice (Fig. 6E).Inflammatory markers were largely comparable in the ARC of Sel1L POMC mice compared to those in WT littermates as measured by phosphorylation and protein levels of c-Jun N-terminal Kinase (JNK) as well as protein levels of I kappa B alpha (IκBα) (Fig. 6F, G).

Chronic HFD feeding mildly increased astrogliosis in the ARC regions of both Sel1L POMC and
Sel1L POMC mice as measured by both Western blot and immunofluorescence staining of astrocyte marker Glial Fibrillary acidic protein (GFAP) and/or microglia marker Ionized calciumbinding adaptor molecule 1 (IBA1) (Fig. 6F-H and Supplementary Fig. 4F).Taken together, these data demonstrate that Sel1L deficiency in POMC neurons triggers leptin resistance, independently of UPR, neuronal cell death and inflammation.
Surprisingly, the protein level of mLepRb was significantly higher in HRD1 -/-HEK293T cells compared to that of WT cells, under both serum-deprived and -supplemented conditions (Fig. 7B, C).Moreover, SEL1L interaction with in LepRb-transfected cells was markedly enhanced in HRD1 -/-cells where substrate-SEL1L interaction is known to be stabilized 29,50,51 (Fig. 7D, E).
LepRb was ubiquitinated in an HRD1-dependent manner (Fig. 7F) and was significantly stabilized in HRD1 -/-cells compared to that in WT cells (Fig. 7G).
We next assess the consequence of ERAD deficiency on LepRb maturation in the ER.
Endoglycosidase H (EndoH) digestion, which cleaves asparagine-linked high mannose or hybrid glycans of the immature glycoproteins predominantly in ER 52 , revealed significantly lower fraction of EndoH resistant form of LepRb that were able to exit the ER for complete maturation in HRD1 -/-HEK293T cells (Fig. 7H).This was further confirmed by surface biotinylation assay followed by immunoprecipitation with streptavidin-beads, which indicated reduced proportion of surface LepRb in HRD1 -/-cells (Fig. 7I and Supplementary Fig. 5A).Moreover, confocal microscopy following immunofluorescence staining further demonstrated an altered distribution of LepRb with increased intracellular, but decreased surface, expression in ERAD-deficient cells (Fig. 7J and Supplementary Fig. 5B).In the absence of SEL1L-HRD1, LepRb protein was prone to form high molecular weight aggregates via disulfide bonds (Fig. 7K) Taken together, our data show that SEL1L-HRD1 ERAD is required for the maturation of LepRb by targeting the misfolding-prone or misfolded LepRb for proteasomal degradation.

SEL1L-HRD1 ERAD degrades and limits the pathogenicity of human LepRb Cys604Ser (C604S) mutant.
To demonstrate the clinical relevance of our findings, we asked whether human LepRb (hLepRb) mutants 53,54 are SEL1L-HRD1 ERAD substrates.Here, we focused on hLepRb mutant C604S, a recessive point mutation due to missense homozygous substitution T > A at position 1810, identified in two brothers at 1-and 5-years old with severely early onset obesity 54,55 .C604-C674 forms a disulfide bond in human LepRb corresponding to C602-C672 in mouse LepRb (Fig. 8A, B) [56][57][58] .This mutation has been predicted as loss-of-function likely due to defects in folding 54,[56][57][58] .C602S mLepRb significantly impaired leptin response compared to WT mLepRb in WT cells, which was further diminished in HRD1 -/-cells (Fig. 8C).Similar to WT mLepRb, C602S mLepRb was stabilized in the absence of HRD1 (Fig. 8D).Notably, C602S mLepRb readily formed HMW aggregates in WT HEK293T cells, and to much greater extent, in HRD1 -/-cells (Fig. 8E).Such aggregates likely formed in the ER as demonstrated by their colocalization with the ER chaperone BiP based on immunostaining (Fig. 8F-I).Hence, SEL1L-HRD1 ERAD is indispensable for the degradation of nascent WT and, at least a subset of, disease mutant LepRb, which ensures the maturation, trafficking and membrane display of functional LepRb.

DISCUSSION
This study not only identifies a novel regulatory mechanism for leptin receptor and signaling, but also reports a key role of hypothalamic ERAD in maintaining energy homeostasis under nutrient overload conditions.SEL1L-HRD1 ERAD defects in POMC neurons predispose mice to DIO and its pathologies, due to hyperphagia and hypothalamic leptin resistance.Our mechanistic studies establish LepRb as a bona fide endogenous substrate of SEL1L-HRD1 ERAD.Pointing to the clinical relevance of our findings, human recessive LepRb C604S variant is trapped in the ER and degraded by SEL1L-HRD1 ERAD (Fig. 9).In the absence of SEL1L-HRD1 ERAD, both WT and C604S LepRb are trapped in the ER in the form of HMW aggregates, with attenuated cell surface expression (Fig. 8E-I and Fig. 9).While this reported effect of ERAD in POMC neurons is in keeping with recent studies demonstrating the profound physiological importance of SEL1L-HRD1 ERAD in vivo 39,40 , it uncovers a novel function of SEL1L-HRD1 ERAD in leptin signaling and a novel regulatory mechanism for leptin biology.
Our data show that hypothalamic SEL1L deficiency markedly increases the progression and pathogenesis of DIO in mice.Sel1L-deficient POMC neurons exhibit mild alterations in ER homeostasis including elevated activation of the IRE1α -XBP1 pathway and expression of ER chaperones, but without any detectable cell death.As previous studies have shown that deficiency of Ire1a or Xbp1 in POMC neurons predispose mice to DIO 21 , while gain-of-function of XBP1s in POMC neurons had an opposite effect 23 , we conclude that the effect of SEL1L-HRD1 ERAD is uncoupled from IRE1α -XBP1 pathway of the UPR and cell death, which is in line with many recent studies of various tissue-specific Sel1L-or Hrd1-deficient models 37,[39][40][41][42]59 .
These findings point to the cellular adaption in response to ERAD deficiency 25 .Such mild UPR activation and chaperone expression are potentially cyto-protective in response to the accumulation of misfolding proteins in the ER.
Previous reports have suggested that UPR may play a causal role in leptin resistance due to impaired leptin signaling 15,17,60 .These studies were performed via the administration of ER stress inducers tunicamycin and thapsigargin which can be fraught with artefacts.Indeed, tunicamycin can inhibit glycosylation of the glycoproteins 61 including LepRb, and thus the impaired leptin signaling can be directly due to defective glycosylation and concomitant functionality of LepRb instead of UPR activation as a general outcome of numerous dysregulation of glycoproteins.Further, high dosage of ER stress inducers included in previous studies may fall far from any physiological relevance 15,17,60 .In our study, thapsigargin treatment induced a range of ER stress response in a dose dependent manner, but failed to alter leptin signaling in WT HEK293T cells transfected with mLepRb even at the high level of UPR.Hence, collective evidence suggests that UPR is likely uncoupled from leptin signaling.The reason for these discrepancies remains unknown.Careful future studies are needed to validate either model.
This study demonstrates an important role of SEL1L-HRD1 ERAD in leptin signaling, at least in part via the regulation of the maturation of nascent LepRb protein.We previously showed that SEL1L-HRD1 ERAD is required for the posttranslational maturation of POMC prohormone in mice on chow diet and that Sel1L deficiency in POMC neurons cause age-associate obesity in mice on chow diet due to the ER retention of POMC prohormone 39 .In DIO mouse models, we found defects in Sel1L POMC mice occurring upstream of POMC transcription as leptin-induced STAT3 phosphorylation is impaired in the absence of SEL1L-HRD1 ERAD [45][46][47][48][49] .Further mechanistic studies identify partial loss-of-function of LepRb resulted from attenuated ER exit of nascent LepRb in SEL1L-HRD1 ERAD deficient cells.This study suggests that nascent LepRb protein is likely misfolding prone in the ER, likely due to multiple glycosylation and the formation of disulfide bonds, and hence relies on SEL1L-HRD1 ERAD to generate an ER environment conducive for the proper folding and conformation of bystander LepRb.
Several human mutants have also been identified as SEL1L-HRD1 ERAD substrates that readily form aggregates and become resistant to and bypassing the quality control mediated by ERAD, leading to loss-of-function disease phenotype.These misfolded substrates with highly reactive cysteine thiols accumulate and promote the formation of inter-or intra-molecular disulfide-bonded aggregates [39][40][41] .Hence, SEL1L-HRD1 ERAD-mediated degradation of nascent unfolded and misfolded substrates, including LepRb in this study, may effectively prevent protein aggregation and maintain the folding environment in the ER.Efforts to target SEL1L-HRD1 ERAD function may represent a viable means for the treatment of certain diseases caused by a dominant-negative disease allele or a general collapse of the folding environment in the ER.

METHODS
Mice.As described previously 39 , POMC-specific Sel1L-deficient mice (Sel1L POMC ) and control littermates (Sel1L f/f ) were generated.The mice were further crossed with Pomc-eGFP reporter mice to generate Sel1L POMC ;POMC-eGFP and control littermates Sel1L f/f ;POMC-eGFP.WT B6 mice were purchased from JAX and bred in our mouse facility.Mice were fed a chow diet (13% fat, 57% carbohydrate and 30% protein, PicoLab Rodent Diet 5L0D) and placed on a high-fat diet (HFD, calories provided by 60% fat, 20% carbohydrate and 20% protein, Research Diet D12492) from 5 weeks of age for 1 week or 8 weeks.All mice were housed in a temperaturecontrolled room with a 12-hour light/12-hour dark cycle.
Food intake measurement and pair-feeding.Food intake were measured as previously described 39 .Briefly, to perform daily food intake measurement, mice were first acclimatized to single housing 24 hours before the experiment.Daily food intake was measured 1 hour before the onset of the dark cycle each day.For the pair-feeding at later stage of HFD feeding, Sel1L POMC and WT littermates had continuous free access to HFD for eight weeks and were then single housed and fed ~2.5 g, which was determined by the average of daily food intake of WT littermates, at the start of the dark cycle.For the pair-feeding at early stage of HFD feeding, 5-week-old Sel1L POMC mice were split into two groups: One group of Sel1L POMC and WT littermates had continuous free access to food; the other group of Sel1L POMC mice (pair-fed) was fed ~2.5 g at the start of dark hours.Weekly bodyweight gains were monitored.
Leptin treatment in mice.Twelve-week-old mice were intraperitoneally (i.p.) injected PBS followed by leptin (2 mg/kg body weight, R&D systems; catalog 498-OB-05M) 1 hour before the onset of dark cycle for three consecutive days as described 39 .Body weight and food intake were monitored daily during the treatment period.For phosphorylated STAT3 staining, 2 mg/kg leptin were i.p. injected to mice, followed by overnight fasting.Mice were anesthetized by isoflurane for fixation-perfusion 30 min after injection.
Tissue and blood collection.These procedures were carried out as previously described 39 .
Briefly, blood was collected from anesthetized mice via cardiac puncture, transferred to 1.5ml microcentrifuge tubes, kept at room temperature for 30 minutes prior to centrifugation at 2,000 g for 15 minutes.Serum was aliquoted and stored at -80°C until analysis.For brain microdissection, Adult Mouse Brain Slicer Matrix (BSMAA001-1, Zivic Instruments) was used to collect coronal brain slices containing ARC region with further microdissection to obtain ARCenriched region.All tissues were snap-frozen in liquid nitrogen and stored at -80°C before use.
All samples were incubated in 65°C for 10min and run with 15-30 μg total lysate on SDS-PAGE gel for separation followed by electrophoretic transfer to PVDF membrane (0.45μm, Millipore; catalog IPFL00010).The blots were incubated in 2% BSA/Tri-buffered saline tween-20 (TBST) with primary antibodies overnight at 4°C, washed with TBST followed by 1hr incubation with goat anti-rabbit or mouse IgG HRP at room temperature.Band density was quantitated using the Image Lab software on the ChemiDOC XRS+ system (Bio-Rad).

Immunostaining and antibodies.
For fluorescent immunostaining in free-floating brain sections, samples were picked out of anti-freezing buffer followed by 3 washes with PBS.Freefloating sections were simultaneously incubated with primary antibodies in blocking buffer (0.3% donkey serum and 0.25% Triton X-100 in 0.1 M PBS) overnight at 4°C.Following 3 washes with PBS, sections were incubated with secondary antibodies for 2 hours at room temperature.Brain sections were then mounted on gelatin-coated slides (Southern Biotech; catalog SLD01-CS).
Counterstaining and mounting were performed with mounting medium containing DAPI (Vector Laboratories; catalog H-1200) and Fisherfinest Premium Cover Glasses (Fisher Scientific; catalog 12-548-5P).For immunostaining in cells, 24 hours after transfection of LepRb-3xFLAG constructs, cells were placed on Poly-L-Lysine (Advanced Biomatrix; catalog 5048) coated Millicell EZ SLIDE 8-well glasses (Millipore; catalog PEZGS0816) for 24 hours before treatment and fixation.For staining surface bound leptin, samples were washed by ice cold PBS for 5 times and fixed by 4% formaldehyde (VWR; catalog 89370-094) for 15 minutes on ice followed by 3 washes with PBS.No permeabilization reagents were involved.For staining other markers, permeabilization was included and the overall process were the same as described above.To quantify immunoreactivity, identical acquisition settings were used for imaging each brain section from all groups within an experiment.The numbers of immunoreactivity-positive soma analysis and intensity of immunoreaction were quantified in 3D stack volumes after uniform background subtraction using the NIS Elements AR software (Nikon) and FIJI (National Institute of Health, USA).

Statistics.
Results are expressed as the mean ± SEM unless otherwise stated.Statistical analyses were performed in GraphPad Prism version 8.0 (GraphPad Software Inc.).
Comparisons between the groups were made by unpaired two-tailed Student's t test for two groups, or one-way ANOVA or two-way ANOVA followed by multiple comparisons test for more than two groups.P value < 0.05 was considered as statistically significant.All experiments were repeated at least twice and/or performed with several independent biological samples, and representative data are shown.

Study Approval.
All experiments performed with mice were in compliance with University of Michigan (Ann Arbor, MI) Institutional Animal Care and Use Committee (#PRO00006888) guidelines.(B) Growth curve of male Sel1L POMC mice fed with either NCD or HFD under ad libitum or pair feeding as indicated (n=3 mice per group, blue solid circles).Male Sel1L f/f mice fed ad libitum with the same diets were included as controls (n=3 mice per group, black open circles) (C) Growth of Sel1L POMC male mice with either ad libitum or pair-feeding of HFD starting at 5 weeks of age (n=3-5 mice per group).

FIGURES AND FIGURE LEGENDS
(D) Body weights of 12-week-old mice put on HFD (at day 0) followed by daily i.p. injected with vehicle (PBS) and leptin (2 mg/kg body weight) for 3 days (n=2 per group for male mice, indicated in dots; n=2-3 per group for female mice, indicated in squares).

Fig. 1 :
Fig. 1: Transient upregulation of SEL1L-HRD1 ERAD expression in the hypothalamus in response to high fat diet (HFD) feeding.(A) Quantitative PCR (qPCR) analysis of Sel1L and Hrd1 mRNA levels in the arcuate nucleus (ARC) of the C57BL/6J male mice fed on normal chow diet (NCD), 1w-and 8w-HFD (n=3-4 mice per group).(B-C) Representative Western blot of SEL1L and HRD1 in the ARC of the C57BL/6J male mice fed on NCD, 1w-and 8w-HFD, with quantitation shown on the right (n=13-15 mice per group).

( E -
F) Percentage of body weight change(E), average daily food intake (F) following 3 daily vehicle and leptin injections of the mice (n=2 per group for male mice, indicated in dots; n=2-3 per group for female mice).% Body weight is calculated based on the body weights at the end point over those at the starting point for each treatment.(G)Serum leptin levels in mice fed on NCD, 1w-and 8w-HFD (n=5-13 mice per group).Values, mean ± SEM. ns, not significant; *p<0.05,**p<0.01,***p<0.001and ****p<0.0001by two-way ANOVA followed by multiple comparisons test (A-G).

Fig. 4 :
Fig. 4: Hypothalamic SEL1L-HRD1 deficiency leads to DIO via leptin signaling.(A) Schematic diagram for parabiosis and pictures (right) of Sel1L f/f and Sel1L POMC female mice after parabiosis HFD for 8 weeks (n=3 pairs in group I, n=1 pair in group II, n=5 pairs in group III).(B-C) Body weights (B) of mice before and after parabiosis and body composition (C) after parabiosis following 8-week HFD for 8 weeks (n=6 mice in group I, n=2 mice in group II, n=5 mice per genotype in group III).(D-E) Serum leptin (D) and insulin (E) levels of mice after parabiosis HFD for 8 weeks (n=6 mice in group I, n=2 mice in group II, n=5 mice per genotype in group III).Values, mean ± SEM. ns, not significant; *p<0.05,**p<0.01,***p<0.001and ****p<0.0001by two-way ANOVA followed by multiple comparisons test (B-E).

Fig. 6 :
Fig. 6: The effect of POMC-specific ERAD in DIO is likely uncoupled from UPR and inflammation.(A) Representative Western blot for the PERK pathway of UPR in the ARC of Sel1L f/f and Sel1L POMC mice fed on 8w-HFD (n=6 mice per group with 3 male mice and 3 female), with quantitation shown on the right.Livers of mice treated with tunicamycin (TM, 1 mg/kg, i.p.) for 24 hours (Liver_TM) or not (Liver_CON), as well as lysates treated with Lambda protein phosphatase, included as controls.

Fig. 9 :
Fig.9: Proposed models for SEL1L-HRD1 ERAD degradation of wildtype LepRb and C604S disease mutant.In the basal conditions, SEL1L-HRD1 ERAD constitutively degrades misfolded LepRb and ensures the proper folding, maturation and surface expression of the LepRb.In the absence of ERAD, the accumulation of misfolded receptors forms aggregates, interferes with the folding and maturation of the nascent LepRb with attenuated surface display.In the context of recessive LepRb C604S mutant, though degraded by SEL1L-HRD1 ERAD, C604S LepRb readily forms aggregates to the extent beyond the capacity of ERAD, resulting in impaired maturation and surface display of the receptors.