1. Cell Biology
Download icon

CB1R regulates soluble leptin receptor levels via CHOP, contributing to hepatic leptin resistance

  1. Adi Drori
  2. Asaad Gammal
  3. Shahar Azar
  4. Liad Hinden
  5. Rivka Hadar
  6. Daniel Wesley
  7. Alina Nemirovski
  8. Gergő Szanda
  9. Maayan Salton
  10. Boaz Tirosh
  11. Joseph Tam  Is a corresponding author
  1. Obesity and Metabolism Laboratory, Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Israel
  2. Laboratory of Physiological Studies, National Institute on Alcohol Abuse & Alcoholism, United States
  3. MTA-SE Laboratory of Molecular Physiology, Department of Physiology, Semmelweis University, Hungary
  4. Department of Biochemistry and Molecular Biology, The Institute for Medical Research Israel-Canada, The Hebrew University of Jerusalem, Israel
  5. The Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Israel
Research Article
  • Cited 0
  • Views 418
  • Annotations
Cite this article as: eLife 2020;9:e60771 doi: 10.7554/eLife.60771

Abstract

The soluble isoform of leptin receptor (sOb-R), secreted by the liver, regulates leptin bioavailability and bioactivity. Its reduced levels in diet-induced obesity (DIO) contribute to hyperleptinemia and leptin resistance, effects that are regulated by the endocannabinoid (eCB)/CB1R system. Here we show that pharmacological activation/blockade and genetic overexpression/deletion of hepatic CB1R modulates sOb-R levels and hepatic leptin resistance. Interestingly, peripheral CB1R blockade failed to reverse DIO-induced reduction of sOb-R levels, increased fat mass and dyslipidemia, and hepatic steatosis in mice lacking C/EBP homologous protein (CHOP), whereas direct activation of CB1R in wild-type hepatocytes reduced sOb-R levels in a CHOP-dependent manner. Moreover, CHOP stimulation increased sOb-R expression and release via a direct regulation of its promoter, while CHOP deletion reduced leptin sensitivity. Our findings highlight a novel molecular aspect by which the hepatic eCB/CB1R system is involved in the development of hepatic leptin resistance and in the regulation of sOb-R levels via CHOP.

eLife digest

When the human body has stored enough energy from food, it releases a hormone called leptin that travels to the brain and stops feelings of hunger. This hormone moves through the bloodstream and can affect other organs, such as the liver, which also help control our body’s energy levels. Most people with obesity have very high levels of leptin in their blood, but are resistant to its effects and will therefore continue to feel hungry despite having stored enough energy.

One of the proteins that controls the levels of leptin is a receptor called sOb-R, which is released by the liver and binds to leptin as it travels in the blood. Individuals with high levels of this receptor often have less free leptin in their bloodstream and a lower body weight. Another protein that helps the body to regulate its energy levels is the cannabinoid-1 receptor, or CB1R for short. In people with obesity, this receptor is overactive and has been shown to contribute to leptin resistance, which is when the brain becomes less receptive to leptin. Previous work in mice showed that blocking CB1R reduced the levels of leptin and allowed mice to react to this hormone normally again, but it remained unclear whether CB1R affects how other organs, such as the liver, respond to leptin.

To answer this question, Drori et al. blocked the CB1R receptor in the liver of mice eating a high-fat diet, either by using a drug or by deleting the gene that codes for this protein. This caused mice to have higher levels of sOb-R circulating in their bloodstream. Further experiments showed that this change in sOb-R was caused by the levels of a protein called CHOP increasing in the liver when CB1R was blocked. Drori et al. found that inhibiting CB1R caused these obese mice to lose weight and have healthier, less fatty livers as a result of their livers no longer being resistant to the effects of leptin.

Scientists, doctors and pharmaceutical companies are trying to develop new strategies to combat obesity. The results from these experiments suggest that blocking CB1R in the liver could allow this organ to react to leptin appropriately again. Drugs blocking CB1R, including the one used in this study, will be tested in clinical trials and could provide a new approach for treating obesity.

Introduction

Leptin, predominantly produced by and secreted from white adipocytes, conveys information regarding the status of energy storage and availability to the brain to maintain energy homeostasis. It binds the leptin receptor in hypothalamic neurons to reduce food intake and increase energy expenditure in coordination with other adipokines and gastric peptides (Allison and Myers, 2014; Pan and Myers, 2018). Molecularly, leptin stimulates the secretion of α-melanocortin stimulating hormone (α-MSH) from proopiomelanocortin (POMC) neurons at the arcuate nucleus (ARC) and inhibits the secretion of the orexigenic peptides neuropeptide-Y (NPY) and Agouti-related protein (AgRP) (Flak and Myers, 2016). Genetic leptin deficiency or lack of functional leptin receptor results in morbid obese and insulin resistance phenotypes in animals (Lepob/ob or Leprdb/db mice, respectively) (Tartaglia et al., 1995; Zhang et al., 1994). In humans, congenital leptin deficiencies are rare, leading to hyperphagia and early-onset obesity, which can be reversed with a leptin replacement therapy (Mantzoros, 1999). However, most cases of obesity are characterized by hyperleptinemia, indicating that obesity is a leptin-resistant state, where leptin signaling is impaired.

Whereas many of the actions of leptin are attributed to its effects in the brain, it also has a broad range of physiological effects in the periphery such as angiogenesis, bone formation, lipid and carbohydrate metabolism, nutrient absorption, and insulin homeostasis (Sáinz et al., 2015). In fact, the lack of a response to leptin due to the development of resistance to the hormone may directly affect the central and peripheral actions of leptin, leading to a dysregulated energy balance. For instance, the liver, a central organ in the regulation of whole-body energy homeostasis, constitutes an important target for leptin as it regulates hepatic gluconeogenesis and insulin sensitivity as well as lipid metabolism (Frühbeck and Nutr, 2002). Therefore, defects in leptin action, which occur in a state of hepatic leptin resistance, impair hepatic function and lead to hyperglycemia, hyperinsulinemia, and dyslipidemia (Frühbeck and Nutr, 2002).

Various mechanisms have been linked to the development of diet-induced obesity (DIO)-related central and peripheral leptin resistance, including limited CNS access of leptin due to saturated transport machinery, uncoupling of leptin from its receptor (due to rare genetic mutations or intracellular modulators), leptin-induced downregulation of its hypothalamic receptor, and several circulating factors such as the soluble isoform of leptin receptor (sOb-R) (reviewed in Engin, 2017; Martin et al., 2008). Both in humans and mice, the leptin receptor gene (LEPR and Lepr, respectively) encodes four membrane-anchored isoforms, which differ in the length of their cytoplasmic tail. The long isoform, Ob-Rb, is considered to convey the most robust cellular response to leptin, while the shorter isoforms (Ob-Ra, Ob-Rc, and Ob-Rd) carry a weaker signal. In addition, sOb-R, which lacks the trans-membrane and intracellular domains, also exists. In humans, sOb-R is exclusively generated via proteolytic shedding of membrane-anchored isoforms (Maamra et al., 2001), whereas in mice, it is produced by both transcription of a designated isoform (Ob-Re) and ectodomain shedding of Ob-Rb and Ob-Ra (Ge et al., 2002; Li et al., 1998). sOb-R, mainly produced by hepatocytes, is the main leptin-binding protein in human plasma, regulating leptin’s bioavailability and bioactivity (Lammert et al., 2001; Yang et al., 2004). In fact, studies have shown that the circulating levels of sOb-R are inversely correlated with body weight and free leptin levels (Ogier et al., 2002). In addition, sOb-R levels are increased following weight loss (Laimer et al., 2002; Reinehr et al., 2005), and its overexpression in mice increases leptin sensitivity (Huang et al., 2001; Lou et al., 2010), supporting the key role of sOb-R in the development as well as the reversal of leptin resistance.

The endocannabinoid (eCB) system, a major regulator of energy homeostasis (Cristino et al., 2014; Fride et al., 2005; Pagotto et al., 2006; Ruiz de Azua and Lutz, 2019; Silvestri and Di Marzo, 2013), evokes various cellular/metabolic pathways via the activation of two G-protein-coupled receptors, cannabinoid type-1 (CB1R) and type-2 (CB2R) receptors, by the main eCBs, N-arachidonoylethanolamine (AEA) and 2-arachidonoylglycerol (2-AG). The eCB/CB1R system is highly overactive during obesity (Engeli, 2008; Engeli et al., 2005; Matias et al., 2006; Monteleone et al., 2005), and both central and peripheral stimulations of this system have been suggested to contribute to the development of the metabolic syndrome, including leptin resistance (Engeli et al., 2005; Matias et al., 2008; Pagotto et al., 2005). Studies have shown that leptin's ability to regulate food intake and peripheral lipid metabolism depends upon hypothalamic CB1Rs (Buettner et al., 2008; Cardinal et al., 2014; Di Marzo et al., 2001; Jo et al., 2005; Malcher-Lopes et al., 2006). Recent evidence demonstrates that peripheral CB1R signaling has the ability to modulate leptin activity too. By using peripherally restricted CB1R blockers, we have recently demonstrated that DIO-related hyperleptinemia is completely reversed by increasing leptin's renal clearance and decreasing its secretion from adipocytes (Tam et al., 2012; Tam et al., 2010). Additionally, we have shown that the reversal of hypothalamic leptin resistance in obese mice treated with the peripherally restricted CB1R blocker, JD5037, is mediated via re-sensitizing the animals to endogenous leptin and re-activating POMC neurons (Tam et al., 2017). Several lines of evidence suggest that hypothalamic neurons, including POMC, undergo endoplasmic reticulum (ER) stress during DIO, which may contribute to the development of leptin resistance (Ozcan et al., 2009; Ramírez and Claret, 2015). We have previously reported that pharmacological inhibition of peripheral CB1Rs (by AM6545) reverses the high-fat diet (HFD)-induced hepatic elevation in the ER stress marker phospho-eIF2α (Tam et al., 2010). Since ER stress strongly affects protein translation and secretion (reviewed in Ron and Walter, 2007), we hypothesized that the eCB/CB1R system plays a direct role in the regulation of sOb-R levels and hepatic leptin signaling involves the ER stress signaling pathway.

Results

Hepatic CB1R regulates sOb-R levels and leptin signaling

To evaluate the direct contribution of CB1R to the regulation of sOb-R levels, we first utilized a pharmacological inhibition paradigm of CB1R in DIO mice by using the peripherally restricted CB1R inverse agonist JD5037. Similar to previous findings (Mazor et al., 2018), a significant reduction in serum levels of sOb-R was documented in obese mice, an effect that was ameliorated by JD5037 treatment (Figure 1A). Since sOb-R is mainly produced by the liver (Lammert et al., 2001), we also analyzed the content of sOb-R in liver homogenates from these animals and found a similar trend as in serum (Figure 1B). Measurements of the Lepr-s (Ob-Re) isoform revealed that JD5037 treatment also affected its transcription and protein levels (Figure 1C–E). Moreover, the protein expression of two additional isoforms of LEPR (Ob-Rb and Ob-Ra) in liver homogenates was also decreased in DIO mice and normalized following JD5037 treatment (Figure 1D–E).

Figure 1 with 1 supplement see all
Hepatic CB1R regulates soluble isoform of leptin receptor (sOb-R) levels and leptin signaling in DIO.

Serum (A, n = 7–21) and liver (B, n = 6–9) levels of sOb-R are reduced following a 14 weeks consumption of high-fat diet (HFD) in wild-type (WT), but not LCB1 cKO mice. JD5037 (3 mg/kg, for 7 days) reverses the reduction in WT mice. The same trend observed in hepatic mRNA levels of Lepr-s (C, n = 3–12) as well as protein levels of Ob-Rb, Ob-Ra, and Ob-Re (D and E, n = 5–10; G and H, n = 8–10). Despite a comparable weight gain following HFD consumption in WT and LCB1 cKO mice (F, n = 4–20), obese mice that lack CB1R in hepatocytes remain leptin sensitive indicated by increased pSTAT3 levels (I–L, n = 3–4). Transgenic mice, expressing CB1R only in hepatocytes, are protected from DIO (M, n = 3–4). An exclusive hepatic overexpression of CB1R is sufficient for HFD feeding to induce reduction in serum, liver mRNA, and protein levels of sOb-R (N–Q, n = 4) as well as to reduce hepatic leptin sensitivity in response to exogenous leptin stimulation (R and S, n = 7). Data represent mean ± SEM of indicated number of replicates in each panel. The blots are representative. *p<0.05 relative to standard diet-fed animals from the same strain. #p<0.05 relative to HFD-fed mice from the same strain. ^p<0.05 relative to the same treatment group of WT mice.

To further establish the contribution of hepatic CB1R to the HFD-induced decrease in sOb-R levels, we utilized the liver-specific CB1R null (LCB1 cKO) mice, a genetic deletion model of mice that lacks CB1R specifically in hepatocytes (mouse model generation is described in Osei-Hyiaman et al., 2008). When fed with a HFD, these mice gain similar weight to their wild-type (WT) littermate controls [(Osei-Hyiaman et al., 2008) and Figure 1F]; however, they are less prone to develop liver steatosis, dyslipidemia, and leptin resistance (Osei-Hyiaman et al., 2008), making hepatic CB1R a central regulator of obesity-related liver complications. We were therefore not surprised to find that the liver specific deletion of CB1R was sufficient to maintain normal circulating levels of sOb-R in obese LCB1 cKO mice (Figure 1A). Similarly, the hepatic gene and protein expression of sOb-R and the other LEPR isoforms were not affected by the HFD feeding (Figure 1B,C and G,H), suggesting that hepatic CB1R most likely regulates sOb-R levels.

To test the functional relevance of our findings to hepatic leptin signaling, we measured the phosphorylation levels of STAT3, the gold-standard measure of leptin signaling (reviewed in Allison and Myers, 2014), in mouse livers following exogenous leptin administration in vivo. Whereas both lean WT and LCB1 cKO mice showed elevated pSTAT3/STAT3 ratio in response to leptin (Figure 1I,J), only obese LCB1 cKO mice remained leptin sensitive (Figure 1K,L). These results are in line with findings from Osei-Hyiaman and colleagues (Osei-Hyiaman et al., 2008), demonstrating that LCB1 cKO mice are resistant to obesity-induced hyperleptinemia.

Additional support for the regulation of sOb-R by hepatic CB1R derived from another transgenic mouse model (hTgCB1 cKO), in which CB1R is expressed only in hepatocytes (mouse model generation is described in Liu et al., 2012; Tam et al., 2010). These mice, despite being resistance to DIO like global CB1R KO mice [(Liu et al., 2012; Tam et al., 2010) and Figure 1M], demonstrate increased circulating leptin levels when fed a HFD (Liu et al., 2012). In accordance with that, the circulating and hepatic sOb-R levels in these mice were decreased by 50% following 14 weeks consumption of a HFD (Figure 1N–Q). Moreover, the hepatic pSTAT3/STAT3 ratio did not respond to exogenous leptin, suggesting reduced hepatic leptin sensitivity (Figure 1R,S). Hence, overexpression of CB1R in the liver alone compromises hepatic leptin sensitivity and recapitulates the HFD-induced downregulation of sOb-R observed in WT mice.

Next, we assessed whether a direct activation of CB1R in hepatocytes induces a reduction in sOb-R levels. To test this, we treated cultured hepatocytes with the synthetic CB1R agonist noladin ether (NE) for 24 hr. We analyzed both culture media and cell lysates and found that, similar to obesity, direct activation of CB1R also decreased sOb-R levels in the culture media. This was also the case with intracellular levels of other LEPR isoforms measured. This CB1R-mediated reduction in ObR levels was completely reversed by blocking CB1R using JD5037 (Figure 2).

Figure 2 with 1 supplement see all
CB1R directly regulates soluble isoform of leptin receptor (sOb-R) levels in hepatocytes.

24 hr treatment with the synthetic CB1R agonist noladin ether (NE, 2.5 μM) induced reduction in sOb-R levels in the culture media of immortalized hepatocytes (blot was quantified using Ponceau staining as a loading control). This was completely ameliorated by 1 hr pretreatment with 100 nM JD5037 (A and B, n = 12–13). Similar results were observed in both mRNA (C, n = 12–17) and protein (D and E, n = 4) levels in hepatocytes lysate (for Ob-Rb, the lower band was quantified). Data represent mean ± SEM of indicated number of biological replicates. Blots are representative. *p<0.05 relative to vehicle-treated cells. #p<0.05 relative to NE-treated cells.

CHOP contributes to the metabolic response to peripheral CB1R blockade

DIO-induced ER stress in the development of leptin resistance has been previously suggested (Ozcan et al., 2009; Ramírez and Claret, 2015). Similar to our previous findings (Tam et al., 2010), treatment of HFD-fed mice with JD5037 normalized p-eIF2α levels (Figure 3—figure supplement 1A,B), suggesting relieved ER stress following CB1R blockade. In agreement with these findings, a comparable ratio of hepatic phospho-to-total eIF2α ratio was documented in lean and obese LCB1 cKO mice (Figure 3—figure supplement 1C,D).

Measuring the expression levels of the ER stress marker C/EBP homologous protein (CHOP) revealed surprising findings, since both the hepatic mRNA (Ddit3) and protein levels of CHOP were downregulated in obese WT mice, despite the suggested ER stress. Its expression levels were reversed above control levels by JD5037, and remained comparable between lean and obese LCB1 cKO mice (Figure 3A–D). In fact, CHOP levels were positively correlated with the levels of sOb-R in both our experimental paradigms, leading us to hypothesize that CHOP may directly be involved in the regulation of sOb-R.

Figure 3 with 2 supplements see all
C/EBP homologous protein (CHOP) contributes to the metabolic benefits of peripheral CB1R blockade.

mRNA (A, n = 5–19) and protein (B–D, n = 4–6) levels of CHOP show reduced expression following 14 weeks on high-fat diet (HFD) in wild-type (WT), but not LCB1 cKO mice. JD5037 (3 mg/kg, for 7 days) treatment reverses the HFD-induced reduction in CHOP levels. Metabolic assessment of mice revealed diminished effect of JD5037 in CHOP KO mice. Weight (E, n = 10–23), fat and lean mass (F and G, n = 5–19), serum HDL, LDL as well as hepatic triglycerides (TG) (H–J, n = 9–26) were comparable in lean and obese WT and CHOP KO mice. JD5037 treatment was significantly more efficient in reducing weight, fat mass, LDL, and TG in WT mice. Data represent mean ± SEM of indicated number of replicates in each panel. *p<0.05 relative to standard diet-fed animals from the same strain. #p<0.05 relative to HFD-fed mice from the same strain. ^p<0.05 relative to the same treatment group of WT mice.

To test our hypothesis, we compared the metabolic efficacy of JD5037 in obese CHOP KO mice and their littermate controls. Whereas JD5037 was almost equieffective in reducing body weight and fat mass in both obese mouse strains (Figure 3E–G), it improved plasma cholesterol levels as well as hepatic steatosis in WT mice only (Figure 3H–J). The reduced ability of peripherally restricted CB1R blockade to improve dyslipidemia and hepatic steatosis in CHOP KO mice led us to measure the hepatic eCB 'tone' in these mice. Strikingly, we found that the basal levels of AEA and 2-AG were markedly higher in CHOP KO mice than in the WT control group. Moreover, the increased eCB levels in CHOP KO mice remained unchanged following a consumption of HFD as well as JD5037 treatment (Figure 3—figure supplement 2A,B). This could be partially explained by the differences documented in the mRNA expression patterns of fatty acid amide hydrolase (Faah), monoacylglycerol lipase (Mgll), N-acyl phosphatidylethanolamine phospholipase D (Napepld), and diacylglycerol lipase alpha (Dagla), the degrading and synthesizing enzymes of both eCBs, respectively (Figure 3—figure supplement 2C–F). Overall, these data indicate that CHOP may play a pivotal role in modulating hepatic eCB 'tone', and that it is required for the beneficial effects of CB1R blockade on dyslipidemia and hepatic steatosis.

CHOP plays a key role in the regulation of sOb-R by the eCB/CB1R system

Measuring the effect of CHOP deficiency on sOb-R levels revealed comparable circulating levels of sOb-R in lean and obese mice in the two mouse strains. However, JD5037 failed to restore sOb-R levels in CHOP KO mice (Figure 4A). The assessment of Lepr-s mRNA expression and sOb-R protein content in the livers of both strains documented reduced baseline levels in CHOP KO mice, compared to WT, which still remained low following HFD consumption and/or JD5037 treatment (Figure 4B,C). A similar trend was observed in the protein level of two more LEPR isoforms (Compare Figure 1D, E to Figure 4D,E). The HFD-induced hyperleptinemia was vastly reduced by JD5037 treatment in WT mice, whereas it was only partially ameliorated by JD5037 in CHOP KO animals (Figure 4F). Interestingly, the hepatic pSTAT3/STAT3 ratio in lean CHOP KO mice was comparable before and after stimulation with exogenous leptin (Figure 4G,H). Taken together, our data suggest that the regulation of sOb-R levels is CHOP-dependent. In addition, regulation of the soluble isoform by CHOP can consequently affect circulating leptin levels and hepatic leptin sensitivity, possibly, in a CB1R-dependent manner.

C/EBP homologous protein (CHOP) plays a key role in regulating soluble isoform of leptin receptor (sOb-R) by the endocannabinoid (eCB)/CB1R system.

Serum (A, n = 11–18) levels of sOb-R were reduced following a 14 weeks consumption of high-fat diet (HFD). JD5037 (3 mg/kg, for 7 days) reversed the reduction in wild-type (WT), but not CHOP KO mice. Basal hepatic levels of sOb-R were lower in CHOP KO mice and did not change following HFD or JD5037 treatment (B, n = 5–8). A similar trend was observed in hepatic mRNA levels (C, n = 9–17) and protein level of Ob-Rb, Ob-Ra, and Ob-Re (D and E, n = 5–6). Whereas DIO-related hyperleptinemia was comparable between WT and CHOP KO mice, JD5037 was more efficacious in reducing it in WT mice (F, n = 8–16). Lean CHOP KO mice failed to increase the hepatic pSTAT3/STAT3 ratio in response to exogenous leptin administration (G and H, n = 5). Western blots are representative. Data represent mean ± SEM of indicated number of replicates in each panel. *p<0.05 relative to standard diet-fed animals from the same strain. #p<0.05 relative to HFD-fed mice from the same strain. ^p<0.05 relative to the same treatment group of WT mice.

To further investigate this concept, we directly activated CB1R (with NE) in immortalized hepatocytes originated from WT or CHOP KO mice. Similar to a HFD consumption in mice (Figure 3A), a direct activation of CB1R downregulated CHOP mRNA expression (Figure 5A). We validated this by measuring the expression levels of Ppp1r15a, a downstream target of CHOP (Hu et al., 2018), and found that its expression was also reduced in NE-treated WT hepatocytes, and remained unchanged in CHOP KO cells (Figure 5B), suggesting that CB1R activation in fact leads to reduced CHOP expression and activity. Whereas NE was able to reduce sOb-R levels in WT hepatocytes, it had the opposite effect in CHOP KO hepatocytes, suggesting that CB1R may regulate sOb-R levels in other mechanisms independently of CHOP (Figure 5C–E).

C/EBP homologous protein (CHOP) regulates LEPR expression in hepatocytes.

In vitro, 24 hr treatment with noladin ether (NE; 2.5 μM) induced reduction in mRNA levels of Ddit3. (ND – not detected) (A), Ppp1r15a (B), and Lepr-s (C) in wild-type (WT), but not CHOP KO hepatocytes. A similar trend was observed in soluble isoform of leptin receptor protein levels secreted into the culture media of hepatocytes (blot was quantified using Ponceau staining as a loading control) (D and E). Data represent mean ± SEM of 8–20 biological replicates. Blots are representative. *p<0.05 relative to vehicle-treated cells in the same genotype. ^p<0.05 relative to same treatment paradigm in WT.

The consistent correlation between CHOP and sOb-R levels implies that CHOP is a positive regulator of Ob-Re. To validate this further, we analyzed Ob-Re levels in WT and CHOP KO hepatocytes treated with tunicamycin (TM), a potent inducer of ER stress. Treatment with TM for 6 hr led to an expected and robust expression of CHOP mRNA and protein in WT cells (Figure 6A,B). Importantly, this was accompanied with elevated mRNA expression levels of Lepr-s as well as secreted levels of sOb-R into the culture media in WT, but not CHOP KO hepatocytes (Figure 6C–E). Increased levels of sOb-R in culture media were also documented when we exogenously overexpressed myc-tagged CHOP in WT hepatocytes (Figure 6F,G), supporting a direct role for CHOP in Lepr gene regulation. By using a luciferase reporter assay, in which the −650 to +850 (relative to transcription start site) region of the LEPR promoter was cloned into firefly luciferase expressing vector, we found that CHOP expression and luciferase activity in transfected cells was induced using TM (Figure 6H), while CB1R activation using NE (which downregulates CHOP expression as seen in Figure 5A) had an opposite effect in WT, but not in CHOP KO cells. These data support the involvement of CHOP in CB1R-dependent regulation of sOb-R.

Figure 6 with 2 supplements see all
C/EBP homologous protein (CHOP) is a positive regulator of Lepr Promoter.

Induction of CHOP mRNA (A, n = 14) and protein (B, n = 3) expression using a 6 hr treatment with tunicamycin (TM; 2.5 µG/mL) was accompanied by elevated soluble isoform of leptin receptor (sOb-R) levels in the culture media (blot was quantified using Ponceau staining as a loading control) (C and D, n = 3) as well as mRNA levels (E, n = 8–21) of wild-type (WT) hepatocytes. A transient CHOP overexpression induced elevation in sOb-R levels in the culture media of WT hepatocytes (F and G, n = 3). Luciferase reporter assay (H, n = 14–16) and chromatin immunoprecipitation (ChIP) (I, n = 2–7) show increased Lepr promoter activity and CHOP binding to this promoter in WT hepatocytes treated with TM. Data represent mean ± SEM of indicated number of biological replicates. Blots are representative. *p<0.05 relative to vehicle-treated cells. ^p<0.05 relative to the same treatment group of WT hepatocytes.

To further explore the possibility that CHOP can directly bind the Lepr promoter and control its expression, we performed several chromatin immunoprecipitation (ChIP) experiments. In silico analysis of the Lepr promoter region revealed a putative binding site, corresponding to five of six nucleotides that compose a core sequence for CHOP binding (GRCm38:CM000997.2. Chromosome 4: 101,717,929–101,717,934) (Ubeda et al., 1996). As seen in the CHOP precipitates (Figure 6I), there was a twofold increase in the recovery of the qPCR product amplified with a primer set flanking the putative CHOP binding site, in cells that were treated with TM. A similar enrichment was seen in Ppp1r15a (GADD34), a well-known target of CHOP. This increase was limited to WT hepatocytes, validating the specificity of CHOP IP. Taken together, our data suggests that CHOP is able to occupy the Lepr promoter and directly regulate sOb-R levels in response to HFD consumption and/or CB1R activation.

The molecular signaling pathway(s) by which eCBs/CB1R regulates CHOP levels calls for further investigation. Nevertheless, one putative mechanism may involve Trib3, a multifunctional protein upregulated during ER stress by the PERK-ATF4-CHOP pathway, which mediates cell death. Trib3 represses its own expression by inhibiting the transcription of both ATF4 and CHOP (Jousse et al., 2007; Mathur et al., 2014; Ohoka et al., 2005). In addition, many studies describe Trib3 as a key factor in mediating the anti-tumor effect of cannabinoids (reviewed in Velasco et al., 2016). Our in vivo data indicate that HFD induces the mRNA and protein expression levels of hepatic Trib3, and that treatment with JD5037 restores these levels. This effect is limited to WT mice, whereas in CHOP KO mice, Trib3 levels did not change in response to HFD nor JD5037 treatment. Similarly, a direct activation of CB1R using NE upregulated Trib3 expression in WT, but not in CHOP KO hepatocytes (Figure 6—figure supplement 1A–E), suggesting that Trib3 is indeed induced via CB1R signaling, and negatively regulates CHOP levels. Further support for this hypothesis comes from our data in Figure 6—figure supplement 2, where we show that ATF4 protein levels are reduced in WT mice following the consumption of HFD, and are normalized by JD5037, whereas they remain unchanged in lean and obese LCB1 cKO mice (Figure 6—figure supplement 2B–D). Moreover, the ATF4 levels were reduced in hTgCB1 cKO mice fed with a HFD, as compared to lean STD-fed mice (Figure 6—figure supplement 2E–G). Overall, the CB1R-related changes in ATF4 expression were found to be well correlated with CHOP as well as with the sOb-R levels, placing ATF4 downstream of CB1R and upstream of CHOP in this cascade.

Discussion

Since only free leptin crosses the blood–brain barrier (BBB) and induces leptin signaling, the sOb-R, which sequesters free leptin in the serum and is considered as the main binding protein for leptin in the circulation, practically regulates leptin’s bioavailability and activity and can potentially affect leptin sensitivity/resistance. This is also true for peripheral tissues, where sOb-R/leptin complexes cannot bind to and activate membrane anchored leptin receptors. Many human and animal studies have demonstrated that sOb-R levels are inversely correlated with plasma levels of leptin, BMI, and adiposity (Chan et al., 2002; Lahlou et al., 2000; Laimer et al., 2002; Ogier et al., 2002; Reinehr et al., 2005), suggesting that low levels of the soluble isoform contribute to obesity-related hyperleptinemia and subsequently, leptin resistance. In contrast to pathological conditions with a positive energy balance (i.e. obesity), human clinical situations associated with energy deficiency (i.e. starvation and/or anorexia nervosa) are characterized by upregulated circulating levels of sOb-R (Monteleone et al., 2002; Reinehr et al., 2005; Stein et al., 2006; Zepf et al., 2012). Moreover, individuals carrying a mutated allele of LEPR, which leads to enhanced shedding of the leptin binding domain, have normoleptinemia and they are not obese (Lahlou et al., 2002). For these reasons, the sOb-R most likely plays a key role in the formation of central and peripheral leptin resistance conditions. Yet, only limited knowledge exists about the molecular mechanisms that regulate sOb-R production and secretion.

Using multiple cultured cell types, Gan and colleagues have shown that TNFα may induce cell surface expression of Ob-Rb as well as sOb-R levels (Gan et al., 2012). In addition, an in vitro study has demonstrated that increasing the concentration of recombinant sOb-R diminishes STAT3 phosphorylation in response to leptin stimulation, but pre-incubation of leptin with recombinant sOb-R forms ligand-receptor complexes do not affect leptin-mediated STAT3 phosphorylation (Yang et al., 2004). In vivo, it has been described that leptin stimulation as well as food deprivation specifically induce the expression of sOb-R in mouse liver (Cohen et al., 2005). It has been also demonstrated that in contrast to mice, the human sOb-R, exclusively generated through proteolytic cleavage of the extracellular domain of membrane-anchored isoforms (Maamra et al., 2001), is shed into the circulation by two well-known proteolytic enzymes, ADAM10 and ADAM 17, belong to the 'ADAM's family' (reviewed in Schaab and Kratzsch, 2015). As we could not detect significant up/down regulation in the expression levels of these proteins in our experimental paradigms (Figure 2—figure supplement 1), and the fact that we detected all changes in both mRNA and protein levels, we reasoned that the observed alterations in the circulating levels of sOb-R result from altered hepatic expression and secretion of the Lepr gene rather than decreased shedding. In fact, the regulation of ADAM10 and ADAM17 is complex and involves transcription, dynamic trafficking, cellular localization, and activity (Edwards et al., 2008; Reiss and Bhakdi, 2017). Whereas the current study is focused on the transcriptional regulation of sOb-R by CB1R, other Ob-R isoforms, also expressed in humans, display (at the gene and protein levels) a trend similar to Ob-Re following either CB1R activation or blockade. These isoforms (Ob-Ra, Ob-Rb) serve as substrates for ectodomain shedding; therefore, their transcriptional regulation may indirectly influence the sOb-R levels and be relevant to human physiology. In addition, numerous stimuli, such as activation of protein kinase C, an increase in intracellular calcium, lipotoxicity, and apoptosis, may contribute to the proteolytic cleavage of the extracellular leptin receptor domain (Maamra et al., 2001; Schaab et al., 2012). However, to the best of our knowledge, a molecular mechanism responsible for the decreased expression/shedding of sOb-R in obesity was never reported. Here we describe, for the first time, the involvement of the eCB/CB1R system in regulating sOb-R levels and consequently leptin's activity.

The importance of the eCB/CB1R system in regulating normal energy homeostasis as well as mediating obesity-related comorbidities is well acknowledged (review in Simon and Cota, 2017). In fact, its pivotal interaction with leptin has been first described in 2001, demonstrating that leptin reduces the content of hypothalamic eCBs (Di Marzo et al., 2001) and attenuates eCB-mediated ‘retrograde' neuronal CB1R signaling (Jo et al., 2005; Malcher-Lopes et al., 2006). On the other hand, activating CB1R by eCBs may, in turn, regulate leptin levels and signaling, as suggested previously in women with anorexia nervosa, whose AEA levels are elevated (Monteleone et al., 2005), whereas their leptin levels are reduced. In fact, we have previously shown that CB1R activation in adipocytes and pre-junctional sympathetic fibers innervating the adipose tissue stimulates leptin biosynthesis and release, and its activation in the proximal tubules of the kidney inhibits leptin degradation and renal clearance (Tam et al., 2012), thus possibly contributing to leptin resistance. In agreement with these findings, peripheral CB1R blockade has been shown to ameliorate obesity-related hyperleptinemia, and subsequently restores leptin sensitivity in obese mice (Tam et al., 2012). By using both pharmacological and genetic approaches that target hepatic CB1R, our findings here suggest another novel mechanism by which the eCB system may regulate hepatic leptin resistance. Specifically, peripheral blockade and hepatic deletion/overexpression of CB1R modulate the expression levels of the sOb-R isoform in hepatocytes and its subsequent release into the circulation, reversing the CB1R-mediated decrease in sOb-R levels and hepatic leptin resistance during obesity. One should point out that although liver-specific CB1R KO mice retained higher levels of circulating sOb-R when fed a HFD, they were equally susceptible to DIO as their WT controls. Similarly, hepatic-specific CB1R transgenic mice in the CB1R KO background remained resistant to DIO while displayed significantly lower circulating sOb-R levels, as compared to their littermates. These data suggest that while liver CB1R expression is a major contributor to circulating sOb-R levels, their roles in regulating systemic/central energy balance will need to be further validated. Nevertheless, the contribution of hepatic CB1R to the regulation of hepatic leptin resistance was clearly demonstrated here by showing that DIO leads to a loss of leptin sensitivity in WT animals, but not in liver-specific CB1R null obese mice. In line with these findings, overexpression of CB1R in hepatocytes of lean mice inhibited leptin-induced STAT3 phosphorylation. These results, linking CB1R with hepatic leptin signaling, may significantly advance our understanding of CB1R’s role in modulating hepatic gluconeogenesis and insulin sensitivity as well as lipid metabolism. Indeed, genetic deletion of CB1R in hepatocytes partially protects mice from developing DIO-related hepatic steatosis, hyperglycemia, dyslipidemia, and insulin resistance (Osei-Hyiaman et al., 2008), whereas its overexpression in hepatocytes contributes to insulin resistance via inhibition of insulin signaling and clearance (Liu et al., 2012). In this sense, given that leptin regulates lipid, glucose, and insulin homeostasis in the liver, and that these metabolic functions are impaired in rodent models of increased eCB/CB1R ‘tone’, a role of CB1R-induced hepatic leptin resistance in regulating these processes can be postulated.

In accordance with our findings, Palomba and colleagues reported that CB1R activation interferes with leptin's activity in hypothalamic ARC neurons (Palomba et al., 2015). On the other hand, opposite findings were reported by Bosier and colleagues, demonstrating that pharmacological or genetic deletion of CB1R in astrocytes downregulates Ob-Rb expression and leptin-mediated functional responses, whereas JZL195 (a dual MAGL and FAAH inhibitor) upregulates these features (Bosier et al., 2013). These differences can be explained by the distinct roles hepatocytes, astrocytes, and neurons play in peripheral and central metabolic regulations, and by cell-specific roles for CB1R in this regulation. As our findings demonstrate a similar effect of hepatic CB1R activation/overexpression or blockade/deletion on the different isoforms of LEPR, it seems equally possible that hepatic CB1R may affect DIO-related hepatic leptin resistance by not only modulating sOb-R levels, which controls leptin’s activity, but also by modulating the expression of Ob-Rb in hepatocytes. Further investigations would allow us to differentiate between these two pathways.

Obesity is often characterized by an ER stress and consequently an adaptive unfolded protein response (UPR), operated by three parallel sensors: activating transcription factor 6 (ATF6), inositol requiring enzyme 1α (IRE1α), and protein kinase R-like ER kinase (PERK) (Walter and Ron, 2011). The activation of the latter induces the phosphorylation of eIF2α, which, in turn, inhibits transcription and protein synthesis (Ron, 2002). In case of an extreme ER stress conditions, CHOP is activated by the PERK signaling pathway and executes ER stress-mediated apoptosis (Hu et al., 2018; Zinszner et al., 1998). In fact, ER stress has been shown to contribute to the development of hypothalamic leptin resistance, by impairing the transport of leptin across the BBB and suppressing STAT3 phosphorylation (El-Haschimi et al., 2000; Hosoi et al., 2008; Ozcan et al., 2009; Zhang et al., 2008). In addition, under physiological conditions, excess nutrients increases the demand for protein synthesis by the liver, leading to ER stress and UPR activation, which resolves the stress within hours (Oyadomari et al., 2008). Nevertheless, chronic ER stress in the liver was demonstrated in both obese mice and humans (Ozcan et al., 2004; Puri et al., 2008). Here we demonstrate that obese mice had elevated levels of phosphorylated eIF2α, indicating increased ER stress, an effect that was reversed by peripheral CB1R blockade and was absent in LCB1 cKO. These findings are in agreement with our previous reports, where we reported that a neutral CB1R antagonist (AM6545) has the ability to reduce the HFD-induced upregulation in hepatic eIF2α (Tam et al., 2010), and that hepatic activation of CB1R induces ER stress and contributes to insulin resistance (Liu et al., 2012). Unexpectedly, we found that the hepatic gene and protein levels of CHOP and its upstream regulator ATF4 were significantly decreased in obese mice, and were upregulated by peripheral CB1R blockade. This observation, although counterintuitive, is conceptually in agreement with several previous reports that describe a modulated UPR signaling with altered sensitivity or output that might implicate conditions of persistence/repeated stress (Chambers et al., 2012; Gomez and Rutkowski, 2016; Preissler et al., 2015; Yang et al., 2015).

Apart from its role in ER stress-mediated apoptosis, CHOP has been implicated in regulating other processes such as inflammation (Endo et al., 2006; Nakayama et al., 2010), insulin resistance (Maris et al., 2012; Song et al., 2008), and adiposity. Specifically in the liver, Chikka and colleagues suggest that CHOP is a suppressor of key regulators of lipid metabolism like Cebpa, Ppara, and Srebf1 (Chikka et al., 2013), and demonstrate that CHOP-deficient mice tend to develop hepatic steatosis in response to bortezomib-induced ER stress. This is in agreement with an earlier report describing higher body weight and adiposity in female CHOP KO mice compared to WT controls (Ariyama et al., 2007). In contrast, we show that male CHOP KO mice and their WT littermate controls gain comparable amount of weight and have similar body composition following exposure to an HFD for 14 weeks. Nevertheless, we did see a trend toward increased liver triglycerides.

An interesting observation was that the eCB ‘tone’ of lean and obese CHOP KO mice was comparable. To the best of our knowledge, a direct regulation of eCB synthesis or degradation by CHOP has never been described. Thus, our data imply a possible link between the two. In fact, it is possible that the higher basal eCB levels seen in CHOP KO mice in comparison with their littermate controls are the consequence of leptin’s reduced ability to inhibit AEA and 2-AG production. This may be due to the reduced level of sOb-R found in the liver of CHOP KO animals. This hypothesis, although not tested here and which needs further experimental corroboration, is in accordance with the findings of others who demonstrated such a mechanism in the hypothalamus and in adipocytes treated with leptin (Di Marzo et al., 2001; Matias et al., 2006). In addition, JD5037 failed to reverse many of the metabolic abnormalities, such as HDL and LDL content as well as liver triglycerides in DIO CHOP KO mice. It also had much smaller effect on total body fat mass then in WT DIO mice, suggesting an obligatory role of CHOP in mediating the metabolic improvements induced by CB1R blockade. Whereas basal circulating levels of sOb-R were comparable between WT and CHOP KO mice, its levels were markedly lower in the liver of CHOP KO mice as well as cultured hepatocytes. Moreover, sOb-R remained low in these mice even on HFD, supporting a role for CHOP in regulating the synthesis of sOb-R in the liver. The failure of JD5037 to elevate sOb-R levels in obese CHOP KO mice places CHOP downstream of CB1R in this molecular cascade.

As mentioned earlier, the molecular signaling pathway(s) by which eCBs or CB1R regulates CHOP levels is outside the scope of this work. However, two possible mechanisms might be relevant. The first putative mechanism may involve Trib3, which is known to be induced in a broad range of cells and in response to multiple forms of cellular stress such as ER stress, excess of free fatty acids, oxidative stress, hypoxia, hyperglycemia, and toxins (reviewed in Ord and Ord, 2017). Interestingly, it has been shown that Δ9-THC as well as synthetic cannabinoid agonists upregulate Trib3 expression (Blázquez et al., 2006; Carracedo et al., 2006; Salazar et al., 2013; Vara et al., 2011) to engage apoptosis in variable cancer models. Moreover, Cinar et al. demonstrated that hepatic CB1R induces ER stress in hepatocytes by increasing de novo synthesis of ceramides (Cinar et al., 2014), which are also involved in Trib3 upregulation following ER stress (Carracedo et al., 2006). In line with our observation that Trib3 levels are negatively correlated with ATF4, CHOP, and sOb-R and are elevated following direct or indirect activation of CB1R, we suggest that Trib3 is a molecular linker between CB1R and the ATF4/CHOP complex. In fact, Trib3 directly interacts with and inhibits ATF4 and CHOP, forming a negative feedback on the regulation of their activity (Ohoka et al., 2005). This Trib3-induced negative modulation of ATF4 and CHOP has been suggested to contribute to the fine-tuning of ATF4- and CHOP-dependent transcription in stressed cells, such as hepatocytes exposed to fatty acid flux. With the extensive body of evidence demonstrating a wide range of molecules that are known to regulate Trib3 expression, our current findings highlight CB1R as a novel possible modulator that regulates Trib3 transcription, thus suggesting that CB1R activation may disrupt the ER stress signaling pathway involving eIF2α, ATF4, and CHOP. However, the molecular events that link CB1R and Trib3 require further assessment, and direct regulation of ATF4 by CB1R cannot be excluded. Second possible mechanism has to do with CHOP being a cAMP responsive protein, so its expression is induced via a cAMP response element (CRE) (Conkright et al., 2003; Pomerance et al., 2003; Ramji and Foka, 2002; Wilson and Roesler, 2002). CB1R, a G-protein coupled receptor (GPCR), which upon activation recruits Gi protein, can inhibit the activity of adenylyl cyclase and reduce the levels of cAMP (Turu and Hunyady, 2010). It is therefore plausible that a decline in cAMP following CB1R activation inhibits CHOP transcription. This hypothesis is more appealing if one considers the pivotal role of cAMP in regulating liver metabolism (Wahlang et al., 2019), and takes into account the fact that reduced levels of cAMP were documented in HFD-fed mice (Zingg et al., 2017). Yet, further studies will need to explore the specific molecular pathways linking together hepatic eCB/CB1R system and CHOP.

In conclusion, we report a new role for the hepatic eCB/CB1R in the development of hepatic leptin resistance, by reducing the expression and/or subsequent release of sOb-R (Figure 7). We show that peripherally restricted CB1R antagonism has the ability to restore sOb-R levels, contributing to the reversal of obesity-induced hyperleptinemia. We also suggest that upon CB1R blockade in hepatocytes, ATF4 as well as CHOP levels are upregulated via reduced Trib3 expression. CHOP, in turn, directly binds the LEPR promoter and promotes the expression of sOb-R.

An illustration that describes the suggested molecular mechanism involving CB1R and C/EBP homologous protein (CHOP) in the regulation of soluble isoform of leptin receptor (sOb-R) levels.

(Left) When overexpressed in the liver, activated by endocannabinoids during diet-induced obesity (DIO) or synthetic cannabinoids, such as noladin ether (NE, green squares), CB1R attenuates cAMP (pink circles) production by inhibiting adenylate cyclase (AC). CB1R also represses the expression of ATF4, as well as upregulates Trib3 expression. As a consequence, CHOP levels are reduced, transcribing less Lepr. (Right) Blocking CB1R in hepatocytes reverses these changes, leading to the activation and translocation of CHOP to the nucleus, which, in turn, directly binds the Lepr promoter and promotes the expression of sOb-R. Red arrows represent downregulation, whereas green arrows represent upregulation. Colored triangles represent activation.

Materials and methods

Key resources table
Reagent type
(species) or resource
DesignationSource or referenceIdentifiersAdditional information
Gene (Mus musculus)Lepre!EnsemblENSMUSG00000057722Leptin receptor
Strain, strain background (Mus musculus)Wild typeEnvigo IsraelC57Bl/6N
Strain, strain background (Mus musculus)LCB1 cKOOsei-Hyiaman et al., 2008Liver-specific CB1R KO
Strain, strain background (Mus musculus)hTgCB1 cKOTam et al., 2010CB1R KO, overexpressing CB1R in hepatocytes
Strain, strain background (Mus musculus)CHOP KOThe Jackson LaboratoryB6.129S(Cg)-Ddit3tm2.1Dron/J,
#005530
RRID:IMSR_JAX:005530
Transfected construct (Mus musculus)mCHOP-9E10AddgeneCHOP6: mCHOP-WT-9E10-pCDNA, #21913 RRID:Addgene_21913
Transfected construct (Firefly,sv40)pGL3PromegapGL3-basic vector,
E1751
Luciferase Assay vector
Transfected construct (Mus musculus)Ad-GFP-mLEPRVECTOR BIOSYSTEMS IncADV-263380
Cell line (Mus musculus)Wild typeUzi et al., 2013Immortalized mouse hepatocytes
Cell line (Mus musculus)CHOP KOUzi et al., 2013Immortalized mouse hepatocytes
AntibodyLEPR (rabbit polyclonal)NovusNB-120–5593
RRID:AB_791038
WB (1:2000)
AntibodypSTAT3 (rabbit monoclonal)Cell signaling#9145
RRID:AB_2491009
Phosphorylated Stat3 (Tyr705),
WB (1:1000)
AntibodySTAT3 (mouse monoclonal)Cell signaling#9139
RRID:AB_331757
WB (1:3000)
Antibodyp-eIF2α (rabbit polyclonal)Cell signaling#9721
RRID:AB_330951
Phosphorylated eIF2α (Ser51),
WB (1:1000)
Antibodyt-eIF2α (rabbit polyclonal)Cell signaling#9722
RRID:AB_2230924
Total eIF2α WB (1:1000)
AntibodyATF4 (rabbit monoclonal)Cell signaling#11815
RRID:AB_2616025
WB (1:1000)
AntibodyCHOP (mouse monoclonal)Cell signaling#2895S
RRID:AB_2089254
WB (1:1000)
ChIP (2.5 µg/sample)
AntibodyTrib3 (rabbit polyclonal)Abcamab137526
RRID:AB_2876352
WB (1:2000)
Antibodyβ-Actin (mouse monoclonal)Abcamab49900 RRID:AB_867494WB (1:30,000)
AntibodyH3 (rabbit polyclonal)Abcamab1791
RRID:AB_302613
ChIP (2.5 µg/sample)
AntibodyAnti-rabbit HRP (donkey polyclonal)Abcamab97085
RRID:AB_10679957
WB (1:10,000)
AntibodyAnti-mouse HRP (donkey polyclonal)Abcamab98799
RRID:AB_10675068
WB (1:10,000)
Commercial assay or kitSLR ELISAShanghai Bluegene BiotechE03S0226
Commercial assay or kitTriglyceride Assay KitAbcamab65336
Commercial assay or kitDual-Glo Luciferase Assay SystemPromegaE2920
Chemical compound, drugJD5037Haoyuan Chemexpress Co., LtdHY-18697
Chemical compound, drugNECayman Chemicals621652-Arachidonyl glycerol ether
Chemical compound, drugTunicamycin
(TM)
Holland Moran11089-65-9
Software, algorithmGraphPad PrismGraphPad SoftwareRRID:SCR_002798

Animals and experimental protocol

Request a detailed protocol

All animal studies were approved by the Institutional Animal Care and Use Committee of the Hebrew University of Jerusalem (AAALAC accreditation #1285; Ethic approval numbers MD-14–14008 and MD-19–15951). Animal studies are reported in compliance with the ARRIVE guidelines (NC3Rs Reporting Guidelines Working Group et al., 2010), and are based on the rule of the replacement, refinement, or reduction. All the animals used in this study were housed under specific pathogen‐free (SPF) conditions, up to five per cage, in standard plastic cages with natural soft sawdust as bedding. The animals were maintained under controlled temperature of 22–24°C, humidity at 55 ± 5%, and alternating 12 hr light/dark cycles (lights were on between 7:00 and 19:00 hr), and provided with food and water ad libitum. C57Bl/6 (Envigo, Israel), LCB1 cKO, and hTgCB1 cKO (kindly provided by Dr. George Kunos, NIH) or B6.129S(Cg)-Ddit3tm2.1Dron/J (CHOP KO, The Jackson Laboratory #005530), and their WT littermate controls were used for in vivo experiments. All mice were male and 8–10 weeks old at the beginning of each experiment. To generate DIO (body weight >42 g), mice were fed with a standard diet (STD; 14% Kcal fat, 24% Kcal protein, 62% Kcal carbohydrates; NIH-31 rodent diet) or a HFD (60% Kcal fat, 20% Kcal protein, and 20% Kcal carbohydrates; Research Diet, D12492) for 14 weeks. Then, obese mice were randomly divided into the experimental groups. Treatment with JD5037 (3 mg/kg, ip) or vehicle (1% Tween80, 4% DMSO, 95% Saline) was conducted for 7 days, and 12 hr following the last dose, the mice were euthanized by a cervical dislocation under anesthesia, and blood and livers were harvested for further analyses. For leptin sensitivity test, mice were fasted for 24 hr before an ip administration of recombinant mouse leptin (3 mg/kg). One hour following leptin administration, mice were euthanized and livers were harvested and processed for phosphorylated STAT3 detection using western blot.

Cell culture

Request a detailed protocol

WT or CHOP KO immortalized hepatocytes (described in Uzi et al., 2013), confirmed to be mycoplasma-negative, were maintained in DMEM (01-100-1A; Biological Industries, Israel) supplemented with 5% FCS, 100 mM glutamine, 100 mM Na-Pyruvate, and Pen/Strep. Cells were cultured at 37°C in a humidified atmosphere of 5% CO2/95% air. To test the effect of CB1R activation, cells were seeded in 6-well plates (25 × 104 cells/well) for 24 hr. Then, growth medium was replaced with a serum-free medium (SFM) for an additional 12 hr. At the morning of the experiment the medium was replaced with fresh SFM containing either vehicle (EtOH), 2.5 µM NE (Cayman Chemicals, Ann Arbor, Michigan) or a combination of 100 nM JD5037 (Haoyuan Chemexpress Co., Ltd) and 2.5 µM NE. After 24 hr, cells were harvested for further analyses as described below.

Measurements of sOb-R

Request a detailed protocol

Soluble leptin receptor was quantified by an ELISA kit, capable to differentiate the soluble isoform from other isoforms, according to manufacturer’s instructions (E03S0226; Shanghai Bluegene Biotech, China). Briefly, for serum measurements, we diluted serum in saline (1:2) and 100 µL from the diluted samples were analyzed. For hepatic measurements, 50–100 mg tissue samples were homogenized in 300 µL of 1× PBS and centrifuged for 5 min in 5000 rpm; 100 µL of cleared lysates were analyzed. Data were normalized to sample protein content, determined using the Pierce BCA Protein Assay Kit (Thermo Scientific, IL).

To measure sOb-R protein content in cell culture media, we used trichloroacetic acid (TCA) precipitation protocol; 350 µL of 100% TCA were added to 1.6 mL culture media, vortexed, and incubated for 30 min on ice. Samples were then centrifuged to pellet proteins (14,000 rpm, 10 min, 4°C). Pellets were washed in 100% acetone, resuspended in 0.1 M NaOH and protein loading dye, and analyzed by western blot. Ponceau staining of the blots was used as loading control for quantification.

Validation of LEPR antibody. The specificity of the anti-LEPR antibody was validated in a control experiment (Figure 1—figure supplement 1), where mouse LEPR was overexpressed in kidney cell line by using a viral infection. The viral vector encoded Ad-GFP-mLEPR (ADV-263380, VECTOR BIOSYSTEMS Inc) was used in a multiplicity of infection of 50, and cells were harvested for western blot analysis 24 hr post infection.

Real-time PCR

Request a detailed protocol

For total mRNA isolation, tissue samples or hepatocytes were washed in 1× PBS and harvested using Bio-Tri RNA lysis buffer (Bio-Lab, Israel). Extracted RNA was treated with DNase I (Thermo Scientific, IL), and reverse transcribed using the Iscript cDNA kit (Bio-Rad Laboratories, CA). Quantitative PCR reactions for Lepr-s, Ddit3, or Ppp1r15a were performed using iTaq Universal SYBR Green Supermix (Bio-Rad Laboratories, CA), and the CFX connect ST system (Bio-Rad Laboratories, CA). Relative quantity (RQ) values of all tested genes were normalized to Ubc. Primers are listed in Supplementary file 1.

Western blot analysis

Request a detailed protocol

Tissue samples or hepatocytes were washed in cold 1× PBS, and harvested in a RIPA buffer (25 mM Tris-HCl pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS), vortexed and incubated for 30 min at 4°C, and then centrifuged for 10 min at 14,000 rpm. Protein concentrations were determined using the Pierce BCA Protein Assay Kit (Thermo Scientific, IL). Cleared lysates were supplemented with protein sample buffer, resolved by SDS-PAGE (4–15% acrylamide, 150 V) and transferred to PVDF membranes using the Trans-Blot Turbo Transfer System (Bio-Rad Laboratories, CA). Membranes were incubated for 1 hr in 5% milk (in TBS-T) to block unspecific binding, washed briefly, and incubated overnight at 4°C with the following primary antibodies: LEPR (NB-120–5593, Novus), phosphorylated STAT3 (9145, Cell Signaling), STAT3 (9139, Cell Signaling), phosphorylated eIF2α (9721, Cell Signaling), eIF2α (9722, Cell Signaling), ATF4 (11815, Cell Signaling), CHOP (2895S, Cell Signaling), Trib3 (ab137526, Abcam), or β-Actin (ab49900, Abcam). Anti-rabbit (ab97085, Abcam) or mouse (ab98799, Abcam) horseradish peroxidase (HRP)-conjugated secondary antibodies were used for 1 hr at room temperature, followed by chemiluminescence detection using Clarity Western ECL Blotting Substrate (Bio-Rad Laboratories, CA). Densitometry was quantified using ImageLab software. Protein RQ was calculated as the ratio between LEPR to total protein signal (ponceau) in culture media supernatants and to β-actin in cell and tissue lysates.

Body composition and biochemical analysis

Request a detailed protocol

Total body fat and lean masses were determined by EchoMRI-100H (Echo Medical Systems LLC, Houston, TX, USA).

HDL and LDL measurements were done using the Cobas C-111 chemistry analyzer (Roche, Switzerland).

Hepatic triglycerides measurements

Request a detailed protocol

Tissue lipids were extracted as described in Folch et al., 1957, and quantified using Triglyceride Assay Kit (ab65336; Abcam). Data were normalized to tissue weight.

eCB measurements by LC-MS/MS

Request a detailed protocol

eCBs were extracted, purified, and quantified in liver homogenates, as described previously (Drori et al., 2019; Udi et al., 2017). LC-MS/MS was analyzed on an AB Sciex (Framingham, MA, USA) Triple Quad 5500 mass spectrometer coupled with a Shimadzu (Kyoto, Japan) UHPLC System. eCBs were detected in a positive ion mode using electron spray ionization (ESI) and the multiple reaction monitoring (MRM) mode of acquisition. The levels of each compound were analyzed by monitoring multiple reactions. The molecular ion and fragment for each compound were measured as follows: m/z 348.3→62.1 (quantifier) and 91.1 (qualifier) for AEA, m/z 379.3→287.3 (quantifier) and 91.1 (qualifier) for 2-AG. The levels of AEA and 2-AG in samples were measured against standard curves and normalized to tissue weight.

Chop overexpression

Request a detailed protocol

WT hepatocytes were transfected with mCHOP-WT-9E10-pCDNA1 vector (Addgene plasmid #21913) using Lipofectamin 3000. Cells were harvested 24 hr post-transfection and CHOP expression was validated by western blot analysis.

Luciferase promoter assay

Request a detailed protocol

Mus musculus Ob-R promoter sequence (GRCm38:CM000997.2. Chromosome 4: 101,716,750–101,718,250 forward strand) was cloned into pGL3-basic vector (E1751, Promega). Reporter vector and Renilla luciferase vector were then co-transfected into WT or CHOP KO hepatocytes using Lipophectamin 3000. Twenty-four hours post transfection, cells were treated with either vehicle, 2.5 µg/mL tunicamycin (11089-65-9; Holland Moran, Israel) or 2.5 µM NE for indicated period. At the end of the experiment, luciferase activity was measured using Dual-Glo Luciferase Assay System (E2920, Promega). Data are presented as the ratio between Firefly and Renilla luciferase activity.

Chromatin immunoprecipitation

Request a detailed protocol

WT or CHOP KO hepatocytes were seeded in 100 mm plate and left to adhere (5 × 106 cells per plate; three plates for each sample). The next day, cells were treated with either DMSO or 2.5 µg/mL tunicamycin for 6 hr to induce CHOP expression. At the end of incubation, cells were washed in 1× PBS, fixed with 1% formaldehyde for 10 min, then quenched with 125 mM glycine and harvested from plates. Following centrifugation, cells were resuspended in a lysis buffer and sonicated for 12 cycles of 30 s pulse followed by 30 s rest in 70% amplitude. Sheared DNA was diluted in a ChIP dilution buffer and pre-cleared with magnetic ProteinG-sepharose beads for 4 hr at 4°C. Ten percent of the lysate was removed and saved as ‘Input’. The rest of the lysate was divided and each part was incubated overnight at 4°C with 2.5 µg of either anti-H3 (ab1791, Abcam), anti-CHOP (2895S, Cell Signaling), or IgG isotype control. Antibody-chromatin complexes were precipitated with magnetic protein G-Sepharose beads, washed with low salt, high salt, lithium chloride, and Tris-EDTA buffers. DNA was then eluted from beads, digested with proteinase K, and purified; 1.5 µL of clean DNA was used in a qPCR reaction using specific primers for Gapdh or Lepr promoter region. For a positive control, Ppp1r15a primers were used. For each sample, we calculated the ratio between the RQ (expressed as % of input) of ObR promoter qPCR product in αCHOP IP relative to αH3 IP. This ratio was normalized to the ratio of Gapdh qPCR product to control for nonspecific binding. Data are expressed as the fold-change of this ratio in tunicamycin-treated compared to vehicle-treated cells. ChIP primers are listed in Supplementary file 1.

Statistics

The data and statistical analysis comply with the recommendations on experimental design and analysis as reported previously (Curtis et al., 2018). Randomization was used to assign samples to the experimental groups and treatment conditions for all in vivo studies. Data collection and acquisition of all in vivo and in vitro experimental paradigms were performed in a blinded manner. Data are presented as mean ± SEM. Unpaired two-tailed Student’s t-test was used to determine variations between two groups. Results in multiple groups were compared by ANOVA followed by a Bonferroni post hoc analysis using GraphPadPrism v6 for Windows. Post hoc tests were conducted only if F was significant, and there was no variance inhomogeneity. Significance was set at p<0.05.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
    A simple method for the isolation and purification of total lipides from animal tissues
    1. J Folch
    2. M Lees
    3. GH Sloane Stanley
    (1957)
    The Journal of Biological Chemistry 226:497–509.
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
  69. 69
  70. 70
  71. 71
  72. 72
  73. 73
  74. 74
  75. 75
  76. 76
  77. 77
  78. 78
  79. 79
  80. 80
    The soluble leptin receptor
    1. M Schaab
    2. J Kratzsch
    (2015)
    Best Practice & Research Clinical Endocrinology & Metabolism 29:661–670.
    https://doi.org/10.1016/j.beem.2015.08.002
  81. 81
  82. 82
  83. 83
  84. 84
  85. 85
  86. 86
  87. 87
  88. 88
  89. 89
  90. 90
  91. 91
  92. 92
  93. 93
  94. 94
  95. 95
  96. 96
  97. 97
  98. 98
  99. 99
  100. 100
  101. 101
  102. 102
  103. 103
  104. 104

Decision letter

  1. David E James
    Senior Editor; The University of Sydney, Australia
  2. Anna Mae Diehl
    Reviewing Editor; Duke University School of Medicine, United States
  3. George Kunos
    Reviewer; NIAAA, United States
  4. Vincenzo Di Marzo
    Reviewer; National Research Council (CNR), Italy
  5. Thomas Scherer
    Reviewer; Medical University of Vienna, Austria

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Thank you for submitting your article "CB1R Regulates Soluble Leptin Receptor Levels via CHOP, Contributing to Hepatic Leptin Resistance" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and David James as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: George Kunos (Reviewer #1); Vincenzo Di Marzo (Reviewer #2); Thomas Scherer (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, we are asking editors to accept without delay manuscripts, like yours, that they judge can stand as eLife papers without additional data, even if they feel that they would make the manuscript stronger. Thus the revisions requested below only address clarity and presentation.

Summary:

Leptin resistance is a major pathogenic factor in obesity and both central and peripheral mechanisms have been implicated in its development. Circulating leptin bound to a short form of soluble leptin receptor (sOb-R encoded by Ob-Re) cannot bind to the functional leptin receptor, resulting in reduced levels of free leptin that can cross the blood-brain-barrier. Circulating levels of sOb-R are inversely related to plasma leptin levels, suggesting that they play a key role in the control of leptin sensitivity. Activation of the endocannabinoid/CB1 receptor system is known to mediate obesity-related leptin resistance in rodents, and Tam and co-workers present elegant evidence for a novel mechanism by which endocannabinoids acting via hepatocyte CB1R down-regulate the hepatic expression and circulating levels of sOb-R, thus contributing to leptin resistance.

Reviewer 1:

1) Pharmacological (CB1 agonist/antagonist) as well as genetic tools (liver-specific CB1 ko and ki) are used to convincingly demonstrate that CB1R activation downregulates Ob-Re mRNA and sOb-R protein in the liver and in the circulation. Surprisingly, CB1R signal via reducing the expression of the pro-apoptotic transcription factor CHOP, an integral component of the PERK/eIF2α/ATF4/CHOP ER-stress pathway, as supported through the use of CHOP-/- mice. While the evidence is convincing, it is counterintuitive. ER-stress induced by a high fat diet was reported to upregulate rather than reduce CHOP expression in the liver (e.g. Bae et al., Molecules 25:2667, 2020) and the authors' own finding of increased eIF2α phosphorylation, which is upstream from CHOP, should result in increased rather than decreased CHOP expression. Could CB1 receptor activation disrupt this pathway at a point between eIF2α and CHOP? Could a Trib3-mediated decrease in ATF4 phosphorylation represent such a 'break'? This could be tested by monitoring ATF4 phosphorylation in the authors' paradigm. At a minimum, this paradox needs to be discussed.

2) In the subsection “CHOP contributes to the metabolic response to peripheral CB1R blockade”, the loss of ability of the CB1 antagonist to reverse the CB1-induced dyslipidemic effects (Figures 3H, I, J) in CHOP-/- mice is attributed to an increase in 'endocannabinoid tone', or more specifically a ~2-3-fold increase in the tissue levels of the endocannabinoids anandamide and 2-AG. In the absence of dose-response data, there is no reason to believe that a modest increase in the local concentration of endogenous agonists would result in the loss of effect of a highly potent competitive antagonist. It is more likely that the genetic deletion of an obligatory downstream mediator (CHOP) of the CB1 response is responsible for the loss of the effect of the antagonist. In this case the unchanged response to HFD may be mediated by a non-CB1 mechanism.

3) A limitation of the impact of these findings is that they may not be extended to human physiology, as in humans circulating sOb-R is generated exclusively by ectodomain shedding catalyzed by the proteolytic enzymes ADAM10 and ADAM17, whose expression remained unchanged (Discussion). However, only mRNA was measured, which does not exclude possible changes in enzyme protein levels or enzyme activity. This limitation needs to be mentioned/discussed.

4) The authors demonstrate that obesity leads to loss of leptin sensitivity in the wild-type, but not in LCB1ko liver, as indicated by the loss of leptin-induced STAT3 phosphorylation in the former but not the latter. These important findings should be more prominently highlighted in the Discussion, which should also briefly touch on what's known about the effects of leptin in the liver (e.g. regulation of hepatic insulin sensitivity).

Reviewer 2:

1) There are points in the Introduction and Discussion that need however to be ameliorated, through minor but nevertheless important corrections, as detailed below:

a) In general, the reader may wonder why the authors decided to investigate the role of CHOP in the effects of CB1R on sObR levels. This should be very briefly mentioned in the Introduction.

b) In the Introduction, the authors mention that the eCB system is overactive in obesity without quoting one of the first paper showing that this is true not only in the hypothalamus but also in peripheral cells, including the major source of leptin, i.e. the adipocytes (Matias et al., 2006). This should be mentioned; additionally, Monteleone et al., 2005, showed probably for the first time the elevation of circulating anandamide levels in overweight/obese women with binge eating disorder, again to be mentioned here;

c) In the Discussion, when referring to leptin signaling in anorexia nervosa, the authors should again mention the Monteleone et al., 2005 study, showing how anandamide levels are also elevated in women with this disorder, possibly due to their low leptin levels;

d) the above mentioned study by Matias et al., 2006, showed how both acute and chronic leptin significantly downregulates both anandamide and 2-AG levels in adipocytes (Figure 1 of that paper). This finding could be related to, or even explain, the authors' finding of higher eCB levels in CHOP-KO mice, which, by producing much less sObR, would counteract this inhibition of eCB levels also in adipocytes, by disinhibiting it. Since the authors did not provide a molecular mechanism for this effect, they may wish to discuss this possibility, especially if they have data showing that also in hepatocytes, like in adipocytes and hypothalamus, leptin decreases endocannabinoid levels via sObR (but also if they don't have such data).

Reviewer 3:

1) From a methodological viewpoint the study is well executed, however what I am missing a bit is the discussion of the physiologic significance and relevance of the findings for the field. This is relevant, because the exact role of the soluble leptin receptor for leptin action is not fully resolved in the current literature. Since the liver-specific CB1R transgenic mice on CB1R-null background are resistant to diet induced obesity, despite lower soluble leptin receptor expression raises the question whether the observed phenomenon is relevant for overall leptin action, which is predominantly mediated via the CNS. In the CB1R transgenic mice central leptin action seems (based on the resistance to DIO) preserved or at least not majorly impaired, despite lower secretion of soluble leptin receptor, questioning the overall physiologic relevance of the observed reduction in soluble leptin receptor for whole body leptin action. Do the authors have additional data on leptin signaling (in liver and brain) in the transgenic mice?

2) The claim that the regulation of soluble leptin receptor secretion via liver CB1R directly relates to liver leptin resistance is in my opinion not fully supported. The authors only present STAT3 phospho western blot data in CB1 KO and WT mice, but not the CB1R transgenic/CHOP KO mice. Direct effects of the endocannabinoid system on STAT3 signaling can thus not be excluded. I suggest deleting the claim from the title/Abstract. The authors should rather expand on this theory in the Discussion.

3) Circulating leptin levels (Figure 4F) in CHOP KO vs. WT JD5037 treated mice on HFD seem to correlate with fat mass (Figure 3F) rather than soluble leptin receptor levels. Are there any correlation data that would support the claim that soluble leptin receptor levels directly affects circulating leptin levels?

https://doi.org/10.7554/eLife.60771.sa1

Author response

Reviewer 1:

1) Pharmacological (CB1 agonist/antagonist) as well as genetic tools (liver-specific CB1 ko and ki) are used to convincingly demonstrate that CB1R activation downregulates Ob-Re mRNA and sOb-R protein in the liver and in the circulation. Surprisingly, CB1R signal via reducing the expression of the pro-apoptotic transcription factor CHOP, an integral component of the PERK/eIF2α/ATF4/CHOP ER-stress pathway, as supported through the use of CHOP-/- mice. While the evidence is convincing, it is counterintuitive. ER-stress induced by a high fat diet was reported to upregulate rather than reduce CHOP expression in the liver (e.g. Bae et al., Molecules 25:2667, 2020) and the authors' own finding of increased eIF2α phosphorylation, which is upstream from CHOP, should result in increased rather than decreased CHOP expression. Could CB1 receptor activation disrupt this pathway at a point between eIF2α and CHOP? Could a Trib3-mediated decrease in ATF4 phosphorylation represent such a 'break'? This could be tested by monitoring ATF4 phosphorylation in the authors' paradigm. At a minimum, this paradox needs to be discussed.

We would like to thank the reviewer for finding our data convincing and important.

As stated by the reviewer, high-fat diet (HFD)-induced obesity is known to upregulate ER-stress in the liver. We have previously demonstrated it (Tam et al., 2010) and are currently reporting it here by showing the increased hepatic phosphorylation of eIF2α (Figure 3—figure supplement 1). Indeed, the reduced transcriptional expression levels of CHOP are counterintuitive. Yet, as suggested by the reviewer, this could be explained via a CB1R-dependent regulation of either ATF4 or Trib3. The latter is upregulated in response to diverse forms of cellular stress (such as ERstress, excess free fatty acid, oxidative stress, hypoxia, hyperglycemia, and toxins; reviewed in Örd and Örd, 2017), and also by cannabinoids. Trib3 is able to directly bind and inhibit ATF4 and CHOP. In fact, our data support such a molecular mechanism, since both HFD and CB1R activation upregulate the mRNA and protein expression of Trib3 in the liver/hepatocytes and peripheral CB1R blockade in HFD-induced obese mice normalizes its expression (Figure 6—figure supplement 1).

Following the comment raised by the reviewer, we expanded our evidence for such a regulation and assessed the hepatic mRNA and protein expression levels of ATF4 in three animal settings: WT mice (STD- and HFD-fed animals treated with/without the peripherally restricted CB1R blocker, JD5037), liver-specific CB1 null mice (LCB1 KO, STD vs. HFD), and animals with overexpression of CB1R only in hepatocytes (htgCB1-/- mice, fed STD, or HFD). As can be seen in the new Figure 6—figure supplement 2, the hepatic ATF4 expression is downregulated in DIO mice, an effect that was completely normalized by peripheral CB1R blockade. Also, whereas nullification of CB1R in hepatocytes prevented the HFD-induced downregulation of ATF4, overexpression of the former in hepatocytes significantly contributes to the reduced expression of the latter. In addition, preliminary studies conducted in our lab revealed that direct CB1R activation by AEA in primary mouse hepatocytes results in a time-dependent reduction in ATF4 and CHOP mRNA levels (please see Author response image 1).

Author response image 1

Taken together, these findings support the idea that although ER stress is induced in the liver by HFD, the activation/blockade (or genetic deletion) of CB1R disrupts the conventional unfolded protein response (UPR) by affecting ATF4 expression, either directly, or indirectly via Trib3. These new findings are now described in the subsection “CHOP plays a key role in the regulation of sOb-R by the eCB/CB1R system”, and are further discussed in the Discussion. We have also amended the illustration that describes the molecular mechanism by which CB1R regulates sOb-R levels (please see the revised Figure 7).

2) In the subsection “CHOP contributes to the metabolic response to peripheral CB1R blockade”, the loss of ability of the CB1 antagonist to reverse the CB1-induced dyslipidemic effects (Figures 3H, I, J) in CHOP-/- mice is attributed to an increase in 'endocannabinoid tone', or more specifically a ~2-3-fold increase in the tissue levels of the endocannabinoids anandamide and 2-AG. In the absence of dose-response data, there is no reason to believe that a modest increase in the local concentration of endogenous agonists would result in the loss of effect of a highly potent competitive antagonist. It is more likely that the genetic deletion of an obligatory downstream mediator (CHOP) of the CB1 response is responsible for the loss of the effect of the antagonist. In this case the unchanged response to HFD may be mediated by a non-CB1 mechanism.

The reviewer is right. The elevated endocannabinoid 'tone' in the CHOP KO mice were not directly linked to the reduced effect of JD5037 in ameliorating dyslipidemia and hepatic steatosis. If our statement was misleading, we apologize for it. Based on our findings, CHOP plays an obligatory role in the CB1R-sOb-R cascade, and therefore, we agree with this reviewer that in CHOP KO animals the inability of JD5037 to ameliorate dyslipidemia is most likely related to the absence of CHOP. We amended the text accordingly (please see subsection “CHOP contributes to the metabolic response to peripheral CB1R blockade” and the Discussion).

3) A limitation of the impact of these findings is that they may not be extended to human physiology, as in humans circulating sOb-R is generated exclusively by ectodomain shedding catalyzed by the proteolytic enzymes ADAM10 and ADAM17, whose expression remained unchanged (Discussion). However, only mRNA was measured, which does not exclude possible changes in enzyme protein levels or enzyme activity. This limitation needs to be mentioned/discussed.

The regulation of ADAM10 and ADAM17 is complex and involves transcription, dynamic trafficking, cellular localization, and activity. Indeed, the current study focused on the transcriptional regulation of sOb-R by the CB1R/CHOP signaling pathway. However, other Ob-R isoforms, also expressed in humans, display (at the gene and protein levels) a trend similar to ObRe following either CB1R activation or blockade. These isoforms (Ob-Ra, Ob-Rb) are substrates for ectodomain shedding and therefore their transcriptional regulation may indirectly influence sObR levels and be relevant to human physiology. Nonetheless, the novel CB1R-induced regulation of CHOP levels and activity is not limited to animals and may also be relevant to humans. We have discussed these issues in the Discussion.

4) The authors demonstrate that obesity leads to loss of leptin sensitivity in the wild-type, but not in LCB1ko liver, as indicated by the loss of leptin-induced STAT3 phosphorylation in the former but not the latter. These important findings should be more prominently highlighted in the Discussion, which should also briefly touch on what's known about the effects of leptin in the liver (e.g. regulation of hepatic insulin sensitivity).

In addition to our findings in WT and LCB1 null mice, we assessed leptin sensitivity in animals with overexpression of CB1R only in hepatocytes (htgCB1-/-), and revealed an inability of leptin to increase STAT3 phosphorylation in these mice (please see the additional data in Figure 1R and S), further supporting the critical role of hepatic CB1R in regulating hepatic leptin sensitivity. As recommended by the reviewer, we have expanded the Introduction and Discussion accordingly.

Reviewer 2:

1) There are points in the Introduction and Discussion that need however to be ameliorated, through minor but nevertheless important corrections, as detailed below:

a) In general, the reader may wonder why the authors decided to investigate the role of CHOP in the effects of CB1R on sObR levels. This should be very briefly mentioned in the Introduction.

We expanded the Introduction to include information about the role of ER stress in the regulation of leptin resistance.

b) In the Introduction, the authors mention that the eCB system is overactive in obesity without quoting one of the first paper showing that this is true not only in the hypothalamus but also in peripheral cells, including the major source of leptin, i.e. the adipocytes (Matias et al., 2006). This should be mentioned; additionally, Monteleone et al., 2005, showed probably for the first time the elevation of circulating anandamide levels in overweight/obese women with binge eating disorder, again to be mentioned here;

The reviewer is right and these two citations as well as Matias et al., 2008 were added to the Introduction.

c) In the Discussion, when referring to leptin signaling in anorexia nervosa, the authors should again mention the Monteleone et al., 2005 study, showing how anandamide levels are also elevated in women with this disorder, possibly due to their low leptin levels;

We discussed this paper in the Discussion.

d) the above mentioned study by Matias et al., 2006, showed how both acute and chronic leptin significantly downregulates both anandamide and 2-AG levels in adipocytes (Figure 1 of that paper). This finding could be related to, or even explain, the authors' finding of higher eCB levels in CHOP-KO mice, which, by producing much less sObR, would counteract this inhibition of eCB levels also in adipocytes, by disinhibiting it. Since the authors did not provide a molecular mechanism for this effect, they may wish to discuss this possibility, especially if they have data showing that also in hepatocytes, like in adipocytes and hypothalamus, leptin decreases endocannabinoid levels via sObR (but also if they don't have such data).

Our data indeed show that CHOP null mice have a higher hepatic eCB 'tone', reduced levels of sOb-R, and a trend toward a reduction in circulating leptin levels (seen only in the STD-fed mice). These observations are in line with the reviewer's suggestion, and as mentioned in the text, deciphering the molecular mechanism by which CHOP regulates eCB production is out of the scope of this paper and will require the utilization of Ob-R-deficient models and/or the use of leptin antagonists to draw such a conclusion. Nevertheless, we amended the Discussion and included a possible explanation to these documented findings.

Reviewer 3:

1) From a methodological viewpoint the study is well executed, however what I am missing a bit is the discussion of the physiologic significance and relevance of the findings for the field.

Thank you for finding our work to be well executed. We added a significance statement as well as discussed the relevance of our findings to human physiology in the Discussion (please also see our response to a similar comment raised by the first reviewer). In addition, the importance of our findings to the regulation of hepatic (peripheral) leptin resistance/sensitivity via CB1R was expanded on in the Discussion.

This is relevant, because the exact role of the soluble leptin receptor for leptin action is not fully resolved in the current literature. Since the liver-specific CB1R transgenic mice on CB1R-null background are resistant to diet induced obesity, despite lower soluble leptin receptor expression raises the question whether the observed phenomenon is relevant for overall leptin action, which is predominantly mediated via the CNS. In the CB1R transgenic mice central leptin action seems (based on the resistance to DIO) preserved or at least not majorly impaired, despite lower secretion of soluble leptin receptor, questioning the overall physiologic relevance of the observed reduction in soluble leptin receptor for whole body leptin action. Do the authors have additional data on leptin signaling (in liver and brain) in the transgenic mice?

Indeed, a soluble leptin receptor is only one of the molecular mechanisms that may contribute to leptin sensitivity/resistance. Here, we reported its role in regulating hepatic leptin resistance. The hTgCB1-/- animals, despite producing less sOb-R, do not become obese because they lack CB1Rs everywhere except the liver. Central CB1Rs are required for the development of diet-induced obesity (DIO, as reported by Pang et al., Obesity, 2012), and are downstream of the effect that sOb-R has on leptin during the development of DIO, as expected. That is why hTgCB1-/- animals, although having reduced sOb-R levels, are resistant to DIO. This is also true for mice globally lacking CB1Rs, which are resistant to DIO and its metabolic consequences, despite a similar caloric intake during the diet period (Ravinet Trillou et al., 2004; Osei-Hyiaman et al., 2005), which also points to energy metabolism being directly regulated by central endocannabinoids.

Nevertheless, this does not, by any means, imply that sOb-R is without relevance for DIO and energy metabolism in WT animals. On the contrary, the observations of Lou et al., 2010 strongly support the importance of sOb-R when CB1Rs are present. Here, we basically used hTgCB1-/- animals to document the relevance of hepatic CB1R for the production of sOb-R, and show that indeed its overexpression in hepatocytes downregulates the expression of sOb-R, and that it contributes to hepatic leptin resistance. Following the reviewer's comment, we expanded our study and measured leptin signaling in the liver of hTgCB1-/- animals, and found an inability of leptin to induce hepatic STAT3 phosphorylation in these mice. These additional results (please see the revised Figure 1R, S, the subsection “Hepatic CB1R regulates sOb-R levels and leptin signalling” and the Discussion) further imply a key role for CB1R in modulating hepatic leptin signaling via regulating sOb-R levels.

2) The claim that the regulation of soluble leptin receptor secretion via liver CB1R directly relates to liver leptin resistance is in my opinion not fully supported. The authors only present STAT3 phospho western blot data in CB1 KO and WT mice, but not the CB1R transgenic/CHOP KO mice. Direct effects of the endocannabinoid system on STAT3 signaling can thus not be excluded. I suggest deleting the claim from the title/Abstract. The authors should rather expand on this theory in the Discussion.

Thank you for this comment. As mentioned above, we have now provided evidence for the lack of hepatic leptin sensitivity in mice overexpressing CB1R in hepatocytes (see the revised Figure 1). We have also tested the ability of exogenous leptin to induce STAT3 phosphorylation in the liver of CHOP KO mice, and found reduced hepatic leptin sensitivity in the absence of CHOP (see the revised Figure 4G, H, and the subsection “CHOP plays a key role in the regulation of sOb-R by the eCB/CB1R system”), that is also in accordance with the reduced levels of sOb-R in these mice. We also discussed the relevance of hepatic leptin sensitivity and its regulation by CB1R (see the Discussion), and therefore decided to leave this claim in the title and Abstract. We hope that the reviewer, in light of the newly added data, also considers our claim now well-founded.

3) Circulating leptin levels (Figure 4F) in CHOP KO vs. WT JD5037 treated mice on HFD seem to correlate with fat mass (Figure 3F) rather than soluble leptin receptor levels. Are there any correlation data that would support the claim that soluble leptin receptor levels directly affects circulating leptin levels?

Many human and animal studies have demonstrated that sOb-R levels are inversely correlated with the plasma levels of leptin, BMI, and adiposity (Chan et al., 2002; Lahlou et al., 2000; Laimer et al., 2002; Ogier et al., 2002; Reinehr et al., 2005). The positive correlation between serum leptin concentrations and fat mass has been well demonstrated (Lonnqvist et al., 1995; Considine et al., 1996; Maffei et al., 1995). And indeed, the circulating leptin levels measured here in WT and CHOP KO mice (under STD, HFD, and HFD+JD5037) seem to be also positively correlated with the fat mass (see Author response image 2, left).

As pointed out by this reviewer, we found a negative correlation between hepatic sOb-R content and circulating leptin levels in both mouse strains (see Author response image 2, right), further supporting the idea that low levels of the hepatic soluble isoform contribute to obesity-related hyperleptinemia and subsequently, leptin resistance.

Author response image 2
https://doi.org/10.7554/eLife.60771.sa2

Article and author information

Author details

  1. Adi Drori

    Obesity and Metabolism Laboratory, Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
    Contribution
    Data curation, Formal analysis, Validation, Investigation, Methodology, Writing - original draft
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-5790-9544
  2. Asaad Gammal

    Obesity and Metabolism Laboratory, Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
    Contribution
    Investigation
    Competing interests
    No competing interests declared
  3. Shahar Azar

    Obesity and Metabolism Laboratory, Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
    Contribution
    Data curation, Investigation
    Competing interests
    No competing interests declared
  4. Liad Hinden

    Obesity and Metabolism Laboratory, Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
    Contribution
    Data curation, Investigation
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0307-4350
  5. Rivka Hadar

    Obesity and Metabolism Laboratory, Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
    Contribution
    Investigation, Project administration
    Competing interests
    No competing interests declared
  6. Daniel Wesley

    Laboratory of Physiological Studies, National Institute on Alcohol Abuse & Alcoholism, Bethesda, United States
    Contribution
    Investigation
    Competing interests
    No competing interests declared
  7. Alina Nemirovski

    Obesity and Metabolism Laboratory, Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
    Contribution
    Methodology
    Competing interests
    No competing interests declared
  8. Gergő Szanda

    MTA-SE Laboratory of Molecular Physiology, Department of Physiology, Semmelweis University, Budapest, Hungary
    Contribution
    Resources, Writing - review and editing
    Competing interests
    No competing interests declared
  9. Maayan Salton

    Department of Biochemistry and Molecular Biology, The Institute for Medical Research Israel-Canada, The Hebrew University of Jerusalem, Jerusalem, Israel
    Contribution
    Resources, Data curation, Formal analysis, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared
  10. Boaz Tirosh

    The Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
    Contribution
    Resources, Formal analysis, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8067-6577
  11. Joseph Tam

    Obesity and Metabolism Laboratory, Institute for Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
    Contribution
    Conceptualization, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Writing - original draft, Writing - review and editing
    For correspondence
    yossi.tam@mail.huji.ac.il
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0948-0093

Funding

Israel Science Foundation (617/14)

  • Joseph Tam

Israel Science Foundation (158/18)

  • Joseph Tam

Obesity Society (Early Career Research Award)

  • Joseph Tam

ERC-2015-StG grant (676841)

  • Joseph Tam

Hungarian National Research, Development, and Innovation Office (NKFI-6/FK_124038)

  • Gergő Szanda

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

This study was supported by grants from the Israeli Science Foundation (ISF; 617/14 and 158/18), The Obesity Society's Early Career Research Award, and an ERC-2015-StG grant (#676841) to J.T, as well as the Hungarian National Research, Development, and Innovation Office (Grant #NKFI-6/FK_124038 to G.S.). This paper is dedicated to the memory of Simon Tam.

Ethics

Animal experimentation: All animal studies were approved by the Institutional Animal Care and Use Committee of the Hebrew University of Jerusalem (AAALAC accreditation #1285; Ethic approval numbers MD-14-14008 & MD-19-15951). Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny et al., 2010), and are based on the rule of the replacement, refinement, or reduction. All the animals used in this study were housed under specific pathogen‐free (SPF) conditions, up to five per cage, in standard plastic cages with natural soft sawdust as bedding. The animals were maintained under controlled temperature of 22-24°C, humidity at 55 ± 5%, and alternating 12-hour light/dark cycles (lights were on between 7:00 and 19:00 hr), and provided with food and water ad libitum.

Senior Editor

  1. David E James, The University of Sydney, Australia

Reviewing Editor

  1. Anna Mae Diehl, Duke University School of Medicine, United States

Reviewers

  1. George Kunos, NIAAA, United States
  2. Vincenzo Di Marzo, National Research Council (CNR), Italy
  3. Thomas Scherer, Medical University of Vienna, Austria

Publication history

  1. Received: July 6, 2020
  2. Accepted: November 17, 2020
  3. Accepted Manuscript published: November 19, 2020 (version 1)
  4. Version of Record published: December 10, 2020 (version 2)

Copyright

© 2020, Drori et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 418
    Page views
  • 68
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Cell Biology
    Aleksandra Z Płochocka et al.
    Research Article Updated

    Robustness of biological systems is crucial for their survival, however, for many systems its origin is an open question. Here, we analyze one subcellular level system, the microtubule cytoskeleton. Microtubules self-organize into a network, along which cellular components are delivered to their biologically relevant locations. While the dynamics of individual microtubules is sensitive to the organism’s environment and genetics, a similar sensitivity of the overall network would result in pathologies. Our large-scale stochastic simulations show that the self-organization of microtubule networks is robust in a wide parameter range in individual cells. We confirm this robustness in vivo on the tissue-scale using genetic manipulations of Drosophila epithelial cells. Finally, our minimal mathematical model shows that the origin of robustness is the separation of time-scales in microtubule dynamics rates. Altogether, we demonstrate that the tissue-scale self-organization of a microtubule network depends only on cell geometry and the distribution of the microtubule minus-ends.

    1. Cell Biology
    2. Immunology and Inflammation
    Leonie Zeitler et al.
    Research Article

    Interleukin-4-induced-1 (IL4i1) is an amino acid oxidase secreted from immune cells. Recent observations have suggested that IL4i1 is pro-tumorigenic via unknown mechanisms. As IL4i1 has homologues in snake venoms (LAAO, L-amino acid oxidases), we used comparative approaches to gain insight into the mechanistic basis of how conserved amino acid oxidases regulate cell fate and function. Using mammalian expressed recombinant proteins, we found venom LAAO kills cells via hydrogen peroxide generation. By contrast, mammalian IL4i1 is non-cytotoxic and instead elicits a cell productive gene expression program inhibiting ferroptotic redox death by generating indole-3-pyruvate (I3P) from tryptophan. I3P suppresses ferroptosis by direct free radical scavenging and through the activation of an anti-oxidative gene expression program. Thus, the pro-tumor effects of IL4i1 are likely mediated by local anti-ferroptotic pathways via aromatic amino acid metabolism, arguing that an IL4i1 inhibitor may modulate tumor cell death pathways.