Central Sfrp5 regulates hepatic glucose flux and VLDL-triglyceride secretion

Journal Pre-proof Central Sfrp5 regulates hepatic glucose flux and VLDLtriglyceride secretion

Yang Li, Mingyuan Tian, Mengliu Yang, Gangyi Yang, Jianrong Chen, Han Wang, Dongfang Liu, Hongyan Wang, Wuquan Deng, Zhiming Zhu, Hongting Zheng, Ling Li PII:

S0026-0495(19)30244-6

DOI:

https://doi.org/10.1016/j.metabol.2019.154029

Reference:

YMETA 154029

To appear in:

Metabolism

Received date:

25 June 2019

Accepted date:

22 November 2019

Please cite this article as: Y. Li, M. Tian, M. Yang, et al., Central Sfrp5 regulates hepatic glucose flux and VLDL- triglyceride secretion, Metabolism(2019), https://doi.org/ 10.1016/j.metabol.2019.154029

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier.

Journal Pre-proof

Central Sfrp5 Regulates Hepatic Glucose Flux and VLDLTriglyceride Secretion Yang Lia,b,h, Mingyuan Tianb,h, Mengliu Yangc, Gangyi Yangb, Jianrong Chena, Han Wangd, Dongfang Liub, Hongyan Wange, Wuquan Deng e, Zhiming Zhuf, Hongting Zhengg, Ling Lia,* a

f

The Key Laboratory of Laboratory Medical Diagnostics in the Ministry of Education and

oo

Department of Clinical Biochemistry, College of Laboratory Medicine, Chongqing Medical

Department of Endocrinology, the Second Affiliated Hospital, Chongqing Medical

e-

b

pr

University, Chongqing, 400016, China

Pr

University and Chongqing Clinical Research Center for Geriatrics, Chongqing, 400010, China c

al

School of Biomedical Sciences, the University of Queensland, Brisbane, 4103, Australia

d

Jo u

China

rn

Department of Laboratory, Children's Hospital of Chongqing Medical University,400015,

e

Chongqing Emergency Medical Center, Chongqing, China

f

Department of Hypertension and Endocrinology, Daping Hospital, Third Military Medical

University, Chongqing Institute of Hypertension, Chongqing, 400010, China g

Department of Endocrinology, Xinqiao Hospital, Third Military Medical University,

Chongqing, 400010, China h

These authors contributed equally to this work.

* Corresponding author: Ling Li, Laboratory of Diagnostic Medicine (Ministry of Education) and Department of Clinical Biochemistry, College of Laboratory Medicine, 1

Journal Pre-proof

Chongqing Medical University, Chongqing, 400016, China. Phone number: +8623-68485240

Jo u

rn

al

Pr

e-

pr

oo

f

E-mail: [email protected]

2

Journal Pre-proof Abstract Objective: Secreted frizzled-related protein 5 (Sfrp5) has been shown to be associated with energy homeostasis and insulin resistance in mouse models of obesity and diabetes. However, its central role in glucose and lipid metabolism is unknown. Methods: HFD-fed rats received ICV infusions of vehicle or Sfrp5 during a pancreatic euglycemic clamp procedure. To delineate the pathway(s) by which ICV Sfrp5 modulates

oo

f

HGP and VLDL-TG secretion, we inhibited the hypothalamic KATP channel using

pr

glibenclamide, the DVC NMDA receptor with MK801, and selectively transected the hepatic branch of the vagal nerve while centrally infusing Sfrp5.

e-

Results: ICV Sfrp5 in HFD-fed rats significantly increased the glucose infusion required to

Pr

maintain euglycemia due to HGP inhibition during the clamp procedure; moreover, hepatic

al

PEPCK and G6Pase expression was decreased, and InsR and Akt phosphorylation was

rn

increased in the liver. ICV Sfrp5 also decreased circulating triglyceride levels via inhibiting

Jo u

hepatic VLDL-TG secretion. These changes were accompanied by the inhibition of enzymes related to lipogenesis in the liver. ICV Sfrp5 significantly increased insulin-stimulated phosphorylation of InsR and Akt in the hypothalamus of HFD-fed rats, and insulin-stimulated immunodetectable PIP3 levels were higher in Sfrp5 group than in control group both in vitro and vivo. The glucose- and lipid-lowering effects of ICV Sfrp5 were eliminated by NMDA receptor or DVC KATP channel inhibition or HVAG. Conclusions: The present study demonstrates that central Sfrp5 signaling activates a previously unappreciated InsR-Akt-PI3k-KATP channel pathway in the hypothalamus and brain-hepatic vagus neurocircuitry to decrease HGP and VLDL-TG secretion.

3

Journal Pre-proof

Keywords: Sfrp5; Insulin resistance; VLDL-TG; Triglyceride

Abbreviations: KATP, ATP-sensitive potassium channel; ICV, intracerebroventricular; HFD,

f

high-fat-diet; HGP, hepatic glucose production; VLDL-TG, very low-density lipoproteins

oo

triglyceride; DVC, dorsal vagal complex; NMDA, N-methyl-D-aspartate; PEPCK,

pr

phosphoenolpyruatecarboxykinase; G6Pase, glucose-6- phosphatase; InsR, insulin receptor;

e-

Akt, Akt kinase; PIP3, phosphatidylinositol 3, 4, 5-trisphosphate; HVAG, hepatic branch

Jo u

rn

al

Pr

vagotomy.

4

Journal Pre-proof

1.

Introduction

Obesity is a worldwide epidemic and a risk factor for metabolic diseases and atherosclerosis, which is often associated with a chronic inflammatory state and insulin resistance (IR). Chronic inflammation of adipose tissue leads to the abnormal release of hormones, inflammatory cytokines and adipokines. Over the last ten years, we and other researchers

f

have reported that the increases in nutrients, hormones and adipokines, such as nesfatin-1,

oo

vaspin and leptin, in the hypothalamus can regulate the peripheral metabolism in rodents [1–

e-

brain-liver axis are not completely understood.

pr

8]. Nevertheless, the molecular signals in the hypothalamus required for the function of the

Pr

Secreted frizzled-related protein 5 (Sfrp5) is a member of the Sfrp family. Sfrp5 is an anti-inflammatory adipocytokine that antagonizes the pro-inflammatory adipokine wingless-

al

type MMTV integration site family member (Wnt) 5a [9]. Sfrp5 has been shown to be

rn

associated with inflammation and IR in mouse models of obesity and type 2 diabetes mellitus

Jo u

(T2DM) [9]. Sfrp5 expression is downregulated in both obesity and diabetic animals. Sfrp5 KO mice fed a high-fat (HFD) and high-glucose diet exhibit impaired glucose tolerance (IGT), IR and lipid accumulation in the liver. Intravenous Sfrp5 administration reverses IR and improves inflammatory states in obese mice [9]. Lv et al. found that Sfrp5 mRNA expression is increased during adipocyte differentiation in vitro and in vivo studies of obese animals [10]. Furthermore, Sfrp5 mRNA expression is markedly associated with increased fat and adipocyte size [11]. In human studies, we and others found that both T2DM and IGT patients exhibited lower circulating Sfrp5 levels than normal controls [12-13]. Sfrp5 has also been associated with markers of adiposity (e.g., body mass index (BMI) and waist-to-hip ratio 5

Journal Pre-proof

(WHR)), IR (e.g., homeostasis model assessment of IR (HOMA-IR)) and lipid profiles [12]. These results indicate that Sfrp5 signaling is associated with metabolic dysfunction and IR. Recently, Sfrp5 was found to be expressed in the brain [9]. Despite compelling evidence that peripheral hormones can cross the blood-brain barrier [14] and can be systemically delivered [9], peripheral and central hormonal signaling are generally considered to be

f

separate regulatory circuits. Therefore, it is important to investigate how the hypothalamic

oo

region of the brain senses circulating nutrients and hormones to regulate energy homeostasis.

pr

Sfrp5 signaling regulates peripheral metabolism and insulin sensitivity, but whether brain

e-

Sfrp5 is sufficient to regulate metabolism remains unknown.

Pr

Given that central cytokine and hormone signaling regulate energy metabolism in the liver [15], we hypothesized that increases in Sfrp5 levels within the hypothalamus may

al

contribute to the regulation of energy metabolism in the liver. Therefore, in this study, we

rn

used molecular, pharmacological, and surgical approaches to investigate whether central

Jo u

Sfrp5 can regulate hepatic glucose production (HGP) and triglyceride-rich very-low-density lipoprotein (VLDL-TG) secretion in vivo.

6

Journal Pre-proof

2.

Materials and Methods

2.1. Experimental animals Sprague-Dawley (SD) rats, C57BL/6J, db/db and adiponectin knockout (Adipoq KO) mice were provided from The Experimental Animal Center of Chongqing Medical University (Chongqing, China) and Shanghai Biomodel Ltd. (Shanghai, China), respectively. Eight-

f

week-old male C57BL/ 6J, Adipoq KO, db/db mice and SD rats were fed either a normal

oo

chow diet (NCD; 10% calories from fat) or an HFD (45% of calories from fat; Medicine Ltd,

pr

Jiangshu, China) for 12 weeks. The mice were sacrificed at the end of the 12-week

e-

experiment, and hypothalamic tissues were harvested for Sfrp5 expression analysis. For

Pr

VLDL-TG and pancreatic-euglycemic clamp (PEC) experiments, rats were stereotaxically implanted with indwelling catheters into the third cerebral ventricle (V3) alone or both V3

al

and dorsal vagal complex (DVC) 12 days before the experiments (1, 5). After full recovery (7

rn

days), the rats were anesthetized and catheters were inserted into the right internal jugular

Jo u

vein and left carotid artery for PECs and VLDL-TG experiments as described previously [7, 16]. All experimental procedures were approved by the Animal Experimentation Ethics Committee (Chongqing Medical University). 2.2. Selective hepatic branch vagotomy (HVAG) HVAG was performed as previously described (Supplementary data) [17]. 2.3. DVC infusion and hypothalamus signaling study Four groups of rats (n = 3 – 6 for each group) were stereotaxically implanted with a bilateral steel guide cannula positioned 2 mm above the caudomedial nucleus of the solitary tract (NTS). Bilateral catheters were inserted into the DVC for MK-801or saline infusion on day 0. 7

Journal Pre-proof

For the insulin signaling study, a total of 4 µU (2 µU per site) of insulin was infused into the DVC (0.4 µU/min). The rats were sacrificed after 15 min, and the DVC tissue was collected immediately and freeze-clamped in situ and stored at -160°C. 2.4. Infusion protocol for VLDL-TG and PEC experiments Five days after intravenous catherization, rats were fasted for 12 h prior to the PEC or

f

VLDL-TG experiments. To investigate the dose-dependent effects of intracerebroventricular

oo

(ICV) Sfrp5 on glucose tolerance test (GTT), PECs and VLDL-TG secretion (Fig. S1A-C),

pr

different concentrations of Sfrp5 (0, 0.5, 1.0, and 1.5 µg for GTT; 1, 10, 100, and 1000 nm/l

e-

for PEC and VLDL-TG experiments) were injected or continuously infused intracerebroventricularly through a V3 catheter. For PECs or VLDL-TG experiments,

Pr

following blood sample collection at t = 0 min, ICV infusions (5 μl/h) in conscious rats were

al

initiated as follows (n = 5- 6 for each group): 1) artificial cerebrospinal fluid (aCSF); 2) aCSF

rn

+ glibenclamide (Gliben, 100 µM); 3) Sfrp5 (10 nmol/L, dissolved in aCSF); 4) Sfrp5

Jo u

(10nmol/L) + Gliben; and 5) diazoxide (46.7 µM). Additionally, in some experimental groups (n = 5- 6 for each group), the following DVC infusions (0.33 µL/h) were also performed: 1) saline with ICV aCSF; 2) MK- 801 (0.06 ng/ min) with ICV aCSF; 3) saline with ICV Sfrp5 (10 nmol/L); 4) MK-801 with ICV Sfrp5. For VLDL-TG experiments, tyloxapol (600 mg/kg) was injected intravenously (i.v.) to inhibit endogenous lipoprotein lipase. This dose of tyloxapol ensured complete inhibition of endogenous lipoprotein lipase and blocked nascent VLDL particle clearance during VLDL-TG experiments. Blood samples were obtained every 30 min until the end of the PEC procedure (t=180 min). The rate of triglyceride (TG) secretion was calculated as the slope of the TG rise over time by linear regression analysis 8

Journal Pre-proof

[18]. For the lipid-induced IR study, a 15% lipid emulsion (Liposyn II, Berlin, Germany) containing heparin (20 U/ml) was infused at a rate of 1.5 ml/h from 120 to 360 min of the PEC procedure. Rats used for the PEC experiments received 16 g of food intake to ensure the same nutritional post-absorptive status the night before clamping. The clamp procedure was performed as reported previously [7, 19]. At the end of the experiments, the rats were

f

anesthetized, and tissues were obtained and stored at -160 °C until use for further analyses.

oo

Blood specimens were centrifuged and stored at - 80ºC until analysis.

pr

2.5. Determination of TG and cholesterol (TC) contents

e-

Total TG and TC contents in the liver were measured as previously described [20].

Pr

2.6. Hepatic H&E and Oil Red O staining

H&E and Oil Red O staining in the liver were performed as previously described [20].

al

2.7. Cell culture and treatment

Jo u

described [21, 22].

rn

SH-SY5Y cells and Primary hypothalamic neurons (PHNs) were cultured as previously

2.8. Immunohistochemistry and immunofluorescence Three hours after Sfrp5 or aCSF was infused intracerebroventricularly, the rats were anesthetized and perfused with saline containing heparin (20 U/ml) for 3 min and 4% paraformaldehyde in 0.1 M PBS for 20 min. The brains were excised, transferred to 4% paraformaldehyde and fixed at 4°C for 24 h. Specimens were embedded in paraffin, and coronal sections of the brain were obtained. Immunohistochemistry for c-Fos protein (rabbit anti-c-Fos, 1:200; Abcam, Cambridge, MA) was performed as described previously [23]. Phosphatidylinositol 3,4,5-trisphosphate (PIP3) immunostaining was performed on arcuate 9

Journal Pre-proof

nucleus (ARC) sections of overnight-fasted rats, which were subjected to either aCSF or insulin injection into the DVC and sacrificed 10 min after stimulation, and on PHNs using an anti-PIP3 antibody co-incubated with anti-mCherry (Abbkine, California, USA), or a fluorescein conjugated anti-PIP3 antibody (Z-G345; Echelon Biosciences, Salt Lake City, UT), respectively [24]. The integrated density quantification of PIP3 in ARC and PHNs was

f

measured using ImageJ software. The slides of PHN were measured including 650–800

oo

cells/group to evaluate PIP3 content.

pr

2.9. Metabolic and biochemical analyses

Protein and mRNA analysis

Pr

2.10.

e-

Metabolic and biochemical analyses are detailed in the Supplementary data [19].

Quantitative RT-PCR was performed as described previously [25]. The primer sequences are

al

shown in Table S1. Western blots were performed with primary and secondary antibodies, as

rn

described previously [26]. The antibodies used are shown in Table S2.

Jo u

2.11. Statistical analysis

Data are expressed as the means ± SME. A two-way ANOVA and least significant difference post hoc test were used to compare the mean values between multiple groups, and a twotailed paired or unpaired Student’s t test was used for two-group comparisons. P values < 0.05 were considered statistically significant.

10

Journal Pre-proof

3.

Results

3.1. Effects of ICV Sfrp5 administration on food intake and energy expenditure As expected, HFD-fed rats developed obesity and metabolic disorder (Table S3). Furthermore, we investigated the effects of ICV Sfrp5 on food intake and energy expenditure. A single injection of ICV Sfrp5 (1 µg/2 µl) produced a temporary decrease in food intake for 24 h in

f

HFD-fed rats (Fig. 1A). Next, we performed indirect calorimetric experiments to investigate

oo

whether ICV Sfrp5 affected energy metabolism. The results demonstrated higher rectal

pr

temperature and energy expenditure in ICV Sfrp5 rats than in ICV aCSF rats fed a HFD (Fig.

e-

1B and F). Consistently, the 24-h VO2 and Vco2 were markedly higher for HFD-fed ICV Sfrp5

Pr

rats than aCSF control rats (Fig. 1C and D), but the respiratory quotient (RER = VCO2/VO2) did not markedly change (Fig. 1E). These data suggest that a single injection of ICV Sfrp5 in

al

HFD-fed rats reduced food intake and increased energy expenditure.

rn

3.2. Hypothalamic Sfrp5 expression in IR mice and c-Fos immunoreactivity induced by

Jo u

central Sfrp5 in hypothalamic regions To investigate whether hypothalamic Sfrp5 expression is changed under the IR-state, we examined the protein levels of Sfrp5 in the hypothalamus of IR and non-IR mice. We found that Sfrp5 protein expression was significantly reduced in the hypothalamus of db/db and Adipoq KO mice compared to WT mice fed a NCD (Fig. S2A and B). In HFD-fed WT and Adipoq KO mice, Sfrp5 protein expression in hypothalamus decreased significantly compared to that in NCD-fed WT mice (Fig. S2B). These results suggest the involvement of hypothalamic Sfrp5 in genetic- and diet-induced IR. Next, to examine whether ICV Sfrp5 leads to specific neuronal activation, we performed 11

Journal Pre-proof

a c-Fos experiment in the hypothalamus. As shown in Figure 1G, ICV Sfrp5 in rats increased the number of c-Fos-positive neurons in the mediobasal hypothalamus (MBH), including the ventromedial hypothalamus (VMH) and ARC, which is related to peripheral energy metabolism. 3.3. ICV Sfrp5 ameliorates IR and improves glucose handling in IR rats

f

To investigate the effects of central Sfrp5 on glucose handling in vivo, we performed GTTs in

oo

rats treated with ICV Sfrp5 or aCSF. A separate, limited dose-response study included the

pr

administration of different doses of ICV Sfrp5 to demonstrate the dose-dependent effects of

e-

central Sfrp5 on blood glucose during GTT. As shown in Figure S1A, 1 and 1.5 μg/2 μl ICV injections of Sfrp5 significantly decreased blood glucose and the area under the curve for

Pr

glucose (AUC glucose). Therefore, 1 μg/2 μl was selected for further experiments. Glucose

al

injection increased blood glucose levels and AUGglucose in NCD-fed rats to a similar extent as

rn

in ICV Sfrp5 and ICV aCSF rats during GTTs (Fig. 2A). However, ICV Sfrp5 infusion in

Jo u

HFD-fed rats markedly decreased glucose levels and AUG glucose compared to control conditions in rats (Fig. 2A). Accordingly, ICV Sfrp5 rats fed an HFD exhibited lower insulin levels and AUGinsulin (Fig. 2B), which suggests that ICV Sfrp5 enhances glucose handling in obese rats. To establish the relationship between ICV Sfrp5 and HGP, we assessed the impact of ICV Sfrp5 on glucose kinetics using an insulin clamp procedure (Fig. 2C). During the clamp study, blood glucose, insulin and free fatty acid (FFA) levels were similar in ICV Sfrrp5 and aCSF rats (Table S4). In the dose-response study, different doses of ICV Sfrp5 infusion in HFD-fed rats were administered to demonstrate the effect of ICV Sfrp5 on the glucose 12

Journal Pre-proof

infusion rate (GIR) during the PEC procedure. ICV infusion of Sfrp5 (10 nmol/L) significantly increased the GIR to the maximum level. Therefore, 10 nmol/L was selected for the clamp study (Fig. S1B). ICV Sfrp5 did not affect glucose kinetics during the PEC procedure in NCD-fed rats (Fig. 2D-G). However, ICV Sfrp5 in HFD-fed rats markedly increased the GIR and glucose disposal rate (GRd) (Fig. 2D and E) and decreased the HGP

f

(by ~ 47%) compared to ICV aCSF (by ~ 28%) (Fig. 2F and G). These data suggest that ICV

oo

Sfrp5 is sufficient to ameliorate glucose kinetics and insulin sensitivity in HFD-induced IR. In

pr

addition, to evaluate the effects of ICV Sfrp5 on acute IR induced by lipid infusion, we

e-

performed both lipid infusion and PEC experiments upon ICV Sfrp5 infusion (Fig. 2H). As

Pr

expected, lipid infusion during the PEC procedure significantly increased FFA levels (Fig. 2I) and HGP (Fig. 2L) and decreased the GIR (Fig. 2J). However, the GRd did not change (Fig.

al

1K), which suggests an acute IR in vivo. However, ICV Sfrp5 attenuated the effects of lipid

rn

infusion on FFA and glucose metabolism (Fig. 2I-L). These data indicated that ICV Sfrp5

Jo u

improved lipid-induced IR.

3.4. Effect of ICV Sfrp5 on hepatic gene expression and insulin signaling Because ICV Sfrp5 significantly suppressed HGP, we explored whether ICV Sfrp5 altered the expression of gluconeogenic genes. As expected, HFD feeding increased the mRNA/protein expression of phosphoenolpyruate carboxykinase (PEPCK) and glucose-6-phosphatase (G6Pase) in the liver and inhibited hepatic insulin signaling. However, ICV Sfrp5 infusion in HFD-fed rats markedly reduced the gene/protein expression of PEPCK and G6Pase compared with ICV aCSF infusion (Fig. 3A and B), which suggests that the HGP-lowering effect of ICV Sfrp5 is due to the suppression of two key gluconeogenic enzymes. 13

Journal Pre-proof

To delineate the mechanisms by which ICV Sfrp5 regulates HGP and gluconeogenic enzymes, we next examined insulin receptor (InsR) and Akt kinase (Akt) phosphorylation in the liver. As shown in Figure 3B, HFD feeding reduced InsR and Akt phosphorylation in the liver. However, ICV Sfrp5 significantly attenuated the effects of HFD feeding on InsR and Akt phosphorylation. These data indicate that the impact of ICV Sfrp5 on glucose metabolism

f

was due to enhanced insulin signaling.

oo

3.5. DVC KATP-channel activation, DVC-NMDA receptors and hepatic vagus nerve are

pr

required for ICV Sfrp5 to regulate glucose metabolism

e-

To investigate whether DVC KATP channel activation was required for the HGP-lowering

Pr

effect of ICV Sfrp5, we used the KATP channel inhibitor glibenclamide and the KATP channel activator diazoxide to inhibit or activate DVC KATP channels during ICV Sfrp5 infusion (Fig.

al

4A). The co-infusion of ICV Sfrp5 and Gliben in HFD-fed rats attenuated the ability of

rn

central Sfrp5 to increase the GIR (Fig. 4B) and GRd (Fig. 4C) and inhibit HGP (Fig. 4D).

Jo u

Interestingly, similar to ICV Sfrp5, ICV diazoxide infusion in HFD rats markedly augmented the GIR (Fig. 4E) and GRd (Fig. 4F) and reduced HGP (Fig. 4G), further suggesting that ICV Sfrp5 may be an activator of the DVC KATP channel. To explore the role of the N-methyl-D-aspartate (NMDA) receptor in the effects of ICV Sfrp5 on HGP, HFD-fed rats were infused with Sfrp5 and MK-801 (an NMDA receptor blocker) into the ICV and DVC simultaneously (Fig. 5A). The concomitant infusion of ICV Sfrp5 with DVC MK-801 fully attenuated the ability of central Sfrp5 to augment the GIR (Fig. 5B) and GRd (Fig. 5C) and decrease the HGP (Fig. 5D). Finally, we infused Sfrp5 into the ICV of HFD-fed rats that underwent HVAG or SHAM to block the neuro-communication 14

Journal Pre-proof

from the brain to the liver to investigate the downstream pathway that NMDA receptors use to relay the signal generated by Sfrp5 (Fig. 5F). The results showed that the effects of ICV Sfrp5 on glucose kinetics were abolished in HFD rats that underwent HVAG compared to SHAM (Fig. 5G-I). Therefore, these data confirm that vagus nerve innervation is important for the effect of central Sfrp5 on HGP. In addition, consistent with changes in glucose metabolism,

f

the ability of ICV Sfrp5 to reduce PEPCK and G-6-Pase protein expression in the liver was

oo

abolished by glibenclamide (Fig. 4H), MK-801(Fig. 5E), and HVAG (Fig. 5J).

pr

3.6. ICV Sfrp5 infusion lowers hepatic VLDL-TG secretion and inhibits the hepatic

e-

expression of genes related to TG metabolism in HFD-fed rats

Pr

We next examined whether HFD feeding affected hepatic VLDL-TG secretion. As indicated in Figure S3A and S3B, HFD-fed rats pretreated with tyloxapol had significantly increased

al

plasma TG levels and VLDL-TG secretion compared to NCD-fed rats. These results indicate

rn

that HFD feeding stimulates VLDL-TG secretion in the liver.

Jo u

To further assess whether ICV Sfrp5 modulates VLDL-TG secretion, we infused the Sfrp5 protein intracerebroventricularly in NCD- or HFD-fed rats after tyloxapol pretreatment (Fig. 6A). Intravenous treatment with an inhibitor of lipoprotein lipase, tyloxapol, suppressed TG clearance from the circulation [18]. In the dose-response study, ICV Sfrp5 at a dose of 10 nmol/L (5 µl/h) was the minimal dose that yielded a significant hypolipidemic effect, and this dose was used for all VLDL-TG secretion experiments (Fig. S1C). This dose did not change plasma TG or VLDL-TG secretion at 3h when administered intravenously (Fig. S4). ICV Sfrp5 after tyloxapol pretreatment did not alter plasma TG levels in NCD-fed rats (Fig. 6B) at different time points and yielded no significant effect on 15

Journal Pre-proof

VLDL-TG secretion (Fig. 6C) or plasma FFA levels (Fig. 6D) compared with aCSF-treatment. Importantly, compared with ICV aCSF, ICV Sfrp5 inhibited the increase in plasma TG levels at the 60-180 time points in HFD-fed rats (Fig. 6E) and the rate of VLDL-TG secretion (Fig. 6F) in tyloxapol-treated rats, but plasma FFA levels were not changed (Fig. 6G).

f

Because ICV Sfrp5 significantly inhibited tyloxapol-induced hepatic VLDL-TG

oo

secretion, and we next examined the hepatic expression of genes related to TG metabolism

pr

using Western blotting. In HFD-fed rats, ICV Sfrp5 significantly decreased the protein

e-

expression of fatty acid synthase (FAS) and stearoyl CoA desaturase-1 (SCD-1), and

Pr

increased acetyl CoA carboxylase (ACC) phosphorylation levels in the liver compared to the ICV aCSF (Fig. 6H). These data indicate that ICV Sfrp5 decreases TG synthesis and hepatic

al

VLDL-TG secretion in HFD rats via suppressing enzymes related to lipogenesis in the liver.

rn

It has been reported that microsomal triglyceride transfer protein (MTP) plays a role in

Jo u

VLDL overproduction in the IR state [18, 27, 28]. Therefore, we measured the protein levels of MTP in the liver. The results showed that under the same dietary conditions, ICV Sfrp5 infusion did not lead to a change in MTP protein expression compared with ICV aCSF infusion (Fig. S5). In addition, H&E and Oil Red O staining showed that there was no significant difference in hepatic lipid accumulation between the ICV aCSF and ICV Sfrp5 infusion groups (Fig. S6 A and B). ICV Sfrp5 infusion also did not affect liver TG and TC contents (Fig. S6 C and D). Therefore, these results showed that acute ICV Sfrp5 infusion for 180 min did not affect MTP protein expression in the liver and hepatic steatosis. 3.7. DVC KATP-channel activation, DVC-NMDA receptor and hepatic vagus nerve are 16

Journal Pre-proof

required for ICV Sfrp5 to regulate VLDL-TG secretion in HFD-fed rats To investigate the role of DVC KATP-channels in the ICV Sfrp5 mediated inhibition of hepatic VLDL-TG secretion, we performed concurrent infusions of ICV Sfrp5 and Gliben or diazoxide in HFD-fed rats (Fig. 7A). The results showed that the co-infusion of ICV Sfrp5 and ICV Gliben blocked the ability of ICV Sfrp5 to inhibit circulating TG and hepatic

f

VLDL-TG secretion (Fig. 7B and C). In addition, the roles of ICV Sfrp5 in increasing ACC

oo

phosphorylation and decreasing FAS and SCD-1 expression were also blocked by ICV Gliben

pr

(Fig. 7D). As expected, ICV diazoxide infusion in NCD-fed rats reduced plasma TG and

e-

hepatic VLDL-TG secretion, and ICV Sfrp5 did not affect these responses (Fig. S7A and

Pr

S7B). However, ICV Sfrp5 and diazoxide infusion significantly decreased plasma TG and hepatic VLDL-TG secretion in HFD-fed rats (Fig. S7C and D). These results suggest that

al

ICV Sfrp5 and the KATP channel activator diazoxide have the similar roles in an IR state and

Jo u

TG production.

rn

that central KATP channel activation is important for ICV Sfrp5 signaling to regulate VLDL-

To explore whether NMDA receptor in the DVC mediates the role of ICV Sfrp5 on VLDL-TG production, ICV Sfrp5 and DVC MK-801 were performed in HFD rats (Fig. 7E). DVC MK-801 eliminated the decreased plasma TG and VLDL-TG secretion induced by ICV Sfrp5, but DVC MK-801 alone had no effect (Fig. 7F and G). The effects of ICV Sfrp5 on hepatic ACC, FAS and SCD-1 expression were also eliminated by DVC MK-801 (Fig. 7H). Next, we further examined the neurocircuitry of the effects of ICV Sfrp5 by repeating the VLDL-TG secretion experiment in HFD-fed rats subjected to HVAG (Fig. 7I). The hypolipidemic effects and inhibition of VLDL-TG secretion mediated by ICV Sfrp5 were 17

Journal Pre-proof

eliminated in HVAG rats compared to SHAM rats (Fig. 7 J and K). As expected, the ability of ICV Sfrp5 to regulate FAS and SCD-1 protein expression and ACC phosphorylation were abolished by HVAG (Fig. 7L). Collectively, these results indicate that ICV Sfrp5 regulates VLDL-TG production in the liver via DVC-hepatic vagal neural circuitry. 3.8. ICV Sfrp5 enhances central insulin sensitivity via an InsR-PI3K-Akt dependent

f

pathway

oo

To investigate the signaling pathway that is activated by ICV Sfrp5 in the hypothalamus, we

pr

first assessed the role of ICV Sfrp5 in InsR and Akt phosphorylation in the hypothalamus of

e-

rats. We found that upon the insulin stimulation, compared to ICV aCSF treatment, ICV Sfrp5

Pr

in HFD-fed rats increased InsR and Akt phosphorylation in the hypothalamus but not in the hypothalamus of NCD-fed rats (Fig. 8A). In parallel, Sfrp5 treatment in SH-SY5Y cells also

al

significantly increased InsR and Akt phosphorylation after insulin stimulation (Fig. 8B).

rn

Importantly, to avoid the shortcomings of SH-SY5Y cells, we treated PHNs isolated from

Jo u

fetal rats with Sfrp5. As shown in Figure 8C, treating PHNs with Sfrp5 also significantly increased insulin-stimulated InsR and Akt phosphorylation. Finally, we investigated the effect of ICV Sfrp5 on insulin’s ability to activate the phosphatidylinositol 3-kinase (PI3K) signaling in the hypothalamic ARC of HFD-fed rats and PHNs. Under insulin stimulation, PIP3 formation was significantly increased in the ARC of ICV Sfrp5 rats fed an HFD compared with ICV aCSF rats fed the same diet (Fig. 8D). In the basal state (saline treatment), immunoreactive PIP3 formation was lower both in both saline- and Sfrp5-treated PHNs. However, insulin-stimulated PIP3 formation was higher in Sfrp5-treated PHNs than those in aCSF-treated PHNs (Fig. 8E). These data suggest that Sfrp5 increases the ability of insulin to 18

Journal Pre-proof

Jo u

rn

al

Pr

e-

pr

oo

f

activate PI3 kinase in ARC neurons.

19

Journal Pre-proof

4.

Discussion

The brain-hepatic circuit is important in the regulation of peripheral metabolism [29-31]. The metabolic impact of insulin and other cytokines in the hypothalamus has garnered considerable interest in the past ten years. In animals, the hypothalamic action of these cytokines regulates peripheral glucose and lipid homeostasis, food intake and body weight

f

[32-34]. However, the signaling pathways of these cytokines in the hypothalamus have not

oo

been fully elucidated. The current study evaluated Sfrp5 in the hypothalamus as a regulator of

pr

peripheral metabolism in vivo. We demonstrated the following results: 1) central Sfrp5

e-

signaling activation reduced the 0- to 24-h cumulative food intake, increased energy

Pr

expenditure, and suppressed HGP and VLDL-TG production; 2) the glucoregulatory role of central Sfrp5 signaling was relative to activation of hepatic insulin signaling; 3) ICV Sfrp5

al

treatment suppressed the activity of enzymes related to lipogenesis in the liver, which

rn

decreased VLDL-TG secretion; and 4) An Sfrp5-dependent neurocircuitry was involved in

Jo u

the activation of the hypothalamic insulin signaling/ PI3K pathway and DVC NMDA receptors and KATP channel-mediated neurotransmission to the hepatic vagus nerve to suppress HGP and VLDL-TG secretion. Therefore, Sfrp5-dependent brain–liver neurocircuitry may have an important function in the maintenance of metabolism. In previous human studies, we and Hu et al. found a decrease in circulating Sfrp5 levels in obese and T2DM patients and a negative correlation between Sfrp5 and IR [12, 13]. An animal study also showed beneficial effects of Sfrp5 on improving hepatic IR [9]. Therefore, combining previous and current studies, we speculate that Sfrp5 may be a new target for the treatment of IR. However, Carstensen, et al. found that circulating 20

Journal Pre-proof

Sfrp5 was positively correlated with IR [35]. Another study in mice reported a detrimental effect on Sfrp5 on metabolism and obesity [36]. In addition, other studies showed that Sfrp5 may be a pro-diabetogenic factor by impairing the function and proliferation of islet β-cells [37-39]. Therefore, the relationship between Sfrp5 and metabolic disorders, as well as IR, is not clear, and there may be tissue specificity.

f

Although Sfrp5 is considered to be an adipocytokine, it is also expressed in the brain [9].

oo

In addition, peripheral Sfrp5 may cross the blood-brain barrier into the brain [14].

pr

Interestingly, a study in humans demonstrated that caloric restriction was related to

e-

circulating levels of Sfrp5 in obese subjects [40]. Therefore, it is important to investigate

Pr

whether brain Sfrp5 is sufficient to regulate food intake and energy homeostasis. Here, we first examined hypothalamic Sfrp5 expression in IR and obese mice. The downregulation of

rn

region under an IR state.

al

Sfrp5 expression in these mice resulted in Sfrp5 signaling attenuation in the hypothalamic

Jo u

We further found that an increase in Sfrp5 signaling in the hypothalamic region led to anorexia, which suggests that Sfrp5 is an anorectic hormone. However, Mori et al. found that Sfrp5 deficiency in the whole body did not alter food intake or energy homeostasis. The reason for this inconsistency is unknown, but it may be due to compensation by other Sfrp proteins and/or cytokines in the circulation or separate regulatory circuits in peripheral and central hormonal signaling systems. Nutrition or hormone signals in the hypothalamus control glucose equilibrium in insulin-sensitive organs [32]. Therefore, we hypothesized that Sfrp5 also exerted its glucoregulatory actions via a hypothalamic region of action. ICV infusion of Sfrp5 in 21

Journal Pre-proof

HFD-fed rats inhibited HGP and gluconeogenic gene expression and increased hepatic insulin signaling, which suggests a role for central Sfrp5 in the mediation of peripheral glucose homeostasis. However, Sfrp5 plays this role in only an IR state, and not under normal conditions in vivo. Therefore, the selective increase in central Sfrp5 signaling under the condition of IR appears to exert a powerful effect on HGP and insulin signaling in the

f

periphery. These data support the hypothesis that increasing Sfrp5 activation in the

oo

hypothalamus may be a novel strategy for the treatment of IR.

pr

HFD-induced obesity and genetically engineered obesity models in rodents are marked

e-

by hyperglycemia and hypertriglyceridemia. Glucoregulatory hormones also modulate VLDL

Pr

production [33]. Therefore, it is important to analyze the dual regulation of hormone signaling in glucose and lipid metabolism. Consistent with previous findings [28], HFD-induced obese

al

rats exhibited VLDL-TG hyperproduction and increased serum VLDL-TG levels. However,

rn

ICV Sfrp5 administration reduced the tyloxapol-induced hepatic VLDL-TG secretion and

Jo u

plasma TG increases associated with the suppression of lipogenetic genes in the liver, especially SCD-1, which is consistent with previous reports [18]. Based on the important role of SCD1 in the regulation of VLDL assembly and secretion, we hypothesized that hepatic SCD-1 suppression blocks the lipid transfer to apoB-lipoproteins during the assembly stages, which contributes to ICV Sfrp5-induced lowering of VLDL-TG production in liver. This hypothesis is supported by other findings showing that hypothalamic fatty acids, glycine and neuropeptide Y signaling regulated SCD-1 activation and VLDL-TG secretion in the liver [16, 29, 41]. Therefore, the cross-talk between the hypothalamus and liver promotes Sfrp5 signaling for lipoprotein production by curtailing hepatic SCD-1 activation. 22

Journal Pre-proof

The activation of DVC KATP channels has been shown to regulate glucose and lipid metabolism in response to nutrient- or hormone-dependent signals, such as leptin, and blockade of the KATP channel with glibenclamide yields the opposite effects [29, 32, 42]. The current study demonstrated that KATP channel blockade with DVC gliben was sufficient to eliminate the effects of ICV Sfrp5 on HGP and VLDL-TG production in the liver of HFD rats.

f

However, the infusion of a KATP-channel activator into the DVC was similar to the effect of

oo

ICV Sfrp5. These results confirmed that the effects of hypothalamic Sfrp5 on HGP and

pr

VLDL-TG production were dependent on the activity of the DVC K ATP channel.

e-

The ICV Sfrp5-induced decreases in HGP and VLDL-TG production raised questions

Pr

about which signaling pathway was related and how the signaling reached the liver to regulate metabolism. We found that ICV Sfrp5 promoted the insulin-mediated activation of InsR/PI3

al

kinase and the subsequent PIP3 formation and Akt activation in the hypothalamus, which

rn

resulted in KATP-channel activation. Therefore, these data provide evidence of the importance

Jo u

of the DVC KATP channel-InsR/PI3 pathway in the regulation of metabolism. Next, we found that the glucose- and lipid-lowering effects of ICV Sfrp5 were blocked by inhibiting the NMDA receptor. This finding indicates that ICV Sfrp5 acts via an NMDA receptor-dependent manner to inhibit HGP and VLDL-TG production and further confirms that receptor-mediated transmission integrates hypothalamic hormone signals to regulate metabolism. We and other laboratories previously demonstrated the particular importance of vagal innervation in the regulation of metabolism in the liver via central mechanisms [7, 29, 43]. Here, we investigated the neuronal circuit downstream of the effect of ICV Sfrp5 by repeating 23

Journal Pre-proof

the PEC and tyloxapol experiments in HFD-fed rats subjected to HVAG. The results indicated that vagal transmission was required for the regulation of glucose and lipid homeostasis by central Sfrp5. Therefore, liver autonomic innervation plays an important role in controlling metabolism. This brain-liver neurocircuitry may represent the glucose- and

Jo u

rn

al

Pr

e-

pr

oo

f

lipid-sensing mechanism from the brain to liver in the regulation of metabolism.

24

Journal Pre-proof

5. Conclusion We revealed Sfrp5 activation of the InsR-Akt PI3 kinase-KATP channel pathway in the hypothalamus as a brain–hepatic neurocircuitry underlying the Sfrp5 regulation of HGP and hepatic VLDL-TG secretion (Fig. 8F). This pathway has potential benefits in decreasing HGP and lipid levels under an IR state. Therefore, we believe that Sfrp5 has potential for the

Jo u

rn

al

Pr

e-

pr

oo

f

treatment of metabolic disorders via its central and peripheral roles.

25

Journal Pre-proof

Funding This study was supported by the Grants from the National Natural Science Foundation of China (No. 81570752, 81670755, 81300670, 81721001 and IRT1216); the Natural Science Foundation Project of CQ (No. cstc2015jcyjA10084 and cstc2013jcyjA10067); the Graduate

oo

f

Research and Innovation Project of CQ (CYB8154).

pr

Author Contributions

e-

Y.L., M.T., M.Y., J.C. and H.W. researched and analyzed the data. H.W. and W.D. provided

Pr

research material. Z.Z., H.Z. and D.L. directed the project, and contributed to discussion. L. L. and G. Y. wrote and edited the manuscript. L. L. is the guarantor of this work and, as such,

al

had full access to all the data in the study and takes responsibility for the integrity of the data

Disclosure

Jo u

rn

and the accuracy of the data analysis.

The authors declare no conflict of interest

26

Journal Pre-proof

References [1] Lam CK, Chari M, Rutter GA, Lam TK. Hypothalamic nutrient sensing activates a forebrain-hindbrain neuronal circuit to regulate glucose production in vivo. Diabetes. 2011;60:107-13. https://doi.org/10.2337/db10-0994 [2] Buettner C, Muse ED, Cheng A, Chen L, Scherer T, Pocai A, et al. Leptin controls adipose tissue lipogenesis

via

central,

STAT3-independent

mechanisms.

Nat

Med.

2008;14:667-75.

https://doi.org/10.1038/nm1775 [3] Knauf C, Cani PD, Perrin C, Iglesias MA, Maury JF, Bernard E, et al. Brain glucagon-like peptide-1 increases insulin secretion and muscle insulin resistance to favor hepatic glycogen storage. J Clin Invest. 2005;115:3554-63. https://doi.org/10.1172/jci25764 [4] Inoue H, Ogawa W, Asakawa A, Okamoto Y, Nishizawa A, Matsumoto M, et al. Role of hepatic

f

STAT3 in brain-insulin action on hepatic glucose production. Cell Metab. 2006;3:267-75.

oo

https://doi.org/10.1016/j.cmet.2006.02.009

[5] Yang M, Zhang Z, Wang C, Li K, Li S, Boden G, et al. Nesfatin-1 action in the brain increases 2012;61:1959-68. https://doi.org/10.2337/db11-1755

pr

insulin sensitivity through Akt/AMPK/TORC2 pathway in diet-induced insulin resistance. Diabetes. [6] Wu D, Yang M, Chen Y, Jia Y, Ma ZA, Boden G, et al. Hypothalamic nesfatin-1/NUCB2

e-

knockdown augments hepatic gluconeogenesis that is correlated with inhibition of mTOR-STAT3 signaling pathway in rats. Diabetes. 2014;63:1234-47. https://doi.org/10.2337/db13-0899

Pr

[7] Luo X, Li K, Zhang C, Yang G, Yang M, Jia Y, et al. Central administration of vaspin inhibits glucose production and augments hepatic insulin signaling in high-fat-diet-fed rat. Int J Obes (Lond) (2005). 2016;40:947-54. https://doi.org/10.1038/ijo.2016.24

al

[8] Taher J, Baker CL, Cuizon C, Masoudpour H, Zhang R, Farr S, et al. GLP-1 receptor agonism ameliorates hepatic VLDL overproduction and de novo lipogenesis in insulin resistance. Mol Metab.

rn

2014;3:823-33. https://doi.org/10.1016/j.molmet.2014.09.005 [9] Ouchi N, Higuchi A, Ohashi K, Oshima Y, Gokce N, Shibata R, et al. Sfrp5 Is an

Jo u

Anti-Inflammatory Adipokine That Modulates Metabolic Dysfunction in Obesity. Science. 2010;329:454-7. https://doi.org/10.1126/science.1188280 [10] Lv C, Jiang Y, Wang H, Chen B. Sfrp5 expression and secretion in adipocytes are up-regulated during differentiation and are negatively correlated with insulin resistance. Cell Biol Int. 2012;36:851-5. https://doi.org/10.1042/cbi20120054 [11] Koza RA, Nikonova L, Hogan J, Rim JS, Mendoza T, Faulk C, et al. Changes in gene expression foreshadow diet-induced obesity in genetically identical mice. PLoS Genet. 2006;2:e81. https://doi.org/10.1371/journal.pgen.0020081 [12] Hu W, Li L, Yang M, Luo X, Ran W, Liu D, et al. Circulating Sfrp5 is a signature of obesity-related metabolic disorders and is regulated by glucose and liraglutide in humans. J Clin Endocrinol Metab. 2013;98:290-8. https://doi.org/10.1210/jc.2012-2466 [13] Hu Z, Deng H, Qu H. Plasma SFRP5 levels are decreased in Chinese subjects with obesity and type 2 diabetes and negatively correlated with parameters of insulin resistance. Diabetes Res Clin Pract. 2013;99:391-5. https://doi.org/10.1016/j.diabres.2012.11.026 [14] Kastin AJ, Akerstrom V, Pan W. Interactions of glucagon-like peptide-1 (GLP-1) with the blood-brain barrier. J Mol Neurosci : MN. 2002;18:7-14. https://doi.org/10.1385/jmn:18:1-2:07 [15] Liang Q, Zhong L, Zhang J, Wang Y, Bornstein SR, Triggle CR, et al. FGF21 maintains glucose 27

Journal Pre-proof homeostasis by mediating the cross talk between liver and brain during prolonged fasting. Diabetes. 2014;63:4064-75. https://doi.org/10.2337/db14-0541 [16] Rojas JM, Bruinstroop E, Printz RL, Alijagic-Boers A, Foppen E, Turney MK, et al. Central nervous system neuropeptide Y regulates mediators of hepatic phospholipid remodeling and very low-density lipoprotein triglyceride secretion via sympathetic innervation. Mol Metab. 2015;4:210-21. https://doi.org/10.1016/j.molmet.2015.01.004 [17] la Fleur SE, Ji H, Manalo SL, Friedman MI, Dallman MF. The hepatic vagus mediates fat-induced inhibition

of

diabetic

hyperphagia.

Diabetes.

2003;52:2321-30.

https://doi.org/10.2337/diabetes.52.9.2321 [18] Stafford JM, Yu F, Printz R, Hasty AH, Swift LL, Niswender KD. Central nervous system neuropeptide Y signaling modulates VLDL triglyceride secretion. Diabetes. 2008;57:1482-90. https://doi.org/10.2337/db07-1702

f

[19] Yuan L, Luo X, Zeng M, Zhang Y, Yang M, Zhang L, et al. Transcription factor TIP27 regulates

oo

glucose homeostasis and insulin sensitivity in a PI3-kinase/Akt-dependent manner in mice. Int J Obes (Lond) (2005). 2015;39:949-58. https://doi.org/10.1038/ijo.2015.5

pr

[20] Zhang C, Luo X, Chen J, Zhou B, Yang M, Liu R, et al. Osteoprotegerin Promotes Liver Steatosis by Targeting the ERK–PPAR-γ–CD36 Pathway. Diabetes. 2019;68:1902-14.https://doi.org/ 10.2337 /db18-1055

e-

[21] Ozek C, Kanoski SE, Zhang ZY, Grill HJ, Bence KK. Protein-tyrosine phosphatase 1B (PTP1B) is a novel regulator of central brain-derived neurotrophic factor and tropomyosin receptor kinase B (TrkB)

Pr

signaling. J Biol Chem. 2014;289:31682-92. https://doi.org/10.1074/jbc.M114.603621 [22] Zhang Q, Yu J, Liu B, Lv Z, Xia T, Xiao F, et al. Central activating transcription factor 4 (ATF4) regulates hepatic insulin resistance in mice via S6K1 signaling and the vagus nerve. Diabetes.

al

2013;62:2230-9. https://doi.org/10.2337/db12-1050 [23] Elmquist JK, Ahima RS, Elias CF, Flier JS, Saper CB. Leptin activates distinct projections from

rn

the dorsomedial and ventromedial hypothalamic nuclei. Proc Natl Acad Sci U S A. 1998;95:741-6. https://doi.org/10.1073/pnas.95.2.741

Jo u

[24] Konner AC, Janoschek R, Plum L, Jordan SD, Rother E, Ma X, et al. Insulin action in AgRP-expressing neurons is required for suppression of hepatic glucose production. Cell Metab. 2007;5:438-49. https://doi.org/10.1016/j.cmet.2007.05.004 [25] Yang M, Wang J, Wu S, Yuan L, Zhao X, Liu C, et al. Duodenal GLP-1 signaling regulates hepatic glucose production through a PKC-delta-dependent neurocircuitry. Cell Death Dis. 2017;8:e2609. https://doi.org/10.1038/cddis.2017.28 [26] Wei Q, Zhou B, Yang G, Hu W, Zhang L, Liu R, et al. JAZF1 ameliorates age and diet-associated hepatic steatosis through SREBP-1c -dependent mechanism. Cell Death Dis. 2018;9:859. https:// doi.org/10.1038/s41419-018-0923-0 [27] Gong Z, Su K, Cui L, Tas E, Zhang T, Dong HH, et al. Central effects of humanin on hepatic triglyceride secretion. Am J Physio Endocrinol Metab. 2015;309:E283-E92.

https://doi.org/ 10.1152/

ajpendo.00043.2015 [28]Kamagate A, Qu S, Perdomo G, Su D, Kim DH, Slusher S, et al. FoxO1 mediates insulin-dependent regulation of hepatic VLDL production in mice. J Clin Invest. 2008;118:2347-64. https://doi.org/10.1172/jci32914 [29] Yue JT, Abraham MA, LaPierre MP, Mighiu PI, Light PE, Filippi BM, et al. A fatty acid-dependent hypothalamic-DVC neurocircuitry that regulates hepatic secretion of triglyceride-rich 28

Journal Pre-proof lipoproteins. Nat Commun. 2015;6:5970. https://doi.org/10.1038/ncomms6970 [30] Lam TK, Gutierrez-Juarez R, Pocai A, Bhanot S, Tso P, Schwartz GJ, et al. Brain glucose metabolism controls the hepatic secretion of triglyceride-rich lipoproteins. Nat Med. 2007;13:171-80. https://doi.org/10.1038/nm1540 [31] Stafford, J.M., Yu, F., Printz, R., Hasty, A.H., Swift, L.L., Niswender, K.D., 2008. Central nervous system neuropeptide Y signaling modulates VLDL triglyceride secretion. Diabetes 57(6):1482-90. https://doi.org/10.2337/db07-1702 [32] Qi Y, Takahashi N, Hileman SM, Patel HR, Berg AH, Pajvani UB, et al. Adiponectin acts in the brain to decrease body weight. Nat Med. 2004;10:524-9. https://doi.org/10.1038/nm1029 [33] Sarruf, D.A., Thaler, J.P., Morton, G.J., German, J., Fischer, J.D., Ogimoto, K., et al., 2010. Fibroblast growth factor 21 action in the brain increases energy expenditure and insulin sensitivity in obese rats. Diabetes 59(7):1817-24. https://doi.org/10.2337/db09-1878

f

[34] Cabou C, Vachoux C, Campistron G, Drucker DJ, Burcelin R. Brain GLP-1 signaling regulates

oo

femoral artery blood flow and insulin sensitivity through hypothalamic PKC-delta. Diabetes. 2011;60:2245-56. https://doi.org/10.2337/db11-0464

pr

[35] Carstensen M, Herder C, Kempf K, Erlund I, Martin S, Koenig W, et al. Sfrp5 correlates with insulin resistance and oxidative stress. Eur J Clin Invest. 2013; 43:350-7. https://doi.org/ 10.1111/ eci.12052

e-

[36] Mori H, Prestwich TC, Reid MA, Longo KA, Gerin I, Cawthorn WP, et al. Secreted frizzled-related protein 5 suppresses adipocyte mitochondrial metabolism through WNT inhibition. J

Pr

Clin Invest. 2012;122:2405-16. https://doi.org/10.1172/JCI63604 [37] Rulifson IC, Majeti JZ, Xiong Y, Hamburger A, Lee KJ, Miao L, et al. Inhibition of secreted frizzled-related protein 5 improves glucose metabolism. Am J Physio Endocrinol Metab.

al

2014;307:E1144-E52. https://doi.org/10.1152/ajpendo.00283.2014 [38] Guan B, Li W, Li F, Xie Y, Ni Q, Gu Y, et al. Sfrp5 mediates glucose-induced proliferation in rat

rn

pancreatic beta-cells. J Endocrinol. 2016;229:73-83. https://doi.org/10.1530/JOE-15-0535 [39] Rebuffat SA, Oliveira JM, Altirriba J, Palau N, Garcia A, Esteban Y, et al. Downregulation of

Jo u

Sfrp5 promotes beta cell proliferation during obesity in the rat. Diabetologia. 2013;56:2446-55. https://doi.org/10.1007/s00125-013-3030-x [40] Schulte DM, Muller N, Neumann K, Oberhauser F, Faust M, Gudelhofer H, et al. Pro-inflammatory wnt5a and anti-inflammatory sFRP5 are differentially regulated by nutritional factors in obese human subjects. PLoS One. 2012;7:e32437. https://doi.org/10.1371/journal.pone.0032437 [41] van den Hoek AM, Voshol PJ, Karnekamp BN, Buijs RM, Romijn JA, Havekes LM, et al. Intracerebroventricular neuropeptide Y infusion precludes inhibition of glucose and VLDL production by insulin. Diabetes. 2004;53:2529-34. https://doi.org/10.2337/diabetes.53.10.2529 [42] Pocai A, Lam TK, Gutierrez-Juarez R, Obici S, Schwartz GJ, Bryan J, et al. Hypothalamic K(ATP) channels control hepatic glucose production. Nature. 2005;434:1026-31. https://doi.org/10.1038/nature 03439 [43] Knight CM, Gutierrez-Juarez R, Lam TK, Arrieta-Cruz I, Huang L, Schwartz G, et al. Mediobasal hypothalamic SIRT1 is essential for resveratrol's effects on insulin action in rats. Diabetes. 2011;60:2691-700. https://doi.org/10.2337/db10-0987

29

Journal Pre-proof

Figure Legends Fig. 1 Effects of central Sfrp5 on energy expenditure and c-Fos immunoreactivity in the hypothalamus. (A-F) NCD- or HFD-fed rats received ICV Sfrp5 (1 μg per 2 μl for 1 min) or aCSF. Food intake (A), rectal temperature (B), Vo2 (C), Vco2 (D), and RER (E) were monitored for 24 h. Energy expenditure was calculated (F). (G) Photomicrographs of coronal

f

brain sections showing c-Fos immunostaining in the hypothalamus. NCD, normal chow diet;

oo

HFD, high-fat diet; ICV, intracerebroventricular; aCSF, artificial cerebrospinal fluid; VO2,

pr

oxygen consumption; RER, respiratory exchange ratio (VCO2/VO2). The data are the mean ±

Pr

e-

SEM (n = 5-6 for each group). *p < 0.05; **p < 0.01 vs. HFD-aCSF.

Fig. 2 Central Sfrp5 administration ameliorates insulin resistance in HFD-fed rats. (A) Blood

al

glucose levels and AUC glucose during GTT. (B) Serum insulin levels and AUCinsulin during GTT.

rn

(C) ICV infusion and insulin clamp protocol. Sfrp5 or aCSF was infused at a rate of 5 µl/h

Jo u

during the clamp through the ICV catheter. (D) Time course of the GIR and the average GIR. (E) GRd. (F) HGP. (G) Percentage of HGP suppression. (H) ICV infusion, lipid infusion and insulin clamp protocol. (I) plasma FFA levels during lipid infusion and insulin clamp procedures. (J) Time course of the GIR and the average GIR. (K) GRd. (L) HGP and the percentage of HGP suppression. NCD, normal chow diet; HFD, high-fat diet; aCSF, artificial cerebrospinal fluid; GTT, glucose tolerance tests; AUC, the area under the curve; ICV, intracerebroventricular; GIR, glucose infusion rate; GRd, glucose disposal rate; HGP, hepatic glucose production; FFA, free fatty acid. The data are the mean ± SEM (n = 5-6 for each group). *p < 0.05; **p < 0.01 vs. HFD-aCSF or aCSF-lipid. 30

Journal Pre-proof

Fig. 3. Central Sfrp5 administration inhibits gluconeogenesis and reinforces insulin signaling in the liver. (A) PEPCK and G6Pase mRNA expression. (B) PEPCK and G6Pase protein expression and InsR and Akt phosphorylation in the liver. NCD, normal chow diet; HFD, high-fat diet; aCSF, artificial cerebrospinal fluid. The data are the mean ± SEM of at least 2-3

f

independent experiments. *p < 0.05; **p < 0.01 vs. HFD-aCSF.

oo

Fig. 4. Activation of DVC KATP channels is required for ICV Sfrp5 to inhibit glucose

pr

production. (A) Experimental procedure and clamp protocol. Eight-week-old rats were fed a

e-

HFD for 12 weeks. An ICV catheter was implanted in the rats on day 1. Internal jugular vein

Pr

and carotid artery cannulations were performed on day 7, and the pancreatic clamp protocol was performed on day 12. aCSF, Sfrp5 or/and glibenclamide or diazoxide were infused

al

during the clamp through the ICV catheter. (B-D) Time course of the GIR and the average

rn

GIR (B), GRd (C), HGP and percentage of HGP suppression (D) during the clamp procedure

Jo u

and Gliben treatment. (E-F) Time course of the GIR and the average GIR (E), GRd (F) and HGP and percentage of HGP suppression (G) during the clamp procedure and diazoxide treatment. (H) Hepatic PEPCK and G6Pase protein expression. ICV, intracerebroventricular; aCSF, artificial cerebrospinal fluid; Gliben, glibenclamide. The data are the mean ± SEM (n = 5-6 for each group). *p < 0.05; **p < 0:01 vs. aCSF or other groups.

Fig. 5. Disruption of the DVC NMDA receptor or HAVG eliminates the ability of ICV Sfrp5 to inhibit glucose production. (A) Schematic representation of the working hypothesis. ICV Sfrp5 rats fed an HFD were administered with the NMDA channel inhibitor MK-801 into the 31

Journal Pre-proof

DVC. (B-D) Time course of the GIR and the average GIR (B), GRd (C), HGP and percentage of HGP suppression (D) during the clamp procedure. (E) Hepatic PEPCK and G6Pase protein expression. (F) Schematic representation of the working hypothesis. Sfrp5 was infused into the ICV of HFD-fed rats subjected to HVAG. (G-I) Time course of the GIR and the average GIR (G), GRd (H), HGP and suppression of HGP (I) during the clamp. (J) Hepatic PEPCK

f

and G6Pase protein expression. ICV, intracerebroventricular; HAVG, hepatic vagotomy;

oo

DVC, dorsal vagal complex; NMDA, N-methyl-D-aspartate; aCSF, artificial cerebrospinal

pr

fluid. The data are the mean ± SEM (n = 3-6 for each group). *p < 0.05; **p < 0.01 vs. other

Pr

e-

groups.

Fig. 6. Effect of central Sfrp5 signaling on hepatic VLDL-TG secretion and gene expression

al

involved in TG metabolism. (A) Experimental protocol. ICV infusion of vehicle or Sfrp5

rn

commenced at 0 min after blood samples were obtained and followed by an intravenous

Jo u

injection of tyloxapol (600 mg/kg) as indicated in the Methods. (B-D) plasma TG levels (B), VLDL-TG secretion (C) and plasma FFA levels (D) in NCD-fed rats. (E-G) plasma TG levels (E), VLDL-TG secretion (F) and plasma FFA levels (G) in HFD-fed rats. (H) Hepatic gene expression involved in TG metabolism. VLDL-TG, triglyceride-rich very-low-density lipoproteins; NCD, normal chow diet; HFD, high-fat diet; aCSF, artificial cerebrospinal fluid; ICV, intracerebroventricular. The data are the mean ± SEM (n = 3-6 for each group). *p < 0.05; **

p < 0.01 vs. HFD-aCSF.

Fig. 7. Activation of KATP channels and NMDA receptors in the DVC and hepatic vagus 32

Journal Pre-proof

nerve are required for central Sfrp5 signaling to inhibit HGP. (A) Experimental protocol. Eight- week-old rats were fed an HFD for 12 weeks. An ICV catheter was implanted on rats on day 1. Internal jugular vein and carotid artery cannulations were performed on day 7, and the tyloxapol experiments were performed on day 12 as indicated in the Methods. (B and C) Plasma TG levels (B) and VLDL-TG secretion (C) after the Gliben treatment. (D) Protein

f

expression of FAS and SCD-1 and the ACC phosphorylation in the liver. (E) Experimental

oo

protocol. The rats were fed an HFD for 12 weeks and stereotaxically implanted with

pr

indwelling catheters into the DVC. ICV aCSF or Sfrp5 and DVC saline or MK-801infusions

e-

commenced at t = 0 min as indicated in the Methods. (F) Plasma TG levels. (G) VLDL-TG

Pr

secretion. (H) Protein expression of FAS and SCD-1 and the ACC phosphorylation in the liver. (I) Experimental protocol for the effect of HVAG on ICV Sfrp5-mediated VLDL-TG

al

secretion in HFD-fed rats subjected to HVAG. (J) Plasma TG levels. (K) VLDL-TG secretion.

rn

(L) Protein expression of FAS and SCD-1, and the ACC phosphorylation in the liver. DVC,

Jo u

dorsal vagal complex; aCSF, artificial cerebrospinal fluid; ICV, intracerebroventricular; NMDA, N-methyl -D-aspartate; HVAG, hepatic vagotomy; Gliben, glibenclamide; VLDL-TG, triglyceride-rich very-low-density lipoproteins. The data are the mean ± SEM (n = 3-5 for each group). **p < 0.01 vs. other groups.

Fig. 8. Hypothalamic Sfrp5 enhances central insulin sensitivity via an InsR-PI3K-Aktdependent pathway. (A) NCD- or HFD-fed rats received an ICV infusion of Sfrp5, and the indicated amounts of insulin were infused for 5 min bilaterally into the DVC, which were sacrificed after 15 min. InsR and Akt phosphorylation were analyzed by Western blotting. (B 33

Journal Pre-proof

and C) SH-SY5Y cells (B) and PHNs (C) were treated as indicated in the Methods. The levels of the InsR and Akt phosphorylation were analyzed and normalized to the total protein levels. IR, cells were incubated with glucosamine to induce insulin-resistance; no-IR, the cells were incubated without glucosamine. The data are the mean ± SEM. *p < 0.05, **p < 0.01 vs. HFD-aCSF or lane 7. (D) PIP3 formation in the ARC neurons. The rats were fed a HFD for

f

12 weeks and received an ICV infusion of aCSF or Sfrp5. Immunohistochemistry was

oo

performed on ARC sections from overnight-fasted rats, which were injected either aCSF or

pr

insulin into the MBH and sacrificed 10 min after stimulation. Blue, DAPI; Red, PIP3 (n = 3).

e-

(E) PIP3 formation in PHNs. The PHNs were cultured and treated as indicated in the Methods.

Pr

The amount of PIP3 was classified as high (arrows) or moderate (arrowheads). Quantification of PIP3 cells displaying moderate to high PIP3 levels (bottom panel). Blue, DAPI; Green,

al

PIP3; PBS, phosphate-buffered saline; DMEM, Dulbecco’s modified Eagle’s medium. The

rn

data are the mean ± SEM. **p < 0.01 vs. PBS + insulin. (F) Schematic representation of the

Jo u

working model. Hypothalamic Sfrp5 signaling activates the insulin-mediated InsR-Akt PI3K-KATP channel pathway in the hypothalamus and initiates communication with the liver via hypothalamus–DVC–hepatic vagus neurocircuitry to regulate HGP and hepatic VLDL-TG secretion.

34

Journal Pre-proof

Figures

Jo u

rn

al

Pr

e-

pr

oo

f

Fig.1

35

Journal Pre-proof

Jo u

rn

al

Pr

e-

pr

oo

f

Fig. 2

36

Journal Pre-proof

Jo u

rn

al

Pr

e-

pr

oo

f

Fig. 3

37

Journal Pre-proof

Jo u

rn

al

Pr

e-

pr

oo

f

Fig. 4

38

Journal Pre-proof

Jo u

rn

al

Pr

e-

pr

oo

f

Fig. 5

39

Journal Pre-proof

Jo u

rn

al

Pr

e-

pr

oo

f

Fig. 6

40

Journal Pre-proof

Jo u

rn

al

Pr

e-

pr

oo

f

Fig. 7

41

Journal Pre-proof

Jo u

rn

al

Pr

e-

pr

oo

f

Fig. 8

42

Journal Pre-proof

Highlights • Secreted frizzled-related protein 5 protein level is down-regulated in the hypothalamus of obese animal models. • ICV sfrp5 in HFD-fed rats can reduce food intake and increase energy expenditure through activating neurons of the hypothalamic BMH.

f

• ICV Sfrp5 improves N-methyl-D-aspartate（NMDA）receptor-mediated transmission of

oo

dorsal vagal complex (DVC) to enhance hepatic glucose metabolism and decrease VLDL-TG

pr

secretion by the hepatic vagus.

Jo u

rn

al

Pr

receptor-PI3K-Akt dependent pathway.

e-

• Central Sfrp5 potentiates insulin sensitivity of hypothalamic BMH via an insulin

43