Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Oct 11:12:e86452.
doi: 10.7554/eLife.86452.

GDF15 is required for cold-induced thermogenesis and contributes to improved systemic metabolic health following loss of OPA1 in brown adipocytes

Jayashree Jena #  1 , Luis Miguel García-Peña #  1 , Eric T Weatherford  1 , Alex Marti  1 , Sarah H Bjorkman  1 , Kevin Kato  1 , Jivan Koneru  1 , Jason H Chen  1 , Randy J Seeley  2 , E Dale Abel  1 , Renata O Pereira  1
Affiliations

GDF15 is required for cold-induced thermogenesis and contributes to improved systemic metabolic health following loss of OPA1 in brown adipocytes

Jayashree Jena et al. Elife. .

Abstract

We previously reported that mice lacking the protein optic atrophy 1 (OPA1 BKO) in brown adipose tissue (BAT) display induction of the activating transcription factor 4 (ATF4), which promotes fibroblast growth factor 21 (FGF21) secretion as a batokine. FGF21 increases metabolic rates under baseline conditions but is dispensable for the resistance to diet-induced obesity (DIO) reported in OPA1 BKO mice (Pereira et al., 2021). To determine alternative mediators of this phenotype, we performed transcriptome analysis, which revealed increased levels of growth differentiation factor 15 (GDF15), along with increased protein kinase R (PKR)-like endoplasmic reticulum kinase (PERK) levels in BAT. To investigate whether ATF4 induction was mediated by PERK and evaluate the contribution of GDF15 to the resistance to DIO, we selectively deleted PERK or GDF15 in OPA1 BKO mice. Mice with reduced OPA1 and PERK levels in BAT had preserved ISR activation. Importantly, simultaneous deletion of OPA1 and GDF15 partially reversed the resistance to DIO and abrogated the improvements in glucose tolerance. Furthermore, GDF15 was required to improve cold-induced thermogenesis in OPA1 BKO mice. Taken together, our data indicate that PERK is dispensable to induce the ISR, but GDF15 contributes to the resistance to DIO, and is required for glucose homeostasis and thermoregulation in OPA1 BKO mice by increasing energy expenditure.

Keywords: GDF15; PERK; biochemistry; brown adipose tissue; chemical biology; integrated stress response; mitochondrial stress; mouse.

PubMed Disclaimer

Conflict of interest statement

JJ, LG, EW, AM, SB, KK, JK, JC, EA, RP No competing interests declared, RS R.J.S. has received research support from Fractyl, Novo Nordisk, AstraZeneca, and Eli Lilly. R.J.S. has served on scientific advisory boards for Novo Nordisk, CinRX, Scohia, Fractyl and Structure Therapeutics. R.J.S. is a stakeholder of Calibrate and Rewind

Figures

Figure 1.
Figure 1.. OPA1 BKO mice have higher resting metabolic rates and are completely resistant to diet-induced obesity (DIO).
(A–D) Data collected on 12-week-old male OPA1 BKO mice. (A) Body mass. (B–G) Indirect calorimetry data represented as the average for the light and dark cycles during the last 48 hr of data recording. (B) Food intake. (C) Locomotor activity. (D) Oxygen consumption (VO2). (E) Carbon dioxide production (VCO2). (F) Respiratory exchange ratio. (G) Energy expenditure. (H) ANCOVA of energy expenditure as a function of body mass. (I–Q) Data collected on 18-week-old male OPA1 BKO mice fed high-fat diet (60% calories from fat) for 12 wk (diet started at 6 wk of age). (I) Body weight curve. (J) Final body weight (cohort used for indirect calorimetry data collection). (K–Q) Indirect calorimetry data represented as the average for the light and dark cycles during the last 48 hr of data recording. (K) Food intake. (L) Locomotor activity. (M) Oxygen consumption (VO2). (N) Carbon dioxide production (VCO2). (O) Respiratory exchange ratio. (P) Energy expenditure. (Q) ANCOVA of energy expenditure as a function of body mass. Data are expressed as means ± SEM. Significant differences were determined by Student’s t-test using a significance level of p<0.05. *Significantly different vs. wild-type (WT) mice.
Figure 1—figure supplement 1.
Figure 1—figure supplement 1.. Food consumption, locomotor activity, and energy expenditure in OPA1 BAT knockout (KO) mice under baseline conditions and following high-fat feeding.
(A–D) Hourly indirect calorimetry data during the light and dark cycles in the last 48 hr of data recording. Data was collected from 12-week-old male OPA1 BAT KO and control mice, consuming normal chow. (A) Hourly food consumed. (B) Locomotor activity. (C) Respiratory exchange ratio. (D) Energy expenditure. (E–I) Hourly indirect calorimetry data during the light and dark cycles in the last 48 hr of data recording. Data was collected from 18-week-old male OPA1 BAT KO and wild-type (WT) control mice fed high-fat diet (HFD – 60% calories from fat) for 12 wk (diet started at 6 wk of age). (E) Hourly food consumption. (F) Cumulative food consumption. (G) Locomotor activity. (H) Respiratory exchange ratio. (I) Energy expenditure. (J) ANCOVA of oxygen consumption as a function of body mass for HFD mice. (K) Serum GDF15 levels in OPA1 BKO mice under baseline conditions, after cold exposure (3 d at 4°C) and after 12 wk of high-fat feeding. Data are expressed as means ± SEM. Significant differences for the group effect were determined by ANCOVA using a significance level of p<0.05. One-Way ANOVA was used to determine the differences in GDF15 serum levels using a significance level of p<0.05. ** p <0.01. **** p <0.0001. Significantly different vs. WT RT.
Figure 2.
Figure 2.. Transcriptome analysis reveals induction of the unfolded protein response (UPR) and GDF15 in mice lacking OPA1 in brown adipose tissue (BAT).
(A–C) Data collected from BAT of 7-week-old male and female OPA1 BKO mice. (A) Bubble plot showing the top 10 canonical pathways from the Ingenuity Pathway Analysis (IPA) database containing genes with a significant overlap (adjusted p-value≤0.05) to those differentially expressed in OPA1 BKO mice. Size of the bubble indicates the -log of the adjusted p-value (Benjamini–Hochberg) from the pathway analysis. Plotted on the x-axis is the ratio of differentially expressed genes relative to the number of genes in the pathway. The bubble color indicates the z-score, which indicates the predicted activation (positive) or inhibition (negative) of the pathway based on the directionality of the gene changes in OPA1 BKO relative to wild-type (WT) mice. (B) Heatmap of the top 25 differentially expressed genes in OPA1 BKO mice. (C) Bar graph showing the Log2 transcript per million (TPM) for the Eif2ka3 gene (encoding PERK). (D) Relative mRNA expression of Gdf15 in BAT of OPA1 BKO and in OPA1/ATF4 BAT DKO mice normalized to tata box protein (Tbp) expression. (E) GDF15 serum levels in OPA1 BKO, OPA1/ATF4 BAT DKO, and OPA1/FGF21 BAT DKO mice. Data are expressed as means ± SEM. Significant differences were determined by Student’s t-test using a significance level of p<0.05. * p <0.05; ** p <0.01. Significantly different vs. WT mice or from OPA1 KO.
Figure 3.
Figure 3.. PERK is dispensable for the integrated stress response (ISR) activation in OPA1 BKO mice.
(A–G) Data collected from 8-week-old OPA1/PERK BAT double-knockout (DKO) male and female mice. (A) Representative immunoblots for OPA1 and PERK protein in brown adipose tissue (BAT) normalized to GAPDH and their respective densitometric quantification. (B) Representative immunoblots for phospho-eIF2α (peIF2α) normalized to total eIF2α protein levels in BAT and the respective densitometric quantification. Optical density (OD). (C) Relative mRNA expression of the ISR genes Atf4, Ddit3, Fgf21, and Gdf15 normalized to Tbp expression in BAT. (D) GDF15 serum levels (ad libitum-fed conditions). (E) Relative mRNA expression of the thermogenic genes normalized to tata box protein (Tbp) expression in BAT. (F) Representative immunoblots for uncoupling protein 1 (UCP1) protein levels in BAT normalized to β-actin and the respective densitometric quantification. (G) ADP-stimulated (state 3) pyruvate-malate-supported oxygen consumption rates (OCRs) in mitochondria isolated from BAT. (H) Relative mRNA expression of the thermogenic genes normalized to Tbp expression in inguinal white adipose tissue (iWAT). (I) Representative immunoblots for UCP1 in iWAT normalized to β-actin and the respective densitometric quantification. (J) Representative immunoblots for tyrosine hydroxylase (TH) protein levels in iWAT normalized to β-actin and the respective densitometric quantification. OD, optical density. (K–M) Data collected in 20-week-old male DKO mice. (K) Body mass. (L) Total fat mass. (M) Total lean mass. Data are expressed as means ± SEM. Significant differences were determined by Student’s t-test using a significance level of p<0.05. *Significantly different vs. wild-type (WT) mice.
Figure 4.
Figure 4.. OPA1/PERK brown adipose tissue (BAT) double-knockout (DKO) mice are resistant to DIO and insulin resistance.
(A–O) Data collected in OPA1/PERK BAT DKO male mice fed a high-fat diet (HFD) for 12 wk. (A) Body weight curve comparing wild-type (WT), OPA1 BKO, and OPA1/PERK BAT DKO mice. (B) Final body weight. (C) Percent fat mass. (D) Percent lean mass. (E–K) Indirect calorimetry data represented as the average for the light and dark cycles during the last 48 hr of data recording. (E) Food intake. (F) Locomotor activity. (G) Oxygen consumption (VO2). (H) Carbon dioxide production (VCO2). (I) Respiratory exchange ratio. (J) Energy expenditure. (K) ANCOVA of energy expenditure as a function of body mass. (L) Glucose tolerance test (GTT). (M) Area under the curve (AUC) for the GTT. (N) Insulin tolerance test (ITT). (O) AUC for the ITT. Data are expressed as means ± SEM. Significant differences were determined by Student’s t-test using a significance level of p<0.05. *Significantly different vs. WT mice.
Figure 4—figure supplement 1.
Figure 4—figure supplement 1.. Food consumption locomotor activity and energy expenditure in OPA1/PERK brown adipose tissue (BAT) double-knockout (DKO) mice.
(A–F) Hourly indirect calorimetry data during the light and dark cycles in the last 48 hr of data recording. Data was collected from 18-week-old male OPA1/PERK BAT DKO and wild-type (WT) control mice fed high-fat diet (60% calories from fat) for 12 wk (diet started at 6 wk of age). (A) Hourly food consumption. (B) Cumulative food consumption. (C) Locomotor activity. (D) Cumulative distance in the cage. (E) Respiratory exchange ratio. (F) Energy expenditure. Data are expressed as means ± SEM.
Figure 5.
Figure 5.. OPA1/GDF15 double-knockout (DKO) mice have similar metabolic phenotype as OPA1 BKO mice under baseline conditions.
(A–O) Data collected in OPA1/GDF15 brown adipose tissue (BAT) DKO mice under baseline conditions (room temperature or 30°C/regular chow). (A) Relative mRNA expression of Opa1 and Gdf15 in BAT normalized to tata box protein (Tbp) expression. (B). GDF15 serum levels (ad libitum-fed conditions). (C) BAT mass relative to body mass in 8-week-old mice. (D) Representative images of H&E-stained histological sections of BAT (n = 3). Scale bar = 50 µm. (E) Electron micrographs of BAT. Scale bar = 2 µm. (F) Relative mRNA expression of the thermogenic genes in BAT normalized to Tbp expression. (G) Representative immunoblots for UCP1 in BAT normalized to β-actin, and their respective densitometric quantification. Optical density (OD). (H) ADP-stimulated (state 3) pyruvate-malate-supported oxygen consumption rates (OCRs) in mitochondria isolated from BAT. (I) Relative mRNA expression of the integrated stress response (ISR) genes Atf4, Ddit3, and Fgf21 normalized to Tbp expression. (J) FGF21 serum levels (ad libitum-fed conditions). (K) Representative images of H&E-stained histological sections of inguinal white adipose tissue (iWAT) (n = 3). Scale bar = 50 µm. (L) Relative mRNA expression of the thermogenic genes in iWAT normalized to Tbp expression. (M) Representative immunoblots for UCP1 in iWAT normalized to β-actin and their respective densitometric quantification. (N) Representative immunoblots for tyrosine hydroxylase (TH) in iWAT normalized to β-actin and their respective densitometric quantification. OD, optical density. (O) Body mass at 8, 10, and 20 wk of age. (P–U) Indirect calorimetry data represented as the average for the light and dark cycles during the last 48 hr of data recording in male mice around 10–12 wk of age. (P) Food intake. (Q) Locomotor activity. (R) Energy expenditure. (S) Oxygen consumption (VO2). (T) Carbon dioxide production (VCO2). (U) Respiratory exchange ratio. Data are expressed as means ± SEM. Significant differences were determined by Student’s t-test using a significance level of p<0.05. *Significantly different vs. wild-type (WT) mice.
Figure 5—figure supplement 1.
Figure 5—figure supplement 1.. GDF15 expression in wild-type thermogenic adipocytes does not influence energy metabolism, glucose homeostasis and core body temperature under baseline conditions.
(A–J) Data collected in 6-week-old GDF15 BKO male mice and their wild-type (WT) control fed regular chow and under ambient temperature conditions. (A) Relative mRNA expression of Gdf15 in brown adipose tissue (BAT) normalized to tata box protein (Tbp). (B) Relative mRNA expression of Gdf15 in inguinal white adipose tissue (iWAT) normalized to Tbp. (C) GDF15 serum levels. (D) Core body temperatures measured with a rectal probe. (E) Body mass. (F) Total fat mass. (G) Total lean mass. (H) Glucose tolerance test (GTT). (I) Area under the curve (AUC) for the GTT. (J) Fasting blood glucose levels. Data are expressed as means ± SEM. Significant differences were determined by Student’s t-test using a significance level of p<0.05. *Significantly different vs. wild-type (WT) mice.
Figure 5—figure supplement 2.
Figure 5—figure supplement 2.. Metabolic phenotyping of OPA1/GDF15 double-knockout (DKO) mice under baseline conditions.
(A–F) Hourly indirect calorimetry data in the light and dark cycles during the last 48 hr of data recording. Data were collected in male DKO and wild-type (WT) control mice between 10–12 wk of age (baseline conditions). (A) Hourly food intake. (B) Cumulative food consumption. (C) Locomotor activity. (D) Energy expenditure. (E) ANCOVA of oxygen consumption as a function of body mass. (F) Respiratory exchange ratio. (G) Glucose tolerance test (GTT). (H) Area under the curve (AUC) for the GTT. (I) Fasting glucose levels. Data are expressed as means ± SEM. Significant differences for the group effect were determined by ANCOVA using a significance level of p<0.05.
Figure 6.
Figure 6.. GDF15 partially mediates the resistance to diet-induced obesity (DIO) and is required to improve glucose homeostasis in OPA1 BKO mice.
(A–O) Data collected in OPA1/GDF15 brown adipose tissue (BAT) double-knockout (DKO) male mice fed a high-fat diet (HFD) for 12 wk. (A) Body weight curve comparing wild-type (WT), OPA1 BKO, and OPA1/GDF15 BAT DKO mice. (B) Final body mass. (C) Total fat mass. (D) Total lean mass. (E–J) Indirect calorimetry data represented as the average for the light and dark cycles during the last 48 hr of data recording. (E) Food intake. (F) Locomotor activity. (G) Energy expenditure. (H) Oxygen consumption (VO2). (I) Carbon dioxide production (VCO2). (J) Respiratory exchange ratio. (K) Glucose tolerance test (GTT). (L) Area under the curve (AUC) for the GTT. (M) Fasting glucose levels. (N) Liver triglycerides levels. (O) Fasting insulin levels. (P) Insulin tolerance test (ITT). (Q) AUC for the ITT. (R) Relative mRNA expression of thermogenic genes in BAT normalized to tata box protein (Tbp) expression. (S) Representative immunoblots for UCP1 in BAT normalized to β-actin and their respective densitometric quantification. Optical Density (OD). (T) Relative mRNA expression of thermogenic genes in inguinal white adipose tissue (iWAT) normalized to Tbp expression. (U) Representative immunoblots for UCP1 in iWAT normalized to β-actin. OD, optical density. (V) Representative immunoblots for Serca1a in gastrocnemius muscle normalized to Ponceau red staining and their respective densitometric quantification. OD, optical density. Data are expressed as means ± SEM. Significant differences were determined by Student’s t-test using a significance level of p<0.05. *Significantly different vs. WT mice.
Figure 6—figure supplement 1.
Figure 6—figure supplement 1.. Indirect calorimetry data and immunoblot for thermogenic markers in white adipose tissue (iWAT) and skeletal muscle.
(A–H) Hourly indirect calorimetry data during the light and dark cycles in the last 48 hr of data recording. Data was collected from 18-week-old male OPA1/GDF15 brown adipose tissue (BAT) double-knockout (DKO) and wild-type (WT) control mice fed high-fat diet (HFD) (60% calories from fat) for 12 wk (diet started at 6 wk of age). (A) Hourly food consumption. (B) Cumulative food consumption. (C) Locomotor activity. (D) Cumulative distance in the cage. (E) Hourly energy expenditure. (F) Cumulative energy expenditure. (G) ANCOVA of oxygen consumption as a function of body mass. (H) Respiratory exchange ratio. (I) Representative immunoblots for UCP1 in iWAT normalized to β-actin in OPA1 BAT KO mice after 12 wk of HFD and their respective densitometric quantification. (J) Representative immunoblots for Serca1a in gastrocnemius muscle normalized to Ponceau red staining in OPA1 BAT KO mice after 12 wk of HFD and their respective densitometric quantification. Optical density (OD). Data are expressed as means ± SEM. Significant differences were determined by ANCOVA for the group effect and by Student’s t-test for the immunoblots using a significance level of p<0.05. *Significantly different vs. WT mice.
Figure 7.
Figure 7.. GDF15 is required to regulate core body temperature in cold-exposed OPA1 BKO mice.
(A) Averaged core body temperature (light and dark cycles) collected from 12-week-old wild-type (WT) and OPA1/GDF15 brown adipose tissue (BAT) double-knockout (DKO) mice after 7 d at 30°C. (B) Averaged core body temperature (light and dark cycles) in mice cold exposed for 24 hr (4°C). (C) Final core body temperature recorded by telemetry in mice exposed to 4°C in the CLAMS system. (D–G) Indirect calorimetry data represented as the average for the light and dark cycles during the first 24 hr of data recording (4°C). (D) Energy expenditure. (E) Respiratory exchange ratio. (F) Food intake. (G) Locomotor activity. (H) Hourly core body temperatures collected from 12-week-old WT and DKO mice during cold exposure (4°C). (I–N) Data collected after 5 hr of cold exposure. (I) Relative mRNA expression of thermogenic genes in BAT after 5 hr of cold exposure normalized to tata box protein (Tbp). (J) Representative immunoblots for UCP1 in BAT normalized to β-actin and their respective densitometric quantification. (K) Pyruvate-malate-supported oxygen consumption rates (OCRs) and UCP1-dependent respirations in mitochondria isolated from BAT (baseline conditions). (L) Relative mRNA expression of thermogenic genes in inguinal white adipose tissue (iWAT) normalized to Tbp expression. (M) Representative immunoblots for UCP1 in iWAT normalized to β-actin and their respective densitometric quantification. (N) Representative immunoblots for Serca1a in gastrocnemius muscle normalized to Ponceau red staining and their respective densitometric quantification. Optical density (OD). Data are expressed as means ± SEM. Significant differences were determined by Student’s t-test using a significance level of p<0.05. * p <0.05; ** p <0.01. Significantly different vs. WT mice.
Figure 7—figure supplement 1.
Figure 7—figure supplement 1.. Indirect calorimetry, food intake, locomotor activity, and skeletal muscle characterization of cold-exposed OPA1/GDF15 brown adipose tissue (BAT) double-knockout (DKO) and OPA1 BAT knockout (KO) mice.
(A–H) Hourly indirect calorimetry data collected in OPA1 BKO and OPA1/GDF15 BAT DKO mice and their respective wild-type (WT) controls during the first 24 hr of cold exposure. (A) Oxygen consumption in OPA1 BKO mice. (B) Oxygen consumption in DKO mice. (C) Respiratory exchange ratio in OPA1 BKO mice. (D) Respiratory exchange ratio in DKO mice. (E) Food consumed in OPA1 BKO mice. (F) Food consumed in DKO mice. (G) Locomotor activity in OPA1 BKO mice. (H) Locomotor activity in DKO mice. (I) Representative immunoblots for tyrosine hydroxylase (TH) in inguinal white adipose tissue (iWAT) from mice exposed to 4°C for 5 hr normalized to β-actin and their respective densitometric quantification. (J) Representative immunoblots for GLUT1 in skeletal muscle from mice exposed to 4°C for 5 hr normalized to β-actin and their respective densitometric quantification. Optical density (OD). Data are expressed as means ± SEM. Significant differences were determined by ANCOVA for the group effect.
See this image and copyright information in PMC

Update of

  • doi: 10.1101/2022.12.23.521796

Similar articles

See all similar articles

Cited by

References

    1. Aguilar-Recarte D, Barroso E, Palomer X, Wahli W, Vázquez-Carrera M. Knocking on GDF15’s door for the treatment of type 2 diabetes mellitus. Trends in Endocrinology and Metabolism. 2022;33:741–754. doi: 10.1016/j.tem.2022.08.004. - DOI - PubMed
    1. Baek SJ, Eling T. Growth differentiation factor 15 (GDF15): A survival protein with therapeutic potential in metabolic diseases. Pharmacology & Therapeutics. 2019;198:46–58. doi: 10.1016/j.pharmthera.2019.02.008. - DOI - PMC - PubMed
    1. Bao XR, Ong S-E, Goldberger O, Peng J, Sharma R, Thompson DA, Vafai SB, Cox AG, Marutani E, Ichinose F, Goessling W, Regev A, Carr SA, Clish CB, Mootha VK. Mitochondrial dysfunction remodels one-carbon metabolism in human cells. eLife. 2016;5:e10575. doi: 10.7554/eLife.10575. - DOI - PMC - PubMed
    1. Bray NL, Pimentel H, Melsted P, Pachter L. Near-optimal probabilistic RNA-seq quantification. Nature Biotechnology. 2016;34:525–527. doi: 10.1038/nbt.3519. - DOI - PubMed
    1. Breit SN, Brown DA, Tsai VW-W. The GDF15-GFRAL Pathway in Health and Metabolic Disease: Friend or Foe? Annual Review of Physiology. 2021;83:127–151. doi: 10.1146/annurev-physiol-022020-045449. - DOI - PubMed

Publication types

MeSH terms

Associated data