Depleting inositol pyrophosphate 5-InsP7 protected the heart against ischaemia-reperfusion injury by elevating plasma adiponectin

Abstract

Aims: Adiponectin is an adipocyte-derived circulating protein that exerts cardiovascular and metabolic protection. Due to the futile degradation of endogenous adiponectin and the challenges of exogenous administration, regulatory mechanisms of adiponectin biosynthesis are of significant pharmacological interest.

Methods and results: Here, we report that 5-diphosphoinositol 1,2,3,4,6-pentakisphosphate (5-InsP7) generated by inositol hexakisphosphate kinase 1 (IP6K1) governed circulating adiponectin levels via thiol-mediated protein quality control in the secretory pathway. IP6K1 bound to adiponectin and DsbA-L and generated 5-InsP7 to stabilize adiponectin/ERp44 and DsbA-L/Ero1-Lα interactions, driving adiponectin intracellular degradation. Depleting 5-InsP7 by either IP6K1 deletion or pharmacological inhibition blocked intracellular adiponectin degradation. Whole-body and adipocyte-specific deletion of IP6K1 boosted plasma adiponectin levels, especially its high molecular weight forms, and activated AMPK-mediated protection against myocardial ischaemia-reperfusion injury. Pharmacological inhibition of 5-InsP7 biosynthesis in wild-type but not adiponectin knockout mice attenuated myocardial ischaemia-reperfusion injury.

Conclusion: Our findings revealed that 5-InsP7 is a physiological regulator of adiponectin biosynthesis that is amenable to pharmacological intervention for cardioprotection.

Keywords: AMPK; DsbA-L; ERp44; Ero1-Lα; IP6K.

Graphical Abstract

Figure 1

Knocking out IP6K1 attenuated myocardial ischaemia–reperfusion injury. (A) Evans blue and TTC double staining of the hearts at 24 h after myocardial ischaemia–reperfusion injury. The non-ischaemic area was indicated by blue, the area at risk (AAR) by red and white, and the infarct area (IA) by white. LV, left ventricle. Data were presented as mean ± SEM, ns = not significant, Student’s t-test, n = 5 mice per group. Scale bar: 5 mm. (B) TUNEL staining of the ischaemia–reperfusion injured hearts in the border zones. Cardiac troponin I was co-stained for cardiomyocytes. Data were presented as mean ± SEM, Student’s t-test, n = 5 mice per group. Scale bar: 20 μm. (C) Plasma levels of hs-cTn-I at 24 h after myocardial ischaemia–reperfusion injury. Data were presented as mean ± SEM, Student’s t-test, n = 5 mice per group. (D) Protein levels of cleaved caspase 3 in the border zones of ischaemia–reperfusion injured hearts. Data were presented as mean ± SEM, Student’s t-test, n = 5 mice per group. (E) Representative echocardiographic images of WT and IP6K1 KO mice at 24 h after the myocardial ischaemia–reperfusion injury. Data were presented as mean ± SEM, Student’s t-test, n = 5 mice per group. (F) Phosphorylation levels of AMPK in the border zones of the ischaemic hearts. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. (G) Phosphorylation levels of AMPK in WT and IP6K1 KO hearts at baseline. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. (H) Phosphorylation levels of AMPK in cultured WT and IP6K1 KO MEF cells. Data were presented as mean ± SEM, ns = not significant, Student’s t-test, n = 6 from three independent repeats. (I) Phosphorylation levels of AMPK in WT MEFs that were treated with plasma from WT or IP6K1 KO mice. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. (J) Plasma adiponectin concentrations that were determined by ELISA. Data were presented as mean ± SEM, Student’s t-test, n = 8 mice per group. (K) Total circulating adiponectin levels in WT and IP6K1 KO mice, n = 3 mice per group. (L) Adiponectin oligomers in the plasma of WT and IP6K1 KO mice. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. (M) Adiponectin oligomers in the plasma of WT and IP6K1 KO mice at 24 h after the myocardial ischaemia–reperfusion injury. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group.

Figure 2

Adipocyte-specific deletion of IP6K1 attenuated myocardial ischaemia–reperfusion injury. (A) Total circulating adiponectin levels in control (IP6K1^f/f) and adipocyte-specific IP6K1 KO mice (adipoq-cre; IP6K1^f/f). n = 3 mice per group. (B) Levels of plasma adiponectin oligomers in control and adipocyte-specific IP6K1 KO mice. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. (C) Phosphorylation levels of AMPK in heart tissues of control and adipocyte-specific IP6K1 KO mice. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. (D) Plasma levels of hs-cTn-I at 24 h after myocardial ischaemia–reperfusion injury in control and adipocyte-specific IP6K1 KO mice. Data were presented as mean ± SEM, Student’s t-test, n = 5 mice per group. (E) Evans blue and TTC double staining of the hearts in control and adipocyte-specific IP6K1 KO mice at 24 h after myocardial ischaemia–reperfusion injury. The non-ischaemic area was indicated by blue, the area at risk (AAR) by red and white, and the infarct area (IA) by white. LV, left ventricle. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. Scale bar: 5 mm. (F) Representative echocardiographic images of control and adipocyte-specific IP6K1 KO mice at 24 h after myocardial ischaemia–reperfusion injury. Data were presented as mean ± SEM, Student’s t-test, n = 5 mice per group. (G) Phosphorylation levels of AMPK in the border zones of the ischaemic hearts of control and adipocyte-specific IP6K1 KO mice at 24 h after myocardial ischaemia–reperfusion injury. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. (H) Protein levels of cleaved caspase 3 in the border zones of the ischaemic hearts of control and adipocyte-specific IP6K1 KO mice at 24 h after myocardial ischaemia–reperfusion injury. Data were presented as mean ± SEM, Student’s t-test, n = 5 mice per group.

Figure 3

IP6K1 bound to adiponectin. (A) Immunoprecipitation of IP6K1 in mice tissues. Silver stain and mass spectrometry revealed that adiponectin (arrow) was co-pulled down by IP6K1 in the heart, adipose tissue, and skeletal muscle, but not in the brain, liver, or kidney. (B) Western blots showed that adiponectin was pulled down by IP6K1 in adipose tissue, heart, and skeletal muscle. (C) Immunoprecipitation of IP6K1 pulled down adiponectin (arrow) in WT but not IP6K1 KO adipose tissues. (D) Pulling down IP6K1 co-immunoprecipitated adiponectin (arrow) in WT but not IP6K1 KO hearts. (E) Myc-tagged IP6K1 and flag-tagged adiponectin were overexpressed together. Pulling down myc-tag IP6K1 co-immunoprecipitated flag-adiponectin. Myc-tagged GFP was used as a negative control. (F) Flag-tagged adiponectin and myc-tagged IP6K1 were overexpressed together. Pulling down flag-tagged adiponectin co-immunoprecipitated myc-IP6K1 (arrow). Flag-tagged GFP was used as a negative control. (G) Adiponectin (arrow) was co-pulled down by IP6K1 but not IP6K2 in WT adipose tissues. (H) Adiponectin (arrow) was co-immunoprecipitated by IP6K1 but not IP6K2 or IP6K3 in WT hearts. (I) Immunoprecipitations of IP6K1 co-pulled down similar amounts of adiponectin in the hearts of control (IP6K1^f/f) and adipocyte-specific IP6K1 KO (adipoq-cre; IP6K1^f/f) mice. Data were presented as mean ± SEM, Student’s t-test, n = 3. (J) Immunoprecipitation of IP6K1 co-pulled down adiponectin (arrow) in adipose tissues of control (IP6K1^f/f) but not adipocyte-specific IP6K1 KO mice (adipoq-cre; IP6K1^f/f).

Figure 4

5-InsP₇ enhanced the binding of adiponectin with ERp44. (A) Immunoprecipitation of adiponectin pulled down less ERp44 in IP6K1 KO adipose tissues. Data were presented as mean ± SEM, Student’s t-test, n = 3 independent repeats. (B) Pulling down ERp44 co-immunoprecipitated less adiponectin in IP6K1 KO adipose tissues. Data were presented as mean ± SEM, Student’s t-test, n = 3 independent repeats. (C) Immunostaining of adiponectin and ERp44 in WT and IP6K1 KO adipocytes. The co-localization of ERp44 and adiponectin was decreased in IP6K1 KO adipocytes. Data were presented as means ± SEM from 30 images of three independent experiments. Student’s t-test. Scale bar: 10 μm. (D, E) 3T3-L1 adipocytes were treated with TNP (3 μM for 24 h) to block 5-InsP₇ synthesis. Immunoprecipitation of adiponectin or ERp44 pulled down less of each other in the TNP-treated cells than in the DMSO-treated control cells. Data were presented as mean ± SEM, Student’s t-test, n = 3 independent repeats. (F) 5PCP-resin pulled down endogenous adiponectin but not ERp44 in whole cell lysates of adipocytes. (G) Compared with InsP₆, 5-InsP₇ enhanced the interaction of adiponectin with ERp44 in an in vitro binding assay. Data were presented as mean ± SEM, Student’s t-test, n = 5 independent repeats. (H) Both 5PCP and CF2 promoted the binding of adiponectin with ERp44 in an in vitro binding assay. Data were presented as mean ± SEM, one-way ANOVA, n = 5 independent repeats. (I) 5-InsP₇ but not InsP₃, InsP₄, or InsP₅ enhanced the binding of adiponectin with ERp44. Data were presented as mean ± SEM, one-way ANOVA, n = 3 independent repeats. (J) 1-InsP₇ and 3-InsP₇ did not display similar effects to 5-InsP₇ in mediating the binding of adiponectin with ERp44. Data were presented as mean ± SEM, one-way ANOVA, n = 3 independent repeats.

Figure 5

5-InsP₇ enhanced the binding of DsbA-L with Ero1-Lα. (A, B) Myc-IP6K1 and flag-DsbA-L were overexpressed together. (A) Immunoprecipitation of flag-tag DsbA-L co-pulled down myc-IP6K1 (arrow). Flag-GFP was used as a negative control. (B) Immunoprecipitation of myc-tag IP6K1 co-pulled down flag-DsbA-L. Myc-GFP was used as a negative control. (C) Immunoprecipitation of DsbA-L pulled down less Ero1-Lα in IP6K1 KO adipocytes than in WTs. Data were presented as mean ± SEM, Student’s t-test, n = 3 independent repeats. (D) Pulling down Ero1-Lα co-immunoprecipitated less DsbA-L in IP6K1 KO adipocytes than in WTs. Data were presented as mean ± SEM, Student’s t-test, n = 3 independent repeats. (E) 5PCP-resin pulled down endogenous DsbA-L but not Ero1-Lα in the whole cell lysates of adipocytes. (F) Compared with InsP₆, 5-InsP₇ enhanced the interaction of Ero1-Lα with DsbA-L in an in vitro binding assay. Data were presented as mean ± SEM, Student’s t-test, n = 5 independent repeats. (G) Both 5PCP and CF2 promoted the binding of Ero1-Lα with DsbA-L in an in vitro binding assay. Data were presented as mean ± SEM, one-way ANOVA, n = 5 independent repeats. (H) 5-InsP₇ and InsP₄ but not InsP₃ or InsP₅ enhanced the binding of Ero1-Lα with DsbA-L. Data were presented as mean ± SEM, one-way ANOVA, n = 3 independent repeats. (I) 1-InsP₇ and 3-InsP₇ did not display similar effects to 5-InsP₇ in mediating the binding of Ero1-Lα with DsbA-L. Data were presented as mean ± SEM, one-way ANOVA, n = 3 independent repeats.

Figure 6

5-InsP₇ elicited adiponectin intracellular degradation. (A, B) 3T3-L1 adipocytes were treated with cycloheximide (CHX, 100 μM) to block protein synthesis. (A) Adiponectin protein levels were determined by western blots. Data were presented as mean ± SEM, one-way ANOVA, n = 3 independent repeats. (B) Cells were co-treated with TNP (3 μM) to block 5-InsP₇ synthesis. Adiponectin protein levels were determined by western blots. Data were presented as mean ± SEM, one-way ANOVA, n = 3 independent repeats. (C) Adiponectin was immunoprecipitated in WT and IP6K1 KO adipose tissues and blotted for ubiquitin. (D) 3T3-L1 adipocytes were treated with TNP (3 μM for 24 h) to block 5-InsP₇ synthesis followed by MG132 treatment (10 μM for 4 h) to block proteasomal degradation. Adiponectin was immunoprecipitated and blotted for ubiquitin. (E) 3T3-L1 adipocytes were treated with TNP (3 μM for 24 h) to block 5-InsP₇ synthesis followed by chloroquine treatment (50 μM for 24 h) to block autophagy. Adiponectin was immunoprecipitated and blotted for ubiquitin. (F) 3T3-L1 adipocytes were treated with TNP for 24 h. Adiponectin (arrow) protein levels were determined by western blots. Data were presented as mean ± SEM, Student’s t-test, n = 3 independent repeats. (G) 3T3-L1 adipocytes were treated with SC-919 for 24 h. Adiponectin (arrow) protein levels were determined by western blots. Data were presented as mean ± SEM, Student’s t-test, n = 3 independent repeats. (H) Adiponectin protein levels in WT and IP6K1 KO adipose tissues. Arrow points to IP6K1. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. (I) Adiponectin protein levels in the adipose tissues of control (IP6K1^f/f) and adipocyte-specific IP6K1 KO mice (adipoq-cre; IP6K1^f/f). Arrow points to IP6K1. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group.

Figure 7

Pharmacological inhibition of 5-InsP₇ biosynthesis attenuated ischaemia–reperfusion injury in WT but not adiponectin KO mice. (A) WT animals were injected with TNP for 1 week. The plasma adiponectin oligomers (arrows) were determined by western blots. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. (B) Evans blue and TTC double staining of the hearts after ischaemia–reperfusion injury. The non-ischaemic area is indicated by blue, the area at risk (AAR) by red and white, and the infarct area (IA) by white. LV, left ventricle. Data were presented as mean ± SEM, Student’s t-test, n = 6 mice per group. Scale bar: 5 mm. (C) Adiponectin protein levels in adipose tissues of WT and adiponectin KO mice, n = 4 mice per group. (D) Plasma adiponectin oligomers (arrows) in WT and adiponectin KO mice, n = 3 mice per group. (E) Adiponectin KO mice were injected with TNP for 1 week and were subjected to ischaemia–reperfusion injury. The ischaemia–reperfusion injured hearts were stained with Evans blue and TTC. The non-ischaemic area is indicated by blue, the AAR by red and white, the IA by white. LV, left ventricle. Data were presented as mean ± SEM, ns = not significant, Student’s t-test, n = 5 mice per group. Scale bar: 5 mm.

Figure 8

Model for 5-InsP₇ regulation of the intracellular fate of adiponectin. Upper panel: IP6K1 physiologically associates with the adiponectin/DsbA-L complex and generates a local pool of 5-InsP₇, which binds adiponectin and DsbA-L to enhance the interaction of the adiponectin/DsbA-L complex with the ERp44/Ero1-Lα complex. As a result, adiponectin is retained and degraded intracellularly. Lower panel: Depleting 5-InsP₇ disrupts the binding between the adiponectin/DsbA-L complex with the ERp44/Ero1-Lα complex, which releases adiponectin into the secretory pathway. Depleting 5-InsP₇ could protect the heart from myocardial ischaemia–reperfusion injury by boosting plasma adiponectin.