Prolonged glucagon exposure rewires lipid oxidation and drives diabetic kidney disease progression

Abstract

Diabetic kidney disease (DKD) is the leading cause of end-stage kidney disease. Tubular abnormalities may precede glomerular pathology and indicate functional progression of DKD. Here, we find glucagon injection exacerbates lipid accumulation and renal injury, in addition to causing morphological changes in proximal tubules, podocytes, and mitochondria in the early phase of DKD in mice. However, the specific knockdown or knockout of Gcgr in renal tubular epithelial cells almost completely halts DKD development. In contrast to the effect of short-term glucagon stimulation, long-term glucagon exposure leads to the reversal of glucagon action (glucagon reversal) in proximal tubular epithelial cells (PTECs), which is characterized by reduced energy production and an increase in lipogenesis through Gcgr-PKA-Creb-mTORC1 pathway. Accordingly, anti-GCGR antibody treatment strongly blocks the pathogenesis of DKD induced by both type 2 and type 1 diabetes. Thus, our results highlight a previously unrecognized role of glucagon/Gcgr signaling in PTEC lipogenesis and DKD.

Introduction

Diabetic kidney disease (DKD) is a serious complication of diabetes and has long been recognized as the main primary cause of chronic kidney disease (CKD) and end-stage kidney disease (ESKD), contributing to kidney failure worldwide¹. Approximately 40% of patients with type 2 diabetes and 30% of individuals with type 1 diabetes develop DKD². It is estimated that by 2030, 14.5 million people will have ESKD³. Symptoms of DKD include an increased amount of urinary protein and a reduced estimated glomerular filtration rate (eGFR). DKD can be divided into stages 1–5 based on the progression from microalbuminuria to macroalbuminuria in the urine and a gradual decrease in the eGFR⁴.

Although proteinuria in diabetic patients has generally been attributed to abnormalities in the glomerulus, tubulointerstitial abnormalities, which may precede pathological glomerular changes, are the best indicator of the functional progression of kidney disease^5,6,7,8. An increasing body of evidence has revealed that lipid metabolism disorders or lipid accumulation in the kidney may contribute to the development and progression of DKD^{9,10,11,12,13,14}. Normally, lipid accumulation does not occur in the kidney since fatty acid uptake, oxidation, and synthesis are tightly balanced. However, lipid droplet accumulation is dramatically increased in the kidneys of humans with DKD and CKD, mostly in proximal tubular epithelial cells (PTECs)¹⁰. Furthermore, in DKD patients at different stages, there is a positive correlation between kidney lipid droplets and the albumin excretion rate (AER)¹⁵; a highly significant correlation between the increased expression of lipogenic genes, including LDLR, CD36 and SCD1, and the progression of human DKD and a decrease in the eGFR; and a highly significant negative correlation between the expression of fatty acid oxidation genes, including PPARα and CPT1, and the eGFR¹³. Altered intracellular nutrient signals are involved in the pathogenesis of lipid accumulation in PTECs. Among these signals, the hyperactivation of mechanistic target of rapamycin complex 1 (mTORC1), which is activated in a hypernutrient state and controls lipid metabolism, has been reported to be upregulated in the kidney and strongly associated with PTECs lipid accumulation^16,17. Excessive lipid accumulation, in turn, exacerbates oxidative stress, inflammation, and endoplasmic reticulum (ER) stress, ultimately leading to kidney cell apoptosis and fibrosis by recruiting proinflammatory and profibrotic factors such as TGF-β1, TNFα, CXCL1, IL-6, and IL-1β^18,19.

Both type 1 and type 2 diabetes are characterized by hyperglucagonemia potentially due to α-cell insulin resistance and β-cell transdifferentiation/dedifferentiation into α-cells in islets^{20,21,22,23,24}. The serum glucagon concentration is 2.6-fold higher in diabetic patients than in healthy individuals^25,26. Glucagon and its receptor, i.e., glucagon receptor (Gcgr), play crucial roles in maintaining glucose homeostasis by promoting glycogenolysis and gluconeogenesis²⁷. Moreover, glucagon is a strong activator of fatty acid oxidation and energy expenditure^{27,28,29,30,31}. GCGR is most highly expressed in the liver, followed by the kidney, heart, adipose tissue, and pancreatic islets³². Intriguingly, a substantial body of evidence highlights a positive relationship between hyperglucagonemia and the pathology of DKD^20,28,33,34. However, whether pathological levels of glucagon have an effect on the pathogenesis of DKD and how glucagon loses the ability to oxidize fatty acids in the kidney during the progression of DKD are unclear.

In our study, we found a role of glucagon in lipogenesis and DKD development. An increase in glucagon directly promoted DKD development in mice. In contrast to the traditional action of glucagon in lipid oxidation, long-term glucagon exposure markedly enhanced lipid accumulation through mTORC1 in PTECs. Importantly, Gcgr could be a new promising target for the treatment of DKD.

Results

Glucagon/Gcgr signaling promoted DKD progression

Glucagon in the blood remains elevated in both type 1 and type 2 diabetes patients. We measured the serum glucagon levels at different time points from day to night in genetically diabetic db/db (leptin receptor deficiency) mice and found that they were sustained elevated (approximately 4-fold) compared to those in the control (db/m) mice (Fig. 1a), which was accompanied by obesity and hyperglycemia (Figure S1a–b). Similarly, in the Unix-HFD/STZ (uninephrectomy followed by high fat diet (HFD) feeding and STZ injection)-induced DKD mouse model, the serum glucagon concentration increased approximately 2-fold, along with an increase of the plasma glucose, body weight was not changed (Fig. 1b, c and S1c-d). Regarding Gcgr, the western blot and qRT-PCR results showed that both the protein and mRNA levels of Gcgr were highest in the liver, followed by the kidney, in C57BL6/J mice (Fig. 1d, e and S1e). Interestingly, there was an increase in Gcgr protein content in the kidneys of DKD mice (Fig. 1f, g). In cultured mammalian cell lines, we also found that both the protein and mRNA levels of Gcgr were greater in primary hepatocytes and kidney-derived cell lines, including NRK-52E (a rat-derived renal proximal tubular epithelial cell line), HK2 (a human-derived renal proximal tubular epithelial cell line), and primary PTECs than in other cell lines, such as A549 cells (an alveolar basal epithelial cell line), C2C12 cells (a myoblast cell line) and 3T3-L1 (a adipocyte-like cell line) (Figure S1f–h). These results indicate that both the serum glucagon and renal Gcgr protein levels are greatly increased in DKD mice.

Fig. 1: Glucagon and Gcgr were elevated in DKD mice.

a Serum glucagon levels in 15-week-old male db/db mice at different time points. b Schematic of UniX-HFD/STZ-induced DKD mouse model: 6-week-old male C57BL6/J mice received HFD for 4 weeks, followed by uninephrectomy, then STZ injection to induce diabetes after 2-week recovery. c Serum glucagon levels in control and DKD mice (16 weeks post-STZ injection). d, e Gcgr protein levels in liver, muscle, eWAT, and kidney of 8-week-old male mice on normal chow. f, g Western blot analysis of renal Gcgr in DKD mice (8 weeks post-STZ). h Co-immunofluorescence of Gcgr with renal cell markers (AQP1 for proximal tubule, LTL for proximal tubules, calbindin-D28K for distal tubules, α-SMA for mesangial cell and podocin for podocytes) in the normal chow-fed C57BL6/J mice (male, 8-week-old). Scale bar: 50 μm. i Schematic of glucagon injection protocol: 300 nmol/kg glucagon i.p. twice daily for 2 weeks starting 1 week after STZ. j–n Serum glucagon, UACR, urine Kim-1 (normalized to creatinine), 24-h urine volume, and kidney weight/body weight ratio. Samples were collected after 2 weeks of glucagon injections. The error bars represent the SEMs; n = 3–6 in each group (a, c–g) and 5-12 in each group (j–n). UniX, uninephrectomized; HFD, high-fat diet; STZ, streptozotocin; UACR, urinary albumin-to-creatinine ratio; Kim-1, kidney injury molecule-1.

Kidney is divided into two sections. One is the granular outer section (renal cortex), containing the glomeruli and convoluted tubules, and this section is responsible for most of the ultrafiltration and reabsorption function of the kidney. The other is the inner section (renal medulla), containing the loops of Henle and the collecting tubules. Renal tubules account for more than 90% of the renal mass. Among these, the proximal tubules account for more than 70% of renal tubule length and 65% of the kidney volume, making up most of the mass of the renal cortex ( ~ 80%)^35,36,37,38. To further determine the cellular localization of Gcgr in the kidney, we performed immunofluorescence staining analysis on kidney cortex sections from C57BL6/J mice. There was almost no Gcgr staining in the kidney cortex from renal tubular epithelial cells specific Gcgr-deficient mice (Gcgr^fl/flKsp^Cre) compared to Gcgr^fl/fl mice, suggesting the specificity of the anti-GCGR antibody (Figure S1i). Then we found that Gcgr colocalized with Aquaporin 1 (Aqp1, proximal tubule marker) and lotus tetragonolobus lectin (LTL, another proximal tubule probe indicating proximal tubule brush border), as well as weak colocalization with Calbindin-D28k (distal tubule marker); while with no colocalization with α-smooth muscle actin (α-SMA, mesangial cell marker) or Podocin (podocyte marker) (Fig. 1h). Taken together, these results suggest that Gcgr is expressed in PTECs in the kidney cortex, which is consistent with the qRT‒PCR results shown in Figure S1f, h.

To investigate the direct effect of glucagon on DKD, we intraperitoneally injected glucagon (300 nmol/kg, twice a day for 2 weeks) into mice in the early phase of DKD (Unix-HFD/STZ-induced DKD mice, one week after STZ injection) (Fig. 1i). Glucagon injection further increased the serum glucagon concentration to levels similar to those found in db/db mice (Fig. 1a), with no effect on plasma glucose or body weight compared to those in the saline-injected DKD group (Fig. 1j, S1j, k). After 2 weeks of injections, the saline-injected DKD mice demonstrated the initiation of kidney injury, as shown by the trend of increase in the urinary albumin-to-creatinine ratio (UACR, a gold-standard diagnostic marker of DKD), significant increase of urine kidney injury molecule-1 (Kim-1, a transmembrane glycoprotein that is recognized as a biomarker for renal proximal tubule injury) protein levels, 24-h urine volume, and kidney weight-to-body weight ratio compared to those of the control group. Importantly, compared to saline injections, 2 weeks of glucagon injections further exacerbated renal dysfunction in the early phase of DKD, as observed by assessing UACR (1.6-fold increase), urinary Kim-1 (1.5-fold increase), and renal weight (1.25-fold increase), suggesting that supplementation with glucagon during the early phase of DKD exacerbates renal dysfunction in DKD mice (Fig. 1k–n). Thus, we found that the sustained upregulation of glucagon signaling directly promoted DKD development.

Gcgr knockdown in renal tubular epithelial cells almost completely halted DKD development

To further investigate the role of glucagon in DKD progression, we knocked down Gcgr in Unix-HFD/STZ-induced DKD mice. Given that Gcgr is expressed in renal tubules of cortex, as shown in Fig. 1h, we generated a Ksp-cadherin (expressed exclusively in tubular epithelial cells in the kidney) promoter-driven Gcgr-specific short hairpin RNA (Gcgr shRNA)³⁹ construct carried by adeno-associated virus-9 (AAV9). This virus or a control virus was injected in situ directly into the kidneys of DKD mice (Fig. 2a). Gcgr shRNA reduced proximal tubule Gcgr mRNA levels by ~80% in PTECs and ~50% in the kidney, with no effect on hepatic Gcgr (Fig. 2b–d and S2a). Next, we evaluated Gcgr downstream signaling by assessing the phosphorylation of cAMP response element-binding protein (Creb) and the expression of phosphoenolpyruvate carboxykinase (Pepck), both of which are glucagon signaling markers, in the kidney. Consistent with the elevated Gcgr expression in the kidneys of DKD mice, compared to normal mice, mice in the Ctrl shRNA group displayed enhanced pCreb activation and Pepck expression in the kidney. In contrast, mice in the Gcgr shRNA group showed a marked decrease in pCreb activation and Pepck expression (Fig. 2c, e and S2b). These findings suggest that Gcgr shRNA dramatically attenuated kidney Gcgr signaling, which is activated in DKD.

Fig. 2: Gcgr-specific knockdown in kidney tubules halted DKD progression.

a Schematic of UniX-HFD/STZ-DKD mouse model with in situ AAV-shRNA injection: modeling as in Fig. 1b, AAV-shRNA injected into kidneys 4 weeks post-STZ injection, sacrificed 4 weeks later. b Gcgr mRNA levels in primary PTECs from control and Gcgr shRNA-injected mice. c–e Renal Gcgr, p-Creb, total Creb protein levels. f–h UACR, urine Kim-1/creatinine, NAG/creatinine in the indicated groups of mice. i 24 h urine volume in the indicated groups. j, k Kidney weight-to-body weight ratio and kidney morphology in the indicated groups. l H&E, Oil Red O and PAS staining of kidneys from the mice with or without Gcgr knockdown. Scale bar: 50 μm. m Relative mRNA levels of inflammatory genes in the kidneys in the indicated groups. n–o Renal Collagen I protein in the indicated groups. p Masson and Sirius red staining. Scale bar: 50 μm. q TEM analyses of kidney proximal tubules. Scale bar: white: 5 μm; yellow: 2 μm; red: 1 μm. Yellow arrow: lipid droplet. Quantitative analysis of the length (r) and number per cell (s) of mitochondria. t Relative mRNA levels of Dpr1 and Mfn1 in the kidneys of mice. u TEM analyses of kidney glomeruli. Scale bar: 5 μm. White arrows: GBM; red asterisk: podocyte fusion. Quantitative analysis of the GBM width (v) and foot process width (w). Mice were male and sampled after 4 weeks of shRNA injections (b–v). The error bars represent the SEMs; n = 3 in each group (c–e, n–o), 4-6 in the control group and 12-20 in the shRNA group (b, f–j, m, r–t, v–w). NAG, N-acetyl-beta-D-glucosaminidase; GBM, glomerular basement membrane. TEM, transmission electron microscopy. PTECs, proximal tubular epithelial cells.

Before control or Gcgr shRNA injection, the DKD mice in these groups presented comparable UACR levels (Figure S2c). Four weeks after injection, mice in the Gcgr shRNA group exhibited lower blood glucose levels with no change in body weight (Figure S2d, e). Notably, compared to Ctrl shRNA, shRNA-mediated Gcgr deficiency resulted in the near normalization of urinary albumin excretion, as determined by the UACR, and proximal tubule injury, as determined by Kim-1 and n-acetyl-beta-D-glucosaminidase (NAG, a lysosomal enzyme that is considered another marker of early proximal tubular damage), in addition to decreased urine volume and kidney weight (Fig. 2f–k and S2f). Taken together, these findings show that Gcgr knockdown in renal tubular epithelial cells via shRNA almost completely stopped DKD development.

We conducted histological examinations to evaluate kidney tubulointerstitial and glomerular lesions. Usually, DKD is accompanied by lipid metabolism abnormalities and hyperglycemia, and H&E and Oil Red O staining of the Ctrl shRNA group showed obvious lipid droplet and vacuole accumulation in the proximal tubules (Fig. 2l and S2g). Moreover, large deposits of glycogen and other polysaccharides (including glycoproteins, glycolipids, and mucin) were detected in tubules and glomeruli by periodic acid–Schiff (PAS) staining. However, Gcgr knockdown in renal tubular epithelial cells distinctly reversed these changes (Fig. 2l and S2h), indicating that Gcgr knockdown strongly blocked lipid and glycogen deposition in PTECs under DKD conditions.

Lipid metabolism disorders play an essential role in modulating inflammation and immune function, further leading to extracellular matrix organization^40,41. Accordingly, we found that Gcgr knockdown in renal tubular epithelial cells ameliorated the expression of inflammatory and fibrotic genes, including Tnfα, Il6, Il1β, and Tgfβ1 (Fig. 2m). Moreover, western blot analysis revealed significantly decreased collagen deposition in the kidneys of the Gcgr shRNA group. Masson and Sirius Red staining also indicated much less fibrosis in mice in the Gcgr shRNA group than in mice in the Ctrl shRNA group (Fig. 2n–p, S2i, j). Thus, these results prove that Gcgr inhibition in renal tubular epithelial cells normalized inflammation and fibrosis in the kidney of DKD model mice.

Under DKD conditions, the glomerular basement membrane (GBM) thickens, and the podocyte foot process effaces, in addition to mitochondrial dysfunction^42,43. Correspondingly, transmission electron microscopy (TEM) revealed significant mitochondrial loss and fission in the Ctrl shRNA-DKD group compared to the control group, while Gcgr knockdown in renal tubular epithelial cells markedly normalized these effects (Fig. 2q–s). Additionally, Gcgr shRNA effectively restored Drp1 (mitochondrial fission) and Mfn1 (mitochondrial fusion) expression levels (Fig. 2t). Consistent with the Oil Red O staining results shown in Fig. 2l, the scanning results revealed obvious lipid droplets in the PTECs of DKD mice, and Gcgr knockdown almost completely eliminated lipid accumulation (Fig. 2q). Moreover, compared with those in the Ctrl shRNA-DKD group, the GBM thickness and podocyte foot depletion in the Gcgr shRNA group were normalized (Fig. 2u–w). Thus, the above results indicate that the inhibition of glucagon/Gcgr signaling in renal tubular epithelial cells notably improved the tubular and glomerular microstructure in DKD.

Genetic knockout of Gcgr in renal tubular epithelial cells rescued the DKD phenotype in mice

To further confirm the role of Gcgr in DKD development, we generated renal tubular epithelial cells specific Gcgr-deficient mice (Gcgr^fl/flKsp^Cre) with a knockout rate of nearly 90% in PTECs and kidneys at the mRNA level and approximately 50% at the protein level in DKD model mice (Fig. 3a–d), without changes in Gcgr expression in the liver, muscle or eWAT (Figure S3a). On a chow diet, knocking out Gcgr in renal tubular epithelial cells led to no changes in GFR, UACR, BUN, serum creatinine, 24 - hour urine volume, or 24 - hour drinking water volume (Figure S3b–g). Additionally, kidney section histology staining revealed no differences in morphology, glycogen, or fibrosis between groups (Figure S3h). These results suggest the safety of blocking Gcgr in PTECs. We then induced the DKD model in both Gcgr^fl/fl and Gcgr^fl/flKsp^Cre mice following the procedure in Fig. 2a, using normal chow-fed wild-type mice as controls. Consistent with previous results (Fig. 2), Gcgr^fl/fl mice exhibited glucagon signal activation, as confirmed by pCreb protein level and Pepck expression, while Gcgr knockout attenuated this effect (Fig. 3c, e, S3i). Similarly, compared with Gcgr^fl/fl mice, Gcgr knockout in renal tubular epithelial cells improved plasma glucose without influencing body weight (Figure S3j, k).

Fig. 3: Renal tubule-specific Gcgr knockout ameliorated renal injury in DKD mice.

a Schematic of CRISPR/Cas9 constructs targeting Gcgr locus before/after Cre recombination. b Gcgr mRNA levels in primary PTECs and kidneys of 6-wk-old male Gcgr^fl/fl and Gcgr^fl/flKsp^Cre mice on Normal Chow. c–e Renal Gcgr, p-Creb, total Creb protein via western blot. f–h UACR, urine Kim-1/creatinine, NAG/creatinine in different groups of mice. i 24 h urine volume in the indicated groups. j–k Kidney weight/body weight ratio and macro renal morphology in the indicated groups. l Kidney H&E, Oil Red O, PAS staining in the Gcgr^fl/fl and Gcgr^fl/flKsp^Cre groups. Scale bar: 100 μm. m–n Renal Collagen I protein levels in kidneys from different groups of mice. o Masson and Sirius red staining of renal lesions in the Gcgr^fl/fl and Gcgr^fl/flKsp^Cre groups. Scale bar: 100 μm. p TEM analyses of kidney proximal tubules. Scale bar: yellow: 5 μm; red: 2 μm; orange: 500 nm. Yellow arrow: lipid droplet. Quantitative analysis of the number of mitochondria per cell (q) and length of mitochondria (r). s TEM analyses of kidney glomeruli. Scale bar: 1 μm. White arrows: GBM; red asterisk: podocyte fusion. t–u Quantitative analysis of the GBM width and foot process width in the indicated groups. Mice were male and sampled after 8 weeks of STZ injections (c–p). The error bars represent the SEMs; n = 3 in each group (c–e, m–n), 5-9 in each group (b, f–j, q) and 12-15 in each group (r, t and u).

We next measured major renal function indicators in the different groups. Compared to the Gcgr^fl/fl group, the Gcgr^fl/flKsp^Cre group exhibited the following improvements: nearly normalization of UACR (Fig. 3f) and tubular injuries, as determined by urine Kim-1 and NAG levels (Fig. 3g, h, S3l); a decrease in 24 h urine volume (Fig. 3i); and kidney weight (Fig. 3j, k). These results further confirm the renoprotective effect of Gcgr deficiency in renal tubular epithelial cells.

Furthermore, H&E, Oil Red O, and PAS staining showed that lipotoxicity and glucotoxicity were dramatically improved in the Gcgr^fl/flKsp^Cre group than in the Gcgr^fl/fl group (Fig. 3l, S3m, n). Subsequently, Gcgr deficiency in renal tubular epithelial cells significantly decreased fibrosis accumulation, as detected by Masson’s trichrome and Sirius Red staining and Collagen I expression (Fig. 3m–o, S3o, p). In addition, Gcgr knockout also ameliorated PTEC mitochondrial dysfunction, GBM thickness, foot process width and podocyte injury, as observed by TEM, indicating functional recovery of the kidney due to Gcgr knockout in renal tubular epithelial cells (Fig. 3p–u).

To further validate that glucagon specifically activates Gcgr in PTECs, causing renal injury, we divided Gcgr^fl/fl and Gcgr^fl/flKsp^Cre mice into three groups after DKD modeling (Figure S3q), treating them with either glucagon or saline. Results showed glucagon notably increased UACR levels only in Gcgr^fl/fl mice, not in Gcgr^fl/flKsp^Cre mice (Figure S3r), indicating a direct glucagon-Gcgr effect on renal function.

Taken together, the above results indicate that glucagon/Gcgr signaling in renal tubular epithelial cells plays a critical role in DKD development.

Gcgr deficiency in renal tubular epithelial cells suppressed renal lipogenesis and promoted lipid oxidation in DKD mice

As proximal tubules are responsible for reabsorbing about 60-70% of water, sodium and chloride, nearly complete of glucose, amino acids and bicarbonate, suggesting a substantial role in kidney filtration⁴⁴. To maintain normal kidney filtration function, proximal tubules require a substantial amount of energy to reabsorb filtered water and solutes, and this process occurs primarily through fatty acid oxidation^45,46. However, in diabetic patients, impaired fatty acid oxidation and lipid accumulation in the kidney have been observed, which results in mitochondrial dysfunction, kidney cortex fibrosis, and apoptosis^11,47. Additionally, under diabetes condition, the lipid accumulation in the cortex was much higher than in medulla^48,49,50. Similarly, we observed a 1.53- to 3.16-fold increase in total triglyceride (TG) and total cholesterol (TC) levels in both the kidney and serum of DKD mice compared to those in control mice. Moreover, 2 weeks of glucagon injection further enhanced TG and TC levels (1.36- to 2.85-fold increase) in the kidney and serum in the early phase of DKD in mice compared to those in saline-injected DKD mice (Fig. 4a–d). Kidney oil red O staining yielded consistent results (Figure S4a). In contrast, renal tubular epithelial cells specific Gcgr knockdown mice (Gcgr shRNA) exhibited extraordinary prevention of lipid accumulation in the kidney and serum compared to that of Ctrl shRNA mice, with kidney TG and TC levels almost normalized to those of the control group (Fig. 4e–i). Similar results were obtained in the renal tubular epithelial cells specific Gcgr knockout group (Figure S4b–f). Furthermore, we used air flow-assisted desorption electrospray ionization mass spectrometry (AFADESI-MS) to visualize the in situ spatial distribution of lipids. Notably, the levels of fatty acids associated with lipid droplet formation, including myristic acid (C14:0), tetradecenoic acid (C14:1), palmitoleic acid (C16:1), oleic acid (C18:1), stearic acid (18:0) and linolelaidic acid (18:2), were dramatically lower in Gcgr shRNA mice than in Ctrl shRNA mice (Fig. 4j). Moreover, the levels of other lipid classes, such as diacylglycerol (DG), phosphatidylcholine (PC), phosphatidylethanolamine (PE), and monoglyceride (MG), also greatly decreased after Gcgr was knocked down (Figure S5). In summary, Gcgr deficiency apparently prevents in situ lipid accumulation in the kidneys of DKD mice.

Fig. 4: Glucagon led to lipid accumulation in the kidney under DKD conditions.

a–d Renal triglyceride (TG, a)/total cholesterol (TC, b) and serum TG (c)/TC (d) levels in Glucagon-injected mouse groups. Renal TG (e)/TC (f) and serum TG (g)/TC (h)/low-density lipoprotein cholesterol (LDL-C, i) levels following renal Gcgr knockdown. j AFADESI-MSI-based spatially resolved lipidomic analysis of kidney tissues from the control, Ctrl shRNA and Gcgr shRNA groups. k The top 10 (ranked by padj) Gene Ontology terms upregulated in the kidneys of mice in the Gcgr shRNA group compared to the Ctrl shRNA group. l Gene set enrichment shows lipid synthesis pathway enrichment and fatty acid oxidation deficiency in Ctrl shRNA vs. Gcgr shRNA kidneys. m-o Renal mRNA expression of lipid biosynthesis/esterification/transport genes (m–n) and fatty acid oxidation genes (o) in different groups. p Immunofluorescence co-localization of LTL (green) with Srebp1/Scd1/Srebp2 (red) in kidney sections. Scale bar: 50 μm. Mice were male and sampled after 2 weeks of glucagon injections (a–d), mice were male and sampled after 8 weeks of STZ injections (e–i, m–o), and mice were male and sampled after 4 weeks of shRNA injections (j–l, p). The error bars represent the SEMs; n = 5–12 in each group (a–i) and 4-9 in each group (m-o). TG, triglyceride; TC, total cholesterol; padj, adjusted p value.

Consistent with the lipid content results, gene ontology analysis of the kidney via RNA-seq revealed significant alterations in lipid metabolism in the Gcgr shRNA group compared to the Ctrl shRNA group (Fig. 4k). Additionally, gene set enrichment analysis (GSEA) confirmed the reduced enrichment of fatty acid synthesis and enhanced lipid oxidation in DKD mice after Gcgr knockdown (Fig. 4l and S4g-h). Furthermore, real-time PCR analysis verified that the expression of lipid biosynthesis-, esterification reaction-, and transportation-related genes, such as Fasn, Agpat2, and CD36, tended to increase in the Ctrl shRNA-treated DKD and Gcgr^fl/fl DKD mice. Conversely, the expression of lipid oxidation-associated genes, such as Cpt1a and Echs1, was attenuated in the Ctrl shRNA and Gcgr^fl/fl DKD mice compared to that in the control group. Gcgr knockdown and knockout significantly reversed these trends (Fig. 4m–o and S4i–k). Similarly, the staining results indicated that there was markedly less co-staining of the PTEC marker LTL with de novo lipid synthesis proteins, such as stearoyl-CoA desaturase 1 (Scd1), sterol regulatory element-binding protein 1 (Srebp1), and sterol regulatory element-binding protein 2 (Srebp2) (Fig. 4p and S4l). Taken together, these findings suggest that renal tubular epithelial cells specific Gcgr deficiency ameliorates DKD mice renal lipotoxicity by suppressing lipid synthesis and restoring lipid oxidation.

Glucagon/Gcgr promotes lipid accumulation in PTECs through mTORC1 signaling

To investigate whether glucagon/Gcgr has a direct effect on lipid accumulation in proximal tubular cells, we treated HK2 cells with 100 nM glucagon and 5 μM PA as a substrate for various durations. The results indicated that 12 h and 24 h of treatment with glucagon significantly promoted TG levels but that short-term glucagon treatment had no effect on TG levels (6 h) (Fig. 5a). We then repeated this experiment in primary PTECs treated with glucagon for 24 h and observed a consistent increase in the intracellular TG concentration (Fig. 5b). Given that high levels of glucagon may activate glucagon-like peptide-1 receptor (GLP-1R) besides Gcgr, we next isolated primary PTECs from Gcgr^fl/flKsp^Cre mice and found that glucagon did not affect lipid accumulation in primary PTECs from Gcgr^fl/flKsp^Cre mice, demonstrating that this effect occurred through Gcgr (Fig. 5b). Consistently, the lipid-inducing function of glucagon in proximal tubular cells was further confirmed by BODIPY lipid probe staining in HK2 cells (Fig. 5c, d). Additionally, we found that prolonged glucagon exposure also facilitated lipid transport within proximal tubules (Figure S6a–c), as shown by the enhanced uptake of BODIPY FL C₁₆ in HK2 cells. Real-time PCR revealed that 4 h of treatment with glucagon increased the expression of fatty acid β-oxidation genes, including EHHADH and CPT1, but did not change the expression of lipid synthesis genes, including SCD1 and ACLY, in primary PTECs (Figure S6d, e). In contrast, prolonged glucagon exposure (48 h) led to the upregulation of lipid synthesis-related gene expression and the downregulation of fatty acid β-oxidation gene expression (Fig. 5e, f). Therefore, long-term glucagon treatment directly promoted lipid synthesis and inhibited lipid oxidation in kidney proximal tubules, which is opposite to the acute effect of glucagon.

Fig. 5: Long-term glucagon treatment promoted lipid accumulation by activating mTORC1 signaling.

a Intracellular TG in HK2 cells treated with 100 nM glucagon for different durations. b Intracellular TG in primary PTECs from Gcgr^fl/fl and Gcgr^fl/flKsp^Cre mice treated with 100 nM glucagon for 24 h using 5 μM PA as a substrate. c–d BODIPY staining of HK2 cells under identical conditions. Scale bar: 20 μm. e–f qRT-PCR analyses of the lipogenesis and fatty acid oxidation genes in HK2 cells after 48 h glucagon treatment. g GSEA shows mTORC1 signaling enrichment in Ctrl shRNA vs. control kidneys. h–j Western blot of p-mTORC1, p-S6 in kidneys (8 wk post-STZ). k Intracellular cAMP in HK2 cells during glucagon treatment. l p-CREB western blot in HK2 cells. m ChIP-qPCR analysis using an anti-CREB antibody. n Luciferase assay of Rraga/S6k promoters in HepG2 cells. o Western blot of S6K/S6 in HK2 cells. p–s BODIPY staining and TG measurement in HK2 cells with PKA/mTORC1 inhibitors. Scale bar: 20 μm. t PTECs derived from Gcgr^fl/fl and Gcgr^fl/flKsp^Cre mice were treated with 10 μM MHY1485 (an mTORC1 activator) for 24 h, and the intracellular TG was subsequently measured. Mice were male and sampled after 4 weeks of shRNA injections (g), and mice were male and sampled after 8 weeks of STZ injections (h–j). The error bars represent the SEMs; n = 3-8 biologically independent cells (a–f, l–n, q–t); n = 3 in each group (h–k, o). PA, palmitate acid; BODIPY, boron-dipyrromethene.

To investigate the molecular mechanism through which glucagon affects lipid accumulation in PTECs, we analyzed RNA-seq data and found that mTORC1 signaling, which stimulates de novo lipogenesis by activating S6 kinase (S6K)-SREBP⁵¹, was activated in DKD mice (Fig. 5g), whereas it was significantly ameliorated in the Gcgr^fl/flKsp^Cre and Gcgr shRNA groups (Fig. 5h–j and S6f–h). These findings suggest a potential interplay between glucagon signaling and the mTORC1 pathway. Glucagon/Gcgr signaling occurs through the Gcgr/cAMP/PKA/Creb pathway⁵². Intracellular cAMP and pCREB activation results showed that both short-term (1 h) and long-term (48 h) glucagon treatment significantly activated downstream signaling in HK2 cells, demonstrating that prolonged exposure to glucagon did not impair signaling (Fig. 5k, l, S6i). However, only long-term glucagon treatment facilitated pS6 activation, indicating that long-term exposure to glucagon may promote mTORC1 signaling activation (Fig. 5o and S6j). According to previous bioinformatic analysis via a whole-genome search of the cAMP response element (CRE), genes related to the mTORC1 pathway might be targets regulated by Creb⁵³. To this end, we performed a ChIP‒qPCR assay to target genes related to mTORC1. The results demonstrated significant enrichment of key mTORC1 regulators, such as Ras-related GTP-binding A (RRAGA), Ras-related GTP-binding B (RRAGB) and the mTORC1 substrate S6K, in the ChIP samples using an anti-CREB antibody compared to the normal IgG control samples (Fig. 5m). To further assess the functional impact of glucagon treatment on potential CREB target genes, we cloned the promoters of RRAGA and S6K into the pGL3-basic vector. Luciferase assays revealed that glucagon treatment (24 h) significantly enhanced luciferase activity driven by the RRAGA and S6K promoters (Fig. 5n). In addition, the expression of genes related to mTORC1 signaling (RRAGA, RRAGB, RHEB, RPTOR, S6K, S6, and mTOR) in HK2 cells was upregulated after 48 h of glucagon treatment but remained the same as that in the short-term glucagon treatment group (Figure S6k, l). Similarly, the expression of proteins related to mTORC1 signaling, such as S6K and S6, significantly increased under prolonged glucagon treatment but remained unchanged under short-term glucagon exposure in HK2 cells (Fig. 5o, S6m, n). In summary, these findings strongly indicated that long-term glucagon treatment markedly promoted the expression of mTORC1-related proteins through glucagon-stimulated Creb.

To further assess the importance of the GCGR downstream kinases PKA and mTORC1 in lipid accumulation within proximal tubular cells, we pretreated HK2 cells with the PKA inhibitor H89 and the mTORC1 inhibitor rapamycin and then exposed them to glucagon for 24 h accompanied by PA as a substrate. BODIPY staining and intracellular TG levels in the inhibitor-treated groups both showed significant prevention of glucagon-induced lipid synthesis (Fig. 5p–s). Consistently, in primary PTECs isolated from Gcgr^fl/fl and Gcgr^fl/flKsp^Cre mice, treatment with the mTORC1 activator MHY1485 (24 h) resulted in significant lipid accumulation in both groups (Fig. 5t), suggesting that mTORC1 is downstream of glucagon signaling and mediates the effect of glucagon on lipid accumulation.

Taken together, these results highlight the substantial contribution of the Gcgr/PKA/Creb/mTORC1 signaling pathway in PTECs to kidney lipid accumulation in the context of DKD.

Prolonged glucagon exposure promotes mitochondrial dysfunction and ECM accumulation

It is commonly believed that glucagon treatment promotes lipid oxidation^29,31; however, our results indicated the opposite effect (increased lipid synthesis and decreased lipid oxidation) under prolonged exposure. To further investigate this observation, we measured the ATP/ADP ratio in HK2 cells at various time points (0, 30 min, 6 h, 24 h, and 48 h) after glucagon treatment. The results showed that the ATP/ADP ratio increased within the first 30 min of glucagon exposure but gradually declined over time, ultimately approaching one-fifth of the initial level after 72 h (Fig. 6a). This result is consistent with the consensus that upregulated lipid catabolism is induced by short-term glucagon treatment, as shown in Figure S6e, and that long-term glucagon treatment dramatically attenuates energy expenditure, as shown in Fig. 5f.

Fig. 6: Long-term glucagon treatment promoted mitochondrial dysfunction.

a ATP/ADP ratio in HK2 cells treated with 100 nM glucagon for varying durations. b Fatty acid oxidation of NRK-52E cells: NRK-52E cells were pretreated with 100 nM glucagon for 24 h or 1 h before the assay, and then, palmitate substrate was added to the wells just before measuring the OCR. Where indicated, etomoxir (4 μM, inhibitor of CPT1), oligomycin (2.5 μM, inhibitor of ATP synthase), FCCP (2 μM, uncoupler of oxidative phosphorylation) and antimycin/rotenone (0.5 μM, inhibitors of respiratory complex) were added. c Basal respiration, maximal respiration, ATP production and respiration capacity were calculated from the OCR in NRK-52E cells. d Images of live mitochondria stained with PKMO in NRK-52E cells. Scale bar: 2 μm. e Mitochondrial superoxide (MitoSOX, red) staining of HK2 cells treated with 100 nM glucagon and 50 μM PA for the indicated period. Scale bar: 100 μm. f–h Representative western blot and quantification of fibronectin and phosphorylated and total S6 in HK2 cells treated with high glucose (30 mM) or 100 nM glucagon with or without the PKA inhibitor H89 (10 μM). i–k Representative western blot and quantification of fibronectin and phosphorylated and total S6 in HK2 cells treated with high glucose (30 mM) or 100 nM glucagon with or without the mTOR inhibitor rapamycin (30 nM). The error bars represent the SEMs; n  = 3-8 biologically independent cells (a‒k). OCR, oxygen consumption rate; FN1, fibronectin.

To further investigate changes in the oxidation function of proximal tubules, we measured the oxygen consumption rate (OCR) in NRK-52E cells using PA-BSA as a substrate and observed an increase in the OCR after 1 h of glucagon treatment but a decrease following 24 h of treatment, suggesting that prolonged glucagon exposure suppresses the utilization of fatty acids in proximal tubular epithelial cells (Fig. 6b, c).

Lipid accumulation leads to mitochondrial dysfunction and fission⁵⁴. To assess changes in mitochondrial morphology in response to glucagon treatment, we stained mitochondria with PKMO, a MitoTracker with photostability and reduced phototoxicity⁵⁵. Notably, we observed that long-term glucagon treatment promoted mitochondrial fission but that short-term treatment did not induce this change (Fig. 6d). As shown in Fig. 2q and Fig. 3p, the proximal tubular epithelial cells of DKD mice exhibited mitochondrial fission and a reduced number of mitochondria.

Mitochondria are known to produce approximately 90% of cellular reactive oxygen species (ROS), and excessive lipid accumulation induces ROS overgeneration and leads to mitochondrial oxidative stress, which causes mitochondrial damage^56,57. Therefore, we measured mitochondrial ROS levels using the MitoSOX probe. Our results indicated that short-term glucagon treatment did not lead to the accumulation of mitochondrial ROS but that prolonged glucagon exposure facilitated ROS production (Fig. 6e). Taken together, these findings indicate that glucagon enhances mitochondrial oxidation in response to short-term stimulation but that this effect is reversed in proximal tubules by long-term treatment.

Mitochondrial dysfunction in tubular epithelial cells leads to extracellular matrix deposition^11,43,58. Lipid accumulation is also an initiator of proximal tubule epithelial-to-mesenchymal transition^59,60. We treated HK2 cells and PTECs with glucagon for 48 h, using high glucose-treated cells as a positive control. Western blot analysis indicated that long-term glucagon treatment markedly promoted fibronectin accumulation by activating mTORC1 signaling and that H89 and rapamycin treatment reversed these changes (Fig. 6f–k). Thus, these results reveal the positive relationship between long-term glucagon treatment and mitochondrial dysregulation in proximal tubules, which is initiated by activating mTORC1 signaling.

Pharmacological Gcgr inhibition improved the pathological process of DKD progression

Given the dramatic effect of glucagon on lipid accumulation and DKD, we explored the therapeutic effects of Gcgr antagonism in the context of DKD. To do so, we treated Unix-HFD/STZ-induced DKD mice with REMD 2.59, a human monoclonal antibody that functions as a competitive antagonist of GCGR^28,61 (Fig. 7a). The GCGR mAb group displayed decreased glucagon signaling in terms of pCreb activation and Pepck expression (Fig. 7b, c, S7a). Before the administration of the REMD 2.59 antibody, all groups exhibited equivalent UACR levels (Figure S7b). After 5 weeks of treatment, compared with those in the DKD model group, renal function in the DKD model group improved markedly, as indicated by reduced urinary albumin excretion, improved proximal tubule injury, and decreased urine volume (Fig. 7d–g). Similarly, REMD 2.59 greatly lower UACR and 24-hour urine volume in db/db mice, with no differences in serum creatinine and urea nitrogen (BUN) levels between the two groups of mice (Figure S7c–f). Consistent with previous results, the levels of serum and kidney lipid accumulation were markedly lower in the GCGR mAb-treated group than in the vehicle-treated group (Fig. 7h–k, S7g–h), as was lipid synthesis and transport gene expression and fatty acid oxidation gene expression in the kidney, as shown by the REMD 2.59 antibody (Fig. 7l–n).

Fig. 7: GCGR mAb treatment improved DKD.

a GCGR mAb treatment protocol for DKD mice: modeled as in Fig. 1b, 5 mg/kg GCGR mAb i.p. injected for 5 week starting 4 week post-STZ, sacrificed 5 week later. b–c Renal p-Creb/total Creb by western blot. d–g UACR, 24-h urine volume, urine Kim-1/NAG/creatinine in the indicated groups of mice. h–j TG and TC levels in the kidney cortex and serum of mice in the indicated groups. k Kidney H&E/Oil Red O staining. Scale bar: black, 100 μm; yellow, 500 μm. l–n qRT‒PCR analyses of the key genes related to lipogenesis, fatty acid oxidation, and lipid transport. o Renal Masson/Sirius red/PAS staining. Scale bar: 100 μm. p Western blot and quantification of Collagen I in the kidneys. q TEM analyses of kidney glomeruli and proximal tubules in the kidney cortex. Scale bar: white, 2 μm; red, 5 μm. r-sGBM width and number of mitochondria. t–v Renal p-mTORC1/p-S6 levels. w GCGR mAb protocol for type 1 diabetic DKD rats: 6-week-old male Sprague‒Dawley rats were intraperitoneally injected with 65 mg/kg STZ, and 4 weeks after STZ injection, 5 mg/kg GCGR Ab was i.p. injected once a week for 7 weeks. x Urine protein levels in rats in the different groups. y–z Serum creatine and BUN levels in rats in the different groups. Mice were male and sampled after 5 weeks of GCGR mAb injection (b–v), and rats were male and sampled after 7 weeks of GCGR mAb injection (x–z). The error bars represent the SEMs; n = 4-10 in each group (d–j, l–n, s, x–z); n = 3 in each group (b, c, p, s–v). BUN, blood urea nitrogen.

We further assessed histologic lesion improvement by Gcgr inhibition via Masson, Sirius Red, and PAS staining, which revealed a substantial reduction in renal fibrosis and glycogen deposition after the injection of the GCGR mAb (Fig. 7o and S7i-k). Collagen accumulation was also notably decreased in the GCGR mAb group, confirming decreased fibrosis deposition after improved lipotoxicity (Fig. 7p). TEM also demonstrated a thinner GBM and a reduction of foot process width in the glomerulus and improved mitochondrial morphology and quantity in PTECs, accompanied by a great reduction in lipid droplet density (Fig. 7q–s and Figure S7l–m). In summary, these results further confirmed that Gcgr antagonism significantly prevents DKD-induced tubulointerstitial and glomerular lesions.

Furthermore, consistent with previous results, mTORC1 and S6 phosphorylation was increased in DKD mice but decreased after Gcgr inhibition, indicating the importance of Gcgr/mTORC1 signaling in DKD development (Fig. 7t–v).

To assess the potential therapeutic benefits of the GCGR mAb in type 1 diabetes-related kidney disease, we established a DKD model by intraperitoneally injecting Sprague‒Dawley (SD) rats with STZ until the model was successfully established. Subsequently, the rats were treated with 5 mg/kg REMD 2.59 or saline once a week (Fig. 7w). After 7 weeks of GCGR mAb administration, we observed notable improvements in urine protein secretion, serum creatinine and serum urea nitrogen (BUN) levels in the DKD rats, and these values were almost normalized to those of the control group without affecting hyperglycemia or body weight (Fig. 7x–z, S7n, o), indicating that the beneficial effect of Gcgr blockade on DKD does not depend on the blood glucose-lowering effect. Moreover, IHC staining of fibronectin and collagen IV and PAS staining demonstrated a significant reduction in fibrosis and glycogen accumulation in the treatment group (Figure S7p–s). TEM revealed improved podocyte morphology and GBM after GCGR mAb injection (Figure S7t–v). In summary, these findings suggest that Gcgr inhibition may offer favorable benefits in the treatment of both type 2 and type 1 diabetes-induced kidney disease and that improving lipid metabolism in PTECs plays a significant role in these therapeutic effects.

Discussion

Here, we found that there is Gcgr in the cortex of kidney, especially in the proximal tubules of the kidney. Sustained elevation of glucagon resulted in lipid accumulation and the development of DKD. Previous RNA-seq and proteomic analysis showed that Gcgr expression in proximal tubules of the kidney is much lower compared to the thick ascending limb of the loop of Henle^62,63. Both WB and qPCR results in this study (Fig. S1a–h) argued a decent amount of Gcgr expression (one-seventh that of hepatocytes) in the PTECs. Most importantly, the Gcgr in PTECs becomes pathogenic when glucagon is sustained elevated, as shown in Fig. 5b in primary PTECs, prolonged glucagon exposure directly promoted lipid accumulation. Although glucagon activates both Gcgr and GLP-1R, its impact on lipid accumulation and renal function is Gcgr dependent, as it has no effect in Gcgr-deficient primary PTECs and Gcgr^fl/flKsp^Cre mice (Fig. 5b and S3). This study also indicates that glucagon and GLP1, both from the proglucagon gene, show dual actions in DKD.

Serum glucagon levels and kidney Gcgr expression were both upregulated under DKD conditions. Additionally, our investigation revealed that the targeted knockdown/deletion of Gcgr in tubular epithelial cells almost completely corrected nephritic lipid deposition in cortex, mitochondrial apoptosis, proximal tubular epithelial cell injury, ECM accumulation, and urinary albumin excretion in DKD mice, suggesting that Gcgr activation may play a critical role in PTEC injury.

The main finding in this study is that elevated glucagon levels may serve as a critical trigger of DKD by expediting lipid toxicity through the glucagon-Gcgr-PKA-Creb-mTORC1 pathway. DKD is accompanied by impaired lipid oxidation and excess lipid accumulation in the kidney^19,40,64. During lipid toxicity, intracellular mitochondria are unable to metabolize excessive lipids, eventually inducing mitochondrial fission and fragmentation and leading to proximal tubular cell apoptosis^43,54. Under physiological conditions, glucagon is involved in lipid oxidation and may increase resting energy expenditure via three major pathways, namely, the Creb-dependent pathway to facilitate the breakdown of long-chain fatty acids and provide substrates for β-oxidation; the blockade of malonyl CoA formation to direct free fatty acids (FFAs) into β-oxidation instead of re-esterification into triglycerides; and the stimulation of peroxisome proliferator-activated receptor-α transcription to increase the transcription of genes involved in β-oxidation. Notably, β-oxidation is a significant source of metabolic energy during interprandial periods and during high energy demand states. PTECs rely largely on fatty acid uptake and oxidation as a fuel source to maintain the reabsorptive capacity of the kidney⁴⁵, as fatty acid oxidation provides more energy than glycolysis, and glucagon strongly promotes fatty acid oxidation^29,30,31, which may explain the high expression of Gcgr in PTECs. However, under pathological conditions, we found that sustained high glucagon levels had the opposite effect on fatty acid metabolism. Long-term glucagon treatment promoted SREBP1-regulated de novo lipogenesis-related gene expression in HK2 cells and primary PTECs through mTORC1 signaling (Fig. 5e). In addition, the transcription of S6K and RRAGA, which are related to mTORC1, was regulated by CREB and downregulated by GCGR (Fig. 5m, n). Experiments in HK2 cells and primary PTECs with mTORC1 and PKA inhibitors confirmed that the lipid synthesis effect of glucagon occurs through mTORC1 signaling (Fig. 5p–s). Therefore, this study establishes a connection among glucagon, lipid accumulation, and DKD. And the underlying mechanism is also distinctive, as this study demonstrates glucagon as a regulator of mTORC1 signaling—an aspect that has not been extensively reported in DKD.

Under physiological conditions, the majority of hormones secreted by glands act on their specific receptors immediately; additionally, they are regulated mostly by feedback regulation to maintain endohomeostasis⁶⁵. However, in individuals with obesity and/or diabetic patients, the secretion of hormones such as insulin, leptin, cortisol, and other growth hormones is disrupted; accordingly, the sensitivity of hormones decreases with diabetes development^66,67. For instance, our previous study showed that insulin promotes the interaction of the insulin receptor (InsR) with EphB4 in the liver, which facilitates the internalization of InsR on the cell membrane and its degradation in lysosomes, eventually leading to insulin resistance⁶⁸. Glucagon normally circulates at low picomolar concentrations, and its level in circulation greatly increases due to β-cell transdifferentiation/dedifferentiation into α-cells in islets or α-cell insulin resistance^{20,21,22,23,25,69}. In contrast to the effect of hyperinsulinemia on InsR degradation, hyperglucagonemia has little effect on the level of its receptor Gcgr in the kidney (Figs. 1f and g, 2c and d, 3c and d). In contrast to its traditional effect on lipid oxidation, prolonged hyperglucagonemia leads to lipid accumulation in PTECs (Figs. 2l and q, 3l and p, 5a-d). In other words, long-term hyperglucagonemia results in glucagon reversal (reversal of the action of glucagon), which is different from glucagon resistance (impaired glucagon action). Thus, the results of the present study demonstrated that the action of a certain hormone is not constant even when the hormone is present in the same cell and that the hormone could exert completely opposite effects on the same tissue over different periods.

One recent study found that deletion of Gcgr in the kidney leads to renal lipid deposition, inflammation, and fibrosis⁶³. The mouse in this paper is on a normal chow or high fat diet. The circulating glucagon levels in these mice are still in the normal range, so the authors studied the physiological function of glucagon/Gcgr. However, in our study, we knocked out Gcgr in diabetic mice, which was characterized by hyperglucagonemia, so we studied the pathological role of glucagon/Gcgr. As we showed here, in diabetic conditions, only the sustained hyperglucagonemia led to the reversal of its effect on fatty acid oxidation, and blockade of Gcgr improved renal lipid deposition, inflammation, and fibrosis in DKD models. Our study is consistent with the clinal observation that there is a positive relationship between hyperglucagonemia and the pathology of DKD^33,34,70. Another study showed that Gcgr knockdown aggravated the progression of kidney injury in db/db mice⁷¹. In this study, the authors injected AAV9-Gcgr shRNA, which was not driven by a specific promoter, indicating the whole kidney was knocked down. This is quite different from our studies (tubular epithelial cell-specific Gcgr knockdown). Additionally, there were no measurements of lipids, kidney function like UACR, or knockdown efficiency, so it’s hard to compare this study to our findings.

Furthermore, our studies showed that glucagon reversal plays a critical role in PTEC lipid synthesis and DKD progression (Figs. 1j–n, 4a–d). Glucagon reversal may also explain the paradox in diabetic patients in which hyperglucagonemia is positively correlated with the pathology of DKD^20,28,33,34. In recent years, several pharmacological therapies for DKD, such as blood pressure control with angiotensin converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), sodium glucose transport protein 2 (SGLT2) inhibitors, and glucagon-like peptide 1 (GLP-1) agonists, have been developed⁷². However, the progression of DKD is not completely controlled even though hypertension and hyperglycemia are tightly controlled. Therefore, there is a pressing need for new and more effective therapeutic targets for DKD^73,74. Here, we found that Gcgr blockade through either genetic KO or antibody treatment nearly completely halted lipid accumulation and the progression of DKD (Figs. 2f–k, 3f–k, 4e–j, S4a–e, S8i). Thus, our study provides a fresh perspective on the pathogenesis of lipotoxicity in DKD and a potential treatment strategy.

Limitation

The pathophysiology of DKD is highly intricate, characterized by heterogeneous lesions that culminate in renal dysfunction. Among patients with T2DM complicated by DKD, only 30%-50% present with classic glomerulopathy, whereas the remaining cases exhibit atypical manifestations—such as tubular or arteriolar abnormalities—often modulated by comorbid conditions like obesity and hypertension. The molecular mechanisms underpinning the pathogenesis of DKD are equally complex. In our study, we primarily employed the Unix-HFD/STZ-induced DKD mouse model which has limitations in fully recapitulating the pathogenic processes of human DKD.

Hyperglycemia is a key driver in the progression of kidney disease. In our study, inhibiting Gcgr in a mouse model of type 2 diabetic DKD reduced blood glucose levels. The extent of its contribution to alleviating the progression of DKD requires further research. However, our experimental evidence shows that injecting Glucagon in the early stage of DKD did not further increase blood glucose levels in model mice, but accelerated the progression of their DKD (Fig. 1 and S1). Moreover, in a rat model of type 1 diabetic nephropathy, the use of GCGR mAb can alleviate the DKD phenotype in model rats without reducing blood glucose (Fig. 7 and S7). These results indicate that the role of Glucagon/Gcgr in DKD is independent of the regulation of blood glucose levels. Additionally, our research demonstrated that inhibition of Gcgr signaling effectively alleviates the progression of DKD in both type 1 and type 2 diabetes. However, our investigations into the underlying molecular mechanisms have focused predominantly on a type 2 DKD mouse model. Whether the same molecular mechanisms exist in the pathogenesis of type 1 DKD needs further verification.

We observed a continuous elevation of glucagon in diabetes, as the serum glucagon levels in 15-week-old male db/db mice at different time points were all increased compared to the controls (Fig. 1a). In both db/db mice and other DKD animal models, we observed not only elevated glucagon levels but also increased renal lipid accumulation (Figs. 1c, 2l, 3l, 4a, b, 4e, f, 7h, i). And inhibiting the glucagon receptor alleviated such lipid accumulation in the kidneys (Figs. 2l, 3l, 4e, f, S4e-f, 7h–i, 7k). Of course, the physiological environment in vivo is more complex than in vitro systems, and such a direct reversal effect requires more elaborate research. Another explanation is that glucagon exerts different effects in healthy and diseased renal tubules.

Methods

Animals and treatments

C57BL/6 J mice were obtained from GemPharmatech (Nanjing, China), and db/db mice were obtained from Huafukang Biotechnology (Beijing, China). All mice were housed in colony cages in a pathogen-free environment with the temperature and relative humidity maintained at 21–23 °C at 50–60%, respectively, on a 12 h:12 h light:dark cycle. The HFD was a rodent diet containing 60 kcal% fat (Research Diets, D12492). Male mice were used in all the experiments. Euthanasia of the mice was performed using the carbon dioxide (CO₂) inhalation method. The key operational points are as follows: in a dedicated airtight container, first infuse 40–50% CO₂, maintain this concentration for 1 minute, then increase the CO₂ concentration to 100% and keep it at this level for 5 minutes. All experiments using animals were performed according to protocols approved by the Animal Experimentation Ethics Committee of the Chinese Academy of Medical Sciences, and all procedures were conducted in accordance with the guidelines of the Institutional Animal Care and Use Committees of the Chinese Academy of Medical Sciences. All animal procedures were consistent with the ARRIVE guidelines.

Generation of tubular epithelial cell-specific Gcgr knockout mice

Gcgr floxed mice (C57BL/6 JGpt-Gcgr^em1Cflox/Gpt) were purchased from GemPharmatech (Nanjing, China), and Ksp-Cre mice (B6. Cg-Tg(Cdh16-Cre)91Igr/J) mice were purchased from Jackson Laboratory. All the Gcgr floxed mice and Ksp-Cre mice were bred on a C57BL/6J background. Gcgr floxed mice were crossed with Ksp-Cre mice to generate renal tubular epithelial cell-specific Gcgr knockout mice (Gcgr^fl/flKsp^Cre mice). The homozygous floxed mice (Gcgr^fl/fl) were used as controls. All mice were genotyped 3 weeks after birth by PCR with specific primers: flox genotyping primers (5’arm F-TATCGCTCAGTTGAGGGAGATG, R-GTACACTAGCATATAGGATCATAGGGC; 3’arm F-TTTGGCTGTTGCTGTGATGC, R-TCCTCAAGTCCCTCCCTGCTAC). The following primers were used for Ksp-Cre genotyping: 9808-GCAGATCTGGCTCTCCAAAG, oIMR8744-CAAATGTTGCTTGTCTGGTG, oIMR8745-GTCAGTCGAGTGCACAGTTT, and oIMR9074-AGGCAAATTTTGGTGTACGG.

DKD mouse model construction

The diabetic nephropathy mouse model was established following the methodology reported by Fu et al. (Ref. ¹²). 6-week-old male mice were fed a HFD for 4 weeks, after which the mice were uni- or sham-nephrectomized. After 2 weeks of recovery from surgery, the mice were injected with STZ, which was dissolved in 50 mmol/L sodium citrate buffer (pH 4.5), at a dosage of 50 mg/kg body weight intraperitoneally daily for 3 consecutive days following an overnight fast. Mice were maintained on a HFD for another 4–8 weeks. UACR was assessed every 2 weeks as a marker of successful establishment of DKD; at the end, urine samples were collected for biochemical analysis. For sacrifice, the mice were anesthetized, blood and kidney samples were harvested for histopathological analysis and subsequent protein or RNA extraction. For the study of Ggcr knockdown mice, 4 weeks after the STZ injections, Ctrl shRNA or Gcgr shRNA was injected in situ into mouse kidneys. For the study of GCGR mAb treatment in DKD mice, 4 weeks after the STZ injections, 5 mg/kg GCGR mAb was intraperitoneally injected once a week for 5 weeks.

For the study of GCGR mAb treatment in DKD rats, 6-week-old male SD rats were injected with STZ, which was dissolved in 50 mmol/L sodium citrate buffer (pH 4.5), at a dosage of 65 mg/kg body weight intraperitoneally, following an overnight fast. 4 weeks after STZ injection, the mice were intraperitoneally injected with the GCGR mAb (5 mg/kg) once a week for 7 weeks.

For the study of glucagon treatment in the early phase of DKD in mice, 6-week-old male C57BL/6 J mice were fed a HFD for 4 weeks, after which the mice were uni- or sham-nephrectomized. After 2 weeks of recovery from uninephrectomy, the mice were injected with STZ, which was dissolved in 50 mmol/L sodium citrate buffer (pH 4.5), at a dosage of 50 mg/kg body weight intraperitoneally daily for 3 consecutive days, following an overnight fast. One week after the STZ injections, the mice were intraperitoneally injected with 300 nmol/kg glucagon twice a day for 2 weeks.

Renal tubular epithelial cell-specific Gcgr knockdown

The sequence of the shRNA targeting Gcgr was 5’- TGGTCAAGTGTCTGTTTGA-3’³⁹. Expression was driven by the cadherin 16 (Cdh16, Ksp) promoter carried by AAV9. The virus (5E + 10 v.g./kidney) was injected into the kidney in situ after 4 weeks of STZ injections.

Primary cell isolation and culture

After the mice were anesthetized, their kidneys were promptly removed and placed in cold PBS at 4 °C. The renal capsule was removed, and each kidney was cut sagittally into two halves. The medulla was dissected and discarded from each half kidney, leaving behind the cortical tissue. The cortical tissue was minced and digested in HBSS containing collagenase type II (Thermo Fisher Scientific, USA; 200 units/mL) and hyaluronidase (Sigma-Aldric, USA; 0.2%). The tissues were incubated at 37 °C while being rotated for 2 h. Following the digestion process, the tubule suspension was filtered through a 100 μm cell strainer and then centrifuged at 100 × g for 2 min. The isolation of PTECs was carried out using a CELLection Biotin Binder Kit (Invitrogen, USA, 11533D) and biotinylated LTL (Vector Laboratories, USA, B-1325-2). The isolated PTECs were resuspended in DMEM/F12 medium (Macgene, China, CM10090) supplemented with 10% FBS, 2 mM GlutaMax (Thermo Fisher, USA, 35050061), and penicillin/streptomycin (Thermo Fisher, USA, 15140122). The cells were then seeded into culture dishes for further experimentation.

Cell culture

All cells were grown at 37 °C in a 5% CO₂ humidified atmosphere. Human tubule epithelial cells (HK-2) were cultured in DMEM/F12 (Macgene, China, CM10090) supplemented with 10% FBS (Gibco, USA, 16000-044), GlutaMax (Gibco, USA, 35050-061) and penicillin/streptomycin (Gibco, USA, 15140122). Rat proximal tubule epithelial cells (NRK-52E) were cultured in DMEM (Gibco, USA, 12800-017) supplemented with 10% FBS (Gibco, USA, 16000-044), GlutaMax (Gibco, USA, 35050-061) and penicillin/streptomycin (Gibco, USA, 15140122). We used PCR assays with species-specific primers to identify the cell line.

ChIP–qPCR

ChIP-qPCR assays were carried out using the Pierce Magnetic ChIP Kit (Thermo Fisher Scientific, USA, #26157) following the manufacturer’s protocol. Briefly, HK2 cells were treated with glucagon (200 nM) for 36 h and then crosslinked with 1% formaldehyde for 20 min at room temperature, after which the reaction was stopped with glycine for 5 min. These cells were subsequently lysed with Membrane Extraction Buffer to isolate the nuclei, which were subsequently resuspended in MNase Digestion Buffer with MNase to cleave the chromatin. After digestion, the nuclei were resuspended and sonicated using a Scientz sonicator. The sonicated samples were immunoprecipitated overnight at 4 °C with an anti-CREB (Cell Signaling Technology, USA, 9197) antibody or normal IgG as a control. After the samples were incubated with antibodies, protein A/G magnetic beads were added to the samples for another 2 h at 4 °C. After standard washes, elution buffer was added to all the immunoprecipitation samples, and the DNA was removed and subjected to qPCR.

Western blot analysis

Tissue or cell samples were homogenized in RIPA buffer (Beyotime, China, P0013B) supplemented with protease and phosphatase inhibitors. The lysates were subjected to western blotting. The protein bands were analyzed using densitometry and ImageJ image analysis software, and the levels of phosphorylated proteins were normalized to the total protein band density. Arbitrary densitometry units were used for quantification, and the data are expressed as the mean ± SEM.

Albumin, creatinine, Kim-1, and NAG measurements

Mouse urine was collected using a metabolic cage and then centrifuged at 800 × g for 15 min at 4 °C. The resulting supernatant was stored at −80 °C for further assays. Mouse urinary albumin concentrations were quantified using a mouse albumin ELISA quantitation kit following the manufacturer’s protocol (Bethyl Laboratories, USA, E99-134). Urinary creatinine levels were determined using a commercial diagnostic kit (Nanjing Jiancheng Bioengineering Institute, China, C011-2-1). Urinary Kim-1 levels were measured using a mouse Kim-1 Quantikine ELISA Kit (R&D Systems, USA, MKM100) following the manufacturer’s instructions. Urinary NAG activity was measured using a commercial kit (Crystal Chem, USA, 80400).

Seahorse analysis

To perform real-time OCR analyses, we analyzed NRK-52E cells using the Seahorse XF24 Analyzer from Agilent Technologies and the XF Mito Stress Test Kit according to the manufacturer’s instructions. A total of 1 × 10⁴ cells were seeded into XF24 cell culture microplates, with n = 6-8 replicates. Cells were treated with 100 nM glucagon for 24 h. Prior to the assay, 85 μL of XF Palmitate-BSA FAO substrate or BSA was added to the appropriate wells (Agilent, USA, 103693-100). The XF cell culture microplates were immediately inserted into the XF24 Analyzer, after which the XF Mito Stress Test was performed. The results were normalized to the total protein in each well.

ATP/ADP ratio measurement in cultured cells

Confluent HK2 cells were treated with 100 nM glucagon for the indicated duration. Subsequently, the cells were harvested using the extraction solution provided in the ATP/ADP Ratio Assay Kit (Abcam, MA, ab65313). The ATP/ADP ratio was subsequently measured using a kit in accordance with the manufacturer’s instructions.

RNA extraction and qRT‒PCR

Total RNA was isolated from cells and tissues using TRIzol reagent (Thermo Fisher Scientific, USA, 15596018). Then, 1 μg of RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, USA, #4368813). qRT‒PCR was carried out using SYBR Green qRT‒PCR Master Mix (Bimake, USA, B21203) on a QuantStudio 3 Real-Time PCR System (Applied Biosystems, USA, QuantStudio™ 3). The primer sequences are listed in Supplementary Table 1. Relative gene expression was determined by normalizing the mRNA levels to those of a reference housekeeping gene (ribosomal protein lateral stalk subunit P0, Rplp0, also known as 36B4) using the ddCt method. The specificity of the PCR amplification was verified by melting curve analysis of the final products using QuantStudio software.

Lipid extraction and analysis

Kidneys were homogenized in 5% Triton X-100 in PBS and incubated at room temperature for 10 min. Then, the supernatant was used for lipid analysis. The nonesterified fatty acid (FUJIFILM, Japan, 633-52001), triglyceride (Biosino, China, 20162400911), and cholesterol (Biosino, China, 20162400910) contents in the kidneys were measured using commercially available kits. HK2 cells and PTECs were homogenized in 5% Triton X-100 in PBS and sonicated until the cells were disrupted. The triglyceride content in the suspension was measured (Biosino, China, 20162400911), and the values were normalized to the protein concentrations.

Immunofluorescence staining

Mouse kidney biopsy paraffin sections were deparaffinized with xylene and ethanol. The sections were blocked with normal sheep serum (ZSGB-BIO, China, ZLI-9056) and incubated with primary antibodies overnight at 4 °C. After washing with PBS, the sections were incubated with secondary antibodies for 1 h at room temperature. All images were obtained by standard or confocal microscopy (Zeiss, Germany, LSM880).

AFADESI-MSI analysis

Using Leica Cryo-Gel embedding medium, tissues were fixed onto sample holders. After the embedding medium solidified, 15 µm kidney tissue sections were prepared using the Leica CM1860 cryostat. Mass spectrometry imaging was performed using an Airflow-assisted Desorption Electrospray Ionization (AFADESI) system, which was equipped with the MassImager Pro advanced mass spectrometry imaging system workstation (version 1.0, Beijing Chemmind Co., Ltd.), a Q-Exactive quadrupole-electrostatic field orbitrap high-resolution mass spectrometer (Thermo Fisher, USA), and the Xcalibur data acquisition and processing system. The specific experimental parameters were set as follows: spray solvent was acetonitrile:water = 8:2 (v/v), spray solvent flow rate was 5 μL/min, spray voltage was 7 kV/−7 kV, ion transfer tube temperature was 350 °C, scanning range was 70–1000 Da, maximum injection time was 200 ms, automatic gain control target was 3E6, resolution was 70,000, spray pressure was 0.7 MPa, exhaust flow rate was 45 L/min, spray needle inner diameter was 20 μm, and scanning step size was 0.1 mm/s * 0.15 mm/step (x * y). After data acquisition, a two-dimensional data matrix with m/z as the x-axis and ion signal intensity as the y-axis was obtained. Then, peak alignment and isotopic ion removal were performed using MarkerView 1.2.1 (AB SCIEX, USA) software. Differential ions were screened using Student’s t-test with a threshold of p < 0.05, and annotation of differential ions was completed using MSIAnnotator⁷⁵.

Serum creatine and BUN measurements

Serum creatine and blood urea nitrogen (BUN) measurements were performed with creatine and BUN diagnostic kits (Nanjing Jiancheng Bioengineering Institute, China, C011-2-1 & C013-2-1).

Glomerular filtration rate (GFR)

The GFR was measured by a noninvasive-clearance transcutaneous measurement based on the elimination kinetics of a fluorescent exogenous marker, FITC-Sinistrin (Fresenius-Kab Austria GmbH). Mice were mildly anesthetized with tribromoethanol was applied to a small area of the back to remove hair. Next, a detecting device (MB-102, Medibeacon) that can measure fluorescent compounds transcutaneously in blood was attatched to the shaved area of the mouse for 5 minutes to subtract background signal. FITC-sinistrin was dissolved in saline solution (35 mg/ml) and intravenous injected at a dose of 7 mg/100 g body weight and monitored for 1 hours after injection. Data was analyzed using MPD Studio software.

Dulaglutide administration

Dulaglutide (Selleck, China, P1115) was intraperitoneally injected at a dose of 0.6 mg/kg body weight once every 7 days for 4 weeks.

Quantification and statistical analysis

The data are presented as the mean ± SEM. The significance of differences between groups was evaluated using ANOVA. A p value < 0.05 was considered to indicate statistical significance. The statistical parameters, including the exact value of n and precision measures (the mean ± SEM), and the statistical significance are reported in the figures and the figure legends. Differences were considered significant at p < 0.05 according to two-tailed Student’s t tests or one-way ANOVA, where appropriate. In the figures, asterisks denote statistical significance (*, p < 0.05; **, p < 0.01; and ***, p < 0.001). Statistical analyses were performed using Graph Pad Prism 9 software.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.