Abstract

   Podocyte injury is a critical event in the pathogenesis of diabetic
   nephropathy (DN). Hyperglycemia, oxidative stress, inflammation, and
   other factors contribute to podocyte damage in DN. In this study, we
   demonstrate that signaling regulatory protein alpha (SIRPα) plays a
   pivotal role in regulating the metabolic and immune homeostasis of
   podocytes. Deletion of SIRPα in podocytes exacerbates, while transgenic
   overexpression of SIRPα alleviates, podocyte injury in experimental DN
   mice. Mechanistically, SIRPα downregulation promotes pyruvate kinase M2
   (PKM2) phosphorylation, initiating a positive feedback loop that
   involves PKM2 nuclear translocation, NF-κB activation, and oxidative
   stress, ultimately impairing aerobic glycolysis. Consistent with this
   mechanism, shikonin ameliorates podocyte injury by reducing PKM2
   nuclear translocation, preventing oxidative stress and NF-κB
   activation, thereby restoring aerobic glycolysis.

   Keywords: SIRPα, Podocyte injury, PKM2, Oxidative stress, Aerobic
   glycolysis, NF-κB

Graphical abstract

   [41]Image 1
   [42]Open in a new tab

High light

     * •
       Podocyte SIRPα is downregulated in DN, and whereas transgenic SIRPα
       reduces podocyte damage, its deletion exacerbates it.
     * •
       SIRPα deficiency promoted PKM2 nuclear translocation, which
       triggers NF-κB activation and oxidative stress.
     * •
       Shikonin relieves SIRPα-deficient podocyte damage by preventing
       PKM2 nuclear translocation.
     __________________________________________________________________

   Abbreviations
   Abbreviations Definition
   DN            Diabetic nephropathy
   SIRPα         signaling regulatory protein alpha
   PKM2          pyruvate kinase M2
   WT            Wild type
   CKO           Podocyte-specific SIRPα-knockout
   KO            Primary podocytes from podocyte-specific SIRPα-knockout
   TG            Podocyte-specific SIRPα-transgenic
   STZ           streptozotocin
   HPC           human podocyte cell line
   PK            Pyruvate kinase
   HG            High glucose
   OCR           Oxygen consumption rate
   FCCP          fluorocarbonyl cyanide phenylhydrazone
   GBM           Glomerular basement membrane
   GSEA          Gene Set Enrichment Analysis
   NOX4          NADPH oxidase 4
   TEMPO         Mito-TEMPO
   TEM           Transmission electron microscope
   PAS           Periodic Acid-Schiff
   OE-SIRPα      Overexpressing SIRPα
   pPKM2         Phosphorylated PKM2 (Tyr105)
   [43]Open in a new tab

1. Introduction

   Diabetic nephropathy (DN) is one of the most severe microvascular
   complications of diabetes mellitus and a leading cause of end-stage
   renal disease (ESRD)[[44]1] [[45][1], [46][2]]. Podocyte injury is a
   central mechanism in the progression of DN. Chronic hyperglycemia
   exacerbates oxidative stress, leading to increased production of
   reactive oxygen species (ROS), mitochondrial dysfunction, and
   activation of inflammatory pathways [[47][3], [48][4], [49][5],
   [50][6], [51][7], [52][8]]. Numerous studies suggest that maintaining a
   high level of ATP is essential for preserving the filtration structure
   and function of podocytes. Both mitochondrial function and glycolysis
   are critical in meeting the high ATP demand in podocytes [[53][8],
   [54][9], [55][10]]. Additionally, podocytes contain components of both
   innate and adaptive immune systems, enabling them to actively
   participate in inflammatory responses [[56]11]. However, the
   interaction between metabolic dysregulation and inflammatory activation
   in podocyte injury under hyperglycemic conditions remains poorly
   understood.

   Signal regulatory protein alpha (SIRPα), a member of the immunoglobulin
   superfamily, is predominantly expressed in leukocytes and plays a
   critical role in modulating various inflammatory responses [[57][12],
   [58][13], [59][14]]. The cytoplasmic tail of SIRPα contains two
   immunoreceptor tyrosine-based inhibitory motifs (ITIMs), which recruit
   and activate SHP-1/2, resulting in the dephosphorylation of
   intracellular phosphorylated substrates and the regulation of various
   downstream signaling pathways [[60]15]. Recent studies have
   demonstrated that SIRPα is also expressed in podocytes [[61]16,[62]17].
   However, while these studies focus on the role of SIRPα in maintaining
   the normal structure of podocytes, there is a notable lack of research
   on the function of SIRPα under pathological conditions such as DN.

   In this study, we identified the reduction of SIRPα as an independent
   risk factor for renal function decline in patients with DN. We further
   validated the protective effects of podocyte-specific transgenic SIRPα
   against podocyte injury and elucidated the underlying protective
   mechanism, which involves the regulation of PKM2 nuclear translocation.

2. Materials and METHODS

Ethics approval

   The collection of renal biopsy samples was approved by the National
   Clinical Research Center of Jinling Hospital (Nanjing, China). All
   animal experiments were conducted in accordance with the National
   Institutes of Health (NIH) Guide for the Care and Use of Laboratory
   Animals.

2.1. Clinical and pathological information

   All individuals underwent renal biopsies between 2013 and 2023.
   Baseline clinical features were collected within one month following
   the biopsy for each patient. Kidney samples were obtained from the
   Renal Biobank at the National Clinical Research Center for Kidney
   Diseases.

2.2. Animals

   Podocyte-specific SIRPα-knockout (CKO) mice were generated as described
   in our previous study [[63]18]. Podocyte-specific SIRPα-transgenic (TG)
   mice were obtained from GemPharmatech. Podocyte-specific
   SIRPα-transgenic (TG) mice were obtained from GemPharmatech.
   SIRPα-transgenic flox homozygous mice were created using the
   CRISPR-Cas9 system. To generate SIRPα-transgenic mice, SIRPα-transgenic
   flox homozygous mice were mated with NPHS2-Cre mice. SIRPα-transgenic
   flox homozygous mice were created using the CRISPR-Cas9 system. To
   generate SIRPα-transgenic mice, SIRPα-transgenic flox homozygous mice
   were mated with NPHS2-Cre mice. For the construction of the DN model,
   Eight-week-old male mice were intraperitoneally injected with
   streptozotocin (STZ) (50 mg/kg, Sigma, S0130, St. Louis, MO, USA) for
   five consecutive days [[64]19]. Blood glucose levels were checked one
   week later, and mice with stable blood glucose levels exceeding
   11.3 mmol/L after one month were included in the follow-up study. Blood
   glucose, urine protein, and urine creatinine levels were measured
   weekly. For treatment, DN model mice received intraperitoneal
   injections of shikonin (MCE, HY-N0822, New Jersey, USA) at a dosage of
   2 mg/kg body weight every other day, beginning at 20 weeks of age
   [[65][20], [66][21], [67][22]]. At 24 weeks of age, mice (n = 6) were
   anesthetized with isoflurane, and samples were collected. All animal
   experiments were conducted in accordance with the guidelines
   established by the Animal Ethics Committee.

2.3. Urinary protein and creatinine

   Mouse urine was collected using a metabolic cage, and samples were
   centrifuged to remove cells, cell debris, and other contaminants.
   Urinary albumin was measured with the Mouse Urinary Albumin Assay Kit
   (Wako, 291–92703, Osaka, Japan), and urinary creatinine was measured
   with a creatinine kit (Wako, 290–65901, Osaka, Japan).

2.4. Transmission electron microscope (TEM)

   Renal tissues were collected, cut into 1 mm³ pieces, and fixed with
   2.5 % glutaraldehyde at 4 °C overnight. After fixation, the tissues
   were washed three times with phosphate-buffered saline and post-fixed
   with 1 % osmium tetroxide at 4 °C for 2 hours. The tissues were then
   dehydrated in a graded series of acetone and ethanol before being
   embedded in epoxy resin. Sections of 80–90 nm thickness were stained
   with 5 % uranyl acetate and 0.1 % lead citrate. Electron micrographs
   were obtained and analyzed using a Hitachi 7500 transmission electron
   microscope.

2.5. RT-qPCR

   Total RNA was extracted from tissues or cells using Trizol reagent
   (Takara, 9109, Kusatsu, Japan). The cDNA was synthesized using the
   HiScript III All-in-one RT SuperMix Perfect for qPCR (Vazyme, R333-01,
   Nanjing, China). Gene expression was measured using SYBR Green
   real-time quantitative PCR (ChamQ Universal SYBR qPCR Master Mix,
   Vazyme, Q711-02, Nanjing, China), with mRNA levels normalized to
   β-actin expression. Primer sequences are listed in [68]Supplementary
   Table 1.

2.6. Histological analysis

   For Periodic Acid-Schiff (PAS) staining (Beyotime, C0142S, Shanghai,
   China) and immunohistochemical assays, renal tissues were embedded in
   paraffin and sectioned at a thickness of 2 μm. For immunofluorescence
   assays, renal tissues were embedded in optimal cutting temperature
   compound (OCT, Tissue-Tek, 4583, Osaka, Japan) and sectioned at a
   thickness of 4 μm. Details of antibodies used are listed in
   [69]Supplementary Table 2. DAPI (1:4000, Beyotime, C1002, Shanghai,
   China) was used to stain nuclei. Stained tissues were imaged using an
   Olympus FluoView 3000 confocal microscope (Olympus, Tokyo, Japan) and
   analyzed with FV3000-ASW software (version 1.3c; Olympus). Fluorescence
   intensity was quantified using ImageJ.

2.7. Western blot analyses

   Proteins were extracted using RIPA lysis buffer (Beyotime, P0013B,
   Shanghai, China) supplemented with 100 μg/mL PMSF (Beyotime, ST507,
   Shanghai, China) and a phosphatase inhibitor cocktail (Thermo, 78442,
   Waltham, US). The extracted proteins were separated by SDS-PAGE,
   transferred to PVDF membranes (Sigma, IPVH00010, St. Louis, MO, US),
   and then blocked with BSA. Membranes were incubated with primary and
   secondary antibodies, as detailed in [70]Supplementary Table 2.

2.8. Extraction and culture of primary mouse podocytes

   Primary mouse podocytes were isolated according to the established
   protocol[[71]23] [[72][23], [73][24]]. In brief, mice were anesthetized
   with isoflurane. Sequentially, 10 mL PBS, 2 mL Perfusion Solution 1
   (containing 50 μL/mL Dynabeads [Dynabeads M450 tosylactivated;
   Invitrogen, 14013, Carlsbad, CA, US] in PBS), and 2 mL Perfusion
   Solution 2 (containing 50 μL/mL Dynabeads, 1 mg/mL Collagenase [Sigma,
   C6885, St. Louis, MO, US], 0.5 mg/mL Pronase E [Sigma, P6911, St.
   Louis, MO, US], and 50 U/mL DNase I [Sigma, D5025, St. Louis, MO, US]
   in PBS) were perfused at a rate of 5 mL/min. The perfused kidneys were
   cut into 1 mm³ pieces and transferred to a 2 mL centrifuge tube.
   Subsequently, 1 mL digestion solution (1 mg/mL Collagenase, 0.5 mg/mL
   Pronase E, and 50 U/mL DNase I in PBS) was added, and the tube was
   centrifuged at 300 rpm for 5 min. The digested solution was filtered
   through a 100 μm sieve, followed by magnetic separation to isolate
   glomeruli. Glomeruli were then washed five times with 1 mL PBS
   containing 10 % FBS and 3 % Penicillin/Streptomycin. Finally, glomeruli
   were resuspended and plated at a density of 300 glomeruli per 10 cm^2
   on type I collagen-coated culture plates or flasks for further
   experiments. Cells were incubated with high glucose (48 hours, 30 mM,
   Sigma, 49163, St. Louis, MO, US) [[74]25], shikonin (1 μM, 24 hours,
   MCE, HY-N0822, New Jersey, US) [[75]26], mito-TEMPO (10 μM, 24 hours,
   MCE, HY-112879, New Jersey, US) [[76]27], or GLX351322 (10 μM,
   24 hours, MCE, HY-100111, New Jersey, US) [[77]28].

2.9. Detection of pyruvate kinase (PK) activity

   A total of 1 × 10^6 podocytes from each group were homogenized in
   200 μL of normal saline (0.9 % NaCl). After centrifugation, the protein
   concentration of the supernatant was determined. Pyruvate kinase
   activity was measured using the Pyruvate Kinase Activity Assay Kit
   (Elabscience, E-BC-K611-M, Shanghai, China). Activity was calculated as
   the amount of enzyme required to consume 1 mmol of NADH per minute per
   gram of cellular protein at room temperature, defining one unit of
   pyruvate kinase activity.

2.10. Detection of total ROS

   ROS levels were quantified using a Total ROS Assay Kit (Invitrogen,
   88-5930-74, Carlsbad, CA, US). Briefly, dilute the ROS assay buffer
   (500 × ) with dimethyl sulfoxide (DMSO). Prepare at least
   5 × 10^5 cells per group. Incubate the cells with the ROS assay buffer
   in a 37 °C incubator with 5 % CO₂ for 60 min. After staining,
   centrifuge and resuspend the cells in phosphate-buffered saline (PBS),
   then immediately analyze on a flow cytometer (Beckman-Coulter Cyan ADP,
   Brea, CA). A total of 1 × 10^4 single cells were analyzed for each
   group, and the average fluorescence intensity was recorded.

2.11. Seahorse assay

   Oxygen consumption rate (OCR) was measured using an XF96 extracellular
   flux analyzer (Seahorse Bioscience) with the Seahorse XF Cell Mito
   Stress Test Kit (Agilent, 103015-100, Santa Clara, CA, US) Following
   previously published protocols [[78]4], cells were seeded on XF96 cell
   culture plates at a density of 1.5 × 10⁴ cells per well (near 90 %
   confluence) on the first day. Since podocytes cannot proliferate, they
   were cultured for 3–5 days to ensure stable attachment to the plate. On
   the day before the assay, hydrate the probe plate by adding calibration
   solution (Agilent, 103059-000, Santa Clara, CA, US) and incubate
   overnight at 37 °C without CO₂. On the experiment day, wash the cell
   culture plate per the manufacturer's instructions, then add analysis
   medium (Agilent, 103576-100, Santa Clara, CA, US) containing 1 mmol/L
   sodium pyruvate (Agilent, 103578-100), 2 mmol/L glutamine (Agilent,
   103579-100), and 10 mmol/L glucose (Agilent, 103579-100). Stabilize the
   cell plates for 1 h at 37 °C in a CO₂-free incubator. After baseline
   measurements, sequentially inject oligomycin (2 μM, Agilent,
   103015-100), FCCP (2 μM, Agilent, 103015-100), and rotenone/antimycin A
   (1.5 μM, Agilent, 103015-100) into each well. Measure and analyze OCR
   on the Agilent Seahorse XF Pro analyzer (US). Finally, calibrate OCR
   data by counting cells stained with the live-cell dye Hoechst (Thermo,
   62249, Carlsbad, CA, US).

2.12. mRNA sequencing (RNA-seq) analysis

   Primary podocytes were isolated from wild-type (n = 3) and
   podocyte-specific SIRPα knockout mice (n = 3). After 7 days of standard
   culture, total RNA was extracted using Trizol reagent (Thermo,
   15596018, Carlsbad, CA, US) according to the manufacturer's protocol.
   RNA quantity and purity were assessed using the Bioanalyzer 2100 and
   RNA 6000 Nano Lab Chip Kit (Agilent, 5067-1511, US). Only high-quality
   RNA samples with an RNA Integrity Number (RIN) greater than 7.0 were
   used for sequencing library construction. Gene expression analysis
   between the two groups was conducted using DESeq2 software, and genes
   with a false discovery rate (FDR) of less than 0.05 and an absolute
   fold change ≥2 were classified as differentially expressed
   ([79]https://www.ncbi.nlm.nih.gov/geo/, [80]GSE274089). Differentially
   expressed genes were then subjected to Kyoto Encyclopedia of Genes and
   Genomes (KEGG) pathway enrichment analysis. To investigate the
   signaling pathways regulated by SIRPα deficiency in podocytes, we
   selected pathways under the signal transduction category in the KEGG
   secondary classification. The top 8 pathways, ranked by enrichment
   scores, are presented in [81]Fig. 5A. Additionally, to ensure key
   molecules were not overlooked, genes with expression levels greater
   than 1 and a fold change ≥1.5 were selected for Gene Set Enrichment
   Analysis (GSEA). Oxidative stress, a biological process associated with
   mitochondrial damage, was chosen from the Gene Ontology (GO) enrichment
   set, and differentially expressed genes involved in oxidative stress
   were identified based on P < 0.05. Lastly, pyruvate metabolism-related
   genes enriched in KEGG were highlighted, and a heat map was generated
   ([82]Supplementary Fig. 4).

Fig. 5.

   [83]Fig. 5
   [84]Open in a new tab

   Deficiency of SIRPα results in formation a positive feedback loop
   between nuclear PKM2, NF-κB activation, and oxidative stress.

   (A) KEGG secondary classification of signal pathway, sorted by
   enrichment scores (fc > 1.5, P < 0.05). (B, C) Proteins levels of
   SIRPα, pP65, P65, and IL-6 were detected by Western blotting (n = 3).
   (D, E) Levels of Nox4 mRNA in podocytes (n = 3). (F–I) Podocytes
   isolated from CKO mice were treated with (KO-Shikonin) or without 1 μM
   shikonin (KO-Ctrl) for 24 hours (n = 3). PKM2 distribution (F, Scale
   bar, 15 μm), protein level of pP65, P65 and IL-6 (G), total ROS
   production (H), mRNA level of Nox4 (I) in podocytes were determined.
   (J–L) NOX4 inhibitor GLX351322 (GLX, 10 μM, 24 hours) and Mito-TEMPO
   (10 μM, 24 hours) were used to inhibit the production or clearance of
   ROS in podocytes. PKM2 distribution (J, Scale bar, 15 μm), total ROS
   production (K), and protein levels of pP65, P65 and IL-6 were evaluated
   (L). (M, N) The primary CKO podocytes were transfected with control
   siRNA (siCtrl) or IL-6 siRNA (siIL-6). The nuclear distribution of PKM2
   (M, Scale bar, 15 μm), and total ROS production were determined (n = 3)
   (N, Scale bar, 15 μm).

2.13. Statistical analysis

   Data were analyzed using GraphPad Prism 9.4. Differences between two
   groups were assessed using Student's t-test, while differences among
   multiple groups were evaluated using one-way or two-way ANOVA followed
   by Tukey's post hoc test. Each group of mice included six subjects, and
   all experiments were conducted at least three times. P value of <0.05
   (∗), <0.01 (∗∗), <0.001 (∗∗∗) and <0.0001 (∗∗∗∗) were considered
   statistically significant.

3. Results

3.1. Podocyte-specific deletion of SIRPα promotes podocyte injury in
experimental diabetic nephropathy (DN) mice

   SIRPα protein levels in podocytes progressively decrease with disease
   advancement in both DN patients and experimental DN mice
   ([85]Supplementary Figs. 1A and B). Similarly, SIRPα expression was
   found to gradually decline in podocytes exposed to serum from DN
   patients over time ([86]Supplementary Fig. 1C). When podocytes were
   treated with high glucose (HG) and various cytokines, including TGF-β,
   TNF-α, IL-6, IL-17, and IFN-γ, SIRPα levels were further reduced
   ([87]Supplementary Figs. 1D–F). To investigate the role of SIRPα under
   DN conditions, previously constructed homozygous SIRPα-flox knockout
   mice were crossed with NPHS2-Cre mice to generate podocyte-specific
   SIRPα knockout mice [[88]18] ([89]Supplementary Fig. 2A). The
   successful deletion of SIRPα in podocytes was confirmed
   ([90]Supplementary Figs. 2B–D), and a STZ-induced DN model was
   established ([91]Fig. 1A). Compared to their WT littermates, CKO mice
   exhibited an increased risk of podocyte injury, including elevated
   urinary protein levels ([92]Fig. 1B), glomerular hypertrophy ([93]Fig.
   1C), podocyte loss ([94]Fig. 1D), and mitochondrial damage ([95]Fig.
   1E). In streptozotocin (STZ) -induced diabetic nephropathy mice, the
   urinary albumin-to-creatinine ratio ([96]Fig. 1B) was significantly
   higher in the CKO-STZ group compared to the WT-STZ group, indicating
   worsened renal dysfunction. Pathological analysis revealed that
   podocyte-specific SIRPα knockdown exacerbated STZ-induced podocyte
   injury and accelerated kidney disease progression. In CKO-STZ mice,
   this was characterized by an enlarged glomerular area ([97]Fig. 1C),
   increased podocyte loss ([98]Fig. 1D), thickened glomerular basement
   membrane (GBM), widened podocyte foot processes, more extensive
   podocyte foot process effacement, and increased mitochondrial damage
   ([99]Fig. 1E). Due to the significant mitochondrial damage observed in
   podocytes, we further examined the mitochondrial ultrastructure and
   number in primary podocytes isolated from WT and CKO mice. The CKO
   podocytes exhibited shorter mitochondrial network length ([100]Fig.
   1F), mitochondrial swelling, and cristae disruption ([101]Fig. 1G), all
   indicative of more severe mitochondrial damage compared to WT
   podocytes. In summary, these findings underscore the critical role of
   SIRPα deficiency in promoting podocyte injury, with a particular
   emphasis on mitochondrial dysfunction.

Fig. 1.

   [102]Fig. 1
   [103]Open in a new tab

   Reduction of SIRPα in Podocytes of Diabetic Nephropathy.

   (A) Schematic representation of wild type (WT-STZ) and CKO mice
   (CKO-STZ) induced by streptozotocin (STZ). (B) Analysis of the urinary
   albumin-to-creatinine ratio in WT-STZ and CKO-STZ mice. (C, D) Periodic
   Acid-Schiff (PAS) staining and WT1 immunohistochemistry (scale bar,
   25 μm). Glomerular area was quantified based on PAS staining (C)
   (n = 6, 50 glomeruli randomly counted per mouse). Podocyte number was
   quantified based on WT1 immunohistochemistry results (D) (n = 6, 50
   glomeruli randomly counted per mouse). (E) Ultrastructural analysis of
   podocytes in WT-STZ and CKO-STZ mice. Glomerular basement membrane
   (GBM) thickness, podocyte foot process width (GBM length/number of foot
   processes), percentage of podocyte foot process fusion (foot process
   fusion length/GBM length), and damaged mitochondria in podocytes were
   quantitatively analyzed based on podocyte ultrastructure (n = 6, 3
   fields randomly selected per mouse). (F, G) Mitochondrial morphology in
   primary mouse podocytes (MitoTracker, scale bar, 15 μm) and
   mitochondrial ultrastructure in mouse podocytes. The mitochondrial
   network was reconstructed using ImageJ, and the mean mitochondrial
   branch length was measured (F) (n = 5). Mitochondrial ultrastructure
   was quantified, and mitochondrial area was analyzed at the same order
   of magnitude (G).

3.2. Podocyte-specific transgenic SIRPα alleviates podocyte injury in
experimental diabetic nephropathy (DN) mice

   We further investigated whether restoring SIRPα expression in podocytes
   could alleviate podocyte injury and slow the progression of
   experimental DN. To achieve this, podocyte-specific SIRPα-transgenic
   (TG) mice were generated using the nephrin promoter, and their
   successful construction was verified ([104]Supplementary Figs. 2E–H).
   TG mice exhibited a significantly lower urinary albumin/creatinine
   ratio compared to wild-type (WT) mice ([105]Supplementary Fig. 3A). To
   assess the protective effects of transgenic SIRPα under pathological
   conditions, STZ was administered. In STZ-induced TG mice (TG-STZ), the
   urinary albumin/creatinine ratio ([106]Fig. 2A), glomerular area
   ([107]Fig. 2B), podocyte loss ([108]Fig. 2C), glomerular basement
   membrane (GBM)thickness, foot process width, foot process fusion
   percentage, and the extent of mitochondrial damage ([109]Fig. 2D and E)
   were significantly reduced compared to STZ-induced WT mice (WT-STZ).
   There were no significant differences in baseline pathological
   histology and podocyte ultrastructure between WT and TG mice under
   non-pathogenic conditions ([110]Fig. 2B–E). There was no significant
   difference in the ultrastructure or quantity of mitochondria between
   primary podocytes from WT and TG mice ([111]Fig. 2F).

Fig. 2.

   [112]Fig. 2
   [113]Open in a new tab

   Podocyte-Specific Transgenic SIRPα protects against
   hyperglycemia-induced podocyte injury.

   (A) Urinary albumin-to-creatinine ratio in WT, WT-STZ, TG, and TG-STZ
   mice (n = 6). (B) PAS staining of renal paraffin sections (scale bar,
   25 μm). Glomerular area was quantified (n = 6, 50 glomeruli randomly
   counted per mouse). (C) Podocyte number was quantitatively analyzed
   based on WT1 immunohistochemistry (n = 6, 50 glomeruli randomly counted
   per mouse; scale bar, 25 μm). (D) Ultrastructural analysis of
   podocytes, including glomerular basement membrane (GBM) thickness,
   podocyte foot process width (GBM length/number of foot processes),
   percentage of podocyte foot process fusion (foot process fusion
   length/GBM length). (E) Damaged mitochondria in podocytes were
   quantitatively assessed (n = 6, 3 fields randomly selected per mouse).
   (F) Mitochondrial morphology in primary mouse podocytes (MitoTracker,
   scale bar, 15 μm) and mitochondrial ultrastructure. The mitochondrial
   network was reconstructed using ImageJ, and the mean mitochondrial
   branch length was measured (n = 5).

   To further elucidate the protective role of SIRPα in podocytes under DN
   conditions, we assessed the expression levels of podocyte-specific
   genes and proteins. In vitro, under stimulation with cytokines (TNF-α,
   IL-6, IL-17, IFN-γ, and TGF-β) and high glucose (HG), TG podocytes
   consistently exhibited stable expression of SIRPα ([114]Supplementary
   Figs. 3B and C). Moreover, the levels of nephrin and synaptopodin were
   significantly higher in TG podocytes compared to WT podocytes
   ([115]Supplementary Figs. 3D and E). In vivo, diabetic TG podocytes
   displayed sustained high expression of SIRPα ([116]Supplementary Figs.
   3F and G). WT-STZ podocytes showed a hyperglycemia-induced reduction in
   nephrin and synaptopodin levels, a decline that was notably mitigated
   in TG-STZ podocytes ([117]Supplementary Figs. 3F and G). Taken
   together, these findings suggest that transgenic expression of SIRPα in
   podocytes provides protection against podocyte injury under diabetic
   conditions.

3.3. SIRPα rescues high glucose-induced oxidative stress and increases
glycolytic flux in podocytes

   The findings mentioned above underscore the critical role of SIRPα in
   maintaining mitochondrial integrity and function. To further elucidate
   the underlying molecular mechanisms, we examined the role of SIRPα in
   these processes. Gene Set Enrichment Analysis (GSEA) of RNA sequencing
   data ([118]https://www.ncbi.nlm.nih.gov/geo/, [119]GSE274089) from
   primary podocytes of WT and CKO mice revealed that numerous genes
   associated with the oxidative stress response, linked to mitochondrial
   damage, were upregulated ([120]Fig. 3A). Genes exhibiting a fold change
   greater than 1.5 (P < 0.05) are detailed in [121]Supplementary Fig. 4A.
   Further validation confirmed that the expression of Il-6, Cd36, and
   Dgkk ([122]Fig. 3A), and the level of mito-ROS ([123]Supplementary Fig.
   5A) and total ROS ([124]Fig. 3B) were significantly upregulated, while
   mitochondrial respiratory capacity (Oxygen consumption rate, OCR) and
   ATP levels ([125]Fig. 3C) were markedly decreased in CKO podocytes
   compared to WT podocytes (WT, KO). These findings indicate pronounced
   mitochondrial damage and dysfunction in CKO podocytes. In the context
   of DN, mitochondrial dysfunction and impaired glycolytic flux are
   believed to play a pivotal role in disease progression
   [[126]29,[127]30], leading us to investigate the relationship between
   mitochondrial dysfunction and aerobic glycolysis. As depicted in
   [128]Supplementary Fig. 4B, genes associated with pyruvate metabolism
   were downregulated in CKO (KO) podocytes. Further validation
   demonstrated that the expression levels of Ppargc1a, Ldha, Pfkl, Pgk1,
   Pdk1, and Slc4a4 were significantly reduced ([129]Fig. 3D).
   Additionally, pyruvate kinase activity in CKO podocytes was found to be
   decreased compared to WT podocytes ([130]Fig. 3E). Thus, both
   mitochondrial respiration and glycolytic flux are disrupted in CKO
   podocytes.

Fig. 3.

   [131]Fig. 3
   [132]Open in a new tab

   SIRPα rescues high glucose-induced oxidative stress and increases
   glycolytic flux in podocytes.

   (A) Gene Set Enrichment Analysis (GSEA) enrichment profiles of
   oxidative stress (NES = 1.80, FDR<0.01, P < 0.001). RT-qPCR was used to
   verify the expression of related genes (n = 6). (B) The total ROS
   levels in primary podocytes from WT and CKO mice (n = 3). (C) The OCR
   and ATP levels in podocytes from WT and CKO mice (n = 6). (D) The mRNA
   levels of pyruvate metabolism-related genes were detected (n = 6). (E)
   The pyruvate kinase activity in podocytes from WT and CKO mice (n = 3).
   Mito-TEMPO (10 μM, 24 hours) was used to clearance of ROS. (F–H) Levels
   of mito-ROS (F), OCR and ATP (G), pyruvate kinase activity (H) in
   podocytes. (I–M) TG and WT podocytes were treated with 30 mM glucose
   for 48 h. Oxidative stress related genes (I), total ROS (J, n=3), OCR
   and ATP levels (K), pyruvate metabolism related genes (L, n = 6), and
   pyruvate kinase activity (M, n = 3) in HG-induced WT and TG podocytes
   were detected.

   ROS are both a product and an indicator of cellular oxidative
   stress[[133]31], [[134][31], [135][32]]. Consequently, we investigated
   the relationship between ROS production and the disruptions in
   glycolysis and mitochondrial respiration observed in CKO podocytes.
   Treatment with the mitochondrial-targeted antioxidant Mito-TEMPO
   (TEMPO) was used to scavenge mito-ROS, thereby reducing ROS levels
   ([136]Fig. 3F), resulted in a significant increase in oxygen
   consumption rate (OCR) and ATP levels in CKO podocytes ([137]Fig. 3G).
   Additionally, pyruvate kinase activity was also enhanced in CKO
   podocytes following TEMPO treatment ([138]Fig. 3H). These findings
   suggest that inhibiting mito-ROS can induce mitochondrial respiration
   and glycolytic flux in CKO podocytes.

   We subsequently explored the impact of transgenic SIRPα on
   mitochondrial respiration and glycolysis in podocytes under high
   glucose (HG) conditions. In HG-induced TG podocytes (TG + HG), the
   expression of oxidative stress-related genes, including Dgkk, Cd36,
   Sod1, and Il-6, was downregulated compared to HG-induced WT podocytes
   (WT + HG) ([139]Fig. 3I). Additionally, TG + HG podocytes exhibited
   reduced mito-ROS ([140]Supplementary Fig. 5B) and total ROS production
   ([141]Fig. 3J), accompanied by increased OCR and ATP levels ([142]Fig.
   3K), relative to WT + HG podocytes. Furthermore, glycolysis-related
   genes, such as Ppargc1a and Ldha, were significantly upregulated
   ([143]Fig. 3L). Pyruvate kinase activity (PK) was also elevated in
   TG + HG podocytes compared to WT + HG podocytes ([144]Fig. 3M).
   Collectively, these findings underscore the crucial role of SIRPα in
   maintaining mitochondrial respiration and glycolytic flux in podocytes
   under hyperglycemic conditions.

3.4. SIRPα deficiency facilitates PKM2 nuclear translocation by encouraging
PKM2 phosphorylation

   SIRPα exerts its regulatory effects by recruiting the protein tyrosine
   phosphatase SHP1 through its immunosuppressive ITIM motif, which
   modulates the phosphorylation of various downstream kinases [[145]33].
   To explore the molecular mechanisms by which SIRPα influences oxidative
   stress, mitochondrial respiration, and glycolytic flux, we constructed
   a FLAG-tagged SHP1 mutant plasmid by mutating the Cysteine 453 to
   serine 453. This mutation inhibited SHP1's dephosphorylation ability
   without impairing its capacity to bind substrates[[146]18] [[147][18],
   [148][34]]. Thus, overexpressing this mutant plasmid enriches
   downstream target molecules. Proteins bound to the mutant SHP1 were
   then isolated via FLAG antibody immunoprecipitation for mass
   spectrometry analysis. By integrating RNA-seq findings on oxidative
   stress and pyruvate metabolism pathways with mass spectrometry results
   ([149]Fig. 3), we identified that PKM2, ENO1, ACACB, HK2, and RANBP2
   proteins were associated with oxidative phosphorylation and pyruvate
   metabolism ([150]Supplementary Fig. 6). PKM2, a well-known
   rate-limiting enzyme catalyzing the final step of glycolysis, also
   possesses protein kinase activity. Phosphorylation of PKM2 at tyrosine
   105 regulates diverse biological processes, including cell inflammation
   and ROS production[[151]35] [[152][35], [153][36]]. Thus, PKM2 was
   selected as a candidate target protein of SIRPα for further validation.
   We subsequently confirmed the binding interaction between SIRPα, PKM2,
   and SHP1 ([154]Fig. 4A and B). Knockdown or overexpression of SIRPα
   resulted in increased or decreased phosphorylation of PKM2,
   respectively ([155]Fig. 4C and D). Stimulation with serum from DN
   patients led to a reduction in SIRPα expression while simultaneously
   increasing phosphorylated PKM2 in HPC ([156]Fig. 4E).
   Immunofluorescence staining of renal tissue sections further revealed a
   significant increase in nuclear pPKM2 in podocytes from DN patients
   ([157]Fig. 4F).

Fig. 4.

   [158]Fig. 4
   [159]Open in a new tab

   SIRPα deficiency facilitates PKM2 nuclear translocation by promoting
   PKM2 phosphorylation.

   (A) Co-immunoprecipitation of SHP1 with SIRPα and PKM2. (B)
   Co-immunoprecipitation of PKM2 with SHP1. (C) Phosphorylation of PKM2
   was induced in SIRPα knockdown podocytes (SIRPα siRNA). (D)
   Phosphorylation of PKM2 was inhibited in SIRPα overexpressing podocytes
   (OE-SIRPα). (E) Protein levels of SIRPα, pPKM2, PKM2 and podocin in HPC
   treated with serum from healthy persons and DN patients (n = 5). (F)
   The expression and distribution of SIRPα (red), and pPKM2 (green) were
   determined by immunofluorescence staining in renal tissues of DN
   patients. Scale bar, 25 μm. Three experiments were repeated
   independently. (G) The expression and distribution relationship between
   SIRPα and PKM2 were determined in primary podocytes from WT, TG, CKO
   mice with or without 30 mM HG treatment. Nucleus (blue, DAPI), F-actin
   (red), and PKM2 (green). Three experiments were repeated independently.
   (H) SIRPα, pPKM2, and PKM2 were measured in podocytes obtained from WT,
   TG, CKO, WT-STZ, TG-STZ, and CKO-STZ mice (Scale bar, 15 μm, n = 6).
   The level of pPKM2 was calculated using Image Pro Plus 6.0.

   To determine whether the nuclear translocation of PKM2 is regulated by
   SIRPα, we isolated podocytes from WT, CKO, and TG mice, and subjected
   them to HG stimulation. CKO podocytes exhibited higher basal levels of
   PKM2 nuclear translocation compared to WT podocytes ([160]Fig. 4G).
   Upon HG stimulation, both WT and CKO podocytes showed increased PKM2
   nuclear translocation, with the translocation in CKO podocytes being
   more pronounced than in WT podocytes ([161]Fig. 4G). In contrast, TG
   podocytes were resistant to HG-induced PKM2 nuclear translocation.
   Notably, under normal conditions, PKM2 was highly enriched in the foot
   processes of both WT and TG podocytes, while this enrichment was absent
   in CKO podocytes. Under HG conditions, the PKM2 enrichment in the foot
   processes of WT podocytes was lost; however, TG podocytes partially
   resisted this HG-induced effect. Similarly, phosphorylated PKM2 levels
   were elevated in the glomeruli of diabetic CKO mice compared to WT
   mice, whereas these levels were decreased in diabetic TG mice compared
   to WT mice ([162]Fig. 4H). Collectively, these findings demonstrate
   that hyperglycemia induces PKM2 phosphorylation and nuclear
   translocation in podocytes via SIRPα.

3.5. SIRPα deficiency facilitates formation of positive feedback loop between
PKM2 nuclear translocation, NF-κB activation, and oxidative stress

   To elucidate the signal transduction pathway underlying PKM2 nuclear
   translocation, we performed KEGG secondary classification analysis
   based on the RNA-seq data from WT and CKO podocytes. The analysis
   identified the NF-κB pathway as the top-ranked pathway according to
   enrichment score ([163]Fig. 5A). Further validation confirmed that
   NF-κB activation was significantly higher in CKO podocytes compared to
   WT podocytes ([164]Fig. 5B), while transgenic SIRPα effectively
   resisted HG-induced NF-κB activation ([165]Fig. 5C). In addition,
   proinflammatory genes Nlrp3, Casp1, Il1b, Il18 were also upregulated
   ([166]Supplementary Figs. 7A and B). As a co-transcription factor,
   NF-κB not only promotes the expression of various cytokines
   [[167]37,[168]38], but also upregulates the expression of several
   members of the NOX family, thereby enhancing oxidative stress and
   inflammatory responses [[169][39], [170][40], [171][41]]. Given that
   NADPH oxidase 4 (NOX4) is the most abundant NOX isoform in the kidney,
   we examined Nox4 expression and found it to be upregulated in CKO
   podocytes compared to WT podocytes ([172]Fig. 5D). Additionally, Nox4
   levels were lower in HG-induced TG podocytes compared to HG-induced WT
   podocytes ([173]Fig. 5E).

   Shikonin, a known PKM2 inhibitor, has demonstrated a preventive effect
   on kidney diseases; however, the underlying mechanism remains
   incompletely understood [[174][42], [175][43], [176][44]]. Larissa et
   al. demonstrated that shikonin can reverse the nuclear translocation of
   PKM2 and restore glycolytic flux in neuronal cells, thereby
   reestablishing the healthy characteristics of neurons [[177]45]. In
   this study, we found that shikonin inhibited PKM2 nuclear translocation
   ([178]Fig. 5F), NF-κB activation ([179]Fig. 5G), total ROS production
   ([180]Supplementary Fig. 7C and Fig. 5H), and Nox4 expression
   ([181]Fig. 5I), while it increased pyruvate kinase activity
   ([182]Supplementary Fig. 7D), mitochondrial respiration, and ATP
   production ([183]Supplementary Fig. 7E) in CKO podocytes. Similar
   effects were observed with the NOX4 inhibitor GLX351322 (GLX) and the
   superoxide scavenger Mito-TEMPO (TEMPO) ([184]Fig. 5J–L,
   [185]Supplementary Figs. 7F–J). Interestingly, the treatment of SIRPα
   KO podocytes with mito-TEMPO also resulted in the downregulation of
   IL-6 and NOX4 ([186]Fig. 5L, [187]Supplementary Fig. 7J. Additionally,
   interfering with IL-6 expression inhibited PKM2 nuclear translocation
   ([188]Supplementary Fig. 7K, Fig. 5M) and total ROS production
   ([189]Supplementary Fig. 7L and Fig. 5N), while enhanced pyruvate
   kinase activity ([190]Supplementary Fig. 7M) and reduced mito-ROS
   ([191]Supplementary Fig. 7N).

   We further validated the effects of PKM2 knockdown in primary podocytes
   with SIRPα knockout. As shown in [192]Supplementary Fig. 8, PKM2
   knockdown partially counteracted the activation of the NF-κB signaling
   pathway induced by SIRPα knockout ([193]Supplementary Figs. 8A–C), but
   it did not reduce overall ROS levels ([194]Supplementary Fig. 8D) or
   mitochondrial ROS production ([195]Supplementary Fig. 8E). The
   differing effects of PKM2 knockdown on ROS production and NF-κB
   signaling pathway activation may be due to a reduction in PKM2 levels
   in both the nucleus and cytoplasm. Notably, neither SIRPα knockout nor
   shikonin treatment affected PKM2 expression levels ([196]Fig. 4,
   [197]Fig. 6E); instead, they promoted the nuclear translocation of
   PKM2, resulting in decreased cytoplasmic PKM2 and increased nuclear
   PKM2. Taken together, these findings suggest the that there may be a
   positive feedback loop between PKM2 nuclear translocation, NF-κB
   activation, and ROS production, with SIRPα playing a regulatory role in
   this loop.

Fig. 6.

   [198]Fig. 6
   [199]Open in a new tab

   Shikonin alleviates hyperglycemia induced podocyte injury in CKO mice
   via inhibiting PKM2 nuclear translocation.

   (A) Schematic diagram of shikonin administration on STZ-induced DN
   experimental mice. WT and CKO mice were given STZ at 8 weeks of age for
   constructing experimental DN model, and shikonin was administrated
   every two days. After four weeks of administration, biochemical and
   pathological indicators of the mice were measured to evaluate the
   effect of shikonin on renal disease progression. (B) Levels of urinary
   albumin/creatinine (n = 6). (C) PAS staining of renal sections and
   glomerular area was counted (n = 6, and 50 glomeruli were counted
   randomly per mouse, Scale bar, 25 μm). (D) Immunohistochemical staining
   of WT1 to calculate the number of podocytes (n = 6, and 50 glomeruli
   were counted randomly per mouse, Scale bar, 25 μm). (E) The expression
   and nuclear distribution (blue, DAPI) relationship between SIRPα
   (purple), Nephrin (red), and PKM2 (green) were determined through
   immunofluorescence staining. Scale bar, 15 μm. (F) The protein levels
   of pP65, P65 and IL-6 in renal cortex of mice.

3.6. Shikonin alleviates hyperglycemia-induced podocyte injury in CKO mice
via inhibiting PKM2 nuclear translocation

   Finally, we investigated the in vivo effects of shikonin on alleviating
   podocyte injury caused by SIRPα deficiency. Shikonin was administered
   to STZ-induced WT and CKO mice every two days ([200]Fig. 6A). Notably,
   shikonin significantly mitigated the increase in urinary
   albumin-to-creatinine ratio ([201]Fig. 6B), glomerular area ([202]Fig.
   6C), and podocyte loss ([203]Fig. 6D) that were induced by SIRPα
   deletion. Importantly, shikonin markedly inhibited the nuclear
   translocation of PKM2 in podocytes ([204]Fig. 6E). Additionally,
   shikonin strongly suppressed NF-κB activation and IL-6 expression in
   the glomeruli of STZ-induced CKO mice ([205]Fig. 6F). These findings
   suggest that shikonin inhibits PKM2 nuclear translocation, NF-κB
   activation, and IL-6 expression, thereby protecting against podocyte
   damage associated with SIRPα deficiency.

4. Discussion

   This study demonstrated that SIRPα acts as a molecular integrator of
   mitochondrial dysfunction, inflammatory response, and aerobic
   glycolysis in podocytes through the regulation of PKM2 nuclear
   translocation. Oxidative phosphorylation (OXPHOS) and glycolysis are
   two key mechanisms by which cells convert glucose into energy. However,
   there is no consensus on the predominant metabolic pathway in
   podocytes. For instance, some studies indicate that kidneys are
   unaffected by the podocyte-specific deletion of genes involved in
   mitochondrial production and dynamics [[206]46]. Conversely, other
   studies have shown that mitochondrial respiration accounts for
   approximately 77 % of podocyte respiration[[207]47] [[208][47],
   [209][48]]. Consistent with our findings, Qi et al. [[210]25] reported
   that a reduction in podocyte-specific PKM2 leads to mitochondrial
   dysfunction and a decrease in glycolytic flux, whereas PKM2 enzyme
   activity enhancers reverse mitochondrial dysfunction and increase
   glycolytic flux in podocytes. These contradictions may result from the
   use of different experimental settings, animal models, cell types, or
   disease models.

   Some studies have hypothesized that OXPHOS provides energy for the
   cytoplasm of podocytes, while glycolysis supplies energy for the foot
   processes of these cells [[211]49]. Due to their distinct architecture,
   mitochondria in podocytes are unevenly distributed [[212]49]. This
   study also offers direct experimental evidence supporting this
   hypothesis: under physiological conditions, PKM2 is enriched in the
   foot processes of podocytes ([213]Fig. 4G); under HG conditions, SIRPα
   deletion promotes PKM2 nuclear translocation. Nuclear PKM2 can function
   as a transcriptional co-activator of glycolysis-related genes
   [[214][50], [215][51], [216][52]]. While we observed that SIRPα
   deletion-induced PKM2 nuclear translocation did not promote the
   expression of glycolytic genes, it inhibited their expression. This
   might be due to the fact that podocytes are terminally differentiated
   cells, and their pathological changes under DN conditions differ
   fundamentally from those of tumor cells. We also revealed that SIRPα
   regulates NF-κB activation by controlling PKM2 nuclear translocation.
   The activation of NF-κB by nuclear PKM2 has been reported in tumor
   cells and cancer-associated fibroblasts[[217]53] [[218][53],
   [219][54]]. However, the underlying mechanism remains unclear, although
   Gu et al. [[220]55] reported that nuclear PKM2 prolongs the retention
   of NF-κB p65 in the nucleus. Moreover, podocytes themselves share many
   innate and acquired immune features[[221]29] [[222][29], [223][56],
   [224][57]] [[225][29], [226][56], [227][57]]. Therefore, the decrease
   in podocyte SIRPα in DN might represent a critical link between
   metabolic imbalance and inflammation activation.

   It is well-established that the dimeric form of PKM2 has low enzymatic
   activity and cannot generate pyruvate at a normal rate, leading to an
   accumulation of upstream glycolytic intermediates [[228]30]. Thus,
   under DN conditions, SIRPα reduction-induced dimerization and nuclear
   translocation of PKM2, along with the excessive glucose utilization,
   may result in the accumulation of numerous by-products [[229]52].
   Multiple studies highlight the importance of PKM2 in regulating
   podocyte metabolism to maintain structure and function[[230]4]
   [[231][4], [232][30], [233][58]] [[234][4], [235][30], [236][58]], as
   well as the significance of PKM2's pyruvate kinase activity in DN
   [[237]59]. However, the regulatory mechanisms of PKM2 pyruvate kinase
   activity in DN remain unclear. Our findings indicate that the reduced
   enzymatic activity of PKM2 in podocytes under DN conditions is due to
   downregulated SIRPα expression, which promotes PKM2 nuclear
   translocation. In our study, both SIRPα knockout and shikonin treatment
   promoted PKM2 nuclear translocation without altering PKM2 expression
   levels, resulting in decreased cytoplasmic PKM2 and increased nuclear
   PKM2. This observation explains why PKM2 knockdown counteracted NF-κB
   pathway activation induced by SIRPα knockout without reducing overall
   or mitochondrial ROS production. Regarding ROS production, Yuan et al.
   reported that PKM2 knockdown in podocytes not only disrupts glycolysis
   but also impairs mitochondrial function, reducing ATP levels in
   podocytes [[238]60]. Nox4, a major ROS source in renal tissue, is
   primarily located in the mitochondria of the renal cortex [[239]61].
   Shanmugasundaram et al. demonstrated that ATP directly binds to NOX4,
   negatively regulating NOX4 activity and ROS production [[240]62]. Thus,
   PKM2 knockdown reduces ATP production in podocytes, relieving ATP's
   inhibitory effect on NOX4, thereby enhancing ROS production and
   offsetting the effects of decreased NOX4 expression.

   It has been reported that in macrophages, nuclear PKM2 recognizes the
   HIF-1α and IL-1β promoters, contributing to inflammation[[241]63]
   [[242][63], [243][64]]. Yang et al. found that lipopolysaccharides
   enhance PKM2 binding to the STAT3 promoter, thereby promoting TNF-α and
   IL-1β expression in colorectal cancer cells [[244]65]. These findings
   suggest that nuclear PKM2 plays a crucial role in activating
   inflammatory responses. Consistent with these studies, our results
   indicate that PKM2 nuclear translocation induced by SIRPα deficiency is
   a significant factor in podocyte inflammation in DN. Specifically, we
   found that SIRPα reduction promotes PKM2 phosphorylation and
   pathological nuclear translocation, which regulates the transcription
   and expression of downstream genes, including NOX4, IL-6, and IL-1β,
   via PKM2's protein kinase and co-transcriptional functions. NOX4,
   IL-1β, and IL-6 are also key players in cellular oxidative stress. For
   instance, IL-6 regulates mitophagy and impacts mitochondrial function
   [[245]66], while IL-1β can increase cellular ROS by upregulating NOX4
   expression [[246]67]. NOX4 is particularly associated with
   mitochondrial function and oxidative stress[[247]68] [[248][68],
   [249][69]]. Karen B. et al. demonstrated that NOX4 is primarily
   localized in renal cortex mitochondria in diabetic mice [[250]61],
   where it modulates cellular redox states. Elevated NOX4 expression
   leads to mitochondrial dysfunction and increased mitophagy [[251]70].
   Our findings suggest that high NOX4 expression in diabetic podocytes is
   a major contributor to ROS production, promoting inflammatory
   activation and oxidative stress in podocytes. This aligns with reported
   roles of NOX4, IL-6, and IL-1β in DN progression [[252][71], [253][72],
   [254][73], [255][74]].

   Targeting PKM2 enzyme activity has been shown to regulate renal
   function [[256]30,[257]58,[258]75]. Interestingly, both the PKM2
   inhibitor shikonin [[259]76,[260]77] and the PKM2 activator TEPP-46
   [[261]78] exhibit protective effects on the kidneys. This may be
   because these regulators have functions beyond directly modulating PKM2
   enzyme activity, or their regulatory roles may differ depending on cell
   type and environmental conditions. Shikonin, a natural drug extract,
   exhibits anti-inflammatory and antioxidant properties by modulating
   multiple molecular pathways. For instance, it selectively inhibits the
   activation of NLRP3 and AIM2 inflammasomes[[262]79] [[263][79],
   [264][80]], reduces oxidative stress and inflammation in tubular cells
   by downregulating the NOX4/PTEN pathway, and decreases the expression
   of downstream pro-inflammatory cytokines IL-1β and IL-6 [[265]81].
   Additionally, shikonin can inhibit chloride channel proteins [[266]20],
   potentially interfering with chloride-related IL-1β production and
   secretion pathways, which subsequently suppresses IL-6 expression
   [[267]82]. Moreover, it prevents the nuclear translocation of PKM2,
   enhances glycolysis, and promotes PKM2 tetramer formation in neuronal
   cells [[268]45]. We found that shikonin reverses PKM2 nuclear
   translocation, thereby restoring mitochondrial function and promoting
   podocyte glycolysis. Our study identified a novel function of shikonin,
   distinguishing it from a mere enzyme activity inhibitor: it regulates
   the subcellular localization of PKM2 in podocytes. Whether shikonin
   directly modulates PKM2 enzyme activity or exerts its effects through
   regulating PKM2 nuclear translocation may depend on the specific cell
   type and extracellular environment. Therefore, our study offers new
   insights into the anti-inflammatory mechanisms of shikonin in DN.

   In summary, reduced SIRPα levels in DN podocytes stimulate PKM2
   phosphorylation, which subsequently forms a positive feedback loop
   involving PKM2 nuclear translocation, NF-κB activation, and oxidative
   stress. This loop promotes mitochondrial dysfunction and inhibits
   glycolytic flux. Shikonin disrupts the PKM2 nuclear translocation
   induced by SIRPα deficiency. SIRPα serves as a crucial molecule linking
   oxidative stress, inflammatory activation, and glycolysis in podocytes,
   and it may represent a potential therapeutic target for modulating PKM2
   nuclear translocation.

CRediT authorship contribution statement

   Yang Chen: Writing – original draft, Visualization, Validation, Project
   administration, Methodology, Investigation, Formal analysis, Data
   curation, Conceptualization. Mingchao Zhang: Visualization, Validation,
   Investigation, Formal analysis, Data curation. Ruoyu Jia: Validation,
   Data curation. Bin Qian: Visualization, Formal analysis. Chenyang Jing:
   Data curation. Caihong Zeng: Resources. Dihan Zhu: Writing – review &
   editing. Zhihong Liu: Writing – review & editing, Supervision. Ke Zen:
   Writing – review & editing, Supervision, Conceptualization. Limin Li:
   Writing – original draft, Investigation, Funding acquisition, Data
   curation, Conceptualization.

Funding

   This work was supported by grants from the National Natural Science
   Foundation of China (32170897).

Declaration of competing interest

   The authors declare the following financial interests/personal
   relationships which may be considered as potential competing interests:
   Limin Li reports financial support was provided by the National Natural
   Science Foundation of China (32170897). Limin Li reports a relationship
   with the National Natural Science Foundation of China (32170897) that
   includes: funding grants. If there are other authors, they declare that
   they have no known competing financial interests or personal
   relationships that could have appeared to influence the work reported
   in this paper.

Footnotes

   ^Appendix A

   Supplementary data to this article can be found online at
   [269]https://doi.org/10.1016/j.redox.2024.103439.

Contributor Information

   Zhihong Liu, Email: liuzhihong@nju.edu.cn.

   Ke Zen, Email: kzen@nju.edu.cn.

   Limin Li, Email: liminli@cpu.edu.cn.

Appendix A. Supplementary data

   The following are the Supplementary data to this article.
   Multimedia component 1
   [270]mmc1.pdf^ (2MB, pdf)
   Multimedia component 2
   [271]mmc2.docx^ (5.3MB, docx)

Data availability

   Data will be made available on request.

References