All animal care and experimental procedures were conducted in compliance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The protocol (Animal protocol # 2011-0401 and # 2018-0971) followed was approved by the Institutional Animal Care and Use Committee at Augusta University. Prior to, during, and after the experiment, welfare assessments, measurements, and interventions were performed. The mice were housed under controlled environmental conditions, including a temperature range of 21-25 °C, humidity between 40-60%, and a 12-hour light/dark cycle. All animals had ad libitum access to food (Teklad global 18% protein rodent diet; 2918-060622M, Envigo, Madison, WI, USA) and water. Floxed Pfkfb3 (Pfkfb3 ^flox/flox) mice were generated by Xenogen Biosciences Corporation (Cranbury, NJ, USA) (44). To achieve cell-specific deficiency of Pfkfb3 in macrophages, Pfkfb3 ^flox/flox mice were crossbred with Lysm-Cre transgenic mice (The Jackson Laboratory, stock no. 004781, Bar Harbor, ME, USA) to generate Pfkfb3 ^flox/flox Lysm-Cre (Pfkfb3 ^ΔMφ) mice, with Pfkfb3 ^WT/WT Lysm-Cre (Pfkfb3 ^WT) mice used as wild type control. All mice were genotyped by PCR amplification of tail-clip samples ( Supplementary Table 1 ).

As described previously (41), Adult male mice weighing between 20-25 grams were used for this study. Under anesthesia induced by i.p. injection of ketamine (100 mg/kg) and xylazine (10 mg/kg), a midline abdominal incision was made to expose the left ureter. A double-ligature technique was utilized to create a complete obstruction by tying a suture around the left ureter at two distinct points, with a 1-mm interval between the ligatures. The ureter was then gently compressed between the ligatures and carefully examined to ensure complete occlusion. The mice were allowed to recover under controlled temperature and humidity conditions. At designated time points, the mice were euthanized using i.p. injection of ketamine (100 mg/kg) and xylazine (10 mg/kg) to collect kidney samples for analysis. The kidneys were carefully dissected and processed for histological examination and gene expression studies. The contralateral kidneys without ligation were used as control.

After euthanizing the mice, femurs and tibias were collected and transected. The bone marrow cells were flushed out from the femurs and tibias. These cells were then passed through a 70 μm cell strainer and centrifuged at 1000 x g for 3 minutes to obtain a single-cell suspension. The collected cells were plated onto a 10 cm² dish and allowed to incubate for 6-8 hours to discard the attached cells. The remaining suspension cells were collected and reseeded into a 6-well plate at a density of 2 x 10⁶ cells/mL. Cells were cultured in RPMI 1640 medium (SH30027.01, Cytiva, Marlborough, MA, USA) supplemented with 10% FBS (F4135, Sigma-Aldrich, St. Louis, MO, USA), 10 ng/mL MCSF (315-02, PeproTech, Cranbury, NJ, USA) and 1% Antibiotic-Antimycotic (15240062, Thermo Scientific, Grand Island, NY, USA) in a humidified incubator with 5% CO2 at 37°C for 7 days to induce macrophage differentiation as described before (45). After 7 days, the differentiated cells were cultured in RPMI 1640 medium supplemented with 1% FBS, 1% Antibiotic-Antimycotic, and 10 ng/mL MCSF. Subsequently, the cells were treated with or without 10 ng/mL mouse TGFβ1 (7666-MB-005, R&D SYSTEMS, Minneapolis, MN, USA) to induce macrophage differentiation for 5 days.

To extract total RNA from cells or tissues, the Trizol Reagent (15596018, Invitrogen, Grand Island, NY) was employed following the manufacturer’s protocol, as previously described (46). For cDNA synthesis, the iScriptTM cDNA synthesis kit (1708891, Bio-Rad Hercules, CA, USA) was utilized. RT-qPCR was performed on a StepOne Plus System (Applied Biosystems, Grand Island, NY) using the Power SYBR Green Master Mix (1725122, Bio-Rad Hercules). The relative gene expression was quantified using the efficiency-corrected 2^−△△CT method, with 18S ribosomal RNA serving as the internal control. The obtained data are presented as fold change relative to the control groups. Supplementary Table 2 includes the list of gene-specific primers employed in our study.

Protein extracts were obtained from kidney tissues and cells by lysing them in RIPA buffer supplemented with protease inhibitor cocktails (05892970001, Roche, SC, USA). The kidney tissues were ground and then lysed to prepare the tissue lysates. The extracts were centrifuged at 12,000 rpm for 10 minutes, and the resulting supernatant was collected. The protein concentration in the supernatant was determined using the BCA assay. Subsequently, the samples were subjected to SDS-PAGE and transferred onto nitrocellulose membrane. After blocking with 5% skim milk for one hour, the blots were probed with the following primary antibodies: PFKFB3 (diluted 1:1000, ab181861, abcam), ACTA2 (diluted 1:1000, sc-56499, Santa Cruz), Vimentin (diluted 1:1000, CST5741, Cell Signaling Technology), Collagen I (diluted 1:1000, NB600-408, NOVUS), Collagen IV (diluted 1:1000, ab6586, abcam), Fibronectin (diluted 1:1000, ab2413, abcam), Cyclophilin B (diluted 1:1000, CST43603, Cell Signaling Technology), HIF1A (diluted 1:500, AF1935-SP, R&D systems). Anti-β-actin (diluted 1:1000, sc47776, Santa Cruz) and GAPDH (diluted 1:1000, 2118, CST) were used as loading controls. The antibodies used are provided in Supplementary Table 3 . The western blots were quantified with a method described previously (47).

Cells were cultured in 8-Chambered Cell Culture Slides (08-774-26, Fisher Scientific). Frozen or paraffin blocks were sectioned at 7 μm thickness using a Microm cryostat or paraffin microtome. For cells or frozen sections, a PBS wash was performed, followed by fixation with 4% paraformaldehyde for 15 minutes. Subsequently, permeabilization was achieved by treating the cells or slides with 0.5% Triton X-100 for 20 minutes at room temperature. In the case of paraffin sections, slides were subjected to deparaffinization and rehydration before permeabilization. For antigen retrieval, slides were heated in citrate acid buffer (10 mM, pH 6.0) by microwaving at 98°C for 10 minutes. Following these steps, cells or sections were blocked with 10% normal goat serum (50062Z, Thermo Scientific) for 1 hour at room temperature. Primary antibodies against PFKFB3, ACTA2, F4/80, Arg1, IL1β, or COL1 were then added and incubated overnight at 4°C in a humidified chamber. After washing, samples were incubated with Alexa Fluor 488, 594, or 647-conjugated secondary antibodies for 1 hour at room temperature. Nuclei were counterstained with DAPI for 10 minutes at room temperature. The imaging process was carried out using an inverted confocal microscope (Zeiss 780, Carl Zeiss). The images were quantified with a method described previously (35). Four images/section/kidney were used for statistic purpose. The primary antibodies used in our study can be found in Supplementary Table 3 .

Paraffin-embedded kidneys were sectioned and underwent deparaffinization using xylene, followed by rehydration with a gradient of ethanol and water solutions. The activity of endogenous peroxidase in the tissue was eliminated by a treatment with methanol containing 30% H₂O₂ for 30 minutes at room temperature. Antigen retrieval was performed by exposing the sections to 10 mM sodium citrate buffer (pH 6.0) at 98°C for 10 minutes, followed by blocking with avidin blocking solution for 1 hour at room temperature. The antibodies were mixed with biotin blocking solution as per the manufacturer’s instructions and applied to the slides, which were then incubated overnight at 4°C. Subsequently, the tissue sections were treated with a biotinylated secondary antibody for 1 hour at room temperature, followed by incubation with ABC solution (PK-6100, Vector Laboratories) for 30 minutes at room temperature. The antibody was visualized using the peroxidase substrate 3, 3’-diaminobenzidine (3468, Dako, Santa Clara, CA, USA). Hematoxylin I (GHS116, Sigma) was used for counterstaining of the kidney sections. Finally, the slides were dehydrated and mounted using a xylene-based mounting medium (8312-4, Richard-Allan Scientific). The image data was quantified with the method described previously (48), using Image J Software (National Institutes of Health, USA, http://imagej.nih.gov/ij). Four images/section/kidney were used for statistic purpose. The primary antibodies used in our study can be found in Supplementary Table 3 .

We performed H & E staining on 7 μm sections of kidney with hematoxylin (22050111, Thermo Scientific) and eosin (220501110, Thermo Scientific) following the manufacturer’s standard protocol. Collagen deposition was visualized using Masson’s Trichrome staining kit (25088-1, Polysciences, PA, USA) and Sirius Red staining kit (ab150681, Abcam, CA, USA) according to the manufacturer’s instructions. Four images of were randomly taken in cortex/outer medulla region per section for each kidney, captured with a Keyence Microscope BZ-X800, and the percentage of collagen deposition area was quantified with Image J Software (National Institutes of Health, USA, http://imagej.nih.gov/ij).

In the experiment, biological triplicate 10 cm² dishes were utilized to cultivate BMDMs in the presence of complete cell medium. Following the specific treatment, metabolites were extracted using 1 mL of ice-cold 80% methanol on dry ice. Subsequently, the samples underwent centrifugation at 14000 rpm for a duration of 5 minutes. To ensure thorough extraction, the cell pellet was subjected to an additional step involving the use of 0.5 mL of 80% methanol. For accurate protein quantitation, the cell pellet was dissolved in an 8 M urea solution. To facilitate convenient shipment or storage, the supernatant obtained from the metabolite extraction was desiccated into a pellet using a SpeedVac from Eppendorf (Hamburg, Germany), employing a heat-free technique. When it was time for analysis, the dried pellets were re-suspended in 20 μL of HPLC grade water in preparation for mass spectrometry as described before (49). A volume of 5-7 μL of the resulting resuspension was injected and subjected to analysis using a cutting-edge hybrid 6500 QTRAP triple quadrupole mass spectrometer from AB/SCIEX (MA, USA), which was coupled to a Prominence UFLC HPLC system from Shimadzu (Kyoto, Japan). The analysis was carried out through selected reaction monitoring (SRM), targeting a comprehensive set of 298 endogenous water-soluble metabolites, enabling a thorough examination of the steady-state characteristics of the samples. The original data has been uploaded to MetaboLights (MTBLS8278) for public data sharing.

The levels of mouse cytokines TNFα and IL1β were measured with ELISA kits (MTA00B and MLB00C, R&D, Minneapolis, MN, USA) according to the manufacturer’s instructions.

Statistical analysis was conducted using GraphPad Prism Software (Version 9.0, RRID: SCR_000306). The significance of differences between two groups was evaluated using the unpaired Student’s t-test or unpaired two-tailed Student’s t-test with Welch’s correction (with unequal variances). Multiple comparisons were performed through one-way ANOVA followed by Bonferroni post hoc test. All results are presented as mean ± SEM. A significance level of P < 0.05 was considered statistically significant. To calculate the Z-score, we used the formula below: Z-score = (x - μ)/σ, x is the value of the data point, μ is the mean of the sample or data set, σ is the standard deviation. Each biological experiment was repeated a minimum of three times, utilizing independent cell cultures or individual animals as biological replications.

In a study by Conway et al., myeloid populations crucial for renal fibrosis in the UUO kidney were characterized using single-cell RNA sequencing (scRNAseq) (50). Among these myeloid populations, the significance of Ly6c2⁺ and Arg1⁺ myeloid cells in renal fibrosis were observed (50). To gain further insights of the early events in kidney obstruction, we conducted a reanalysis of the inflammation, TGFβ signaling, and glycolysis related mRNA expression in these myeloid cells and compared them to resident macrophages in UUO 2-day condition. Our findings revealed substantially higher levels of mRNA expression in inflammation, TGFβ signaling, and glycolysis pathways in Ly6c⁺ and Arg1⁺ myeloid cells than those of resident macrophages ( Figure 1A , Supplementary Figure 1 ). Consistent with these altered pathways, the corresponding genes involved in these pathways were upregulated in Ly6C⁺ and Arg1⁺ myeloid cells compared to resident macrophages. Notably, these genes included Il1, Tgfb1, and Hif1 ( Figure 1B , Supplementary Figure 2 ). Intriguingly, Ly6C⁺ and Arg1⁺ myeloid cells also exhibited elevated levels of Fn1 and Vim in comparison to resident macrophages ( Figure 1B ), suggesting a potential role of myeloid glycolysis in renal fibrosis initiation.

To investigate the role of myeloid PFKFB3 in renal fibrosis, we generated myeloid-specific Pfkfb3-deficient mice (Pfkfb3 ^ΔMϕ ) and their controls (Pfkfb3 ^WT) by breeding floxed Pfkfb3 mice with Lysm-Cre mice ( Supplementary Figures 3A, B ). Deletion of Pfkfb3 in myeloid cells was confirmed through Real time PCR, Western blot analysis and immunostaining of PFKFB3 in Bone marrow derived macrophages ( Supplementary Figures 3C-E ). We conducted unilateral ureter obstruction (UUO) surgery in these mice ( Figure 2A ). The change in mouse body weight following the UUO procedure did not show a significant difference between the two groups ( Figure 2B ). After 14 days, the weight of obstructive kidneys was significantly reduced compared to that of control kidneys, indicating obvious kidney atrophy ( Figures 2C, D , Supplementary Figure 4 ). And hydronephrosis was observed at kidney harvesting, indicating the success of ureter obstruction. Meanwhile, there was no significant difference of kidney weight between groups of Pfkfb3 ^ΔMϕ and Pfkfb3 ^WT mice ( Figures 2C, D ). The co-staining of PFKFB3 and F4/80 confirmed the induction of PFKFB3 in Pfkfb3 ^WT macrophages and the depletion of Pfkfb3 in Pfkfb3 ^ΔMϕ macrophages in kidneys after UUO injury ( Supplementary Figure 5 ). Using qRT-PCR, we examined the mRNA levels of multiple fibrotic factors. The expression of Acta2, Col1 and 3, Mmp2 and 9, and Tgfb were significantly increased in UUO kidneys compared to control kidneys ( Figure 3A ). Notably, the upregulation of these fibrotic factors in UUO kidneys of Pfkfb3 ^ΔMϕ mice was markedly reduced compared to Pfkfb3 ^WT mice ( Figure 3A ). We also assessed the protein levels of a few fibrotic factors through Western blot analysis ( Figure 3B ). The levels of α-smooth muscle actin (ACTA2), vimentin (VIM), collagen (COL) I and IV and fibronectin (FN) were significantly elevated in UUO kidneys compared to control kidneys. However, the increased levels of these fibrotic factors in UUO kidneys of Pfkfb3 ^ΔMϕ mice were significantly reduced compared to Pfkfb3 ^WT mice ( Figure 3B ).

Furthermore, we verified the interstitial renal fibrosis by staining renal sections of UUO and control kidneys with Sirius Red and Masson’s trichrome. In the UUO kidneys from Pfkfb3 ^WT mice, characterized by pronounced tubular dilation and thinning on hematoxylin and eosin (HE) stained sections, there was evident collagen deposition with Masson’s trichrome and Sirius Red staining, indicating a significant interstitial renal fibrosis ( Figures 4A, B , Supplementary Figure 4 ). In contrast, there were fewer tubular destruction and less collagen deposition shown by Masson’s trichrome and Sirius Red staining in the sections of UUO kidneys from Pfkfb3 ^ΔMϕ mice compared to Pfkfb3 ^WT mice ( Figures 4A, B ).

We also performed immunostaining to examine the presence of specific myofibroblast marker and components of extracellular matrix. As shown in Figure 4C , the levels of ACTA2, Collagen I and IV, which were low but detectable in the control kidneys from both groups of mice, were remarkably elevated in the UUO kidneys of Pfkfb3 ^WT mice. However, these increased levels of myofibroblast marker and extracellular matrix proteins were significantly reduced in the UUO kidneys from Pfkfb3 ^ΔMϕ mice ( Figure 4C ).

We investigated the infiltration of macrophages in the UUO kidney of Pfkfb3 ^ΔMϕ mice by performing immunostaining on renal sections using the F4/80 antibody. The presence of F4/80 positive cells was not noticeable in control kidneys from both Pfkfb3 ^WT and Pfkfb3 ^ΔMϕ mice ( Figures 5A, B ). However, an obvious increase of F4/80-positive cells was observed in the UUO Pfkfb3 ^WT kidneys ( Figures 5A, B ). Remarkably, the F4/80 positive area was significantly reduced in the UUO Pfkfb3 ^ΔMϕ kidney sections compared to the UUO Pfkfb3 ^WT kidney ( Figures 5A, B ). These findings strongly suggest that myeloid Pfkfb3 deficiency markedly inhibits the infiltration of macrophages into the UUO kidneys.

To assess the impact of myeloid Pfkfb3 deficiency on the infiltration of M1 and M2 macrophages in the UUO kidneys, we examined the mRNA and protein levels of M1/M2 markers and cytokines associated with M1/M2 macrophages. As shown in Figure 6A , the expression levels of M2 markers Arg1, Cd206, and M1 markers Cd80 were detectable but similar in the control kidneys of both groups of mice. However, in the UUO Pfkfb3 ^WT kidney, there was a substantial upregulation of both M2 and M1 markers ( Figure 6A ). In contrast, these markers were significantly reduced in the UUO Pfkfb3 ^ΔMϕ kidney compared to the UUO Pfkfb3 ^WT kidney ( Figure 6A ). Consistent with the levels of M1/M2 markers, the levels of cytokines, including Il10, Mgl2, Retnla, Il6, Mcp1, Tnfa, Il1b, Nos2, and Cxcl10, were significantly lower in the UUO Pfkfb3 ^ΔMϕ kidney compared to the UUO Pfkfb3 ^WT kidney ( Figure 6A ). To validate the expression of some genes at protein levels, we examined the expression of Arg1 and Il1β on F4/80 positive cells by immunostaining in renal sections with their specific antibodies. The staining intensity of Arg1 and Il1β on F4/80-positive cells were much lower in sections of the UUO Pfkfb3 ^ΔMϕ kidney compared to the UUO Pfkfb3 ^WT kidney ( Figure 6B ). These findings indicate that Pfkfb3 deficiency leads to a significant reduction in the number of M1 and M2 macrophages, as well as the levels of cytokines associated with these macrophage phenotypes in UUO kidneys.

To investigate whether myeloid Pfkfb3 deficiency affected macrophage differentiation to obtain myofibroblast phenotype in the UUO kidney, we quantified the number of macrophages expressing ACTA2 through co-immunostaining of the renal sections. In the control kidneys, the presence of ACTA2 and F4/80 positive cells was rare ( Figures 7A, B ). Conversely, the UUO Pfkfb3 ^WT kidneys exhibited a significant induction of F4/80 positive cells, with about 6% of them co-stained with ACTA2 ( Figures 7A, B ), suggesting their transition to myofibroblasts. Remarkably, the UUO Pfkfb3 ^ΔMϕ kidneys showed a substantial reduction of F4/80 positive area as well as the double-positive area of F4/80 and ACTA2 compared to the UUO Pfkfb3 ^WT kidney ( Figures 7A, B ). Further analysis on the lineage of myofibroblasts revealed that, in the UUO Pfkfb3WT kidneys, about 32% of myofibroblasts (ACTA2 positive) expressed F4/80. This percentage dropped to approximately 15% in the UUO Pfkfb3ΔMϕ kidneys ( Figure 7B ). These findings indicate that myeloid-specific Pfkfb3 deficiency leads to a decrease of differentiated macrophage with myofibroblast characteristics in the UUO kidney.

After observing the phenotypic changes of M1 and M2 markers, cytokines, and myofibroblast maker in the UUO Pfkfb3 ^ΔMϕ kidney, we further examined whether Pfkfb3 deficiency could induce similar changes in macrophages in vitro. We first isolated bone marrow cells from Pfkfb3 ^WT and Pfkfb3 ^ΔMϕ mice and cultured bone marrow-derived macrophages (BMDMs). Upon TGFβ1 treatment, the expression of Pfkfb3 at both mRNA and protein levels was enhanced in BMDMs from Pfkfb3 ^WT mice ( Figures 8A, B ). Subsequently, TGFβ1 incubation significantly upregulated the mRNA levels of M1 and M2 markers and associated cytokines, such as Tnfa and Arg1 ( Figure 8C ). In contrast, Pfkfb3 deficient BMDMs exhibited significantly lower levels of these mRNAs compared to WT BMDMs with TGFβ1 ( Figure 8C ). Furthermore, TGFβ1 treatment resulted in the differentiation of BMDMs, as evidenced by the expression of Acta2 at both mRNA and protein levels in the TGFβ1-treated group but not in the vehicle-treated group, and it was associated with Col1a1 mRNA change ( Figures 8D-G ). Remarkably, TGFβ1-induced Acta2 expression was nearly abolished in Pfkfb3 deficient BMDMs ( Figures 8D-G ). We subsequently analyzed the release of pro-inflammatory cytokines, TNFα and IL1β, by the BMDMs into the culture medium. As depicted in Figure 8H , TGFβ1 triggered a substantial release of these cytokines, which was markedly attenuated by the PFKFB3 knockout.

We assessed glucose metabolism in cultured WT and Pfkfb3-deficient BMDMs using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Pfkfb3-deficient BMDMs displayed a modest decrease in the levels of some glycolytic metabolites compared with vehicle-treated WT BMDMs ( Figure 9A ). Upon TGFβ1 treatment, WT BMDMs exhibited a substantial increase in glycolytic metabolites, whereas Pfkfb3-deficient BMDMs failed to show similar increases ( Figure 9A ). Most importantly, the increase of the glycolysis end-product lactate relied on PFKFB3 induction ( Figure 9A ). Additionally, we observed the PFKFB3-dependent glycolysis level change after TGFβ1 by Seahorse analysis of ECAR (data not shown).

Since glycolytic metabolites are known to stabilize HIFs and modulate the phenotypic change of vascular cells (51), we investigated the protein level of HIF1α in our cultured BMDMs. The level of HIF1α was comparable between vehicle treated WT and Pfkfb3-deficient BMDMs ( Figure 9B ). However, TGFβ1 treatment significantly elevated the level of HIF1α in WT BMDMs ( Figure 9B ), while this upregulation was compromised in Pfkfb3-deficient BMDMs ( Figure 9B ). To determine whether this decreased HIF1α in TGFβ1-treated Pfkfb3-deficient BMDMs is responsible for the decreased expression of M1 and M2 markers, cytokines and Acta2, we transduced Pfkfb3-deficient BMDMs with an adenovirus-packaged non-degradable mutant HIF1α, aiming to restore the HIF1α level in these cells ( Figure 9C ). We measured mRNA levels of M1, M2 macrophage markers, Acta2, and cytokines with qPCR and observed that the decreased expression of these molecules in Pfkfb3-deficient BMDMs was enhanced following HIF1α overexpression ( Figure 9D ). Furthermore, it correlated with the cytokine release changes in the culture medium ( Figure 9E ). The ECAR measurement also indicated that HIF1α restoration in Pfkffb3-deficent BMDMs partially reversed the glycolysis level (data not shown). These findings indicate that HIF1α, at least partially, participates in PFKFB3-mediated macrophage phenotypic change.