Mice were used in accordance with the protocol approved by the Institutional Animal Care and Use Committee of Augusta University. The floxed Pfkfb3 (Pfkfb3^flox/flox) mice were generated by Xenogen Biosciences Corporation (Cranbury, NJ, USA) (14). Myeloid-specific Pfkfb3 knockout was achieved by cross-breeding Pfkfb3^flox/flox mice with Lysm-Cre mice (stock no. 004781, The Jackson Laboratory, Bar Harbor, ME, USA) to generate Pfkfb3^flox/flox; Lysm-Cre (Pfkfb3^ΔMϕ) mice. All mice were on a C57BL/6J background.

For LPS-induced sepsis, age (8- to 12-week-old)- and sex-matched Pfkfb3^ΔMϕ and Pfkfb3^WT mice were administered with LPS (L2630, Sigma-Aldrich, St. Louis, MO, USA) at 12.5 mg/kg body weight by intraperitoneal injection. The mice had normal access to food and water and were monitored twice a day over the course of 10 days. The rectal temperature of mice was measured with a model BAT-12 digital thermocouple thermometer (Physitemp Instruments, Clifton, NJ, USA) before and 6 h, 24 h after LPS injection.

Mouse blood pressure (BP) was measured at the same time of day using a noninvasive tail-cuff blood pressure measurement system (Coda 6, Kent Scientific, Torrington, CT, USA). In brief, mice were trained for BP measurement conditions on a daily basis for 1 week. After training, mouse BP was measured twice before the injection of LPS and at indicated times following LPS injection. For BP measurements, we placed conscious mice in tail-cuff restrainers over a warmed surface. Twenty consecutive BP measurements were taken, and the last fifteen readings per mouse were averaged and used for analysis.

Echocardiograms were performed in isoflurane-anesthetized mice using the Vevo 2100 high-frequency ultrasound imaging platform (FUJIFILM VisualSonics, Toronto, Ontario, Canada) at baseline and 6 h after LPS injection. After sedation, mice were laid on the heating platform (37 °C) to maintain normothermia and continuously delivered a gas inhalation of two percent isoflurane. The prewarmed ultrasonic gel (4238, Chattanooga, Vista, CA, USA) was added to the mouse's chest after removal of hair with depilatory cream. Two-dimensional echocardiography was performed with a 40 MHz ultrasound probe (MS-400). The percentage of left ventricular (LV) ejection fraction and fractional shortening, LV stroke volume and cardiac output were obtained from the parasternal short axis view using M-mode.

Lung wet-to-dry weight ratio was used as an index of pulmonary edema formation that served as a gauge for measuring pulmonary permeability. The lung was weighed immediately after its excision (wet weight), and then placed into an oven at 60 °C for 48 h, and reweighed as dry weight. The ratio of the lung weight before and after drying was calculated.

Pulmonary permeability was also evaluated with Evans Blue dye. Briefly, LPS (1 mg/kg body weight) or saline was intratracheally injected to mice via a 20-gauge catheter. In order to assess vascular leak, Evans blue (30 mg/kg body weight, E2129, Sigma-Aldrich, St. Louis, MO, USA), was injected to mice via the tail vein 2 h before mice were sacrificed. The lungs were perfused with PBS, then dried with tissue papers, and left lung were imaged.

After euthanization of mice, femurs and tibias were isolated and transected. Bone marrow cells were flushed from the femurs and tibias. The cell suspension was pipetted repeatedly to obtain a single-cell suspension, which was then filtered with a 70 μm cell strainer and centrifuged at 500 g for 8 min. The acquired cells were plated at a density of 2 × 10⁶/mL and cultured in RPMI 1640 medium (SH30027.01, Cytiva, Marlborough, MA, USA) supplemented with 10% FBS (F4135, Sigma-Aldrich, St. Louis, MO, USA), 20% L929-conditioned medium, and 1 × Antibiotic-Antimycotic (15240062, Thermo Scientific, Grand Island, NY, USA) in a humidified incubator with 5% CO₂ at 37°C for 6 days to induce macrophage differentiation. In some experiments, cells were incubated with 100 ng/mL LPS (L3137, Sigma-Aldrich, St. Louis, MO, USA) or 10 μm AZ26 (HY-101971, MedChemExpress, Monmouth Junction, NJ, USA) at various time points as indicated.

Serum samples were collected using a standard protocol following intraperitoneal injection with PBS (control) or LPS (12.5 mg/kg body weight) into Pfkfb3^WT and Pfkfb3^ΔMϕ mice for 6 h. Cell culture supernatants were collected from Pfkfb3^WT and Pfkfb3^ΔMϕ BMDMs treated with LPS for 16 h (100 ng/mL). Cell culture supernatants and sera were measured for mouse Il1b and Il6 with ELISA kits (Il1b, MLB00C; Il6, M6000B, R&D Systems, Minneapolis, MN, USA). All assays were carried out following the instructions provided by the manufacturer.

NO release was measured using a Sievers NOA 280i chemiluminescence analyzer (Analytix, Sunderland, UK) as previously described (15). Briefly, 20 μl of supernatant or 100 μL of serum was injected into a nitrogen-purge vessel containing a 1% solution of sodium iodide in glacial acetic acid. The output was recorded using a Labchart program (ADInstruments, Colorado Springs, CO, USA) and the area under the curve was converted to picomole NO using a calibration curve constructed after the analysis of a series of sodium nitrite standards ranging from 2.5 to 100 pmol.

Real-time changes in extracellular acidification rate (ECAR) of BMDMs were analyzed with an XF24 Extracellular Flux Analyzer as described previously (16). Briefly, BMDMs were seeded at 2 × 10⁵ per well onto Seahorse XF24 culture microplates (100777-004, Seahorse Bioscience, North Billerica, MA, USA) and incubated in a humidified incubator with 5% CO₂ at 37°C overnight. The next day, after 100 ng/mL LPS treatment for 6 h, the medium was changed to a XF base Medium (102353-100, Seahorse Bioscience, North Billerica, MA, USA) supplemented with 2 mM glutamine, pH adjusted to 7.4 with 0.1 M NaOH, and then the plate was incubated for 45 min in a non-CO₂ incubator at 37°C. The ECAR assay was performed on the XF24 extracellular flux analyzer (Seahorse Bioscience, North Billerica, MA, USA), and the ECAR values were normalized using protein concentration. Inhibitors and activators were used in these tests at the following concentrations: glucose (10 mM), oligomycin (2 μM) and 2-DG (50 mM).

Total RNA of cells and tissues were extracted using TRIzol reagent (15596018, Invitrogen, Grand Island, NY, USA) according to the manufacturer's instructions, and RT-qPCR was done as described previously (17). Briefly, one microgram of total RNA was used to synthesize first stranded cDNA with the iScript^TM cDNA synthesis kit (1708891, Bio-Rad, Hercules, CA, USA). RT-qPCR was performed on a QuantStudio^TM 3 Real-Time PCR System (Applied Biosystems, Grand Island, NY, USA) using iTaq^TM Universal SYBR Green Supermix (1725122, Bio-Rad, Hercules, CA, USA) with the respective gene-specific primers listed in Supplementary Table 1. Quantification of relative gene expression was calculated with the 2^−ΔΔCt method using 18S rRNA as the internal control, and data were expressed as fold change relative to control groups.

Western blot was performed as previously described (18). Cells and tissues were homogenized in RIPA buffer (R0278, Sigma-Aldrich, St. Louis, MO, USA) supplemented with 1% protease inhibitor cocktail (05892970001, Sigma-Aldrich, St. Louis, MO, USA) and 1% phosphatase inhibitor (4906845001, Sigma-Aldrich, St. Louis, MO, USA). After measuring protein concentration using BCA Protein Assay Kit (23225, Thermo Scientific, Grand Island, NY, USA), equal amounts of denatured proteins were loaded and separated onto a 7.5–10% SDS-PAGE gel, and transferred on nitrocellulose membranes (10600015, Cytiva, Marlborough, MA, USA). Membranes were blocked with 5% non-fat milk, then incubated with specific antibodies. The antibodies used were as follows: rabbit Pfkfb3 (1:1000, ab181861, Abcam, Cambridge, MA, USA), rabbit Nos2 (1:1000, sc-650, Santa Cruz, Dallas, TX, USA), rabbit p-Erk (1:1000, 4370S, Cell Signaling Technology, Danvers, MA, USA), rabbit Erk (1:1000, 4695S, Cell Signaling Technology, Danvers, MA, USA), rabbit p-Jnk (1:1000, 4668S, Cell Signaling Technology, Danvers, MA, USA), rabbit Jnk (1:1000, 9252S, Cell Signaling Technology, Danvers, MA, USA), rabbit p-p38 (1:1000, 9215S, Cell Signaling Technology, Danvers, MA, USA), rabbit p38 (1:1000, 9212S, Cell Signaling Technology, Danvers, MA, USA), rabbit p-p65 (1:1000, 3033S, Cell Signaling Technology, Danvers, MA, USA), rabbit p65 (1:1000, 8242S, Cell Signaling Technology, Danvers, MA, USA), and mouse Actb (1:1000, sc-47778, Santa Cruz, Dallas, TX, USA). Images were taken with the ChemiDoc Imaging System (Bio-Rad, Hercules, CA, USA), and band densities were identified using ImageJ (National Institutes of Health, Bethesda, MD, USA).

Murine lungs were flushed with PBS and 4% PFA, fixed and embedded in paraffin according to a standard protocol. Five micrometer paraffin-embedded sections were cut and stained with hematoxylin and eosin (H&E) routinely, the lung injury score was assessed based on the method as reported previously (19). Immunohistochemical staining was performed as described previously (20), briefly, lung sections were first deparaffinized and rehydrated, endogenous peroxidase activity was destroyed with H₂O₂ (3 mL 30% H₂O₂ in 200 mL methanol) for 30 min at room temperature. After antigen retrieval with Antigen Unmasking Solution (H-3301, Vector Laboratories, Burlingame, CA, USA) at 98°C for 10 min, sections were blocked with avidin solution with 10% normal rabbit serum for 1 h at room temperature, and incubated in biotin blocking solution with primary antibodies against Mac2 (3 μg/mL, CL8942F, Cedarlane, Burlington, NC, USA), or Ly6G (3 μg/mL, 551459, BD biosciences, San Jose, CA, USA) at 4°C overnight. Sections were then incubated with a biotinylated rabbit anti-rat IgG secondary antibody (1:200, BA-4001-0.5, Vector Laboratories, Burlingame, CA, USA) for 1 h at room temperature, followed by incubation with ABC solution (PK-6100, Vector Laboratories, Burlingame, CA, USA) for 30 min at room temperature. Peroxidase substrate 3, 3′-diaminobenzidine (3468, Dako, Santa Clara, CA, USA) was used to detect the antibodies according to the manufacturer's instructions. The sections were counterstained with hematoxylin for 30 sec, then dehydrated and mounted with xylene-based mounting medium (8312-4, Thermo Scientific, Grand Island, NY, USA). Quantification of Mac2 and Ly6G staining was performed using the ImageJ (National Institutes of Health, Bethesda, MD, USA).

For immunofluorescence staining of lung sections, 5 μm paraffin-embedded sections were deparaffinized and rehydrated. After antigen retrieval with Antigen Unmasking Solution (H-3301, Vector Laboratories, Burlingame, CA, USA) at 98°C for 10 min, sections were blocked and incubated at 4°C overnight with primary antibodies against Pfkfb3 (1:100, ab181861, Abcam, Cambridge, MA, USA) and Adgre1 (1:100, ab6640, Abcam, Cambridge, MA, USA). For immunofluorescence staining of BMDMs, cells seeded on culture slides (354108, Corning, Glendale, AZ, USA) were fixed in 4% PFA for 20 min, permeabilized with 0.5% Triton X-100 for 20 min, and then blocked and incubated at 4°C overnight with primary antibodies against Pfkfb3 (1:100, ab181861, Abcam, Cambridge, MA, USA) and Adgre1 (1:100, MCA497R, Bio-Rad, Hercules, CA, USA). Sections and slices were incubated at room temperature for 1 h with Alexa Fluor 594-labeled goat anti-rabbit IgG (8 μg/mL, A11012, Molecular Probes, Grand Island, NY, USA) and Alexa Fluor 488-labeled goat anti-rat IgG (8 μg/mL, A11006, Molecular Probes, Grand Island, NY, USA), stained with 4′,6-diamidino-2-phenylindole (1 μg/mL, D1306, Invitrogen, Grand Island, NY, USA) for 5 min, and immersed in mounting medium (H-1000, Vector Laboratories, Burlingame, CA, USA). Images were acquired with a confocal microscope (Zeiss 780 Upright Confocal, Carl Zeiss, Oberkochen, Germany), and analyzed using ImageJ (National Institutes of Health, Bethesda, MD, USA).

Statistical analyses were performed with GraphPad Prism 9 software (GraphPad, La Jolla, CA, USA). Categorical data are shown as percentages, and Mantel-Cox test was used to compare differences between the two groups (survival curves). Continuous variables are presented as the means ± SEM, two-tailed unpaired Student's t-test was used to compare the difference between two groups, and one-way analysis of variance (ANOVA) followed by Bonferroni's post-hoc test was performed for comparison among more than two groups when appropriate. Differences were considered statistically significant at P < 0.05 (^*P < 0.05, ^**P < 0.01 and ^***P < 0.001).

Increased glycolytic flux has been shown in LPS-stimulated macrophages. We first analyzed the expression of glycolytic molecules in BMDMs treated with LPS for 4 h using genome mRNA expression profiling data (GSE125036) from the GEO database. As shown in Figure 1A, the expression of Slc2a1, Hk1, Pfkfb3 and Pfkp was upregulated more than 2-fold in LPS-stimulated macrophages compared to untreated macrophages, and Pfkfb3 is the highest expressed gene among the upregulated glycolytic genes. To further confirm the upregulation of Pfkfb3 in LPS-stimulated macrophages, we first analyzed the mRNA levels of Pfkfb3 in LPS-stimulated BMDMs using real-time PCR, and found that LPS increased Pfkfb3 mRNA expression in a time-dependent manner (Figure 1B). Consistently, the protein levels of Pfkfb3 were also significantly upregulated in BMDMs 2 h after LPS stimulation (Figure 1C). Moreover, increased Pfkfb3 expression in macrophages was also observed in lung from LPS-challenged mice compared with that from control mice (Figure 1D). Together, these findings suggested that inflammatory stimulator LPS reprograms the expression of Pfkfb3 in macrophages, and this may be a critical link in severe inflammatory responses in septic mice.

To determine the contribution of Pfkfb3-mediated glycolysis in macrophages to LPS-induced sepsis, we generated myeloid-specific Pfkfb3 deficient mice (Pfkfb3^ΔMϕ) and control mice (Pfkfb3^WT) by intercrossing Pfkfb3^flox/flox mice with Lysm-Cre mice (Figures 2A,B). The protein level of Pfkfb3 was obvious decreased in main organs (including lung and liver) and bone marrow from Pfkfb3^ΔMϕ mice compared with those from Pfkfb3^WT mice (Figure 2C), indicating that Pfkfb3 is highly expressed in myeloid cells. The mRNA and protein analysis showed that Pfkfb3 expression was barely detectable in BMDMs cultured with bone marrow from Pfkfb3^ΔMϕ mice compared with those cultured with bone marrow of Pfkfb3^WT mice (Figures 2D–F), confirming effective deletion of Pfkfb3 in the myeloid cells of Pfkfb3^ΔMϕ mice.

To examine the role of myeloid Pfkfb3 in LPS-induced cardiac dysfunction (one of the features of multi-organ dysfunction in sepsis), Pfkfb3^WT and Pfkfb3^ΔMϕ mice were intraperitoneally administered with LPS, and cardiac assessment was used to record the cardiac parameters using M-mode echocardiography. As shown in Figures 3A–E, compared with the control mice, Pfkfb3^WT mice exhibited cardiac dysfunction 6 h after LPS challenge, as evidenced by decrease in ejection fraction, fractional shortening, left ventricle stroke volume and cardiac output. However, myeloid-specific Pfkfb3 deficiency significantly prevented the decrease in cardiac function of Pfkfb3^WT mice observed after LPS injection (Figures 3A–E). Moreover, Pfkfb3^ΔMϕ mice exhibited significantly increased blood pressure compared to Pfkfb3^WT mice at 12 h after LPS injection, as indicated by increasing systolic blood pressure, diastolic blood pressure and mean blood pressure (Figures 3F–H). Together, these findings suggest that myeloid Pfkfb3 is a critical mediator in LPS-induced cardiac dysfunction in vivo.

To investigate the effect of myeloid-specific Pfkfb3 deficiency on LPS-induced acute lung injury, we first determined the histological alterations of lungs from LPS-challenged Pfkfb3^ΔMϕ and Pfkfb3^WT mice by hematoxylin and eosin (H&E) staining. The H&E staining results indicated a lower degree of recruitment of inflammatory cells into the lungs of Pfkfb3^ΔMϕ mice compared to those of Pfkfb3^WT mice (Figure 4A). To further examine the protective effect of myeloid-specific Pfkfb3 deficiency on lung injury, the pulmonary permeability was assessed in vivo. The lung wet-to-dry ratio, which was used as an index to evaluate lung edema, was decreased in the lungs of Pfkfb3^ΔMϕ mice compared with those of Pfkfb3^WT mice after LPS injection (Figure 4B). Next, we injected Evans Blue in Pfkfb3^ΔMϕ and Pfkfb3^WT mice with or without LPS treatment. As shown in Figure 4C, Evans Blue leakage majorly decreased in the lungs of Pfkfb3^ΔMϕ mice compared to those of Pfkfb3^WT mice after LPS injection. In addition, Pfkfb3^ΔMϕ mice exhibited higher body temperature after LPS challenge compared to Pfkfb3^WT mice (Figure 4D). Moreover, we observed a marked improvement in mortality: 67% Pfkfb3^ΔMϕ mice survived after 96 h of the LPS-challenge whereas none of Pfkfb3^WT mice survived in the control group (Figure 4E). Collectively, these observations suggest that myeloid-specific Pfkfb3 deficiency protects mice against LPS-induced lethality and tissue injury.

Recruitment of myeloid cells into the lung represents a key feature of LPS-induced tissue injury. We first exposed Pfkfb3^ΔMϕ and Pfkfb3^WT mice to 12.5 mg/kg LPS for 6 h and then detected the infiltration of neutrophils and macrophages in the lung tissue with immunostaining. The results showed that the infiltration of Ly6G positive cells and Mac2 positive cells was significantly decreased in the lungs of Pfkfb3^ΔMϕ mice compared to those of Pfkfb3^WT mice (Figures 5A,B). Given that the recruitment of myeloid cells is mainly dependent on the excessive expression of proinflammatory mediators, including interleukin 1b (Il1b), interleukin 6 (Il6) and inducible nitric oxide synthase 2 (Nos2), we next analyzed the effect of myeloid-specific Pfkfb3 deficiency on the expression of proinflammatory genes in lungs using real-time PCR. The mRNA levels of Il1b, Il6 and Nos2 were significantly increased in the lungs of LPS-challenged Pfkfb3^WT mice compared to those of control mice, whereas these mRNA levels were decreased in the lungs of LPS-challenged Pfkfb3^ΔMϕ mice compared to LPS-challenged Pfkfb3^WT mice (Figures 5C–E). Consistently, the serum levels of Il1b, Il6 and NO were significantly increased in LPS-challenged Pfkfb3^WT mice compared to control mice, whereas the serum levels of Il1b, Il6 and NO were decreased in LPS-challenged Pfkfb3^ΔMϕ mice compared to LPS-challenged Pfkfb3^WT mice (Figures 5F–H). Thus, these data suggest that myeloid-specific Pfkfb3 deficiency protects mice from LPS-induced excessive inflammatory responses in vivo.

Having observed that myeloid Pfkfb3 deficiency reduced LPS-induced inflammation in vivo, we further investigated whether Pfkfb3 regulates the expression of proinflammatory cytokines and Nos2 in macrophages in vitro. We first analyzed the glycolytic flux in BMDMs cultured with bone marrow from Pfkfb3^ΔMϕ mice or corresponding wild-type Pfkfb3^WT mice using the Seahorse Extracellular Flux analyzer. As shown in Figures 6A–C, Pfkfb3^ΔMϕ BMDMs displayed significantly reduced glycolysis and glycolysis capacity in response to LPS compared to Pfkfb3^WT BMDMs. Importantly, Pfkfb3^ΔMϕ BMDMs displayed reduced mRNA levels of Il1b, Il6 and Nos2 in response to LPS compared with Pfkfb3^WT BMDMs (Figures 6D–F). Consistent with mRNA levels, Pfkfb3^ΔMϕ BMDMs displayed significantly reduced protein levels of Nos2 in response to LPS compared with Pfkfb3^WT BMDMs (Figures 6G–I). Moreover, the secretion of Il1b, Il6 and NO was significantly decreased in Pfkfb3^ΔMϕ BMDMs compared to that in Pfkfb3^WT BMDMs (Figures 6J–L). Thus, these data suggest that inhibition of glycolysis with Pfkfb3 deletion in macrophages reduces LPS-induced expression of proinflammatory cytokines and Nos2 in vitro.

We next investigated the molecular mechanisms by which LPS-induced signal transduction is regulated by Pfkfb3 in macrophages. The activation of nuclear factor-κB (NF-κB) and mitogen-activated protein kinase (MAPK) by LPS is the key signaling pathways for the induced expression of proinflammatory genes in macrophages. We analyzed the phosphorylation of Erk, Jnk, p38 and p65 in LPS-stimulated Pfkfb3^ΔMϕ and Pfkfb3^WT BMDMs by Western blot. As shown in Figures 7A–E, the protein levels of p-Erk, p-Jnk, p-p38 and p-p65 were markedly increased in Pfkfb3^WT BMDMs stimulated with LPS. However, of the above observed signaling molecule changes, only the upregulation of p-p65 by LPS was suppressed in Pfkfb3^ΔMϕ BMDMs treated with LPS. Moreover, similar to Pfkfb3 knockout, AZ26, which is an inhibitor that abolishes the enzyme activation of Pfkfb3 (21), could antagonize the upregulation of proinflammatory cytokine Il1b and Il6 expression in LPS-treated wild type BMDMs (Figures 7F,G). Furthermore, the phosphorylation of p65 was also antagonized by AZ26 in LPS-treated BMDMs (Figure 7H). These data suggest that the NF-κB signaling pathway is involved in the downregulation of proinflammatory mediators in macrophages by Pfkfb3 inactivation.