All mice were housed in specific-pathogen-free (SPF) facility under conditions of a 12 h light/dark cycle, controlled temperature (23 ± 1 °C), relative humidity (55% ± 5%) with food and water ad libitum. The animal experiments were approved by the Institutional Animal Care Use Committees at Army Medical University (AMUWEC20207065).

Mouse breeding and genotyping

Wild-type male C57BL/6 J and DBA/1 J mice were obtained from ENSIWEIER Company. The Hspa5f/f and Irf2bp2f/f mice were generated using the CRISPR/Cas9 system in the C57BL/6 J mouse background. Briefly, for Hspa5, an Hspa5 donor vector containing flox sites flanking exons 5-9 of Hspa5 was created. Two sgRNA sites targeting intron 4 and the 3’UTR were transcribed in vitro. The sgRNA targeting intron 4 was 5’-AGTTAAGATTGAAAGGTTTCTGG-3’, and the sgRNA targeting the 3’UTR was 5’-ATTGAGAAAAAGGTGGGTCAGGG-3’. The strategy for Irf2bp2 was similar to that for Hspa5. floxed mice, in which exons 1-2 of Irf2bp2 were flanked with loxP recombination sites, were generated. Briefly, an Irf2bp2 donor vector containing flox sites flanking exons 1-2 of Irf2bp2 was created. Two sgRNA sites targeting the 5’ upstream sequence and 3’UTR were transcribed in vitro. The sgRNA targeting the 5’ upstream sequence was 5’-GCTGCACACCTGCGGCGCTAGGG-3’, and the sgRNA targeting the 3’UTR was 5’-GAAATGTAGACTTCATAACATGG-3’. Thereafter, the donor vector with two gRNAs and Cas9 mRNA was microinjected into fertilized C57BL/6 J eggs. F0 generation mice, which were positive for homologous recombination, were identified by long PCR spanning the 5’ or 3’ homologous arm. The positive founder mice were mated to WT C57BL/6 J mice to obtain Hspa5 or Irf2bp2 flox heterozygous mice. The mice were generated by Shanghai Model Organisms Center, Inc. (Shanghai, China). Hspa5f/f; Lyz2-Cre and Irf2bp2f/f; Lyz2-Cre mice were obtained by crossing with Lyz2-Cre mice from the Jackson Laboratory. Mice between 6 and 12 weeks of age were used for the experiments.

Reagents

Metabolite candidates in screening assay were used at the concentration of 6 mM. Phenyllactate, L-malate, cis-aconitate, leucine, citrate, uracil, guanosine, serine, oxaloacetate, fructose 6-phosphate, pyruvate, uridine, UMP, phosphorylethanolamine, ornithine, threonine, glutamine, ATP, GDP, argininosuccinate, citrulline, cysteine, isocitrate, GMP, IMP, α-ketoglutarate, ribose 5-phosphate, aspartate, NAD + , PRPP, GTP, and Flavin mononucleotide were purchased from MCE. PRPP was purchased from Cayman. Alanine, methylmalonate, and glucose were purchased from Selleck. BPG, acetyl-CoA, argininosuccinate, and glucose 6-phosphate were purchased from Sigma. Small molecule compounds used in SPR assay: L-malate (#02288, Sigma), D-malate (#46940-U, Sigma), Succinate (#14079, Sigma), 3-Hydroxybutyrate (#HY-113378, MCE), L-2-hydroxybutyrate acid (#HY-W018499, MCE), L-2-hydroxyglutarate (#90790, Sigma), Malonate (#792535, Sigma).

Generation and treatment of BMDMs

Bone marrow cells from the femur and tibia of 6–10-week-old male C57BL/6 J mice were differentiated in the presence of recombinant mouse M-CSF (20 ng/mL; PeproTech) in DMEM (containing 10% fetal calf serum and 1% penicillin/streptomycin) for 5–7 days. For experiments, the cells were seeded at a concentration of 0.5 × 106 cells per ml in tissue culture plates of various formats. The cells were treated with L-malate (0.5 - 6 mM, #02288, Sigma), D-malate (0.5 mM, M109342, Aladdin), Dimethyl L-malate (0.5 -3 mM D806475, MACKLIN), monensin sodium salt (0.1 - 1 μM, #HY-N0150, MCE), NH4Cl (15 - 30 mM, #HY-Y1269, MCE), sodium acetate (20 mM, #S2889, Sigma), H89 (20 μM, #S1582, Selleck), Tunicamycin (0.1 μM, #654380, Sigma), Thapsigargin (0.1 μM, #HY-13433, MCE), Arachidonic acid (200 μM, #HY-109590, MCE), DCPIB (100 μM, #HY-103371, MCE), CHX (5 μg/mL, #S7418, Selleck), MG-132 (5 μM/0.5 μM, #HY-13259, MCE), bafilomycin A1 (100 nM, #HY-100558, MCE), heptelidic acid (1 μM, #14079, Cayman Chemical), fumarate hydratase-IN-1 (25 μM, #HY-100004, MCE), LPS (100 ng/mL, #L4524, Sigma), 4μ8C (IRE1 Inhibitor III) (100 μM, #HY-19707, MCE), GSK2606414 (10 μM, #S7307, Selleck), and ceapin-A7 (8 μM, #HY-108434, MCE) as indicated.

Generation and treatment of peritoneal macrophages

6–10-week-old male C57BL/6 J mice were intraperitoneally injected with 3 mL 3% thioglycolate solution 3 days in advance. Mice were euthanized with carbon dioxide. Then, the peritoneum was opened to maximize ascites collection using instruments under sterile conditions. Then erythrocytes were removed from the ascites, and the remaining cells were centrifuged and resuspended in DMEM (containing 10% fetal calf serum and 1% penicillin/streptomycin). And these cells were cultured at 37 °C with 5% CO2. After 2 h, the adherent cells were perceived as peritoneal macrophages. The cells were treated with MG-132 (5 μM, #HY-13259, MCE), cellular pH calibration buffer (#P35379, thermo scientific), Nigericin sodium salt (10 μM, # HY-100381, MCE), Valinomycin (10 μM, #HY-N6693, MCE).

PBMC isolation

Peripheral blood was collected in EDTA-coated tubes. After centrifugalization at 3000 × g for 10 min at room temperature, plasm was collected and stored. The remaining blood cells were diluted one-fold with phosphate-buffered saline (PBS) and were layered gently on top of half amount of Ficoll-Paque PREMIUM 1.084 (#17544602, GE Healthcare) in a 15 mL tube. After centrifugation (450 g, 20 min), PBMCs were concentrated in the second layer from the top to the bottom and was collected with pipettes and then stored in another 15 mL tube. Collected PBMCs were washed twice with PBS and centrifuged at 300 x g for 10 min. The isolated PBMCs were resuspended in freezing solution (90% FBS, 10% DMSO) and stored under liquid nitrogen for future use.

Intracellular pH measurement

Intracellular pH was determined using the ratiometric fluorescent intracellular pH probe 2′,7′-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein acetoxymethyl ester (BCECF-AM)141 according to the protocol of the commercial kit (#B1150, Thermo Scientific). Before the measurement of Intracellular pH, BMDMs and Raw264.7 cells were implanted in 96 Wells dedicated to fluorescence intensity measurement at 70%-80% confluence. After the treatments as described aforementioned, 2 μM BECEF-AM was added into 96 Wells and incubated for 45 min at 37 °C. The fluorescence intensity was quantified by alternating excitation at two different wavelengths under λ1 = 490 nm and λ2 = 440 nm, while fixed emission at 535 nm respectively. Then the fluorescence intensity was used to calculated the absolute value of intracellular pH based on the standard curve which was established as described below. With the aim of quickly balancing and equalizing intracellular and extracellular pH values, the cellular pH calibration buffer, valinomycin (10 μM) and nigericin (10 μM) were used to establish the standard curve in accordance with the protocol of the commercial kit (#P35379, Thermo Scientific).

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) activity assay

The GAPDH activity analysis was compliant with the instructions provided with the kit (#ab204732, Abcam). BMDMs were cultured in 6-well plates at 1 ×106 cells per well and treated with heptelidic acid and LPS for 16 h. After cracking and centrifugation, 20 µL of supernatant was used as the sample to be tested. The 50 µL reaction system was proportionally configured in advance before the formal measurements, and this system was composed of 46 µL of assay buffer IV/GAPDH assay buffer, 2 µL of developer solution III/GAPDH developer, and 2 µL of glyceraldehyde-3-Phosphate/GAPDH substrate. Then, 20 µL of sample was combined with 50 µL of the reaction mix, and the mixture was loaded onto a 96-well plate to measure the absorbance at OD450 nm in the kinetic mode at 37 °C.

Quantitative real-time reverse transcription PCR

Total RNA was extracted from cells or tissues with a RNeasy Micro Kit (QIAGEN). cDNA synthesis was carried out with a RevertAid First Strand cDNA Synthesis Kit (#K1622, Thermo Scientific) according to the manufacturer’s instructions. Real-time quantitative PCR was performed with PowerUP SYBR Green Master Mix (Thermo Fisher Scientific) and the CFX96 Touch System (Bio-Rad). 18S rRNA was used as an internal control for normalization. The primers for qRT‒PCR are listed below: 5’-TGCCACCTTTTGACAGTGATG-3’ and 5’-TGATGTGCTGCTGCGAGATT-3’ for Il1b; 5’-CTCCCAACAGACCTGTCTATAC-3’ and 5’-CCATTGCACAACTCTTTTCTCA-3’ for Il6; 5’-CCCTCACACTCACATCATCTTCT-3’ and 5’-GCTACGTGGGCTACAG-3’ for Tnfa; 5’-CACTGTCGAGTCGCGTCC-3’ and 5’-TCATCCATGGCGAACTGGTG-3’ for Actb; 5’-AGGTCGGTGTGAACGGATTTG-3’ and 5’-TGTAGACCATGTAGTTGAGGTCA-3’ for Gapdh; 5’-AGCAAGTTCAAGAAGGAGCCG-3’ and 5’-GCTGTCCTTGCAACTGCTTTA-3’ for Irf2bp2; 5’-CCTGAGCCCGGAGGAGAA-3’ and 5’-CTCGAGCAGTCTGCGCTG-3’ for Xbp1; 5’-GAGTCCGCAGCAGGTG-3’ and 5’-GTGTCAGAGTCCATGGGA-3’ for Xbp1s.

Immunoblot analysis

Tissues or cell proteins were extracted with M-PER (Thermo Fisher Scientific) supplemented with 1× cocktails containing protease and phosphatase inhibitors (Roche). The protein concentrations were determined using a BCA kit (Thermo Fisher Scientific). The extracted proteins were resolved by 4-12% SDS‒PAGE and then transferred onto a polyvinylidene difluoride (PVDF) membrane using either a wet or semidry transfer system. The membranes were blocked with 5% (w/v) dried milk in TBS-Tween (TBST) for 60 min at room temperature. The membranes were incubated with primary antibodies [anti-beta-actin (#8457, Cell Signaling; R23613, ZENBIO), anti-IRF2BP2 (#18847-1-AP, Proteintech), anti-beta-tubulin (#2146, Cell Signaling), anti-IL-1-beta (#12507, Cell Signaling), anti-BiP (#ab21685, Abcam), anti-ATF6 (#24169-1-AP, Proteintech), anti-eIF2α (#ab169528, Abcam), anti-P-eIF2α (9721S, Cell Signaling)] at 4 °C overnight and then with horseradish peroxidase-conjugated secondary antibodies (CST) for 2 h at room temperature. The blots were developed with a Western blotting detection kit (Advansta) for 1 min, and the bands were visualized using Fusion FX6-XT (VILBER).

ELISA assay

Levels of IL-1β, IL-6 and TNF-α were detected in BMDMs medium or serums isolated from murine models using ELISA kits, in accordance with the manufacturer’s instructions [BGK10749 (IL-1β), BGK08505 (IL-6), BGK06804 (TNF-α) Biogem, originally known as a part of Peprotech].

Immunofluorescence

BMDMs were washed in PBS and fixed with 4% paraformaldehyde at room temperature for 15 min. The cells were then blocked for 30 min with 5% bovine serum albumin (BSA) diluted with 0.3% Triton X–100. The cells were incubated with antibodies against BiP (1:100, ab212054, Abcam) and Irf2bp2 (1:100, HPA027815, Sigma) at 1:50 dilution overnight at 4 °C in a dark humidified chamber. After washing, the cells were incubated with a fluorophore-conjugated secondary antibody for 60 min. The cells were then counterstained with DAPI in the dark for 5 min. The stained samples were visualized and photographed by laser-scanning confocal microscopy (Olympus, IXplore SpinSR).

Proximal ligation assay

A proximal ligation assay was conducted as previously described.142 Briefly, BMDMs were seeded on a chamber slide. The cells were fixed with 4% PFA, and washed three times with PBS before performing the PLA experiment according to the manufacturer’s instructions provided with the Duolink In Situ Red Starter Kit Mouse/Rabbit (Sigma Aldrich). Subsequently, the nonspecific cellular antigens were blocked with blocking solution for 1 h at 37 °C and then incubated with antibodies against BiP (1:100, ab212054, Abcam) and Irf2bp2 (1:100, HPA027815, Sigma) alone or simultaneously in the Probemaker PLA Probe Diluent overnight at 4 °C. Thereafter, the PLA probes were linked to corresponding antibodies and amplified in relevant buffer at 37 °C. Duolink In Situ Mounting Medium with DAPI (Sigma) was used to mount the slides with a cover slip. Images were acquired using a confocal microscope (Olympus, IXplore SpinSR).

DARTS

Drug affinity responsive target stability (DARTS) assay was adapted from Zhang et al..45 Raw 264.7 (washed by pre-chilled PBS twice) were lysed using M-PER (Thermo Scientific, 78501) supplemented with the addition of protease inhibitors (Roche) and phosphatase inhibitors (Roche). TNC buffer (50 mM Tris-HCl pH 8.0, 50 mM NaCl, 10 mM CaCl2) was added to the lysate and centrifuged at 14,000 at 4°C for 10 min. Protein concentration of the supernatant was determined using the BCA Protein Assay kit. Proteins were diluted to the same final volume and protein concentration. To add L-malate (1 mM) to the supernatant, 100 mM L-malate solution with pH adjusted to 7.0 or vehicle control (H2O, pH 7.0) was added to the supernatant at 99 times the volume, followed by incubation with vehicle or L-malate for 1 h at room temperature. Digestion was performed using Pronase (Roche) at room temperature for 15 min and was stopped using excess protease inhibitors with immediate transfer to ice. The resulting digests were separated by SDS-PAGE 10% gel. Bands with increased staining were excised from the lane of L-malate-treated group and the matching area of the vehicle lane, and then the gels were prepared for LC-MS/MS analysis.143 LC separation was performed online on an EASY-nLC 1000 (Thermo Scientific). Peptides were gradient eluted directly to an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher). High resolution full MS spectra were acquired every second with a resolution of 240,000, an AGC target of 1e6, with a maximum ion injection time of 50 ms, and scan range of 400 to 1500 m/z. Following each full MS data-dependent HCD MS/MS scans were acquired in the ion trap using the rapid scan mode with an AGC target of 6e4, maximum ion time of 18 ms, one microscan, 0.7 m/z isolation window, normalized collision energy (NCE) of 30, fixed first mass 110 m/z and dynamic exclusion for 20 s. Only ions with a charge state of 2-7 were allowed to trigger an MS2 scan. The MS/MS spectra were searched against a Uniprot (www.uniprot.org) mouse protein database with common lab contaminants added using Sequest within Proteome Discoverer 1.4 (Thermo Fisher). The search parameters were as follows: mass accuracy better than 10 ppm for MS1 and 0.4 Da for MS2, two missed cleavages, fixed modification carbamidomethyl on cysteine, variable modification of oxidation on methionine and deamidation on asparagine and glutamine. The data was filtered using a 1% FDR cut off for peptides and proteins against a decoy database and only proteins with at least 2 unique peptides were reported.

Macrophage polarization assay

The polarization of BMDMs was induced for M1 activation by treatment with LPS (100 ng/mL, Sigma) for 24 h or the indicated periods before further analysis.

Flow cytometry

Flow cytometry was conducted as previously described.144 For the detection of macrophage polarization, mouse claws or colorectal resection samples were mechanically cut into 1-mm3 pieces with scissors and digested with PBS containing Ca/Mg, 10 μg/ml DNase I (Sigma Aldrich), and 1 mg/ml collagenase IV (Sigma Aldrich) for 30 min at 37 °C. The solution was stirred on a magnetic stirrer. After digestion, the tissue suspension was filtered through a 70-μm filter and centrifuged at 500 × g for 5 min. The blood cells were lysed in RBC lysis buffer (Sigma Aldrich) for 10 min and then mixed with PBS to stop the reaction. The samples were then resuspended in FACS buffer (PBS with 2% FBS) for staining. The spleen was separated by a 70-μm filter and washed with FACS buffer. After centrifugation at 500 × g for 5 min, the blood cells were lysed as described above. The spleen sample was eventually resuspended in FACS buffer. The following antibodies were used: PerCP/Cy5.5 anti-mouse CD45 antibody (#103132, BioLegend), APC/Cy7 anti-mouse F4/80 (#123118, BioLegend), APC anti-mouse CD206 (#141708, BioLegend), FITC anti-mouse CD45R (#103206, BioLegend), FITC anti-mouse CD3 (#100203, BioLegend), and PE anti-mouse CD11c (#117308, BioLegend). Macrophages were marked based on the phenotype CD3-CD45R-CD45+ F4/80+ . M1-like cells were defined by the phenotype CD3-CD45R-CD45+F4/80+CD206-CD11+ .

In the cell apoptosis assay, BMDMs were collected, washed twice with cold PBS and stained with the APC Annexin V Apoptosis Detection Kit with PI (#640932, BioLegend). The apoptosis rate was detected by flow cytometry after incubation at room temperature for 15 min in darkness.

Collagen antibody induced arthritis (CAIA) model

CAIA was induced as previously described.145 On day 0, 6–8-week-old male DBA/1 mice were injected intravenously with 80 μg/mg 5-clone cocktail (Catalog # 53100, Chondrex, LLC, Seattle, WA). CAIA mice were given an intraperitoneal injection of PBS, 200 mg/kg malate, or 400 mg/kg malate every day, starting on day 2. On day 3, an additional injection of 1.5 μg/mg LPS injection was given. Severity of arthritis was assessed by the qualitative clinical score and the swelling score via determining paw thickness using a Mitutoyo loop handle dial thickness gauge with a caliper.

For histopathological assessments, the ankle joints of CAIA mice treated with PBS, 200 mg/kg malate, or 400 mg/kg malate were sectioned, deparaffinized, rehydrated, and stained with hematoxylin and eosin. Slides were scanned using an Olympus VS200 slide scanner and visualized with OlyVIA version 3.2 software (Olympus).

Endotoxin-induced model of sepsis

For survival studies, the mice were intragastrically administered L-malate (50, 100, 200, or 400 mg/kg) or PBS for 2 h, sepsis was induced by the injection of 50 mg/kg LPS, and the survival was monitored every 3 h for up to 24 h. The mice were culled immediately at a humane end-point.

Induction and assessment of experimental (acute) colitis

Colitis was induced by colitis-grade dextran sodium sulfate (DSS; 36–50 kDa; 3% weight/volume; MP Biomedicals) added to the drinking water for 7 days. The control mice received standard drinking water. C57 mice were intragastrically administered PBS, 50 mg/kg L-malate, 100 mg/kg L-malate, or 200 mg/kg L-malate every day starting 2 days after the addition of DSS. The disease activity index (DAI) was calculated by scoring the percent weight loss, intestinal bleeding [no blood, occult blood (hemoccult + ), or gross blood], and stool consistency (normal stool, loose stool, or diarrhea), as previously described (28). After 7 days of DSS treatment, the mice were sacrificed. For the histological assessment of colitis, colon specimens were fixed in 4% paraformaldehyde (PFA) and embedded in paraffin. Four-micrometer tissue sections were stained with H&E.

Coimmunoprecipitation (Co-IP)

Co-IP was conducted as previously described.146 Briefly, cell lysis buffer (87787, Thermo) for Western and IP was used for the protein extraction of Raw264.7 on ice for 15 min, and then the lysate was quantified. Next, magnetic beads (Sera-Mag SpeedBead Protein A/G, 17152104010150, GE) which were incubated with 4-25 μg of antibodies against BiP (PA5-19503, Invitrogen) or IRF2BP2 (18847-1-AP, ProteinTech) at 4 °C overnight, were washed three times with lysis buffer, and then added into 500-1000 μg of cellular protein extract and co-incubate with extract at 4 °C overnight. Thereafter, these magnetic beads were washed three times with lysis buffer. The bound protein and input samples were examined by Western blotting with anti-BiP (PA5-19503, Invitrogen), anti-IRF2BP2 (PA5-55700, Invitrogen), anti-ubiquitin (3933, CST) or anti-beta-actin (#8457, Cell Signaling; R23613, ZENBIO) antibodies.

Human proteome microarray

Analysis of protein‒protein interactions using a Human Proteome Microarray (CDI Laboratories) was performed as described previously with some modifications.147 In brief, a total of 25 μg of recombinant human BiP protein (#ab78432, Abcam) was added to 67 μL of labeling buffer, and then 3 μL of 10 μg/μL Biotin Reagent diluted in N, N-Dimethylformamide was added. Subsequently, the above mixture was shaken at room temperature for 2 h, once every 10 min. Finally, 30 μL of stop solution was added to the mixture, followed by shaking at room temperature for 30 min. Three slides of the proteome microarray were preblocked with blocking buffer (PBS with 0.1% Tween 20 and 5% BSA) for 1.5 h and then incubated with biotin (control), biotin-labeled BiP protein at 1 μg/ml, or a mixture of L-malate (1 mM) and biotin-labeled BiP protein for 1 h at room temperature. The slides were then subjected to three 5-min washes with 1× phosphate-buffered saline buffer containing Tween-20 (PBST) and subsequently to two 5-min washes with distilled water. The microarrays were further incubated with Cy5-conjugated streptavidin diluted in 1× detection buffer (1:1000) for 20 min at room temperature and washed. The microarrays were then spun to dryness and scanned with GenePix 4000B (Axon Instruments, Sunnyvale, CA, USA) for visualization. The proteome microarray data analysis included four steps: image scan, background correction, interchip normalization and differential spot statistics.60 (1) Image scan: The median foreground and background intensities for each spot in the protein microarrays were obtained with GenePix Pro 6.0 software. We manually checked all the spots on the chips and adjusted the size and position of the spots with dust or specks. (2) Background correction: We calculated the signal intensity of each spot, which was defined as the foreground median intensity divided by its local background median intensity, named the “I” value. (3) Interchip normalization: The values of “I” between the “BiP” and “BiP+MA” samples were normalized using the R package “limma” with the function “normalizeBetweenArrays”.148 The aim of normalization is to reduce the effect of systematic errors caused by experimental factors and make the microarrays comparable in order to allow the discovery of actual biological differences. (4) Differential spot statistics: Differential spots were defined as those with a difference in normalized I values >2 between the “BiP+MA” (normalized) and “BiP” (normalized) samples.

Measurement of L-malate and fumarate by LC–MS/MS

The sample extracts were collected from BMDMs or PBMCs washed with ice cold DPBS for three times, and were analyzed by tandem mass spectrometry (MS/MS, QTRAP® 6500+ with ultra-performance liquid chromatography (UPLC, ExionLC™ AD)). L-malate was analyzed by scheduled multiple reaction monitoring. The metabolites were isolated using an ACQUITY UPLC amino column (2.1 mm × 100 mm, 1.7 μm) with a maintained column temperature of 40 °C. The mobile phase comprised 10 mmol/L ammonium acetate, 0.3% ammonia solution, and 90% acetonitrile water. The flow rate was set to 0.4 mL/min with a sample volume of 2 μL. The MS operating parameters were as follows: ion spray voltage, -4500 V; temperature, 550 °C°C; curtain gas, 35; ion source gas 1, 50; ion source gas 2, 60; collision gas, 8; entrance potential, 10; collision cell exit potential, 10; and scan type: multiple reaction monitoring (MRM). The quantitative standard curve was prepared for LC‒MS analysis. Standard linear regression curves were drawn with the mass of the analyte peak area on the vertical axis and the analyte concentration on the horizontal axis.

L-malate metabolite analysis and stable isotope tracing by GC-MS

LPS-stimulated BMDMs were treated with deuterium-labeled L-malate (6 mM, 8 h) (MCE, HY-Y1069S) in a pHe-controlled condition. After incubation with medium, cells were washed with ice-cold PBS for three times and frozen with liquid nitrogen. For the cultured cell tracing experiment, samples preparation and methods of metabolomics analysis were performed as described previously.149,150 Natural isotope abundance was corrected using IsoCor v.2.0.151

Measurement of glycolytic rate by Seahorse assay

BMDMs extracellular acidification rate (ECAR) were measured using the Seahorse XFe96 Extracellular Flux Analyzer (Seahorse Bioscience, Agilent) according to manufacturer’s instructions of Seahorse XF glycolytic rate assay kit (103344, Agilent). BMDMs were seeded into 96-well XF cell culture microplate in induction medium (3.3 × 104 per well). Before the Seahorse assay, the cells were first treated by LPS with or without L-malate (6 mM, pH 6.7) for different times, and then remove the medium. Wash once with warmed assay medium (XF DMEM medium, pH 7.4, with 1 mM pyruvate, 2 mM glutamine and 10 mM glucose), and incubate with warmed assay medium at 37 °C in a non-CO2 incubator for 60 min. Then, remove the assay medium, and add fresh, warm assay medium and injected Rot/AA and 2-DG (2-deoxy-glucose) sequentially. The glycolytic proton efflux rate (glycoPER), basal glycolysis rate and % PER from glycolysis were calculated using Wave Desktop 2.6 (Agilent Technologies).

Surface plasmon resonance (SPR) analysis

The SPR experiments were performed using a Biacore 8 K system at 25 °C with a CM5 sensor chip. The recombinant BiP protein (#ab78432, Abcam) or BiP fragments for SPR analysis were diluted into NaAc (pH 4.0) and immobilized on the CM5 chip using EDC/NHS regents; a blank channel was used as a negative control. Small molecule compounds and recombinant IRF2BP2 protein (OriGene) diluted with HBS-EP (or HBS-EP with small molecule compounds) were allowed to flow over the chip surface, and the response units were measured. The binding kinetics constants were analyzed using Bia-evaluation analysis software (Cytiva).

Preparation of BiP protein

The appropriate plasmid pET28a-HSPA5 was transformed into calcium chloride-competent E. coli BL21 (DE3), and the cells were then induced by 1 mM IPTG at 37 °C for 3 h. Recombinant protein was purified from the supernatant of the bacterial lysate using an affinity HisTrapTM HP column (Cytiva). After the column was washed with 10 mL of equilibrium buffer, the protein was eluted by imidazole and dialyzed in PBS (pH 7.4). The purified product was analyzed by SDS‒PAGE.

Molecular docking simulation

The molecular docking of malate to HSPA5 was completed with the software package described by Schrödinger (2015). The crystal structure (PDB ID: 5E84) was first prepared using the Protein Preparation Wizard module. During the process, nonessential components in the crystal structure were deleted, and only the protein structure was retained. The force field of OPLS3 was selected to perform restricted energy optimization. The RMSD value of hydrogen atom variation was limited to 0.30 Å. The obtained protein structure was used for further docking simulation. The binding sites were detected using the Sitemap module. According to the identified sites, grid files were generated. Malate was then processed to generate low-energy 3D structures using the LigPrep module, and possible protonation states were generated at pH 7.0 ± 0.5 with the force field of OPLS3. The original chirality of malate was maintained, and isomers were not generated. The structure was docked into each binding site using the Glide extra precision (XP) mode with default settings. The docking conformations were output based on their docking score.

The structure of IRF2BP2 was predicted using RoseTTAFold according to its amino acid sequence. Based on the evaluation results and error estimate values, Profiles-3D and Ramachandran plot, model 1 was ultimately selected for protein‒protein docking with HSPA5 using ZDOCK implemented in Discovery Studio 2019. The force field of CHARMM was set for protein‒protein docking. The top 2000 poses of 100 clusters were set to output based on the ZDOCK score. The other parameters were set as defaults.

Generation of Mdh2 −/− and Hspa5+/− Raw264.7 cells by CRISPR‒Cas9

Hspa5+/− and Mdh2−/− cell lines derived from Raw264.7 cells were constructed using the CRISPR/Cas9 system. Gene-specific sgRNAs were designed to target gene coding regions as follows: Hspa5 sgRNA sense, 5′-AGCGCCTCATCGGACGCACT-3′ and antisense, 5′-AGTGCGTCCGATGAGGCGCT-3′; and Mdh2 sgRNA sense, 5′-GTGTGTGAGCGATATCGTAG-3′ and antisense, 5′-CTACGATATCGCTCACACAC-3′. The sgRNA was cloned and inserted into the pHBLV-U6-gRNA-EF1-CAS9-PURO vector (Hanbio Biotechnology). The cotransfection of CRISPR plasmids, psPAX2 and pMD2G into HEK293T cells (ATCC) was performed to produce the lentivirus. Raw264.7 cells were then infected with lentivirus using polybrene (6 μg/mL) for 48 h and then selected with 2 μg/mL puromycin (Gibco) for 2 days.

NaOH-induced colitis in mice

The procedure was performed as described86 with some modifications. Before the start of the study, preliminary experiments were used to determine the dose of NaOH. Twenty-four hours before the procedure, food was removed from the cage, and the mice were only allowed to drink water. On the day of induction, all the mice were anesthetized, and a 4-mm-diameter enema tube was inserted rectally until the tip was 4 cm proximal to the anus. Then, 0.3 ml of 0.9% saline or 0.3% NaOH solution was administered slowly to the Hspa5f/f or Hspa5f/f; Lyz2-Cre mice. Thereafter, all the mice were maintained in a head-down position for 5 min to limit expulsion of the solution. On the morning of the 3rd or the 6th day, all the mice were sacrificed. Laparotomy and total colectomy were performed. The lumen of the resected specimen was irrigated with 0.9% NaCl. The distal colon segment was then split longitudinally into two pieces and preserved for histological and biochemical analyses.

For histopathological assessment, the colons were fixed in 4% paraformaldehyde for 48 h. Blocks were serially sectioned at a thickness of 5 μm along the cephalocaudal axis of the gut to the level of the lumen. Hematoxylin and eosin (H&E) staining procedures were performed according to the standard protocol provided with the stain kit (G1120, Solarbio). IHC staining of IRF2BP2 was performed with anti-IRF2BP2 antibody (1:200, 18847-1-AP, ProteinTech) and detected using horseradish peroxidase-conjugated secondary antibodies. Digital light microscopy images were recorded with a Zeiss Axio Imager.A1 microscope (Thornwood, NY), AxioCam MRc5 camera, and AxioVision 4.7.1 imaging software (Carl Zeiss Microimaging).

RNA sequencing

Total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). All RNA samples were determined to have A260/280 values ≥ 1.8 (Nanodrop); the samples for RNA-seq had RIN values >7 (BioAnalyzer, Agilent). Poly (A) RNA was purified from 1 μg of total RNA, fragmented, and then reverse-transcribed to create cDNA using SuperScript™ II Reverse Transcriptase (Invitrogen, cat. 1896649, USA) to construct a library. Then, 2×150-bp paired-end sequencing (PE150) on an Illumina NovaSeq™ 6000 (LC-Bio Technology Co., Ltd., Hangzhou, China) was performed following the vendor’s recommended protocol.

Bioinformatics analysis

Fastp software was used to remove the reads that contained adaptor contamination, low-quality bases, and undetermined bases with the default parameters. HISAT2 was used to map reads to the reference genome of the mouse: Mus_musculus_Ensemble_101. StringTie was used to determine the expression levels of mRNAs by calculating FPKM values (FPKM = [total_exon_fragments / mapped_reads (millions) × exon_length (kB)]). We considered genes to be significantly differentially expressed based on a threshold of <0.05 in P value obtained from a parametric F test comparing nested linear models using R package edgeR, and a threshold of >1 in log1.3 |fold-change| or log2 |fold-change| when performing the GO analysis or all other analysis, respectively. A volcano plot was generated to visualize the differentially expressed genes between groups, and genes associated with the UPR were labeled. KEGG pathway enrichment analysis was performed. Canonical pathway analysis and upstream regulator predictions were performed using IPA (Qiagen).

Quantification and statistical analysis

The data are presented as the means ± SEMs. Statistical analyses were performed in Excel, R, and Prism. For comparisons of treatment groups, two-tailed Student’s t-test, one-way ANOVA or two-way ANOVA was performed. The data that did not adhere to a normal distribution were analyzed by a two-sided Mann‒Whitney U test. Mantel‒Cox was used for survival analysis. p < 0.05 was considered to indicate statistical significance: *p < 0.05, **p < 0.01, and ***p < 0.005. Details of the statistical analyses and “n” values are found in the Figure Legends. For in vivo experiments, “n” indicates the number of mice, and in vitro experiments, “n” indicates the number of biological replicates.