Human samples
Human samples were collected from 44 CRC patients: (20 with liver metastasis) and 20 HCC patients who underwent radical resection in the First Affiliated Hospital of Nanjing Medical University. The fresh tissue samples were collected immediately after tumor resection and cryopreserved in liquid nitrogen. Patients had not received any treatment including neoadjuvant radiotherapy, chemotherapy, or traditional Chinese medicine prior to collection, and had no other malignant tumors. The study was approved by the Institutional Ethics Committee of the First Affiliated Hospital of Nanjing Medical University. Informed consent for tissue analysis was obtained before the surgery. The clinical characteristics of patients with CRC are listed in Supplementary Table S1. All the research was performed in accordance with government policies and the Helsinki declaration.
Mice
Wild-type (WT), FloxP-RIG-I (RIG-IFL/FL), FloxP-Nlrp3 (Nlrp3FL/FL), Lyz2-Cre RIG-I knockout (RIG-IΔMØ), and Lyz2-Cre Nlrp3 knockout (Nlrp3ΔMØ) 6–8-week-old male mice on a C57BL/6 background were used in the experiments. The floxed allele was bred into homozygosity to generate RIG-IFL/FL and Nlrp3FL/FL mice. RIG-IΔMØ and Nlrp3ΔMØ mice were generated by crossing RIG-IFL/FL or Nlrp3FL/FL mice with Lyz2-Cre mice (all on C57BL/6 background). Littermates with floxed alleles without Cre were used as WT controls (RIG-IFL/ FL, Nlrp3FL/FL). Foxp3-DTR mice were purchased from the Gempharmatech Co., Ltd. Male C57BL/6 mice of 6-8 weeks were anesthetized, and a left upper abdominal transverse incision was made, the abdominal cavity was opened, and the spleen was separated and exposed in the ventral lateral direction. A 1 mL syringe was used to inject approximately 3 mm of tumor cells mixed with E.coli suspension from the lower pole of the spleen into the spleen capsule. The concentration of tumor cells in the suspension was 2×106/ml and of E.coli was 5×105 CFUs. When the spleen capsule was white and swollen at the injection site, the needle was pulled out, the blood was stopped by compression for 2 minutes, and the abdomen was closed layer by layer. Mice were divided into five groups: Control group (Ctrl); Antibiotic group (Abx): Antibiotic complex (ampicillin 1 g/L, vancomycin 500 mg/L, neomycin 1 g/L, metronidazole 1 g/L) was added to the drinking water of mice and mixed, and drinking was continued until the end of modeling; E.coli group (E.coli): 100 μL of E.coli (1×109 CFUs) was administered orally on day 8–14 of modeling; lactate dehydrogenase inhibitor (LDHi) group: gavage was performed at 6 mg/kg at day 7 and continued until the end of modeling; and clodronate liposome group: 200 μL was intraperitoneally injected one day before modeling and then administered at 3-day intervals until the end of modeling. Lactylation inhibitor C646 group: intraperitoneal injection was performed at 2.5, 5, or 10 μM at day 7 and continued until the end of modeling; and Class I histone deacetylases HDAC3 group: 10, 50, or 100 nM were intraperitoneally injected at day 7 and then until the end of modeling. Liver tissues were collected on Day 21, tumor number were recorded, and the sample size of each experimental group was 5. All the mice were housed under specific ventilated, pathogen-free, and thermostatic conditions with a 12-hour light-dark cycle at 24℃. Food and water were not limited. Then serum and tissues were harvested as described in a previous study13, 14, 15. All the animal studies were performed according to the guidelines of the Institutional Animal Use and the Animal Experimentation Ethics Committee of The First Affiliated Hospital of Nanjing Medical University. Natural small molecule compound combined with 5-FU treated mice model: In total 24 C57BL/6 mice bearing tumors were randomly divided into four groups (n=10 per group) and treated as follows: (1) Control group; (2) 5-FU group that received intraperitoneal injection of 5-FU (5 mg/kg) every other day; (3) 5-FU+compound group that received intraperitoneal injection of 5-FU (5 mg/kg) every other day and compound was administered i.p. daily on days 1–10 at 200 mg/kg. Doses and schedules were determined by tolerance studies.
Cell lines
Mouse colon cancer cell line MC38(ZQ0933) was purchased from the Shanghai Zhong Qiao Xin Zhou Biotechnology Co. Ltd (Shanghai, China). The E.coli: O6 (ATCC 25922) and mCherry-E.coli BS00F1-84 strain were purchased from Bio Sci Biotechnology Co. Ltd (Hangzhou, China). Tumor cells were maintained in RPMI 1640 (HyClone, SH30255, GEHealthcare, Chicago, IL) containing 10% (v/v) fetal calf serum (Fcs) (HyClone, SH3007003HI) and 1% (v/v) pen/strep (GIBCO, 15140-122) for 30 generations or more than 3 months before performing experiments. All cell lines in our laboratory are routinely tested for mycoplasma contamination, and the cells used in this study were negative for mycoplasma. None of our cell lines are on the list of commonly misidentified cell lines (International Cell Line Authentication Committee). Human colon tumor and liver tumor tissue samples were obtained from patients who underwent radical colon cancer surgery and partial hepatectomy for simultaneous colon cancer liver metastasis at the First Affiliated Hospital of Nanjing Medical University.
16S rDNA sequencing
In total 40-70 mg of tumor tissues were taken, and total DNA was extracted by the Cetyltrimethylammonium Bromide (CTAB) method, and then the quality and concentration of genomic DNA were checked using Qubit.
Absolute quantification-PCR (AQ-PCR) technology uses the standard product with known copy number to make a standard curve, by measuring the cycle threshold (Ct) value of the unknown sample, calculating the starting concentration of the sample from the standard curve, and realizing the absolute quantification of the measured sample gene mRNA, DNA molecular copy number. When assessing the sample, the standard underwent the PCR cycle simultaneously with the unknown sample, and the starting copy number of the unknown sample was obtained according to the Ct value of the unknown sample and combined with the standard curve. Ct refers to the number of amplification cycles passed when the fluorescence signal of the amplification product reaches the set threshold value during PCR amplification, which is reproducible. The fluorescence threshold was taken on the fluorescence signal from the first 15 cycles of the PCR reaction as the fluorescence background signal, and the general fluorescence threshold was defined as 10 times the standard deviation of the fluorescence signal of 3–15 cycles. The standard can be purified plasmid DNA, in vitro transcribed RNA, or ssDNA synthesized in vitro.
Copy number (X0) calculation method:
Ct=-K logX0+b
Note: k represents the standard curve slope, b represents the standard curve intercept=copy number per g sample =X0 * 50 uL (eluted volume) / weight assuming the standard curve is: y=-3.481x + 40.123 substitute Ct (18.45) into the linear equation 18.45=-3.481 X + 40.123 X=(18.45-40.123) / (-3.481)=6.226 X0=QuantityUnknown=10 6.226=1682674
Virtual screening of Specs database performed based on Lysine No. 852 (K852) on RIG-I protein
Approximately, 220,000 compound molecules in the Specs database and RIG-I protein in 6KYV were screened by high throughput virtual screening (HTVS), standard precision screening (SP) and high precision screening (XP). The docking results of the three docking modes were 10%, 10%, and 10% respectively. The results showed that 130 compound molecules were obtained from the Specs database, and 19 small molecules were selected based on the ligand-protein 2D map, Docking score, Glide energy, number of hydrogen bonds, drug-like parameters, and pharmacokinetic prediction results. The properties of intestinal absorption and water solubility of these 19 small molecules were predicted, and 3 of them were screened out with high property scores.
Kupffer cells extraction, culture, and grouping
The mice were anesthetized and opened, and the portal vein was separated with T-shaped fixation. There was tension in pulling the lower left adipose tissue moderately to straighten the portal vein. An intravenous indwelling needle (24 g) was inserted 3–4 mm in the portal vein while bending the tip naturally. Then, the needle was slowly withdrawn followed by clamping with hemostatic clips. The perfusion machine was opened, and CMF HBSS was lavaged at 5 mL/min. When the liver was swollen and white, the inferior vena cava was cut, and a large amount of blood was drained out. Lavage was continued with intermittent compression of the inferior vena cava to fill the liver 5–10 times. HBSS containing 0.05% collagenous IV (preheated at 37 °C) was infused at 5 mL/min, and the inferior vena cava was pressed intermittently. The liver was freed and put in a petri dish, and forceps were used to break the liver capsule. The cell suspension was filtered with 70 um grinding filter and centrifuged at 50×g for 3 min at 4°C. The rate of increase was 7, and the rate of decrease was 0. The supernatant was centrifuged at 500×g for 8 min and resuspended in 3 mL 1640 medium to precipitate. Layup: 15 mL centrifuge tube bottom 3 mL of 50% Percoll solution, middle 3 mL of 25% Percoll solution, top 3 mL of cell suspension. After centrifugation at 800×g for 15 min at 4 degrees, the ascending speed was 1, and the descending speed was adjusted to 0 by a brake. Cells in the 2/3 layer were absorbed into a 50mL centrifuge tube (i.e., cells in the 25% and 50% layers), filled with PBS, washed twice, and centrifuged at 500×g for 8 min. The 1640 resuspended precipitate was inoculated into the culture dish, and the non-adherent cells were removed by changing the solution 30 min later. Kupffer cells complete culture medium containing M-CSF was added and spread on a 6-well plate at a density of 1×106/mL. The culture medium was added and placed in an incubator with 5% CO2 at 37℃. The treatments were divided into 6 groups: Control group (Ctrl), lactate group (15 mmol/L Lac), RIG-IK852Ab group (15 mmol/L Lac+ RIG-IK852Ab 3.7 mg/mL), HMGB1K88Ab group (15 mmol/L Lac+HMGB1 K88Ab 3.8 mg/mL), HMGB1 K114Ab group (15 mmol/L Lac+HMGB1 K114Ab 1.6 mg/mL), and TNF-α group (15 mmol/L Lac +50 ng/mL TNF-α). The cells were incubated in a 5% CO2 incubator at 37℃ for 24 h, and then the proteins or RNA were collected for the next experiment.
Bone marrow-derived macrophage (BMDM) extraction
The mice were euthanized, and the femur and tibia were separated. The attached tissue on the bone was removed, and the bone marrow was flushed into a 50 mL centrifuge tube with DMEM complete medium and centrifuged for 5 min at 1250 rpm. Red blood cell lysate was added, mixed, and then centrifuged again.
Treg differentiation in vitro
Mice: We obtained suspensions of murine leukocytes from lymph nodes and spleens. Naive CD4+ T cells were then acquired by auto-MACS (Miltenyi, San Diego, CA, USA) based on their CD4+/CD62L+ surface marker. Then, naive T cells were activated in 96-well plates with completed media supplemented with anti-CD3/28 beads (Dynal beads, 1:1). For the differentiation of p-Tregs, cells were activated in the presence of IL-2 (100 U/mL, R&D Systems) and TGF-β (5 ng/mL, R&D Systems). The completed media is RPMI supplemented with 10% heat-inactivated FBS, 1% antibiotics, 1% HEPES and 0.1% 2-mercaptoethanol. Human: Peripheral blood (PB) leukapheresis products were obtained from volunteers at Nanjing Medical University. Naive human CD4+ T cells (CD4+CD25+) were sort-purified from PB mononuclear cells (PBMCs) (Ficoll-Hypaque, Amersham Biosciences) in a two-step procedure. p-Tregs were stimulated with anti-CD3/CD28 mAb-coated Dynabeads (Life Technologies, Carlsbad, CA) at 1:1 (cell to bead) ratios in the presence of recombinant IL-2 (100 U/mL, R&D Systems). TGF-β (5 ng/mL, R&D Systems) in X-Vivo-15 (BioWhittaker, Walkersville, MD) media supplemented with 10% human AB serum (Valley Biomedical) on day 0. Cells were counted and cultured at the concentration of 0.5×106 cells/mL, and IL-2 (300 U/mL) was renewed every 2 or 3 days. On point days (day 0, 3 or 6), cells were re-suspended at 0.5×106 cells/mL and renewed together with IL-2. Cells were harvested and assayed as listed.
Co-culture system
Macrophages MØ from WT mice, RIG-IΔMø mice or Nlrp3ΔMø mice were cultured for 48 h using completed media is DMEM supplemented with. supplementation with 10% heat-inactivated FBS, 1% antibiotics, 1% HEPES 0.1% 2-mercaptoethanol, IL-4 (20 ng/mL) and IL-13 (20 ng/mL) to induce differentiation to M2 macrophages. After obtaining a certain purity of M2 macrophages, 5×105/mL of M2 cells and 5×105/mL of naive T cells were co-cultured with or without Transwell, and Treg characteristic activation markers and functional markers were assayed after 3 days of co-culture.
Bacterial infection of MC38 cells
MC38 cells were cultured with RPMI complete medium. E.coli (5×105 CFUs) were co-cultured with 2×106 MC38 for 6 h in a 5% CO2 incubator at 37℃.
CCK8 experiment
MC38 cells were seeded in 96-well plates, and 100 uL of cell suspension (approximately 5×103 cells) was added to each well. The co-culture model of E.coli-infected MC38 cells was established as the experimental group, and MC38 cells without bacterial infection were used as the control group. Five multiple wells were set up in each group, and 10 μL CCK8 detection solution was added to each well, and the cells were incubated for 2 h. The light absorbance value was detected at 450 nm using a microplate reader.
ECAR
The above co-cultured cells were prepared into a cell suspension, and 80 μL of cells (approximately 1×104 cells) were seeded into a 96-well Seahorse plate per well. It stood at room temperature for 1 h, and then was put into the incubator for further cultivation for 23 h, 25 uL glucose, oligomycin and 2-deoxyribose were added to probe plate A, B and C, respectively. After correction of probe plate, cell culture plate was added for detection.
Determination of lactate concentration
The tumor tissue was prepared into tissue homogenate, and the enzyme working solution and chromogenic agent were prepared according to the instructions. The samples, enzyme working solution and chromogenic agent were added to the plate successively and incubated at 37 °C for 10 minutes, followed by the addition of terminator to terminate the reaction. The reaction solution was 250 μL, and the wavelength of microplate reader was set to 530 nm.
Histopathology
Liver tissues were excised and fixed in 10% formalin buffer and subsequently embedded in paraffin. Liver tumor tissue sections were stained with hematoxylin eosin (HE).
Immunofluorescence and confocal microscopy
After paraffin sections of liver tumor specimens were dewaxed, antigen repaired, and blocked with serum, the samples were incubated with primary antibodies at 4°C overnight and then incubated with corresponding HRP-labeled secondary antibodies for 50 min at 15–25°C. After washing with PBS three times, Bovine serum albumin was added, and the cells were incubated for 10 min at room temperature in the dark and then washed with Tris buffered saline with Tween-20 three times. 4',6-diamidino-2-phenylindole was dropped for counterstaining, the slices were sealed, and the images were observed and collected under a fluorescence microscope.
Measurement of lactate
The tumor tissue was prepared into tissue homogenate, and the supernatant was obtained by centrifugation. The enzyme working solution and chromogenic agent (Nanjing Jiancheng Institute of Biological Engineering) were added to the incubation supernatant. After adding the terminator, the absorbance was measured at the wavelength of 530 nm on the enzyme labeling instrument. The concentration of lactate in the tumor was calculated according to the absorbance.
MAVS aggregation assays
For in vitro MAVS aggregation, crude mitochondria were isolated, and RIG-I activation was detected as previously described16. Briefly, each 1 ml of mixture contained 100 ng GST-RIG-I(N) and 50–100 ng ubiquitin chains (K63–Ub4 from Boston Biochem UC-310B) in buffer containing 20 mM HEPES-KOH (pH 7.0) and 10% (v/v) glycerol. After incubation at RT for 10 min in total 10 μL reaction system, 1 μL of reaction mixture was mixed with 10 μg of mitochondrial fraction in 10 μL Buffer B (20 mM HEPES-KOH [pH 7.0], 5 mM MgCl2, and 0.25 M D-mannitol) at 30°C for 30 min. The mitochondria fraction was then pelleted at 10,000×g for 10 min and washed twice with Buffer C (20 mM HEPES-KOH at pH 7.4, 0.5 mM EGTA, 0.25 MD-mannitol, and EDTA-free protease inhibitor cocktail) and then subjected semi-denaturing detergent agarose gel electrophoresis.
Generating RIG-Iarg852 RAW264.7 cell lines through CRISPR-Cas9 genome engineering
To clarify that RIG-Ik852 is the key lactylation modification site that determines the direction of macrophage polarization, we built our editing strategy by using CRISPR-Cas9 system to introduce that mutation in an ATCC-certified RAW264.7 cell line. The overview of our protocol is shown in Figure 4F. First of all, we screened 2 sgRNAs (single guide RNAs) for Mouse Ddx58 (p.K852L) gene and selected sgRNA1(TACAACTGAAGGAACCCCACAGG) and sgRNA2(ACAACTGAAGGAACCCCACAGGG) as the candidate because a Cas9 protein with this sgRNA guide creates a site-specific double-strand break downstream of the genomic site of the target mutation. After the cell test is qualified, the passaging ability is normal, and the genotype test is qualified, the gRNA single chain is synthesized with Oligo, the cells of the formal project group were mixed with the system composed of gRNA, Cas9 protein complex and Oligo, and the cells of the positive control group were transferred into the EGFP plasmid. After 24 hours, the efficiency of the electrotransfection and whether there was any error in the process of electrotransfection were roughly judged according to the fluorescence expression of the positive control group. Twenty-four hours after electrotransfection, the cell poll was sent to detect the efficiency of gRNA and Oligo in the transfection, and the mutation peak was determined according to the sequencing peak map. If the cell pool test result was not satisfactory, the next step was determined after considering whether the power transfer operation, cell, and protocol were abnormal. Then, the cells obtained after monoclonal expansion of positive clones were screened.
CRISPR/Cas9 editing
A point mutation p.K852L (AAG to TTA) project was introduced into the mouse Ddx58 gene exon 18 of mouse monocyte macrophage leukemia cells RAW 264.7 using CRISPR/Cas9 technology. Single clones were selected after electrotransfection, and the homozygous cells with the point mutation p.K852L (AAG to TTA) in mouse Ddx58 gene exon 18 of mouse macrophage leukemia cells RAW 264.7 were obtained by PCR and sequencing. Single clones were selected after electric conversion, and the homozygous cells with point mutation p.K852L (AAG to TTA) in mouse Ddx58 gene exon 18 of mouse macrophage leukemia cells RAW 264.7 were obtained by PCR and sequencing.
RIG-I and MAVS: Protein−Protein Docking with RosettaDock
We used RosettaDock to obtain initial structures for the RIG-I−MAVS complexes. It employs a Monte Carlo (MC) search with low-resolution runs followed by high-resolution refinements. The starting orientations of RIG-I with respect to MAVS were generated by ZDOCK (https://zdock.umassmed.edu/). The docking protocol started with 15 000 steps of a low-resolution rigid-body MC search followed by 500 cycles of refinement. The best binding mode was selected based on Score and visualized with pymol2.2. The lactated Lys852 of RIG-I and MAVS binding mode were generated on wild-type binding mode.
STAT1 and MAVS: Protein−Protein Docking with RosettaDock
The method is similar with previous version.
qPCR
The treated cells were collected and RNA was extracted. The concentration and purity of RNA were detected by Nanodrop. After reverse transcription into cDNA, Q-PCR system (10 uL) was prepared according to the ratio of pure water: SYBR Mix: pre-primer: post-primer: cDNA= 3.6:5.0:0.2:0.2:1.0, and qPCR was performed. After the reaction, the data were copied for subsequent analysis.
Western blot analysis
The treated cells were collected, the radio immunoprecipitation assay lysate was added to extract total protein, and the Bicinchoninic Acid (BCA) method was used for quantification. The same amount of protein was taken for sodium dodecyl sulfate⁃polyacrylamide gel electrophoresis, membrane transformation, and 5% skim milk powder solution blocking for 1 h. The primary antibody (1:1000) was added and incubated at 4℃ overnight. After cleaning with TBST, HRP-labeled secondary antibody (1:2000) was added and incubated at room temperature for 2 h.
ELISA
IL-2, IL-10, IFN-γ and TGF-β levels were measured using commercially available enzyme-linked immunosorbent assay (ELISA) kits (BioLegend) according to the manufacturer’s instructions.
Surface and intracellular staining with flow cytometry
For intracellular staining of cytokines, cells were stimulated with phorbol myristate acetate (0.25 mg/mL) and ionomycin (0.25 mg/mL) for 5 hours and with brefeldin A (5 mg/mL) for 4 hours. Then, the surface markers, including CD4 and CD25, were stained for 30 min which was followed by further fixation, permeabilization, and stained with Foxp3, IL-10, and TGF-β.
CO-IP
The cells were collected and washed twice. The cells were lysed with RIPA lysate, and then the protein concentration was determined via BCA. The lysates were clarified with agarose resin for antibody fixation, and the protein content in the filtrate was analyzed after elution of the complex. Western Blotting analysis was performed to determine the expression of binding proteins.
Database analysis
The public STRING database collects the proteins and their sites that can undergo lactylation modification and shows the interaction between these proteins. With ‘F4/80’ as the keywords, we screened out some interacting proteins and their sites in the database and then selected three proteins to customize the specific antibodies for lactylation sites.
Virtual screening
The Specs database (~204,000 compounds) was used as the screening library. All compounds were first treated with Pipeline Pilot v7.5 (Accelrys) to generate 3D coordination, strip salt, minimize molecule energy, and standardize the chemical table coding. FILTER 2.1.1 from OpenEye was then applied to filter unfavorable compounds using the recommended filter criteria documented in filter_blockbuster.txt provided by OpenEye. Conformations were generated by OMEGA 2.4.6 using the optimized parameters with the FLIPPER module turned on. To explore the clinical potential of inhibiting the lactylation of RIG-I combined with the administration of chemotherapeutic drugs, we performed virtual screening of the Specs database based on RIG-IK852 and identified approximately 220,000 compounds. We then screened these using high-throughput virtual screening, standard precision screening, and high-precision screening docking methods. Compounds with 10% docking results in all three modes were retained. Overall 130 compounds were obtained. Nineteen compounds were selected based on the ligand-protein 2D map, docking score, glide energy, number of hydrogen bonds, drug-like parameters, and pharmacokinetic prediction results. Intestinal absorption and water solubility properties were predicted, and three compounds that could inhibit the lactylation of RIG-IK852 were identified
Statistical analyses
Experimental data were expressed as mean ± standard error (SD). T-test was used for comparison between two groups, and one-way analysis of variance (ANOVA) was used for comparison between multiple groups. P < 0.05 was considered statistically significant.