Electrophoresis is a technique that uses an electric field to cause charged particles (such as DNA, RNA, or proteins) to migrate through a gel. Smaller molecules or molecules that are more heavily charged migrate faster.

MFG-E8 does not have a publicly available crystal structure, so Alphafold3 prediction was used. In this study, we employed a semi flexible docking method. Firstly, we downloaded the crystal structure of the target protein MFG-E8 from a protein database or modeled it using AlphaFold3. Then, we used Maestro computational software to prepare the structures of proteins and small molecules separately. During the protein preparation process, optimize the hydrogen bond network and minimize the energy of the protein system. Small molecule preparation: Generate three-dimensional structures and minimize energy, and generate multiple conformations (15) to increase the likelihood and accuracy of successful docking. Next, we use the sitemap module of Maestro software to predict the binding pocket of small molecules (default parameters) and generate a receiver grid file. Based on the predicted pocket location, size, and small molecule docking score, we determine the most likely binding pocket. The other parameters remained at their default values. During molecular docking, select the prepared grid file and small molecules in the small molecule docking ligand docking module for molecular docking. Based on the scoring of docking results and the rationality of small molecule binding conformation, the optimal binding mode is selected through comprehensive analysis and judgment. The docking results were visualized using Pymol software. The human sequence of MFG-E8 was submitted to the server along with oleic acid, and the generated computational model was used for molecular docking and MM-GBSA. Oleic acid combining pattern analysis: The binding mode of the three molecules is the same. The carboxyl group interacts with GLN314 through hydrogen bonding and with ARG377 through salt bridge interaction (also known as electrostatic interaction). The alkyl chain of the molecule occupies the hydrophobic pocket composed of MFG-E8 proteins PHE260, ASN274, ASN275, and ASN309. Linoleic acid combined with pattern analysis: The binding mode of the three molecules is the same. The carboxyl group interacts with GLN314 through hydrogen bonding and with ARG377 through salt bridge interaction (also known as electrostatic interaction). The alkyl chain of the molecule occupies the hydrophobic pocket composed of MFG-E8 proteins PHE260, ASN274, ASN275, and ASN309. Linolenic acid combined with pattern analysis: The binding mode of the three molecules is the same. The carboxyl group interacts with GLN314 through hydrogen bonding and with ARG377 through salt bridge interaction (also known as electrostatic interaction). The alkyl chain of the molecule occupies the hydrophobic pocket composed of MFG-E8 proteins PHE260, ASN274, ASN275, and ASN309.

For MFG-E8 is a soluble protein that can be expressed in Escherichia coli (E. coli), which is designed as follows. Predicted locationi Secreted, Intracellular (different isoforms)
Introduction of the protein: https://www.proteinatlas.org/ENSG00000140545-MFGE8
mRNA Nucleotide Sequence (1008 nt):
ATGCCGCGCCCCCGCCTGCTGGCCGCGCTGTGCGGCGCGCTGCTCTGCGCCCCCAGCCTCCTCGTCGCCC
TGGATATCTGTTCCAAAAACCCCTGCCACAACGGTGGTTTATGCGAGGAGATTTCCCAAGAAGTGCGAGG
AGATGTCTTCCCCTCGTACACCTGCACGTGCCTTAAGGGCTACGCGGGCAACCACTGTGAGACGAAATGT
GTCGAGCCACTGGGCCTGGAGAATGGGAACATTGCCAACTCACAGATCGCCGCCTCGTCTGTGCGTGTGA
CCTTCTTGGGTTTGCAGCATTGGGTCCCGGAGCTGGCCCGCCTGAACCGCGCAGGCATGGTCAATGCCTG
GACACCCAGCAGCAATGACGATAACCCCTGGATCCAGGTGAACCTGCTGCGGAGGATGTGGGTAACAGGT
GTGGTGACGCAGGGTGCCAGCCGCTTGGCCAGTCATGAGTACCTGAAGGCCTTCAAGGTGGCCTACAGCC
TTAATGGACACGAATTCGATTTCATCCATGATGTTAATAAAAAACACAAGGAGTTTGTGGGTAACTGGAA
CAAAAACGCGGTGCATGTCAACCTGTTTGAGACCCCTGTGGAGGCTCAGTACGTGAGATTGTACCCCACG
AGCTGCCACACGGCCTGCACTCTGCGCTTTGAGCTACTGGGCTGTGAGCTGAACGGATGCGCCAATCCCC
TGGGCCTGAAGAATAACAGCATCCCTGACAAGCAGATCACGGCCTCCAGCAGCTACAAGACCTGGGGCTT
GCATCTCTTCAGCTGGAACCCCTCCTATGCACGGCTGGACAAGCAGGGCAACTTCAACGCCTGGGTTGCG
GGGAGCTACGGTAACGATCAGTGGCTGCAGATCTTCCCTGGCAACTGGGACAACCACTCCCACAAGAAGA
ACTTGTTTGAGACGCCCATCCTGGCTCGCTATGTGCGCATCCTGCCTGTAGCCTGGCACAACCGCATCGC
CCTGCGCCTGGAGCTGCTGGGCTGTTAG
Translation (335 aa):
MPRPRLLAALCGALLCAPSLLVALDICSKNPCHNGGLCEEISQEVRGDVFPSYTCTCLKGYAGNHCETKC
VEPLGLENGNIANSQIAASSVRVTFLGLQHWVPELARLNRAGMVNAWTPSSNDDNPWIQVNLLRRMWVTG
VVTQGASRLASHEYLKAFKVAYSLNGHEFDFIHDVNKKHKEFVGNWNKNAVHVNLFETPVEAQYVRLYPT
SCHTACTLRFELLGCELNGCANPLGLKNNSIPDKQITASSSYKTWGLHLFSWNPSYARLDKQGNFNAWVA
GSYGNDQWLQIFPGNWDNHSHKKNLFETPILARYVRILPVAWHNRIALRLELLGC
Expression vector: pET28A HIS tag at the C-terminus of MFG-E8. NcoI+xhoI enzyme digestion was performed, and the plasmids were cloned into pET283a and Dh5a strains for amplification; Transfer to protein expression using E. coli.

MFG-E8 does not have a publicly available crystal structure, so Alphafold3 prediction was used. In this study, we employed a semi flexible docking method. Firstly, we downloaded the crystal structure of the target protein MFG-E8 from a protein database or modeled it using AlphaFold3. Then, we used Maestro computational software to prepare the structures of proteins and small molecules separately. During the protein preparation process, optimize the hydrogen bond network and minimize the energy of the protein system. Small molecule preparation: Generate three-dimensional structures and minimize energy, and generate multiple conformations (15) to increase the likelihood and accuracy of successful docking. Next, we use the sitemap module of Maestro software to predict the binding pocket of small molecules (default parameters) and generate a receiver grid file. Based on the predicted pocket location, size, and small molecule docking score, we determine the most likely binding pocket. The other parameters remained at their default values. During molecular docking, select the prepared grid file and small molecules in the small molecule docking ligand docking module for molecular docking. Based on the scoring of docking results and the rationality of small molecule binding conformation, the optimal binding mode is selected through comprehensive analysis and judgment. The docking results were visualized using Pymol software. The human sequence of MFG-E8 was submitted to the server along with oleic acid, and the generated computational model was used for molecular docking and MM-GBSA. Oleic acid combining pattern analysis: The binding mode of the three molecules is the same. The carboxyl group interacts with GLN314 through hydrogen bonding and with ARG377 through salt bridge interaction (also known as electrostatic interaction). The alkyl chain of the molecule occupies the hydrophobic pocket composed of MFG-E8 proteins PHE260, ASN274, ASN275, and ASN309. Linoleic acid combined with pattern analysis: The binding mode of the three molecules is the same. The carboxyl group interacts with GLN314 through hydrogen bonding and with ARG377 through salt bridge interaction (also known as electrostatic interaction). The alkyl chain of the molecule occupies the hydrophobic pocket composed of MFG-E8 proteins PHE260, ASN274, ASN275, and ASN309. Linolenic acid combined with pattern analysis: The binding mode of the three molecules is the same. The carboxyl group interacts with GLN314 through hydrogen bonding and with ARG377 through salt bridge interaction (also known as electrostatic interaction). The alkyl chain of the molecule occupies the hydrophobic pocket composed of MFG-E8 proteins PHE260, ASN274, ASN275, and ASN309.

The treated cells in Part 1 were collected and treated with lysis buffer, 300 μL of the 5× loadding buffer was added into each cell lysis sample, and we boiled and denatured the samples for 15 min. To quantify protein expressions of MFG-E8, associated proteins of mTOR, TNF-α and IL-1β, with respect to the control protein β-actin, western blot experiment was performed. 1× Tris-glycine electrophoresis buffer was filled into the electrophoresis unit, samples and pre-stained rainbow protein marker were loaded onto the gel plates and electrophoresed initially at 110V for 20 min, then at 135V for about 30 min until the protein marker was completely separated. The PVDF membranes were prepared by 40 s immersion in anhydrous methanol. The membranes, sponges, and filter paper were submerged in a 1× transfer buffer, and the assembly sequences within the transfer apparatus was organized from the negative to positive pole as follows: sponge, filter paper, gel, membrane, filter paper, and sponge, and the membranes then underwent a 4W transfer for 1.5 h. After transfer, the membranes were rinsed and incubated in 10% skimmed milk for an hour. The membranes were then washed using 1× TBST with 5 min for three times, which were further incubated with the primary antibodies for two hours. Later, the membranes were washed using 1× TBST buffer for three times, and next incubated with goat anti-rabbit IgG-HRP and 1× TBST for 1 h. After three times of 1× TBST washing, 1 mL ECL developer for the membrane was prepared in darkness and added to the membrane, and chemiluminescence signals were captured with a designated imaging system, and the relative amounts of proteins in each group were quantified concerning β-actin expression. Results demonstrated that the treatment of MFG-E8 protein complex with the concentration of 1 μM significantly alleviated inflammatory reaction, embodying on the down-regulation of inflammatory factors, including TNF-α and IL-1β proteins.

All data are presented as mean ± standard deviation (SD). Statistical analyses were performed using SPSS version 19.0 and GraphPad Prism version 7.0 (GraphPad Software Inc., San Diego, CA, USA). Comparisons between control and intervention groups were made using t-tests, with statistical significance set at p < 0.05.

Oleanolic acid (OA) nanoparticles were prepared by ultrasonic emulsification method, and the morphology was observed by SEM. Overexpressed MFG-E8 protein sample was successfully prepared into OA nanoparticles. Agilent 1290 liquid chromatography was used to detect encapsulated protein samples, with the following liquid phase conditions: Mobile phase: 0.5% formic acid water (A) - acetonitrile (B), flow rate of 0.3 mL/min Gradient elution conditions: 0-1 min, 30% B; 1-3 min, 30-80% B; 3-5 min，80-98% B; 5-6 min，98% B; 6-7.5 min, 98-30% B; 7.5-9.5 min, 30% B; 9.5 min, stop. Injection volume: 2 μL, column temperature: 35 ℃, detection wavelength: 350 nm. Production of chemical tablets: the concentration of OA embedded MFG-E8 protein was 1 μM, the accessory ingredient was natural peppermint extract.