Marine antibacterial peptide based on machine learning mining and application thereof

Abstract

The invention belongs to the field of antibacterial peptides, and particularly relates to a marine antibacterial peptide based on machine learning mining and application thereof. The invention discloses a method for mining marine antibacterial peptides based on machine learning, and 10 marine antibacterial peptides are screened, and the amino acid sequence of the marine antibacterial peptides is shown as any one of SEQ ID NO: 1-10. Wherein the Peptide 5 and the Peptide 8 have a broad-spectrum antibacterial characteristic, and the drug-resistant bacteria MIC (Minimum Inhibitory Concentration) of the Peptide 5 and the Peptide 8 is as low as 2-8 mu g / mL and is superior to that of The Peptide 2 and the Peptide 7 belong to strain specific antibacterial peptides, and are highly specific to bacillus subtilis. The antibacterial peptide has a wide application prospect in the field of preparation of antibacterial dressings and antibacterial drugs.

Description

Technical Field

[0001] The present invention belongs to the field of antimicrobial peptides, and specifically relates to marine antimicrobial peptides mined based on machine learning and their applications. Background Art

[0002] Antimicrobial peptides (AMPs) are a class of natural defense molecules found widely in organisms, originating from plants, animals, and microorganisms. Marine organisms, a key source of AMPs, have evolved unique antimicrobial molecules due to their unique marine habitats. The shell mucus of marine mollusks is rich in defensin-like AMPs, while marine microorganisms, such as Bacillus and actinomycetes, secrete cyclic peptides with membrane-lytic activity.

[0003] Structurally, marine antimicrobial peptides usually have: 1. Amphiphilic structure: composed of positively charged amino acids and hydrophobic amino acids, which enables them to bind to negatively charged microbial cell membranes through electrostatic interactions; 2. Diversity of secondary structures: including α-helix (tachypleuronidase), β-fold (plant defensins) and random coil (some marine-derived linear peptides). The structural flexibility gives them diverse modes of action; 3. Sequence specificity: Some marine antimicrobial peptides have special functional domains, regions rich in proline and arginine (ascidian antimicrobial peptides) or disulfide bond cross-linking structures (sponge antimicrobial peptides), which can enhance cell penetrability or stability.

[0004] Functionally, antimicrobial peptides, as highly promising "new antibiotics", have unique antimicrobial mechanisms and relatively low drug resistance. Antimicrobial peptides not only directly kill pathogens such as bacteria, fungi, and viruses, but can also exert indirect antimicrobial effects by regulating the host immune response, promoting the secretion of inflammatory factors, and activating phagocytes. Unlike traditional antibiotics that act on a single target such as inhibiting bacterial metabolic enzymes or interfering with nucleic acid synthesis, the antimicrobial mechanism of antimicrobial peptides has multi-target and physical destructive characteristics. The mechanism of action of marine antimicrobial peptides is even more innovative. For example, the hybrid antimicrobial peptide Def-Prp from Pacific oysters has dual activities of membrane perforation and nucleic acid interference by fusing the sequences of defensins and proline-rich peptides. It exhibits high inhibitory ability against both Gram-negative bacteria and fungi, and its hemolytic toxicity is significantly lower than that of traditional antimicrobial peptides. The broad-spectrum antimicrobial activity of various marine antimicrobial peptides shows unique advantages against multidrug-resistant bacteria and is not easy to induce drug resistance. They are considered to be ideal candidates for the next generation of antibiotics.

[0005] Traditional antimicrobial peptide screening relies on chemical synthesis and biological activity testing, which is plagued by long lead times, high costs, and limited coverage. Machine learning technology significantly improves the efficiency and accuracy of antimicrobial peptide discovery through data integration, feature extraction, and intelligent optimization. With the advancement of algorithmic models (such as integrating graph neural networks for three-dimensional structure analysis) and the accumulation of ocean data, machine learning will become a core driver for discovering marine antimicrobial peptide patents, driving disruptive innovation in the biopharmaceutical field. Summary of the Invention

[0006] To identify novel broad-spectrum marine antimicrobial peptides with high antimicrobial activity, this study utilized a machine learning-driven screening strategy, integrating a global marine microbiome database with multi-dimensional prediction tools. Through redundancy removal and multi-model cross-validation (confidence level > 0.95), the team efficiently mined novel marine antimicrobial peptides. Ten highly active antimicrobial peptides were identified from 67,363 candidate peptides, covering Gram-negative and Gram-positive bacteria and fungi, addressing the time-consuming and limited coverage of traditional methods.

[0007] Based on the secondary structural characteristics of relevant antimicrobial peptides such as α-helix and random coil, as well as physical and chemical properties such as isoelectric point, net charge and hydrophobicity arithmetic mean, the present invention preferentially screens antimicrobial peptides with stable structures and amphipathic characteristics.

[0008] The present invention first provides a method for screening marine antimicrobial peptides based on machine learning, comprising: using six prediction tools, such as Antimicrobial Peptide Scanner v2 and AMPIify, to screen non-redundant antimicrobial peptides from the global marine microbiome database, and then filtering the results by deduplication and E-value (consistency ≥ 75%, E-value ≤ 10 -5 ) and multi-tool confidence screening (>0.95), and finally obtained the novel antimicrobial peptide sequence.

[0009] The present invention also provides a marine antimicrobial peptide mined based on machine learning, whose amino acid sequence is shown in any one of SEQ ID NOs: 1-10.

[0010] The names and sequence characteristics of the 10 marine antimicrobial peptides are shown below:

[0011] Peptide1: EAIFFKKKKKRVAKKVVKIARRFQGTLWDLMPPMMA (SEQ ID NO: 1);

[0012] Peptide2: VIGRNICKISKKKLLFFPNYGHRMKLKIGVNMA (SEQ ID NO: 2);

[0013] Peptide3: ILAGRAINDHMPKHVAQMMIKGLNEVGKVIKGS (SEQ ID NO: 3);

[0014] Peptide4: AGYSKTPLVKKLGIKENFKVRLIHPPANFLDLG (SEQ ID NO: 4);

[0015] Peptide5:AALKAGKIAIFTYYPMAGHHNGKRIFRQRLPHST (SEQ ID NO: 5);

[0016] Peptide6: IIGKRIVRVKKKKLCAASGNGAGLIKGSEKPPSFEY SEQ ID NO: 6);

[0017] Peptide7:AGSLGKVPKNLQEKMVSHFTKADPAYGKGVAKSLGLV (SEQ ID NO: 7);

[0018] Peptide8: LVSFKCFFEKLVSNEVGKVSKKGPKVLHSGLVNLG (SEQ ID NO: 8);

[0019] Peptide9: IIVVSGTPGTGKTTLSKKLAKNLGYKYLDVAGL (SEQ ID NO: 9);

[0020] Peptide10: VLCGKYVLIIKLFFHRKGADAGSTIGHPNVELTR (SEQ ID NO: 10).

[0021] Preferably, the amino acid sequence of the marine antimicrobial peptide is shown in SEQ ID NO: 2, 5, 7 or 8.

[0022] The present invention also provides the use of the marine antimicrobial peptide in preparing medical antimicrobial dressings or antimicrobial drugs.

[0023] Furthermore, the medical antibacterial dressing or antibacterial drug is used to inhibit one or more of the following bacteria: Escherichia coli, Bacillus subtilis, Klebsiella pneumoniae, and Staphylococcus epidermidis.

[0024] The present invention also provides a medical antibacterial dressing, the active ingredient of which is one or more marine antibacterial peptides.

[0025] Furthermore, the medical antibacterial dressing contains at least one marine antibacterial peptide as follows: Peptide 5 with an effective concentration of ≥2 μg / mL, Peptide 8 with an effective concentration of ≥8 μg / mL, Peptide 2 with an effective concentration of ≥8 μg / mL, and Peptide 7 with an effective concentration of ≥8 μg / mL.

[0026] The present invention also provides an antibacterial drug, the active ingredient of which is one or more of the marine antibacterial peptides.

[0027] Furthermore, the antibacterial drug contains at least one marine antibacterial peptide as follows: Peptide 5 with an effective concentration of ≥2 μg / mL, Peptide 8 with an effective concentration of ≥8 μg / mL, Peptide 2 with an effective concentration of ≥8 μg / mL, and Peptide 7 with an effective concentration of ≥8 μg / mL.

[0028] The present invention also provides the use of the antimicrobial peptide in the preparation of antimicrobial products such as disinfectants, detergents and feed additives.

[0029] Furthermore, the disinfectant, detergent or feed additive contains Peptide 2 at an effective concentration of ≥8 μg / mL, and / or Peptide 7 at an effective concentration of ≥8 μg / mL.

[0030] Compared with the prior art, the beneficial effects of the present invention include at least:

[0031] The present invention screened 10 antimicrobial peptides that can inhibit the growth of a variety of bacteria; among them, Peptide 5 and Peptide 8 have broad-spectrum antimicrobial properties, with MICs as low as 2-8 μg / mL against drug-resistant bacteria, which are superior to traditional antibiotics and suitable for use as medical dressings or drug development; among them, Peptide 2 and Peptide 7 are species-specific antimicrobial peptides, which are specific and highly effective against Bacillus subtilis (Gram-positive), and are suitable for the development of antimicrobial drugs and antimicrobial products (such as disinfectants and detergents) for specific scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 Structural prediction of antimicrobial peptides containing α-helical domains.

[0033] Figure 2 Structure prediction of antimicrobial peptides containing random coil domains. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solutions and beneficial effects of the present invention more clear, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. Examples of the embodiments are shown in the accompanying drawings. It should be understood that the specific embodiments described in the following embodiments of the present invention are only exemplary descriptions of the specific embodiments of the present invention, intended to be used for explaining the present invention, and do not constitute a limitation of the present invention.

[0035] The endpoints of ranges and any values ​​disclosed herein are not limited to the exact range or value, and these ranges or values ​​should be understood to include approximations to these ranges.

[0036] Example 1: Screening of antimicrobial peptides

[0037] In order to obtain highly active marine antimicrobial peptides, the present invention uses six prediction tools, including Antimicrobial Peptide Scanner v2, AMPlify and amPEPpy, to predict 67,363 new non-redundant antimicrobial peptides from the Global Marine Microbiome Database and establish a comprehensive antimicrobial peptide database. The Global Marine Microbiome Database is the largest marine resource database to date, containing 24,195 species-level non-redundant genomes, of which nearly 10,000 genomes mainly come from marine habitats such as the deep sea, sediments, and symbiotic organisms. First, Prodigal is used to predict short open reading frames (smORFs, 33-303 base pairs) from overlapping groups. The redundant smORFs are then deduplicated, and Macrel is used to predict the deduplicated smORFs. Except for those singleton sequences that have significant matches with sequences in the Antimicrobial Peptide Data Repository (DRAMP) (amino acid identity ≥ 75%, E value ≤ 10 -5 ), and all other singleton sequences were removed. Finally, 67,363 non-redundant predicted antimicrobial peptides were obtained. Based on the results of multiple prediction software, 10 antimicrobial peptide sequences with a confidence level greater than 0.95 were selected for functional verification.

[0038] The antimicrobial peptide sequence is shown below:

[0039] Peptide1: EAIFFKKKKKRVAKKVVKIARRFQGTLWDLMPPMMA (SEQ ID NO: 1);

[0040] Peptide2: VIGRNICKISKKKLLFFPNYGHRMKLKIGVNMA (SEQ ID NO: 2);

[0041] Peptide3: ILAGRAINDHMPKHVAQMMIKGLNEVGKVIKGS (SEQ ID NO: 3);

[0042] Peptide4: AGYSKTPLVKKLGIKENFKVRLIHPPANFLDLG (SEQ ID NO: 4);

[0043] Peptide5:AALKAGKIAIFTYYPMAGHHNGKRIFRQRLPHST (SEQ ID NO: 5);

[0044] Peptide6: IIGKRIVRVKKKKLCAASGNGAGLIKGSEKPPSFEY SEQ ID NO: 6);

[0045] Peptide7:AGSLGKVPKNLQEKMVSHFTKADPAYGKGVAKSLGLV (SEQ ID NO: 7);

[0046] Peptide8: LVSFKCFFEKLVSNEVGKVSKKGPKVLHSGLVNLG (SEQ ID NO: 8);

[0047] Peptide9: IIVVSGTPGTGKTTLSKKLAKNLGYKYLDVAGL (SEQ ID NO: 9);

[0048] Peptide10: VLCGKYVLIIKLFFHRKGADAGSTIGHPNVELTR (SEQ ID NO: 10).

[0049] Example 2: Physicochemical properties and structure prediction of antimicrobial peptides

[0050] The tertiary structure of the 10 antimicrobial peptides screened out was predicted using an online model prediction tool. Figure 1 and Figure 2 As shown in Figure 1, Peptide 1, Peptide 3, Peptide 4, Peptide 7, and Peptide 9 all contain α-helical structures and are typical α-helical peptides. Peptide 2, Peptide 5, Peptide 6, Peptide 8, and Peptide 10 all contain random coil structures and are typical linear extended peptides. The physicochemical properties of the 10 antimicrobial peptides described in the present invention are shown in Table 1, including the number of amino acids, extinction coefficient, molecular weight, isoelectric point, net charge, and arithmetic mean hydrophobicity. Many of the antimicrobial peptides exhibit amphiphilic properties.

[0051] Table 1 Physicochemical properties of antimicrobial peptides

[0052] name Number of amino acids <![CDATA[Extinction coefficient (M -1 cm -1 )]]> Molecular weight (MW) Isoelectric point Net charge GRAVY Peptide1 36 5500 4292.32 12.34 9.98 -0.28 Peptide2 33 1280 3818.71 10.74 11.18 0.02 Peptide3 33 0 3557.25 10.55 4.45 0.04 Peptide4 34 1280 3778.47 10.46 6.21 -0.02 Peptide5 34 2560 3852.46 11.51 9.68 -0.39 Peptide6 36 1280 3844.56 10.68 9.94 -0.31 Peptide7 37 1280 3827.43 10.35 6.21 -0.17 Peptide8 35 0 3789.47 10.27 6.18 0.23 Peptide9 33 2560 3406.99 10.28 6.97 0.22 Peptide10 34 1280 3754.39 10.05 6.42 0.26

[0053] GRAVY: The arithmetic mean of hydrophobicity. Negative values ​​represent hydrophilicity. The smaller the value, the more hydrophilic it is.

[0054] Example 3: Broad-spectrum antibacterial application of antimicrobial peptides

[0055] All 10 antimicrobial peptides in Example 1 were synthesized by solid-phase chemistry at Sangon Biotech (Shanghai) Co., Ltd. to obtain antimicrobial peptide samples with a purity greater than 95%, which were then used to determine their antibacterial effects. The minimum inhibitory concentration (MIC) assay used multiple bacterial strains, including two Gram-negative bacterial standard strains, Escherichia coli ATCC25922 and Bacillus subtilis ATCC6633, and two Gram-positive bacterial standard strains, Klebsiella pneumoniae ATCC13883 and Staphylococcus epidermidis ATCC35984. The MIC of the antimicrobial peptides was determined using the broth microdilution method, as follows:

[0056] (1) Strain activation and bacterial suspension preparation: The above strains were inoculated into LB solid medium and cultured at 37°C for 18-24 hours. Single colonies were picked and suspended in physiological saline and adjusted to a McFarland turbidity of 0.5 (approximately 1×10 8 CFU / mL). Diluted with MHB medium to a final concentration of 1×10 7 CFU / mL was used for MIC determination.

[0057] (2) Antimicrobial peptide dilution: Prepare the antimicrobial peptide stock solution (128 μg / mL) in MHB medium and filter sterilize. Use the two-fold dilution method to dilute the antimicrobial peptide in a 96-well plate: add 100 μL of stock solution to the first well, and then add 50 μL of medium + 50 μL of the previous well solution to each well. The final concentrations are 64, 32, 16, 8, 4, 2, 1, 0.5, 0.25, and 0.125 μg / mL, respectively. Set up 3 replicates for each concentration, with the 12th column as the negative control (medium only) and the 1st column as the positive control (bacteria + medium).

[0058] (3) Inoculation and culture: Add 100 μL of bacterial suspension to each well (final inoculation volume 1×10 5 After mixing, the plate was sealed with plastic wrap and incubated at 37°C for 16-20 hours.

[0059] (4) Reading results: The bacterial growth in the 96-well plate was directly observed with the naked eye. The lowest concentration of the antimicrobial peptide solution in which no bacteria grew in the well was the minimum inhibitory concentration of the antimicrobial peptide under these conditions. The results are shown in Table 2.

[0060] antimicrobial peptides E. coli B. subtilis K.pneumoniae S.epidermidis Peptide1 8 16 32 64 Peptide2 4 8 16 32 Peptide3 64 >64 64 >64 Peptide4 16 32 64 128 Peptide5 2 4 8 16 Peptide6 32 16 32 64 Peptide7 16 8 16 32 Peptide8 8 8 8 8 Peptide9 4 16 32 64 Peptide10 64 >64 >64 >64

[0061] Table 2: Minimum inhibitory concentration of antimicrobial peptides against four strains (μg / mL)

[0062] Among them, Peptide 5 and Peptide 8 have broad-spectrum antimicrobial peptide properties, the lowest MIC, and are highly effective against the four strains, especially against the highly resistant K. pneumoniae ATCC 13883; they exhibit broad-spectrum activity and may be suitable for development as broad-spectrum antimicrobial drugs.

[0063] Peptide 2 and Peptide 7 are species-specific antimicrobial peptides that are highly effective in inhibiting Bacillus subtilis (Gram-positive) but less effective against other bacteria.

[0064] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and do not constitute a limitation of the present invention. Within the technical concept of the present invention, the technical solutions of the present invention may be subjected to various simple modifications, including combining the various technical features in any other appropriate manner. These simple modifications and combinations should also be considered as disclosed in the present invention and fall within the scope of protection of the present invention.

Claims

1. Marine antimicrobial peptides mined based on machine learning, characterized by: The amino acid sequence thereof is shown in any one of SEQ ID NOs: 1-10.

2. The marine antimicrobial peptide according to claim 1, characterized in that The amino acid sequence of the marine antimicrobial peptide is shown in SEQ ID NO: 2, 5, 7 or 8.

3. Use of the marine antimicrobial peptide according to claim 1 or 2 in the preparation of medical antimicrobial dressings or antimicrobial drugs.

4. The use according to claim 3, characterized in that The medical antibacterial dressing or antibacterial drug is used to inhibit one or more of the following bacteria: Escherichia coli, Bacillus subtilis, Klebsiella pneumoniae, and Staphylococcus epidermidis.

5. A medical antibacterial dressing, characterized in that: The active ingredient is at least one marine antimicrobial peptide according to claim 1 or 2.

6. The medical antibacterial dressing according to claim 5, characterized in that: The medical antibacterial dressing contains at least one marine antibacterial peptide as follows: Peptide 5 with an effective concentration of ≥2 μg / mL, Peptide 8 with an effective concentration of ≥8 μg / mL, Peptide 2 with an effective concentration of ≥8 μg / mL, and Peptide 7 with an effective concentration of ≥8 μg / mL; The amino acid sequence of Peptide 5 is shown in SEQ ID NO: 5, the amino acid sequence of Peptide 8 is shown in SEQ ID NO: 8, the amino acid sequence of Peptide 2 is shown in SEQ ID NO: 2, and the amino acid sequence of Peptide 7 is shown in SEQ ID NO:

7.

7. An antibacterial drug, characterized in that: The active ingredient is at least one marine antimicrobial peptide according to claim 1 or 2.

8. The antibacterial drug according to claim 7, wherein The antimicrobial drug contains at least one marine antimicrobial peptide as follows: Peptide 5 with an effective concentration of ≥2 μg / mL, Peptide 8 with an effective concentration of ≥8 μg / mL, Peptide 2 with an effective concentration of ≥8 μg / mL, and Peptide 7 with an effective concentration of ≥8 μg / mL; The amino acid sequence of Peptide 5 is shown in SEQ ID NO: 5, the amino acid sequence of Peptide 8 is shown in SEQ ID NO: 8, the amino acid sequence of Peptide 2 is shown in SEQ ID NO: 2, and the amino acid sequence of Peptide 7 is shown in SEQ ID NO:

7.

9. Use of the marine antimicrobial peptide according to claim 1 or 2 in the preparation of disinfectants, detergents and feed additives.

10. The use according to claim 9, characterized in that The disinfectant, detergent or feed additive contains Peptide 2 at an effective concentration of ≥8 μg / mL and / or Peptide 7 at an effective concentration of ≥8 μg / mL; the amino acid sequence of Peptide 2 is shown in SEQ ID NO: 2, and the amino acid sequence of Peptide 7 is shown in SEQ ID NO: 7.

Images