A paracasei subsp. that has a significant hypoglycemic effect

Abstract

This invention relates to the field of biomedical technology, and more particularly to a *Lactobacillus paracasei* strain with significant hypoglycemic effects. This invention develops an engineered bacterium that continuously secretes a mutant GLP-1 protein, capable of expressing a fusion peptide containing a signal peptide, an enhancing peptide, and the mutant GLP-1 protein. It exhibits good passage stability and can exert a dual therapeutic effect on type 2 diabetes through continuous secretion of the mutant GLP-1 protein and regulation of the gut microbiota. The engineered bacterium of this invention can be used to prepare oral pharmaceutical formulations. Therefore, this invention provides a new technical means for developing non-invasive, cost-effective, and efficient drug delivery strategies targeting GLP-1, and has broad application prospects.

Description

Technical Field

[0001] This invention relates to the field of biomedical technology, and in particular to a Lactobacillus paracasei with a significant hypoglycemic effect. Background Technology

[0002] The global burden of diabetes continues to increase. Of all people with diabetes, over 90% have type 2 diabetes (T2DM). As a core contributing factor to a variety of serious complications, type 2 diabetes can lead to cardiovascular disease, retinopathy, dementia, and other conditions.

[0003] Currently, the management of type 2 diabetes mellitus (T2DM) relies on a combination of lifestyle modifications and drug therapies such as metformin and SGLT2 inhibitors, but clinical outcomes are often unsatisfactory. The efficacy of lifestyle interventions is frequently limited by genetic heterogeneity and socioeconomic barriers, while patient adherence to medications is impaired by gastrointestinal side effects and the burden of injection regimens. Although GLP-1 receptor agonists (GLP-1RAs), such as semaglutide and telpoxetine, offer superior glycemic control and weight loss benefits, their widespread use is limited by invasive administration methods, low oral bioavailability, and tolerability issues. Therefore, developing non-invasive, cost-effective administration strategies is crucial to overcoming these obstacles and fully realizing the therapeutic potential of GLP-1.

[0004] The pharmacological activity of GLP-1 stems from its activation of the glucagon-like peptide-1 receptor (GLP-1R) and subsequent recruitment of the Gαs subunit; the recruited Gαs subunit further initiates the cAMP / PKA signaling cascade. In pancreatic β cells, this pathway enhances glucose-stimulated insulin secretion (GSIS) by regulating KATP and voltage-gated calcium channel dynamics, ultimately promoting exocytosis of insulin granules. In addition to its acute secretory activity, GLP-1R signaling also converges with the PI3K / AKT axis, resulting in durable metabolic benefits: this cross-interaction maintains the functional quality of β cells and improves insulin resistance in key metabolic reservoirs, particularly the liver and skeletal muscle, thereby optimizing systemic glucose homeostasis.

[0005] Lactic acid bacteria (LAB) are ideal hosts for oral protein delivery due to their inherent tolerance to gastrointestinal stress (acid and bile) and their ability to colonize mucosa. Recent advances in synthetic biology have transformed LAB from traditional probiotics into engineered in vivo therapies. This use has been validated in the field of metabolism; for example, recombinant Lactococcus lactis secreting proinsulin and IL-10 successfully restored glucose homeostasis in a diabetic model. In oncology, the efficacy of Lactobacillus plantarum-mediated delivery of antitumor peptides has also been demonstrated. Furthermore, as probiotics with multiple beneficial functions, LAB can also assist in the healing process by regulating the gut microbiota and physiological metabolism. This functional plasticity makes engineered LAB a promising platform for the non-invasive, targeted delivery of metabolic regulators such as GLP-1. Summary of the Invention

[0006] Previous research by the inventors on *Lactobacillus paracasei* A4-2 showed that it could produce glucagon-like peptide-1 (GLP-1) intracellularly, but due to the lack of a secretion mechanism, it could not interact therapeutically with the host receptor. To construct a secretory chassis, this invention genetically engineered *Lactobacillus paracasei* through chromosome integration, fusing signal peptides and enhancing peptides with a mutant GLP-1 protein (mGLP-1) to form an expression cassette. This genetic modification conferred a secretory phenotype on the strain, enabling it to continuously release bioactive GLP-1 protein extracellularly. After verifying genome integration and peptide secretion, this invention passaged the strain 50 times consecutively, obtaining a stable derivative strain LCglp1v1-50. Subsequently, its in vivo efficacy, including intestinal colonization and glycemic control, was evaluated in a type 2 diabetes (T2DM) mouse model. It was found that the engineered bacteria expressing the fusion peptide could exert a dual therapeutic effect on type 2 diabetes by continuously secreting mutant GLP-1 protein and regulating the gut microbiota. Based on this, the following technical solution is proposed.

[0007] First, the present invention provides a fusion peptide containing a signal peptide, an enhancing peptide, and a mutant GLP-1 protein; the mutant GLP-1 protein has mutations at the Ala8Gly, Lys26Gln, and Lys34Asp sites on the basis of the wild-type GLP-1 protein.

[0008] Preferably, the amino acid sequence of the wild-type GLP-1 protein is shown in SEQ ID No. 1.

[0009] Preferably, the amino acid sequence of the mutant GLP-1 protein is shown in SEQ ID No. 2.

[0010] Preferably, the nucleotide sequence encoding the wild-type GLP-1 protein is shown in SEQ ID No. 3.

[0011] Preferably, the nucleotide sequence encoding the mutant GLP-1 protein is shown in SEQ ID No. 4.

[0012] SEQ ID No.1:HAEGTFTSDVSSYLEGQAAKEFIAWLVKGR SEQ ID No.2: HGEGTFTSDVSSYLEGQAAQEFIAWLVDGR SEQ ID No.3: CATGCTGAAGGCACATTTACATCAGATGTTTCATACTTGGAGGGCCAAGCTGCTAAAGAATTTATTGCTTGGTTGGTCAAGGGCCGT SEQ ID No.4: CATGGCGAAGGCACATTTACATCAGATGTTTCATACTTGGAGGGCCAAGCTGCTCAAGAATTTATTGCTTGGTTGGTTGATGGCCGT Preferably, the signal peptide is the Usp45 signal peptide; and / or, the amino acid sequence of the enhancing peptide is LEISSTCDA (SEQ ID No. 6).

[0013] Preferably, the amino acid sequence of the Usp45 signal peptide is shown in SEQ ID No. 5.

[0014] SEQ ID No.5: KKKIISAILMSTVILSAAAPLSGVYADTNSD Furthermore, the present invention provides an engineered bacterium capable of expressing the aforementioned fusion peptide.

[0015] Preferably, the engineered bacteria shown contain an expression cassette; the expression cassette contains a sequence encoding the Usp45 signal peptide, a sequence encoding the enhancing peptide, and a sequence encoding the mutant GLP-1 protein.

[0016] Preferably, the sequence encoding the Usp45 signal peptide is shown in SEQ ID No. 7.

[0017] Preferably, the sequence encoding the enhancing peptide is shown in SEQ ID No. 8.

[0018] SEQ ID No.7: AAGAAGAAGATCATCTCAGCTATCCTGATGTCAACCGTCATTTTGTCAGCTGCTGCTCCATTGTCAGGCGTTTATGCTGATACAAATTCAGAC SEQ ID No.8: TTGGAAATTTCATCAACATGCGACGCT Preferably, the originating strain of the engineered bacteria is Lactobacillus paracasei.

[0019] Preferably, the present invention provides a *Lactobacillus paracasei* (engineered strain), which is LCglp1v1, and its accession number is CGMCC No. 37511.

[0020] This strain was deposited on January 23, 2026, at the China General Microbiological Culture Collection Center (CGC-MSC). Address of the depository: Institute of Microbiology, Chinese Academy of Sciences, No. 3, Courtyard 1, Beichen West Road, Chaoyang District, Beijing; Postcode: 100101; Classification and nomenclature: *Lactobacillus paracasei*. Lacticaseibacillus paracasei The accession number is CGMCCNo. 37511.

[0021] Furthermore, the present invention provides a pharmaceutical product containing the aforementioned fusion peptide, or the aforementioned engineered bacteria, or the aforementioned Lactobacillus paracasei engineered bacteria.

[0022] Preferably, the medicine further includes pharmaceutically acceptable excipients.

[0023] Preferably, the drug is used to treat type 2 diabetes.

[0024] Preferably, the drug achieves the purpose of treating type 2 diabetes by continuously secreting mutant GLP-1 protein and regulating the intestinal flora.

[0025] Preferably, the drug is an oral preparation.

[0026] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention develops an engineered bacterium that continuously secretes mutant GLP-1 protein. It exhibits good passage stability and can exert a dual therapeutic effect on type 2 diabetes through continuous secretion of mutant GLP-1 protein and regulation of the gut microbiota. This engineered bacterium can be used to prepare oral pharmaceutical formulations. Therefore, this invention provides a new technical means for developing non-invasive, cost-effective, and efficient drug delivery strategies targeting GLP-1, and has broad application prospects. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the pTRST-mGLP-TA gene editing plasmid structure.

[0028] Figure 2The results represent the genetic stability and physiological characteristics of the engineered bacteria; lane a shows the PCR validation of plasmid elimination targeting the erythromycin resistance gene; lane 1: wild-type *Lactobacillus paracasei*; lane 2: LCglp1v1 strain after 50 consecutive passages (LCglp1v1-50); lane 3: initial engineered strain LCglp1v1 (before passage); lane 4: gene-edited plasmid (positive control); lane b shows the extracellular mGLP-1 level in the culture supernatant quantified by ELISA; lane c compares the growth kinetics of wild-type, LCglp1v1, and LCglp1v1-50 strains cultured in MRS broth; lanes d and e show the physiological robustness of LCglp1v1-50 compared to the wild-type under acid and bile salt stress, respectively. Data are expressed as mean ± standard deviation (n=3), and statistical significance was determined by Student's t-test.

[0029] Figure 3 This study investigated the effect of engineered L. LCglp1v1-50 on glucose homeostasis in type 2 diabetic (T2DM) mice. Figure a shows the experimental design and timeline of the mouse model; figure b shows fasting blood glucose (FBG) levels during treatment; figures c and d show the oral glucose tolerance test (OGTT), with figure c representing the glucose fluctuation curve and figure d representing the area under the curve (AUC); figures e and f show the insulin tolerance test (ITT), with figure e representing the glucose level relative to baseline (percentage of initial value) and figure f representing the area under the curve; groups: A, normal control group (NC); B, model control group (MC, i.e., type 2 diabetes group); C, wild-type L. paracasei; D–F, low / medium / high dose L. LCglp1v1-50 groups; G, positive control group (metformin); data are expressed as mean ± standard deviation (n=6); different letters indicate statistical significance (…). P <0.05, one-way ANOVA and Tukey's test).

[0030] Figure 4 This data represents the results of the engineered bacterium LCglp1v1-50 restoring metabolic homeostasis and structural integrity of the liver; a–h represent systemic biochemical analyses, where a is the renal indicator creatinine (CRE), b is blood urea nitrogen (BUN), c is the lipid metabolism parameter total cholesterol (T-CHO), d is high-density lipoprotein cholesterol (HDL-C), e is low-density lipoprotein cholesterol (LDL-C), f is triglycerides (TG), g is the liver-damaging enzyme aspartate aminotransferase (AST), and h is the serum level of alanine aminotransferase (ALT); i represents the histological assessment of liver tissue, with representative H&E stained sections shown at 200x and 400x magnification (scale bars: 50 μm and 20 μm); data are expressed as mean ± standard deviation (n=3); bars labeled with different letters are shown in… PSignificant difference was found when <0.05 (Tukey post-hoc test was performed after one-way ANOVA).

[0031] Figure 5 This study investigated the regulatory effects of engineered LCglp1v1-50 on the hepatic GLP-1R and insulin signaling cascade. A–E represent the immunoblotting detection of the GLP-1R / cAMP / PKA axis; b, c, d, and e quantified the expression of GLP-1R, cAMP, PKA, and p-CREB in liver lysates. F–i assessed the downstream PI3K / AKT / FoxO1 pathway; g, h, and i showed the relative phosphorylation ratios of PI3K (p-PI3K / total PI3K), AKT (p-AKT / total AKT), and FoxO1 (p-FoxO1 / total FoxO1). Treatment groups: A, normal control (NC); B, model control (MC); C, wild-type (WT); F, high-dose LCglp1v1-50; G, positive control. Data are expressed as mean ± standard deviation (n=5). Different letters indicate statistical significance. P <0.05, ANOVA and Tukey's test). Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of this invention, not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0033] In the embodiments provided in this specification, unless specific techniques or conditions are specified, the techniques or conditions described in the literature in this field, or the product instructions, shall be followed. Reagents or instruments whose manufacturers are not specified are all conventional products that can be purchased from legitimate channels.

[0034] This invention relates to molecular biology experiments. Unless otherwise specified, reference can be made to the book *Molecular Cloning* (J. Sambrook, E.F. Fritsch, and T. Maniatis, Science Press, 1994). This book and its subsequent editions are the most commonly used and guiding reference books for those skilled in the art when performing experiments related to molecular biology. In addition, depending on the experimental purpose, those skilled in the art complete the corresponding experiments under the guidance of the operating manuals accompanying various commercial reagent kits or entrust them to specialized companies, such as primer synthesis and gene sequencing.

[0035] The testing methods for the oral glucose tolerance test (OGTT) and insulin tolerance test (ITT) in the following examples are as follows: Oral glucose tolerance test (OGTT) and intraperitoneal test (ITT) were performed 14 days and 7 days before the study endpoint, respectively. For the OGTT, mice were fasted overnight (12 hours) and administered glucose solution (2.0 g / kg body weight, orally). For the ITT, mice were fasted for 6 hours, followed by intraperitoneal injection of recombinant human insulin (1.0 IU / kg, intraperitoneal). In both experiments, tail vein blood glucose was quantified using a glucometer at 0, 30, 60, 90, and 120 minutes. The area under the curve (AUC) was calculated to assess glycemic variability.

[0036] The method for continuous blood glucose monitoring in the following examples is as follows: Longitudinal blood glucose monitoring was performed to assess the long-term efficacy of the intervention, with fasting blood glucose (FBG) levels monitored weekly at days 7, 14, 21, and 28 post-treatment. At each time point, mice (n=6 per group) were randomly selected and fasted overnight. Blood samples were collected via tail tip sampling, and glucose concentrations were quantified using a glucometer.

[0037] The blood biochemical analysis method in the following examples is as follows: At the end of the experiment, mice were anesthetized by inhalation with 2% isoflurane. Once the anesthesia depth met the blood collection criteria, peripheral blood samples were collected via retro-orbital venous plexus puncture. All animal experimental procedures strictly followed the "Guidelines for Ethical Review of Laboratory Animal Welfare" and were approved by the institution's Laboratory Animal Ethics Committee (Approval No.: Ethics 2025-1). To separate serum, samples were allowed to coagulate at room temperature for 2 hours, followed by centrifugation at 450×g for 20 minutes. Quantitative analysis was performed on key metabolic markers, including liver enzymes (alanine transaminase, aspartate transaminase), renal function indicators (blood urea nitrogen, creatinine), and lipid profiles (triglycerides, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol).

[0038] The methods for histopathological examination in the following examples are as follows: Liver and kidney tissue samples were fixed in 4% paraformaldehyde (PFA / POM) for 24 hours, treated with a gradient of alcohol and xylene, embedded in paraffin, and cut into 10-micrometer-thick sections. The sections were then stained with hematoxylin and eosin (H&E) for morphological evaluation. The stained sections were then examined under a microscope and imaged.

[0039] The following examples illustrate the method for quantitative analysis of the persistence of engineered bacteria in intestinal cultures: After a one-week acclimatization period, C57BL / 6 mice (n=10 per group) were randomly divided into three groups, each receiving a single oral gavage (0.2 mL): (1) 10% skim milk; (2) wild-type Lactobacillus paracasei (10 mL) 10(CFU / mL); (3) Recombinant strain LCglp1v1-50 (10 10 CFU / mL). To track bacterial persistence, fecal particles were collected longitudinally via abdominal massage from 0 to 168 hours post-inoculation. Total DNA (Tiangen Biotech) was extracted and qPCR was performed against the integrated mGLP-1 expression cassette (mGLP-1-F: 5ʹ-TCATCAACATGCGACGCTCA-3ʹ, mGLP-1-R: 5ʹ-TTAACGGCCATCAACCAACC-3ʹ). Bacterial load (copy number / gram of feces) was quantified using a standard curve generated from LCGLP1v1 genomic DNA.

[0040] The metagenomic analysis method in the following examples is as follows: Fecal genomic DNA was extracted (Tiangen Biotech) and shotgun metagenomic sequencing was performed on the Illumina NovaSeq 6000 platform (Shanghai Meiji Biotechnology). Clean reads were de novo assembled into contigs using MEGAHIT, and ORF prediction was performed using Prodigal to generate a coding sequence catalog. A non-redundant gene set was constructed using CD-HIT (≥95% consistency) and annotated by alignment with DIAMOND, NR (for taxonomy), KEGG, and eggNOG databases. Statistical modeling, including LEfSe and pathway enrichment analysis, was performed on the Meiji Biotechnology cloud platform.

[0041] The methods for metabolite extraction and liquid chromatography-mass spectrometry analysis in the following examples are as follows: Metabolites were extracted from 100 mg of tissue homogenized in liquid nitrogen using 500 μL of 80% methanol. After incubation on ice for 5 min, the supernatant was centrifuged at 20,000 × g for 20 min at 4 °C. The supernatant was diluted with HPLC-MS grade water to a final methanol concentration of 53% and centrifuged again. The final supernatant was analyzed using a Q Exactive™ HF-X mass spectrometer with a Hypersil Gold C18 column (100 × 2.1 mm, 1.9 μm; Thermo Fisher Scientific). Chromatography employed a linear gradient of 0.1% formic acid (A) and methanol (B) at a flow rate of 0.2 mL / min, an autosampler temperature of 10 °C, and an injection volume of 5 μL. Mass spectrometry scans covered m / z 100–1500. Raw data were aligned using Compound Discoverer 3.1 (Thermo Fisher Scientific) and annotated according to the mzCloud / KEGG database. The difference feature is defined as VIP > 1 and P <0.05.

[0042] The transcriptome analysis methods in the following examples are as follows: Total RNA was extracted from liver tissue (n=5 per group) using the RNAprep Pure Tissue Kit (Tiangen Biotech, Beijing, China). Its integrity was assessed using an Agilent 2100 before constructing strand-specific libraries using the NEBNext Ultra II Directed RNA Library Preparation Kit (NEB, Massachusetts, USA). Sequencing was performed on an Illumina NovaSeq 6000 to generate paired-end reads. After trimming (Trimmomatic v0.39), clean reads were aligned to a GRCm38 / Ensembl100 reference genome. Gene counts were generated using featureCounts (v2.0.3). Differential expression was determined using DESeq2 based on FDR < 0.05 and |log2FC| > 1. Pathway enrichment (GO / KEGG) analysis was performed using the clusterProfiler (v4.10.0) software package.

[0043] The methods for protein blot analysis in the following examples are as follows: Tissue samples (approximately 100 mg) were homogenized in ice-cold RIPA buffer containing protease inhibitors. Lysates were clarified by centrifugation at 12,000 × g for 30 minutes at 4°C and quantified using the BCA method. Equal volumes of protein samples were separated by SDS-PAGE, transferred to PVDF membranes, and blocked with 5% skim milk. The membranes were incubated overnight at 4°C with specific primary antibodies (Table 1), followed by incubation for 1 hour with HRP-conjugated secondary antibodies. Protein bands were developed using ECL and analyzed by ImageJ.

[0044] Table 1

[0045] All data in the following examples were analyzed using GraphPad Prism statistical software (Prism version 8.0, https: / / www.graphpad.com / ). Differences between two independent samples were analyzed using Student's t-test; statistical differences between three or more samples were analyzed using one-way or two-way ANOVA combined with Tukey's multiple comparison test. Data are expressed as mean ± standard deviation (SD). Different letters indicate different values. P There was a statistically significant difference at the <0.05 level.

[0046] Example 1: Construction of engineered bacteria and testing of their genetic stability and physiological characteristics (1) Strains, plasmids and culture conditions *Escherichia coli* DH5α was used as the cloning host for all plasmid constructions. *E. coli* DH5α was cultured in LB (Luria-Bertani) broth at 37°C, while *Lactobacillus paracasei* was cultured in MRS (Man Rogosa Sharpe) medium at 37°C. The *E. coli*-*Lactobacillus paracasei* shuttle vector pTRST was used as the gene backbone for all recombinant constructs. Transformants were screened on antibiotic agar plates supplemented with erythromycin (100 μg / mL for *E. coli*; 10 μg / mL for *Lactobacillus paracasei*). All plasmids used in this study are shown in Table 2.

[0047] Table 2

[0048] (2) Construction of engineered strain LCglp1v1 Add the lactate dehydrogenase promoter to the pTRST vector via Gibson assembly. P ldh The pTRST-mGLP-TA gene editing plasmid was constructed using four parts: a sgRNA module (designed for the 5'-TGAAA-3' PAM site, using overlap extension PCR (OE-PCR) to fuse the spacer sequence with a tracrRNA scaffold to construct a chimeric sgRNA expression cassette), two 1000 bp homologous arms flanking the genomic target site, and an insert fragment. The specific steps are as follows: like Figure 1 As shown, to achieve exocrine expression of peptides with defined biological activity, we designed and constructed a fusion expression cassette containing mutant GLP-1 (mGLP-1). Specifically, based on the nucleotide sequence of wild-type GLP-1 (7-36) amide (SEQ ID NO 3), we introduced base substitutions corresponding to the three amino acid sites Ala8Gly, Lys26Gln, and Lys34Asp into its coding sequence using site-directed mutagenesis to obtain the mGLP-1 fragment (its nucleotide sequence is shown in SEQ ID NO 4).

[0049] To confer the ability of this mutant to be efficiently secreted and expressed in the host bacteria, we sequentially fused a gene derived from Lactococcus lactis to the upstream (5' end) of the mGLP-1 coding sequence. Lactococcus lactis The sequence encoding the Usp45 signal peptide (SPusp45, SEQ ID NO 7) and the sequence encoding the LEISSTCDA secretion-enhancing peptide (SEQ ID NO 8) are combined. The nucleotide sequence of the final secretory mGLP-1 complete expression cassette is shown in SEQ ID NO 9.

[0050] SEQ ID NO.9: TTGGAAATTTCATCAACATGCGACGCTCATGGCGAAGGCACATTTACATCAGATGTTTCATCATACTTGGAGGGCCAAGCTGCTCAAGAATTTATTGCTTGGTTGGTTGATGGCCGTTAA.

[0051] The primers required for plasmid construction in this embodiment are shown in Table 3.

[0052] Table 3

[0053] Competent cells of *Lactobacillus paracasei* were prepared by pretreatment with 3% glycine and washing with 10% glycerol, and transformed by electroporation at 2000 V, 400 Ω, and 25 μF. After recovery at 37°C for 3 hours, transformants were screened on MRS agar supplemented with erythromycin (72 hours). Colony PCR was performed using primers YZ-QR-F / R2 to confirm genome integration, which spanned the tpiA-mglp-1 linker region. To obtain a label-free strain, transformants were serially passaged in antibiotic-free MRS to induce plasmid loss. Using wild-type (WT) and plasmid-removed strains as controls, the final engineered strain was validated by PCR to confirm the presence of the mGLP-1 expression cassette and the absence of the erythromycin resistance gene. The extracellular mGLP-1 level in the supernatant was quantified using a human GLP-1 ELISA kit (catalog number MM-1393W1, MEIMIAN).

[0054] (3) Stability test of gene-edited strains The engineered bacteria were passaged 50 times in antibiotic-free MRS medium to induce plasmid loss, aiming to remove the gene-editing plasmid from the engineered bacteria. Stability was subsequently confirmed by sequencing and ELISA. In the stress assay, the activated culture (OD) was... 600 Approximately 0.5 g of taurine (after two 18-hour subcultures) was inoculated into MRS broth adjusted to pH 2.0, 3.0, or 4.0 with hydrochloric acid, and simultaneously into MRS broth supplemented with 0%, 0.1%, 0.3%, and 0.5% (w / v) taurine. All media were sterilized at 115°C for 20 minutes. Viability was determined by CFU counting using the standard serial dilution plate method.

[0055] Figure 2 The results represent the genetic stability and physiological characteristics of the engineered bacteria.

[0056] PCR verification confirmed that the erythromycin resistance gene was completely deleted, ensuring successful plasmid loss. Figure 2a). The mGLP-1 secretion of strain LCglp1v1 reached 99.85 pg / mL ( Figure 2 (b) This level was maintained even after 50 consecutive passages, indicating excellent stability of genome integration. Comparative growth analysis showed that mGLP-1 expression imposed minimal metabolic burden on strain LCglp1v1, and its growth kinetics were similar to those of wild-type *Lactobacillus paracasei* (entering the exponential phase at approximately 8 hours and the stationary phase at 16-20 hours). LCglp1v1-50 exhibited a prolonged lag phase, entering the exponential phase at 12 hours and reaching saturation at 24 hours. However, the final biomass of LCglp1v1-50 was comparable to that of the wild-type, indicating that the host could tolerate this metabolic load. Figure 2 c). Furthermore, this engineered bacterium retains the inherent probiotic resilience of its parent strain and exhibits strong tolerance to simulated gastric acid and bile salt stress. Figure 2 (d, e). These data confirm that LCglp1v1 and LCglp1v1-50 possess the genetic stability and physiological adaptability required for oral delivery applications.

[0057] Example 2 Experimental design and animal treatment of a type 2 diabetes model The experiment was conducted with approval from the Institutional Animal Care and Use Committee (IACUC) of Beijing Technology and Business University (No. 2025-1). Male C57BL / 6 mice (8 weeks old, 28-30 g; purchased from Beijing Saibaifu Biotechnology Co., Ltd.) were housed in an SPF environment (22±2℃, 12:12 hour light / dark cycle). After a one-week acclimatization period, the mice were divided into a normal control group (Group A, n=12; standard diet) and a model group (fed a high-fat diet (HFD; basal diet containing 10% lard, 20% sucrose, 2.5% cholesterol, and 0.5% sodium cholate; Saibaifu No. SFD019)). After 12 weeks, type 2 diabetes was induced by intraperitoneal injection of streptozotocin (STZ) (80 mg / kg daily for 3 consecutive days). Diabetic mice with a confirmed blood glucose level ≥16.7 mmol / L were randomly divided into groups (n=12 per group): (B) negative control group, (C) wild-type Lactobacillus paracasei strain group, (DF) mGLP-1 producing strain group (i.e., the engineered strain LCglp1v1-50 of Example 1, D. low-dose group, E. medium-dose group, F. high-dose group), and (G) positive drug control group. Administered via gavage daily starting from week 13 (dosing regimen shown in Table 4).

[0058] Table 4

[0059] The overall design diagram and timeline of the animal experiment are as follows: Figure 3As shown in a. All experiments followed applicable institutional and / or national guidelines regarding animal care and use.

[0060] We evaluated the therapeutic effects of the engineered bacterium LCglp1v1-50 in a type 2 diabetic mouse model. Figure 3 As shown in b, high-dose LCglp1v1-50 treatment (group F) significantly alleviated hyperglycemia in type 2 diabetic mice, and its efficacy was comparable to that of the metformin-positive control group (group G). Figure 3 b) Comparable. Specifically, fasting blood glucose (FBG) in group F decreased from 17.7 ± 5.2 mmol / L at week 1 (baseline) to 10.0 ± 3.5 mmol / L at week 4, a 43.2% reduction compared to the model group (17.6 ± 1.5 mmol / L). This therapeutic effect was dose-dependent: the medium-dose LCGLP1v1-50 (group E) gradually reduced blood glucose to 15.4 ± 4.5 mmol / L, while the low-dose (group D) and wild-type (group C) strains maintained the endpoint fasting blood glucose levels at 12.2 ± 7.2 mmol / L and 18.3 ± 6.2 mmol / L, respectively. Notably, group F elicited the strongest response, with a rapid reduction in blood glucose starting as early as week 2 (13.6 ± 2.8 mmol / L, compared to the model group, P <0.05), and continued until the end of the experiment.

[0061] In the oral glucose tolerance test (OGTT), compared with the model group, group F suppressed the peak glucose level by 37.3% at 30 minutes (16.7±2.4 vs. 26.7±2.3 mmol / L) and reduced the area under the curve (AUC0-120min) to 29.6±4.9 mmol / L·h. Similarly, in the insulin tolerance test (ITT), the glucose level in group F decreased by 52.5±13.8% within 60 minutes after insulin injection, and its AUC ITT was significantly lower than that in the wild-type (WT) group (12.4±3.0 vs. 20.7±9.0 mmol / L). P <0.05).

[0062] Overall, the engineered bacterium LCglp1v1-50 was able to improve blood glucose levels, glucose tolerance, and insulin tolerance in type 2 diabetic mice, indicating that LCglp1v1-50 treatment can effectively reverse the hyperglycemic state in type 2 diabetic (T2DM) mice.

[0063] Based on these findings, we will now attempt to elucidate the molecular mechanisms driving this therapeutic phenotype.

[0064] Further whole-body biochemical analysis showed ( Figure 4Compared with the normal control group (Group A), the type 2 diabetes induction group (Group B) showed severe organ dysfunction. This manifested as widespread metabolic dysregulation, characterized by a surge of approximately 3.3 times in kidney injury markers (creatinine: 90.9±5.0 vs. 27.4±2.9 μmol / L; blood urea nitrogen: 12.1±0.9 vs. 6.7±0.6 mmol / L). P <0.001). Liver function was also impaired, with aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels elevated to 49.7±3.2 U / L and 15.9±3.2 U / L, respectively. Furthermore, the model group exhibited severe hyperlipidemia, with significantly elevated levels of low-density lipoprotein cholesterol (LDL-C, 8.05±0.67 mmol / L), triglycerides (TG, 1.59±0.22 mmol / L), and total cholesterol (TC, 3.82±0.23 mmol / L) compared to the control group.

[0065] The administration of engineered bacteria successfully alleviated these pathological changes in a dose-dependent manner. Specifically, the high-dose regimen (Group F) elicited the most potent therapeutic response, with no significant difference compared to the metformin control group (Group G). Treatment in Group F significantly reduced creatinine (CRE) and blood urea nitrogen (BUN) levels to 54.0±2.9 μmol / L and 7.8±0.6 mmol / L, respectively, implying a recovery of renal function of 40.6% and 35.6%, respectively. Similarly, lipid disorders were effectively resolved, with low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), and total cholesterol (TC) levels decreasing by 52.8%, 35.6%, and 24.3% (to 3.80±0.33, 1.02±0.08, and 2.89±0.31 mmol / L, respectively), indicating a significantly better restoration of systemic homeostasis than the wild-type (Group C) and low-dose (Group D) cohorts. P <0.05).

[0066] Further H&E staining confirmed ( Figure 4 i) The normal control group mice (group A) retained an intact liver structure with regularly arranged hepatocytes. Compared with the control group, the model group (group B) showed severe structural disorder, characterized by extensive necrosis, extensive inflammatory infiltration, and bile duct hyperplasia. In contrast, the wild-type group (group C) and the low-dose group (group D) showed no significant improvement, with persistent necrotic foci and inflammation still present, while the medium- or high-dose LCglp1v1-50 treatment groups (groups E and F) and the metformin group (group G) significantly alleviated these pathological changes. The liver structure of these effective treatment groups was substantially restored, with only mild signs of fibrosis and a small amount of inflammatory infiltration.

[0067] The above results indicate that the engineered bacterium LCglp1v1-50 can correct dyslipidemia, reduce liver damage, and decrease lipid accumulation in type 2 diabetic mice.

[0068] Furthermore, this invention investigated the colonization characteristics and regulatory effects of engineered bacteria LCglp1v1-50 in the intestinal flora of mice.

[0069] To characterize in vivo colonization kinetics, we tracked the retention of LCglp1v1-50 after a single gavage administration. Longitudinal qPCR analysis of mGLP-1 in feces showed strong bacterial persistence. The strain was still detectable on day 7 post-administration, confirming its ability to stably and persistently colonize the mouse intestine.

[0070] Metagenomic analysis was performed to assess the impact of LCglp1v1-50 on microbial community structure. Alpha diversity analysis revealed a significantly higher microbial richness (Chao1 index) in the model group (Group B) compared to the normal control group. P <0.05). Notably, *Lactobacillus paracasei* intervention effectively reversed this increase, restoring species richness to near-normal physiological levels. There was no significant difference in community evenness (Shannon index) among the experimental groups. P >0.05). However, in mice treated with LCglp1v1-50, a trend toward baseline levels similar to those in the healthy control group was observed.

[0071] β-diversity analysis using principal coordinate analysis (PCoA) revealed a significant separation between the dysregulated model group (Group B) and the normal control group (Group A). ​​Crucially, *Lactobacillus paracasei* intervention restructured the microbial community, driving its transformation towards the healthy control phenotype. This restorative effect was confirmed by hierarchical clustering analysis, which showed high concordance between high-dose LCglp1v1-50 (Group F) and the positive control group (Group G) (correlation coefficient = 0.97), and lowest affinity with the model group (correlation coefficient = 0.88). At the compositional level, Venn diagram analysis identified a stable core of 9482 shared operational taxonomic units (OTUs). However, distinct characteristics emerged among the groups: Group F had the most unique OTUs (385), while the metformin group had the fewest (120), suggesting that LCglp1v1-50 promotes a specific bacterial enrichment pattern distinct from standard drug interventions.

[0072] Phylogenetic analysis indicated that type 2 diabetes mellitus (T2DM) induced (group B) severe dysbiosis, characterized by a significant increase in potentially pathogenic Proteobacteria compared to the control group. P <0.01) and a reduction in the protective wart microbes ( P<0.001). Although Bacteroidetes remained stable, *Lactobacillus paracasei* caused a resurgence of Firmicutes. P =0.058), and significantly reversed the changes in Proteobacteria and Verruciformis. At the genus level, this strain exhibited selective regulatory effects: the model group Akkermansia The genera *Lactobacillus* and *Bacillus* were almost entirely absent, while opportunistically pathogenic *Corynebacterium* overgrowth. However, intervention with *Lactobacillus paracasei* effectively counteracted these changes, significantly increasing beneficial taxa while strictly suppressing *Corynebacterium* (*Bacillus*). P <0.01).

[0073] KEGG pathway enrichment analysis based on reporter gene scores revealed significant functional remodeling in LCglp1v1-50 treated mice. The upregulated functional modules were primarily biased towards xenobiotic biodegradation and amino acid metabolism, particularly the degradation of aromatic compounds and benzoic acid. Notably, the valine, leucine, and isoleucine degradation pathway (mmu00280) was significantly enriched, highlighting enhanced branched-chain amino acid (BCAA) catabolism as a potential mechanism for improved insulin sensitivity. Conversely, pathways regulating genetic information processing (such as ribosome and aminoacyl-tRNA biosynthesis) and pathogenicity-related pathways (Staphylococcus aureus infection) were significantly weakened, reflecting a shift towards a beneficial metabolic phenotype in mice.

[0074] Furthermore, the regulatory role of GLP-1 on blood glucose in mice was revealed through integrated metabolomics and transcriptomics analysis.

[0075] We performed multi-omics analyses combining transcriptomics and metabolomics on groups A, B, C, G, and F, prioritizing the high-dose LCglp1v1-50 cohort (group F) due to its superior hypoglycemic efficacy. Non-targeted LC-MS / MS analysis of plasma samples showed significant metabolic differences between the model and intervention groups, confirmed by principal component analysis (PCA). A supervised PLS-DA model validated by permutation tests (VIP > 1.0, FC > 1.2 or < 0.83) was used. P <0.05) identified metabolites with discriminative function.

[0076] Volcano plots revealed significant metabolic changes in the high-dose LCglp1v1-50 group, with 114 significant signatures identified in negative ion mode (24 upregulated, 90 downregulated) and 142 significant signatures identified in positive ion mode (95 upregulated, 47 downregulated). In contrast, the wild-type strain (Group C) caused only minor perturbations, with 43 signatures identified in negative ion mode (6 upregulated, 37 downregulated) and 30 signatures identified in positive ion mode (8 upregulated, 22 downregulated), highlighting the stronger regulatory potency of LCglp1v1-50 compared to the wild-type strain. Pathway enrichment analysis revealed the different regulatory profiles between the two interventions. Intervention with the engineered bacterium LCglp1v1-50 specifically targets the core signal transduction network, characterized by specific regulation of the mTOR (mmu04150), PI3K-Akt (mmu04151), FoxO (mmu04068), and cAMP (mmu04024) pathways. In contrast, metabolic changes in the wild-type group (Group C) were mainly limited to general metabolic processes, such as 2-oxocarboxylic acid and amino sugar metabolism. This effect on upstream regulatory pathways indicates that LCglp1v1-50 exerts its regulatory capacity far exceeding that of the wild-type control group through a multi-layered mechanism in synergy with transcriptome reprogramming.

[0077] RNA-seq analysis revealed that, compared with the model group, the high-dose LCglp1v1-50 group had 8204 differentially expressed genes (DEGs) (4267 upregulated and 3937 downregulated), and compared with the wild-type (WT) group (6898 DEGs: 3449 upregulated and 3449 downregulated), the regulatory range was significantly expanded.

[0078] These data indicate that LCglp1v1-50 administration induced more significant transcriptomic changes than WT intervention. Hierarchical clustering revealed that group F reproduced the gene expression characteristics of the metformin-positive control group (group G), effectively distinguishing it from the disease state. Functional enrichment analysis using KEGG correlated these changes with key insulin-sensitive networks, particularly the PI3K-Akt and FoxO signaling pathways. Metabolically, high-dose LCglp1v1-50 treatment upregulated fatty acid degradation and cholesterol metabolism, especially the degradation of valine, leucine, and isoleucine, compared to the model group. This transcriptional enrichment is fully consistent with our metabolomics data, reinforcing the view that enhanced branched-chain amino acid (BCAA) clearance is a core mechanism for alleviating insulin resistance. Furthermore, pathways associated with diabetic cardiomyopathy, reactive oxygen species, and protein processing in the endoplasmic reticulum were significantly regulated, suggesting a broad protective role against cellular stress and tissue damage. KEGG analysis of the WT group compared to the model group revealed a more dispersed regulatory pattern, primarily focused on stress responses and cell survival (e.g., "chemical carcinogenesis-reactive oxygen species," "FoxO signaling pathway," and various cancer pathways), rather than metabolic homeostasis. Unlike the engineered strain, the WT group showed negligible involvement in fatty acid degradation pathways, suggesting that its therapeutic effect is mediated through broad, non-specific signaling rather than precise metabolic intervention.

[0079] Combined pathway enrichment and multi-omics network analysis identified the PI3K-Akt / mTOR signaling axis as a central hub coordinating gene expression and metabolic flux. Regarding lipid metabolism, the transcriptional upregulation of fatty acid degradation was supported by a strong correlation (Pearson |r|>0.96) between upregulated transcripts (e.g., Gene_199, Gene_012) and acylcarnitine, consistent with enhanced mitochondrial β-oxidation. Furthermore, this analysis revealed a simultaneous restoration of branched-chain amino acid (BCAA) homeostasis: genes involved in valine, leucine, and isoleucine degradation (e.g., Gene_389, Gene_301, Gene_456) showed extensive associations with free BCAAs and dipeptides (e.g., valine-isoleucine-leucine). Overall, these multidimensional results suggest that LCglp1v1-50 reduces glycemic index by simultaneously promoting lipid degradation and BCAA clearance.

[0080] Furthermore, we found that LCglp1v1-50 regulates and maintains the number of functional β cells through the cAMP / PKA-dependent PI3K / Akt / FoxO1 pathway.

[0081] like Figure 5As shown, Western blot analysis revealed that the healthy control group (Group A) maintained baseline biomarker levels. Conversely, the type 2 diabetes model (Group B) exhibited a marked pathological phenotype in the liver. GLP-1R protein expression was significantly suppressed compared to the control group, indicating reduced receptor stability in a diabetic state. Figure 5 a, b). This downregulation was also accompanied by a significant decrease in downstream PKA and CREB phosphorylation levels, indicating damage to the GLP-1R signal axis ( Figure 5 a, d, e). Furthermore, although total protein levels remained unchanged, the phosphorylation levels of PI3K, Akt, and FoxO1 were significantly reduced in group B. Figure 5 These results indicate severe impairment in hepatic insulin signaling. In summary, these data confirm that the current model effectively reproduces the key pathological features of hepatic insulin resistance and GLP-1 signaling deficiency unique to human type 2 diabetes.

[0082] We observed that GLP-1R protein expression and PKA phosphorylation levels in group F mice were significantly restored, returning to the levels of the healthy control group (group A). Figure 5 (a, d). Conversely, these markers did not show significant changes in the wild-type *Lactobacillus paracasei* control group (Group C) compared to the type 2 diabetes model group (Group B). This confirms that pathway activation is caused by mGLP-1 secretion, rather than the probiotic strain itself. Furthermore, LCglp1v1-50 treatment enhanced the insulin signaling cascade, as evidenced by increased phosphorylation levels of PI3K, AKT, and FoxO1 in Group F compared to Group B. Figure 5 These results indicate that LCGLP1v1-50 can activate the classical GLP-1 pathway and restore the insulin-mediated PI3K / AKT / FoxO1 axis.

[0083] Although the metabolic efficacy of metformin (group G) was comparable to that of group F, their underlying signaling mechanisms differed significantly. Metformin treatment significantly increased the levels of p-PI3K, p-AKT, and p-FoxO1, confirming the restoration of hepatic insulin sensitivity. Figure 5 c, d). However, compared with group F, metformin failed to alter GLP-1R protein expression or PKA phosphorylation ( Figure 5 (a, b) This situation, which does not involve the GLP-1R / PKA axis, is consistent with the established pharmacological properties of metformin, whose action is independent of receptor stimulation. Therefore, this comparison further confirms that the GLP-1R pathway activation observed in group F is specifically driven by mGLP-1 secreted by engineered bacteria.

[0084] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fusion peptide, characterized in that, It contains a signal peptide, an enhancing peptide, and a mutant GLP-1 protein; the mutant GLP-1 protein has mutations at the Ala8Gly, Lys26Gln, and Lys34Asp sites on the basis of the wild-type GLP-1 protein.

2. The fusion peptide according to claim 1, characterized in that, The signal peptide is the Usp45 signal peptide; and / or, the amino acid sequence of the enhancing peptide is LEISSTCDA.

3. An engineered bacterium, characterized in that, It is capable of expressing the fusion peptide described in claim 1 or 2.

4. The engineered bacteria according to claim 3, characterized in that, It contains an expression cassette; the expression cassette contains a sequence encoding the Usp45 signal peptide, a sequence encoding the enhancing peptide, and a sequence encoding a mutant GLP-1 protein.

5. The engineered bacteria according to claim 3 or 4, characterized in that, The origin of the engineered bacteria is *Lactobacillus paracasei*.

6. A type of *Lactobacillus paracasei*, characterized in that, It is LCglp1v1, and its accession number is CGMCC No.37511.

7. A medicine, characterized in that, It contains the fusion peptide as described in claim 1 or 2, or the engineered bacteria as described in any one of claims 3 to 5, or the engineered Lactobacillus paracasei as described in claim 6.

8. The pharmaceutical product according to claim 7, characterized in that, The drug is an oral preparation; and / or, the drug further includes pharmaceutically acceptable excipients.

9. The pharmaceutical product according to claim 7 or 8, characterized in that, The drug is used to treat type 2 diabetes.

10. The pharmaceutical product according to claim 9, characterized in that, The drug aims to treat type 2 diabetes by continuously secreting mutant GLP-1 protein and regulating gut microbiota.

Images