Vaginal flora protein marker group related to cervical high-grade squamous intraepithelial lesions and application thereof

Abstract

The invention discloses a vaginal flora protein marker group related to cervical high-level squamous intraepithelial lesions and application thereof, and adopts a metaproteomics method to construct vaginal flora protein markers for cervical HSIL diagnosis for the first time; the marker combination not only contributes to early recognition of cervical HSIL patients and optimization of clinical intervention strategies, but also provides an important basis for deep understanding of functional mechanisms of vaginal flora in cervical precancerous lesions and development of novel micro-ecological regulation therapy.

Description

Technical Field

[0001] The present invention belongs to the field of biomedical testing and relates to a vaginal flora protein marker group associated with high-grade squamous intraepithelial lesions (HSIL) of the cervix and its application. Background Art

[0002] Cervical cancer is a malignant tumor that originates in the cervix and is the most common gynecological malignancy. The female vaginal microbiome plays a key role in the development and progression of human papillomavirus (HPV) infection and cervical cancer. Under normal physiological conditions, the vaginal microbiome maintains a dynamic balance, forming a healthy microenvironment. However, persistent disturbances in this microbiome significantly increase the risk of HPV infection and promote the development of cervical squamous intraepithelial lesions (SIL). As SIL progresses, patients typically progress from low-grade SIL (LSIL) to high-grade HSIL, ultimately developing cervical cancer.

[0003] Currently, research on the role of vaginal microbiota in cervical precancerous lesions mainly focuses on changes in microbial community structure and species abundance (J Transl Med, 2024, 22(1):396; BMC Women's Health, 2025, 25(1):65). However, there is still a lack of in-depth understanding of the regulatory mechanisms of microbial functional activity (such as protein expression profiles) during the transition from LSIL to HSIL. As clinical management strategies and cancer risk vary significantly between different levels of SIL, exploring the dynamic changes of the vaginal microbiota proteome is of great value in understanding the progression mechanism of precancerous lesions, developing new diagnostic markers, and discovering potential therapeutic targets.

[0004] Traditional microbiome research methods (such as 16S rRNA sequencing) are difficult to directly reflect the functional activity of the microbiome, while vaginal secretions contain rich microbial protein information, which can more directly reveal the interaction mechanism between the microbiome and the host. Metaproteomics technology based on liquid chromatography-mass spectrometry can perform high-precision qualitative and quantitative analysis of proteins in complex samples and has been widely used in the field of disease mechanism research and biomarker mining. Rafael Valdés-Mas et al. used metaproteomics methods to conduct a comprehensive analysis of human intestinal microbial communities, hosts, and dietary exposures, and discovered some microbial proteins associated with inflammatory bowel disease. These microbial proteins can be used as biomarkers to assess the functional status of intestinal microorganisms (Cell, 2025, 188(4):1062-1083).

[0005] Based on this, the present invention uses a metaproteomics approach to construct, for the first time, a panel of vaginal microbiome protein markers for the diagnosis of cervical HSIL. This marker combination not only facilitates early identification of cervical HSIL patients and optimizes clinical intervention strategies, but also provides an important basis for a deeper understanding of the functional mechanisms of the vaginal microbiome in cervical precancerous lesions and the development of novel microecological modulation therapies.

[0006] The information disclosed in this background technology section is only intended to enhance understanding of the overall background of the invention and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the Invention

[0007] In view of the shortcomings of the existing technology, the present invention provides a vaginal flora protein marker group related to high-grade squamous intraepithelial lesions (HSIL) of the cervix and its application.

[0008] To achieve the above object, the present invention adopts the following technical solutions:

[0009] Application of vaginal flora protein marker panel as biomarker in the preparation of products for diagnosis of cervical HSIL:

[0010] The biomarkers include one or more of protein combination 1, protein combination 2, protein combination 3 or protein combination 4;

[0011] The protein combination 1 is selected from the genus Prevotella, including proteins I4ZBF9, A0A134B2X0, A0A134BDM5, I4Z7G8, and I4ZAD6;

[0012] The protein combination 2 is selected from the genus Lactobacillus, including proteins C2E762, A0A109DEN7, and A0A120DKA4;

[0013] The protein combination 3 is selected from the genus Megasphaera, including proteins A0A2J8BAI4;

[0014] The protein combination 4 is selected from the genus Streptobacillus moniliformis and includes protein D1AYN8.

[0015] Preferably, the product comprises a reagent or a kit.

[0016] Preferably, among the biomarkers, I4ZBF9, C2E762, A0A134B2X0, A0A2J8BAI4, A0A134BDM5, A0A120DKA4, and I4ZAD6 are upregulated in patients with cervical HSIL, while A0A109DEN7, D1AYN8, and I4Z7G8 are downregulated in patients with cervical HSIL.

[0017] By adopting the above-mentioned technical solution, the establishment of this marker panel fills the gap in the existing technology for the lack of functional protein markers, and provides a new molecular detection target for the accurate diagnosis of cervical precancerous lesions. By analyzing the expression characteristics of these protein markers, not only can early identification of HSIL patients be achieved and guidance can be provided for the development of individualized intervention plans in clinics, but the functional changes in the vaginal microecology reflected by these markers also provide important clues for revealing the molecular regulatory mechanisms of the microbiome in the occurrence and development of cervical lesions, and also lay a theoretical foundation for the development of therapeutic strategies that target the regulation of the vaginal microenvironment. This discovery elevates the diagnosis of cervical lesions from traditional morphological observation to the level of functional proteomics, and has important clinical translational value.

[0018] Preferably, the diagnosis of cervical HSIL is achieved by liquid chromatography-mass spectrometry.

[0019] The above technical solution is adopted to diagnose whether a patient has cervical HSIL by detecting biomarkers through liquid chromatography-mass spectrometry, which has the detection characteristics of high sensitivity, high accuracy and strong anti-interference.

[0020] The method for screening a vaginal flora protein marker panel comprises the following steps:

[0021] 1) Vaginal flora samples were collected from patients with cervical HSIL and a control group and pretreated. Vaginal flora samples were collected, centrifuged to obtain precipitates, and proteins were extracted and enzymatically hydrolyzed to obtain peptides.

[0022] 2) Detecting the vaginal flora sample pretreated in step 1) using liquid chromatography-mass spectrometry;

[0023] 3) Processing the raw data obtained in step 2) to screen out all differentially expressed proteins associated with cervical HSIL;

[0024] 4) Performing machine learning feature weight analysis on the differentially expressed proteins screened in step 3) to select the top 10 vaginal flora proteins with the largest relative weights and containing annotation information as the vaginal flora protein marker group associated with cervical HSIL.

[0025] Preferably, in step 1), the vaginal flora sample is vaginal secretions; and the control group sample is derived from vaginal secretions of patients with cervical LSIL.

[0026] Preferably, in step 2), the liquid chromatography-mass spectrometry is specifically ultra-high performance liquid chromatography-tandem mass spectrometry; the liquid chromatography uses a reverse phase chromatography column to separate and detect peptides produced by protein hydrolysis.

[0027] Furthermore, in step 3), the data processing method includes ① protein database construction and annotation; ② protein database retrieval; ③ missing value filling; ④ searching for differential vaginal flora proteins.

[0028] Furthermore, in step 4), the machine learning feature weight analysis uses SelectFromModel + random forest (SFM + RF) as the classification model, and determines the protein marker group based on the product value of the frequency of features retained in leave-one-outcross validation (LOO-CV) and the average feature importance.

[0029] By adopting the above technical solutions, more precise and effective screening methods can screen protein markers more accurately and quickly. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 The volcano plot group includes Figure A, which is a volcano plot of vaginal flora protein expression in cervical HSIL / cervical LSIL; Figure B is a volcano plot of human protein expression in cervical HSIL / cervical LSIL;

[0031] Figure 2 Schematic diagram of the vaginal flora protein marker panel screening scheme;

[0032] Figure 3 This is the total ion chromatogram (TIC) of vaginal samples detected by liquid chromatography-mass spectrometry;

[0033] Figure 4 The fragment ion chromatograms and mass spectra of peptides detected in vaginal samples using liquid chromatography-mass spectrometry, taking the _TWGYNAPLLGK_.2 peptide of A0A109DEN7 protein as an example. DETAILED DESCRIPTION

[0034] The present invention is further described below by specific examples. In order to make the invention purpose, technical solution and beneficial technical effect of the present invention clearer, the present invention is further described in detail below with reference to the examples. It should be understood that the embodiments described in this specification are only for explaining the present invention and are not intended to limit the present invention.

[0035] Unless otherwise specified, the experimental methods used in the examples of this invention are conventional methods; all materials and reagents used are commercially available. The data acquisition system used in the examples primarily consisted of an ultra-high performance liquid chromatography (UPLC) and a tandem mass spectrometer (timsTOF Pro2) from Bruker. The chromatographic column used was an IonOpticks AUR3-15075 reversed-phase column (15 cm × 75 μm, 1.7 μm, 120 Å). Mass spectrometry-grade water, acetonitrile, and formic acid were purchased from Thermofisher.

[0036] Example 1

[0037] ① Vaginal flora sample collection, protein extraction and enzymatic hydrolysis

[0038] A total of 172 subjects were recruited, including 103 with cervical LSIL and 69 with cervical HSIL. Recruitment criteria included premenopausal women aged 18 years or older, HPV positive, not pregnant, breastfeeding, or menstruating, and not suffering from hepatitis, diabetes, alcoholism, or drug addiction.

[0039] The subject was positioned in the lithotomy position. A sterile speculum was used to expose the posterior vaginal fornix. A sterile brush was used to collect a secretion sample, avoiding the use of lubricants. The sample was immediately suspended in 10 mL of pre-chilled phosphate-buffered saline (PBS). After transport to the laboratory, the suspended sample was centrifuged at 12,000 rpm for 10 minutes at 4°C to obtain a pellet, which was washed twice with deionized water and stored at -80°C until use.

[0040] Preferably, during the last wash, for the samples in the LSIL and HSIL groups, 10 μL of each sample is taken from each group and mixed into one tube to serve as the mixed samples of the LSIL and HSIL groups, respectively.

[0041] Protein extraction and enzymatic digestion procedures were as follows: 300 μL of lysis buffer (containing 1% SDS and 8 M urea) was added to each precipitate sample (excluding the pooled sample) and sonicated on ice for 5 minutes (90 W power, 2 seconds on / 2 seconds off). After centrifugation at 12,000 rpm at 4°C for 10 minutes, the supernatant was collected and protein concentration was determined using the BCA assay. A 200 μg protein solution was adjusted to 180 μL with 8 M urea solution, and 20 μL of 1 M TEAB and 4.08 μL of 0.5 M TCEP reducing agent were added, followed by shaking at 37°C and 600 rpm for 1 hour. In the dark, 8.50 μL of 1 M IAA alkylating agent was added, and the reaction was continued at 25°C and 600 rpm for 45 minutes. Proteins were precipitated with 1 mL of pre-chilled acetone (-20°C) for 4 hours. After centrifugation, the protein was washed twice with 90% cold acetone and reconstituted in 0.1 M TEAB to a final concentration of 1 μg / μL. Trypsin was added at an enzyme-substrate ratio of 1:50, and enzymatic digestion was performed at 37°C and 750 rpm for 16 hours. The resulting peptides were desalted on a C18 column, and their concentrations were determined using a Pierce quantitative colorimetric peptide kit. The peptides were then lyophilized and stored at -80°C.

[0042] ②Ultra-performance liquid chromatography-tandem mass spectrometry was used to detect vaginal flora samples

[0043] Lyophilized peptides were reconstituted in mobile phase A (0.1% formic acid in water) to a final concentration of 100 ng / μL. After centrifugation at 15,000 rpm for 30 minutes at 4°C to remove insoluble matter, 1 μL of the supernatant was injected. Separation was performed on an IonOpticks AUR3-15075 reversed-phase column (15 cm × 75 μm, 1.7 μm, 120 Å). The column temperature was maintained at 60°C at a flow rate of 300 nL / min. A gradient elution was constructed using phases A and B (0.1% formic acid in acetonitrile): 0-3-16-24-27-30 min, with a gradient of 2%-6%-22%-31%-72%-72% phase B.

[0044] Mass spectrometry analysis was performed on a timsTOF Pro2 instrument in diaPASEF mode with a 30-minute acquisition cycle. For primary signal acquisition, the m / z scan range was set from 100 to 1700 Da, and the ion mobility scan range was set from 0.64 to 1.40 1 / K0. For secondary signal acquisition, the m / z scan range was set from 400 to 1000 Da, with an isolation window of 24 × 25 Da defined; the ion mobility scan range was set from 0.64 to 1.37 1 / K0. Collision energy increased linearly with mobility, from 0.6 1 / K0 at 20 eV to 1.6 1 / K0 at 59 eV.

[0045] ③Data processing and screening of differential vaginal flora proteins

[0046] The steps for constructing a protein database are as follows: For the human protein database, download the standard database directly from the UniProt website (https: / / www.uniprot.org). For the vaginal microbiome protein database, preferably, perform metagenomic sequencing on the mixed sample obtained in step 1.1, translate the resulting base sequences, after removing human genes, into amino acid sequences, and annotate them for database construction. Combine the human and vaginal microbiome protein databases to construct a mixed database.

[0047] Protein database searches were performed as follows: raw data were processed using the Spectronaut 19 platform (Biognosys, Switzerland), and standard proteomics analysis workflows were used to identify all proteins and peptides. The protein database was the hybrid database described above, with the enzyme digestion parameters set to Trypsin / P (allowing cleavage at the N-terminus of proline). The fixed modification was designated as cysteine ​​carboxylmethylation [Carbamidomethyl (C)], and the variable modifications included protein N-terminal acetylation [Acetyl (Protein N-term)] and methionine oxidation [Oxidation (M)].

[0048] Based on the above identification results, 1 / 5 of the minimum identification intensity of each protein was used to fill the missing value of the protein identification. Furthermore, the average expression level of each protein in the LSIL and HSIL groups was calculated, and the fold change (FC) was calculated. The t-test was used to calculate the significance of the expression difference (p value) of each protein in the LSIL and HSIL groups. Furthermore, proteins with |log2FC| > 0.585 and p < 0.05 were screened as differentially expressed proteins associated with cervical HSIL ( Figure 1 ), a total of 690 proteins were obtained.

[0049] ④Machine learning feature weight analysis and marker panel screening

[0050] The SFM + RF model was constructed using Python (version 3.8.19) and the scikit-learn machine learning package (version 1.3.0). The model construction steps included calling the make_pipeline function in scikit-learn to construct the pipeline; first, applying a log10 transformation and a Pareto transformation to the selected protein data; then, calling the SelectFromModel function in scikit-learn to perform feature selection, using the SGDClassifier (alpha=0.05, penalty="l1") model; and finally, calling the RandomForestClassifier function in scikit-learn for machine learning training.

[0051] The frequency and average feature importance of each protein retained in LOO-CV were calculated, and the two were multiplied and sorted from large to small. The top 10 vaginal flora proteins with the largest relative weight and annotated information were selected as the vaginal flora protein marker group associated with cervical HSIL (Table 1). Table 1 shows the specific information of the vaginal flora protein marker group for the diagnosis of cervical HSIL, as follows:

[0052]

[0053] In the cervical HSIL group, I4ZBF9, C2E762, A0A134B2X0, A0A2J8BAI4, A0A134BDM5, A0A120DKA4, and I4ZAD6 were upregulated, while A0A109DEN7, D1AYN8, and I4Z7G8 were downregulated.

[0054] In summary, this study identified a panel of vaginal microbial protein markers for the diagnosis of cervical HSIL by analyzing vaginal microbial samples from 172 subjects using liquid chromatography-mass spectrometry, statistical analysis, and machine learning. This study is crucial for identifying patients with cervical HSIL and optimizing clinical intervention strategies. It also provides valuable insights into the functional mechanisms of vaginal microbiome in cervical precancerous lesions.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit the scope of protection of the present invention. Those skilled in the art should understand that the technical solution of the invention can be modified or replaced by equivalents without departing from the essence and scope of the technical solution of the present invention.

Claims

1. Application of vaginal flora protein marker panel as biomarker in the preparation of products for diagnosis of cervical HSIL: The biomarkers include one or more of protein combination 1, protein combination 2, protein combination 3 or protein combination 4; The protein combination 1 is selected from the genus Prevotella, including proteins I4ZBF9, A0A134B2X0, A0A134BDM5, I4Z7G8, and I4ZAD6; The protein combination 2 is selected from the genus Lactobacillus, including proteins C2E762, A0A109DEN7, and A0A120DKA4; The protein combination 3 is selected from the genus Megasphaera, including proteins A0A2J8BAI4; The protein combination 4 is selected from Streptobacillus moniliformis and includes protein D1AYN8.

2. The use according to claim 1, characterized in that Among the biomarkers, I4ZBF9, C2E762, A0A134B2X0, A0A2J8BAI4, A0A134BDM5, A0A120DKA4, and I4ZAD6 were upregulated in patients with cervical HSIL, while A0A109DEN7, D1AYN8, and I4Z7G8 were downregulated in patients with cervical HSIL.

3. The use according to claim 1, characterized in that The diagnosis of cervical HSIL is achieved by liquid chromatography-mass spectrometry.

4. A method for screening a panel of vaginal flora protein markers, characterized in that: The steps include: 1) Vaginal flora samples were collected from patients with cervical HSIL and a control group and pretreated. Vaginal flora samples were collected, centrifuged to obtain precipitates, and proteins were extracted and enzymatically hydrolyzed to obtain peptides. 2) Detecting the vaginal flora sample pretreated in step 1) using liquid chromatography-mass spectrometry; 3) Processing the raw data obtained in step 2) to screen out all differentially expressed proteins associated with cervical HSIL; 4) Performing machine learning feature weight analysis on the differentially expressed proteins screened in step 3) to select the top 10 vaginal flora proteins with the largest relative weights and containing annotation information as the vaginal flora protein marker group associated with cervical HSIL.

5. The use according to claim 1, characterized in that The product includes a reagent or a kit.

Images