Mycobacterium tuberculosis latent infection detection method

Abstract

The invention discloses a method for detecting latent infection of tubercle bacillus, and belongs to the field of tuberculosis detection. According to the method, detection of latent infection of tubercle bacillus is completed through the steps of sample collection, sample treatment, marker detection, data analysis, result judgment and the like. According to the invention, mycobacterium tuberculosis specific extracellular RNA (exoRNA) is selected as a novel biomarker. Compared with the traditional protein or DNA marker, the exoRNA has higher sensitivity and specificity, and can reflect the existence and activity state of the pathogen earlier and more accurately. The application of the marker effectively reduces the risk of cross reaction caused by bacillus calmette guerin vaccine inoculation or nontuberculous mycobacterium infection, thereby improving the reliability of the detection result. According to the invention, a high-throughput sequencing technology and an artificial intelligence algorithm are combined, and an intelligent diagnosis model is constructed. The method not only can realize qualitative judgment of LTBI, but also can finely distinguish infection states by analyzing the combination and expression profile of the exoRNA markers.

Description

Technical Field

[0001] This invention relates to tuberculosis detection, and more particularly to a method for detecting latent tuberculosis infection. Background Technology

[0002] Tuberculosis (TB) is a major infectious disease that seriously threatens human health, and individuals with latent tuberculosis infection (LTBI) represent a potential reservoir for TB development. Accurate and efficient identification of LTBI individuals and subsequent preventative treatment are crucial for controlling TB transmission. Currently, the main clinical methods for LTBI detection include the tuberculin skin test (TST) and the interferon-gamma release assay (IGRA). While the TST is a long-established and relatively inexpensive technique, it suffers from poor specificity and is susceptible to BCG vaccination history and infection by many environmental mycobacteria, leading to frequent false-positive results. Furthermore, it requires subjects to return to a medical institution 48-72 hours later for result interpretation, which is inconvenient. The IGRA technique, on the other hand, determines infection status by detecting the level of interferon-gamma released by T cells after stimulation with Mycobacterium tuberculosis-specific antigens. Its specificity is significantly improved compared to the TST, and it is largely unaffected by BCG vaccination. However, the IGRA test procedure is complex, requiring specialized laboratory equipment and technicians. The collection, transportation, and preservation of blood samples demand stringent conditions, and the testing costs are high, which significantly limits its application in primary healthcare institutions and large-scale population screening. Furthermore, neither TST nor IGRA can effectively differentiate between active tuberculosis and latent infection, thus presenting certain limitations in clinical decision-making.

[0003] Existing technologies are constantly seeking breakthroughs. For example, Chinese patent CN102925541A discloses a nucleic acid detection method based on real-time quantitative fluorescence amplification to detect characteristic genes. However, it still relies on comparing gene expression differences before and after tuberculin stimulation, and the operation remains cumbersome. Patents such as CN104805063A focus on using multiple protein antigen stimulation to detect immune responses, but they are essentially still within the scope of in vitro immunological detection and have not solved the problems of invasive blood collection and complex operation. Summary of the Invention

[0004] Purpose of the invention: The purpose of this invention is to provide a new method for LTBI detection that is non-invasive, convenient, highly specific, and can effectively distinguish different infection states.

[0005] Technical solution: A method for detecting latent tuberculosis infection, comprising the following steps: S1. Sample collection: Collect non-blood body fluid samples from the subject, including urine or sweat; S2. Sample Processing: Extracellular RNA is extracted from the non-blood body fluid sample to obtain an extract. The extraction is performed using an extraction kit that specifically binds to extracellular RNA derived from Mycobacterium tuberculosis complex. The extraction kit contains magnetic beads for enriching exosomes and a lysis buffer for lysing exosomes and releasing RNA. The extraction uses magnetic beads containing antibodies that specifically bind to exosome surface antigens for immunoaffinity capture of exosomes, and the captured exosomes are lysed using the lysis buffer to release RNA. S3. Biomarker detection: High-throughput sequencing analysis was performed on the Mycobacterium tuberculosis complex-specific extracellular RNA biomarkers in the extract, wherein the Mycobacterium tuberculosis complex-specific extracellular RNA biomarkers were selected from the mRNA fragments of the esxG, esxH, espA, cfp10, and esat6 genes. S4. Data Analysis: The sequence data obtained from high-throughput sequencing is compared with a pre-established database containing characteristic extracellular RNA expression profiles of healthy individuals, BCG recipients, latent tuberculosis-infected individuals, and active tuberculosis patients. The results are analyzed using a machine learning model trained on a random forest algorithm to calculate the probability that a subject is in a latent tuberculosis infection state. S5. Result Interpretation: The infection status is interpreted based on the probability value. When the probability value is higher than the preset threshold, it is determined to be a positive case of latent tuberculosis infection. Furthermore, based on the combination and expression level differences of the extracellular RNA markers, the subjects are further distinguished as BCG vaccination reaction, latent tuberculosis infection, or active tuberculosis.

[0006] Furthermore, in step S1, morning urine samples are collected first, and RNA stabilizers are immediately added to the collected urine samples, which are then stored and transported at 4°C; the sweat samples are obtained by wiping the subject's skin surface with a sterile swab.

[0007] Furthermore, in step S2, the extraction kit further includes a purification column for removing impurities and inhibitors from the sample, the lysis buffer contains guanidine salt and surfactant, and the extraction step specifically includes: performing immunoaffinity capture of exosomes using the magnetic beads, lysing the captured exosomes using the lysis buffer, and finally purifying the released RNA using the purification column to obtain a high-purity RNA extract.

[0008] Furthermore, in step S3, the high-throughput sequencing is real-time single-molecule sequencing based on the nanopore sequencing principle, and the portable sequencer used can complete sequencing on-site within 2 hours. Before sequencing, reverse transcription is performed using reverse transcription primers designed for the Mycobacterium tuberculosis complex-specific extracellular RNA markers to generate cDNA. Implementations based on nanopore sequencing may include either a "direct RNA sequencing" mode or a "cDNA sequencing" mode.

[0009] Furthermore, in step S3, the Mycobacterium tuberculosis complex-specific extracellular RNA markers include specific mRNA fragments of the esat6, cfp10, and espA genes.

[0010] Furthermore, in step S4, the machine learning model analyzes the expression levels and ratios of multiple extracellular RNA markers to output a comprehensive risk score. The preset threshold is 0.7. When the comprehensive risk score is higher than 0.7, it is interpreted as positive. At the same time, the model also outputs an infection status classification, indicating whether it is "possibly a BCG reaction", "possibly a latent infection" or "possibly active tuberculosis".

[0011] Furthermore, the method also includes a test result report generation step, which automatically generates a standardized test report from the conclusions of the result interpretation step, the information of the markers used, the probability value and the risk score, and transmits it to a designated medical information system or user terminal via the network.

[0012] Beneficial Effects: This invention achieves truly non-invasive testing by using non-blood fluids such as urine and sweat as test samples, greatly improving subject acceptance and sampling convenience. It is particularly suitable for children, special populations, and large-scale community screening, avoiding the discomfort and infection risks associated with traditional venous blood collection. This invention selects Mycobacterium tuberculosis-specific extracellular RNA (exoRNA) as a novel biomarker. Compared with traditional protein or DNA biomarkers, exoRNA has higher sensitivity and specificity, reflecting the presence and activity of pathogens earlier and more accurately. The application of this biomarker effectively reduces the risk of cross-reactions caused by BCG vaccination or non-tuberculous mycobacterial infection, improving the reliability of test results. This invention combines high-throughput sequencing technology with artificial intelligence algorithms to construct an intelligent diagnostic model. This method not only enables qualitative assessment of LTBI but also allows for precise differentiation of infection status by analyzing the combination and expression profiles of exoRNA biomarkers, such as identifying BCG vaccination reactions, latent infection, and active tuberculosis. This provides more comprehensive and accurate decision support for clinical practice, something existing single-indicator detection methods cannot achieve. The sample processing workflow and portable sequencing technology involved in this invention facilitate standardization and automation of the testing process, reducing reliance on operator skills and potentially significantly lowering testing costs and time. This will promote the widespread application of this technology in resource-scarce areas, demonstrating significant social value and broad market prospects. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0014] To make the technical solution of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0015] Example 1 This embodiment provides a complete workflow for high-precision LTBI detection and genotyping using urine samples in a central laboratory environment. Details are as follows: (1) Sample collection and stabilization Collect approximately 50 mL of midstream morning urine from the subject and immediately mix it with an equal volume of pre-chilled urine preservation solution (such as the Norgen Biotek Urine Preservation Kit). This solution effectively inhibits RNase activity and stabilizes exoRNA. The mixed sample can be temporarily stored at 4°C and transferred to the laboratory within 24 hours. For long-term storage, it should be placed in an ultra-low temperature freezer at -80°C.

[0016] (2) Extraction of extracellular RNA (exoRNA) The procedure was performed using an optimized exosome RNA extraction kit. The specific steps were as follows: First, urine samples were centrifuged at 3000×g for 15 minutes at 4°C to remove cell debris. The supernatant was collected, and magnetic beads specifically binding to exosomes (such as magnetic beads coated with CD63 and CD81 antibodies) were added. The mixture was incubated at room temperature for 30 minutes, and exosomes were captured using a magnetic rack. The supernatant was discarded, and the magnetic bead-exosome complex was washed twice with PBS buffer. Subsequently, lysis buffer containing a dissociation salt (such as guanidine isothiocyanate) and a surfactant (such as Triton X-100) was added, and the mixture was vortexed to completely lyse the exosomes, releasing total RNA (including exoRNA). Finally, the lysis buffer was transferred to a silica gel purification column, and the RNA was purified by centrifugation, washing, and elution with RNase-free water to obtain a high-purity RNA extract.

[0017] (3) Library construction and high-throughput sequencing A certain amount of RNA extract was taken and reverse transcribed using sequence-specific reverse transcription primers designed for Mycobacterium tuberculosis complex-specific genes (including but not limited to esat6, cfp10, espA, esxG, and esxH) to generate cDNA. Subsequently, these specific cDNA fragments were amplified using multiplex PCR, with sequencing adapters and sample index sequences introduced during the PCR process to complete the construction of the sequencing library. After the library passed quality control, paired-end high-throughput sequencing was performed on the Illumina NovaSeq 6000 sequencing platform to ensure sufficient sequencing data for each sample.

[0018] (4) Data analysis and intelligent interpretation The raw data (FASTQ format) generated after sequencing was quality-controlled and filtered, then compared with the Mycobacterium tuberculosis complex reference genome to quantify the expression level (e.g., TPM value) of each specific exoRNA marker. This quantitative data was then input into a pre-trained machine learning model (in this example, a random forest algorithm was used, trained on a large cohort including healthy individuals, BCG-inoculated individuals, LTBI-diagnosed individuals, and active tuberculosis patients). The model input consisted of the normalized expression levels (e.g., TPM values) of each specific exoRNA marker (e.g., esat6, cfp10, espA, etc.) and their relative ratios (e.g., esat6 / cfp10 ratio, espA / esat6 ratio). After receiving the input exoRNA expression profile, the model integrates the prediction results of multiple decision trees to obtain the final probability and outputs a comprehensive risk score (PS) between 0 and 1. At the same time, it provides an infection status classification suggestion (e.g., when the expression level of the espA gene is significantly higher than the expression level combination of esat6 / cfp10, the model tends to classify it as active tuberculosis).

[0019] (5) Results generation and reporting The testing system automatically generates standardized reports. These reports include: subject information, a list of biomarkers detected and their expression levels, a comprehensive risk score (PS) calculated by the model, an infection status classification conclusion, and clinical interpretation suggestions. For example, when PS > 0.7, the system classifies it as LTBI positive and further indicates the infection status based on the expression ratio pattern of esat6 / cfp10 and espA. The reports can be directly transmitted to clinicians through the hospital information system (HIS).

[0020] Example 2 This embodiment provides a portable and rapid testing solution suitable for primary healthcare institutions or on-site screening points. Details are as follows: (1) Sample collection Using a sterile sweat collection swab (e.g., the HIDEF sweat collection kit), vigorously rub the skin of the subject's forearm or back for about one minute to fully absorb the sweat. Immediately afterwards, insert the swab tip into a microcentrifuge tube containing 0.5 mL of RNA stabilizer (such as RNA later), and vortex to dissolve the sweat components in the stabilizer. This sample can be stably stored at room temperature and transported to the testing site.

[0021] (2) Rapid nucleic acid extraction and amplification At the on-site testing site, an integrated miniature nucleic acid extraction and purification instrument and its accompanying reagent kit (such as the Oxford Nanopore VolTRAX system) are used. Sample solutions are added to the wells of a disposable sequencing chip pre-loaded with reagents for exosome lysis, RNA capture, and purification. The device automatically completes all extraction and purification steps, yielding purified exoRNA in approximately 15 minutes.

[0022] (3) Real-time sequencing This embodiment utilizes the MinION portable sequencer from Oxford Nanopore Technologies. The purified RNA is directly mixed with sequencing adapters, eliminating the need for complex PCR amplification steps (direct RNA sequencing mode), and added directly to the sequencing chip. Once the sequencer is started, real-time sequencing can begin.

[0023] (4) Real-time data analysis and interpretation The electronic signals generated by the sequencer are converted into nucleotide sequence data in real time. The device's built-in dedicated software (developed on the MinKNOW platform) analyzes the sequence data in real time. Once a sequence matching a preset Mycobacterium tuberculosis-specific marker (such as esxG, esxH, espA) is identified, counting begins. A simplified machine learning model embedded in the software (such as a logistic regression model) calculates and updates the risk score in real time based on the number and combinations of target sequences read within a short period (e.g., within one hour of sequencing starting). A preliminary report can be issued in advance when the accumulated data reaches a preset confidence level.

[0024] (5) Results Report Test results (such as "LTBI negative" or "further testing recommended") and key indicators (such as the name of the detected marker and the preliminary risk value) are displayed directly on the connected tablet or mobile phone, and a simplified report can be printed immediately, which greatly shortens the testing cycle and meets the needs of rapid on-site screening.

[0025] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for detecting latent tuberculosis infection, characterized in that, Includes the following steps: S1. Sample collection: Collect non-blood body fluid samples from the subject, including urine or sweat; S2. Sample Processing: Extracellular RNA is extracted from the non-blood body fluid sample to obtain an extract. The extraction is performed using an extraction kit that specifically binds to extracellular RNA derived from Mycobacterium tuberculosis complex. The extraction kit contains magnetic beads for enriching exosomes and a lysis buffer for lysing exosomes and releasing RNA. The extraction uses magnetic beads containing antibodies that specifically bind to exosome surface antigens for immunoaffinity capture of exosomes, and the captured exosomes are lysed using the lysis buffer to release RNA. S3. Biomarker detection: High-throughput sequencing analysis was performed on the Mycobacterium tuberculosis complex-specific extracellular RNA biomarkers in the extract, wherein the Mycobacterium tuberculosis complex-specific extracellular RNA biomarkers were selected from the mRNA fragments of the esxG, esxH, espA, cfp10, and esat6 genes. S4. Data Analysis: The sequence data obtained from high-throughput sequencing is compared with a pre-established database containing characteristic extracellular RNA expression profiles of healthy individuals, BCG recipients, latent tuberculosis-infected individuals, and active tuberculosis patients. The results are analyzed using a machine learning model trained on a random forest algorithm to calculate the probability that a subject is in a latent tuberculosis infection state. S5. Result Interpretation: The infection status is interpreted based on the probability value. When the probability value is higher than the preset threshold, it is determined to be a positive case of latent tuberculosis infection. Furthermore, based on the combination and expression level differences of the extracellular RNA markers, the subjects are further distinguished as BCG vaccination reaction, latent tuberculosis infection, or active tuberculosis.

2. The method for detecting latent tuberculosis infection according to claim 1, characterized in that, In step S1, morning urine samples are collected first, and RNA stabilizers are added to the collected urine samples immediately. The samples are then stored and transported at 4°C. Sweat samples are obtained by wiping the subject's skin surface with a sterile swab.

3. The method for detecting latent tuberculosis infection according to claim 1, characterized in that, In step S2, the extraction kit further includes a purification column for removing impurities and inhibitors from the sample. The lysis buffer contains guanidine salt and surfactant. The extraction step specifically includes: performing immunoaffinity capture of exosomes using the magnetic beads, lysing the captured exosomes using the lysis buffer, and finally purifying the released RNA using the purification column to obtain a high-purity RNA extract.

4. The method for detecting latent tuberculosis infection according to claim 1, characterized in that, In step S3, the high-throughput sequencing is real-time single-molecule sequencing based on the nanopore sequencing principle, and the portable sequencer used can complete the sequencing on-site within 2 hours; before sequencing, the process also includes reverse transcription using reverse transcription primers designed for the Mycobacterium tuberculosis complex-specific extracellular RNA markers to generate cDNA.

5. The method for detecting latent tuberculosis infection according to claim 1, characterized in that, In step S3, the Mycobacterium tuberculosis complex-specific extracellular RNA markers include specific mRNA fragments of the esat6, cfp10, and espA genes.

6. The method for detecting latent tuberculosis infection according to claim 1, characterized in that, In step S4, the machine learning model analyzes the expression levels and ratios of multiple extracellular RNA markers and outputs a comprehensive risk score. The preset threshold is 0.7, and a positive result is obtained when the comprehensive risk score is higher than 0.

7.

7. The method for detecting latent tuberculosis infection according to claim 1, characterized in that, It also includes a test result report generation step, which automatically generates a standardized test report from the conclusions of the result interpretation step, the information of the markers used, the probability value and the risk score, and transmits it to the designated medical information system or user terminal via the network.

Images