Blood biomarker for diagnosing hypoovarian reserve function, kit and application of blood biomarker

Abstract

The invention relates to the technical field of biomarkers, in particular to a blood biomarker combination for prediction and auxiliary diagnosis of ovarian reserve hypofunction (DOR), a corresponding detection kit and application of the blood biomarker combination. According to the invention, the content change of the nicotinamide mononucleotide (NMN) and the metabolite combination thereof in serum or plasma is screened for the first time and can be used as a detection index for predicting the ovarian reserve function decline. Based on the biomarker combination, the invention further provides an in-vitro detection kit containing a related standard substance, an internal standard solution and an extracting solution, and a data analysis method combined with a random forest model, so that an individual with a normal ovarian reserve function and an individual with hypofunction can be accurately distinguished. The method is noninvasive, simple, convenient, high in sensitivity and suitable for early recognition and layered screening of clinical ovarian reserve hypofunction.

Description

Technical Field

[0001] The present invention relates to the field of biomarkers, and in particular, to a urine biomarker, a kit and applications thereof for diagnosing decreased blood reserve function. Background Art

[0002] With the trend of delaying pregnancy in modern society, diminished ovarian reserve (DOR) has become one of the most common challenges in clinical reproductive medicine. DOR refers to a decrease in oocyte quantity and / or quality, leading to insufficient ovarian function. While DOR occurring after age 40 is typically physiological, experiencing DOR earlier in life can lead to a premature decline in reproductive function and infertility. The most common clinical manifestations of DOR are menstrual irregularities, endocrine disorders, poor response to ovarian stimulation, and infertility. Left untreated, DOR can progress to premature ovarian failure within a few years. DOR accounts for approximately 20% of ovarian diseases in women and approximately 10% of infertile individuals. Its incidence is increasing annually, and the age of patients is trending younger, severely impacting the reproductive health and quality of life of many women. Due to its unclear mechanisms, complex clinical manifestations, and difficulty in prevention, as well as the lack of early screening markers, targeted treatment is difficult and the clinical prognosis is poor. Therefore, how to early identify and restore ovarian function in DOR patients and precisely address the bottleneck of declining fertility is an urgent issue of concern to human well-being.

[0003] Currently, DOR is primarily diagnosed based on a decrease in antral follicle count (AFC) and anti-Müllerian hormone (AMH), as well as elevated serum basal follicle-stimulating hormone (FSH) levels. However, DOR diagnosed by these methods already presents a severely diminished ovarian reserve, making existing diagnostic methods incapable of early identification and intervention. These issues have made the diagnosis and treatment of DOR a hot topic and a challenge in the fields of gynecology and reproductive medicine. Therefore, developing an easily detectable, noninvasive biomarker diagnostic method to predict, stratify, and diagnose diminished ovarian reserve is of great significance for the early diagnosis and intervention of this condition.

[0004] Diminished ovarian reserve is closely associated with mitochondrial dysfunction. NMN and its metabolites participate in multiple mitochondrial metabolic processes and can effectively reflect mitochondrial function at an early stage. Recently, new research has shown that NMN can reverse oocyte quality and effectively improve mitochondrial function in oocytes. Therefore, changes in the abundance of nicotinamide mononucleotide and its metabolites in vivo are expected to provide a new path for the diagnosis of diminished ovarian reserve.

[0005] Currently, there are no commercially available diagnostic kits that can detect DOR early and with high sensitivity using nicotinamide mononucleotide and its metabolites in blood samples, hindering the identification, stratification, and treatment of DOR. This method, for the first time, combines metabolomics dynamic networks with machine learning to address the challenges of early warning and classification of DOR. Therefore, using an easily detectable and accessible substance for diagnosis and prediction of DOR is of great significance for its early diagnosis and intervention. Summary of the Invention

[0006] The present invention provides a group of markers that can be used to predict diminished ovarian reserve function, and further provides a kit for detecting diminished ovarian reserve function comprising the markers and a method for using the kit.

[0007] In one aspect, the present invention provides a set of biomarkers useful for diagnosing diminished ovarian reserve. The biomarkers include multiple metabolites selected from nicotinamide mononucleotide and its metabolite components in blood samples, which are differentially expressed in target plasma or serum compared with a control group. These biomarkers can be used to diagnose diminished ovarian reserve in a subject.

[0008] In one embodiment, a biomarker of the present invention refers to a metabolite component present in a biological sample from a subject. The subject can be a human or a mammal. The biological sample can be selected from plasma or serum derived from the subject. Preferably, the biological sample is serum.

[0009] The marker is used to prepare a kit for testing the ovarian reserve function of a subject. The method of using the kit includes testing the level of the marker in the serum or plasma of the subject to test the ovarian reserve function of the subject.

[0010] In another aspect of the present invention, a diagnostic kit for diminished ovarian reserve function is provided, comprising 15 metabolites. The kit comprises nicotinamide mononucleotide (NMN) and its related metabolites, including standard solutions of nicotinamide riboside (NAR), quinolinic acid (QA), tryptophan (TRP), aspartic acid (ASP), kynurenine (KYN), 3-hydroxyanthranilic acid (3HAA), nicotinic acid mononucleotide (NAMN), dihydropyridine mononucleotide (DHMONAD), nicotinamide adenine dinucleotide (NAAD), methylnicotinamide (MNAM), nicotinic acid (NA), 6-hydroxynicotinic acid (6-Hydroxynicotinic acid), nicotinic acid riboside (Nicotinate-D-ribonucleoside), and L-aspartate.

[0011] Another aspect of the present invention provides a diagnostic kit for ovarian reserve deficiency with 15 metabolites, the kit comprising a standard solution of nicotinamide mononucleotide and its metabolites and an internal standard solution, the internal standard solution being an isotope-labeled nicotinamide mononucleotide and its metabolites, the isotope labeling method being 2 H or 13 C.

[0012] In one embodiment, the kit further comprises an extraction solution composed of methanol and acetonitrile in a volume ratio of 1:1 to 5:1. Furthermore, the kit further comprises a 700 μL 96-well plate, a 350 μL V-shaped 96-well plate, a plate sealing silica gel, and a 96-well sealing aluminum film.

[0013] Another aspect of the present invention provides a method for using the above-mentioned kit for diagnosing diminished ovarian reserve function, specifically, by measuring the levels of nicotinamide mononucleotide and its metabolites in the serum or plasma of the subject, and inputting these measured values into a random forest model to obtain a score cutoff value for judgment.

[0014] In one embodiment, the kit is used for diagnosing diminished ovarian reserve, comprising the following steps: a) preparing metabolite standard solutions of varying concentrations: preparing solutions of varying concentrations using standard solutions of nicotinamide mononucleotide and its metabolites, respectively, and placing the solutions together with a blank control in centrifuge tubes. The solutions are centrifuged in a tabletop centrifuge at 4,000-10,000 rpm for 10-30 minutes; adding 200 μL of freshly prepared deionized water to each centrifuge tube, shaking vigorously, and dissolving the solution at 800-1200 rpm for 10-15 minutes, allowing the solution to stand for use; b) preparing an internal standard solution: adding 3 mL of methanol as an internal standard diluent to the internal standard solution, capping the tube, shaking vigorously, and allowing the solution to stand for approximately 15 minutes to dissolve. Dilute the internal standard solution and add it to a 96-well microplate. c) Serum or plasma sample preparation: Remove the 700 μL microplate provided in the kit and add 5 μL of Standards 1 to 7 and a blank control to wells A1 to A8, in that order. Add 5 μL of serum (or plasma) sample or 5 μL of low-, medium-, or high-concentration quality control to the remaining wells. Add 25 μL of the internal standard solution to each well, cover with a silicone lid, and shake at 1000 rpm for 10 minutes. Centrifuge at 2000 g for 2 minutes. Gently remove the silicone cover to prevent liquid from splashing out of the microplate, properly place the silicone pad to prevent contamination, and set aside; react at 30°C and 1450 rpm for 60 minutes, carefully remove the silicone pad, properly place the silicone pad to prevent contamination, and set aside; add 350 μl of loading buffer to each well, cover with the silicone cover and shake vigorously; place at -20°C for 20 minutes, and centrifuge at 2000g for 20 minutes; remove the silicone cover, carefully pipette 150 μl of supernatant into a clean V-bottom microplate, cover with aluminum foil, and place in an automatic sampler; d) measure the sample prepared in c) by liquid chromatography and mass spectrometry, and calculate the concentration of metabolites in the sample; e) input the concentration of the biomarker calculated in d) into the random forest model for calculation, and determine the ovarian reserve function impairment of the subject according to the score.

[0015] In one embodiment, the random forest model used in the present invention can be software available on the market, which can also be optionally used as a part of the above-mentioned kit. As a preferred embodiment, the software package is used as a part of the kit. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 : NAD Metabolic pathway (part) and annotation of key metabolites in the pathway; Figure 2 : ROC curve of the diagnostic value of nicotinamide mononucleotide and its metabolite combination; Figure 3 : ROC curve of the diagnostic model in the external validation set; Figure 4 : ROC curve of the diagnostic value of the clinical indicator combination in the external validation set; Figure 5 : ROC curves for the diagnostic ability of the nicotinamide mononucleotide combination; DETAILED DESCRIPTION

[0017] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0018] Example 1 Screening of differential biomarkers The experimental samples used in this study were approved by the local ethics committee, and informed consent was obtained from all subjects. A total of 110 subjects were enrolled in this study. Serum samples of nicotinamide mononucleotide and its metabolites were measured using pyrolysis gas chromatography-mass spectrometry (Py-GCMS) in 59 patients with normal ovarian reserve (NOR) and 51 patients with confirmed diminished ovarian reserve (DOR). All subjects were recruited from the Reproductive Center of Peking University Third Hospital. The diagnostic criteria for diminished ovarian reserve (DOR) in women were as follows: a basal follicle-stimulating hormone (FSH) level ≥ 10 IU / L; an anti-Müllerian hormone (AMH) level < 1.1 ng / mL; and a bilateral antral follicle count (AFC) < 7. The control group consisted of women from the general community or those seeking medical attention for physical examinations, with regular menstrual cycles, normal ovarian morphology, and normal hormone levels. Exclusion criteria for either the DOR or control groups included: 1) age younger than 20 or older than 40 years. 2) Body mass index (BMI) less than or greater than 3. 3) Pregnant, lactating, or postmenopausal women. 4) History of surgery related to the uterus, ovaries, or fallopian tubes. 5) Polycystic ovary syndrome, hyperprolactinemia, hyperandrogenism, diabetes, thyroid dysfunction, adrenal dysfunction, or other endocrine diseases that affect ovarian function. 6) Women with cancer or who have received radiotherapy or chemotherapy.

[0019] The detection method is as follows: 1. Serum Sample Collection and Preparation 5 mL of fasting venous blood was collected from the subjects and placed in a plastic centrifuge tube.

[0020] Serum preparation: 1) Slowly invert the serum preparation tube 5 times.

[0021] 2) Place the test tubes vertically in a test tube rack at room temperature (approximately 25 degrees Celsius) for 1.5 hours.

[0022] 3) Centrifuge the tube at 2500 rpm for 10 minutes (4 degrees Celsius).

[0023] 4) Use a pipette to distribute the supernatant (about 2.5 ml) into plastic centrifuge tubes (Eppendorf, 1.5 ml centrifuge tubes), with 0.5 ml of serum in each cryotube.

[0024] 5) Label the sample number on the centrifuge tube.

[0025] 6) Place in a -80 degree Celsius refrigerator immediately.

[0026] 2. Serum clinical marker detection Hematology and biochemistry testing were performed using the LH750 hematology analyzer and the Synchron DXC800 clinical system according to the manufacturer's protocols. Specific test parameters are listed in Table 1.

[0027] Table 1. Basic characteristics of the included population

[0028] 3. Detection of Nicotinamide Mononucleotide and Its Metabolites in Serum Sample preparation: Slowly thaw serum samples at 4°C. Transfer 100 μL of serum to a 1.5 mL centrifuge tube and add 150 μL of methanol (containing the internal standard nicotinamide). Vortex and mix for 10 minutes, let stand for 10 minutes, and then centrifuge at 13,500 rpm for 20 minutes at 4°C. The supernatant was analyzed by UPLC-TQMS (ultra-performance liquid chromatography-triple quadrupole mass spectrometry).

[0029] Analytical instrument testing: UPLC-TQMS: A Waters ultra-high performance liquid chromatography system (Waters, USA) equipped with a binary solvent controller and sample control chamber was used. A Waters XEVO triple quadrupole mass spectrometer (Waters, USA) equipped with a dual electrospray ionization source was used.

[0030] Chromatographic conditions: UPLC BEH C18 column (100mm x 2.1mm, 1.7m); column temperature 45℃; mobile phase A: water (0.1% formic acid), B: acetonitrile (0.1% formic acid); flow rate 0.4mL / min; injection volume 5uL; gradient elution conditions: 0-1min (5%B), 1-5min (5-25%B), 5-15.5min (25-40%B), 15.5-17.5min (40-95%B), 17.5-19min (95%B), 19-19.5min (95-5%B), 19.6-21min (5%B).

[0031] Mass spectrometry conditions: The electrospray ionization source was used in negative ion scanning mode (ESI-). Specific conditions were as follows: capillary voltage 1.2 kV, cone voltage 55 V, extraction cone voltage 4 V, ion source temperature 150°C, desolvation temperature 550°C, reverse cone gas flow 50 L / h, desolvation gas flow 650 L / h, low mass resolution 4.7, high mass resolution 15, and data were collected in multiple reaction detection mode. The results are shown in Table 2: Table 2. Sample test results

[0032]

[0033]

[0034]

[0035]

[0036]

[0037]

[0038] In the DOR index, "1" represents confirmed DOR and "0" represents healthy volunteers; all data are processed by log2(x+1) Example 2 Serum marker combination to distinguish between normal and decreased ovarian reserve groups 2.1 Model establishment: To distinguish patients with normal ovarian reserve from those with diminished ovarian reserve, we used t-tests to select and identify candidate biomarkers. We selected 15 metabolites (listed in Table 2) that showed significant differences between the two groups as candidate variables and used a random forest model to evaluate the candidate variables and build a model.

[0039] Random Forest Model Establishment Method and Related Parameter Selection: The relevant models were established in the R-Studio environment using the R language. The raw data were first cleaned and preprocessed, and the data were divided into a 7:3 ratio to create training and validation sets. Furthermore, serum metabolites with significant differences between the two groups were selected as input features for constructing the random forest model. A diagnostic model based on the randomForest function (Ntree=500 and mtry=4) was established. The pROC was used to plot the receiver operating characteristic (ROC) curve and assess the 95% confidence interval (CI) of the area under the curve (AUCs) to measure the predictive performance of the random forest in the training and external validation sets.

[0040] The random forest model labeling method includes the following steps: 1) Serum metabolome data were collected from individuals in the normal ovarian reserve (NOR) group and the decreased ovarian reserve (DOR) group. The obtained raw metabolite characteristic data were uniformly processed, including missing value filling, standardization, and normalization, and then divided into training and test sets. 2) Build a random forest model based on the training set. Specifically, a decision tree-based learner is used to build the forest model. Each tree generates a training subset based on bootstrap sampling. At each split node, a preset number of features are randomly selected from all features for splitting. This process is repeated to build multiple trees (the number of trees T in this example is 500). 3) For the trained random forest model, calculate the predicted probability value for the input sample. That is, summarize the voting results of all decision trees to obtain the predicted probability that a certain input sample is a DOR individual: 4) Based on the ROC curve analysis results, the optimal diagnostic threshold τ is determined. In this embodiment, τ is set to 0.495. If the ovarian reserve is ≥ 1 / 2, the individual is judged to have diminished ovarian reserve (DOR); otherwise, the individual is judged to have normal ovarian reserve (NOR).

[0041] 5) AUC (area under the curve) was used to evaluate model performance, and confidence interval (95% CI) was used for statistical evaluation to verify the diagnostic efficacy and clinical promotion value of the model.

[0042] 60 patients with normal ovarian reserve (NOR) and 50 patients with confirmed diminished ovarian reserve (DOR) were selected. The random forest model based on biomarker combination was used to output the probability that the above subjects had diminished ovarian reserve. The optimal cutoff value was found through the Youden optimal point in the ROC analysis, and the overall ability of the model to distinguish patients with diminished ovarian reserve from healthy people was evaluated.

[0043] The results are as follows Figure 2 As shown: the area under the ROC curve is 1, the optimal cutoff value is 0.514, and the sensitivity and specificity percentages at the optimal cutoff value are 1 and 1 respectively. After these measured values are brought into the random forest model, the specific calculation is as follows: Assume that the biomarker measured value of a subject is represented as a feature vector: The model consists of 500 decision trees, and the output of each tree is:

[0044] Then the DOR risk probability of this individual is:

[0045] Where x represents the biomarker measurement value of the subject, and each x i Corresponds to a specific biomarker; F t represents the feature subset used by the t-th tree, i.e., 4 features are randomly selected from all n features. ) is greater than 0.514, indicating that the individual has a higher risk of suffering from diminished ovarian reserve; if the individual score threshold is less than or equal to 0.514, it indicates that the individual has a lower risk of suffering from diminished ovarian reserve.

[0046] 2.2 Verification of the model’s diagnostic (differentiation) capabilities: To further verify the diagnostic capability of the above model, we further included 16 DOR patients and 30 normal subjects. All subjects were recruited from the Reproductive Center of Peking University Third Hospital. The inclusion and exclusion criteria were the same as above. The subjects were tested for the concentration of related metabolites in their serum samples using the kit of the present invention. The test results are shown in Table 3. After these measured values were brought into the diagnostic model, the results were as follows: Figure 3 As shown in the figure, the area under the receiver operating characteristic (ROC) curve was 0.990, and the sensitivity and specificity at the optimal cutoff were 0.938 and 0.867, respectively. The test results were consistent with our expectations, and the diagnostic model was still able to distinguish DOR patients from normal subjects very well in the external validation set.

[0047] Table 3. Kit test results

[0048]

[0049]

[0050]

[0051] Example 3: Specific Nicotinamide Mononucleotide Combination Distinguishes Normal Ovarian Reserve Group from Reduced Ovarian Reserve Group To further verify the predictive performance of our nicotinamide mononucleotide combination, the sample and data sources were the same as in Example 1. A random forest model was used to evaluate candidate variables and establish a model. The model establishment method was the same as in Example 2. The predictive ability of the clinical indicators FSH, AFC, and AMH combination for predicting DOR was verified. The results are shown in Figure 2. Figure 4 As shown: The area under the ROC curve is 0.965, which is lower than our nicotinamide mononucleotide combination, indicating that our prediction combination has better prediction performance.

[0052] Among all the nicotinamide mononucleotide components displayed by the random forest model, we found that nicotinamide mononucleotide, tryptophan, aspartic acid, and 6-hydroxynicotinic acid played the most important role in predicting ovarian dysfunction. In order to further verify the predictive function of these four components, we used LASSO regression from the initial metabolite combination, specifically including: 1) Z-score normalization of metabolite concentrations; 2) using DOR diagnostic status as the dependent variable, and selecting the optimal penalty coefficient λ through 10-fold cross-validation; 3) retaining non-zero coefficient metabolites to form the core combination. In Example 3, LASSO regression screened out four key metabolites: nicotinamide mononucleotide, tryptophan, aspartic acid, and 6-hydroxynicotinic acid. We continued to use the random forest model to evaluate candidate variables and establish a model. The model establishment method was the same as in Example 2 to verify the predictive ability of the combination of nicotinamide mononucleotide, tryptophan, aspartic acid, and 6-hydroxynicotinic acid for predicting DOR. The results are as follows. Figure 5 As shown in the figure, the area under the ROC curve is 0.979, and the sensitivity and specificity percentages at the optimal cutoff value are 1 and 0.867, respectively. The results show that the combination of these four metabolites also has a good predictive effect when compared with existing test indicators.

[0053] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. Use of a blood biomarker in the preparation of an in vitro detection kit for predicting diminished ovarian reserve, wherein the blood biomarker is nicotinamide mononucleotide and its metabolites, wherein the metabolites include at least one or more selected from the group consisting of nicotinamide riboside, quinolinic acid, tryptophan, aspartic acid, kynurenine, 3-hydroxyanthranilic acid, nicotinic acid mononucleotide, dihydropyridine mononucleotide, nicotinamide adenine dinucleotide, methylnicotinamide, nicotinic acid, 6-hydroxynicotinic acid, nicotinic acid riboside, and L-aspartic acid.

2. The use according to claim 1, characterized in that The in vitro detection kit comprises a standard solution of the above-mentioned metabolite and an internal standard solution, wherein the internal standard solution is isotope-labeled nicotinamide mononucleotide and its metabolites.

3. The use according to claim 2, characterized in that The isotope label is 2H or 13C label.

4. The use according to claim 2, characterized in that The kit also includes an extraction solution consisting of methanol and acetonitrile, with the volume ratio of methanol to acetonitrile being 1:1 to 5:

1.

5. The use according to any one of claims 1 to 4, characterized in that: The method for using the kit comprises the following steps: (1) Collect serum or plasma samples from subjects; (2) using the in vitro detection kit to detect the content of nicotinamide mononucleotide and its metabolites in the sample; (3) The test results are input into the random forest model for calculation, and the auxiliary judgment results of ovarian reserve function decline are output.

6. The use according to claim 5, wherein the detection method in step (2) is liquid chromatography-mass spectrometry analysis.

7. The application according to claim 6, wherein the determination result in step (3) is based on a preset score cutoff value.

8. The application according to claim 7, wherein the random forest model used for calculation is a multi-feature combination prediction model established based on a training set.

Images