Glioma treatment regimen prediction system based on gap43 phase separation and phosphorylation detection

Abstract

The application relates to a glioma drug scheme prediction system based on GAP43 phase separation and phosphorylation detection, relates to the field of medical artificial intelligence, and aims to solve the problem that existing detection platforms or diagnosis systems cannot accurately evaluate glioma drug resistance. The system comprises a GAP43 phosphorylation detection module for detecting the GAP43 protein S41 site phosphorylation level in glioma tissue; a GAP43 phase separation imaging module for quantitatively evaluating the clustering rate of GAP43 condensates through fluorescence confocal microscopic imaging and dynamic analysis; a joint score calculation module for jointly calculating the GAP43 phosphorylation level and the clustering rate of GAP43 condensates to calculate a joint score; and a scheme prediction module for predicting a glioma drug scheme based on the score result. The application improves the accuracy and specificity of drug resistance prediction by combining GAP43 phosphorylation and phase separation. The application is used for predicting a glioma drug scheme.

Description

Technical Field

[0001] This invention relates to the field of medical artificial intelligence, and in particular to a glioma medication prediction system based on GAP43 phase separation and phosphorylation detection. Background Technology

[0002] Glioma is one of the most common primary malignant tumors of the central nervous system, with glioblastoma (GBM) being particularly aggressive and having a very poor prognosis. Standard treatment typically includes maximal surgical resection, postoperative radiotherapy, and chemotherapy. Temozolomide (TMZ), as a first-line chemotherapy drug, is widely used for postoperative maintenance therapy in GBM patients. However, numerous clinical studies have shown that the efficacy of TMZ in glioma treatment is limited by its rapidly developing resistance mechanisms. Existing biomarkers cannot accurately and comprehensively predict patient resistance and are also insufficient to guide the selection of combination therapy regimens.

[0003] In recent years, liquid-liquid phase separation (LLPS), as an important intracellular spatial regulatory mechanism, has been increasingly found to be closely related to drug resistance and metabolic regulation in tumor cells. Meanwhile, GAP43 (Growth Associated Protein 43), traditionally considered a neurodevelopmental protein, has been found to be aberrantly expressed in gliomas in recent years. Recent studies have shown that GAP43 is phosphorylated in drug-resistant glioma cells, thereby mediating cellular tolerance to TMZ. Conversely, when GAP43 is dephosphorylated and undergoes phase separation, it can lead to the misrecruitment of MICU1, causing mitochondrial calcium overload and cell death, suggesting that it may act as a key switch for reversing drug resistance.

[0004] Although this mechanism is becoming increasingly clear, there is currently no mature detection platform or diagnostic system that can apply GAP43 phosphorylation status and phase separation behavior to clinical drug resistance assessment and medication guidance. The lack of quantitative detection tools also limits the development of precision treatment strategies. Summary of the Invention

[0005] The present invention aims to address the problem that existing detection platforms or diagnostic systems cannot accurately assess drug resistance in gliomas, and provides a glioma drug regimen prediction system based on GAP43 phase separation and phosphorylation detection.

[0006] This invention relates to a glioma drug regimen prediction system based on GAP43 phase separation and phosphorylation detection, comprising:

[0007] The GAP43 phosphorylation detection module is used to detect the phosphorylation level of the S41 site of the GAP43 protein.

[0008] The GAP43 phase separation imaging module uses fluorescence confocal microscopy and dynamic analysis techniques to quantitatively assess the quantity and distribution of GAP43 condensates.

[0009] The joint scoring calculation module is used to combine the phosphorylation level of GAP43 and the separation state of the phosphorylation phase of GAP43 to calculate the score;

[0010] Treatment plan prediction module: Predicts glioma treatment plans based on scoring results.

[0011] The beneficial effects of this invention are:

[0012] This invention constructs a multidimensional drug resistance scoring system that incorporates both GAP43 phosphorylation and phase separation characteristics by combining these two states. The glioma drug regimen prediction system built upon this combined scoring system can effectively predict drug response, improving the accuracy and specificity of drug resistance prediction.

[0013] The glioma medication prediction system built on this joint scoring system can more accurately identify the response characteristics of different patients or different tumor subgroups to specific drugs, thereby realizing intelligent recommendation and optimization of individualized treatment plans and significantly improving the accuracy and specificity of drug efficacy prediction.

[0014] Furthermore, this invention further validated the significant effect of the combination drug regimen recommended by this prediction system in inhibiting glioma cell proliferation in animal experiments, confirming the feasibility and effectiveness of the system in guiding preclinical drug screening and combination strategy formulation. This invention not only provides a new perspective and analytical tool for studying glioma drug resistance mechanisms, but also provides a scientific basis and technical support for personalized clinical treatment decisions, demonstrating promising clinical application prospects and widespread value. Attached Figure Description

[0015] Figure 1 HE staining results of glioma tissue microarray;

[0016] Figure 2 Immunohistochemical results of p-GAP43 on a microarray of glioma tissue;

[0017] Figure 3 To analyze the relationship between p-GAP43 levels in glioma tissue and overall survival (OS) of patients;

[0018] Figure 4 To analyze the relationship between p-GAP43 levels and TMZ (temozolomide) resistance in glioma tissue;

[0019] Figure 5ROC curve analysis for predicting the sensitivity of gliomas to TMZ based on p-GAP43 levels;

[0020] Figure 6 Confocal microscopy observation of GAP43 aggregates in glioma tissue microarray;

[0021] Figure 7 To analyze the relationship between pathological stage of glioma and GAP43 aggregate expression level;

[0022] Figure 8 ROC curve analysis for predicting glioma prognosis based on GAP43 condensate expression levels;

[0023] Figure 9 This is a comparison chart of tumor size in mice;

[0024] Figure 10 Statistical analysis of changes in tumor weight in mice;

[0025] Figure 11 Confocal microscopy observation of GAP43 aggregates in subcutaneous tumor sections of mice;

[0026] Figure 12 Statistical analysis of the number of GAP43 aggregates in mouse subcutaneous tumor sections;

[0027] Figure 13 Immunohistochemical (IHC) staining results for Ki-67, GAP43, and p-GAP43 in glioma organoids;

[0028] Figure 14 Statistical analysis of the Ki-67 positivity rate in glioma organoids;

[0029] Figure 15 Statistical analysis of the positive rate of p-GAP43 in organoids of gliomas;

[0030] Figure 16 Confocal microscopy observation of GAP43 aggregates in glioma organoids;

[0031] Figure 17 Statistical analysis of the number of aggregates in glioma organoids. Detailed Implementation

[0032] The embodiments of the present invention will be described in detail below. The following embodiments are implemented based on the technical solution of the present invention, and detailed implementation schemes and specific operation processes are given. However, the protection scope of the present invention is not limited to the following embodiments.

[0033] Example 1: Immunohistochemical detection of GAP43 phosphorylation level

[0034] This embodiment provides a method for detecting the phosphorylation level of GAP43 protein at the S41 site in human glioma tissue using immunohistochemistry (IHC). HE staining and p-GAP43 immunohistochemistry (IHC) results of 144 glioma tissue sections were analyzed to preliminarily determine the resistance of tumor cells to temozolomide (TMZ).

[0035] (1) Antigen retrieval in tissue microarrays or paraffin-embedded tissues

[0036] After dewaxing tissue microarrays or paraffin tissues, they were hydrated with graded ethanol, placed in sodium citrate buffer at pH 6.0, and microwaved at high temperature for 10 min, then cooled to room temperature.

[0037] (2) Blocking and Closure

[0038] Incubate with 3% hydrogen peroxide for 10 minutes to block endogenous peroxidase activity; then block with 5% BSA for 30 minutes to reduce nonspecific binding.

[0039] (3) Primary antibody incubation

[0040] Add anti-pS41-GAP43 rabbit monoclonal antibody (recommended concentration 1:100, catalog number such as Abcam ab194929) dropwise and incubate overnight at 4°C.

[0041] (4) Secondary antibody and colorimetric assay

[0042] The next day, after restoring to room temperature for 30 minutes, the rabbit was washed three times with PBS. HRP-labeled anti-rabbit IgG secondary antibody was added and the rabbit was incubated at room temperature for 30 minutes. The rabbit was then reacted with DAB chromogenic solution for 5 minutes, and the degree of color development was observed under a microscope.

[0043] (5) Nucleus staining and mounting

[0044] Counterstain cell nuclei with hematoxylin for 30 seconds, rinse with running water; dehydrate, clear, and mount with resin.

[0045] (6) Setting up positive and negative control

[0046] Negative control: Use rabbit IgG isotype control antibody of equal concentration instead of primary antibody;

[0047] Positive control: Glioma tissue sections with high levels of GAP43 S41 phosphorylation (tissue samples from patients with known drug resistance).

[0048] (7) Results Analysis and Interpretation

[0049] Microscopic observation of the brownish-yellow DAB deposits revealed that GAP43 phosphorylation was mainly located in the cytoplasm.

[0050] The semi-quantitative scoring method (H-score) was used:

[0051] ImageJ software was used to obtain staining intensity scores for immunohistochemical staining images of glioma tissue. First, the image type was converted to RGB Stack using the Type tool. Then, the image threshold was adjusted to 80 / 180 using the Adjust-Threshold tool to clearly distinguish the positive areas from the background. Analyze-measure was then used to divide the final values ​​into four segments: 0 (values ​​0-2), 1 (values ​​3-8), 2 (values ​​9-14), and 3 (values ​​14-30), representing the four levels of staining intensity.

[0052] Staining intensity: 0 (negative), 1 (weak), 2 (moderate), 3 (strong);

[0053] The percentage of positive cells refers to the proportion of cells stained in a field of view of an immunohistochemical staining image, with a value range of 0 to 100.

[0054] Formula for calculating GAP43 phosphorylation level: H-score = (Staining intensity × percentage of positive cells); Full marks: 300 points.

[0055] For each sample, at least three independent representative fields were randomly evaluated, and the average of their H-scores was used as the final quantitative value of GAP43 phosphorylation level for that sample.

[0056] The observed minimum H-score in the 144 samples was Min. Hscore =0.1 × 4% = 0.4, the maximum observed H-score in the 144 samples is Max. Hscore =3×98%=294.

[0057] (8) Clinical validation data

[0058] Analysis of HE staining and p-GAP43 immunohistochemical (IHC) results from tissue microarrays of 144 glioma cases revealed that the staining intensity of p-GAP43 significantly increased with increasing clinical stage. Figure 1 and Figure 2 Kaplan–Meier survival curves showed that the overall survival of patients in the low p-GAP43 expression group (n=76) was significantly higher than that in the high p-GAP43 expression group (n=68) (p=0.009, hazard ratio R=2.094). Figure 3 The results suggest that elevated p-GAP43 levels may be associated with poor prognosis and drug resistance.

[0059] Furthermore, the relationship between p-GAP43 levels and TMZ (temozolomide) resistance in glioma tissue was analyzed. The results showed that patients with low p-GAP43 levels exhibited significantly lower TMZ resistance compared to those with high p-GAP43 levels (p=0.0036). Figure 4 ).

[0060] Receiver operating characteristic (ROC) curves were used to assess the predictive power of p-GAP43 expression levels on TMZ treatment sensitivity, such as... Figure 5 As shown in the figure. The results showed that p-GAP43 level could effectively predict the sensitivity of gliomas to TMZ. Specifically, the AUC value of the ROC curve was 0.635 (p=0.0067), and the TPR (true positive rate) was 78%, while the TNR (true negative rate) was 54%.

[0061] Example 2: Detection of GAP43 phase separation state

[0062] This embodiment uses immunofluorescence staining combined with confocal microscopy to observe the liquid-liquid phase separation state (condensates) of GAP43 protein, which is used to quantitatively assess the phase separation behavior of GAP43 in glioma cells, thereby inferring their tendency to resist chemotherapy drugs.

[0063] (1) Preparation of tissue microarray or paraffin tissue

[0064] FFPE paraffin-embedded tissues that were pathologically confirmed as gliomas were selected and baked at a constant temperature of 60℃ for 2 hours.

[0065] (2) Immunofluorescence staining

[0066] Sections were dewaxed with xylene, rehydrated with graded alcohols, and then subjected to antigen retrieval using pH 6.0 citrate buffer (microwave method, 98℃, 10 min). Endogenous fluorescence background was removed by treatment with 3% hydrogen peroxide for 10 min, followed by washing with PBS. Non-specific binding sites were blocked with 5% BSA and incubated at room temperature for 30 min.

[0067] Add primary antibody against GAP43 (Rabbit anti-GAP43, CST #8945, dilution 1:100) and incubate overnight at 4°C. The next day, wash with PBST and incubate with Alexa Fluor 488-labeled goat anti-rabbit IgG secondary antibody (1:200) at room temperature in the dark for 60 minutes. Counterstain cell nuclei with DAPI (5 μg / mL, 5 min).

[0068] (3) Confocal imaging and GAP43 condensate analysis

[0069] Immunofluorescence staining was followed by imaging using laser confocal microscopy. Random field-of-view images were acquired, and then processed using ImageJ software. First, the image type was converted to 8-bit grayscale using the Type tool. Then, the image threshold was adjusted to 50 / 150 using the Adjust-Threshold tool, converting the image to a binary image. In this case, the phase-separated regions appeared white, and the background was black. A specified threshold algorithm was used to identify granular regions with a signal-to-noise ratio higher than the background as GAP43 aggregates.

[0070] Parameter settings:

[0071] Roundness limit: 0.30-1.00;

[0072] Particle size: >0.2 μm 2 .

[0073] (4) Quantitative evaluation criteria for phase separation state

[0074] GAP43 condensates are defined in cells as multiple clear, round fluorescent clusters appearing in the cytoplasm or protrusions.

[0075] The following parameters are recorded for each field of view:

[0076] Average number of aggregates ( / cell), average area of ​​aggregates (μm) 2 Clustering Rate: The percentage (%) of cells that have ≥3 aggregates.

[0077] Judgment criteria:

[0078] ① High-level Condensates

[0079] Number threshold: ≥3 aggregates / cell (vesicle structures with a diameter <0.1 μm need to be excluded)

[0080] Morphological criteria:

[0081] Individual aggregate area ≥ 0.3 μm 2 (Calibrated via super-resolution imaging)

[0082] Circularity ≥ 0.5 (excluding non-specific protein aggregation)

[0083] Spatial organization characteristics: Clustering rate ≥60%.

[0084] ②Low-level Condensates

[0085] Quantity threshold: <3 aggregates / cells or only diffuse signal present in a single cell

[0086] Morphological criteria: Maximum aggregate area <0.3μm 2

[0087] Spatial organization characteristics: clustering rate ≤30% (or exhibiting random spatial distribution).

[0088] (5) Preliminary conclusions on drug resistance correlation

[0089] In tissue samples from 72 glioma patients, GAP43 aggregate levels were significantly correlated with TMZ sensitivity (p<0.001, Fisher's exact test):

[0090] High-level phase separation group (n=18):

[0091] Stage distribution: Stage I-II accounted for 44.4% (8 / 18), Stage III for 22.2% (4 / 18), and Stage IV for 33.3% (6 / 18).

[0092] Clinical significance: The proportion of low-grade (stage I-II) tumors was significantly higher than that of the low-level group (p=0.0001), suggesting that GAP43 aggregates may inhibit tumor malignant progression through phase separation or enhance TMZ sensitivity.

[0093] Low-level phase separation group (n=54):

[0094] Stage distribution: Stage I-II 0% (0 / 54), Stage III 24.1% (13 / 54), Stage IV 75.9% (41 / 54, revised from the original data 42.6%→23 / 54 to 75.9%).

[0095] Clinical significance: All were high-grade (stage III-IV), and 75.9% were stage IV, suggesting that low GAP43 aggregate levels are strongly correlated with tumor malignancy and TMZ resistance.

[0096] Confocal microscopy observations of GAP43 aggregates in glioma tissue microarrays are as follows: Figure 6 As shown. Figure 6 The results indicate that GAP43 can undergo phase separation under pathological conditions in clinical cases.

[0097] The relationship between glioma pathological stage and GAP43 aggregate expression level is as follows: Figure 7 As shown, statistical analysis (χ² test) revealed that the level of GAP43 aggregates was significantly correlated with pathological stage.

[0098] This embodiment uses ultra-high resolution imaging to quantitatively analyze the characteristics of GAP43 functional aggregates, finding that GAP43 aggregate expression levels are significantly correlated with the prognosis of glioma patients. ROC curve analysis showed that GAP43 aggregate expression levels have significant predictive value for patient prognosis (AUC=0.785, p=0.004). Figure 8 The results suggest that GAP43 aggregates may serve as a potential biomarker for glioma prognosis. Based on this, and combined with the phosphorylation detection results from Example 1, a multidimensional drug resistance scoring system incorporating GAP43 phosphorylation and aggregate characteristics can be further constructed to improve predictive efficacy.

[0099] Example 3: GAP43 Joint Scoring System

[0100] This embodiment provides a scoring method that combines the phosphorylation level of GAP43 S41 site with the phase separation behavior of GAP43 to establish a glioma medication regimen prediction system.

[0101] (a) Setting of scoring dimensions

[0102] The combined score in the glioma treatment regimen prediction system includes two indicators: phosphorylation level of GAP43 protein at S41 site and GAP43 phase separation state (i.e., GAP43 aggregate clustering rate).

[0103] 1. Detection of phosphorylation level of GAP43 protein at S41 site in glioma tissue:

[0104] Immunohistochemistry was used to identify the S41 phosphorylation site in glioma tissue sections using a specific monoclonal antibody (the specific method is the same as in Example 1), and the expression level was quantified using a semi-quantitative scoring method to obtain the GAP43 phosphorylation level H-score, where H-score = (Staining intensity × positive cell percentage), full score 300 points. The average score is calculated by determining the staining intensity and positive cell percentage in three different regions of the sample.

[0105] The semi-quantitative scoring method is as follows: ImageJ software is used to obtain the staining intensity score of immunohistochemical staining images of glioma tissue. First, the image type is converted to RGB Stack using the Type tool in the software. Then, the image threshold is adjusted to 80 / 180 using the Adjust-Threshold tool in the software to clearly distinguish the positive area from the background. Analyze-measure is selected, and the final value is divided into four segments: 0 (value 0-2), 1 (value 3-8), 2 (value 9-14), and 3 (value 14-30), which are the four levels of staining intensity.

[0106] Staining intensity: 0 (negative), 1 (weak), 2 (moderate), 3 (strong);

[0107] The percentage of positive cells refers to the proportion of cells stained in a field of view of an immunohistochemical staining image, with a value range of 0 to 100.

[0108] 2. GAP43 aggregate analysis:

[0109] Glioma tissue sections were stained with immunofluorescence and imaged using a laser confocal microscope. Random field-view images were acquired, and the images were processed using ImageJ software. First, the image type was converted to 8-bit grayscale using the Type tool. Then, the image threshold was adjusted to 50 / 150 using the Adjust-Threshold tool, converting the image to a binary image. In this case, the phase-separated regions appeared white, and the background was black. A specified threshold algorithm was used to identify granular regions with a signal-to-noise ratio higher than the background as GAP43 aggregates. The clustering rate of the sample was calculated by counting the proportion of cells with ≥3 aggregates among all analyzed cells, with a maximum score of 100.

[0110] Table 1

[0111]

[0112] (2) Data standardization processing

[0113] To construct a combined drug resistance score, the two indicators are first normalized (Z-score or Min-Max standardization):

[0114] The standardized value of GAP43 phosphorylation level is set as X. phos X phos = (H-score - Min Hscore ) / (Max Hscore - Min Hscore ); where Min Hscore To observe the minimum value, Min Hscore =0.4; Max Hscore Max is the observed maximum value. Hscore =294.

[0115] This method employs the Min-Max standardization approach for normalization. Based on 144 samples from Example 1, covering different pathological types including normal tissue and gliomas grade 1-4, a standardization transformation formula was constructed to ensure that new samples can be effectively mapped to the same standardized space. Through the extreme value (minimum and maximum) formula, all samples are compared on a uniform scale, thus ensuring the consistency of the results.

[0116] The theoretical range of the H-score is 0–300. This method sets Min=0.4 and Max=294, which are within the theoretical domain and close to the boundary, and are therefore reasonable. Comparing the training set parameters (Min=0.4, Max=294) with the theoretical extreme values ​​(Min=0, Max=300), the maximum deviation introduced by the standardized values ​​is only 0.0204, far below the resolution required for clinical decision-making. Therefore, it has no substantial impact on the calculation of combined drug resistance scores and subsequent predictive efficacy.

[0117] In actual testing, the H-score of GAP43 phosphorylation in real biological samples is almost never zero. Therefore, during the prediction system construction phase, negative control slides were included in each batch of staining, and the obtained Min=0.4 was used as the lower limit of system signal detection. This processing effectively subtracts system background noise, making X... phos This allows for a more accurate reflection of the relative level of specific signals. Similarly, through positive control experiments, the upper limit of detection was determined to be Max=294. This is an inherent characteristic of the detection system and was determined through optimization using negative / positive control experiments.

[0118] The standardized value of GAP43 aggregate clustering rate is set as X. LLPS X LLPS = Clustering rate value / 100.

[0119] The formula for calculating the combined score is as follows:

[0120] Resistance Score = X phos -X LLPS

[0121] Example 4: Validating the effectiveness of combined scoring in guiding combination therapy using a mouse tumor-bearing model

[0122] In this embodiment, a mouse subcutaneous tumor-bearing model of the human glioma cell line U87-MG was constructed to further verify the predictive and guiding value of the GAP43 phosphorylation + phase separation combined scoring system in Example 3 for combination drug regimens through in vivo experiments.

[0123] (1) Experimental materials and grouping

[0124] Laboratory animals: SPF-grade female C57BL / 6 nude mice, 4-6 weeks old, weighing 18-22 g, provided by a reputable animal center. The animals were randomly divided into 4 groups of 5 mice each.

[0125] Glioma cell preparation: U87-MG cell line, cell number 5 × 10⁶ 6 100 μL.

[0126] Glioma cells were implanted under the skin of nude mice, and the specific procedure was as follows:

[0127] U87-MG cells in the logarithmic growth phase were digested with 0.25% trypsin, centrifuged, resuspended in PBS, and adjusted to 5 × 10⁻⁶ cells / mL. 7 Single-cell suspensions were prepared at a concentration of cells / mL. Six- to eight-week-old C57BL / 6 nude mice were acclimatized for one week. Before injection, the skin in the upper part of the back and groin was disinfected with alcohol. The skin was then held up with forceps to form a triangular cavity between the skin and the body. Each mouse was subcutaneously injected with 5 × 10⁻⁶ cells / mL. 6 A cell suspension of 100 μL was used to slowly inject cells into a cavity using a 1 mL syringe to create skin papules. The group design was as follows:

[0128] Group A: DMSO control group

[0129] Group B: TMZ monotherapy group (25 mg / kg, ip)

[0130] Group C: PKC inhibitor (IDE, 50 mg / kg, ip)

[0131] Group D: TMZ + PKC inhibitor

[0132] Each group was injected with the corresponding medication, starting from a tumor diameter of approximately 80-100 mm. 3 Initially, injections were given every 3 days until the 12th day after tumor transplantation.

[0133] (2) Efficacy observation indicators

[0134] Tumor volume measurement: The tumor volume V is measured every two days using vernier calipers. The formula is: V = 1 / 2 × L × W 2 Where L is the long diameter of the tumor and W is the short diameter of the tumor.

[0135] Experimental endpoint: Mice were euthanized 12 days after treatment, and tumor tissue was removed, weighed, and fixed.

[0136] (3) Observation of results

[0137] The average tumor volume in group D (combined drug treatment) was significantly smaller than that in groups A and B, and the body weight of the mice remained stable. Figure 9 and Figure 10 );

[0138] The average tumor mass is as follows:

[0139] Table 2

[0140]

[0141] Confocal microscopy observation results of GAP43 aggregates in mouse subcutaneous tumor sections are as follows: Figure 11 As shown in the figure, the statistical analysis of the number of GAP43 aggregates in mouse subcutaneous tumor sections is as follows: Figure 12 As shown. Figure 11 Green aggregates, which are GAP43 aggregates, were observed in the cells of the Con (Group A), IDE (Group C) and combined drug group (Group D), while no obvious green spots appeared in the TMZ group (Group B). Figure 12 The statistics focused on the formation of aggregates from the GAP43 phase separation. The results show that TMZ induces GAP43 dispersion, significantly reducing the number of droplets compared to the control group. However, after using a PKC inhibitor, the number of droplets recovered to or even exceeded the control group level. Furthermore, the combined use of TMZ and the PKC inhibitor reversed the GAP43 droplet dispersion caused by TMZ. This indicates that the PKC inhibitor can maintain the aggregated state of GAP43 and enhance the cell's sensitivity to TMZ.

[0142] Treatment with different drugs induces different states in GAP43: TMZ reduces GAP43 aggregates, leading to TMZ tolerance; conversely, PKC inhibitors increase GAP43 aggregates, resulting in TMZ sensitivity. Increased GAP43 aggregates make tumor cells more sensitive. Sensitivity is represented here by the increase in subcutaneous tumor volume and final weight in mice. Tumor reduction demonstrates drug effectiveness.

[0143] Example 5: A Glioma Drug Regimen Prediction System Based on GAP43 Phase Separation and Phosphorylation Detection

[0144] This embodiment provides a method for applying the combined scoring results of GAP43 phosphorylation level and phase separation behavior to the decision-making of combined chemotherapy for glioma, and the method is validated by in vitro cell experiments and retrospective patient data.

[0145] Combination therapy decision-making process:

[0146] Obtain the GAP43 phosphorylation normalized value X from the patient's glioma tissue sample. phos and the normalized value of clustering rate of GAP43 aggregates X LLPS Calculate the joint score: Resistance Score = X phos - X LLPS .

[0147] Based on the scoring results, the tiered decision-making is as follows:

[0148] Table 3

[0149]

[0150] To further verify the clinical application value of the combined scoring system in guiding individualized combination therapy for glioma, this embodiment uses a glioma organoid model for functional verification.

[0151] The organoids were purchased from Aiming Medical Company, and were cultured using a specially purchased culture medium. The organoids were spherical tumor spheres. Organoids of similar size were divided into four groups: A, B, C, and D, with five organoids in each group.

[0152] (1) Group design:

[0153] Group A: DMSO control group

[0154] Group B: TMZ monotherapy (200 μM)

[0155] Group C: PKC inhibitor (IDE, 20 μM)

[0156] Group D: TMZ + PKC inhibitor

[0157] The treatment method involves preparing four specific concentrations of drugs using specialized culture media, immersing the organoids in the corresponding drug-containing culture media, and treating for 72 hours.

[0158] (2) Efficacy observation indicators

[0159] ①IHC was used to detect the expression of GAP43 phosphorylation in each group, and a semi-quantitative scoring method (H-score) was used:

[0160] The staining intensity was measured using ImageJ and then divided according to the IHC score. All grouped images were processed in the same way and set with the same threshold (80 / 180). The final value was the staining intensity, which was then divided into 4 segments: 0 (IHC score 0-2), 1 (IHC score 3-8), 2 (IHC score 9-14), and 3 (IHC score 14-30).

[0161] Staining intensity: 0 (negative), 1 (weak), 2 (moderate), 3 (strong);

[0162] Staining positivity rate: 0~100%;

[0163] Calculation formula: H-score = (Staining intensity × percentage of positive cells); Full marks: 300 points.

[0164] ② Confocal imaging and GAP43 aggregate analysis:

[0165] Glioma tissue sections were stained with immunofluorescence and imaged using a laser confocal microscope, acquiring images of at least five random tumor fields. The following parameters were recorded for each field:

[0166] Average number of aggregates ( / cell), average area of ​​aggregates (μm²), and clustering rate: the percentage of cells with ≥3 aggregates (%).

[0167] Image processing was performed using ImageJ software. First, the image type was converted to 8-bit grayscale using the Type tool. Then, the image threshold was adjusted to 50 / 150 using the Adjust-Threshold tool, converting the image to a binary image. In this case, the phase-separated regions appeared white, and the background was black. A specified threshold algorithm was used to identify granular regions with a signal-to-noise ratio higher than the background as GAP43 aggregates. The proportion of cells with ≥3 aggregates among all analyzed cells was calculated and output as a percentage, representing the clustering rate. The clustering rate was scored out of 100.

[0168] (3) Evaluation of drug efficacy response

[0169] Immunohistochemical (IHC) staining results of Ki-67, GAP43, and p-GAP43 in glioma organoids are as follows: Figure 13 As shown in the figure, the statistical analysis of the Ki-67 positivity rate in glioma organoids is as follows: Figure 14 As shown. Figure 15 The phosphorylation levels of GAP43 in organoids of gliomas in groups A, B, C, and D are represented by their scores.

[0170] The H-score was calculated by averaging the staining intensity and positive rate in three different regions of the sample. (Intensity × Proportion of Positive Cells) = (3 × 90 + 3 × 90.5 + 3 × 90) / 3 = 270.5, which means that the H-score for GAP43 phosphorylation expression in group A organoids is 270.

[0171] Calculate X based on H-score phos X phos =270-0.4 / 294-0.4=0.92.

[0172] GAP43 aggregate clustering rate: Of the 42 cells in the field of view, 21 had aggregates with ≥3 cells, resulting in a clustering rate of 21 / 42 × 100 = 50%. LLPS =50 / 100=0.5.

[0173] Joint score Resistance Score = X phos - X LLPS=0.4 points. The combined score range for this type of organoid is 0.3-0.6. The recommended combined trial regimen is: TMZ + low concentration PKC inhibitor (20μM).

[0174] Different drug treatments induce different states in GAP43: TMZ enhances p-GAP43 and reduces GAP43 aggregates; conversely, PKC inhibitors weaken p-GAP43 and enhance GAP43 aggregates. Increased GAP43 aggregates make organoids more sensitive. Ki-67, a marker of cell proliferation, is used to represent sensitivity; a decrease in Ki-67 positivity indicates slower organoid growth and proliferation, and higher organoid sensitivity. Combined drug treatment significantly reduces Ki-67, demonstrating the effectiveness of GAP43-targeted drug treatment.

[0175] The results showed that combined treatment with TMZ and IDE significantly reduced the Ki-67 positivity rate (Group D: 6.6% ± 1.9%, Group A: 12.7% ± 1.4%, p = 0.01) and restored the number of agglomerates (Group B: 4.6 ± 1.3 dots / cell, Group D: 10.6 ± 0.9 dots / cell, P < 0.0001). These results demonstrate that the combination therapy decision model based on this scoring system can effectively predict drug response, and the experimental verification of the anti-proliferative effect of the combination therapy regimen confirms the feasibility and effectiveness of this scoring system in guiding preclinical drug selection.

[0176] This result is consistent with the in vitro experimental results, suggesting that tumor tissues with high GAP43 scores are more sensitive to combination therapy.

[0177] Figure 16 Confocal microscopy observation of GAP43 condensates in glioma organoids. Figure 17 Statistical analysis of the number of aggregates in glioma organoids.

[0178] Figure 16 Green dot-like aggregates were observed in the cells of the Con (Group A), IDE (Group C), and combined drug group (Group D), which are phase-separated condensates formed by GAP43. Figure 17 The statistics focused on the formation of GAP43 aggregates through phase separation. It can be concluded that GAP43 aggregates were present in group A, i.e., the organoids themselves. TMZ induced GAP43 diffusion, resulting in a significant reduction in droplet numbers compared to the control group. After using a PKC inhibitor, the droplet number recovered to or even exceeded the control group level. The combined use of the PKC inhibitor also reversed the TMZ-induced GAP43 droplet dissipation. This indicates that PKC inhibitors can enhance the aggregate state of GAP43 and reverse cellular resistance to TMZ.

Claims

1. A glioma drug regimen prediction system based on GAP43 phase separation and phosphorylation detection, characterized in that, The system includes: The GAP43 phosphorylation detection module is used to detect the phosphorylation level of GAP43 protein at the S41 site in glioma tissue. The GAP43 phase separation imaging module uses laser confocal microscopy to quantitatively assess the clustering rate of GAP43 aggregates. The joint score calculation module is used to combine the phosphorylation level of GAP43 and the clustering rate of GAP43 aggregates to calculate the joint score; Treatment plan prediction module: Predicts glioma treatment plans based on scoring results.

2. The glioma drug regimen prediction system based on GAP43 phase separation and phosphorylation detection according to claim 1, characterized in that, The specific method for detecting the phosphorylation level of GAP43 protein at the S41 site is as follows: Immunohistochemistry is used to identify the S41 phosphorylation site with a specific monoclonal antibody, and the expression level is quantified using a semi-quantitative scoring method.

3. The glioma drug regimen prediction system based on GAP43 phase separation and phosphorylation detection according to claim 2, characterized in that, The use of a semi-quantitative scoring method to quantify the expression level specifically refers to: ImageJ software was used to obtain the staining intensity scores of immunohistochemical staining images of glioma tissue. First, the image type was converted to RGB Stack. Then, the image threshold was adjusted to 80 / 180 using the Adjust-Threshold tool to clearly distinguish the positive area from the background. Analyze-measure was selected to divide the final values ​​into four segments, which are the four levels of staining intensity: 0: value 0-2, 1: value 3-8, 2: value 9-14, 3: value 14-30. Staining intensity is categorized as follows: 0: negative, 1: weak, 2: moderate, 3: strong; The H-score of GAP43 phosphorylation level was obtained, and H-score = (Staining intensity × Proportion of positive cells); The positive cell ratio refers to the proportion of cells stained in a field of view of an immunohistochemical staining image, with a value range of 0 to 100.

4. The glioma drug regimen prediction system based on GAP43 phase separation and phosphorylation detection according to claim 1, characterized in that, The method for quantitatively evaluating the clustering rate of GAP43 aggregates using laser confocal microscopy is as follows: Glioma tissue sections were stained with immunofluorescence and imaged using a laser confocal microscope. Random field-view images were acquired, and the images were processed using ImageJ software. First, the image type was converted to 8-bit grayscale using the Type tool. Then, the image threshold was adjusted to 50 / 150 using the Adjust-Threshold tool. After application, the image was converted to a binary image. A specified threshold algorithm was used to identify granular regions with a signal-to-noise ratio higher than the background as GAP43 aggregates. The proportion of cells with ≥3 aggregates in all analyzed cells was counted as the clustering rate of the sample.

5. The glioma drug regimen prediction system based on GAP43 phase separation and phosphorylation detection according to claim 1, characterized in that, The formula for calculating the joint score, Resistance Score, is as follows: Resistance Score = X phos -X LLPS X phos = (H-score - Min Hscore ) / (Max Hscore - Min Hscore ); X LLPS = Clustering rate / 100 Where X phos Min is the standardized value of GAP43 phosphorylation level. Hscore =0.4, Max Hscore =294; X LLPS This is the standardized value for the clustering rate of GAP43 aggregates.

6. The glioma drug regimen prediction system based on GAP43 phase separation and phosphorylation detection according to claim 1, characterized in that, The specific treatment regimens for gliomas based on the scoring results are as follows: for a combined score <0.3, temozolomide monotherapy is used; for a combined score of 0.3~0.6, temozolomide plus a low-concentration PKC inhibitor is used; for a combined score >0.6, temozolomide plus a standard-concentration PKC inhibitor is used; the low concentration is 10g / kg and the standard concentration is 20g / kg.

Images