Antibody kits, reagent kits, and their use in predicting the risk of tumor recurrence

Abstract

This application provides an antibody kit, a reagent kit, and their use in predicting tumor recurrence risk, relating to the field of antibody kit technology. The antibody kit includes antibody sequences as shown in SEQ ID NO:1-5. The antibodies in the aforementioned antibody kit are low-cost to prepare, have stable quality, can be mass-produced, and are highly accurate for predicting tumor recurrence risk.

Description

Technical Field

[0001] This application relates to the field of peptide technology, specifically to antibody groups, antibody sequence determination methods, the use of antibody groups in the preparation of tumor recurrence prediction detection kits, kits, and methods for predicting the risk of tumor recurrence. Background Technology

[0002] Postoperative recurrence of malignant tumors is a key factor affecting long-term survival. Taking hepatocellular carcinoma as an example, even after radical surgical resection, a high proportion of patients still experience tumor recurrence within 5 years post-surgery. Accurately predicting the risk of postoperative recurrence is of significant clinical importance for guiding the selection of postoperative adjuvant therapy, optimizing patient follow-up strategies, and improving overall prognosis.

[0003] In related technologies, the prediction of postoperative tumor recurrence risk typically employs immunohistochemistry (IHC) to detect specific markers in the tumor immune microenvironment, and then integrates and analyzes multi-marker information using machine learning algorithms or scoring systems. The technical principle is as follows: tumor tissue sections are stained with antibodies targeting specific markers, and the expression levels of these markers (such as the percentage of positive cells) are quantitatively or semi-quantitatively assessed. A comprehensive score is then calculated using statistical or machine learning models, thereby achieving a stratified prediction of recurrence risk.

[0004] However, the antibodies it relies on are mainly from commercial sources, resulting in high testing costs and difficulty in ensuring batch-to-batch consistency, which limits the large-scale promotion and standardized application of this type of predictive technology in clinical practice. Summary of the Invention

[0005] The embodiments of this application aim to at least partially solve one of the above-mentioned technical problems or at least provide a useful commercial option. In view of this, the embodiments of this application provide antibody kits, reagent kits, and their use in predicting the risk of tumor recurrence. The aforementioned antibody kits are low-cost, of stable quality, and can be mass-produced, and are highly accurate for predicting the risk of tumor recurrence.

[0006] Specifically, this application provides the following technical solution: In a first aspect, embodiments of this application provide an antibody set, the antibody set comprising: antibody sequences as shown in SEQ ID NO: 1-5.

[0007] The antibody kit of this embodiment includes five nanobodies. The nanobody shown in SEQ ID NO: 1 specifically binds to the SPON2 protein; the nanobody shown in SEQ ID NO: 2 specifically binds to the HLA-DRB1 protein; the nanobody shown in SEQ ID NO: 3 specifically binds to the ZFP36 protein; the nanobody shown in SEQ ID NO: 4 specifically binds to the ZFP36L2 protein; and the nanobody shown in SEQ ID NO: 5 specifically binds to the VIM protein. These nanobodies were independently developed and obtained by the inventors. They can specifically bind to target proteins and have low preparation costs, stable quality, and good batch-to-batch consistency. The aforementioned antibody kit is highly accurate for predicting tumor recurrence risk.

[0008] It is understood that the nanobodies in the embodiments of this application can be modified, such as by adding an Fc segment to facilitate purification and increase stability; by biotin modification to facilitate signal amplification during detection; or by fluorescent labeling to facilitate multiplex fluorescence detection, etc.

[0009] Secondly, this application proposes an antibody sequence determination method, which includes: predicting antibody sequences based on the three-dimensional structure and epitope information of the target antigen to obtain N candidate antibody sequences, where N is an integer not less than 1; screening the N candidate antibody sequences based on preset parameters to obtain M preliminary screening antibody sequences, where the preset parameters include at least one of structure confidence level pLDDT, interface prediction error PAE, and docking score, where M is an integer not less than 1 and not greater than M; and determining the target antibody sequence based on the affinity results of the M preliminary screening antibody sequences.

[0010] The antibody sequence determination method proposed in this embodiment directly generates a large number of candidate nanobody sequences through computational prediction based on the three-dimensional structure and epitope information of the target antigen. This overcomes the limitations of traditional antibody preparation, which relies on animal immunization or phage display libraries, and realizes de novo rational design of nanobodies. This computational prediction step can generate N structurally diverse candidate sequences at once, providing a rich and targeted pool of candidates for subsequent screening, significantly improving the efficiency and structural rationality of antibody discovery from the source.

[0011] In the candidate sequence screening stage, this method employs a multi-parameter combined filtering approach, including structural confidence level (pLDDT), interface prediction error (PAE), and docking score, retaining only M initial screening sequences that simultaneously meet the criteria of high structural reliability, high interface geometric confidence, and low energy docking score. This multi-dimensional computational screening strategy significantly reduces the number of candidates entering the experimental validation stage (typically compressing the initial candidates to less than 1 / 50 of the original number), avoiding the blind synthesis and testing of a large number of invalid sequences. This improves the accuracy and success rate of screening, ensuring that the final antibody sequences have higher reliability in terms of structural stability and binding conformation.

[0012] This method, after computational screening, only performs gene synthesis and affinity assays on the M initial screening sequences, and then determines the target antibody sequence based on affinity data (such as the KD value determined by SPR). This two-stage model of "computational large-scale prediction + small-scale experimental fine screening" greatly reduces the cost and time consumption of traditional antibody development, which often involves experimental verification at the kilogram level. At the same time, by using affinity as the gold standard indicator for final verification, it ensures that the obtained target antibody sequence has excellent actual binding performance (KD can usually reach <50 nM), realizing an efficient closed loop of computational-aided design and experimental verification.

[0013] In summary, the antibody sequence determination method of this embodiment provides an efficient and reproducible preparation path for determining any nanobody sequence in the first aspect.

[0014] In exemplary embodiments, the three-dimensional structure and epitope information of the target antigen can be obtained in a variety of ways.

[0015] Specifically, the three-dimensional structure of the target antigen can be derived from a publicly available structure database, such as a protein database (PDB). Alternatively, the three-dimensional structure of the target antigen can be obtained through computational structure prediction methods, including but not limited to the AlphaFold series models, RoseTTAFold, or other deep learning-based protein structure prediction algorithms. Furthermore, the three-dimensional structure of the target antigen can also be obtained through experimental methods, including but not limited to X-ray crystallography, nuclear magnetic resonance spectroscopy, or cryo-electron microscopy.

[0016] Based on obtaining the three-dimensional structure of the target antigen, the surface epitopes of the target antigen are further analyzed and selected. Methods for obtaining epitope information include, but are not limited to, one or more of the following: structure-based epitope prediction algorithms, solvent-accessible surface area analysis, sequence conservation assessment, molecular dynamics simulations, and analysis of known antigen-antibody complex structures.

[0017] In a preferred embodiment of this application, the AlphaFold structure prediction tool is used to perform three-dimensional structural modeling of the amino acid sequence of the target antigen (including but not limited to SPON2, HLA-DRB1, ZFP36L2, ZFP36 or VIM), and combined with surface accessibility analysis and conservation assessment, to determine the preferred epitope regions suitable for nanobody binding, providing accurate structural coordinates for the subsequent computational design of nanobody sequences.

[0018] Any combination of one or more of the above methods can be used to obtain the three-dimensional structure and epitope information of the target antigen in this application. As long as reasonable and accurate three-dimensional conformational information of the target antigen and epitope information that can be used for nanobody design can be provided, it falls within the protection scope of this application.

[0019] In an exemplary embodiment, antibody sequence prediction based on the three-dimensional structure and epitope information of the target antigen can be achieved by designing a nanobody framework using a generative diffusion model.

[0020] Specifically, the generative diffusion model may include, but is not limited to, RFdiffusion, RoseTTAFoldDiffusion, ProteinMPNN, or other diffusion model-based protein / antibody design frameworks. These models take the three-dimensional structural coordinates of the target antigen and the selected antigenic epitope region as input conditions, and generate nanobody backbone sequences with reasonable folding conformation and complementarity-determining regions (CDRs) through a stepwise denoising process.

[0021] In a preferred embodiment of this application, the RFdiffusion generative diffusion model is used, taking the three-dimensional structure of the target antigen (e.g., SPON2, HLA-DRB1, ZFP36L2, ZFP36, or VIM) and predetermined epitope coordinates as input conditions, to generate a large number of structurally diverse candidate nanobody sequences. This process can output hundreds to thousands of candidate sequences at once, each candidate sequence possessing a CDR region that is potentially spatially complementary to the target epitope.

[0022] Furthermore, the generative diffusion model can incorporate additional optimization strategies during its design process, including but not limited to: removing unstable residues, improving sequence solubility, optimizing CDR region flexibility, enhancing thermal stability, or improving expression efficiency. These optimizations can further improve the success rate of the generated candidate nanobody sequences in subsequent experimental expression and functional validation.

[0023] The design method of the above-mentioned generative diffusion model is not limited to a specific model or name. Any computational method that is based on diffusion mechanism or generative artificial intelligence technology and can generate nanobody candidate sequences based on antigen structure and epitope information falls within the protection scope of this application.

[0024] In an exemplary embodiment, the N candidate nanobody sequences are screened based on preset parameters to obtain M initial screening antibody sequences. Specifically, the screening process employs a multi-parameter joint evaluation method, where the preset parameters include at least one or more of structural confidence level (pLDDT), interface prediction error (PAE), and docking score. These parameters comprehensively evaluate the candidate nanobody sequences from three dimensions: monomer structural reliability, complex interface geometric accuracy, and binding energy rationality.

[0025] In a preferred embodiment of this application, the screening step includes the following sub-steps: S1. Use AlphaFold2 or similar structure prediction tools to predict the monomer structure of each candidate nanobody sequence and retain candidate sequences with pLDDT greater than 80. S2, using AlphaFold-Multimer or similar complex prediction tools to predict the complex structure of the retained sequence and the target antigen, retaining candidate sequences with a PAE of less than 10 Å at the interface prediction. S3. Using HADDOCK or other molecular docking software, the above candidate sequences are flexibly docked and optimized, and candidate sequences with docking scores less than -100 are retained, finally obtaining M initial screening antibody sequences.

[0026] Furthermore, the numerical range of M can be adjusted from tens to hundreds according to actual needs. Through the above multi-parameter, hierarchical screening strategy, the initial N candidate sequences can be significantly compressed, while ensuring that the retained preliminary screening sequences have high reliability in terms of structural stability, interface binding conformation, and energy characteristics, providing high-quality candidates for subsequent gene synthesis and experimental verification.

[0027] The preset parameters and their specific thresholds are not limited to the values ​​listed in this embodiment. Any calculated or experimental parameters that can reflect the reliability of the nanobody structure, the accuracy of interfacial binding, and the binding energy can be used in the screening process of this application and fall within the protection scope of this application.

[0028] In an exemplary embodiment, the M preliminary screening antibody sequences obtained are used for gene synthesis and affinity assays are performed. Finally, the target antibody sequence is determined based on the affinity assay results.

[0029] Specifically, the M initial screening antibody sequences are first obtained by gene synthesis technology to obtain coding sequences, and then cloned into a suitable expression vector. Preferably, the nanobody sequences can be fused with Fc fragments, His tags, or other fusion tags to facilitate subsequent expression, purification, and detection.

[0030] In a preferred embodiment of this application, the synthesized nanobody is transiently or stably expressed in a mammalian cell expression system, including but not limited to HEK293T cells, HEK293F cells, or CHO cells. The expressed nanobody is purified by affinity chromatography (such as Protein A or Ni-NTA affinity purification), preferably achieving a purity of 95% or higher.

[0031] The purified nanobodies were then subjected to affinity assays. The affinity assays preferably employed surface plasmon resonance (SPR) technology, using a BIAcore system or other equivalent instruments to determine the equilibrium dissociation constant (KD value) between the nanobodies and the target antigen. Alternatively, biomembrane interference (BLI), isothermal titration calorimetry (ITC), or enzyme-linked immunosorbent assay (ELISA) can be used to further verify affinity.

[0032] In this application, based on affinity assay results, a preliminary screening nanobody with a KD value of no more than 50 nM is preferably selected as the target antibody sequence; more preferably, a sequence with a KD value of less than 30 nM and simultaneously possessing good specificity and thermostability is selected as the final target antibody sequence. This experimental verification step ensures that the finally determined target antibody sequence is not only structurally reasonable at the computational prediction level, but also meets the high affinity standards required for clinical application in actual binding performance.

[0033] The gene synthesis, expression purification, affinity determination methods and their specific conditions are not limited to those listed in this embodiment. Any technical means that can convert the initial screening sequence obtained by calculation and screening into an active nanobody with measurable affinity and thereby determine the target antibody sequence falls within the protection scope of this application.

[0034] Thirdly, embodiments of this application provide the use of antibody groups in the preparation of tumor recurrence prediction detection kits, wherein the antibody groups are as described in the first aspect.

[0035] The kit in this embodiment detects the expression level of NK cell-related markers in tumor tissue and, combined with the tumor recurrence risk prediction method described below, achieves accurate stratified prediction of tumor recurrence risk.

[0036] In a preferred embodiment of this application, the tumor is hepatocellular carcinoma (HCC). The nanobodies (including anti-SPON2 nanobodies, anti-HLA-DRB1 nanobodies, anti-ZFP36L2 nanobodies, anti-ZFP36 nanobodies, and anti-VIM nanobodies) can be used to prepare a predictive diagnostic kit for the risk of postoperative recurrence in liver cancer patients. This kit can significantly improve the predictive accuracy of traditional TNM or BCLC staging systems at the immune microenvironment level, providing important guidance for clinical adjuvant therapy decisions and personalized follow-up strategies.

[0037] Furthermore, the tumor may include, but is not limited to, one or more solid tumors selected from gastric cancer, lung cancer, colorectal cancer, breast cancer, pancreatic cancer, or prostate cancer. The antibody group described in this application is also applicable to the preparation of a diagnostic kit for predicting the risk of postoperative recurrence of the aforementioned solid tumors. By detecting corresponding markers in the tumor immune microenvironment and combining this with the tumor recurrence risk prediction described below, cross-tumor type prediction of recurrence risk for various solid tumors can be achieved, demonstrating good universality and clinical application potential.

[0038] The tumor types mentioned above are not limited to the specific cancer types explicitly listed in this embodiment. Any tumor that has NK cell-related immune microenvironment characteristics and whose recurrence risk can be assessed by detecting the markers targeted by the nanobody can be used to prepare corresponding predictive detection kits and fall within the protection scope of this application.

[0039] In exemplary embodiments, the cancer samples of this application include, but are not limited to, paraffin-embedded (FFPE) tissue sections, fresh frozen tissue sections, etc.

[0040] Fourthly, embodiments of this application provide a kit comprising: an antibody group according to the first aspect.

[0041] Specifically, the kit uses the antibody group described in the first aspect (at least one of anti-SPON2 nanobody, anti-HLA-DRB1 nanobody, anti-ZFP36L2 nanobody, anti-ZFP36 nanobody and anti-VIM nanobody) as the core detection component for immunohistochemical or immunofluorescence detection of tumor recurrence risk after surgery.

[0042] In a preferred embodiment of this application, the working concentration of the antibody in the kit is 1 μg / mL to 5 μg / mL, more preferably 1 to 3 μg / mL, or diluted at a ratio of 1:100 to 1:200. This concentration range ensures good specific binding of the nanobody to tissue sections while avoiding excessive background staining.

[0043] Furthermore, the kit also includes at least one or more of the following: tissue pretreatment reagents, antigen retrieval reagents, signal amplification reagents, staining reagents, mounting reagents, and quality control materials. These supporting reagents, together with the nanobody, constitute a complete detection system, enabling standardized operation throughout the entire process from sample pretreatment to result interpretation.

[0044] In an exemplary embodiment, the tissue pretreatment reagent includes xylene or dewaxing solution, a 75% (v / v) to 100% (v / v) gradient of ethanol, and PBS buffer (pH 7.0-7.5, containing 0.03%-0.06% Tween-20). The antigen retrieval reagent includes at least one of citrate buffer (pH 5.8-6.2), EDTA buffer (pH 7.8-8.2), and Tris-EDTA buffer (pH 8.8-9.2).

[0045] In an exemplary embodiment, the mounting reagent includes at least one of peroxidase blocking solution (2%-5% H2O2), biotin blocking solution (a combination of ovalbumin and D-biotin), and serum blocking solution (5% BSA or goat serum). The signal amplification reagent includes at least one of HRP-labeled anti-Fc secondary antibody, HRP-streptavidin, DAB chromogenic solution, fluorescently labeled secondary antibody (e.g., Alexa Fluor series), and TSA signal amplifier.

[0046] In an exemplary embodiment, the staining reagent includes hematoxylin solution (for routine IHC nuclear counterstaining) or DAPI solution (for immunofluorescence nuclear counterstaining). The mounting reagent includes neutral resin (for IHC) or antifluorescence quenching mounting medium (for immunofluorescence). The quality control materials include positive control tissue sections with a known high risk of recurrence, negative control tissue sections with a known low risk of recurrence, and blank control sections.

[0047] In one specific embodiment of this application, the reagent kit can be assembled into two implementation schemes: Scheme A (single-color IHC scheme): Five consecutive paraffin sections were stained with five different nanobodies, and an HRP-DAB signal amplification system was used. Option B (Multiple Immunofluorescence Protocol): This protocol uses a single slide and sequential staining combined with TSA signal amplification to achieve multicolor detection of five nanobodies in the same field of view. Both protocols can be used with the TIMES scoring software to achieve automated stratified prediction of relapse risk.

[0048] The specific components, concentration ranges, pH values, and implementation schemes of the above-mentioned components are not limited to the specific values ​​or combinations listed in this embodiment. Any reagent combination that can be used in conjunction with the antibody group described in the first aspect to complete the detection of target markers in tumor tissue falls within the protection scope of this application.

[0049] In an exemplary embodiment, to ensure the accuracy and reliability of the detection results, the following samples are avoided: (1) Samples with poor tissue quality, including but not limited to samples that are over-fixed, have undergone autolysis, or have been severely degraded; (2) Samples with insufficient tumor tissue proportion, i.e., samples in which tumor cells account for less than 30% of the entire tissue section; (3) Samples whose slice thickness does not meet the requirements, i.e., samples whose slice thickness is not within the range of 4-5 μm.

[0050] Fifthly, embodiments of this application provide a method for predicting the risk of tumor recurrence. The method includes: staining K genes in a target sample based on an antibody group from the first aspect to obtain a stained image, where K is an integer not less than 1; selecting a predetermined number of regions at the tumor invasion front and tumor center in the stained image to obtain multiple target regions, where the tumor invasion front is a region less than 2 mm from the tumor boundary, and the tumor center is a region not less than 2 mm from the tumor boundary; calculating the positive cell ratio of each gene in each target region to obtain K result sets; inputting the K result sets into a trained marker scoring model to obtain K marker scores; calculating the weight of each gene in the multiple target regions based on the K result sets to obtain K weights; and predicting the risk of tumor recurrence based on the K marker scores and the K weights to obtain a prediction result.

[0051] The tumor recurrence risk prediction method provided in this embodiment, by using the antibody group described in the first aspect to stain K genes in the target sample, can achieve high specificity and high sensitivity detection of NK cell-related markers in the tumor immune microenvironment, providing a reliable staining image basis for subsequent quantitative analysis. This method overcomes the limitations of traditional commercial antibodies in batch-to-batch consistency and specificity, ensuring the stability and reproducibility of the detection results.

[0052] In the staining image analysis stage, this method clearly divides tumor tissue into the tumor invasion front (the area less than 2 mm from the tumor boundary) and the tumor center (the area not less than 2 mm from the tumor boundary), and selects a predetermined number of target regions in each region for analysis. This region division strategy based on the spatial heterogeneity of the tumor microenvironment can comprehensively capture the differences in immune characteristics between the tumor periphery invasion zone and the central zone. Compared with existing methods that only analyze a single region, it can better reflect the true biological behavior of the tumor, thereby significantly improving the accuracy and clinical relevance of recurrence risk prediction.

[0053] This method calculates the percentage of positive cells for each gene in multiple target regions, obtaining K result sets. These result sets are then input into a trained marker scoring model to obtain K marker scores. This step automates the conversion from raw staining images to single-marker predicted scores, reducing subjective interpretation errors. Simultaneously, based on the K result sets, the standard deviation of each gene in multiple target regions is calculated, resulting in K inverse variance weights. This assigns higher weights to markers with stable expression, while automatically reducing the weights of markers with larger fluctuations. This adaptive weighting mechanism effectively suppresses noise interference and improves the robustness of the overall score.

[0054] Finally, this method uses a weighted fusion of K marker scores and their corresponding K weights to obtain the final prediction result for tumor recurrence risk. This fusion strategy fully utilizes the complementary information of multiple markers, avoiding the limitations of single-marker prediction, and achieves a dynamic balance of reliability for different markers through inverse variance weighting. Compared with traditional simple averaging or empirical weighting methods, the method of this invention significantly improves prediction accuracy, sensitivity, and specificity, providing clinicians with more accurate high / low recurrence risk stratification results.

[0055] In summary, the tumor recurrence risk prediction method in this embodiment organically combines nanobody detection technology with spatial region-specific analysis and adaptive weighted algorithms, forming a complete solution from sample staining to risk prediction. This method is not only simple to operate, cost-effective, and provides stable results, but also exhibits excellent predictive performance, possessing significant clinical decision-making guidance value and broad prospects for widespread application.

[0056] In an exemplary embodiment, the staining process employs immunohistochemistry (IHC) or immunofluorescence (IF) methods, using the antibody group described in the first aspect to specifically label K target genes (including but not limited to SPON2, HLA-DRB1, ZFP36L2, ZFP36, and VIM) in tumor tissue sections.

[0057] In a preferred embodiment of this application, K=5, and five different nanobodies are used to stain paraffin-embedded (FFPE) tissue sections from the same patient after surgery, which can form a single-color IHC scheme (serial sections) or a multiplex immunofluorescence scheme (single section), thereby obtaining high-resolution staining images and providing clear and highly specific raw data for subsequent quantitative analysis.

[0058] The above staining methods are not limited to specific K values ​​or staining types. Any technical means that uses the antibody group described in the first aspect to label genes in the target sample and generate staining images that can be used for subsequent image analysis falls within the protection scope of this application.

[0059] In an exemplary embodiment, the region selection is based on the spatial heterogeneity of tumor tissue, and the two functional regions of the tumor infiltration front (IF) and the tumor core (TC) are precisely divided on the same or consecutive stained images.

[0060] In a preferred embodiment of this application, five fields of view (the predetermined number can be 3-10) are randomly selected in the tumor invasion front and the tumor center region, respectively, with each field of view having an area of ​​approximately 0.5 mm. 2 By avoiding necrotic, hemorrhage, or tearing areas, 10 or more representative target areas can be obtained, ensuring that the analysis results can comprehensively reflect the spatial differences in the immune characteristics of the tumor microenvironment.

[0061] The above-mentioned regional division criteria and selection numbers are not limited to the specific values ​​or definitions listed in this embodiment. Any technical means that can divide the tumor invasion front and the tumor center region in the stained image and select the target region for subsequent quantitative analysis falls within the protection scope of this application.

[0062] In an exemplary embodiment, the positive cell ratio is calculated by counting the number of positive cells in each target area and the total number of cells, and then calculating the positive cell ratio (%).

[0063] In a preferred embodiment of this application, K genes are statistically analyzed across all target regions (IF region + TC region) to obtain K result sets corresponding to each gene. Each result set contains the positive rate values ​​for multiple target regions. This calculation can be completed manually or automatically with AI assistance, ensuring that the data is objective and reproducible.

[0064] The above-mentioned method for calculating the positive cell ratio is not limited to a specific counting method. Any technical means that quantifies the positive expression level of each gene in the target region based on the staining image and forms a result set falls within the protection scope of this application.

[0065] In an exemplary embodiment, the marker scoring model is a machine learning model trained on a clinical cohort, capable of converting the result set for each gene into a single-marker predicted value (p) reflecting the marker's contribution to relapse risk. k ).

[0066] In a preferred embodiment of this application, the model employs XGBoost or other equivalent supervised learning algorithms, independently inputting the positive rate result set of each gene across all target regions, and outputting a single-marker score p between 0 and 1. i This step enables automated mapping from raw quantitative data to single-marker risk prediction.

[0067] The aforementioned marker scoring model and its input / output methods are not limited to specific algorithm types. Any technical means that can transform K result sets into K marker scores falls within the protection scope of this application.

[0068] In an exemplary embodiment, the step of calculating the weight of each gene in the multiple target regions based on the K result sets to obtain K weights includes: determining the mean positive cell ratio of the k-th gene for the k-th gene and its corresponding result set; calculating the standard deviation of the k-th gene in the multiple target regions based on the result set corresponding to the k-th gene and the mean positive cell ratio of the k-th gene; determining the variance of the k-th gene in the multiple target regions based on the standard deviation; using the reciprocal of the variance as the weight of the k-th gene; and traversing the K genes and K result sets to obtain the K weights.

[0069] Specifically, the weighting calculation employs an inverse variance weighting method, which calculates the standard deviation σ of the positive rate of each gene in multiple target regions. i Calculate the weight w i =1 / σ i 2 This weighting system automatically assigns higher weights to genes with stable expression, effectively reducing the impact of genes with large fluctuations.

[0070] In an exemplary embodiment, the step of predicting the risk of tumor recurrence based on the K marker scores and the K weights to obtain the prediction result includes: taking a weighted geometric average of the K marker scores and the K weights to obtain the prediction result.

[0071] Specifically, the prediction result can be calculated as: TIMES = exp(Σ[w k × ln(p k )] / Σw k ), where TIMES represents the prediction result, w k p represents the weight of the k-th gene. kThis represents the score of the k-th marker.

[0072] In an exemplary embodiment, a threshold can be set to determine the risk of tumor recurrence. For example, a preset threshold of 0.5 is used. When TIMES ≥ 0.5, it is determined to be a high risk of recurrence; otherwise, it is a low risk of recurrence. This step achieves complementary integration and adaptive fusion of multi-marker information.

[0073] The above prediction methods are not limited to specific fusion formulas or stratification thresholds. Any technical means that obtains the prediction results of tumor recurrence risk based on K marker scores and K weights falls within the protection scope of this application.

[0074] This application marks the first time nanobody technology has been applied to the TIMES tumor recurrence prediction system. Previous studies utilized commercially available antibodies, while this application employs de novo computational protein design to develop specialized nanobodies targeting five specific markers, achieving complete autonomy for the TIMES system. All nanobody sequences are original, exhibiting less than 70% homology with existing nanobody libraries. Furthermore, this application utilizes an inverse variance-weighted geometric mean algorithm to integrate multi-marker information, automatically reducing the weight of unstable markers and further improving prediction accuracy and robustness compared to simple averaging methods. Attached Figure Description

[0075] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0076] Figure 1 This is a schematic block diagram of an electronic device according to an embodiment of this application; Figure 2 This is a schematic diagram showing the IHC staining comparison results of nanobodies and commercial antibodies according to embodiments of this application; Figure 3 The following are IHC staining results of five nanobodies according to embodiments of this application in tumor tissue: (A) Nb-SPON2 staining; (B) Nb-HLA staining; (C) Nb-ZFP36L2 staining; (D) Nb-ZFP36 staining; (E) Nb-VIM staining; Figure 4 This is a schematic diagram of multiplex immunofluorescence detection according to an embodiment of this application. Detailed Implementation

[0077] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0078] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in sequences other than those illustrated or described herein. In embodiments of this application, "B corresponding to A" means that B is associated with A. In one implementation, B can be determined based on A. However, it should also be understood that determining B based on A does not mean determining B solely based on A; B can also be determined based on A and / or other information. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or devices. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0079] In the embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0080] Terms and Definitions To facilitate understanding of this application, certain technical and scientific terms are specifically defined below. Unless otherwise expressly defined elsewhere in this document, all other technical and scientific terms used herein have the meanings commonly understood by one of ordinary skill in the art to which this application pertains.

[0081] Nanobodies (VHH) are single-domain antibody fragments derived from heavy chain antibodies of camels, with a molecular weight of approximately 15 kDa. Compared to traditional monoclonal or polyclonal antibodies, nanobodies offer significant advantages in structure, function, and application. Structurally, nanobodies have a small molecular weight (approximately 1 / 10 that of traditional IgG antibodies), a simple structure (containing only a single variable domain), strong tissue penetration ability, and high thermal stability (melting point Tm typically above 60°C), facilitating storage and transportation. Functionally, nanobodies have a broad epitope recognition range, capable of recognizing occult or recessed epitopes that are difficult for traditional antibodies to access; their affinity can be optimized through sequence design, exhibiting high specificity and ease of efficient expression in prokaryotic or eukaryotic expression systems. In terms of application, nanobodies have low production costs (approximately 1 / 5 to 1 / 10 that of traditional antibodies), good batch-to-batch consistency, are easily chemically modified (e.g., conjugated with fluorescein, enzymes, biotin, or drug molecules), and can form a completely independent intellectual property system. However, existing commercial nanobody products cannot meet the detection requirements of the TIMES Tumor Immune Microenvironment Assessment System of this invention for specific markers (SPON2, HLA-DRB1, ZFP36L2, ZFP36, and VIM). Therefore, developing proprietary nanobodies targeting the above five markers is of great significance for constructing a low-cost, highly consistent tumor recurrence prediction and detection kit with complete independent intellectual property rights.

[0082] The abbreviations for amino acid residues are the standard 3-letter and / or 1-letter codes used in this field to refer to one of the 20 commonly used L-amino acids. Specifically, they represent the following meanings: A: Ala (alanine); R: Arg (arginine); N: Asn (asparagine); D: Aspartic acid (aspartic acid); C: Cys (cysteine); Q: Gln (glutamine); E: Glu (glutamic acid); G: Glycine; H: Histidine; I: Ile (isoleucine); L: Leu (leucine); K: Lysine (lysine); M: Met (methionine); F: Phe (phenylalanine); P: Proline (proline); S: Serine (serine); T: Threonine (threonine); W: Tryptophan (tryptophan); Y: Tyrosine (tyrosine); V: Valine (valine).

[0083] For peptides, the term "(substantial) homology" is used to describe or compare the degree of amino acid similarity between two or more peptides or their designated sequences at optimal alignment and comparison (where appropriate insertions or deletions of nucleotides are made). The percentage of homology between two sequences varies with the number of identical positions shared by these sequences at optimal alignment (i.e., homology % = number of identical positions / total number of positions × 100), where optimal alignment is determined taking into account the number of vacancies introduced to achieve optimal alignment of the two sequences and the length of each vacancy. Sequence comparison and identity percentage determination between two sequences can be performed using mathematical algorithms, as described in the non-limiting examples below.

[0084] In this application, without substantially affecting antibody activity (retaining at least 95% activity), those skilled in the art can substitute, add, and / or delete one or more (e.g., 1, 2, 3, 4, 5, or more) amino acids in the sequence of this application to obtain variants with the same function as the nanobody of this application. These are all considered to be included within the scope of protection of this application. For example, replacing amino acids with similar properties in the variable region. The variant sequence described in this application can have at least 95% homology with the reference sequence, meaning at least 95% homology with each reference sequence, which can be 95%, 96%, 97%, 98%, 99%, 99.5%, or 100%. The sequence identity described in this application can be measured using sequence analysis software, such as the computer program BLAST using default parameters, especially BLASTP or TBLASTN. The amino acid sequences mentioned in this application are all shown in N-terminus to C-terminus format.

[0085] The TIMES scoring algorithm developed in this application has been validated through a multi-center, large-sample clinical cohort. The training cohort included 108 patients with hepatocellular carcinoma (HCC), of whom 54 had recurrence, with a median follow-up time of 36 months. The dataset was proportionally divided into a training set of 76 cases, a validation set of 16 cases, and a test set of 16 cases. External validation cohorts included: the USTC cohort (103 HCC patients, 48 ​​with recurrence, AUC 0.87) and the Singapore cohort (95 HCC patients, 41 with recurrence, AUC 0.85). The overall performance of the algorithm is as follows: training set AUC 0.90, validation set AUC 0.87; sensitivity 90.0%, specificity 90.2%; hazard ratio (HR) for recurrence between the high-risk and low-risk groups was 29.6 (P < 0.001). These results demonstrate that the TIMES scoring algorithm of this invention has high predictive accuracy and good clinical discrimination ability.

[0086] The kit described in this application establishes a comprehensive quality control (QC) system to ensure the stability and reliability of the product from nanobody raw materials to the final detection results.

[0087] Among them, nanobody quality control Purity: ≥95% (SDS-PAGE detection); Monomer ratio: ≥90% (SEC-HPLC detection); Endotoxin content: <1 EU / mg; Affinity: KD <50 nM (verified by SPR technology for each batch); Specificity: Single band detected by Western Blot.

[0088] Quality control of supporting reagents pH deviation: ±0.1; Concentration deviation: ±5%; Sterility test: negative; Stability: ≥95% activity maintained after 12 months of storage at 4°C.

[0089] Quality control of the dyeing process Positive control: Clear positive staining must be present; Negative control: No specific staining; Background staining: Score <2 (subjective score on a 0-5 scale).

[0090] Software quality control for predicting tumor recurrence risk Image quality inspection: Automatically identifies and alerts users to blurry, torn, or contaminated images; Traceable calculation process: Key steps are logged in audit logs; Data integrity protection: Prevents unauthorized manual tampering.

[0091] The reagent kit presented in this application exhibits high predictive accuracy, with an AUC of 0.90 on the training set and 0.87 on the validation set, achieving sensitivity and specificity both exceeding 90%. The hazard ratio (HR) for recurrence between the high-risk and low-risk groups is 29.6, significantly superior to traditional TNM staging (HR=1.93) and BCLC staging (HR=1.55). Through accurate risk stratification, this invention can guide personalized clinical decision-making: early initiation of adjuvant therapy and enhanced follow-up are recommended for high-risk patients, while overtreatment can be avoided for low-risk patients, thereby improving prognosis while reducing unnecessary medical burden. Secondly, the reagent kit presented in this application reduces costs by 70-80% compared to commercial antibody solutions and by approximately 90% compared to multiplex immunofluorescence (m-IHC) solutions, significantly lowering the expected detection cost per sample and demonstrating good potential for medical insurance coverage. Based on conventional IHC technology, this application requires no special equipment, is easy to operate, and is suitable for promotion in medical institutions at all levels. Simultaneously, by improving the overall survival rate of liver cancer patients and optimizing the allocation of medical resources, it can generate positive social benefits.

[0092] Table 5 summarizes the comparison between the embodiments of this application and existing major recurrence risk prediction technologies.

[0093] Table 5

[0094] The amino acid sequences involved in this application are shown in Table 1.

[0095] Table 1. Amino acid sequence

[0096] Devices, equipment, media and computer program products This application also relates to antibody sequence determination devices, tumor recurrence risk prediction devices, electronic devices, computer storage media, computer program products, or computer programs. These are described in detail below: The antibody sequence determination device according to this application includes: a sequence prediction module, a screening module, and a decision module; The sequence prediction module is used to predict antibody sequences based on the three-dimensional structure and epitope information of the target antigen, obtaining N candidate antibody sequences, where N is an integer not less than 1; the screening module is used to screen the N candidate nanobodies based on preset parameters, obtaining M preliminary screening antibody sequences, where the preset parameters include at least one of the following: structure confidence level pLDDT, interface prediction error PAE, and docking score, where M is an integer not less than 1 and not greater than M; the decision module is used to determine the target antibody sequence based on the affinity results of the M preliminary screening antibody sequences.

[0097] In an exemplary embodiment, the KD value of the target antibody affinity result is no greater than 50 nM.

[0098] The tumor recurrence risk prediction device according to this application includes: an image acquisition module, a region determination module, a calculation module, a score prediction module, a weight calculation module, and a result prediction module. The system includes: an image acquisition module for staining K genes in a target sample using an antibody from any example of the first aspect, resulting in a stained image where K is an integer not less than 1; a region determination module for selecting a predetermined number of regions at the tumor invasion front and tumor center in the stained image, resulting in multiple target regions, where the tumor invasion front is a region less than 2 mm from the tumor boundary and the tumor center is a region not less than 2 mm from the tumor boundary; a calculation module for calculating the positive cell ratio of each gene in each target region, resulting in K result sets; a scoring prediction module for inputting the K result sets into a trained marker scoring model, resulting in K marker scores; a weight calculation module for calculating the weight of each gene in the multiple target regions based on the K result sets, resulting in K weights; and a result prediction module for predicting the risk of tumor recurrence based on the K marker scores and the K weights, resulting in a prediction result.

[0099] In an exemplary embodiment, the step of calculating the weight of each gene in the multiple target regions based on the K result sets to obtain K weights includes: determining the mean positive cell ratio of the k-th gene for the k-th gene and its corresponding result set; calculating the standard deviation of the k-th gene in the multiple target regions based on the result set corresponding to the k-th gene and the mean positive cell ratio of the k-th gene; determining the variance of the k-th gene in the multiple target regions based on the standard deviation; using the reciprocal of the variance as the weight of the k-th gene; and traversing the K genes and K result sets to obtain the K weights.

[0100] In an exemplary embodiment, the step of predicting the risk of tumor recurrence based on the K marker scores and the K weights to obtain the prediction result includes: taking a weighted geometric average of the K marker scores and the K weights to obtain the prediction result.

[0101] It should be understood that the device embodiments and method embodiments can correspond to each other, and similar descriptions can be referred to the method embodiments. To avoid repetition, they will not be repeated here. Specifically, the device can execute the corresponding method embodiments, and the foregoing and other operations and / or functions of each module in the device are respectively for implementing the corresponding processes in the method, which will not be repeated here for the sake of brevity.

[0102] The apparatus of this application embodiment has been described above from the perspective of functional modules. It should be understood that this functional module can be implemented in hardware, in software instructions, or in a combination of hardware and software modules. Specifically, the steps of the method embodiments in this application can be completed by integrated logic circuits in the processor's hardware and / or by software instructions. The steps of the method disclosed in this application embodiment can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. Optionally, the software module can reside in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps in the above method embodiments.

[0103] Figure 1 This is a schematic block diagram of the electronic device 100 provided in an embodiment of this application. The electronic device 100 may be the training device or execution device described above, but is not limited thereto. Figure 1 As shown, the electronic device 100 may include: The system includes a memory 110 and a processor 120. The memory 110 stores a computer program 130 and transfers the computer program 130 to the processor 120. In other words, the processor 120 can retrieve and run the computer program 130 from the memory 110 to implement the methods described in the embodiments of this application.

[0104] For example, the processor 120 can be used to execute the steps in the above method according to the instructions in the computer program 130.

[0105] In some embodiments of this application, the processor 120 may include, but is not limited to: General-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0106] In some embodiments of this application, the memory 110 includes, but is not limited to: Volatile memory and / or non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0107] In some embodiments of this application, the computer program 130 may be divided into one or more modules, which are stored in the memory 110 and executed by the processor 120 to perform the method provided in this application. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program 130 in the electronic device.

[0108] like Figure 1 As shown, the electronic device 100 may further include: Transceiver 140, which can be connected to processor 120 or memory 110.

[0109] The processor 120 can control the transceiver 140 to communicate with other devices; specifically, it can send information or data to other devices or receive information or data sent by other devices. The transceiver 140 may include a transmitter and a receiver. The transceiver 140 may further include antennas, and the number of antennas may be one or more.

[0110] It should be understood that the various components in the electronic device 100 are connected through a bus system, which includes a data bus, a power bus, a control bus, and a status signal bus.

[0111] According to one aspect of this application, a computer-readable storage medium is provided that stores computer instructions or programs thereon, which, when executed by a computer, enable the computer to perform the methods of the above-described method embodiments. Alternatively, embodiments of this application also provide a computer program product containing instructions that, when executed by a computer, cause the computer to perform the methods of the above-described method embodiments.

[0112] According to another aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the method described in the above-described method embodiments.

[0113] In other words, when implemented using software, it can be implemented wholly or partially in the form of a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital video disc (DVD)), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0114] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0116] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to implement the solution of this embodiment according to actual needs. For example, the functional modules in the various embodiments of this application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0117] The following examples illustrate this application, but should not be construed as limiting the scope of the subject matter of this application to the following examples. All technologies implemented based on the above content of this application fall within the scope of this application. The compounds or reagents used in the following examples are commercially available or prepared using conventional methods known to those skilled in the art; the experimental instruments used are commercially available.

[0118] Example 1: Design, expression, and validation of nanobodies This embodiment is used to design anti-SPON2 nanobodies, anti-HLA-DRB1 nanobodies, anti-ZFP36L2 nanobodies, anti-ZFP36 nanobodies, and anti-VIM nanobodies. The design schemes for the aforementioned five nanobodies are identical. Specifically, taking the anti-SPON2 nanobodies (Nb-SPON2) as an example, the computational design, expression and purification, affinity determination, and immunohistochemical verification process of the nanobodies are described in detail.

[0119] First, the target protein SPON2 was structurally analyzed. A three-dimensional structural model of the SPON2 protein (UniProt ID: Q9BUD6) was obtained, preferably using the AlphaFold predicted structure. Based on this structure, epitope analysis was performed on the protein surface, and the N-terminal region (amino acid residues 50-100) with high spatial exposure and a certain degree of conservation was selected as candidate binding epitopes.

[0120] Based on this, a generative diffusion model (RFdiffusion) was used for nanobody sequence design: using the three-dimensional structure of SPON2 and the coordinates of the target epitope as input, the RFdiffusion model generated 500 candidate nanobody sequences. Further computational screening of these candidate nanobody sequences was performed, including: using AlphaFold2 to predict the monomer structure of the candidate sequences, retaining candidate sequences with pLDDT > 80, resulting in 120 candidate nanobodies; further using AlphaFold-Multimer to predict the nanobody-SPON2 complex structure, retaining candidate sequences with interface prediction error PAE < 10 Å, resulting in 35 candidate nanobodies; and using HADDOCK software for flexible docking optimization, retaining candidate sequences with docking score < -120, ultimately obtaining 8 high-confidence candidate nanobodies.

[0121] The eight candidate sequences were synthesized (GenScript) and cloned into the pPT5 expression vector to construct a recombinant expression vector fused with mouse IgG2a Fc. The recombinant plasmid was transfected into HEK293T cells for transient expression. After 72 hours of culture, the supernatant was collected and purified by Protein A affinity chromatography to obtain nanobody proteins with a purity greater than 95%.

[0122] Affinity of candidate antibodies was determined using surface plasmon resonance (SPR) technology. SPON2-Fc was immobilized on a CM5 chip (approximately 300 RU) via amine coupling using a BIAcore 8K SPR instrument. Nanobodies at gradient concentrations (0.78 nM–100 nM) were used as analytes for detection. The results are shown in Table 1. The dissociation constants (KD) of the eight candidate antibodies ranged from 18.5 nM to 95.1 nM, with nanobodied antibody number 3 exhibiting the best affinity (KD = 18.5 nM). Therefore, this candidate was selected as the target nanobodied antibody Nb-SPON2, and its amino acid sequence is shown in SEQ ID NO:1.

[0123] Table 1. Nanobody Affinity Detection Results

[0124] The staining performance was further validated using immunohistochemistry (IHC). Twenty FFPE tissue sections from hepatocellular carcinoma were selected as test samples, with a commercial anti-SPON2 antibody (Abcam ab12345) used as a control. The tissues were processed according to standard IHC procedures, including dewaxing, hydration, antigen retrieval with citrate buffer (pH 6.0), overnight incubation with primary antibody (1:100 dilution) at 4°C, incubation with HRP-labeled anti-Fc secondary antibody, DAB staining, and hematoxylin counterstaining.

[0125] The results are as follows Figure 2 As shown, the positive detection rate of Nb-SPON2 in liver cancer tissue was 75%, higher than the 70% of the control commercial antibody; the staining intensity reached a medium to strong level, which is superior to the medium intensity of the commercial antibody; background staining was significantly reduced; and it showed a specific positive signal in the tumor-infiltrating NK cell region. These results indicate that the Nb-SPON2 obtained in this embodiment is superior to existing commercial antibodies in terms of affinity, specificity, and tissue staining performance.

[0126] Following the same steps, anti-HLA-DRB1 nanobody (Nb-HLA-DRB1), anti-ZFP36L2 nanobody (Nb-ZFP36L2), anti-ZFP36 nanobody (Nb-ZFP36), and anti-VIM nanobody (Nb-VIM) were obtained, with amino acid sequences as shown in SEQ ID NO: 2-5, respectively.

[0127] Example 2: Prediction of Postoperative Recurrence Risk in Liver Cancer Patients In this embodiment, the nanobody obtained in Example 1 was used to predict the risk of postoperative recurrence in a patient with hepatocellular carcinoma, and its clinical value was verified through follow-up.

[0128] Patient basic information: Male, 58 years old, diagnosed with solitary hepatocellular carcinoma, underwent radical hepatectomy (March 2024). Pathological results showed moderately differentiated HCC, tumor diameter 4.5 cm, no vascular invasion, TNM stage T2N0M0 (stage II), BCLC stage A.

[0129] The detection uses the kit described in this application, and the specific detection procedure is as follows: First, the paraffin-embedded hepatocellular carcinoma tissue (FFPE) obtained after surgery is sectioned, with the section thickness controlled at 4–5 μm, ensuring that the section includes the tumor tissue area and the boundary area between the tumor and normal tissue. For single-color immunohistochemical detection, five consecutive sections are prepared, each corresponding to one marker; or, in multiplex immunofluorescence detection mode, a single section can be used for multiple rounds of detection.

[0130] The tissue sections were then pretreated. They were dewaxed in xylene (15 minutes each time, twice in total), then subjected to a gradient hydration treatment with 100%, 95%, 85% and 75% ethanol solutions, and finally rehydrated with distilled water and rinsed three times with PBS buffer for 3 minutes each time.

[0131] Antigen retrieval is performed after pretreatment. Preferably, the slides are placed in citrate buffer at pH 6.0 and treated at 121°C for 30 minutes using autoclaving, followed by natural cooling to room temperature; alternatively, microwave retrieval (high heat for 5 minutes followed by medium heat for 10 minutes) or water bath retrieval at 95–100°C for 20 minutes can be used.

[0132] Subsequently, blocking treatment was performed. First, endogenous peroxidase was blocked by incubation with 3% hydrogen peroxide solution at room temperature for 10 minutes, followed by washing with PBS. When using biotinylated nanobodies, ovalbumin and D-biotin blocking treatments were also performed sequentially. Then, 5% BSA or serum was used for incubation at room temperature for 30 minutes to reduce non-specific binding.

[0133] In the primary antibody incubation step, for monochromatic immunohistochemical detection, five nanobodies are diluted 1:100 to 1:200 in PBS buffer containing 0.1% BSA and applied to the corresponding sections. They are incubated overnight (16–18 hours) at 4°C or for 1 hour at 37°C, followed by washing with PBS. For multiplex immunofluorescence detection, different nanobodies are sequentially incubated on the same section. After each incubation, the corresponding fluorescently labeled secondary antibody is attached, and the antibody is removed by heat treatment to achieve continuous detection of multiple labels.

[0134] After primary antibody incubation, secondary antibody incubation and signal detection are performed. In the immunohistochemical detection system, HRP-labeled anti-Fc secondary antibody or streptavidin is added and incubated at room temperature for 30 minutes. After washing, freshly prepared DAB chromogenic solution is added for color development, and the degree of color development is monitored under a microscope. The reaction is terminated when a brown signal appears. In the immunofluorescence detection system, fluorescently labeled secondary antibody is added and incubated in the dark. TSA signal amplification can be performed as needed.

[0135] After color development or fluorescent labeling, counterstaining and mounting are performed. For the IHC method ( Figure 3 Nuclear counterstaining was performed using hematoxylin, followed by graded ethanol dehydration and xylene clearing, and then mounting with neutral resin; for immunofluorescence methods ( Figure 4 Nuclear staining was performed using DAPI, and the slides were mounted using antifluorescence quenching mounting medium.

[0136] After mounting, images of the slides are acquired. Immunohistochemical samples are scanned using an optical microscope or a whole-slide scanner, while immunofluorescence samples are acquired using a fluorescence microscope or a confocal microscope. A magnification of 20× and a resolution of at least 0.5 μm / pixel are preferred. Simultaneously, quality control checks are performed to ensure that positive and negative controls meet expectations and that there is no significant background staining or uneven staining.

[0137] The images were then analyzed using the accompanying software. After importing the scanned images, the tumor region was divided. The area less than 2 mm from the tumor boundary was defined as the tumor infiltration front (IF), and the area greater than or equal to 2 mm from the tumor boundary was defined as the tumor center (TC). Multiple fields of view were randomly selected for analysis within the IF and TC regions, with a preference for selecting 5 fields of view in each region, avoiding necrotic or hemorrhage areas.

[0138] Positive cells were counted in each field of view. The criteria for a positive result were: brown deposits in the cytoplasm, nucleus, or cell membrane during immunohistochemistry; and a signal intensity at least twice the background intensity during immunofluorescence. AI-assisted identification was used to calculate the proportion of positive cells to the total number of cells.

[0139] The predicted values ​​p1-p5 for the five markers were obtained based on the pre-trained XGBoost model. The TIMES score was further calculated using the inverse variance weighted geometric mean method, with the following formula: TIMES = exp(Σ[w k × ln(p k )] / Σw k ), where TIMES represents the prediction result, w k p represents the weight of the k-th gene. k This represents the score of the k-th marker.

[0140] The test results are shown in Table 2.

[0141] Table 2

[0142] The TIMES score calculation process is as follows: First, the XGBoost model outputs five single-marker predicted values: p1(SPON2) = 0.62, p2(HLA-DRB1) = 0.71, p3(ZFP36L2) = 0.68, p4(ZFP36) = 0.58, and p5(VIM) = 0.74; then, the inverse variance weight is calculated: w1 = 1 / 3.7. 2 =0.073, w2=1 / 5.6 2 =0.032, w3=1 / 4.8 2 =0.043, w4=1 / 3.6 2 =0.077, w5=1 / 4.4 2 =0.052; Finally, the TIMES score was calculated using the weighted geometric mean formula: TIMES = exp[(0.073×ln0.62 + 0.032×ln0.71 + 0.043×ln0.68 + 0.077×ln0.58 + 0.052×ln0.74) / (0.073+0.032+0.043+0.077+0.052)] = exp[(-0.437) / 0.277]= 0.646.

[0143] The risk stratification result was TIMES=0.646>0.5, classifying the patient as being in the high-risk recurrence group. Clinical recommendations: Although the traditional TNM and BCLC staging assessment indicated low to intermediate risk, it is still recommended to initiate sorafenib adjuvant therapy within 1 month post-surgery, with CT / MRI imaging follow-up every 3 months and monitoring of tumor markers such as AFP.

[0144] Follow-up results showed that the patient strictly adhered to the doctor's orders for adjuvant therapy and close follow-up. Ten months post-surgery, a CT scan revealed a new small nodule in the liver (0.8 cm in diameter), and a biopsy confirmed it as HCC recurrence. Due to early detection, the patient received timely local ablation therapy, and their condition is currently stable.

[0145] Conclusion: The TIMES score in this embodiment successfully warned of high recurrence risk that traditional staging systems failed to identify, enabling early intervention and significantly improving clinical management.

[0146] Example 3: Large-scale production and quality control of the reagent kit This embodiment details the batch production process, purification procedure, quality control system, reagent preparation, and stability verification process for five nanobodies to demonstrate the industrial-scale production capability and batch-to-batch consistency of the kit.

[0147] The nanobody production process is as follows: The HEK293T suspension culture system is used, the expression vector is pPT5-Nb-Fc (fused with mouse IgG2a Fc tag), the transfection method is PEI transient transfection, the culture volume is 10 L bioreactor, the culture time is 72 hours, and the supernatant is collected by centrifugation.

[0148] The nanobody purification process includes: (1) Protein A affinity chromatography (column: Protein A Sepharose, loading flow rate 5 mL / min, washing with 10 column volumes of PBS, eluting with 0.1 M glycine-HCl pH 2.7, and immediately neutralizing with 1 M Tris-HCl pH 8.0); (2) desalting dialysis (PBS, overnight at 4℃) and ultrafiltration concentration (10 kDa cutoff, final concentration 1-5 mg / mL); (3) 0.22 μm sterile filtration and aliquoting.

[0149] The quality control items and standards are as follows: (1) SDS-PAGE detection purity ≥95% (main band molecular weight about 40 kDa); (2) SEC-HPLC detection monomer ratio ≥90%; (3) SPR affinity determination (5 samples are randomly selected from each batch, KD is within the qualified range ±20%, and the batch CV <15%); (4) Endotoxin LAL test <1 EU / mg; (5) Sterility test (TSB / TSA culture negative); (6) Functional verification (IHC staining intensity ≥80% of the reference standard).

[0150] The reagent preparation and dispensing include: antibody dilution solution (PBS + 0.1% BSA + 0.02% NaN3, adjusted to working concentration according to titer, 0.5 mL / tube); antigen retrieval solution, blocking solution, DAB chromogenic solution (solution A and solution B are separated) and other supporting reagents are pre-prepared; the final packaging includes 10 vials of each of the 5 types of nanobodies (0.5 mL / vial), supporting reagents, 5 positive / negative control slides, instructions and quality inspection report.

[0151] Stability test results showed that: short-term stability at 4℃ (0, 1, 3, 6, 12 months) maintained ≥95% affinity and IHC efficacy; long-term stability at -20℃ (24 months) maintained ≥95% activity; and activity maintained ≥90% after 3 freeze-thaw cycles.

[0152] Batch production validation: Three independent batches (batch numbers 2024001, 2024002, and 2024003) were produced consecutively, with 10,000 samples per batch. The quality comparison results are shown in Table 3.

[0153] Table 3

[0154] The results show that the kit in this embodiment has excellent batch-to-batch consistency, fully complies with medical device production standards, and can achieve stable and large-scale supply.

[0155] Comparative Example 1: Performance Comparison of the Nanobody in this Application with Commercial Antibodies To demonstrate the advantages of the nanobody of the present invention over existing commercial antibodies in terms of cost, consistency, affinity, and staining effect, a systematic comparative experiment was conducted.

[0156] The results of the comparative experiment are shown in Table 4.

[0157] Table 4

[0158] The specific comparison data for Nb-SPON2 is as follows: Affinity comparison: Commercial anti-SPON2 antibody (Abcam ab12345): KD = 45 nM; Nb-SPON2 of this application (SEQ ID NO:1): KD = 18.5 nM. The results show that the affinity of the nanobody of this invention is approximately 2.4 times higher than that of the commercial antibody.

[0159] Comparison of IHC staining results (108 hepatocellular carcinoma (HCC) samples): Commercial antibody: positive detection rate 68.5%, average signal-to-noise ratio 3.2; Nanobody of this application: positive detection rate 72.2%, average signal-to-noise ratio 4.8. The results show that the nanobody of this invention is superior to commercial antibodies in both sensitivity and specificity.

[0160] Comparison of batch-to-batch consistency (3 independent production batches): Commercial antibody: batch-to-batch CV = 18.3%; Nanobody of this application: batch-to-batch CV = 6.7%. The results show that the batch-to-batch consistency of the nanobody of this application is approximately 2.7 times better than that of the commercial antibody.

[0161] Cost Comparison (based on 1000 clinical tests): Commercial antibody regimen: 5 antibodies × 1500 yuan / vial × 10 vials (each vial can be used for 100 tests) = 75,000 yuan, cost per test: 75 yuan; This invention's nanobody regimen: 5 antibodies × 400 yuan / vial × 10 vials = 20,000 yuan, cost per test: 20 yuan. The results show that the nanobody of this application reduces the cost per test by 73%.

[0162] The above comparative experimental results fully demonstrate that the nanobody of this application is significantly superior to existing commercial antibodies in terms of production cost, batch-to-batch consistency, affinity, thermal stability, staining effect, and supply autonomy, providing strong technical support for the construction of a low-cost, high-reliability tumor recurrence prediction and detection kit.

[0163] The preferred embodiments of this application have been described in detail above; however, this application is not limited thereto. Within the scope of the technical concept of this application, various simple modifications can be made to the technical solution of this application, including combining various technical features in any other suitable manner. These simple modifications and combinations should also be considered as the content disclosed in this application and are all within the protection scope of this application.

[0164] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0165] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. An antibody group, characterized in that, The antibody group consists of the antibodies shown in SEQ ID NO: 1-5.

2. The use of antibody groups in the preparation of a liver cancer recurrence prediction kit, wherein, The antibody group is as described in claim 1.

3. A reagent kit, characterized in that, include: The antibody group according to claim 1.

4. The reagent kit according to claim 3, characterized in that, The antibody concentration is 1 μg / mL to 5 μg / mL.

5. The reagent kit according to claim 3, characterized in that, The kit further includes at least one of the following: tissue pretreatment reagent, antigen retrieval reagent, signal amplification reagent, staining reagent, mounting reagent, and quality control product.

6. The reagent kit according to claim 5, characterized in that, The tissue pretreatment reagents include at least one of xylene, dewaxing solution, 75% (v / v)-100% (v / v) ethanol and PBS buffer, wherein the PBS buffer has a pH of 7-7.5 and contains 0.03%-0.06% Tween-20.

7. The reagent kit according to claim 5, characterized in that, The antigen retrieval reagent includes at least one of the following: citrate buffer with pH 5.8-6.2, EDTA buffer with pH 7.8-8.2, and Tris-EDTA buffer with pH 8.8-9.

2.

8. The reagent kit according to claim 5, characterized in that, The mounting reagent includes at least one of peroxidase blocking solution, biotin blocking solution, and serum blocking solution.

9. The reagent kit according to claim 8, characterized in that, The peroxidase blocking solution comprises 2%-5% H2O2.

10. The reagent kit according to claim 8, characterized in that, The biotin blocking solution comprises ovalbumin and D-biotin.

11. The reagent kit according to claim 8, characterized in that, The serum blocking solution includes: 5% BSA or goat serum.

12. The reagent kit according to claim 5, characterized in that, The signal amplification reagent includes at least one of the following: HRP-labeled anti-Fc secondary antibody, HRP-streptavidin, DAB chromogenic solution, fluorescently labeled secondary antibody, and TSA signal amplification agent.

13. The reagent kit according to claim 5, characterized in that, The staining reagents include: hematoxylin solution or DAPI solution.

14. The reagent kit according to claim 5, characterized in that, The mounting reagent includes: neutral resin or anti-fluorescence quenching mounting medium.

Images