A biomarker for predicting the responsiveness of multiple myeloma to a bortezomib-based treatment regimen, and methods of detecting and using the same

Abstract

This invention provides a biomarker for predicting the response of multiple myeloma to bortezomib-based treatment regimens, along with its detection method and application. This invention discovers that CD14 in the bone marrow microenvironment... + IL1B + The abundance of tumor-associated monocytes (TAMO_IL1B) was significantly positively correlated with patient responsiveness to bortezomib-based treatment regimens, with TAMO_IL1B abundance significantly higher in the best responder group than in the second-best responder group. This invention also found that TAMO_IL1B interacts with malignant plasma cells via a THBS1-SDC1 ligand-receptor relationship, and the strength of this interaction was positively correlated with good responsiveness to bortezomib-based treatment regimens. This invention further provides a combination of gene markers comprising B4GALT1, C3AR1, HCLS1, COMMD8, NAGA, SAMHD1, and ZBTB16 for assessing the prognosis of patients with multiple myeloma. This invention also provides corresponding detection kits and predictive systems. This invention can be used for personalized medication guidance and prognostic assessment of patients with multiple myeloma.

Description

Technical Field

[0001] This invention relates to the field of multiple myeloma technology, specifically to a biomarker for predicting the responsiveness of multiple myeloma to bortezomib-based treatment regimens, its detection method, and its application. Background Technology

[0002] Multiple myeloma (MM) is a common hematologic malignancy characterized by the clonal proliferation of malignant plasma cells in the bone marrow, leading to progressive end-organ damage, manifested as osteolytic lesions, anemia, renal impairment, and hypercalcemia. Despite significant advancements in recent years with novel treatment strategies such as proteasome inhibitors, immunomodulators, and monoclonal antibodies, MM remains an incurable disease, and most patients eventually relapse due to acquired drug resistance.

[0003] Bortezomib (BTZ), a proteasome inhibitor, inhibits the degradation of target proteins by the 26S proteasome, thereby preventing the degradation of pro-apoptotic factors and triggering programmed cell death in tumor cells. Bortezomib-containing regimens induce significant anti-tumor responses in multiple myeloma (MM) and are considered first-line treatments, widely used for induction, consolidation, maintenance, or salvage therapy. However, bortezomib resistance remains a significant clinical challenge.

[0004] In recent years, significant progress has been made in research on the mechanisms of bortezomib resistance. Shi et al., through analysis of the GEO dataset GSE9782, discovered that NPM1 (nucleophosphorus protein 1) mediates bortezomib resistance in MM cells by upregulating CXCR4 expression. MM patients with high NPM1 expression showed a significantly reduced remission rate with bortezomib-containing regimens, suggesting NPM1 as a potential new prognostic biomarker and therapeutic target. Liao et al. found that IFI16 drives bortezomib resistance by activating the Wnt / β-catenin signaling pathway. IFI16 expression was significantly elevated in newly diagnosed and relapsed MM, and it was associated with bone lesions, renal impairment, and poor survival. Furthermore, exosome-derived HSPA9 can propagate bortezomib resistance via the USP1 / TRIP13 signaling axis. Exosomes secreted by BTZ-resistant MM cells, such as HSPA9, can transfer resistance characteristics to sensitive cells via endocytosis. TRIP13 overexpression has also been shown to alter mitochondrial function, reduce BTZ-induced apoptosis, and alleviate G2 / M cell cycle arrest. 6 A-methylation-mediated H19 upregulation promotes BTZ resistance by regulating the miR-184 / CARM1 axis. These studies demonstrate that the molecular mechanisms of bortezomib resistance are highly complex, involving the aberrant activation of multiple signaling pathways and regulatory networks.

[0005] However, the occurrence and progression of multiple myeloma (MM) depend not only on the genetic damage of tumor cells themselves but also on the profound influence of the bone marrow microenvironment (BME). The BME is a complex ecosystem composed of malignant plasma cells, immune cells (including T cells, NK cells, monocytes, macrophages, dendritic cells, etc.), hematopoietic stem cells, mesenchymal stromal cells, and the extracellular matrix. Numerous studies have shown that various immune cell subsets in the BME can promote myeloma cell growth, immune escape, and drug resistance through multiple mechanisms such as cytokine secretion, expression of immune checkpoint molecules, and metabolic reprogramming. Among these, monocytes, as key innate immune cells, exhibit phenotypic heterogeneity, including classic (CD14) ++ CD16 - ), intermediate type (CD14) + CD16 + ) and non-classical (CD14) + CD16 ++ Subgroups of MM, as well as cytokine secretion (such as IL-6 and TGF-β) and immunosuppressive mechanisms, play important roles in the pathogenesis and drug response of MM.

[0006] In recent years, the emergence of single-cell RNA sequencing (scRNA-seq) technology has provided a revolutionary tool for analyzing cellular heterogeneity and molecular interaction networks in the bone marrow microenvironment. This technology can characterize the transcriptomic features of different cell types unbiasedly at the single-cell level, reveal rare cell subpopulations, and reconstruct cell differentiation trajectories and intercellular communication networks. The application of scRNA-seq in MM research has initially revealed the molecular evolution from precancerous plasma cell disease to overt malignant tumors, mapped the cellular composition of the bone marrow microenvironment, and begun to explore molecular features associated with bortezomib sensitivity and resistance. For example, Seung-Hyun et al. identified three monocyte subpopulations associated with bortezomib resistance using scRNA-seq analysis. The An Gang / Qiu Lugui team used single-cell mapping to discover that monocytes in MM patients exhibited a significantly enhanced type I interferon response, revealing the key role of this signal-driven abnormal developmental pathway in the formation of immunosuppression. Other studies have identified PDIA4 as a biomarker for MM progression and treatment vulnerability using single-cell transcriptomics.

[0007] Despite the significant progress made in the above research, the following shortcomings still exist in the existing technology:

[0008] First, no studies have yet reported specific tumor-associated monocyte subsets, particularly CD14. + IL1B +Association between the abundance of tumor-associated monocytes (TAMO_IL1B) and the responsiveness of multiple myeloma patients to bortezomib-based therapy regimens. IL1B + The functional role of monocytes / macrophages in MM remains controversial, and their specific mechanisms of action in response to bortezomib-based treatment regimens have not been fully elucidated.

[0009] Second, the role of the THBS1-SDC1 ligand-receptor in intercellular communication between tumor-associated monocytes and malignant plasma cells, and its impact on the responsiveness to bortezomib-based treatment regimens, has not been reported in the current technology. THBS1 (thromboretin-1), as a key activator of TGF-β, has complex immune functions in the tumor microenvironment; SDC1 (CD138) is a transmembrane heparan sulfate proteoglycan highly expressed on the surface of myeloma cells. However, whether the ligand-receptor interaction between the two participates in regulating the therapeutic response to bortezomib-based treatment regimens remains unresearched.

[0010] Third, current technologies lack systematic single-cell-level analysis of the association between the composition of monocyte subsets in the pre-treatment bone marrow microenvironment and treatment outcomes of bortezomib-based regimens. Most studies focus on intrinsic tumor cell resistance mechanisms, while insufficient attention is paid to the value of immune cells (especially monocytes) in the microenvironment in predicting treatment response.

[0011] Fourth, there is a lack of reliable biomarkers, detection methods, and kits in the current technology that can effectively predict the response of multiple myeloma patients to bortezomib-based treatment regimens. Summary of the Invention

[0012] To address the shortcomings of existing technologies, this invention provides a class of biomarkers that can be used to predict the response of multiple myeloma to bortezomib-based treatment regimens; this invention also provides a method for detecting such biomarkers and their applications.

[0013] To achieve the above objectives, the present invention provides the following technical solution:

[0014] A biomarker for predicting the response of multiple myeloma to bortezomib-based treatment regimens, comprising the detection of CD14 in ex vivo bone marrow samples. + IL1B + Reagent for tumor-associated monocyte (TAMO_IL1B) abundance;

[0015] The CD14 + IL1B +Tumor-associated mononuclear cells are a subset of bone marrow-derived mononuclear cells that co-express CD14 and IL1B, identified by single-cell transcriptome sequencing or flow cytometry.

[0016] The abundance is CD14. + IL1B + The percentage or absolute number of tumor-associated monocytes in the total monocyte population of an ex vivo bone marrow sample.

[0017] Preferably, the reagent comprises:

[0018] Antibodies or antigen-binding fragments thereof that specifically bind to the CD14 protein; and

[0019] Antibodies or antigen-binding fragments thereof that specifically bind to IL1B protein;

[0020] Each antibody independently carries a detectable marker, which includes, but is not limited to, fluorescent dyes, enzymes, biotin, radioisotopes, or quantum dots.

[0021] Preferably, the biomarker composition further comprises reagents for detecting THBS1 expression levels and / or reagents for detecting SDC1 expression levels.

[0022] This invention also discloses a kit for predicting the response of multiple myeloma patients to bortezomib-based treatment regimens, comprising:

[0023] The first reagent capable of specifically binding to CD14;

[0024] A second reagent capable of specifically binding to IL1B;

[0025] A third reagent capable of specifically binding to THBS1;

[0026] A fourth reagent capable of specifically binding to SDC1;

[0027] Instructions for use, which include the detection of CD14 in ex vivo bone marrow samples using the first and second reagents. + IL1B + Operational guidelines for using tumor-associated mononuclear cell abundance to predict patient responsiveness to bortezomib-based treatment regimens;

[0028] The instruction manual also includes operational guidelines for detecting the expression levels of THBS1 and SDC1 using the third and fourth reagents, respectively, wherein the expression level of THBS1 is derived from CD14. + IL1B +Tumor-associated mononuclear cells, wherein the expression level of SDC1 was derived from malignant plasma cells in the same sample, and the expression levels of THBS1 and SDC1 were used to assess CD14. + IL1B + THBS1-SDC1 communication inference score with tumor-associated monocytes as the sending cell population and malignant plasma cells as the receiving cell population;

[0029] The communication inference score is calculated from the expression levels of THBS1 and SDC1 according to a predetermined cell communication analysis rule. The positive correlation between the communication inference score and treatment responsiveness is established in advance under a determined treatment regimen, sampling time point, and efficacy evaluation criteria.

[0030] Preferably, the first reagent is an anti-CD14 antibody with a detectable label, and the second reagent is an anti-IL1B antibody with a detectable label;

[0031] The kit also includes auxiliary reagents for flow cytometry detection, which are selected from one or more of erythrocyte lysis buffer, cell fixative, flow cytometry gating control antibody, and isotype control antibody.

[0032] Preferably, the instruction manual includes the following prediction criteria:

[0033] When CD14 in isolated bone marrow samples + IL1B + When the abundance of tumor-associated mononuclear cells is above a predetermined threshold, it indicates that the patient has a good response to bortezomib-based treatment regimens;

[0034] The predetermined threshold is determined by CD14 in the suboptimal responder population. + IL1B + The distribution of tumor-associated mononuclear cell abundance was determined by statistical analysis.

[0035] This invention also discloses a system for predicting the response of multiple myeloma patients to bortezomib-based treatment regimens, comprising:

[0036] A sample processing unit for processing isolated bone marrow samples to prepare single-cell suspensions;

[0037] Detection unit for detecting CD14 in isolated bone marrow samples. + IL1B + Abundance of tumor-associated monocytes; and optionally, detection of THBS1 and SDC1 expression levels;

[0038] Storage unit, which stores reference level data;

[0039] The analysis unit is used to compare the abundance measured by the detection unit with the reference level, and to calculate CD14 based on the expression levels of THBS1 and SDC1. + IL1B + The THBS1-SDC1 communication inference score, with tumor-associated mononuclear cells as the sending cell population and malignant plasma cells as the receiving cell population, is used to generate predictive information based on the comparison and calculation results.

[0040] Output unit for displaying or reporting the patient's predicted response to bortezomib-based treatment regimens.

[0041] This invention also discloses CD14 + IL1B + The use of tumor-associated monocytes as a biomarker in the development of kits for predicting the response of multiple myeloma patients to bortezomib-based treatment regimens.

[0042] The kit contains reagents for detecting CD14 expression levels and reagents for detecting IL1B expression levels;

[0043] The kit also includes reagents for detecting THBS1 expression levels and / or reagents for detecting SDC1 expression levels.

[0044] This invention also discloses a method for detecting the strength of THBS1-SDC1 ligand-receptor interaction in bone marrow samples in vitro, comprising the following steps:

[0045] Step S1: Obtain an ex vivo bone marrow sample;

[0046] Step S2: Detect CD14 in bone marrow samples in vitro. + IL1B + Expression level of THBS1 in tumor-associated monocytes;

[0047] Step S3: Detect the expression level of SDC1 in malignant plasma cells in bone marrow samples in vitro;

[0048] Step S4: Calculate the strength value of the THBS1-SDC1 ligand-receptor interaction based on the detection results of steps S2 and S3.

[0049] The present invention also discloses a set of gene biomarkers for assessing the prognosis of patients with multiple myeloma, the set of gene biomarkers including one or more genes selected from the following: B4GALT1, C3AR1, HCLS1, COMMD8, NAGA, SAMHD1 and ZBTB16;

[0050] The combination of genetic markers is used to assess the prognosis of patients with multiple myeloma by detecting the expression levels of the genes in ex vivo bone marrow samples.

[0051] This invention constructs a single-cell atlas of the bone marrow microenvironment in multiple myeloma, encompassing 15 cell types. This not only provides fundamental data for understanding the cellular composition and heterogeneity of the bone marrow microenvironment in multiple myeloma, but also offers an important reference framework and value for subsequent research on the role of monocyte subsets in disease progression and treatment response.

[0052] This invention provides a biomarker for predicting the response of multiple myeloma to bortezomib-based treatment regimens, with the following beneficial effects: identification of a specific monocyte subset CD14 associated with a favorable response at the single-cell level. + IL1B + Tumor-associated monocytes (TAMO_IL1B) were identified, and their abundance was confirmed as a positive predictor. This finding reveals the functional heterogeneity of monocyte subsets in response to bortezomib-based treatment regimens.

[0053] This study is the first to discover and validate the crucial role of the THBS1-SDC1 ligand-receptor pair in intercellular communication between TAMO_IL1B cells and malignant plasma cells (TPCs). Specifically, THBS1 is highly expressed in TAMO_IL1B cells, and SDC1 is highly expressed in malignant plasma cells, forming a ligand-receptor interaction pair. The interaction strength of the THBS1 signaling pathway between TAMO_IL1B cells and TPCs was significantly higher in the OR group than in the SOR group. THBS1-SDC1 exhibited a stronger interaction in the OR group than the known THBS1-CD47 interaction. The expression levels of THBS1 in TAMO_IL1B cells and SDC1 in TPC cells were significantly higher in the OR group than in the SOR group. These findings provide a novel perspective for understanding the mechanisms of drug sensitivity mediated by the bone marrow microenvironment and lay a theoretical foundation for developing novel therapeutic strategies targeting the THBS1-SDC1 axis.

[0054] This study reveals for the first time the unique position of TAMO_IL1B in the monocyte differentiation lineage: mono_CD14 cells are in the early pseudo-time stage, while TAMO_IL1B cells are in the late pseudo-time stage; TAMO_IL1B cells are located at the end of the differentiation trajectory and are only present in the OR group; increased B4GALT1 expression was observed in the late stage, while effector molecules such as COMMD8, HCLS1, and SAMHD1 were gradually lost; in the OR group, B4GALT1 was highly expressed, while C3AR1, HCLS1, COMMD8, NAGA, SAMHD1, and ZBTB16 were lowly expressed. These results indicate that TAMO_IL1B represents one of the terminal states of monocyte differentiation, and its unique transcriptomic characteristics are closely related to a good response to bortezomib-based treatment regimens.

[0055] For the first time, the THBS1-SDC1 ligand-receptor pair has been proposed as a drug target for screening candidate drugs that can enhance the sensitivity of multiple myeloma cells to bortezomib-based treatment regimens. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in this invention or the prior art, the accompanying drawings used in the description of this invention or the prior art will be briefly introduced below.

[0057] Figure 1 This is a UMAP image of the single-cell microenvironment of bone marrow microenvironments in multiple myeloma (MM) and healthy donors (HD) in Embodiment 3 of the present invention. Figure 1 A is a UMAP plot showing the source information of cell samples (HD, OR, and SOR groups are marked with different colors). Figure 1 B is the UMAP projection map of bone marrow cells in the HD, OR, and SOR groups; Figure 1 C represents the UMAP diagram for 15 cell types; Figure 1 D is a bubble heatmap of the expression of characteristic marker genes for each cell type; Figure 1 E is a stacked bar chart showing the proportion of each cell type in the three groups: HD, OR, and SOR. Figure 1 F is a quantitative comparison graph of each cell type in the three groups of HD, OR, and SOR;

[0058] Figure 2 This is a schematic diagram illustrating cell type screening based on drug response in Embodiment 4 of the present invention. Figure 2 A shows the UMAP density map of cells in the HD, OR, and SOR groups; Figure 2 B is CD14 + Monocytes, CD16 + Principal component analysis (PCA) plots of monocytes and T cells in the HD, OR, and SOR groups.

[0059] Figure 3 This is a characterization diagram of plasma cell subsets in Embodiment 5 of the present invention. Figure 3 A is a UMAP diagram of the six plasma cell subsets (PB, PC, TPB, TPB_CIN, TPC, TPC_CIN); Figure 3 B represents the UMAP projection of plasma cell division in the HD, OR, and SOR groups; Figure 3 C is a bubble heatmap showing the expression of characteristic marker genes for each plasma cell subset; Figure 3 D is a bar chart showing the proportions of plasma cell subsets in the HD, OR, and SOR groups; Figure 3 E represents the quantitative comparison of each plasma cell subset in the HD, OR, and SOR groups; Figure 3 F is the UMAP diagram of plasma cells carrying high-risk cytogenetic alterations; Figure 3 G is a heatmap of the average expression levels of 10 MM driver genes in each patient; Figure 3 H is a characteristic map of the expression levels of 10 MM driver genes in plasma cell subsets.

[0060] Figure 4 This is a characterization diagram of a monocyte subset in Embodiment Six of the present invention. Figure 4 A is a UMAP diagram of six monocyte subsets (mono_CD14, TAMO_IL1B, TAMO_PADI4, TAMO_MX1, mono_CD16, and macrophages); Figure 4 B represents the UMAP projection of monocyte division in the HD, OR, and SOR groups; Figure 4 C is a bubble heatmap of the expression of characteristic marker genes for each monocyte subset; Figure 4 D is a density map of the expression of characteristic marker genes of each monocyte subset; Figure 4 E is a bar chart showing the proportion of monocyte subsets in the three groups: HD, OR, and SOR. Figure 4 F represents the quantitative comparison of each monocyte subset in the HD, OR, and SOR groups.

[0061] Figure 5 This is a diagram illustrating the signaling pathway and cell activity analysis in Embodiment Seven of the present invention. Figure 5 A represents the pathways that were differentially enriched in monocytes of the OR and SOR groups according to GSVA analysis; Figure 5 B is a GSVA heatmap of 50 hallmark gene sets among different subgroups in MM; Figure 5 C- Figure 5 E represents the superimposed distribution of single-cell AUC scores for selected classical pathways (TGF-β signaling pathway, inflammatory response pathway, TNFA signaling pathway) on the three UMAP maps of HD, OR, and SOR.

[0062] Figure 6 This is a schematic diagram of the interaction pattern between OR and SOR cells in Embodiment 8 of the present invention. Figure 6 A shows the differential interaction weights / strengths between specified cell types in the OR group (left) and the SOR group (right) using CirclePlot; Figure 6 B represents CellChat network analysis, showing the contribution of each cell type to the signaling network output and input in the OR and SOR groups; Figure 6 Figure C shows the differences in intercellular communication mediated by the THBS signaling axis in the OR group (left) and the SOR group (right); Figure 6 D and Figure 6 E represents the quantitative comparison of THBS1 and SDC1 expression levels in the OR group and the SOR group (t-test). Figure 6 F is a heatmap showing the differences in intercellular communication between specified cell types in the OR group (left) and the SOR group (right); Figure 6 G is a bubble diagram comparing important ligand-receptor pairs from TAMO_IL1B to various target cells (macrophages, PCs, TAMO_IL1B, TAMO_PADI4, TAMO_MX1, TPB, TPB_CIN, TPC, TPC_CIN) in the OR and SOR groups, with the THBS1-SDC1 ligand-receptor pair marked by a red solid box.

[0063] Figure 7 These are the pseudo-time trajectory analysis diagrams of monocytes in Example 9 and the survival analysis diagrams of prognostic gene markers in Example 10 of the present invention. Figure 7 A and Figure 7 B is a Monocle2 pseudo-time trajectory plot, displaying the pseudo-time curve ( Figure 7 A) and the dynamic distribution of monocyte subsets ( Figure 7 B); Figure 7 C represents the pseudo-temporal distribution of monocytes in the three groups of samples: HD, OR, and SOR. Figure 7 D represents the heatmap hierarchical clustering showing genes differentially expressed along the pseudo-time curve; Figure 7 E represents the expression changes of selected genes that are pseudo-time-dependent; Figure 7 F is a volcano diagram showing the differentially expressed genes between the OR group and the SOR group monocyte subsets; Figure 7 G represents a quantitative comparison of the expression levels of B4GALT1, C3AR1, HCLS1, COMMD8, NAGA, SAMHD1, and ZBTB16 in the OR and SOR groups (t-test). Figure 7 H represents the Kaplan-Meier survival analysis based on the differential expression of the above seven characteristic genes.

[0064] Figure 8This is a supplementary analysis diagram of the single-cell atlas of the bone marrow microenvironment in multiple myeloma (MM) and healthy donor (HD) in Embodiment 3 of the present invention. Among them, Figure 8 A is a UMAP projection of the division of MM cells in three groups: HD (n=10), OR (n=10), and SOR (n=8), and is colored according to the cell type of transcriptome expression annotation; Figure 8 B is a stacked bar chart showing the proportion of each cell type in 28 samples (HD, OR, SOR); Figure 8 C represents the overlay of cell type-specific gene expression levels on the UMAP plot.

[0065] Figure 9 This is a multivariate logistic regression analysis diagram showing the association between TAMO_IL1B abundance and treatment response in Embodiment Six of the present invention. Figure 9 A represents the effect of TAMO_IL1B score on the predictive probability of treatment response after stratification by high-risk cytogenetics; Figure 9 B is the ROC curve of the multivariate logistic regression model, with an area under the curve (AUC) of 0.767.

[0066] Figure 10 This is a verification diagram of the THBS1 and SDC1 protein levels in Example 11 of the present invention. Figure 10 A represents the validation results of THBS1 protein expression in the THP-1 cell line (data sourced from the Human Protein Atlas (HPA) public database). Figure 10 B represents the validation results of SDC1 protein expression in the LP-1 cell line; Figure 10 C is a ranking plot of SDC1 and THBS1 proteins in myeloma samples, sorted by the median raw intensity value of each protein in the myeloma samples. The color intensity represents the percentage of effective detection values ​​of that protein in the myeloma dataset.

[0067] Figure 11 This diagram illustrates the structural similarity analysis of THBS signaling axis-mediated intercellular communication and signaling pathways between the OR and SOR groups in Embodiment 8 of the present invention. Figure 11 A is a Circle diagram showing differential intercellular communication between specified cell types mediated by the THBS signal axis in the OR group (left) and the SOR group (right), with the side width representing the communication probability; Figure 11 B is a density map of THBS1 gene expression in various cell types; Figure 11 C is a characteristic graph of normalized RNA expression levels of THBS1 in the OR and SOR groups; Figure 11 D is a density map of SDC1 gene expression in various cell types; Figure 11 E is a characteristic map of normalized RNA expression levels of SDC1 in the OR and SOR groups; Figure 11F is a clustering graph that clusters the signal pathways of the OR group and the SOR group into a shared two-dimensional manifold based on structural similarity. The circles and squares represent the signal networks of the OR group and the SOR group, respectively, and different colors represent different signal pathway groups. Figure 11 G is Figure 11 Enlarged views of each pathway group in F.

[0068] Figure 12 This is an analysis diagram of the monocyte subset diffusion (DC) in Embodiment 9 of the present invention. Among them, Figure 12 A represents a three-dimensional representation of the terminal differentiation characteristic scores of monocyte subsets based on diffusion components 1-3; Figure 12 B represents the distribution of terminal differentiation feature scores on the DC embedding map; Figure 12 C represents the analysis of monocyte subsets (DCs) stained by group (HD, OR, SOR); Figure 12 D represents the DC analysis of monocyte subsets stained by cell type.

[0069] Figure 13 This is a verification diagram showing the source of the prognostic value of the TAMO_IL1B characteristic gene in Example 10 of the present invention. Among them, Figure 13 A- Figure 13 G represents the Kaplan-Meier survival analysis (based on deconvolution analysis of the GSE136324 dataset) of the specific expression levels of B4GALT1, C3AR1, HCLS1, COMMD8, NAGA, SAMHD1 and ZBTB16 in monocyte compartments. All genes showed no statistically significant association (p>0.05). Figure 13 H is a forest plot of univariate Cox regression analysis for the above 7 genes in purified MM cells from the CoMMpass cohort, showing COMMD8 (HR=1.667, p=3×10). -9 ), HCLS1 (HR=1.690, p=1.69×10 -6 ) and NAGA (HR=1.480, p=9.44×10) -7 () is the strongest predictor of poor prognosis. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0071] Example 1: Acquisition and Processing of Bone Marrow Samples

[0072] Research subjects:

[0073] The single-cell transcriptome sequencing data used in this invention were obtained from two publicly available GEO databases. Bone marrow sample data from healthy donors were obtained from the GSE120221 dataset (Oetjen et al., JCI Insight 2018), comprising 10 healthy donors. Bone marrow sample data from multiple myeloma patients were obtained from the GSE189460 dataset (Jung et al., Exp Mol Med 2022), comprising 18 treatment-naïve multiple myeloma patients, including 10 in the best responder (OR) group and 8 in the second-best responder (SOR) group. Detailed information on the two single-cell transcriptome sequencing datasets used in this invention (including dataset number, source literature, tissue type, sample type, number of patients, sequencing platform, and total number of cells after quality control) is summarized in Table 1 below:

[0074] Table 1 Information on the GEO dataset used in this study

[0075]

[0076] All 18 patients received bortezomib-melphalan-prednisone (VMP) combination therapy after bone marrow aspiration. According to the International Myeloma Working Group (IMWG) unified response criteria, patients were divided into two groups: those achieving complete remission (CR) or very good partial remission (VGPR) were classified as best responders (OR group, n=10); those achieving partial remission (PR), stable disease (SD), or progressive disease (PD) were classified as second-best responders (SOR group, n=8). Ten healthy donors served as the control group (HD group, n=10). Detailed clinical information for each sample, including sample number, age, sex, genotype, FISH test results, treatment regimen, treatment response, and cell count, is summarized in Table 2 below:

[0077] Table 2. Single-cell RNA sequencing data from 10 healthy subjects and 18 patients with multiple myeloma.

[0078]

[0079] Example 2: Preprocessing and Quality Control of Single-Cell RNA Sequencing Data

[0080] 2.1 Data Acquisition and Preliminary Processing:

[0081] The raw gene expression matrix was obtained from the Gene Expression Omnibus (GEO) database using the GEOquery Bioconductor package. The raw gene expression matrix was then processed using the Seurat (v4.0) package.

[0082] Initial quality control involves filtering cells according to the following criteria:

[0083] (1) Cells with a total UMI count of less than 1,000;

[0084] (2) Cells with fewer than 500 detected genes;

[0085] (3) Cells with a mitochondrial gene content greater than 5%;

[0086] (4) Cells with less than 1% of red blood cell-related genes.

[0087] The single-cell transcriptome sequencing data involved in this invention comes from two publicly available GEO datasets: the healthy donor bone marrow sample dataset GSE120221 (Oetjen et al., 2019) and the multiple myeloma patient bone marrow sample dataset GSE189460 (Jung et al., 2022). Detailed information on the two datasets is shown in Table 1.

[0088] 2.2 Dual-cell removal:

[0089] Two-cell detection was performed using the DoubletFinder (v2.0.3) software package with the following parameters: pN=0.25, pK=0.09 (PCA neighborhood size parameter), nExp=0.08×nCells 2 / 10,000 (expected number of double cells). Double cell detection is performed independently for each sample, and then cells identified as double cells are removed.

[0090] After completing the double-cell removal, all samples that have passed quality control and double-cell filtering are merged into a single Seurat object.

[0091] 2.3 Data Normalization and Batch Effect Correction:

[0092] (1) Data normalization: Data was processed according to the Seurat standard procedure. The LogNormalize method (scaling factor = 10000) was used to normalize the data. The normalized data are the log-normalized expression values ​​of each gene in each cell.

[0093] (2) Identification of hypervariable genes: 2000 hypervariable genes (variable features) were identified using Seurat’s “FindVariableFeatures” function, and the selection method was vst (variance stable transformation).

[0094] (3) Data scaling: Use Seurat’s ScaleData function to linearly scale the data, adjusting the expression values ​​of all genes to a mean of 0 and a variance of 1.

[0095] (4) Principal Component Analysis (PCA): Perform principal component analysis (PCA) on the scaled data. Calculate the first 50 principal components (PCs).

[0096] (5) Batch effect correction: Batch effects between samples were corrected using the Harmony (v0.1.1) software package. The PCA dimensionality reduction results were used as input, and Harmony eliminated the differences between batches through iterative clustering and linear correction.

[0097] 2.4 Cell Clustering and Annotation:

[0098] The KNN graph was constructed using Seurat's FindNeighbors function (dims=1-50), and the FindClusters function (resolution=2.0) was used to cluster the cells of different types.

[0099] The SingleR (v1.6.1) software package was used for automatic annotation, with a human primary cell dataset as the reference database. Manual annotation and validation were performed in conjunction with known cell type-specific marker genes. Annotated cell types included: plasma cells, B cells, pre-B cells, early pre-B cells, NK cells, T cells, and CD14. + Monocytes, CD16 + Monocytes, myeloid dendritic cells (mDCs), plasmacytoid dendritic cells (pDCs), hematopoietic stem cells (HSCs), granulocyte-monocyte progenitor cells (GMPs), megakaryocytes, megakaryocyte-erythroid progenitor cells (MEPs), and stromal cells.

[0100] UMAP dimensionality reduction and visualization were performed using Seurat's UMAP (Uniform Manifold Approximation and Projection) function. The number of each cell type in each sample was counted, and their percentage of the total number of cells was calculated.

[0101] After undergoing the aforementioned quality control (QC), double cell removal, data normalization, and batch effect correction, the raw sequencing data yielded high-quality single-cell gene expression matrices. Following this quality control process, a total of 56,568 high-quality cells were obtained for subsequent analysis, including 24,500 cells from 10 healthy donors and 32,068 cells from 18 MM patients.

[0102] 2.5 Cell distribution characteristics of samples from different sources:

[0103] like Figure 1As shown in Figure A, to illustrate the distribution of cells from different sources, the cell source information for the HD group (healthy donors, n=10), OR group (best responders, n=10), and SOR group (second-best responders, n=8) is indicated on the UMAP plot using different colors. Figure 1 As can be seen, cells from different patient sources exhibit certain distribution characteristics in the UMAP space, with some cell sources forming relatively independent clusters. This suggests that there is significant heterogeneity in cell composition of the bone marrow microenvironment among different individuals and different disease states (healthy, best response, and suboptimal response). It also verifies that after Harmony batch effect correction, the sequencing data of samples from different sources were effectively integrated, and the batch effect was controlled within an acceptable range. Figure 1 A and Figure 1 B corroborates each other, among which Figure 1 B further illustrates the overall overlap in the distribution of the three sample groups in the UMAP space, suggesting that although individual heterogeneity exists, there are still common cell type clusters among the sample groups. Based on this unsupervised clustering analysis, cell type annotation was further performed on all cells using classic marker genes, as shown in the results. Figure 1 As shown in C.

[0104] Example 3: Construction of a bone marrow microenvironment cellular atlas and identification of cell types

[0105] 3.1 Identification of fifteen cell types:

[0106] like Figure 1 As shown in Figure C, UMAP dimensionality reduction visualization, combining unsupervised clustering with classical marker gene annotation, reveals that fifteen cell types form independent clusters in two-dimensional space. Fifteen cell types were identified from bone marrow samples: plasma cells, B cells, pre-B cells, pro-B cells, NK cells, T cells, and CD14. + Monocytes, CD16 + Monocytes, myeloid dendritic cells (mDCs), plasmacytoid dendritic cells (pDCs), hematopoietic stem cells (HSCs), granulocyte-monocyte progenitor cells (GMPs), megakaryocytes, megakaryocyte-erythroid progenitor cells (MEPs), and stromal cells.

[0107] like Figure 1 As shown in the bubble heatmap, each cell type expresses its own characteristic marker genes (such as CD14 being a marker gene for monocytes, CD3D being a marker gene for T cells, etc.). Figure 1B shows the distribution of bone marrow cells in the UMAP space for the HD, OR, and SOR groups, and it can be seen that some cell populations have different distributions among the three groups.

[0108] 3.2 Supplementary analysis of cell type distribution in different groups:

[0109] like Figure 1 As shown in Figure C, fifteen cell types were identified from bone marrow samples using unsupervised clustering combined with classical marker gene annotation. To further illustrate the distribution of each cell type across different groups (HD, OR, SOR), Figure 8 Figure A uses split UMAP projection to show the distribution characteristics of various cell types in the three groups of samples. As can be seen from the figure, some cell types (such as plasma cells, CD14 cells, etc.) show distribution characteristics. + The distribution of monocytes differed significantly among the three groups. Figure 1 The overall UMAP diagram of C corroborates each other.

[0110] Figure 8 B presents the proportion of each cell type in 28 samples (10 HD, 10 OR, and 8 SOR) in the form of stacked bar charts. As can be seen from the figure, there is significant individual heterogeneity in cell type composition among different samples, but the proportion of plasma cells in the OR and SOR groups is significantly higher than that in the HD group. Figure 1 The summary results for E are consistent.

[0111] Figure 8 C shows the superposition of expression levels of cell type-specific marker genes (such as CD14, CD3D, CD79A, etc.) on the UMAP map, further validating the... Figure 1 The expression characteristics of cell type marker genes in the D-bubble heatmap confirmed the accuracy of cell type annotation.

[0112] 3.3 Intergroup comparison of cell type ratios:

[0113] Statistical analysis was performed on the proportions of each cell type in the HD, OR, and SOR groups. For example... Figure 1 As shown in the stacked bar chart of E, there are significant differences in cell type composition among the HD, OR, and SOR groups. Specifically, as... Figure 1 As shown in Figure F, the proportion of malignant plasma cells was significantly increased in both the OR and SOR groups compared to the HD group. B cells, T cells, and pDCs were reduced in both the OR and SOR groups, while pro-B cells, NK cells, mDCs, and megakaryocytes were significantly depleted in the SOR group. HSCs, GMPs, pre-B cells, and CD14 were also significantly reduced. + Monocytes, CD16 +Monocytes, stromal cells, and MEPs showed no significant overall changes across groups, but certain cell types (such as CD14) were observed in some individual patients. + The proliferation of monocytes and stromal cells indicates a significant remodeling of the bone marrow microenvironment in MM patients.

[0114] Example 4: Screening Cell Types Based on Drug Response

[0115] 4.1 Cell density analysis:

[0116] like Figure 2 As shown in Figure A, the distribution density of each cell type in the HD, OR, and SOR groups was analyzed using UMAP density views. The results showed that CD14... + Monocyte density (cells / mm) 2 The levels of CD14 were significantly higher in the OR group than in the SOR group. Notably, CD14 levels were significantly higher in the OR group. + The density of monocytes was greater in the SOR group than in the CD14 group, suggesting that CD14 + Monocytes may improve the response to bortezomib-based treatments to block MM growth.

[0117] 4.2 Three-dimensional principal component analysis of transcriptome profiles:

[0118] like Figure 2 As shown in B, when comparing CD14 of the OR group and the SOR group + Monocytes, CD16 + In transcriptomic profiling of monocytes and T cells, three-dimensional principal component analysis (3D-PCA) showed that CD14 in the OR group... + Monocytes and CD16 + Monocytes were the furthest apart along the PC1, PC2, and PC3 axes, exhibiting clear separation. This result indicates that monocytes (especially CD14) are highly segregated. + Transcriptomic characteristics of monocytes are closely related to bortezomib treatment response.

[0119] Example 5: Identification and Characterization of Plasma Cell Subpopulations

[0120] 5.1 Identification of plasma cell subsets:

[0121] like Figure 3 As shown in Figure A, subpopulation analysis was performed on cells annotated as plasma cells. Using UMAP visualization, six distinct plasma cell subpopulations at the transcriptional level were identified: PB, PC, TPB, TPC, TPB_CIN, and TPC_CIN. Figure 3 As shown in the C-bubble heatmap, the bubble heatmap displays the expression of characteristic marker genes in each subgroup.

[0122] 5.2 Intergroup comparison of plasma cell subset ratios:

[0123] like Figure 3 As shown by the UMAP projection of B cells, PB and PC cells are enriched in HD samples, while TPB, TPB_CIN, TPC, and TPC_CIN are more prevalent in MM samples. Figure 3 As shown in the DE, the proportion of TPC cells increased in both the OR and SOR groups, with the proportion of TPC_CIN cells significantly higher in the OR group than in the SOR group (P=0.023). Both PB and PC cells were severely depleted in both the OR and SOR groups. There were no significant differences in TPB and TPB_CIN cells between the OR and SOR groups and the HD group, but TPB and TPB_CIN cell types were amplified in some individual patients.

[0124] 5.3 Identification of high-risk cytogenetic plasma cells:

[0125] like Figure 3 As shown in F, plasma cells carrying high-risk cytogenetic alterations were annotated based on transcriptomic features. For translocations and amplifications (e.g., t(4;14), t(11;14), 1q gain), malignant plasma cells (TPCs) highly expressing characteristic oncogenes (NSD2 / FGFR3, CCND1, and BCL9 / MCL1 / UBE2Q1, respectively) were identified as high-risk TPCs. For deletions (e.g., del17p, del13q), TPCs with significantly reduced or absent expression of the affected tumor suppressor genes (TP53 / FXR2 for del17p, RB1 / DIS3 for del13q) were defined as high-risk TPCs.

[0126] like Figure 3 The heatmap shown in G displays the average expression levels of 10 MM driver genes (including NSD2, FGFR3, CCND1, BCL9, MCL1, UBE2Q1, TP53, FXR2, RB1, and DIS3) in each patient, with the corresponding cytogenetic information labeled at the top. Figure 3 As shown in the feature map of H, the expression levels of the above 10 MM driver genes in plasma cell subsets are visualized using UMAP feature maps.

[0127] Example 6: CD14 + IL1B + Identification and Quantitative Analysis of Tumor-Associated Monocytes (TAMO_IL1B)

[0128] 6.1 Reclustering of monocyte subsets:

[0129] Mononuclear cell tags were screened from all annotated cells, and re-clustering analysis was performed using Seurat, following the method of Jung et al. (Exp Mol Med 2022), with clustering parameters consistent with those in Example 2 above.

[0130] 6.2 Identification of TAMO_IL1B:

[0131] Based on clustering results and characteristic gene expression profiles, subpopulation analysis was performed on cells annotated as monocytes. For example... Figure 4 As shown in Figure A, six distinct monocyte subsets at the transcriptional level were identified using UMAP visualization: mono_CD14, TAMO_IL1B, TAMO_PADI4, TAMO_MX1, mono_CD16, and macrophages. Characteristic marker genes for each subset are presented using bubble heatmaps and density maps.

[0132] The TAMO_IL1B subset is characterized by CD14 positivity and IL1B positivity. For example... Figure 4 As shown in the C-bubble heatmap, each monocyte subset expresses its own characteristic marker genes. For example... Figure 4 As shown in the D-density diagram, each characteristic marker gene exhibits high-density expression in its corresponding monocyte subset.

[0133] like Figure 4 As shown by the splitting UMAP projection of B, Mono_CD14 and Mono_CD16 cells were enriched in the HD group, while TAMO_IL1B, TAMO_PADI4, TAMO_MX1 and macrophages were more prevalent in the MM sample.

[0134] 6.3 Quantitative analysis and intergroup comparison of TAMO_IL1B abundance:

[0135] Calculate the proportion of TAMO_IL1B cells to the total number of monocytes in each sample. The calculation formula is:

[0136] TAMO_IL1B abundance = (number of TAMO_IL1B cells / total number of monocytes) × 100%;

[0137] The total number of monocytes includes all CD14 cells. + Monocytes (containing mono_CD14, TAMO_IL1B, TAMO_PADI4, TAMO_MX1), CD16 + Monocytes (mono_CD16) and macrophages.

[0138] Comparison of TAMO_IL1B abundance in different response groups:

[0139] The abundance of TAMO_IL1B in the best responder (OR) group and the second-best responder (SOR) group was compared as follows:

[0140] (1) Perform statistical tests on the abundance of TAMO_IL1B for each sample in the OR and SOR groups. If the data are normally distributed and have homogeneous variance, use the independent samples t test; if they are not normally distributed, use the Wilcoxon rank-sum test (Mann-Whitney U test).

[0141] (2) The statistical significance threshold was set to P < 0.05.

[0142] The results are as follows Figure 4 As shown in Figure F, the proportion of TAMO_IL1B in the OR group was significantly higher than that in the SOR group (P=0.045). This finding indicates that the TAMO_IL1B subgroup is closely related to the good responsiveness of bortezomib.

[0143] like Figure 4 As shown in the bar chart for E, the proportions of TAMO_IL1B, TAMO_PADI4, TAMO_MX1 and macrophages were increased in MM patients, while mono_CD14 and mono_CD16 cells were severely depleted in both the OR and SOR groups.

[0144] 6.4 Multivariate logistic regression analysis of TAMO_IL1B abundance and treatment response:

[0145] To assess the association between TAMO_IL1B abundance and treatment response, multivariate logistic regression analysis was performed in a subgroup of patients (n=11) with high-risk cytogenetic data. Treatment response grouping (OR=1, SOR=0) was the dependent variable, and TAMO_IL1B abundance and high-risk cytogenetic status were the independent variables.

[0146] The results are as follows Figure 9 As shown in Figure A, after stratification by high-risk cytogenetic factors, an increase in TAMO_IL1B score was positively correlated with an increase in the predicted probability of treatment response. Figure 9 B shows the receiver operating characteristic (ROC) curve of the multivariate logistic regression model, with an area under the curve (AUC) of 0.767. The multivariate model adjusted for high-risk cytogenetic factors showed a positive correlation between TAMO_IL1B abundance and treatment response, but due to the small sample size, this effect did not reach statistical significance (p=0.157). This result suggests that increased TAMO_IL1B abundance may be associated with improved treatment response.

[0147] 6.5 Flow cytometry detection method for TAMO_IL1B:

[0148] Based on the findings of this invention, a specific method for detecting the abundance of TAMO_IL1B in bone marrow samples using flow cytometry is provided:

[0149] (1) Sample preparation: Take fresh bone marrow samples, add red blood cell lysis buffer, incubate at room temperature in the dark, centrifuge and discard the supernatant, and wash the cells once with PBS.

[0150] (2) Cell surface staining: Resuspend the cells in FACS buffer (PBS + 2% FBS) and add fluorescently labeled anti-CD14 antibody (used to label CD14). + Mononuclear cells were used, and anti-CD45 antibody (for leukocyte gating) and anti-CD16 antibody were optionally added. Isotype control tubes were prepared. The cells were incubated at 4°C in the dark, and the supernatant was discarded after washing.

[0151] (3) Intracellular IL1B staining: After surface staining, cells were treated with cell fixation / permeabilization reagents. Fluorescently labeled anti-IL1B antibody was added, and the cells were incubated at 4°C in the dark. After washing, the cells were resuspended in FACS buffer.

[0152] (4) Flow cytometry data acquisition: Data is acquired on a flow cytometer, and gating is performed sequentially to exclude cell debris and adherent cells, and to delineate leukocyte populations and monocyte populations (CD14). + ), in CD14 + Analysis of IL1B in monocyte populations + The proportion of cells.

[0153] (5) Data analysis: Data analysis was performed using flow cytometry software to calculate the abundance of TAMO_IL1B: (CD14) + IL1B + Cell count / CD14 + (Total number of monocytes) × 100%.

[0154] Example 7: Signaling Pathway and Cell Activity Analysis

[0155] 7.1 Gene Set Variation Analysis (GSVA):

[0156] Gene set variation analysis (GSVA) ​​was performed on 50 hallmark gene sets from the MSigDB database using the GSVA package (v1.46.0). Subsequently, LIMMA (v3.54.2) was used to assess differences in pathway activity between the OR and SOR groups.

[0157] like Figure 5As shown in Figure A, 38 pathways were significantly enriched in monocytes of the OR group, including the TGF-β signaling pathway, the TNFA signaling pathway (via NFκB), and the inflammatory response signaling pathway. Four pathways were significantly enriched in the SOR group, including interferon α response, KRAS signal downregulation, MYC target V2, and the pancreatic β-cell signaling pathway.

[0158] 7.2 Signaling pathways enriched in various monocyte subsets:

[0159] like Figure 5 The GSVA heatmap of TAMO_IL1B cells shows that the number of signaling pathways enriched in TAMO_IL1B cells is greater than that in other monocyte subsets. Pathways significantly upregulated in TAMO_IL1B cells include: TGF-β signaling pathway, TNFA signaling pathway (via NFκB), IL2-STAT5 signaling pathway, inflammatory response pathway, KRAS downregulation, angiogenesis, early estrogen response, Hedgehog signaling pathway, UV response upregulation, WNT-β-catenin signaling pathway, and coagulation signaling pathway (11 pathways in total). MYC target V1, MYC target V2, and unfolded protein response pathways are significantly upregulated in macrophages. HEME metabolism, myogenesis, exogenous substance metabolism, and glycolysis signaling pathways are significantly upregulated in TAMO_PADI4 cells. Interferon α response, interferon γ response, peroxisome, PI3K-AKT-mTOR signaling pathway, complement, and IL6-JAK-STAT3 signaling pathways are significantly upregulated in TAMO_MX1 cells.

[0160] 7.3 Single-cell pathway activity score:

[0161] Cellular pathway activity was quantified using the AUCell R package (v1.24.0). First, the AUCell_buildRankings function (default parameters) was used to generate gene expression rankings for each cell based on the transcriptome matrix. Then, using the Broad Institute's classic pathway definition, the area under the curve (AUC) score was calculated using AUCell_calcAUC to quantify the proportion of top-ranked genes for the corresponding pathway component in each cell. The Kolmogorov-Smirnov test was used to compare the AUC score distributions among the HD, OR, and SOR groups.

[0162] like Figure 5 C- Figure 5 As shown in E, the TGF-β signaling pathway ( Figure 5 C) Inflammatory response pathways ( Figure 5 D) and the TNFA signaling pathway (via NFκB) Figure 5E) There were significant differences in the distribution of AUC scores in single cells of the HD, OR and SOR groups, especially in the OR group, which showed higher pathway activity.

[0163] Example 8: Analysis of intercellular communication network and verification of the THBS1-SDC1 axis

[0164] 8.1 Inference of intercellular communication:

[0165] Cell-cell communication networks were inferred and analyzed using the CellChat (v1.6.1) software package. CellChat objects were created using a normalized gene expression matrix and cell metadata (including cell type tags). The built-in human ligand-receptor interaction database of CellChat was loaded, integrating three sub-databases: "Secretory Signals," "Extracellular Matrix-Receptor," and "Cell-Cell Contact." The probability of communication between each pair of cell types via a specific ligand-receptor pair was calculated using the computeCommunProb function. The statistical significance of the communication was assessed using a permutation test. The cell-cell communication networks were visualized using the visualization tools provided by CellChat.

[0166] In the CellChat analysis results, ligand-receptor interaction data related to the THBS1 signaling pathway were extracted to obtain the communication probability value between TAMO_IL1B and malignant plasma cells (TPC) via the THBS1-SDC1 ligand-receptor pair. The strength of the THBS1-SDC1 interaction is defined as this communication probability value, which ranges from 0 to 1, with a larger value indicating a stronger interaction.

[0167] 8.2 Intergroup comparison of the strength of cell-cell interactions:

[0168] Compare the differences in the strength of THBS1-SDC1 interaction between the best responder (OR) group and the second-best responder (SOR) group.

[0169] like Figure 6 As shown in Figure A, CirclePlot displays the differential interaction weights / strengths between specified cell types in the OR group (left) and the SOR group (right). The interaction weights / strengths of TAMO_IL1B-TPC and TAMO_PADI4-TPC were significantly higher in the OR group, while the interaction weights / strengths of TPC_CIN-TPB, TPC_CIN-TPB_CIN, and TPC_CIN-TPC were higher in the SOR group. This result indicates that intercellular communication between TAMO_IL1B and malignant plasma cells is more active in the OR group.

[0170] like Figure 6As shown in Figure B, CellChat network analysis revealed that TPC and TAMO_IL1B played important roles in signal output and input in the OR group, while TPC and TPC_CIN played the main roles in the SOR group. The cell distribution in two-dimensional space further indicated that the contribution of TAMO_IL1B cells to the signaling network was significantly enhanced in the OR group.

[0171] 8.3 Identification of THBS1-SDC1 ligand-receptor:

[0172] Analysis revealed that the THBS signaling pathway is CD14. + The main source of ligands in monocytes, these ligands act on malignant plasma cells. For example... Figure 6 As shown in Figure C (left: OR group, right: SOR group), compared with the SOR group, the THBS signaling pathway in the OR group is significantly upregulated, and CD14... + The interaction between monocytes and malignant plasma cells is stronger. Figure 11 A further illustrates this difference in the form of a Circle diagram, showing that the communication side width between TAMO_IL1B and TPC in the OR group is significantly larger than that in the SOR group.

[0173] Further analysis of the signal flow in the THBS pathway revealed that CD14 + Monocytes are the primary target cells for receiving THBS1 ligands, while SDC1 is secreted by plasma cells.

[0174] 8.4 THBS1-SDC1 interaction between TAMO_IL1B and TPC:

[0175] like Figure 6 As shown in C, the THBS signaling pathway is the main source of ligands in TAMO_IL1B, and these ligands act on TPC cells. Compared with the SOR group ( Figure 6 Compared to group C (right), group OR (right) Figure 6 In the left side of C, the THBS signaling pathway between TAMO_IL1B and TPC cells is significantly upregulated and the interaction is stronger (side width represents the communication probability).

[0176] like Figure 6 As shown in DE, the THBS1 expression level of TAMO_IL1B in the OR group ( Figure 6 D) and SDC1 expression level in TPC cells ( Figure 6 E) was significantly higher than that of the SOR group (t test).

[0177] like Figure 6 As shown in the F heatmap, there are significant differences in intercellular communication between TAMO_IL1B and TPC in the OR group (left) and the SOR group (right).

[0178] like Figure 6 As shown in G, the THBS signaling pathway involves three ligand-receptor pairs: THBS1-SDC1, THBS1-CD47, and THBS1-CD36. THBS1-CD36 was relatively stable across different groups, while THBS1-SDC1 and THBS1-CD47 showed significant differences between the OR and SOR groups. More importantly, THBS1-SDC1 exhibited a stronger interaction than THBS1-CD47 in the OR group. Figure 6 (marked by a red solid box in G).

[0179] 8.5 Validation of THBS1-SDC1 axis expression and similarity analysis of signaling pathway structure:

[0180] To further verify the expression characteristics of the THBS1-SDC1 axis at the single-cell level and the classification of its signaling pathways, the inventors conducted supplementary analysis.

[0181] like Figure 11 As shown in Figure A, there are significant differences in THBS signaling axis-mediated intercellular communication between the OR group (left) and the SOR group (right). As can be seen from the figure, the communication boundary between TAMO_IL1B and TPC in the OR group is significantly wider than that in the SOR group, indicating a higher probability of THBS signaling pathway communication in the OR group. Figure 6 The results of the main plot in group C are consistent, further confirming that there is stronger intercellular communication between TAMO_IL1B and TPC through the THBS signaling axis in the OR group.

[0182] like Figure 11 As shown in Figure B, the expression density distribution of the THBS1 gene in monocyte subsets shows that THBS1 is highly expressed in the TAMO_IL1B subset, while its expression level is low in other monocyte subsets such as mono_CD14 and mono_CD16. Figure 11 C further illustrates the characteristic maps of normalized RNA expression levels of THBS1 in the OR and SOR groups, showing that the expression level of THBS1 in the OR group was significantly higher than that in the SOR group. These results are consistent with... Figure 6 Consistent with the main plot results, D further confirmed at the single-cell level that THBS1 is specifically highly expressed in TAMO_IL1B, and that this expression characteristic is associated with a good response to bortezomib-based treatment regimens.

[0183] like Figure 11 As shown in Figure D, the expression density distribution of the SDC1 gene in plasma cell subsets shows that SDC1 is highly expressed in the malignant plasma cell (TPC) subset, while the expression level is low in normal plasma cells (PB, PC). Figure 11E further illustrates the characteristic maps of normalized RNA expression levels of SDC1 in the OR and SOR groups, showing that the expression level of SDC1 in the OR group was significantly higher than that in the SOR group. These results are consistent with... Figure 6 Consistent with the main plot results of E, this further confirms at the single-cell level that SDC1 is specifically highly expressed in TPC, and that this expression characteristic is associated with good responsiveness to bortezomib.

[0184] like Figure 11 As shown in F, based on the structural similarity of the signal paths, the signal paths of the OR group and the SOR group are further clustered into a shared two-dimensional manifold. Here, circles represent the signal networks of the OR group, squares represent the signal networks of the SOR group, the size of the dots or squares is proportional to the total communication probability, and different colors represent different signal path groups. Figure 11 G is an enlarged view of each pathway group.

[0185] Of particular note is that the THBS signaling pathway and the TGF-β signaling pathway were grouped into the same group (Group 1) in the structural similarity clustering. This finding has significant technical implications: the TGF-β signaling pathway is known to play a crucial role in the regulation of the bone marrow microenvironment in multiple myeloma (MM), supporting MM progression by promoting bone resorption and immune dysfunction. THBS1 has been identified as a key regulator of latent TGF-β activation, capable of increasing the level of biologically active TGF-β in MM models. The grouping of THBS and TGF-β into the same signaling pathway group supports, from a computational biology perspective, the mechanism model in this invention that "TAMO_IL1B activates the TGF-β signaling pathway through the THBS1-SDC1 axis, thereby shaping a bone marrow microenvironment favorable to bortezomib response."

[0186] The above Figure 11 AG results and Figure 6 The main plot results of AG corroborate each other, together forming a complete chain of evidence that the THBS1-SDC1 axis plays a key role in the communication between TAMO_IL1B and TPC cells.

[0187] 8.6 Immunohistochemical-based method for co-localization detection of THBS1 and SDC1:

[0188] An immunohistochemical (IHC)-based multiplex staining method is provided for detecting the expression and spatial distribution of THBS1 and SDC1 in bone marrow biopsy tissues.

[0189] (1) Preparation of tissue sections: After fixing and embedding the bone marrow biopsy tissue, tissue sections are prepared.

[0190] (2) Antigen retrieval: Place the slide in an appropriate buffer solution for antigen retrieval.

[0191] (3) Peroxidase blocking: Incubation with H2O2 to block endogenous peroxidase activity.

[0192] (4) Immunohistochemical staining: Immunohistochemical staining was performed sequentially using anti-THBS1 antibody and anti-SDC1 antibody, and detection was performed using different enzyme-labeled secondary antibodies and different chromogenic substrates (such as DAB and VectorBlue).

[0193] (5) Image acquisition and analysis: Images were acquired using a digital pathology scanner, and THBS1 was calculated using image analysis software. + Cells and SDC1 + Cell density and spatial overlap coefficient in tissue sections.

[0194] Example 9: Pseudo-time trajectory analysis of monocyte subsets

[0195] 9.1 Monocle2 Pseudo-Time Analysis:

[0196] Differentiation trajectories of monocyte subsets were inferred using the Monocle (v2.26.0) package. The raw UMI counting matrix processed by Seurat and cell metadata were input into a CellDataSet object (expressionFamily=negbinomial.size()). Genes with a mean expression level ≥0.1 were retained. Differentially expressed genes (FDR < 1%) were selected and dimensionality was reduced using the DDRTree algorithm (max_components = 2). Cell trajectories were reconstructed and visualized using the orderCells and plot_cell_trajectory functions, respectively. The plot_pseudotime_heatmap function was used to analyze dynamic genes (q < 10). -10 Clustering was performed on branch-specific genes (BEAM analysis, q < 10) using the plot_genes_branched_heatmap function. -10 Mapping is performed.

[0197] 9.2. Diffusion diagram analysis:

[0198] The Destiny (v3.12.0) software package was used to perform nonlinear dimensionality reduction and diffusion map analysis on monocyte subpopulations, and the diffusion component (DC) was calculated using the DiffusionMap function. The results of the diffusion component analysis were cross-validated with the results of the Monocle pseudotime analysis.

[0199] like Figure 12As shown in Figure A, the diffusion components DC1, DC2, and DC3 together illustrate the three-dimensional distribution of the terminal differentiation characteristic scores of each monocyte subset. As can be seen from the figure, Mono_CD14 and Mono_CD16 cells have lower terminal differentiation characteristic scores, placing them at the beginning of their differentiation trajectories; while TAMO_IL1B cells have the highest terminal differentiation characteristic scores, placing them at the end of their differentiation trajectories. Figure 12 B further validated this trend, showing that the terminal differentiation feature score gradually increases along a specific direction on the DC embedding map.

[0200] like Figure 12 As shown in Figure C, after staining by group (HD, OR, SOR), it can be seen that cells from the OR group occupy a specific area in the DC space and are partially separated from the HD and SOR groups, suggesting that the enrichment of TAMO_IL1B is group-specific. Figure 12 After staining by cell type, D showed that TAMO_IL1B cells formed independent clusters in the DC space, and their positions were far from Mono_CD14 and Mono_CD16 cells, suggesting that they were in the final stage of monocyte differentiation.

[0201] The above diffusion plot analysis results are consistent with the Monocle2 pseudo-time trajectory analysis results. Figure 7 The cross-validation of the two independent algorithms (AC) showed a high degree of consistency, confirming the differentiation trajectory of the mononuclear cell subset: starting from mono_CD14, passing through a transitional state, and finally differentiating into TAMO_IL1B, with TAMO_IL1B being enriched in the OR group.

[0202] 9.3 Construction of Differentiation Trajectories:

[0203] like Figure 7 As shown in Figure AB, the Monocle2 pseudo-time trajectory plots indicate that mono_CD14 cells are in the early pseudo-time phase, TAMO_IL1B is in the late pseudo-time phase, and TAMO_PADI4, TAMO_MX1, and macrophages are in a transitional state. The trajectory of the monocyte subsets originates from mono_CD14 cells in the HD group, subsequently differentiating into two trajectories (Trajectory I and Trajectory II), with Trajectory II further differentiating into two new trajectories (Trajectory III and Trajectory IV).

[0204] like Figure 7 As shown in Figure C, most mono_CD14 and mono_CD16 cells (trajectory I) originated from the HD group, TAMO_IL1B cells were located at the end of trajectory III and were only present in the OR group, while macrophages were enriched at the end of trajectory IV. This result is consistent with the results of diffusion map analysis ( Figure 12 CD) consistent.

[0205] 9.4. Differentially expressed genes with pseudo-time correlation:

[0206] The plot_pseudotime_heatmap function of the Monocle package was used to identify genes that showed significant changes over pseudotime. Dynamic genes were clustered according to their expression patterns into three gene clusters: early expression, mid-term expression, and late expression.

[0207] The results are as follows Figure 7 As shown in the D heatmap, clustering of all genes that showed significant covariance with pseudotime (FDR < 1%) identified early, mid / mid-late, and late-term expressed genomes.

[0208] like Figure 7 As shown in E, molecules associated with a good response in the OR group, such as B4GALT1, were observed to be highly expressed in the late stage, while COMMD8, HCLS1, and SAMHD1 were gradually lost in the late stage.

[0209] like Figure 7 The F-volcano plot shows the differentially expressed genes in monocyte subsets between the OR and SOR groups. Figure 7 As shown in G, B4GALT1 was highly expressed in the OR group, while C3AR1, HCLS1, COMMD8, NAGA, SAMHD1 and ZBTB16 were lowly expressed (t-test).

[0210] Example 10: Validation of prognostic gene markers

[0211] 10.1 Survival analysis of the CoMMpass dataset:

[0212] To validate the prognostic value of TAMO_IL1B-related characteristic genes, survival analysis was performed using an independent large multiple myeloma cohort (MMRF CoMMpass, n=837):

[0213] Download RNA-seq expression data and clinical survival data from the MMRF CoMMpass dataset using the Xena browser. Normalize the raw RNA-seq counts. For each target gene (B4GALT1, C3AR1, HCLS1, COMMD8, NAGA, SAMHD1, ZBTB16), use the median expression level of that gene in the patient population as the cutoff value to divide patients into high-expression and low-expression groups. Plot overall survival curves using the Kaplan-Meier method and compare the differences in overall survival between the two groups using the log-rank test.

[0214] The results are as follows Figure 7As shown in Figure H, MM patients with high B4GALT1 expression (>median, n=418) had significantly longer overall survival than those with low B4GALT1 expression (<median, n=419) (P<0.033). Conversely, high expression of C3AR1, HCLS1, COMMD8, NAGA, SAMHD1, and ZBTB16 was significantly associated with shorter overall survival.

[0215] 10.2 Deconvolution analysis of monocyte-specific expression:

[0216] To clarify whether the prognostic value of TAMO_IL1B-related genes stems from expression in monocytes or malignant cells, a computational deconvolution method was employed. Deconvolution analysis was performed using bulk RNA-seq data from bone marrow samples in the GSE136324 dataset. A reference expression matrix containing characteristic genes for each cell type was constructed, and deconvolution tools were used to infer the cell type-specific expression level of each target gene in each cell type.

[0217] like Figure 13 A- Figure 13 As shown in G, Kaplan-Meier survival analysis was performed based on the inferred monocyte-specific expression levels. Patients were divided into high-expression and low-expression groups using the median monocyte-specific expression level of each gene as a threshold. The results showed that when the expression of these seven genes was localized to monocytes, none of them showed a statistically significant association with overall survival (all p-values ​​> 0.05). This result indicates that the prognostic value of these genes does not derive from their expression in monocytes.

[0218] 10.3 Univariate Cox regression analysis of purified MM cells:

[0219] To identify the prognostic value of genes derived from monocyte subsets in purified MM cells, the prognostic value of these genes was analyzed in purified MM cells from the CoMMpass cohort.

[0220] like Figure 13 As shown in the H-forest plot, univariate Cox regression analysis revealed that 6 out of 7 genes were significantly associated with overall survival. Among them, COMMD8 (HR=1.667, p=3×10⁻⁶) was significantly associated with overall survival. -9 ), HCLS1 (HR=1.690, p=1.69×10 -6 ) and NAGA (HR=1.480, p=9.44×10) -7 B4GALT1 was the strongest predictor of poor prognosis. Only B4GALT1 did not show a significant association (p=0.181). Detailed hazard ratios (HR), 95% confidence intervals (CI), and p-values ​​for each gene are summarized in Table 3 below.

[0221] Table 3. Univariate Cox regression analysis based on the prognostic value of genes derived from monocyte subsets of purified MM cells in the CoMMPASS cohort.

[0222]

[0223] comprehensive Figure 13 A- Figure 13 G and Figure 13 The results of H indicate that the expression of the above 7 genes in purified MM cells (malignant plasma cells) has significant clinical prognostic value. Figure 13 H), but its expression in monocytes has no prognostic value ( Figure 13 (AG). This finding suggests that these genes may represent "transcriptional imprints" left by the interaction between MM cells and monocytes; that is, after monocytes interact with MM cells through the THBS1-SDC1 axis, they induce prognostic-related gene expression changes in MM cells. The above results validate the significant clinical prognostic value of seven genes derived from monocytes in purified MM cells and indicate that COMMD8, HCLS1, and NAGA are particularly promising biomarkers.

[0224] Example 11: Validation at the proteomic, transcriptomic, and reported literature levels

[0225] 11.1 Analysis of Public Proteomics Datasets:

[0226] To validate the expression of THBS1 and SDC1 at the protein level, protein expression data from the public proteomics database (The Human Protein Atlas, HPA) were analyzed.

[0227] like Figure 10 As shown in Figure A, specific expression signals of THBS1 protein were detected in the THP-1 cell line (human monocyte cell line), verifying the protein-level expression of THBS1 in the monocyte / macrophage lineage. Figure 10 As shown in Figure B, a specific expression signal of SDC1 protein can be detected in the LP-1 cell line (human multiple myeloma cell line), verifying the protein level expression of SDC1 in myeloma cells.

[0228] like Figure 10 As shown in Figure C, the detection of THBS1 and SDC1 proteins in independent multiple myeloma proteogenomic resources was further analyzed. Figure 10Figure C ranks proteins by their median raw intensity detected in at least one myeloma sample, with color intensity representing the percentage of valid detections for that protein in the myeloma dataset. As shown in the figure, SDC1 and THBS1 were both effectively detected in multiple myeloma samples and ranked highly in intensity, further supporting the detectability and clinical operability of THBS1 and SDC1 as multiple myeloma-related proteins.

[0229] The analysis results of the above public proteomics data are consistent with the findings of the single-cell transcriptome sequencing of this invention (i.e., THBS1 is expressed in TAMO_IL1B and SDC1 is expressed in malignant plasma cells) at the protein level.

[0230] 11.2 Validation of Independent Transcriptome Datasets:

[0231] The transcriptomic findings were further validated by analysis of the published MM dataset (GSE124310). This dataset details the expression of THBS1 in patient-derived monocytes, consistent with the observations of this invention.

[0232] 11.3. Reported documentary evidence:

[0233] The study by Omatsu et al. in colorectal cancer (Nat Commun 14, 5534, 2023) clearly demonstrated through protein-level co-staining and genetic models that bone marrow-derived monocyte-like cells are the main source of THBS1 in the tumor microenvironment. The high conservation of THBS1 cellular origin and immunomodulatory function between colorectal cancer and multiple myeloma (MM) suggests a potential universal tumor mechanism.

[0234] Example 12: Preparation and Use of the Reagent Kit

[0235] 12.1 Flow cytometry detection kit:

[0236] A kit for predicting the response of multiple myeloma patients to bortezomib-based treatment regimens, comprising the following components:

[0237] (1) Fluorescently labeled anti-CD14 antibody (such as FITC-anti-CD14).

[0238] (2) Fluorescently labeled anti-IL1B antibodies (such as PE-anti-IL1B).

[0239] (3) Fluorescently labeled anti-THBS1 antibodies (such as APC-anti-THBS1).

[0240] (4) Fluorescently labeled anti-SDC1 antibodies (such as PerCP-anti-SDC1).

[0241] (5) Red blood cell lysis buffer;

[0242] (6) Cell fixative;

[0243] (7) Flow cytometry with gate control antibodies;

[0244] (8) Isotype control antibody;

[0245] (9) Instruction manual.

[0246] How to use:

[0247] (1) Obtain bone marrow aspiration samples from patients with multiple myeloma and place them in collection tubes containing anticoagulant;

[0248] (2) Take an appropriate amount of bone marrow sample, add red blood cell lysis buffer, incubate at room temperature for 10-15 minutes, centrifuge and discard the supernatant;

[0249] (3) Resuspend the cells in PBS and wash twice;

[0250] (4) Add fluorescently labeled anti-CD14 antibody and anti-IL1B antibody, and incubate at 4°C in the dark for 30 minutes;

[0251] (5) After washing with PBS, resuspend the cells in cell fixation solution;

[0252] (6) Flow cytometry analysis was performed, with FSC / SSC gating to exclude cell debris and dead cells, and CD14 analysis was performed. + Screening for mononuclear cell populations using phylogenetic methods, specifically CD14. + IL1B analysis in the population + Cell ratio;

[0253] (7) Calculate CD14 + IL1B + Cells account for CD14 + Percentage of the total number of monocytes;

[0254] (8) Compare the measured percentage with a predetermined threshold; the predetermined threshold is determined by the percentage of CD14 in the suboptimal responder population. + IL1B + The distribution of tumor-associated monocyte abundance was determined by statistical analysis;

[0255] (9) When the percentage is higher than the reference level, it is predicted that the patient will respond well to bortezomib-based treatment regimens.

[0256] (10) Optionally, the same sample was surface-stained with fluorescently labeled anti-THBS1 antibody and anti-SDC1 antibody, and the expression levels of THBS1 in TAMO_IL1B and SDC1 in malignant plasma cells were analyzed by flow cytometry as an auxiliary indicator for assessing the strength of intercellular communication between TAMO_IL1B and malignant plasma cells.

[0257] 12.2 Single-cell transcriptome sequencing detection kit:

[0258] A single-cell transcriptome sequencing kit for predicting the response of multiple myeloma patients to bortezomib-based treatment regimens, comprising the following components:

[0259] (1) Single-cell capture chip or microfluidic device;

[0260] (2) Cell lysis buffer;

[0261] (3) Reverse transcription reaction mixture;

[0262] (4) cDNA amplification reaction mixture;

[0263] (5) Library construction reagents;

[0264] (6) Specific primers or probes for CD14, IL1B, THBS1 and SDC1 genes;

[0265] (7) Instruction manual.

[0266] How to use:

[0267] (1) Obtain bone marrow aspiration samples from patients with multiple myeloma and prepare single-cell suspensions;

[0268] (2) Single-cell capture and library construction were performed using the 10x Genomics platform or the Drop-seq platform;

[0269] (3) Perform quality control and data normalization on sequencing data;

[0270] (4) Cell clustering was performed based on the expression levels of CD14, IL1B, THBS1, and SDC1 genes to identify CD14. + IL1B + Cell subsets and malignant plasma cell subsets (TPC);

[0271] (5) Calculate CD14 + IL1B + Cells account for CD14 + Percentage of the total number of monocytes;

[0272] (6) Compare the percentages to a reference level to predict patient response to bortezomib-based treatment regimens.

[0273] 12.3 Real-time quantitative PCR detection kit:

[0274] A real-time quantitative PCR assay kit for assessing the prognosis of patients with multiple myeloma, comprising the following components:

[0275] (1) B4GALT1 gene-specific primer pair;

[0276] (2) C3AR1 gene-specific primer pair;

[0277] (3) HCLS1 gene-specific primer pair;

[0278] (4) COMMD8 gene-specific primer pair;

[0279] (5) NAGA gene-specific primer pairs;

[0280] (6) SAMHD1 gene-specific primer pair;

[0281] (7) ZBTB16 gene-specific primer pair;

[0282] (8) Specific primer pairs for internal reference genes (GAPDH or ACTB);

[0283] (9) Reverse transcriptase, DNA polymerase, dNTP mixture and PCR buffer;

[0284] (10) Instruction manual.

[0285] How to use:

[0286] (1) Extract total RNA from bone marrow or peripheral blood samples from patients with multiple myeloma;

[0287] (2) cDNA was synthesized using a reverse transcription kit;

[0288] (3) Using cDNA as a template, perform real-time quantitative PCR amplification using gene-specific primer pairs;

[0289] (4) Using GAPDH or ACTB as internal reference genes, the relative expression levels of each gene were calculated using the Livak method;

[0290] (5) Compare the expression levels of each gene with the median expression value in the multiple myeloma patient population;

[0291] (6) Prognostic assessment criteria: B4GALT1 expression level above the median indicates that the patient has a longer overall survival; C3AR1, HCLS1, COMMD8, NAGA, SAMHD1 or ZBTB16 expression level above the median indicates that the patient has a shorter overall survival.

[0292] Example 13: A method for detecting the strength of THBS1-SDC1 ligand-receptor interaction in vitro

[0293] 13.1 Detection methods based on single-cell transcriptome sequencing:

[0294] A method for detecting the strength of THBS1-SDC1 ligand-receptor interaction in isolated bone marrow samples in vitro, comprising the following steps:

[0295] (1) Obtaining ex vivo bone marrow samples;

[0296] (2) Single-cell transcriptome sequencing was performed on bone marrow samples to obtain the gene expression matrix of each cell;

[0297] (3) Identify TAMO_IL1B cell population based on CD14 and IL1B expression, and identify TPC cell population based on characteristic genes;

[0298] (4) Extract the expression levels of THBS1 in TAMO_IL1B and SDC1 in TPC;

[0299] (5) Calculate the communication probability of THBS1-SDC1 ligand-receptor interaction between TAMO_IL1B and TPC using the CellChat software package;

[0300] (6) The formula for calculating the communication probability is:

[0301] Communication probability = (THBS1 expression level in TAMO_IL1B × SDC1 expression level in TPC) / total expression level of ligand-receptor pair in all cell types;

[0302] (7) When the communication probability is ≥0.1, it indicates the presence of significant THBS1-SDC1 ligand-receptor interaction;

[0303] (8) Compare the measured communication probability with the reference level to assess the strength of the interaction.

[0304] 13.2 Validation method based on immunoprecipitation:

[0305] To further verify the physical interaction between THBS1 and SDC1, a co-immunoprecipitation (Co-IP) experiment can be used:

[0306] (1) Isolate monocytes and plasma cells from bone marrow samples, or co-culture them in vitro using monocyte cell lines expressing THBS1 (such as THP-1) and MM cell lines expressing SDC1 (such as LP-1);

[0307] (2) Collect cells and lyse them using RIPA lysis buffer containing protease inhibitors;

[0308] (3) Centrifuge and collect the supernatant. Use anti-THBS1 antibody for immunoprecipitation, with normal IgG as a negative control.

[0309] (4) After the precipitated complex was separated by SDS-PAGE, it was detected by Western blot using anti-SDC1 antibody;

[0310] (5) If the SDC1 protein signal is detected in the immunoprecipitation complex of anti-THBS1 antibody, it confirms that there is a physical binding between THBS1 and SDC1.

[0311] 13.3. Affinity determination method based on surface plasmon resonance:

[0312] The binding affinity between THBS1 and SDC1 can be determined using surface plasmon resonance (SPR) technology.

[0313] (1) The recombinant SDC1 protein was immobilized on the sensor chip;

[0314] (2) Recombinant THBS1 protein at different concentrations (e.g., 0.1 nM, 1 nM, 10 nM, 100 nM, 1 μM) is passed through the chip surface;

[0315] (3) Record the binding and dissociation curves, and calculate the binding constant (Ka) and dissociation constant (Kd);

[0316] (4) Calculate the equilibrium dissociation constant (KD) between THBS1 and SDC1 using Scatchard analysis.

[0317] Example 14: Drug Screening Method

[0318] 14.1 Drug screening based on THBS1-SDC1 interaction:

[0319] A method for screening candidate drugs that can enhance the sensitivity of multiple myeloma cells to bortezomib-based treatment regimens, comprising the following steps:

[0320] (1) Provide monocytes expressing THBS1 (such as THP-1 cell line or primary monocytes) and malignant plasma cells expressing SDC1 (such as MM.1S, RPMI-8226 or LP-1 multiple myeloma cell lines).

[0321] (2) Co-culture monocytes and malignant plasma cells in a certain ratio (e.g., 1:1 or 1:5);

[0322] (3) Add the candidate drug to the co-culture system at different concentrations (e.g., 0.1 μM, 1 μM, 10 μM);

[0323] (4) After incubation for 24-48 hours, the interaction strength between THBS1 and SDC1 can be detected using the following methods:

[0324] (4a) Detection of the amount of THBS1-SDC1 complex formed by co-immunoprecipitation (Co-IP);

[0325] (4b) Enzyme-linked immunosorbent assay (ELISA) was used to detect the binding amount of THBS1 and SDC1;

[0326] (4c) Biolayer interferometry (BLI) detects combined kinetic parameters;

[0327] (5) Simultaneously detect the activity of the TGF-β signaling pathway (e.g., detect the phosphorylation level of SMAD2 / 3).

[0328] (6) Select candidate drugs that can enhance THBS1-SDC1 interaction and activate the TGF-β signaling pathway;

[0329] (7) Further verification: MM cells were treated with the candidate drug in combination with bortezomib, and cell viability was detected by CCK-8 or MTT assay. Apoptosis rate was detected by Annexin V / PI double staining to assess whether the candidate drug enhances the sensitivity of MM cells to bortezomib.

[0330] 14.2 Drug screening based on increased TAMO_IL1B abundance:

[0331] A method for screening candidate drugs that can mimic the beneficial effects of TAMO_IL1B:

[0332] (1) Establish an in vitro co-culture system of TAMO_IL1B and TPC;

[0333] (2) The interaction strength of THBS1-SDC1 in the TAMO_IL1B and TPC co-culture group was used as a positive control;

[0334] (3) Add the candidate drug to the TPC single culture system;

[0335] (4) To detect whether candidate drugs can upregulate the expression of SDC1 in TPC or enhance its response to THBS1;

[0336] (5) Select candidate drugs that can simulate the co-culture effect of TAMO_IL1B.

[0337] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A biomarker for predicting the response of multiple myeloma to bortezomib-based treatment regimens, characterized in that, Including the detection of CD14 in ex vivo bone marrow samples + IL1B + Reagent for tumor-associated monocyte abundance; The CD14 + IL1B + Tumor-associated mononuclear cells are a subset of bone marrow-derived mononuclear cells that co-express CD14 and IL1B, identified by single-cell transcriptome sequencing or flow cytometry. The abundance is CD14. + IL1B + The percentage or absolute number of tumor-associated monocytes in the total monocyte population of an ex vivo bone marrow sample.

2. The biomarker composition according to claim 1, characterized in that, The reagents include: Antibodies or antigen-binding fragments thereof that specifically bind to the CD14 protein; Antibodies or antigen-binding fragments thereof that specifically bind to the IL1B protein; Each antibody independently carries a detectable marker, which includes, but is not limited to, fluorescent dyes, enzymes, biotin, radioisotopes, or quantum dots.

3. The biomarker composition according to claim 1, characterized in that: It also contains reagents for detecting THBS1 expression levels and / or reagents for detecting SDC1 expression levels.

4. A kit for predicting the response of multiple myeloma patients to bortezomib-based treatment regimens, characterized in that, Include: The first reagent capable of specifically binding to CD14; A second reagent capable of specifically binding to IL1B; A third reagent capable of specifically binding to THBS1; A fourth reagent capable of specifically binding to SDC1; Instructions for use, which include the detection of CD14 in ex vivo bone marrow samples using the first and second reagents. + IL1B + Operational guidelines for using tumor-associated mononuclear cell abundance to predict patient responsiveness to bortezomib-based treatment regimens; The instruction manual also includes operational guidelines for detecting the expression levels of THBS1 and SDC1 using the third and fourth reagents, respectively, wherein the expression level of THBS1 is derived from CD14. + IL1B + Tumor-associated mononuclear cells, wherein the expression level of SDC1 was derived from malignant plasma cells in the same sample, and the expression levels of THBS1 and SDC1 were used to assess CD14. + IL1B + THBS1-SDC1 communication inference score with tumor-associated monocytes as the sending cell population and malignant plasma cells as the receiving cell population; The communication inference score is calculated from the expression levels of THBS1 and SDC1 according to a predetermined cell communication analysis rule. The positive correlation between the communication inference score and treatment responsiveness is established in advance under a determined treatment regimen, sampling time point, and efficacy evaluation criteria.

5. The reagent kit according to claim 4, characterized in that: The first reagent is an anti-CD14 antibody with a detectable label, and the second reagent is an anti-IL1B antibody with a detectable label; The kit also includes auxiliary reagents for flow cytometry detection, which are selected from one or more of erythrocyte lysis buffer, cell fixative, flow cytometry gating control antibody, and isotype control antibody.

6. The reagent kit according to claim 4, characterized in that: The instruction manual outlines the following prediction criteria: When CD14 in isolated bone marrow samples + IL1B + When the abundance of tumor-associated mononuclear cells is above a predetermined threshold, it indicates that the patient has a good response to bortezomib-based treatment regimens; The predetermined threshold is determined by CD14 in the suboptimal responder population. + IL1B + The distribution of tumor-associated mononuclear cell abundance was determined by statistical analysis.

7. A system for predicting the responsiveness of multiple myeloma patients to bortezomib-based treatment regimens, characterized in that, include: A sample processing unit for processing isolated bone marrow samples to prepare single-cell suspensions; Detection unit for detecting CD14 in isolated bone marrow samples. + IL1B + Abundance of tumor-associated monocytes; and detection of THBS1 and SDC1 expression levels; The storage unit stores cell recognition rules, reference cell population definitions, and abundance thresholds corresponding to specific treatment plans and sampling times. The analysis unit is used to compare the abundance measured by the detection unit with the reference level, and to calculate CD14 based on the expression levels of THBS1 and SDC1. + IL1B + The THBS1-SDC1 communication inference score, with tumor-associated mononuclear cells as the sending cell population and malignant plasma cells as the receiving cell population, is used to generate predictive information based on the comparison and calculation results. Output unit for displaying or reporting the patient's predicted response to bortezomib-based treatment regimens. 8.CD14 + IL1B + The use of tumor-associated monocytes as a biomarker in the preparation of a kit for predicting the response of multiple myeloma patients to bortezomib-based treatment regimens, characterized by: The kit contains reagents for detecting CD14 expression levels and reagents for detecting IL1B expression levels; The kit also contains reagents for detecting THBS1 expression levels and / or reagents for detecting SDC1 expression levels.

9. A method for detecting the strength of THBS1-SDC1 ligand-receptor interaction in bone marrow samples in vitro, characterized in that, Includes the following steps: Step S1: Obtain an ex vivo bone marrow sample; Step S2: Detect CD14 in bone marrow samples in vitro. + IL1B + Expression level of THBS1 in tumor-associated monocytes; Step S3: Detect the expression level of SDC1 in malignant plasma cells in bone marrow samples in vitro; Step S4: Calculate the strength value of the THBS1-SDC1 ligand-receptor interaction based on the detection results of Step S2 and Step S3.

10. A set of genetic markers for assessing the prognosis of patients with multiple myeloma, characterized in that, The combination of genetic markers includes one or more genes selected from the following: B4GALT1, C3AR1, HCLS1, COMMD8, NAGA, SAMHD1, and ZBTB16. The combination of genetic markers is used to assess the prognosis of patients with multiple myeloma by detecting the expression levels of the genes in ex vivo bone marrow samples.

Images