A marker for detecting prostate cancer and its use in detecting prostate cancer

Abstract

The application discloses a marker for detecting prostate cancer and application of the marker in detection of prostate cancer. The marker is FST antigen or an autoantibody combined with the FST antigen. The prostate cancer is detected by detecting the FST level in serum, AR non-dependent disease progression can be accurately recognized even when the PSA level is reduced or normal, a clinical trap that PSA reduction is misjudged as "treatment effective" is avoided, and different tumor subtypes can be distinguished in combination diagnosis of PSA.

Description

Technical Field

[0001] This invention belongs to the field of biology, specifically relating to a biomarker for detecting prostate cancer and its application in the detection of prostate cancer. Background Technology

[0002] The emergence of resistance to androgen receptor (AR) targeted therapy is one of the most pressing clinical challenges in prostate cancer treatment. Despite the significant efficacy of newer generation anti-AR drugs such as enzalutamide and apalutamide in initial treatment, almost all patients with advanced prostate cancer eventually develop resistance to these drugs. This resistance often manifests as the development of an AR-independent prostate cancer phenotype, fundamentally altering the biological characteristics of the disease. This shift from AR-dependent to AR-independent growth represents a crucial evolutionary step, rendering standard hormone therapy ineffective and resulting in limited treatment options and significantly worse prognosis for patients.

[0003] AR-independent prostate cancer can arise through various mechanisms, one important pathway being lineage plasticity, where prostate adenocarcinoma cells can transdifferentiate into alternative cell lineages that no longer depend on androgen signaling for survival and proliferation. This process can generate several AR-independent variants, including neuroendocrine prostate cancer (NEPC) and double-negative prostate cancer (DNPC). DNPC is characterized by the simultaneous loss of both AR and neuroendocrine markers, representing a particularly aggressive form of treatment-resistant disease, exhibiting enhanced stemness, enhanced metastatic potential, and being associated with the worst clinical prognosis. Although molecular lineage analysis has increasingly clarified the characteristics of DNPC in recent years, the underlying molecular drivers of DNPC formation and maintenance remain poorly understood. This knowledge gap hinders the development of mechanism-based therapeutics and non-invasive biomarkers for identifying newly emerging resistant patients.

[0004] Prostate-specific antigen (PSA, encoded by the KLK3 gene) is currently the most widely used serological biomarker in the clinical management of prostate cancer. Its core principle is that normal prostate epithelial cells express PSA and secrete it into semen under the drive of the androgen receptor (AR) signaling pathway. When the prostate becomes cancerous, the glandular structure is damaged, causing PSA to leak into the bloodstream, thus increasing serum PSA levels. Clinically, a screening threshold of 4 ng / mL is typically used for early prostate cancer screening, post-treatment efficacy monitoring, and recurrence early warning. The widespread use of PSA testing has indeed promoted the early detection of prostate cancer, especially showing good detection sensitivity for adenocarcinomas with active AR signaling pathways (i.e., ARPC). It is mainly used for screening, efficacy monitoring, and recurrence early warning. PSA expression is directly regulated by the transcriptional regulation of the androgen receptor (AR) signaling pathway; therefore, PSA is essentially an indirect readout indicator of AR pathway activity.

[0005] However, as prostate cancer undergoes spectrum plasticity shifts under the pressure of AR-targeted therapies (such as enzalutamide, apalutamide, and abiraterone), a significant proportion of patients develop an AR-independent phenotype, primarily including neuroendocrine prostate cancer (NEPC) and double-negative prostate cancer (DNPC). DNPC simultaneously loses both AR and neuroendocrine markers and is the subtype with the worst prognosis. Due to the loss of the AR signaling pathway, these tumors no longer produce PSA or produce only extremely low levels of PSA, resulting in a fundamental blind spot in PSA testing for these patients. Specifically: (1) The decrease in PSA level is misinterpreted as "treatment effectiveness": In patients receiving AR-targeted therapy, the decrease in PSA may not reflect tumor shrinkage, but rather that AR-positive clones are suppressed while AR-negative (FST-high expression) clones are selectively proliferating; (2) Missed detection of DNPC and mixed tumors: When ARPC and DNPC components are present in the tumor at the same time, the PSA signal contributed by ARPC components may mask the presence and amplification of DNPC components, causing serious bias in disease assessment. (3) Lack of early warning capability: There are currently no biomarkers available for serum testing that can provide early warning to clinicians in the early stages of AR independence transition.

[0006] (4) PSA is a direct transcriptional target of the AR signaling pathway. PSA expression depends entirely on the transcriptional activation function of AR. When tumor cells lose AR expression (such as DNPC), the transcriptional driving force of PSA disappears fundamentally, and serum PSA levels decrease or even drop to the normal range. This means that PSA detection has a fundamental "blind spot" for AR-negative tumor subtypes—not a problem of detection sensitivity, but because it cannot be detected by biological mechanisms.

[0007] (5) For mixed tumors (i.e., the presence of both ARPC and DNPC components in the same patient), PSA testing is more misleading. The PSA signal contributed by ARPC components may mask the presence of DNPC components. Clinically, a decrease in PSA in patients receiving AR-targeted therapies such as enzalutamide is traditionally interpreted as a sign of treatment effectiveness, but in reality, it may indicate that while AR-positive clones are being suppressed, AR-negative FST-overexpressing clones are selectively amplifying. If these patients rely solely on PSA monitoring, they will completely miss the window of opportunity for early intervention.

[0008] The bone morphogenetic protein (BMP) signaling pathway, a key branch of the TGF-β superfamily, plays a profound regulatory role in prostate cancer development. For example, BMP6 expression is higher in prostate cancer than in benign tissue and is associated with high-grade disease and tumor progression. Pten In the missing mouse model Bmpr2The deletion of certain genes reveals that BMP signaling drives prostate cancer progression and promotes local inflammation in early-stage disease. Conversely, other BMP ligands exhibit tumor-suppressive functions. BMP10 inhibits the growth and invasiveness of prostate cancer cells by inducing apoptosis. BMP7 induces senescence in prostate cancer stem cell-like cells, although its anti-apoptotic effects have also been reported. Furthermore, it has been reported that reduced BMP receptor expression occurs in more aggressive prostate cancers, increasing the complexity of BMP signaling regulation in the cancer environment. These conflicting findings suggest that BMP signaling may have stage-specific or environment-dependent functions, particularly regarding treatment resistance and lineage plasticity, which remain not fully elucidated. Summary of the Invention

[0009] To address the inherent limitations of existing PSA testing in detecting AR-independent prostate cancer and its inability to provide early warning of AR-independent transitions, this invention provides a biomarker for prostate cancer and the application of its FST protein in the detection of AR-independent prostate cancer. This invention overcomes the inherent deficiencies of traditional PSA testing, enabling early diagnosis and efficacy monitoring of AR-independent prostate cancer, and providing a new technical approach for the clinical management of prostate cancer.

[0010] To solve the above-mentioned technical problems, the present invention adopts the following technical means: The first aspect of the present invention provides a biomarker for detecting prostate cancer, said biomarker being the FST antigen or an autoantibody that binds to it.

[0011] A second aspect of the present invention provides a combination of biomarkers for detecting prostate cancer, the biomarker combination comprising the biomarkers described in the first aspect of the present invention.

[0012] In some preferred embodiments of the present invention, the FST protein is used to detect AR-independent prostate cancer or mixed prostate cancer containing AR-independent subtypes.

[0013] In some preferred embodiments of the present invention, the prostate cancer is neuroendocrine prostate cancer or double-negative prostate cancer.

[0014] In some preferred embodiments of the present invention, the biomarker combination further comprises any one or more of the following biomarkers: selected from FST antigen or autoantibodies bound thereto, AR antigen or autoantibodies bound thereto, PSA / KLK3 antigen or autoantibodies bound thereto, SOX9 antigen or autoantibodies bound thereto, and PSA antigen or autoantibodies bound thereto.

[0015] In some further preferred embodiments of the invention, the biomarker combination comprises an FST antigen or an autoantibody binding thereto and a PSA antigen or an autoantibody binding thereto; or, The biomarker combination includes FST antigen or its binding autoantibody, AR antigen or its binding autoantibody, PSA / KLK3 antigen or its binding autoantibody, SOX9 antigen or its binding autoantibody, and PSA antigen or its binding autoantibody.

[0016] A third aspect of the present invention provides a kit for detecting AR-independent prostate cancer, the kit comprising reagents for detecting biomarkers as described in the first aspect of the present invention or in combinations of biomarkers as described in the second aspect of the present invention; for example, the reagents are autoantibodies for antigen detection or antigens for autoantibody detection.

[0017] In some preferred embodiments of the present invention, the antigen or autoantibody in the biomarker combination further contains a tag peptide; the tag peptide preferably includes one or more of the following: His tag, streptavidin tag, avidin tag, biotin tag, GST tag, C-myc tag, Flag tag, and HA tag; the biomarker is more preferably expressed by Escherichia coli, yeast, insect cells, or animal cells, and / or the biomarker is purified by Ni affinity chromatography, ion exchange chromatography, molecular sieve, dialysis, ultrafiltration, or hydrophobic chromatography.

[0018] In some preferred embodiments of the present invention, the kit further includes one or more of the following: sample diluent, calibrator diluent, washing solution, analytical buffer, anti-human IgG secondary antibody, calibrator, and quality control product; the kit preferably also includes a 96-well plate.

[0019] A fourth aspect of the present invention provides a detection method, wherein the detection method uses a biomarker as described in the first aspect of the present invention, a combination of biomarkers as described in the second aspect of the present invention, or a kit as described in the third aspect of the present invention to detect a corresponding antigen or an autoantibody binding thereto or an mRNA encoding thereto in a sample. The specific steps include: contacting the biomarker with the sample; if binding is detected, it indicates the presence of the corresponding antigen or an autoantibody binding thereto; the detection method is selected from any one or more of ELISA detection, multiplex immunohistochemistry detection, chemiluminescent immunoassay, electrochemiluminescent immunoassay, lateral flow immunochromatography, RT-qPCR or liquid biopsy, and mass spectrometry-based proteomics methods.

[0020] In some preferred embodiments of the present invention, the detection method is ELISA detection, the sample is serum, and the interpretation criteria include: FST levels <5.5 ng / mL, 5.5 - 7.0 ng / mL and >7.0 ng / mL are interpreted as low level, medium level and high level, respectively; and / or, PSA levels <4.0 ng / mL, 4.0 - 9.0 ng / mL and >9 ng / mL are interpreted as low level, medium level and high level, respectively; Alternatively, the detection method may be multiplex immunohistochemical detection, the sample may be a tissue section, and the interpretation criteria may include: FST being negative, positive, or focally positive; and / or AR being negative, strongly positive, or focally positive; and / or PSA being negative / weakly positive, strongly positive, or heterogeneous.

[0021] In some more preferred embodiments of the present invention, the detection method is for non-diagnostic purposes or for diagnostic purposes.

[0022] In some further preferred embodiments of the present invention, the diagnostic purpose is to diagnose prostate cancer subtypes, and the diagnostic criteria include: in the ELISA test, low FST levels indicate low risk, predominantly AR-dependent; medium FST levels indicate intermediate risk / gray zone, with a tendency for early AR-independent transformation; high FST levels indicate high risk, suggesting AR-independent / DNPC progression; or, low FST levels + high PSA levels indicate ARPC-predominant, AR-dependent prostate cancer; high FST levels + low or medium PSA levels indicate a high probability of transitional / DNPC, with caution regarding AR-independent progression; high FST levels + high PSA levels indicate mixed / transitional type, possibly with mixed ARPC and DNPC prostate cancer; or... In the multiplex immunohistochemical assay, the following are indicated for AR-dependent prostate cancer: FST negative + strong AR positive, FST negative + strong PSA positive; the following are indicated for AR-independent prostate cancer: FST positive + AR negative, FST positive + PSA negative / weak positive; the following are indicated for mixed prostate cancer: FST focal positive + AR focal positive, FST focal positive + heterogeneous PSA.

[0023] The fifth aspect of the present invention provides the use of biomarkers as described in the first aspect of the present invention, combinations of biomarkers as described in the second aspect of the present invention, or kits as described in the third aspect of the present invention in the preparation of reagents for detecting prostate cancer.

[0024] In some preferred embodiments of the present invention, the reagent is used to diagnose or assist in the diagnosis of AR-independent prostate cancer or mixed prostate cancer containing AR-independent subtypes, or for recurrence monitoring and / or prognostic monitoring after treatment for AR-independent prostate cancer; for example, monitoring the efficacy of enzalutamide treatment and early warning of drug resistance.

[0025] In some more preferred embodiments of the invention, the prostate cancer further includes AR-dependent prostate cancer.

[0026] A sixth aspect of the present invention provides a prostate cancer diagnostic system, the prostate diagnostic system comprising the following modules: (1) An input module, which is used to input the detection results of the markers in the sample to be tested as described in the first aspect of the present invention or in the combination of markers as described in the second aspect of the present invention; (2) An interpretation module interprets the detection result using the interpretation criteria in the detection method as described in the fourth aspect of the present invention; (3) Prostate cancer subtype indication module, indicating prostate cancer subtype according to the diagnostic criteria in the detection method as described in the fourth aspect of the present invention.

[0027] In some preferred embodiments of the present invention, the prostate cancer diagnostic system further includes a recommendation module that provides recommendations for disease monitoring and / or treatment based on the prostate cancer subtype indicated in (3); for example, in the ELISA test, a low level of FST indicates low risk, and routine follow-up is recommended; a medium level of FST indicates intermediate risk / gray zone, and a shorter follow-up interval is recommended, combined with a comprehensive interpretation of PSA trends. If a downward trend in PSA occurs during the same period, high vigilance is required, and early combination with docetaxel chemotherapy should be considered; a high level of FST indicates high risk, and a tissue biopsy is recommended for confirmation and assessment of whether the treatment strategy needs to be adjusted; or, a low level of FST + a high level of PSA indicates ARPC as the predominant risk, and standard AR targeted therapy is recommended. Routine PSA monitoring; high FST levels + low or intermediate PSA levels suggest a high probability of transitional / DNPC, recommending tissue biopsy. AR-targeted therapy may have failed or is about to fail, suggesting a chemotherapy-based regimen and assessing the possibility of combining immune checkpoint therapy with FST inhibitors; high FST levels + high PSA levels suggest a mixed / transitional type, suggesting that relying solely on PSA monitoring is unreliable and close monitoring of FST dynamics is necessary; or, if the multiple immunohistochemical tests indicate AR-dependent prostate cancer, AR-targeted therapy is recommended; if they indicate AR-independent prostate cancer, AR-targeted therapy may be ineffective; if they indicate mixed prostate cancer, the presence of coexisting ARPC and DNPC areas is suggested.

[0028] In some preferred embodiments of the present invention, the prostate diagnostic system further includes a login module and / or a printing module; the login module requires input of a username and password, and the printing module can print the results generated by the input module, the interpretation module, the prostate cancer subtype prompting module, and the suggestion module.

[0029] A seventh aspect of the present invention provides a computer-readable medium storing a computer program that, when executed by a processor, can perform the functions of the prostate cancer diagnostic system as described in the sixth aspect of the present invention.

[0030] An eighth aspect of the present invention provides a prostate cancer diagnostic device, comprising: (1) The computer-readable medium as described in the seventh aspect of the present invention; (2) A processor for executing computer programs to implement the functions of the prostate cancer diagnostic system.

[0031] Based on common knowledge in the field, the above-mentioned preferred conditions can be combined arbitrarily to obtain various preferred embodiments of the present invention.

[0032] The reagents and raw materials used in this invention are all commercially available.

[0033] Based on common knowledge in the field, the above-mentioned preferred conditions can be combined arbitrarily to obtain various preferred embodiments of the present invention.

[0034] The reagents and raw materials used in this invention are all commercially available.

[0035] The positive and progressive effects of this invention are as follows: 1. Overcoming the fundamental blind spot in PSA testing: This invention is the first to discover FST as a biomarker for AR-independent prostate cancer, solving the technical problem that traditional PSA testing cannot detect AR-negative tumor subtypes such as DNPC due to the loss of the AR signaling pathway. FST expression is independent of the AR pathway, and can accurately identify the progression of AR-independent disease even when PSA levels are decreased or even normal, avoiding the clinical trap of misjudging a decrease in PSA as "treatment effectiveness".

[0036] 2. Early Warning of AR Independence Transition: The FST level stratification criteria established in this invention (<5.5 ng / mL for low levels, 5.5-7.0 ng / mL for intermediate levels, and >7.0 ng / mL for high levels) can provide early warning signals at the early stage of AR independence transition. When patients show elevated FST levels, it indicates a possible early tendency for AR independence transition, providing a valuable window for clinical intervention, which is significantly better than the limitation of existing technologies that can only detect it in the late stage of the disease.

[0037] 3. Establishment of a combined FST and PSA diagnostic model: This invention achieves precise assessment of prostate cancer heterogeneity through the combined interpretation of FST and PSA. Low FST level + high PSA level suggests typical AR-dependent prostate cancer; high FST level + low / moderate PSA level suggests high suspicion of AR-independent progression; high FST level + high PSA level suggests mixed tumors. This combined model can accurately distinguish different prostate cancer subtypes, guiding the individualized selection of clinical treatment plans.

[0038] 4. Applicable to efficacy monitoring and drug resistance early warning: The FST detection of this invention can be used for efficacy monitoring and drug resistance early warning of AR targeted therapies such as enzalutamide. Clinical data show that high FST expression is associated with a significant shortening of survival in patients treated with enzalutamide. By dynamically monitoring changes in FST levels, the occurrence of AR targeted therapy resistance can be detected in a timely manner, guiding clinicians to adjust treatment strategies promptly.

[0039] 5. Non-invasive testing facilitates clinical application: FST is a secreted protein that can be detected through serum samples. It offers advantages such as being non-invasive, easy to operate, and repeatable, making it convenient for routine clinical use. The detection kit and diagnostic system provided by this invention can be directly applied in clinical laboratories, demonstrating promising industrialization prospects and clinical application value. Attached Figure Description

[0040] Figures 1-14 FST expression is elevated in DNPC.

[0041] Figure 1 Heatmap of BMP / TGFβ pathway activity and expression of specified genes in the CPGEA-2020 prostate cancer cohort (ARPC, n=84; NEPC, n=23; DNPC, n=27).

[0042] Figure 2 Heatmap of BMP / TGFβ pathway activity and expression of specified genes in the SU2C-2015 prostate cancer cohort (ARPC, n=78; NEPC, n=25; DNPC, n=15).

[0043] Figure 3 and Figure 4 For CPGEA-2020 ( Figure 3 ) and SU2C-2015 ( Figure 4 Bar charts (mean ± SD) showing BMP / TGFβ pathway activity and specified gene expression in ARPC, NEPC, and DNPC subtypes in the cohort.

[0044] Figure 5Forest plots were used to depict the prognostic value of FST in prostate cancer patients in the SU2C-2019 cohort, stratified by treatment type: all patients (n=81), abiraterone (Abi) treatment (n=58), and enzalutamide (Enza) treatment (n=22), based on Cox regression analysis.

[0045] Figure 6 For prostate cancer cell lines in the Encyclopedia of Cancer Cell Lines (32) AR , FST and key BMP Receptor mRNA expression level (measured in RPKM).

[0046] Figure 7 LNCaP cells were treated with enzalutamide (10 μM) for a specified time. BMPR1A , BMPR1B , BMPR2 and FST Relative mRNA expression levels (bar chart, mean ± SD, n=3).

[0047] Figure 8 LNCaP cells were treated with DHT (100 nM) for a specified time. BMPR1A , BMPR1B , BMPR2 and FST Relative mRNA expression levels (bar chart, mean ± SD, n=3).

[0048] Figure 9 LNCaP cells were treated with LDN-193189 (100 nM) for a specified time. AR The relative mRNA expression levels of their target genes (bar chart, mean ± SD, n=3).

[0049] Figure 10 A scatter plot depicting 26,686 tumor cells from 11 prostate cancer patients. FST and AR / KLK3 The expression of [unclear]. Mutual exclusion was assessed by contingency tables based on dichotomous expression (positive was defined as UMI count > 0), expressed by odds ratio and Fisher's exact test p-value.

[0050] Figure 11 Representative hematoxylin-eosin (H&E) and multiplex immunohistochemical (mIHC) staining of clinical prostate cancer tissue sections collected by Changhai Hospital (scale bar, 50 μm).

[0051] Figure 12Serum FST levels in CPGEA-2020 patients classified as DNPC (n=7) and ARPC (n=10) were measured by ELISA (left). Paired comparisons of FST and PSA concentrations in the cohort (n=17) were performed using a paired two-tailed Student's t-test (right).

[0052] Figure 13 Serum samples collected from patients at Changhai Hospital were analyzed for FST levels stratified by PSA levels (<7 ng / mL, n=47; >7 ng / mL, n=55) (left). Paired comparisons of FST and PSA concentrations in the cohort (n=102) were performed using a paired two-tailed Student's t-test (right).

[0053] Figure 14 A graph showing the accuracy and specificity of FST as a biomarker for prediction.

[0054] Figures 15-29 FST inhibits BMP11-mediated AR expression.

[0055] Figure 15 The top 10 enriched Gene Ontology-Bioprocess (GO-BP) entries in the comparison between FST-overexpressing VCaP cells and empty vector control.

[0056] Figure 16 The GSEA diagram shows the AR signaling pathway activity in FST-overexpressing VCaP cells compared to the empty vector control.

[0057] Figure 17 and Figure 18 To stably express LNCaP and VCaP cells with the specified vector, RT-qPCR was used. Figure 17 ) and Western blot ( Figure 18 Analysis was conducted to assess the expression levels of AR and its target genes.

[0058] Figure 19 and Figure 20 The expression levels of AR and its target genes in LNCaP cells transfected with specified sgRNA were determined by RT-qPCR. Figure 19 ) and Western blot ( Figure 20 )Evaluate.

[0059] Figure 21 and Figure 22 RT-qPCR for infecting LNCaP cells with a specified shRNA vector ( Figure 21 ) and Western blot ( Figure 22 )analyze.

[0060] Figure 23Western blot analysis of AR and p-SMAD1 / 5 / 8 expression in VCaP cells stably expressing the specified vector.

[0061] Figure 24 Venn diagram of potential FST-binding proteins identified from VCaP conditioned medium (CM) by mass spectrometry.

[0062] Figure 25 and Figure 26 RT-qPCR for AR and its target gene expression in BMP16 knockdown prostate cancer cell lines Figure 25 ) and Western blot ( Figure 26 )analyze.

[0063] Figure 27 Western blot analysis was performed on LNCaP and VCaP cells that stably expressed the specified vectors after treatment with recombinant BMP11 (200 ng / mL) for 1 or 40 hours.

[0064] Figure 28 Western blot analysis was performed on VCaP cells treated with FST and increasing concentrations of BMP11 for 40 hours.

[0065] Figure 29 Western blot analysis was performed on VCaP cells that stably expressed the specified vector after treatment with BMP11 for 40 hours.

[0066] Figures 30-41 FST-driven AR-targeted therapy for drug resistance

[0067] Figures 30-33 To express a specified carrier ( Figure 30 It is an FST overexpression vector. Figure 31 To exhaust SMAD1 / 5 / 9, Figure 32 To deplete BMPR1A / B, Figure 33 To activate the BMPR1A Q233D mutant, VCaP cells overexpressing VCaP were seeded at a density of 1,000 cells / well in 96-well plates and treated with enzalutamide for 7 days. Cell viability was normalized to the untreated control. The drug concentration (IC50 μmol / L) that induced 50% survival inhibition is indicated.

[0068] Figure 34 The half-maximal inhibitory concentration (IC50 μmol / L) of enzalutamide was determined for VCaP and LNCaP cells infected with BMP11 shRNA.

[0069] Figure 35 Western blot analysis of AR and PSA expression in VCaP cells expressing a specified vector.

[0070] Figure 36 The half-maximal inhibitory concentration (IC50 μmol / L) of enzalutamide was determined for VCaP cells expressing the specified vector.

[0071] Figure 37 Growth analysis of VCaP cells expressing the specified vector after treatment with solvent (DMSO) or enzalutamide (10 μM).

[0072] Figure 38 Colony formation analysis of VCaP cells expressing the specified vector after treatment with solvent (DMSO) or enzalutamide (bar chart, mean ± SD, n=4).

[0073] Figure 39 Prostate globules formed in LNCaP and VCaP cells expressing the specified vectors after treatment with solvent (DMSO) or enzalutamide. Representative images (scale bar, 100 μm) and quantitative analysis of prostate globule counts from biologically independent experiments (bar chart, mean ± SD, n=4).

[0074] Figure 40 and Figure 41 Tumor growth and endpoint tumor weight were measured in castrated SCID mice carrying xenografts expressing specified vectors of LNCaP(K) or VCaP(L). Mice were given enzalutamide (10 mg / kg) from day 1 after transplantation. Data are presented as mean ± SEM (VCaP: n=12 empty vector and FST317, n=7 FST344; LNCaP: n=14 empty vector, n=10 FST317, n=8 FST344).

[0075] Figures 42-56 To target FST to reverse anti-AR resistance and inhibit DNPC metastasis

[0076] Figure 42 A schematic diagram of the design for FST-Trap.

[0077] Figure 43 This study describes the co-immunoprecipitation of FST317 (FLAG tag) and mature BMP11 (V5 tag) in 293T cells with or without FST-Trap (Fc tag) treatment.

[0078] Figure 44 and Figure 45 VCaP cells were treated with a specified protein for 1 hour. Figure 44 ) or 40 hours ( Figure 45 Western blot analysis following the initial analysis.

[0079] Figure 46Cell growth analysis of VCaP cells expressing empty vector or FST under enzalutamide (10 μM) and / or FST-Trap (5 μg / mL) treatments (bar chart, mean ± SD; n=6).

[0080] Figure 47 Experimental design for tracking the growth of VCaP and LNCaP xenografts in SCID mice during castration ± FST-Trap treatment.

[0081] Figure 48 and Figure 49 VCaP in male castrated SCID mice ( Figure 48 ) or LNCaP ( Figure 49 Tumor growth and weight of xenograft tumors. Mice were treated with enzalutamide (10 mg / kg) from day 1 post-transplantation, followed by either solvent or FST-Trap (300 μg / mouse) when the tumor reached 100 mm³. Data are presented as mean ± SEM (VCaP: solvent n=6, FST-Trap n=7; LNCaP: n=8).

[0082] Figure 50 and Figure 51 for Figure 48 and Figure 49 The mouse VCaP ( Figure 50 ) and LNCaP ( Figure 51 Representative immunohistochemical staining of xenograft tumors (left; scale bar, 50 μm) and their quantitative analysis (right; bar chart, mean ± SEM; n=5).

[0083] Figure 52 Prostate bulb formation after PC3M cells were treated with FST-Trap (5 μg / mL). Representative images (left) (scale bar, 100 μm) and quantitative analysis of prostate bulb counts (right) are shown (bar plot, mean ± SD, n=4).

[0084] Figure 53 Representative bioluminescent images (left) and luciferase counting quantification (right) of male nude mice injected intracardiacly with PC3M cells and treated with solvent or FST-Trap (300 μg / mouse) (bar chart, mean ± SEM; n=6).

[0085] Figure 54 Tumor growth (left) and tumor weight (right) of RM1 xenograft tumors in male C57BL / 6J mice. When the tumor volume reached 50 mm³, mice were treated with solvent (n=6), anti-PD-1 (n=6), FST-Trap (300 μg / mouse, n=6), or combination therapy (n=7) (bar chart, mean ± SEM).

[0086] Figure 55 for Figure 54 Representative immunohistochemical staining (top; scale bar, 50 μm) and quantitative analysis (bottom; bar chart, mean ± SEM, n=5) of the mouse xenograft tumors.

[0087] Figures 56-61 Increased FST expression in DNPC

[0088] Figure 56 Ridge maps were used to show the expression levels of TGFβ / BMP pathway-related genes in ARPC (n=84), NEPC (n=23), or DNPC (n=27) subtypes of CPGEA-2020. Statistical analysis was performed using unpaired two-sided Student's t-test.

[0089] Figure 57 For Kaplan-Meier analysis, disease-free survival was based on the alteration status of specified genes (alteration groups were defined as low expression or deep deletion), and data were obtained from the Taylor et al. 2010 cohort. Statistical analysis was performed using Cox regression.

[0090] Figure 58 Disease-free survival was analyzed for Kaplan-Meier studies of BMPR1A-modified states (defined as low expression or deep deletion) in the TCGA dataset. P-values ​​were calculated using Cox regression.

[0091] Figure 59 Pearson correlations between KLK3 expression and specified genes in the SU2C-2015 cohort (ARPC, n=78; NEPC, n=25; DNPC, n=15) (25).

[0092] Figure 60 To show the cell type distribution in integrated human prostate cancer scRNA-seq data for UMAP.

[0093] Figure 61 Visualizing single-cell resolution for UMAP FST The expression.

[0094] Figures 62-78 FST inhibits BMP11-mediated AR expression

[0095] Figure 62 AR and PSA protein levels in VCaP cells after 40 hours of treatment with increasing concentrations of FST317.

[0096] Figure 63 and Figure 64 VCaP cells transfected with the specified sgRNA ARThe expression of its target genes was measured by RT-qPCR ( Figure 63 ) and Western blot ( Figure 64 )analyze.

[0097] Figure 65 LNCaP and VCaP cells infected with specified shRNA vectors SMAD5 and SMAD9 Expression RT-qPCR analysis.

[0098] Figure 66 and Figure 67 The expression of AR and its target genes in VCaP cells stably expressing the specified shRNA was determined by Western blot analysis. Figure 66 ) and RT-qPCR ( Figure 67 )analyze.

[0099] Figure 68 Western blot analysis of AR and PSA expression in LNCaP and VCaP cells after 40 hours of treatment with SB-431542.

[0100] Figure 69 Western blot analysis of AR and PSA expression in LNCaP and VCaP cells infected with specified shRNA vectors.

[0101] Figure 70 LNCaP and VCaP cells infected with specified shRNA vectors ACVR1B , TGFBR1 and TGFBR2 Expression RT-qPCR analysis.

[0102] Figure 71 Western blot analysis of AR and PSA expression in LNCaP and VCaP cells infected with specified shRNA vectors.

[0103] Figure 72 Western blot analysis of VCaP cells stably expressing the specified vector.

[0104] Figure 73 To validate the efficacy of RT-qPCR in AR-positive prostate cancer cell lines BMP11 Reduce efficiency.

[0105] Figure 74 and Figure 75 for GDF8 Knockdown LNCaP and VCaP cells were analyzed by RT-qPCR ( Figure 74 ) and Western blot ( Figure 75 )analyze.

[0106] Figure 76 and Figure 77 for INHBB Knockdown LNCaP and VCaP cells were analyzed by RT-qPCR ( Figure 76 ) and Western blot ( Figure 77 )analyze.

[0107] Figure 78 Co-immunoprecipitation (Co-IP) of FST317 and mature BMP11 in conditioned medium (CM) for 293T cells.

[0108] Figures 79-89 FST drives resistance to AR-targeted therapy.

[0109] Figure 79 The half-maximal inhibitory concentration (IC50 μmol / L) of enzaluramide in LNCaP cells expressing the specified vector was determined.

[0110] Figure 80 Western blot analysis of AR and PSA expression in LNCaP and VCaP cells expressing the specified vectors.

[0111] Figures 81-87 The half-maximal inhibitory concentration (IC50 μmol / L) of enzalutamide was determined for LNCaP / VCaP cells expressing the specified vector.

[0112] Figure 81 FSTL1 failed to induce enzalutamide resistance.

[0113] Figure 82 The IC50 of enzalutamide was significantly increased by depleting SMAD1 / 5 / 9 in VCaP and LNCaP cells.

[0114] Figure 83 The depletion of BMPR1A / B in VCaP and LNCaP cells significantly increased the IC50 of enzalutamide.

[0115] Figure 84 The loss of SMAD2 / 3 receptors had no effect on enzalutamide sensitivity (and). Figure 85 The loss of TGFβ receptors has no effect on enzalutamide sensitivity.

[0116] Figure 86 The loss of GDF8 has minimal impact on the sensitivity of enzalutamide.

[0117] Figure 87 The loss of activin has minimal impact on the sensitivity of enzalutamide.

[0118] Figure 88Growth analysis of LNCaP cells expressing a specified vector after treatment with solvent (DMSO) or enzalutamide.

[0119] Figure 89 Colony formation analysis of LNCaP cells expressing the specified vector after treatment with solvent (DMSO) or enzalutamide (bar chart, mean ± SD, n=3).

[0120] Figures 90-97 To target FST to reverse anti-AR resistance and inhibit DNPC metastasis.

[0121] Figure 90 Western blot analysis of whole-cell lysate (WCL) and conditioned medium (CM) from 293T cells.

[0122] Figure 91 Co-immunoprecipitation of FST317 (FLAG tag) and wild-type activin A or M418A mutant (Fc tag) in 293T cells CM.

[0123] Figure 92 Coomassie Brilliant Blue staining of FST-Trap protein purified from Expi293F cells.

[0124] Figure 93 VCaP cell growth was analyzed in the presence or absence of FST-Trap (5 μg / mL) after treatment with recombinant FST317 (500 ng / mL) and enzalutamide (10 μM), as shown in bar charts (mean ± SD, n=6). Statistical analysis was performed using one-way ANOVA.

[0125] Figure 94 LNCaP cells expressing the empty vector or FST were analyzed for cell growth with or without enzalutamide (10 μM) and / or FST-Trap (5 μg / mL) treatment (bar chart, mean ± SD; n=6). Statistical analysis was performed using one-way ANOVA.

[0126] Figure 95 for Figure 48 and Figure 49 Weight analysis of the mice (bar chart, mean ± SEM).

[0127] Figure 96 Cell growth analysis of PC3M cells treated with solvent or FST-Trap (5 μg / mL) (bar chart, mean ± SD; n=6). Statistical analysis was performed using unpaired two-tailed Student's t-test.

[0128] Figure 97 for Figure 54Weight analysis of the mice. Detailed Implementation

[0129] The present invention is further illustrated below by way of embodiments, but the invention is not limited to the scope of the embodiments described herein. Experimental methods in the following embodiments that do not specify specific conditions were performed according to conventional methods and conditions, or as selected according to the product instructions.

[0130] Experimental Materials and Methods

[0131] The reagents and materials used in this invention are shown in Table 1.

[0132] Table 1 Reagents and Materials

[0133]

[0134]

[0135]

[0136] Example 1: FST as a biomarker for prostate cancer, and a key factor mediating the progression of treatment-resistant and aggressive diseases through tumor intrinsic mechanisms and microenvironment remodeling.

[0137] 1. Clinical Samples

[0138] 1.1 Human prostate cancer specimen

[0139] The experiments involving human prostate cancer specimens were approved by the Ethics Committee of Changhai Hospital (Approval No. CHEC2019-012). The collected samples included 7 formalin-fixed paraffin-embedded (FFPE) tissue blocks, 64 serum samples, and 4 tissue specimens for single-cell RNA sequencing.

[0140] 2. Animal experiments

[0141] 2.1 Animal Models and Feeding

[0142] All mouse experiments were conducted in accordance with protocols approved by the Institute of Laboratory Animal Management and Use (IACUC) of Zhejiang University. Mice were purchased from Gempharmatech and housed in a controlled environment (temperature approximately 22°C, humidity 40-60%, 12-hour light / 12-hour dark cycle) with free access to food and water.

[0143] 2.2 Subcutaneous xenograft tumor model

[0144] For xenograft tumor studies involving enzalutamide treatment, 2×10 6LNCaP or VCaP cells were resuspended in a 1:1 mixture of 100 μL PBS and Matrigel (ABW, #082704) and subcutaneously inoculated into the lateral ventral regions of 8-week-old castrated male NOD / SCID mice. Starting from the day of cell inoculation, enzalutamide (10 mg / kg; TargetMol, #T6002) or a solvent control (containing 1% carboxymethyl cellulose, 0.1% Tween-80, and 5% DMSO) was administered daily by gavage.

[0145] In other local tumor growth experiments, 5×10 5 Myc-CaP cells were inoculated into male FVB / NJ or NCG mice at a dose of 2 × 10⁻⁶. 4 RM1 cells were inoculated into male C57BL / 6J or NCG mice.

[0146] For evaluation Figure 7 The treatment effects shown were observed in mice receiving FST-Trap (300 μg / mouse, retroorbital injection, every two days), anti-PD1 antibody (200 μg / mouse, intraperitoneal injection, twice weekly), or a combination of both. Tumor size was measured using calipers, and volume was calculated using the formula: length × width² × 0.5.

[0147] 2.3 Experimental Transfer Model

[0148] 1×10 6 PC3M cells were resuspended in 100 μL PBS and injected intracardiacly via the left ventricle using a 26G tuberculin syringe. Transfer load was detected using non-invasive bioluminescence imaging with a Spectral Instruments Imaging system. Mice were observed daily throughout the experiment for signs of disease or death.

[0149] 3. Cell Culture

[0150] 3.1 Cell line origin

[0151] LNCaP, VCaP, 22Rv1, C4-2B, DU145, PC3, Myc-CaP, RM1, 293T, and RAW 264.7 cells were purchased from the American Type Culture Collection (ATCC). Expi293F cells were purchased from Thermo Fisher Scientific. The parental CWR22Pc and PC3M cell lines were kindly provided by Dr. Filippo Giancotti's laboratory, and the CWR22Pc-2 subline was established in our laboratory through single-cell cloning. All cell lines were cultured in basal medium supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, and 100 IU / mL penicillin / streptomycin. Specifically, LNCaP, 22Rv1, PC3M, and CWR22Pc-2 cells were cultured in RPMI-1640 medium; PC3 cells were cultured in DMEM / F12 medium; RAW 264.7 cells were cultured in α-MEM medium; and the remaining cell lines were cultured in DMEM medium.

[0152] 3.2 Lentiviral Packaging and Construction of Stable Cell Lines

[0153] Lentiviral particles were generated by co-transfecting 293T cells with transfer plasmids encoding shRNA, FST, SOX9, or BMP11, along with packaging plasmid psPAX2 (Addgene, #12260) and envelope plasmid pMD2.G (Addgene, #12259). The plasmids were mixed in Opti-MEM serum-depleted medium (Thermo Fisher Scientific, #31985-070). Linear polyethyleneimine (PEI) transfection reagent (BIOHUB, #78PEI40000) was diluted separately with Opti-MEM. The diluted plasmids and PEI solution were gently mixed and incubated at room temperature for 20 minutes to form a complex, then added dropwise to 293T cells. The medium was changed 12–16 hours post-transfection. The viral supernatant was collected 36 hours after medium change, filtered through a 0.45 μm syringe filter, and stored for later use.

[0154] At infection, target cells were exposed to a mixture of 50% viral supernatant and 50% fresh culture medium (with 2 μg / mL polybrene added) for 24 hours. To establish stable cell lines, selection was initiated 48 hours post-infection with 2 μg / mL puromycin for 3 days. All shRNA and sgRNA sequences used in this study are detailed in Table 2.

[0155] Table 2 shRNA and sgRNA sequences

[0156] 4. Cellular functional experiments

[0157] 4.1 Cell viability and growth assay

[0158] In the enzalutamide sensitivity assay, LNCaP and VCaP cells were seeded at a density of 1,000 cells per well in 96-well plates and treated for 7 days with gradient concentrations of enzalutamide (0.1, 0.3, 1, 3, 10, 30, 100 μM) or solvent (DMSO). Cell viability was determined by the CCK-8 assay, and the results were standardized with a solvent control. The half-maximal inhibitory concentration (IC50) was determined. 50 The nonlinear regression was calculated using GraphPadPrism (v9.0.0.121).

[0159] Cell proliferation was simultaneously assessed: LNCaP and VCaP cells were seeded at 1,000 cells / well, while PC3M, RM1, and Myc-CaP cells were seeded at 800 cells / well, all using the same plate size. Cell viability was monitored at specified time points using the CCK-8 assay, and the OD values ​​at each time point were normalized to the OD value of day 1 to plot growth curves.

[0160] 4.2 Cloning experiment

[0161] LNCaP and VCaP cells were seeded at a density of 1,000 cells / well in 12-well plates and allowed to adhere overnight. After 14 days of enzalutamide treatment, cells were washed with PBS, fixed with ice-cold methanol, and stained with 0.1% crystal violet for 10 minutes. After thorough rinsing and air-drying, clones were imaged and quantified.

[0162] 4.3 Prostasphere Formation Experiment

[0163] PC3M or RM1 cells were prepared into single-cell suspensions at a density of 1,000 cells / mL and seeded onto ultra-low adsorption plates. Cells were maintained in a serum-free ProstaLife medium complete kit (Lifeline, #LL-0041) supplemented with 1:50 B-27, 20 ng / mL basic fibroblast growth factor (bFGF), and 40 ng / mL epidermal growth factor (EGF) for 7–10 days.

[0164] For LNCaP and VCaP cells, tumor spheroids were generated using a matrix gel-based method. In short, 1,000 cells per well were resuspended in 10 μL of medium, mixed with 30 μL of cold matrix gel, and seeded into 24-well plates. After polymerization at 37°C for 20 minutes, the same supplemental medium was added, and the cells were cultured for 14–20 days before imaging and quantification.

[0165] 5. Immunological experiments

[0166] 5.1 FACS analysis of tumor-infiltrating immune cells

[0167] Tumors were removed, weighed, and representative fragments of appropriate size were dissected for processing. Single-cell suspensions were prepared by mechanical dissociation in FACS buffer (PBS containing 2% FBS). After filtration through a 40 μm filter and centrifugation, cells were stained for analysis. Myeloid cells were stained with a surface antibody (1:200 dilution) in 50 μL FACS buffer at 4°C for 30 min, followed by washing. After surface staining, lymphoid cells were fixed, permeabilized, and stained with intracellular targets using the eBioscience™ Foxp3 / transcription factor staining buffer kit (Thermo Fisher Scientific, #00-5523-00) according to the manufacturer's protocol. Samples were resuspended in FACS buffer and filtered before collection. Samples were analyzed on a full-spectrum flow cytometer with appropriate voltage and compensation settings. Data were collected and processed using FlowJo software (v10.8.1) to determine the proportions of different cell subpopulations. Antibody details are shown in Table 1.

[0168] 6. Imaging examination

[0169] 6.1 Bioluminescence Imaging

[0170] In a mouse bone transfer model established via intracardiac injection, bioluminescence was performed according to the established protocol (reference PMID: 31327655).

[0171] 7. Molecular biology experiments

[0172] 7.1 RNA extraction and RT-qPCR

[0173] Total RNA was extracted from cells using Trizol reagent (Vazyme, #R401-01). Subsequently, 1 μg of isolated RNA was reverse transcribed into cDNA using HiScript II Q RT SuperMix (Vazyme, #R312-02) according to the manufacturer's instructions. Real-time quantitative PCR (qPCR) was then performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme, #Q711-03). Gene expression levels were compared between 2... ^-ΔΔCt Quantitative methods were employed, with normalization performed using the endogenous control gene GAPDH. Differential expression data were visualized as bar charts using GraphPad Prism (v9.0.0.121) software, with fold changes normalized to their respective control cell lines. All qPCR primer sequences are detailed in Table 3. Each experiment comprised three independent biological replicates (n=3), and data are presented as mean ± SD.

[0174] Table 3 qPCR primer sequences

[0175] 7.2 Protein Expression Analysis (Western Blot)

[0176] After washing with PBS, the cell pellet was lysed on ice for 20 min in buffer A (50 mM HEPES, 150 mM NaCl, 2.5 mM EDTA, 1% Triton X-100) containing protease and phosphatase inhibitors (Thermo Fisher Scientific, #A32961). The lysates were centrifuged at 20,000 × g for 15 min at 4°C, and the supernatant was collected for protein quantification using the Pierce BCA Protein Assay Kit (Yeasen, #20201ES86). Similarly, for analysis of conditioned media, samples were centrifuged at 600 × g for 5 min to remove cell debris. All protein samples were denatured by boiling in loading buffer and then analyzed by Western blot to detect the target proteins.

[0177] 7.3 Co-immunoprecipitation (Co-IP) analysis of conditioned media

[0178] 293T cells were seeded in 6 cm culture dishes and transfected with the specified plasmid using PEI at approximately 80% confluence. 24 hours post-transfection, the medium was replaced with serum-free medium. After another 24 hours, the conditioned medium was collected and incubated with anti-FLAG M2 magnetic beads (Sigma-Aldrich, #M8823) or V5-coupled agarose beads (Sigma-Aldrich, #A7345) at 4°C with gentle stirring for 3 hours. The beads were then washed with buffer A, and the bound proteins were eluted with glycine buffer (pH 2.6) and neutralized with Tris-HCl (pH 8.9). The eluent was denatured in loading buffer at 95°C for 15 minutes and analyzed by Western blot.

[0179] 8. Histology and Immunohistochemistry

[0180] 8.1 H&E staining, IHC and mIHC

[0181] Fresh tissue samples (xenograft tumors or bone tissue) were washed with cold PBS and fixed in 4% PFA (Biosharp, #E672002) at 4°C for 24 hours. Bone tissue was then decalcified in EDTA decalcification solution (Sangon Biotech, #E671001-0500) for 10 days, with the solution changed every 2-3 days, and then refixed in 4% PFA.

[0182] After washing with PBS, all tissues were dehydrated by a series of ethanol gradients, cleared with xylene, infiltrated with paraffin, and then embedded into blocks. Continuous sections of 3-5 μm thickness were cut, floated in a 40°C water bath, attached to adhesive slides, and dried overnight at 37°C.

[0183] H&E staining: After dewaxing and rehydration, the sections were stained with hematoxylin (Servicebio, #G1004) and eosin (Servicebio, #G1001).

[0184] Immunohistochemistry (IHC): Antigen retrieval was performed in EDTA buffer (pH 9.0; ZSGB-BIO, #ZLI-9079). Endogenous peroxidase was quenched with 3% H2O2 in methanol. Sections were blocked with 3% BSA and incubated overnight at 4°C with primary antibody (see Table 1). Subsequently, sections were incubated with HRP-conjugated secondary antibodies of suitable species—goat anti-rabbit (Jackson ImmunoResearch, #111-035-003), goat anti-rat (Jackson ImmunoResearch, #112-035-003), or rabbit anti-goat (Abbkine, #A21030). Signal detection was performed using the DAB substrate kit (Gene Tech, #GK347011).

[0185] 9. ELISA testing

[0186] Serum samples from prostate cancer patients were obtained from Changhai Hospital. Serum was diluted with the dilution buffer provided with the FST ELISA kit (RayBiotech, #ELH-Follistatin) and processed according to the manufacturer's instructions. Absorbance was measured at 450 nm, and FST concentration was determined based on a standard curve.

[0187] 10. Recombinant protein preparation

[0188] 10.1 Protein Expression and Purification

[0189] FST-Trap is a fusion protein linking human activin A (M418A mutant) to the Fc domain of human IgG1. To prevent unwanted Fc-receptor interactions, key residues were mutated: L234A, L235A, and P329G. The recombinant FST-Trap protein was expressed in Expi293F cells using the Harvest™ 293F expression system (Opmm, #TR01) according to the manufacturer's instructions. For the FST317 recombinant protein, a construct with an 8×His tag was constructed in the pcDNA3.1 vector and expressed under the same conditions.

[0190] Culture supernatants were collected and purified by affinity chromatography: FST-Trap was separated using Protein A resin (GenScript, #L00210) at 4°C for 3 hours, and FST317 was purified using Ni-NTA resin (GenScript, #L00250) according to the manufacturer's instructions. In both cases, after thorough washing, bound proteins were eluted according to their respective recommended protocols. Eluted proteins were confirmed by SDS-PAGE and Coomassie Brilliant Blue staining (Sangon Biotech, #A610037), then transferred to PBS, concentrated, flash-frozen, and stored at -80°C.

[0191] 11. Mass spectrometry analysis

[0192] 11.1 Sample Preparation and Mass Spectrometry Analysis

[0193] VCaP cells stably expressing the FLAG tags FST317 or FST344 were cultured serum-free for 24 hours after reaching 90% confluence. Conditioned medium was collected, concentrated, and pre-cleaned with Protein A / G magnetic beads (Pierce, #88803) on a rotating shaker at 4°C for 30 minutes. The supernatant was then incubated with anti-FLAG M2 magnetic beads (Sigma-Aldrich, #M8823) at 4°C for 3 hours. After thorough washing with pre-chilled buffer A (as described above), bound proteins were eluted with glycine buffer (pH 2.6), neutralized with Tris-HCl (pH 8.9), mixed with loading buffer, and denatured at 95°C for 15 minutes. Proteins were separated by SDS-PAGE. The gels were sent to the LC-MS / MS analysis platform of the Institute of Life Sciences, Zhejiang University, as detailed below.

[0194] Gel sections were destained, reduced with dithiothreitol, alkylated with iodoacetamide, and digested overnight at 37°C with sequencing-grade trypsin. The resulting peptides were extracted, desalted using a C18 StageTip, and concentrated. The peptide mixture was analyzed by nano-liquid chromatography-tandem mass spectrometry on a Q Exactive HF-X instrument (Thermo Fisher) equipped with an Easy-nLC1200 system. The mobile phase consisted of 0.1% formic acid, water (A), and 80% acetonitrile (B). The digested peptides were directly loaded onto an analytical column (75 μm × 15 cm, 1.9 μm C18, 5 μm head) at a flow rate of 300 nanoliters / min.

[0195] Peptide separation was performed using the following linear gradient: 4-6% B over 1 minute, 6-26% B over 42 minutes, 26-35% B over 12 minutes, 35-90% B over 2 minutes, and finally, 90% B held for 3 minutes. Full-scan mass spectrometry was performed in the range of 400 to 1400 m / z with a resolution of 60,000 (full width at half maximum) and an automatic gain control target of 3 × 10⁻⁶. 6 The maximum injection time was 45 milliseconds. The instrument was set to top-speed mode, with a scan and MS / MS scan cycle of 1 second. After each full scan, the strongest precursor ions (charge states 2-7, intensity >1×10⁻⁶) were analyzed. 5 Tandem MS analysis was performed in a quadrupole with a 1.6 m / z window isolation. High-energy collisional dissociation (HCD) was performed at a 27% normalized collision energy, and the resulting fragments were analyzed in Orbitrap at a resolution of 15,000. The maximum injection time was set to 22 ms, with dynamic exclusion enabled for 30 s, and a mass tolerance of 10 ppm around the precursor. All mass spectrometry data were processed using Proteome Discoverer (Thermo Fisher Scientific).

[0196] 12. Transcriptomics Analysis

[0197] 12.1 RNA Sequencing and Data Processing

[0198] VCaP cells stably overexpressing FST317 and FST344 isoforms were cultured in 6% decarbonized serum for 7 days, and PC3M cells were harvested 4 days after infection with sh-FST lentivirus. Total RNA was extracted from the cells using Trizol reagent (Vazyme, #R401-01). RNA samples were sent to Annoroad Genetics for sequencing. After quality assessment (Nanodrop 2000; Labchip GXTouch), mRNA was enriched using Oligo(dT) magnetic beads to construct libraries. After mRNA fragmentation, first- and second-strand cDNAs were synthesized. After end repair, A-tailing, and adapter ligation, fragments of approximately 350 bp were selected. Library quality was validated (Qubit 3.0; Agilent 2100), and effective concentrations were quantified by qPCR (Bio-Rad CFX 96). Qualified libraries were mixed and sequenced at 150 bp paired ends on an Illumina platform.

[0199] Raw sequencing reads were processed and filtered using fqtools_plus to obtain high-quality clean data. The cleaned reads were then aligned to the human reference genome (GRCh38) using HISAT2 (v2.1.0). Gene expression quantification was performed against GENCODE annotations using featureCounts (v2.0.1), generating a count matrix. This matrix was then converted to a TPM (transcripts per million) expression matrix for subsequent downstream analysis.

[0200] 12.2 Differential Expression Analysis

[0201] Differential gene expression analysis was performed using the DESeq2 R package (v1.38.3). The raw count expression matrix, with a total gene count greater than 10 for all samples, was used as input data. Differentially expressed genes (DEGs) were identified based on the following criteria: |log2 fold change (log2FC)| > 1 and Benjamini-Hochberg corrected p-value < 0.05.

[0202] 12.3 Gene Set Enrichment Analysis (GSEA)

[0203] Gene lists sorted by fold change in gene expression were subjected to gene set enrichment analysis (GSEA) using the clusterProfiler R package (v4.7.1.003). Gene sets related to WikiPathways and Gene Ontology (GO) biological processes were obtained from the Molecular Signature Database (MSigDB) using the msigdbr package (v7.5.1) in R. GSEA was used to assess the enrichment of these GO biological process gene sets, with entries having a p-value less than 0.05 considered statistically significant.

[0204] 13. Survival Analysis

[0205] Survival analyses were performed using the Kaplan-Meier method and Cox regression modeling. All analyses were performed in R using the `survival` package (v3.5-8), and graphical visualizations were provided by the `survminer` package (v0.4.9).

[0206] 14. Prostate Cancer Subtypes (AR / NE / DNPC) Classification

[0207] Prostate cancer (PCa) subtype classification was performed according to the method described by Su et al. 2019 (PMID: 31327655). In short, samples with AR and / or KLK3 mRNA z-score > 0 were classified as ARPC (AR-positive prostate cancer), and samples with SYP and / or CHGA mRNA z-score > 0 were classified as NEPC (neuroendocrine prostate cancer). For samples expressing both AR and NE markers, AR scores were compared with NE scores, and subtype assignment was based on the higher score. All remaining samples were classified as DNPC (double-negative prostate cancer).

[0208] 15. Single-sample gene set enrichment analysis (ssGSEA)

[0209] Single-sample gene set enrichment analysis (ssGSEA) was performed using the R package GSVA (v1.46.0).

[0210] The AR scoring gene set is: KLK3, KLK2, TMPRSS2, FKBP5, NKX3-1, PLPP1, PMEPA1, PART1, ALDH1A3, STEAP4.

[0211] The NE scoring gene set is: SYP, CHGA, CHGB, ENO2, CHRNB2, SCG3, SCN3A, PCSK1, ELAVL4, NKX2-1.

[0212] The BMP activity-specific gene set includes: AHSG, CER1, CTDSP1, CTDSP2, CTDSPL, GSK3B, MAP3K7, MAPK1, NOG, NUP214, PPM1A, PPP1CA, PPP1R15A, SKI, SMURF1, SMURF2, SOSTDC1, TAB1, TAB2, XIAP, and ZFYVE16.

[0213] The active gene set of TGFβ is: PMEPA1, SERPINE1, VMP1, IER3, TNFRSF12A, SPHK1, FZD8, CTGF, ST3GAL1, DACT1, GADD45B, IL11, SKIL, SAMD11, SLC20A1, CBFB, DYRK2, B4GALT1, SOX4, FSTL3, PRPS1L1, PRPS1, ARHGEF40, SH3PXD2A, AQP1, GSTT2, GPR68.

[0214] 16. Single-cell RNA sequencing

[0215] 16.1 Construction and Sequencing of Single-Cell RNA Libraries from Patient Samples

[0216] Seven samples were from a previous study (PMID: 35789834). Four new samples were mapped using the 10x Genomics Chromium Next GEM Single Cell 3' Kit v3.1. Adapters and poly-A tails were trimmed, reads were aligned to GRCh38 and annotated using a genome annotation file (GENCODE V32). Additionally, the raw count matrices for dp2, dp5, and dp6 were downloaded from the GEO database, accession number GSE137829 (PMID: 33328604).

[0217] Next, basic single-cell data analysis was performed using the Seurat R package. For quality control, cells with <500 features, <1000 counts, or >20% mitochondrial percentage were excluded from downstream analysis as unqualified cells. The top 50 principal components were selected for unsupervised clustering and visualization. Initial annotation distinguished epithelioid and mesenchymal-like states using EPCAM and VIM; considering prostatic epithelial origin, mesenchymal-like cells were further refined. Batch effects were checked for visualization and assessed using CellMixS. Cluster identities were assigned using FindAllMarkers (adjusted P < 0.05, log2FC > 0.25, Δprop > 0.25); clusters lacking clear markers were labeled "unknown," with ambiguous cases marked accordingly. For FST differential analysis, Fisher's exact test was used to test the association between FST± and AR± or KLK3± in malignant cells. Differential expression analysis used Seurat's FindMarkers to compare FST. + vs FST - and FST + AR - vs FST - AR + DEG is defined by adjusting P < 0.05 and |log2FC| > 0.5.

[0218] 17. Statistical Methods

[0219] Statistical analyses were performed using R (v4.2.3) and GraphPad Prism (v9.0.0.121). All analyses included at least three biologically independent samples. In animal studies, for the subcutaneous injection model, each subcutaneous tumor was considered a biologically independent sample; for the intracardiac injection model, each mouse was considered a biologically independent sample. Data are presented as mean ± SD or mean ± SEM, as indicated in the figure captions. Differences between two groups were assessed using an unpaired two-sided Student's t-test, and a p-value <0.05 was considered statistically significant. One-way ANOVA was used for comparisons among multiple groups. Pearson correlation coefficients were used for correlation analysis. Unless otherwise stated, all experiments were performed at least three times independently.

[0220] 18. Data Availability

[0221] All data generated in this study have been stored in public databases, as detailed below: RNA sequencing data of VCaP and PC3M cells, single-cell RNA sequencing data of VCaP cells, and single-cell RNA sequencing data of four prostate cancer cases are stored in the GSA database, accession numbers HRA013060, HRA014829, and HRA014657, respectively; mass spectrometry proteomics data have been stored on ProteomeXchange via the iProX platform, identifier PXD070388.

[0222] This study used several publicly available prostate cancer datasets, including: the SU2C dataset from Robinson et al. (2015, dbGaP: phs000915.v.p1) and Abida et al. (2019, dbGap: phs000915.v2.p2); the TCGA prostate cancer dataset from The Cancer Genome Atlas Research (2018) (https: / / www.cbioportal.org / study / summary?id=prad_tcga_pan_can_atlas_2018); the Taylor dataset from Taylor et al. (2010, GEO: GSE21032); and the Changhai dataset from Li et al. (2020, GSA: PRJCA001124). In addition, single-cell sequencing data from seven patients included four datasets from GSA (HRA002145) and three datasets from GEO (GSE137829).

[0223] result: Increased FST expression in DNPC Despite Pten In the missing mouse model Bmpr2 The absence of [a specific gene] prolonged survival and established its tumor-promoting role in prostate cancer initiation, but analysis of clinical datasets revealed the opposite pattern in advanced disease. Patients in the Chinese Prostate Cancer Genome and Epigenome Atlas (CPGEA) and SU2C cohorts were classified into ARPC, DNPC, and NEPC subtypes. Figure 1 and Figure 2 Single-sample gene set enrichment analysis (ssGSEA) showed that in CPGEA (… Figure 1 and Figure 3 ,as well as Figure 56 ) and SU2C queue ( Figure 2 and Figure 4 In the DNPC patients, BMP signaling activity was significantly downregulated compared to AR-sensitive prostate cancer (ARPC) samples. In contrast, TGFβ signaling activity did not show significant changes among the three prostate cancer subtypes. Figures 1-4 Detailed analysis of the core pathway components revealed BMPR1A and BMPR1B Expression was significantly reduced in DNPC ( Figure 3 and Figure 4 In addition, TGFβ superfamily antagonists FST Elevated expression in DNPC ( Figure 3 and Figure 4 This is related to the mesenchymal marker vimentin and prostate stem cell markers. ITGB4 Increase the relevance of the expression ( Figure 3 This aligns with the dedifferentiation phenotype. Conversely, neither TGFβ receptors nor SMADs showed downregulation ( Figure 56 ).

[0224] Given that DNPC is associated with the worst clinical prognosis among prostate cancer subtypes, patient datasets were examined to validate the prognostic importance of BMP pathway gene expression in prostate cancer progression. From the mCRPC genomic landscape study with relevant longitudinal clinical outcome data, 81 CRPC patients treated with abiraterone or enzalutamide were identified, and these patients had tumor whole-exome and RNA sequencing (RNA-seq) data within 30 days prior to treatment initiation. Cox proportional hazards analysis revealed that high... FST mRNA levels were associated with significantly reduced survival rates. Figure 5 Interestingly, Gao FST The expression of adverse clinical outcomes was more pronounced in patients treated with enzalutamide than in those treated with abiraterone, a finding confirmed by Pearson correlation analysis. Figure 5 This suggests a possible difference in drug resistance mechanisms between tumors that directly block AR and those that deplete androgens. Furthermore, BMPR1A , SMAD1 , SMAD9 or SMAD4 Decreased expression was associated with shortened disease-free survival in the Taylor dataset. Figure 57 Analysis of the TCGA dataset shows that... BMPR1A Patients with missing [features] exhibited shorter disease-free survival. Figure 58 ).

[0225] To investigate whether BMP signaling pathway inactivation promotes AR-independent progression, a group of prostate cancer cell lines molecularly classified as ARPC, NEPC, and DNPC were examined. Gene expression analysis revealed... FST and AR There was a negative correlation between expression levels, with PC3 cells exhibiting the highest level of expression characteristic of DNPC. FST level( Figure 6 Because long-term use of next-generation AR pathway inhibitors can lead to lineage shifts in cancer cells, resulting in anti-AR resistance, the inventors treated AR-positive LNCaP cells with enzalutamide to detect whether changes in BMP signaling were accompanied by AR inhibition. Notably, AR blockade led to… FST The expression gradually increased over time and then decreased. BMPR1A and BMPR1B Instead BMPR2 The expression ( Figure 7 Conversely, activation of AR signaling with dihydrotestosterone (DHT) can induce... BMPR1A and BMPR1B Simultaneous expression and inhibition FST Transcription ( Figure 8 To achieve BMP signal inhibition in the same system, the inventors treated LNCaP cells with LDN-193189 and observed... AR and its target genes include KLK2 , KLK3 and TMPRSS2 The downward adjustment ( Figure 9 These results demonstrate the existence of a positive feedback loop supporting cavity surface spectrum identity between AR and BMP signals, where interference with one pathway affects the normal function of the other.

[0226] To further verify FST The association between expression, BMP activity, and AR dependence was revealed by Pearson correlation analysis on the SU2C cohort. FST Expression and KLK3 negative correlation between them Figure 59 ).also, BMPR1A and BMPR1B and KLK3 The positive correlation further supports the positive modulation relationship between BMP and AR signals. Figure 59Consistent with these correlations, the inventors analyzed single-cell RNA sequencing (scRNA-seq) data from 11 prostate cancer patients, including 3 cases of metastatic hormone-sensitive prostate cancer (mHSPC), 6 cases of metastatic castration-resistant prostate cancer (mCRPC), 1 case of non-metastatic HSPC with mixed neuroendocrine features, and 1 patient of unclassified origin. Four samples were collected from Changhai Hospital as part of this study, and seven samples were from previously published cohorts. Notably, FST and AR / KLK3 The expression of these substances in the cancer cells of these patients exhibited a mutually exclusive pattern. Figure 60 and Figure 61 , Figure 10 To further confirm this correlation at the protein level, the inventors collected tumor and serum samples from patients at Changhai Hospital, naming it the Changhai Cohort, and compared FST and AR / PSA levels. In formalin-fixed prostate tumor tissue, the inventors detected FST and AR / PSA expression using multiplex immunohistochemical staining. FST staining was observed in tumor cells in 4 out of 7 cases, and FST-positive tumor areas were AR-negative, with weak or absent PSA expression. Figure 11 The remaining three FST-negative cases showed strong AR and PSA expression (). Figure 11 This confirms the mutually exclusive expression pattern.

[0227] For patient serum analysis, the test cohort included 7 DNPC patients and 10 ARPC patients from the CPGEA dataset, with matched RNA sequencing data and defined patient subtypes. Serum FST levels were significantly higher in DNPC samples than in ARPC samples, with mean values ​​of 5.3 ng / mL and 3.8 ng / mL, respectively. Figure 12 This supported the gene expression results from the dataset analysis. Subsequently, the inventors included 102 patients in a larger cohort (Changhai cohort) to assess the potential of FST levels as a biomarker indicating AR-independent progression. Patients with PSA < 7 ng / mL had significantly higher serum FST levels (mean 12.7 ng / mL), while patients with PSA > 7 ng / mL had a mean of 5.59 ng / mL. Figure 13 Both cohorts showed a significant negative correlation between serum FST and PSA levels. Figure 12 and Figure 13 Furthermore, FST as a biomarker has good specificity and sensitivity. Figure 14 In summary, these clinical correlations between BMP activity, FST expression, and the development of AR-independent prostate cancer propose a compelling hypothesis: FST may play a key role in mediating resistance to AR-targeted therapy.

[0228] FST inhibits BMP11-mediated AR expression.

[0229] Since FST is negatively correlated with AR dependence, the inventors hypothesized whether FST affects AR signaling in a prostate cancer cell line model sensitive to the AR antagonist enzalutamide. Using LNCaP and VCaP cells, the inventors overexpressed two independent FST isoforms: a full-length version encoding a 344-amino acid preprotein (FST344), and a 317-amino acid truncated form lacking exon 6 (FST317). Batch RNA sequencing and pathway analysis of FST-overexpressing VCaP cells showed that the upregulated pathways mainly involved genes related to lineage development and metabolic regulation, while the downregulated pathways were related to immune responses and cell transport. Figure 15 Interestingly, the inventors observed a significant decrease in AR signaling activity after FST expression. Figure 15 and Figure 16 This was verified by the downregulation of AR and its target gene mRNA and protein levels in LNCaP and VCaP cells, and was confirmed by both RT-qPCR and Western blot analyses. Figure 17 and Figure 18 Treatment of VCaP cells with recombinant FST protein also resulted in a dose-dependent decrease in AR and PSA levels. Figure 62 This confirms that both endogenous and exogenous FST can suppress AR signals.

[0230] To assess whether the effect of FST on AR signaling is dependent on BMP, the inventors first analyzed the expression of AR and target genes after BMP receptor knockout. In LNCaP cells, shRNA-mediated... BMPR1A / 1B Knockdown leads to decreased expression of AR and target genes. Figure 19 and Figure 20 ).also, SMAD1 / 5 / 9 Knockdown also reduced the expression of AR and downstream targets. Figure 21 and Figure 22 Similar effects were observed in silencing BMP receptors and SMADs in VCaP cells. Figures 63-67 These results indicate that the intact BMP signaling pathway is essential for maintaining AR expression and activity in prostate cancer cells. In contrast, inactivation of the TGFβ signaling pathway did not affect AR expression. Treatment with the selective TGFβ type I receptor inhibitor SB-431542 failed to inhibit AR expression. Figure 68 The depletion of SMAD2 / 3 or TGFβ receptors also had no effect. Figures 69-71These findings suggest that FST may specifically inhibit AR signaling by suppressing the BMP pathway rather than through general inhibition of the TGFβ superfamily. To further verify this hypothesis, the inventors used a constitutively activated BMPR1A mutant (Q233D). Overexpression of the Q233D mutant, rather than wild-type BMPR1A, induced elevated p-SMAD1 / 5 / 8 and AR levels in VCaP cells. Figure 72 ), and reversed FST-mediated downregulation of AR expression ( Figure 23 In summary, the inventors' data identify FST as a negative regulator of AR signaling by specifically inhibiting the BMP pathway.

[0231] As a secreted protein, most of the effects of FST are mediated by its ability to bind and neutralize various members of the TGFβ superfamily ligands. To determine which specific ligand FST regulates AR signaling, the inventors performed mass spectrometry analysis on FST-binding proteins in VCaP cells. The inventors identified BMP11 and GDF8 as TGFβ ligand proteins that bind to FST in conditioned medium. Figure 24 Among the TGFβ family members, BMP11 (also known as GDF11) and GDF8 are two highly homologous proteins, sharing approximately 90% sequence identity. Mature BMP11 and GDF8 can both bind to activin type II receptors and activate downstream signaling via SMAD2 / 3 phosphorylation. BMP11 has also been reported to stimulate BMP signaling via SMAD1 / 5 / 8 phosphorylation. To verify whether these two ligands are involved in regulating AR expression, the inventors examined the expression of AR and target genes after inactivation of the two ligand genes. Knockdown was performed in multiple AR-positive prostate cancer cell lines, including LNCaP, VCaP, and CWR22Pc cells. BMP11 Both lead to downregulation of AR and target gene expression. Figure 73 , Figure 25 and Figure 26 In contrast, GDF8 depletion failed to reproduce this effect. Figure 74 and Figure 75 As an additional control, the inventors knocked down its subunit statin β chain ( INHBB It disrupted activin expression but had no effect on AR expression. Figure 76 and Figure 77 These results further support the conclusion that inhibiting BMP activity, rather than TGFβ activity, leads to AR signal inactivation, and specifically identify BMP11 as a key ligand in this regulatory axis.

[0232] The inventors then focused on characterizing the role of the FST-BMP11 axis in AR regulation. Co-immunoprecipitation confirmed the binding between FST secreted in the supernatant of 293T cells and BMP11. Figure 78This validated the mass spectrometry findings. Recombinant BMP11 treatment increased AR and PSA levels in LNCaP and VCaP cells, while FST overexpression counteracted this stimulatory effect. Figure 27 Conversely, BMP11 treatment counteracted FST-mediated downregulation of AR and PSA levels in a dose-dependent manner. Figure 28 This demonstrated the antagonistic relationship between the two factors. Finally, the inventors constructed the FST C56Y mutant to disrupt the BMP11 antagonistic effect and demonstrated that this mutant failed to antagonize BMP11-induced AR expression ( Figure 29 In summary, the data demonstrate that FST antagonizes BMP11-mediated BMP signaling activation, which is essential for maintaining AR expression and activity in prostate cancer cells. This mechanistic discovery reveals a previously unknown regulatory loop whereby FST may promote AR-independent growth by disrupting BMP11-dependent AR signaling maintenance, potentially facilitating the development of treatment-resistant prostate cancer phenotypes.

[0233] FST drives resistance to AR targeted therapy

[0234] FST-mediated downregulation of AR expression may lead to loss of anti-AR therapy targets, thereby conferring resistance to AR-targeted therapies. To verify this hypothesis, AR-positive prostate cancer cells stably expressing FST were constructed and their response to enzalutamide treatment was evaluated. In VCaP cells, FST overexpression significantly enhanced cell resistance to enzalutamide, with a half-maximal inhibitory concentration (IC50) of approximately 16 μM, compared to 0.12 μM in control cells. Figure 30 Increased resistance to enzalutamide was also observed in LNCaP cells, with FST-overexpressing cells showing an IC50 value exceeding 6 μM, compared to 0.28 μM in control cells. Figure 79 To test the specific effect of FST against AR resistance, the inventors also expressed follicle-stimulating factor 1 (FSTL1) in two cell types. FSTL1 is a member of the secretory follicle-stimulating factor family and also has an inhibitory effect on BMP signaling. However, unlike FST, FSTL1 failed to induce enzalutamide resistance. Figure 80 and Figure 81 This demonstrates the specificity of the FST resistance-promoting effect.

[0235] To determine whether FST-induced resistance to anti-AR therapy is mechanistically related to BMP signaling inhibition, the inventors examined the effect of BMP pathway inactivation on sensitivity to AR inhibitors. Depletion of SMAD1 / 5 / 9 or BMPR1A / B in VCaP and LNCaP cells significantly increased the IC50 of enzalutamide (…). Figure 31 and Figure 32 , Figure 82 and Figure 83The effect of FST overexpression was reproduced. In contrast, loss of SMAD2 / 3 or TGFβ receptors had no effect on enzalutamide sensitivity. Figure 84 and Figure 85 This confirmed the specificity of the BMP pathway in mediating drug resistance. Furthermore, the introduction of a constitutively activated BMPR1A Q233D mutant resensitized FST-overexpressing VCaP cells to AR inhibition. Figure 33 This demonstrates that restoring BMP signaling can overcome FST-mediated drug resistance. Consistent with the inventors' discovery that BMP11 is a key ligand mediating the FST effect, BMP11 knockdown confers enzalutamide resistance. Figure 34 The loss of GDF8 or activin has minimal impact. Figure 86 and Figure 87 Furthermore, the BMP11-binding defective FST C56Y mutant showed impaired ability to induce enzalutamide resistance. Figure 35 and Figure 36 These results collectively demonstrate that FST antagonizes BMP11-induced BMP signaling activation, which is essential for FST-mediated anti-AR resistance.

[0236] Tumor cell growth analysis showed that although FST overexpression had no effect on baseline VCaP or LNCaP cell proliferation ( Figure 37 , Figure 88 However, under AR inhibitor treatment, it significantly promoted cell growth ( Figure 37 and Figure 38 , Figure 88 and Figure 89 Under three-dimensional organoid culture conditions, FST enhanced spheroid formation in both the presence and absence of enzalutamide. Figure 39 This indicates that FST not only promotes drug resistance but also enhances the stem cell-like characteristics of prostate cancer cells. To evaluate the in vivo effects of FST on resistance to AR-targeted therapy, the inventors used VCaP and LNCaP xenograft models. Tumor cells were inoculated after castration (androgen deprivation therapy, ADT) and immediately followed by enzalutamide treatment to simulate a clinical scenario of combined hormone therapy. Notably, under combined ADT and enzalutamide treatment, wild-type VCaP and LNCaP tumors were completely suppressed, while increased FST expression led to complete resistance to enzalutamide in castrated mice. Figure 40 and Figure 41 In summary, these comprehensive in vitro and in vivo results demonstrate that FST confers significant resistance to AR-targeted therapies through its ability to inhibit BMP signaling activity.

[0237] Targeting FST reverses anti-AR resistance and inhibits DNPC metastasis

[0238] To explore the therapeutic potential of FST inhibition in DNPC models, the inventors utilized protein engineering techniques to create FST-Trap (SEQ ID NO: 1)—a novel formulation designed to neutralize FST function. Considering that BMP11 activates the BMP and TGFβ signaling pathways, and that FST binds to multiple TGFβ superfamily members (with activin as the highest affinity ligand), the inventors engineered a fusion protein consisting of human activin A linked to the Fc domain of human immunoglobulin G1 (IgG1). To prevent unwanted Fc-FcR interactions, positions 234 and 235 were mutated to alanine, and position 329 was mutated to glycine. Figure 42 Furthermore, the inventors introduced the M418A mutation to eliminate the ability of activin-Fc to activate TGFβ signaling, while retaining its FST-binding ability (FST-Trap). Figure 42 Validation studies confirmed that wild-type, but not M418A, mutants activate phosphorylated SMAD2 in 293T cells. Figure 90 The M418A mutant showed FST binding affinity similar to that of wild-type activin-Fc. Figure 91 ).

[0239] SEQ ID NO: 1: MDWTWRILFLVAAATGAHSSPTPGSEGHSAAPDCPSCALAALPKDVPNSQPEMVEAVKKHILNMLHLKKRPDVTQPVPKAALLNAIRKLHVGKVGENGYVEIEDDIGRRAEMNELMEQTSEIITFAESGTARKTLHFEISKEGSDLSVVERAEVWLFLKVPKANR TRTKVTIRLFQQQKHPQGSLDTGEEAEEVGLKGERSELLLSEKVVDARKSTWHVFPVSSSIQRLLDQGKSSLDVRIACEQCQESGASLVLLGKKKKKEEEGEGKKKGGGEGGAGADEEKEQSHRPFLMLQARQSEDHPHRRRRRGLECDGKVNICCKKQFFVSFKD IGWNDWIIAPSGYHANYCEGECPSHIAGTSGSSLSFHSTVINHYRMRGHSPFANLKSCCVPTKLRPMSMLYYDDGQNIIKKDIQNAIVEECGCSGGGGSEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDG VEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALGAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK The inventors expressed and purified FST-Trap ( Figure 92 The FST neutralizing activity was tested in vitro. In 293T cells co-expressing BMP11 and FST, FST-Trap effectively eliminated the binding of FST to BMP11 in conditioned medium. Figure 43 FST-Trap treatment alone had no effect on BMP11-mediated BMP signaling activation and AR upregulation, but completely reversed the inhibitory effect of FST on BMP11-induced BMP activation and AR upregulation. Figure 44 and Figure 45 These findings demonstrate that FST-Trap provides potent FST neutralizing activity while maintaining efficient FST binding, and does not interfere with normal TGFβ signaling. Figure 90 ).

[0240] To test the efficacy of FST-Trap in reversing drug resistance, the inventors first evaluated its effect in an enzalutamide-resistant cell model. In FST-overexpressing VCaP cells, FST-Trap treatment significantly resensitized the cells to enzalutamide. Figure 46 In castrated SCID mice, mice carrying 100 mm³ FST-expressing VCaP or LNCaP tumors received intravenous FST-Trap or solvent control combined with enzalutamide every other day. Figure 47 Compared with the enzalutamide-only group, the combination therapy of FST-Trap and enzalutamide resulted in significant tumor growth inhibition in both VCaP and LNCaP xenografts. Figure 48 and Figure 49 ), without affecting the mouse's weight ( Figure 95 Immunohistochemical analysis showed that tumor angiogenesis was significantly inhibited in the FST-Trap combined group, as evidenced by a decrease in CD31-positive staining. Figure 50 and Figure 51 We also observed increased staining for p-SMAD1 / 5 / 8 and cleavage-type caspase 3, with minimal effect on tumor cell proliferation. Figure 50 and Figure 51 This indicates that FST-Trap treatment effectively reversed the FST-mediated effects in vivo.

[0241] To further investigate the therapeutic effect of FST-Trap in the DNPC model, the inventors utilized PC3M and RM1 models. FST-Trap significantly inhibited tumor spheroid formation in PC3M cells. Figure 52 At concentrations as high as 5 μg / mL, it does not affect 2D cell growth. Figure 96 This indicates selective targeting of stem cell-like properties rather than general cytotoxicity. To test the preclinical activity of FST-Trap as a single drug in the metastatic setting, the inventors intracardiacly injected PC3M cells into mice, and FST-Trap was administered at a dose of 300 μg starting on day 7. Administration of FST-Trap starting on day 7 effectively prevented the formation of bone metastases in nude mice. Figure 53 This demonstrates its anti-metastatic efficacy as a monotherapy. Given the inventors' discovery that FST depletion can induce the accumulation of effector T cells in tumors, they tested the hypothesis that targeting FST could improve the efficacy of checkpoint immunotherapy. C57BL / 6J mice were subcutaneously inoculated with RM1 cells and treated with FST-Trap, anti-PD-1 antibody, or a combination thereof. Combination therapy significantly enhanced the inhibition of tumor growth, superior to FST-Trap or PD-1 antibody alone. Figure 54 , Figure 97Endpoint tumor analysis showed a significant accumulation of CD8-positive T cells, an increase in apoptotic tumor cells, and a decrease in CD31-positive endothelial cell infiltration in the combination therapy group. Figure 55 These findings demonstrate that FST-Trap can be effectively combined with checkpoint immunotherapy to inhibit prostate cancer progression by remodeling the tumor immune microenvironment.

[0242] The inverse correlation between FST and AR / PSA levels suggests that patients with high FST and high PSA may represent a mixed tumor subtype requiring early intensive treatment. Given that FST promotes stem cell-like characteristics and lineage plasticity—features associated with chemotherapy responsiveness—elevated FST levels may identify patients who will benefit from early multi-chemotherapy. This represents a paradigm shift from a reactive approach after resistance develops. The secretory nature of FST makes it an ideal biomarker for clinical implementation, complementing existing transcriptome-based classifiers. Future prospective studies should evaluate whether FST levels can guide treatment selection toward early docetaxel combination therapy and improve patient outcomes.

[0243] In summary, this work establishes FST as a key factor mediating the progression of treatment-resistant and aggressive diseases through tumor intrinsic mechanisms and microenvironment remodeling. The stage-specific function of BMP signaling helps resolve long-standing controversies and provides a framework for therapeutic targeting. The FST-Trap demonstrates the feasibility of precise targeting of the TGFβ superfamily. Most importantly, FST, as a non-invasive biomarker, opens new possibilities for precision medicine approaches that can prevent rather than address treatment resistance, potentially fundamentally improving the prognosis of patients with advanced prostate cancer.

[0244] Example 2: FST Serum ELISA Detection Kit

[0245] 2.1 Reagent Kit Components

[0246] This embodiment provides an enzyme-linked immunosorbent assay (ELISA) kit for detecting the level of follicle-staphylin (FST) protein in human serum, comprising the following components: (1) Pre-coated microplate: 96-well polystyrene microplate, with the bottom of each well pre-coated with anti-human FST monoclonal capture antibody (recognizing the N-terminal follistatin domain FSD1-FSD2 region of FST). This antibody can simultaneously recognize the two major subtypes, FST317 and FST344. The coating concentration is 2-8 μg / mL, and the plate is stored at 4°C.

[0247] (2) Detection antibody: Biotin-labeled anti-human FST polyclonal detection antibody (recognizes the C-terminal region of FST), which can distinguish between free FST and FST-ligand complex, and preferentially detects free FST. It is provided in liquid form and should be diluted 1:100 before use.

[0248] (3) Standards: Recombinant human FST317 protein standard, in lyophilized powder form, provided with 7 concentration gradients (0, 0.5, 1.0, 2.0, 4.0, 8.0, 16.0 ng / mL) for plotting standard curves. The reason for choosing FST317 as the standard is that FST317 is the major subtype in circulating blood, and recombinant FST317 protein has been used as a reference in ELISA validation experiments.

[0249] (4) Enzyme label: horseradish peroxidase (HRP) labeled streptavidin.

[0250] (5) Chromogenic substrate: TMB (3,3',5,5'-tetramethylbenzidine) chromogenic solution A and B.

[0251] (6) Termination solution: 2M sulfuric acid solution.

[0252] (7) Buffer: concentrated wash buffer (20×PBS-T), sample dilution buffer (PBS buffer containing 1% BSA).

[0253] (8) Quality control materials: High-value quality control materials (FST concentration of approximately 6.0 ng / mL, simulating serum levels in DNPC patients) and low-value quality control materials (FST concentration of approximately 3.5 ng / mL, simulating baseline levels in ARPC patients). The concentration of quality control materials was set based on CPGEA cohort data: the average serum FST in DNPC patients was 5.3 ng / mL, and the average serum FST in ARPC patients was 3.8 ng / mL.

[0254] 2.2 Detection Method

[0255] Sample collection and processing: Collect 3-5 mL of peripheral venous blood from the patient into a coagulation tube. After coagulation at room temperature for 30 minutes, centrifuge at 1000×g for 15 minutes to separate the serum. Serum samples can be stored at -80°C, avoiding repeated freeze-thaw cycles (no more than 3 times). Before testing, bring the samples to room temperature and dilute them 1:2 with sample diluent.

[0256] Testing procedure: Step 1: Add 100 μL of standard, quality control or diluted sample to each well of the pre-coated microplate and incubate at 37°C for 2 hours.

[0257] Step 2: Discard the liquid and wash the plate 5 times with washing buffer, soaking for 30 seconds each time.

[0258] Step 3: Add 100 μL of biotin-labeled detection antibody working solution to each well and incubate at 37°C for 1 hour.

[0259] Step 4: Wash the plate 5 times. Add 100 μL of HRP-streptavidin working solution to each well and incubate at 37°C for 30 minutes.

[0260] Step 5: Wash the plate 5 times. Add 100 μL of TMB developer to each well and develop at 37°C in the dark for 15-20 minutes.

[0261] Step 6: Add 50 μL of stop solution to each well and read the absorbance at 450 nm on a microplate reader within 5 minutes (reference wavelength 630 nm).

[0262] Result calculation: A four-parameter Logistic curve was plotted with the standard concentration as the x-axis and the OD value as the y-axis to fit the standard curve. The FST concentration in the sample was calculated based on this curve and multiplied by the dilution factor to obtain the actual concentration.

[0263] 2.3 Detection Performance Indicators

[0264] The detection performance indicators are shown in Table 4.

[0265] Table 4 Detection Performance Indicators

[0266] 2.4 Clinical interpretation criteria (basis for determining cut-off value)

[0267] Based on serum test data from two independent clinical cohorts, this invention proposes clinical interpretation reference intervals as shown in Table 5.

[0268] Table 5 Clinical Interpretation Reference Intervals

[0269] Cut-off value setting basis: (1) In the CPGEA cohort, the average serum FST of ARPC patients (n=10) was 3.8 ng / mL, and the average serum FST of DNPC patients (n=7) was 5.3 ng / mL. Figure 12 (2) In the Changhai cohort (n=102), the mean serum FST in the PSA<7 ng / mL group (indicating low AR signaling activity / AR-independent tendency) was 12.7 ng / mL, and in the PSA>7 ng / mL group it was 5.59 ng / mL. Figure 13 (3) Both cohorts showed a significant negative correlation between FST and PSA levels. Based on the above data, using 5.5 ng / mL as the low / intermediate risk cutoff and 7 ng / mL as the intermediate / high risk cutoff can effectively distinguish between ARPC and DNPC-prone patients. The final cutoff value should be determined through larger-scale prospective clinical validation.

[0270] Example 3: FST / PSA Combined Detection Kit

[0271] 3.1 Experimental Design

[0272] Given that FST and PSA exhibit mutually exclusive expression patterns in prostate cancer subtypes ( Figure 10 , Figure 11 Direct evidence at both the single-cell and tissue levels), and a significant negative correlation between serum levels of both. Figure 12 , Figure 13 This embodiment provides a combined detection kit that integrates FST and PSA on the same detection platform, which can assist in the determination of prostate cancer subtype by the ratio or combination pattern of the two markers.

[0273] 3.2 Reagent Kit Composition

[0274] This kit employs a dual-channel ELISA design. The upper half (AD row, 48 wells) of a 96-well microplate is coated with anti-FST capture antibody, and the lower half (EH row, 48 wells) is coated with anti-PSA capture antibody. Each plate can simultaneously detect FST and PSA levels (including standards and controls) in up to 20 patient samples. The remaining components are similar to those in Example 1, with the addition of PSA standards (0-100 ng / mL concentration gradient) and anti-PSA detection antibody.

[0275] 3.3 Joint Interpretation Model

[0276] The FST / PSA joint interpretation model established based on the invention data is shown in Table 6.

[0277] Table 6 FST / PSA Joint Interpretation Model

[0278] Special Note – Clinical Value of the “Elevated FST + Decreased PSA” Pattern: Figure 7 It was clearly shown that FST expression in LNCaP cells gradually increased over time while BMP receptor expression decreased under enzalutamide treatment; Figure 42 Co-culture experiments demonstrated that FST-positive cells had a competitive advantage under enzalutamide pressure (expanding from 6% to over 20%). This means that during AR-targeted therapy, continuous monitoring of the dynamic changes in FST / PSA is more clinically significant than the absolute values ​​at a single time point. If a patient exhibits a "scissor-shaped" trend of gradually increasing FST and gradually decreasing PSA during treatment, even if the absolute values ​​of both indicators have not yet reached a clear abnormal threshold, it should raise a high degree of suspicion that the disease is undergoing an AR-independent transition, and adjustments to the treatment strategy should be considered in advance.

[0279] Example 4: FST detection for monitoring the efficacy of enzalutamide treatment and early warning of drug resistance.

[0280] 4.1 Applicable Scenarios

[0281] This embodiment relates to the specific application of the FST test kit in patients with advanced prostate cancer who are receiving enzalutamide (or other AR-targeted therapy drugs).

[0282] Criteria for scene selection: Figure 5 Cox regression analysis clearly showed that high FST mRNA levels were significantly associated with poor survival outcomes in patients treated with enzalutamide (HR > 1), but no equally significant association was observed in patients treated with abiraterone. This difference suggests that the predictive value of FST as a biomarker is particularly prominent in patients treated with direct AR antagonists (such as enzalutamide), possibly due to the difference in driving lineage transition between direct AR blockade and androgen synthesis inhibition.

[0283] 4.2 Monitoring Plan

[0284] Baseline detection: Before initiating enzalutamide treatment, serum samples were collected from patients to simultaneously measure FST and PSA levels, establishing individualized baseline values. Patients with baseline FST > 5.5 ng / mL suggest the possible presence of AR-independent components, which should be considered when developing a treatment plan (e.g., early combination with docetaxel chemotherapy).

[0285] Dynamic monitoring during treatment: Serum FST levels should be measured every 4-6 weeks in conjunction with PSA levels. Pay close attention to the following warning signs: (a) FST absolute value increases by more than 50% from baseline; (b) FST shows a continuous upward trend (two consecutive measurements show an increase compared to the previous one), especially when accompanied by a decrease or plateau in PSA; (c) FST exceeds the 5.5 ng / mL threshold.

[0286] Drug resistance early warning and intervention decision-making: When FST monitoring reaches the above warning criteria, it is recommended to: (a) perform imaging reassessment (paying particular attention to changes in bone metastases to confirm FST-driven osteolytic bone metastases); (b) consider tissue re-biopsy to confirm the tumor phenotype; and (c) assess the necessity of treatment regimen adjustment—data suggest that high FST expression is associated with enhanced stemness characteristics, and such patients may benefit from early docetaxel combination therapy.

[0287] 4.3 Clinical validation data support

[0288] The clinical feasibility of this embodiment is supported by the following data: (1) Cox analysis of 81 CRPC patients in the SU2C-2019 cohort confirmed that high FST mRNA levels were significantly associated with shortened overall survival, especially in the enzalutamide treatment subgroup. Figure 5 ); (2) Both the CPGEA cohort and the Changhai cohort, two independent cohorts, verified the negative correlation between FST and PSA at the serum protein level. Figure 12 , Figure 13 This demonstrates that discoveries at the gene level can be translated into serological testing. (3) In vitro cell experiments demonstrated that enzalutamide treatment directly induced an increase in FST expression. Figure 7 This provides direct evidence for the biological rationale of dynamic monitoring of FST during treatment; (4) FST overexpression increased enzalutamide IC50 from 0.12 μM to approximately 16 μM (VCaP cells, Figure 30 The FST level increased from 0.28 μM to over 6 μM (LNCaP cells), confirming a direct association between FST levels and enzalutamide resistance.

[0289] Example 5: Stratification Method for Prostate Cancer Patients Based on FST Detection

[0290] 5.1 Method Overview

[0291] This embodiment provides a method for stratifying prostate cancer patients using serum FST levels to aid in treatment decision-making. The method includes the following steps: Step 1: Collect serum samples from the patient; Step 2: Detect serum FST protein concentration using the ELISA kit described in Example 1 or other immunoassay methods (such as chemiluminescence immunoassay, electrochemiluminescence immunoassay, etc.); Step 3: Optionally, serum PSA concentration can be detected simultaneously; Step 4: Based on the FST level (and / or the FST / PSA combination pattern), patients are classified into the following levels (as shown in Table 7).

[0292] Table 7 Patient Tiers

[0293] 5.2 Biological mechanisms support

[0294] The biological rationale for the above stratification strategy is based on the discovery of the following mechanism: AR-BMP positive feedback loop: This confirms the existence of a positive feedback loop between AR and BMP signals to maintain the luminal differentiation identity of prostate cancer. DHT activation of AR signaling can upregulate BMPR1A / 1B and inhibit FST expression. Figure 8 Conversely, BMP signaling activation maintains AR expression through BMP11 (); Figure 27 , Figure 28 This bidirectional regulation means that disruption of any pathway will lead to the cascading inactivation of the other pathway.

[0295] Example 6: Multiplex Immunohistochemistry (mIHC) Detection Reagent Combination

[0296] 6.1 Experimental Design

[0297] In addition to serological testing, this invention also provides a combination of multiplex immunohistochemical (mIHC) detection reagents for prostate cancer tissue sections, which can directly determine the subtype composition of the tumor at the histological level by simultaneously detecting the in situ expression of FST, AR and PSA proteins on the same tissue section.

[0298] 6.2 Reagent Combination

[0299] This combination includes primary antibodies for the following biomarkers and corresponding fluorescently labeled secondary antibodies / detection systems: (1) Anti-FST antibody: used to label FST-positive (AR-independent) tumor regions; (2) Anti-AR antibodies: used to label AR-positive (AR-dependent) tumor regions; (3) Anti-PSA / KLK3 antibody: as a functional biomarker for verifying the activity of the AR signaling pathway; (4) Optional: Anti-SOX9 antibody: as an auxiliary indicator of lineage plasticity.

[0300] 6.3 Interpretation Criteria

[0301] based on Figure 11 Histological validation data: In 7 prostate cancer tissue samples, FST-positive tumor regions were observed in 4 cases, and these regions were all AR-negative and PSA weakly positive or negative; the remaining 3 FST-negative cases showed strong AR positivity and PSA positivity. This mutually exclusive expression pattern constitutes the core principle of mIHC interpretation, as shown in Table 8.

[0302] Table 8 Core Principles of mIHC Interpretation

[0303] Example 7: Extended Form of the Detection Method

[0304] The core of the FST detection described in this invention lies in detecting the level of FST protein or mRNA in biological samples. The detection method is not limited to ELISA, but may include, but is not limited to: (1) Chemiluminescence immunoassay (CLIA): Suitable for high-throughput clinical laboratories, it can realize a fully automated testing process and is suitable for large-scale routine screening.

[0305] (2) Electrochemiluminescence immunoassay (ECLIA): It can be integrated into existing mainstream immunoassay platforms such as Roche Cobas or Abbott Architect, which facilitates clinical promotion.

[0306] (3) Lateral flow immunochromatography (LFIA / colloidal gold test strip): It can be developed into a point-of-care testing (POCT) product to achieve semi-quantitative rapid interpretation of FST levels, which is suitable for primary healthcare and rapid intraoperative assessment.

[0307] (4) RT-qPCR or liquid biopsy (ctRNA / cfRNA): Detection was achieved by measuring FST mRNA levels in circulating tumor cells or cell-free RNA. RT-qPCR was used to verify changes in FST expression under different cell lines and treatment conditions. Figures 7-9 This provides a methodological basis for mRNA-level detection.

[0308] (5) Mass spectrometry-based proteomics methods: such as targeted proteomics using multiple reaction monitoring (MRM) or parallel reaction monitoring (PRM), can achieve highly specific quantification of FST in complex serum matrices. In this invention, mass spectrometry has been successfully used to identify the binding of FST to BMP11 in VCaP conditioned medium. Figure 24 This provides preliminary verification of the feasibility of the mass spectrometry method.

Claims

1. A biomarker for detecting prostate cancer, characterized in that, The marker is the FST antigen or an autoantibody that binds to it.

2. A combination of biomarkers for detecting prostate cancer, characterized in that, The combination of markers includes the markers as described in claim 1; Preferably, the FST protein is used to detect AR-independent prostate cancer or mixed prostate cancer containing AR-independent subtypes, such as neuroendocrine prostate cancer or double-negative prostate cancer. More preferably, the biomarker combination further comprises any one or more of the following biomarkers: AR antigen or autoantibody bound thereto, PSA / KLK3 antigen or autoantibody bound thereto, SOX9 antigen or autoantibody bound thereto, and PSA antigen or autoantibody bound thereto. More preferably, the biomarker combination comprises an FST antigen or an autoantibody binding thereto and a PSA antigen or an autoantibody binding thereto; or, The biomarker combination includes FST antigen or its binding autoantibody, AR antigen or its binding autoantibody, PSA / KLK3 antigen or its binding autoantibody, SOX9 antigen or its binding autoantibody, and PSA antigen or its binding autoantibody.

3. A kit for detecting AR-independent prostate cancer, characterized in that, The kit contains reagents for detecting the biomarker as described in claim 1 or the combination of biomarkers as described in claim 2; for example, the reagents are autoantibodies for antigen detection or antigens for autoantibody detection.

4. The kit according to claim 3, characterized in that, In the biomarker combination, the antigen or autoantibody further contains a tag peptide; the tag peptide preferably includes one or more of the following: His tag, streptavidin tag, avidin tag, biotin tag, GST tag, C-myc tag, Flag tag, and HA tag; the biomarker is more preferably expressed by Escherichia coli, yeast, insect cells, or animal cells, and / or the biomarker is purified by Ni affinity chromatography, ion exchange chromatography, molecular sieve, dialysis, ultrafiltration, or hydrophobic chromatography.

5. The kit as described in claim 3 or 4, characterized in that, The kit further includes one or more of the following: sample diluent, calibrator diluent, washing solution, buffer, anti-human IgG secondary antibody, calibrator, and quality control product; the kit preferably also includes a 96-well plate.

6. A detection method, characterized in that, The detection method uses the biomarker as described in claim 1, the combination of biomarkers as described in claim 2, or the kit as described in any one of claims 3 to 5 to detect the corresponding antigen or its binding autoantibody or the mRNA encoding it in the sample. The specific steps include: contacting the biomarker with the sample; if binding is detected, it indicates the presence of the corresponding antigen or its binding autoantibody; the detection method is selected from one or more of ELISA detection, multiplex immunohistochemistry detection, chemiluminescent immunoassay, electrochemiluminescent immunoassay, lateral flow immunochromatography, RT-qPCR or liquid biopsy, and mass spectrometry-based proteomics methods. Preferably, the detection method is ELISA detection, the sample is serum, and the interpretation criteria include: FST levels <5.5 ng / mL, 5.5 - 7.0 ng / mL and >7.0 ng / mL are interpreted as low level, medium level and high level, respectively; and / or, PSA levels <4.0 ng / mL, 4.0 - 9.0 ng / mL and >9 ng / mL are interpreted as low level, medium level and high level, respectively; Alternatively, the detection method is multiplex immunohistochemical detection, the sample is a tissue section, and the interpretation criteria include: FST being negative, positive, or focally positive; and / or AR being negative, strongly positive, or focally positive; and / or PSA being negative / weakly positive, strongly positive, or heterogeneous. More preferably, the detection method is for non-diagnostic purposes or for diagnostic purposes; More preferably, the diagnostic purpose is to diagnose prostate cancer subtypes, and the diagnostic criteria include: in the ELISA test, low FST levels indicate low risk, predominantly AR-dependent; medium FST levels indicate intermediate risk / gray zone, with a tendency for early AR-independent transformation; high FST levels indicate high risk, suggesting AR-independent / DNPC progression; or, low FST levels + high PSA levels indicate ARPC predominant, AR-dependent prostate cancer; high FST levels + low or medium PSA levels indicate a high probability of transitional / DNPC, with caution for AR-independent progression; high FST levels + high PSA levels indicate mixed / transitional type, possibly with mixed ARPC and DNPC prostate cancer; or... In the multiplex immunohistochemical assay, the following are indicated for AR-dependent prostate cancer: FST negative + strong AR positive, FST negative + strong PSA positive; the following are indicated for AR-independent prostate cancer: FST positive + AR negative, FST positive + PSA negative / weak positive; the following are indicated for mixed prostate cancer: FST focal positive + AR focal positive, FST focal positive + heterogeneous PSA.

7. The use of the biomarker as described in claim 1, the combination of biomarkers as described in claim 2, or the kit as described in any one of claims 3 to 5 in the preparation of reagents for detecting prostate cancer; Preferably, the reagent is used for the diagnosis or auxiliary diagnosis of AR-independent prostate cancer or mixed prostate cancer containing AR-independent subtypes, or for recurrence monitoring and / or prognostic monitoring after treatment for AR-independent prostate cancer; for example, monitoring the efficacy of enzalutamide treatment and early warning of drug resistance. More preferably, the prostate cancer also includes AR-dependent prostate cancer.

8. A prostate cancer diagnostic system, characterized in that, The prostate diagnostic system includes the following modules: (1) An input module, which is used to input the detection results of the markers as described in claim 1 or the combination of markers as described in claim 2 in the sample to be tested; (2) An interpretation module, which interprets the detection result according to the interpretation criteria in the detection method as described in claim 6; (3) Prostate cancer subtype indication module, which indicates prostate cancer subtype using the diagnostic criteria in the detection method as described in claim 6; Preferably, the prostate cancer diagnostic system further includes a suggestion module that provides suggestions for disease monitoring and / or treatment based on the prostate cancer subtype indicated in (3); for example, in the ELISA test, a low level of FST indicates low risk and routine follow-up is recommended; a medium level of FST indicates medium risk / gray zone and it is recommended to shorten the follow-up interval. Combined with the PSA trend, if the PSA shows a downward trend at the same time, high vigilance is required, and early combination with docetaxel chemotherapy should be considered. High FST levels indicate high risk; a tissue biopsy is recommended for confirmation and assessment of whether a treatment strategy adjustment is necessary. Alternatively, low FST levels combined with high PSA levels suggest predominantly ARPC; standard AR-targeted therapy is recommended, along with routine PSA monitoring. High FST levels combined with low or moderate PSA levels suggest a high probability of transitional / DNPC; a tissue biopsy is recommended, as AR-targeted therapy may have failed or is about to fail; chemotherapy-based regimens are recommended, and the possibility of combining immune checkpoint therapy with FST inhibitors should be assessed. High FST levels combined with high PSA levels suggest a mixed / transitional type; relying solely on PSA monitoring is unreliable, and close monitoring of FST dynamics is necessary. The multiple immunohistochemical assays indicated AR-dependent prostate cancer, suggesting that AR-targeted therapy is sensitive; they indicated AR-independent prostate cancer, suggesting that AR-targeted therapy may be ineffective; and they indicated mixed prostate cancer, suggesting the presence of coexisting ARPC and DNPC regions. More preferably, the prostate diagnostic system further includes a login module and / or a printing module; The login module requires a username and password, and the printing module can print the results generated by the input module, the interpretation module, the prostate cancer subtype prompting module, and the suggestion module.

9. A computer-readable medium, characterized in that, The computer-readable medium stores a computer program that, when executed by a processor, enables the functionality of the prostate cancer diagnostic system as described in claim 8.

10. A prostate cancer diagnostic device, characterized in that, include: (1) The computer-readable medium as claimed in claim 9; (2) A processor for executing computer programs to implement the functions of the prostate cancer diagnostic system.

Images