Therapeutic potential of SHCBP1 inhibitor AZD5582 in pancreatic cancer treatment

Abstract

Pancreatic cancer (PC) is a highly aggressive and deadly malignancy with limited treatment options and poor prognosis. Identifying new therapeutic targets and developing effective strategies for PC treatment is of utmost importance. Here, we revealed that SHCBP1 is significantly overexpressed in PC and negatively correlated with patient prognosis. Knockout of SHCBP1 inhibits the proliferation and migration of PC cells in vitro, and suppresses the tumor growth in vivo. In addition, we identified AZD5582 as a novel inhibitor of SHCBP1, which efficiently restrains the growth of PC in cell lines, organoids, and patient‐derived xenografts. Mechanistically, we found that AZD5582 induced the apoptosis of PC cells by inhibiting the activity of PI3K/AKT signaling and preventing the degradation of TP53. Collectively, our study highlights SHCBP1 as a potential therapeutic target and its inhibitor AZD5582 as a viable agent for PC treatment strategies.

In this study, the potential inhibitors of SHCBP1 were screened using computer virtual docking, cell activity detection, and micro‐heat surging, and the identified inhibitors showed good antitumor effects in tumor organoids, patient‐derived xenograft, and other models.

Abbreviations

DFS

Disease‐free survival

EdU

5‐Ethynyl‐2′‐deoxyuridine

gRNAs

guide RNAs

GTEx

Genotype‐tissue expression

IAPs

Inhibitor of apoptosis proteins

KPC

LSL‐KrasG12D/+, LSL‐Trp53R172H/+, Pdx1‐Cre

MOE

Molecular operating environment

OS

Overall survival

PC

Pancreatic cancer

PFI

Progression‐free interval

PLB

Protein–ligand binding

SHC

Src homolog and collagen homolog

SHCBP1

SHC SH2 domain‐binding protein 1

SPR

Surface plasmon resonance

TFBG

Theaflavin‐3,3′‐digallate

1. INTRODUCTION

Pancreatic cancer, also known as pancreatic ductal carcinoma, accounts for approximately 90% of primary malignant tumors in the pancreas. It has abundant mucins and a dense collagen fiber matrix, which contributes to local invasion and distant metastasis. ¹ The prognosis of PC is extremely poor, with a 5‐year survival rate of only 10%, ² despite targeted therapy and immunotherapy, such as olaparib, larotrectinib, entrectinib, and pembrolizumab, approved for subsequent therapy for metastatic PC. However, these personalized therapies are only applicable in a minority of patients with rare molecular subtypes involving BRCA1/2 mutant, NTRK gene fusion or high‐microsatellite instability. ³ , ⁴ Consequently, PC remains a lethal malignancy, and the available treatment options are still limited. Therefore, there is an urgent imperative to explore novel therapeutic targets and develop more effective and tolerable agents for PC.

The SHC protein, a scaffold protein primordially found in mammals, is associated with cell surface receptors. ⁵ As a downstream component of multiple tyrosine kinase receptors, SHC can respond to these receptor signals to regulate Ras/MAPK, PI3K/Akt and other signaling pathways, thereby affecting cell growth, differentiation, and prognosis. ⁶ SHCBP1 serves as a crucial adapter protein that interacts with the SH2 domain of the SHC protein, and is highly conserved across different species. ⁷ Emerging evidence indicates that dysregulation of SHCBP1 is implicated in the development and metastasis of various cancers, including breast, stomach, bladder, lung and pancreatic. ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² Overexpression of SHCBP1 has been associated with increased cell proliferation, migration, and invasion, as well as decreased apoptosis in cancers. Furthermore, our previous investigations have demonstrated that hyperactivation of the SHC1/SHCBP1/PLK1 signaling pathway plays a crucial role in the development of resistance to Herceptin, a HER2‐targeted drug, in gastric cancer. Importantly, TFBG has been identified as an inhibitor of the SHCBP1–PLK1 complex, which can effectively enhance the therapeutic effect of Herceptin on gastric cancer in vitro and in vivo. ⁹ Therefore, targeting SHCBP1 presents an attractive strategy to modulate an aberrant carcinogenic signal. However, the precise role and underlying mechanism of SHCBP1 in PC remain unclear.

In the present study, we investigated the expression levels and clinical significance of SHCBP1 in PC. Our findings unequivocally demonstrate the oncogenic role of SHCBP1 in PC. Notably, the downregulation of SHCBP1 expression yields significant inhibition of PC cell proliferation, migration, invasion and tumorigenesis. What is more, we screened a new inhibitor, AZD5582, that directly targets SHCBP1 from a library of over 9000 bioactive small molecules utilizing virtual screening and microscale thermophoresis. Importantly, the molecule demonstrated potent inhibitory effects on PC cells in both in vivo and in vitro experiments. These results suggest that the SHCBP1 targeted drug AZD5582 can be used as a potential drug for the treatment of PC, with significant clinical application value.

2. MATERIALS AND METHODS

2.1. SHCBP1 expression and survival analysis based on bioinformatics data

GEPIA2 (http://gepia.cancer‐pku.cn/index.html) is an online analysis tool that includes 9736 tumors and 8587 normal samples in The Cancer Genome Atlas (TCGA) and Genotype‐Tissue Expression (GTEx). ¹³ It can analyze the relationship between mRNA expression and the prognosis of various tumors. We selected “Expression DIY” in GEPIA2 to examine the expression of SHCBP1 in PAAD. Microarray data were downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/), specifically the dataset GSE28735. We performed an analysis using GEO2R to assess the paired expression of SHCBP1 in PC. The GSE28735 dataset consists of 45 pairs of matched normal samples and cancer samples. Additionally, we utilized TCGA database (https://portal.gdc.cancer.gov), which encompasses 183 PC samples and four normal samples. Our analysis involved downloading TCGA‐PAAD expression data along with corresponding clinical information. The stats package and the car package were selected for expression analysis of paired samples and unpaired samples. The pROC package was used for receiver operating characteristic curve (ROC) analysis of the data. The survival package was used for proportional hazard hypothesis testing and fitting survival regression to explore the relationship between SHCBP1 expression and OS, DSS, and PFI. The results were expressed in the survminer package, and all results were visualized using the ggplot2 package.

2.2. Human PC specimens

In total, 114 PC samples were obtained, and the corresponding clinical data were collected to analyze the expression of SHCBP1 in PC. None of these patients have received preoperative chemotherapy or radiotherapy. All specimens were obtained using protocols approved by the Human Research Ethics Committee of Lanzhou University Second Hospital. Informed consent was obtained from all the patients.

2.3. Cell culture

Human PC cell lines PANC‐1, MIA PaCa‐2, CFPAC‐1, and SW1990 were obtained from the cell bank of the Chinese Academy of Sciences (China). Human PC cell line Capan‐2 was purchased from Guangzhou Saiku Biotechnology Co., Ltd. (China), and BXPC‐3 and PANC02 were provided from Shanghai Fuheng Biotechnology Co., Ltd. (China). PANC‐1, MIA PaCa‐2, CFPAC‐1, and PANC02 were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum. BXPC‐3 and SW1990 were cultured in RPMI medium. All cell lines were authenticated using short tandem repeat (STR) profiling. Lipofectamine 2000 (Invitrogen), pLenti‐CRISPR‐V2, pSPAX2, and pMD2G plasmids were transiently transfected into the HEK293T cells. After 60 h of transfection, the supernatant in the culture plate was collected and centrifuged at 13,778 g. Then, 0.22‐μm filters were used to obtain and sterilize the lentivirus suspensions, which were then used to infect PANC‐1, CFPAC‐1 and PANC02 PC cells at a concentration of 1 × 10⁶/mL for 48 h. Subsequently, puromycin was chosen to identify successfully transfected cells, which were subjected to western blot assays to detect the efficiency of SHCBP1 knockout. The precise sequences of the guide RNAs (gRNAs) used for the CRISPR/Cas9‐mediated SHCBP1 knockout have been provided below. Notably, these distinct gRNA sequences were applied for human and mouse SHCBP1, respectively (for human SHCBP1: sgRNA1: CCCTTGCAATTGAGCATGTC; for mouse SHCBP1: sgRNA2: CATCCCTGAAGGCGGTCGTA).

2.4. Organoids studies

PC samples from KPC mice were minced and incubated at 37°C in a digestion medium (DMEM/F12 with 10% fetal bovine serum, 2 mM glutamine, 10.5 mM Y‐27632, 500 nM A83‐01, 100 mg/mL primocin, and 5 mg/mL collagenase II) for 1–2 h. Digested cells were collected by centrifugation and resuspended in Cultrex UltiMatrix RGF BME (Bio‐Techne, #BME001‐05). Organoids were cultured in PancreaCult™ Organoid Medium (STEMCELL Technologies, #100‐0820/1).

2.5. Immunohistochemistry

Tissue microarrays (TMAs) of human PC specimens were deparaffinized and rehydrated, followed by antigen retrieval. After anti‐SHCBP1 (Sigma, diluted at 1:200) antibody or anti‐Ki67 (Abcam, diluted at 1:200) antibody incubation, the slides were dehydrated and stabilized with mounting medium and the images were acquired with a KF‐PRO‐120 scanner (Konfoong, China). Staining intensity (0, 1, 2, and 3) and percentage of positive cells among cancer ducts (0%–100%) were evaluated by a pathologist at the Lanzhou University Second Hospital. The final histoscore (H‐score) was calculated by multiplying the staining intensity and the percentage of positive cells.

2.6. Western blotting

Cells were lysed with RIPA buffer and centrifuged at 13,000 g for 10 min. Protein lysates were separated using SDS/PAGE, blotted onto a PVDF membrane, and probed overnight at 4°C with antibodies against rabbit anti‐SHCBP1 (1:1000; Sigma), mouse anti‐GAPDH (1:1000, Proteintech) or rabbit anti‐BCL‐2 (1:1000; Proteintech), rabbit anti‐caspase 3 (1:1000; Abcam), rabbit anti‐caspase 3 cleave (1:1000; Abcam), rabbit anti‐caspase 9 (1:1000; Abcam), rabbit anti‐E‐cadherin (1:1000; Cell Signaling Technology), rabbit anti‐vimentin (1:1000; Cell Signaling Technology), rabbit anti‐AKT (1:1000; Cell Signaling Technology), rabbit anti‐p‐AKT (1:1000; Cell Signaling Technology), rabbit anti‐ERK (1:1000; Proteintech), rabbit anti‐p‐ERK (1:1000; Cell Signaling Technology), rabbit anti‐PI3K (1:1000; Cell Signaling Technology), rabbit anti‐p‐PI3K (1:1000; Cell Signaling Technology). Stripped membranes were probed with a secondary antibody of goat anti‐mouse or anti‐rabbit IgG (1:10,000, Bioworld) and then visualized with enhanced chemiluminescence.

2.7. Cell viability and colony formation analysis

Cell viability was determined using the MTT assay. Approximately 2 × 10³ cells were seeded into a 96‐well plate and cultured for 0, 24, 48, or 72 h. Subsequently, 20 μL MTS reagent was added to each well and incubated with the cells at 37°C for 2 h. Finally, a microplate reader was used to measure the absorbance values of the liquid in the wells at 490 nm. All experiments were performed in duplicate with three replicate wells per group. In addition, approximately 2000 PC cells were evenly seeded in 35 mm Petri dishes. After culturing for approximately 2 weeks, the wells were washed with PBS, fixed with 4% paraformaldehyde and stained with 0.05% crystal violet. The acquired images were processed using ImageJ software and analyzed with GraphPad Prism 8.0.

2.8. Transwell invasion and cell scratch migration analysis

Approximately 200 μL PC cell suspension (1 × 10⁵ cells/mL) was carefully seeded into the Matrigel glue‐precoated upper chamber of a Transwell. Serum‐free medium was added to the upper chamber, and medium supplemented with 10% FBS was placed in the lower chamber. After 48 h of incubation, the cells were washed with PBS, wiped gently with cotton swabs, fixed with 4% paraformaldehyde, and stained overnight with 0.05% crystal violet. The stained cells at the bottom of the invasion chamber were photographed using a microscope. The data were processed using ImageJ software and GraphPad Prism 8.0. In addition, approximately 3 × 10⁵ PC cells were seeded in a 6‐well plate. The next day, the plate was scraped with a 200 μL pipette tip to form parallel linear wounds. Cells were cultured with serum‐free medium and photographed using a microscope. After 48 h of cell culture, pictures were taken through the same field of view using the same microscope, and the scratch healing rate was calculated using ImageJ software and GraphPad Prism 8.0.

2.9. 5‐Ethynyl‐2′‐deoxyuridine staining assay

Cells were seeded in 20‐mm confocal dish and incubated with EdU for 2 h. They were then stained using an EdU analysis kit (BeyoClick, Shanghai, China, #C0078S) according to the manufacturer's instructions. The cells were detected using a Zeiss LSM 880 laser microscope (Carl Zeiss, Germany).

2.10. Virtual screening of inhibitor targeting SHCBP1

The AlphaFold‐predicted structure of SHCBP1 AF‐Q9BZQ2‐F1‐model_v4 was downloaded from https://alphafold.ebi.ac.uk/entry/Q8NEM2. ¹⁴ Virtual screening, protein docking pocket finding, and molecular docking simulations were conducted by the commercial software MOE (version 2022.02, Chemical Computing Group, Canada). The binding affinities of small molecules and proteins were predicted based on the Generalized‐born volume integral/Weighted surface area (GBVI/WSA dG), which is a scoring function of the free energy of the protein bound to each molecule. For small molecular docking with SHCBP1 protein, 1000 random poses of small molecular were generated and docked with the selected‐binding pocket of SHCBP1 with the rigid receptor model, and the pose with the lowest estimated free energy was analyzed and displayed.

2.11. Tumor xenograft experiments

All procedures involved in the animal experiment have been approved by the Ethics Committee of the Second Hospital of Lanzhou University. 1 × 10⁶ murine pancreatic adenocarcinoma cells (PANC02) were suspended in a mixture of 100 μL PBS and 100 μL Matrigel (obtained from Corning Incorporated, New York, NY, USA). Subsequently, this cell suspension was subcutaneously injected into female C57BL/6 mice aged 6–8 weeks (procured from Charles River in Beijing, China). Body weights were monitored every 3 days for 3 weeks until euthanasia, and the ethical point was defined as tumor size approaching 1500 mm³.

2.12. Patient‐derived xenograft (PDX) studies

Six‐week‐old female NOD‐PrkdcscidIl2rgem1/Smoc mice were purchased from the Shanghai Model Organisms Center, Inc. Samples from patients with PC were cut into small pieces (approximate diameter: 3 mm) and subcutaneously implanted into the axilla of mice. When the tumor volume reached 150 mm³, the mice were administered vehicle or 3 mg/kg AZD5582 by i.v. once a week for 3 weeks, and the tumors and body weights were monitored and measured every 3 days for 3 weeks until euthanasia. ¹⁵ The ethical point was defined as tumor size of approximately 1500 mm³.

2.13. Statistical analysis

SPSS 24.0 (IBM, Armonk, NY, USA) and GraphPad Prism 8.0 (San 1055 Diego, CA, USA) were used for statistical analysis and graphing. A two‐tailed Student's t‐test was used for the comparison of two groups and one‐way ANOVA with Tamhane's T2 or least significant difference (LSD) post hoc tests were performed for multiple groups. Bioinformatics data were processed in R (v3.6.3) language. The data are shown as the means ± standard deviation. Statistical significances are denoted as follows: *p < 0.05, **p < 0.01, ***p < 0.001.

3. RESULTS

3.1. SHCBP1 was overexpressed in PC and correlated with poor prognosis of patients with PC

First, based on bioinformatics analysis, we explored the expression level of SHCBP1 in PC and its relationship with the prognosis of patients. The results showed that SHCBP1 was overexpressed in both unpaired and paired tissues in PC (Figure 1A,B). ROCs indicated that the expression of SHCBP1 showed high sensitivity and specificity for detecting PC, yielding an impressive area under the curve (AUC) value of 0.972 (CI = 0.954–0.990) (Figure 1C). This underscores the potential role of SHCBP1 as a candidate tumor marker for PC. Furthermore, Kaplan–Meier analysis revealed that high SHCBP1 expression was correlated with unfavorable OS (HR = 1.57, 1.04–2.37, p = 0.035), DFS (HR = 1.8, 1.21–2.28, p = 0.015), and PFI (HR = 1.57, 1.04–2.37, p = 0.035) (Figure 1D–F). To validate the overexpression of SHCBP1 protein in clinical PC samples, we performed immunohistochemical analysis on 114 cases of PC tissues and 100 cases of adjacent tissues. The outcomes unequivocally revealed that SHCBP1 was significantly upregulated in the PC samples (p = 0.007) (Figure 1G,H). We further analyzed the correlation between SHCBP1 levels and clinicopathological characteristics of 114 patients with PC. Our findings indicated that there were significant correlations between SHCBP1 levels and AJCC stage (p = 0.034), CA199 value (p = 0.029) (Table 1). Western blot analysis of normal and PC tissues from patients, along with immunohistochemistry (IHC) of normal and PC tissues from KPC mice, consistently affirmed the high expression of SHCBP1 in PC (Figure 1I–K). Furthermore, we found that SHCBP1 was expressed lowly in normal pancreatic ductal cells (HPNE) and human pancreatic duct epithelial (HPDE) cells, but significantly overexpressed in PC cells, especially in PANC‐1 and CFPAC‐1 cells (Figure 1L). Therefore, these findings suggested that SHCBP1 is prominently overexpressed in PC, which may play a role in PC development.

FIGURE 1.

SHCBP1 was overexpressed in PC and correlated with the poor prognosis of patients with PC. (A) SHCBP1 was overexpressed in the PC tissue data from TCGA database compared with normal tissues. (B) SHCBP1 was overexpressed in the PC tissue data from the GEO database compared with adjacent normal tissues (GSE28735). (C) ROC curve analysis depicting the predictive performance of SHCBP1 in PC using data from TCGA database. (D–F) Kaplan–Meier survival analysis of overall survival (OS), disease‐free survival (DFS), and progression‐free interval (PFI) in PC patients with high and low SHCBP1 tumor expression. (G, H) Representative H&E and IHC staining of SHCBP1 in PC TMAs. (I, J) Representative H&E and IHC staining of SHCBP1 in the pancreas of KPC spontaneous tumor mice and normal C57 mice. (K) SHCBP1 western blotting analysis of PC and adjacent normal tissues from four patients was revealed. (L) SHCBP1 expression was detected from HPDE, HPNE, and six PC cell lines by western blotting.

TABLE 1.

The correlation between SHCBP1 level and age, sex, AJCC stage, differentiation degree, diameter of tumor and CA19‐9 value.

Clinico‐pathologic	SHCBP1 low (≤110), N = 60		SHCBP1 high (>110), N = 54		p‐value	Chi‐squared
Characteristic	Numbers	Percentage	Numbers	Percentage
Age
≤60	36	60%	34	63%	0.746	0.105
>60	24	40%	20	37%
Gender
Male	37	61.70%	26	48.1%	0.147	2.101
Female	23	38.30%	28	51.9%
AJCC stage
I–IIA	27	45%	14	25.9%	0.034	4.49
IIb–III	33	55%	40	74.1%
Differentiation degree
Poorly	22	36.70%	24	44.4%	0.398	0.714
Moderately and well	38	63.3%	30	55.60%
Diameter of tumor
≤4	35	58.30%	30	55.6%	0.765	0.089
>4	25	41.70%	9	44.4%
CA19‐9
≤500	45	75%	30	55.6%	0.029	4.774
>500	15	25%	24	44.4%

3.2. SHCBP1 knockout inhibited the invasion, migration, and proliferation of PC cells

To gain further insights into the biological function of SHCBP1 in the progression of PC, PANC‐1 and CFPAC‐1 cells were selected for further investigation due to their high levels of SHCBP1 expression. Subsequently, SHCBP1 knockout cell lines were successfully generated for both PANC‐1 and CFPAC‐1 using CRISPR/Cas9‐mediated gene editing, and the efficiency of the knockout was confirmed through western blotting (Figure 2A). We also conducted Transwell cell invasion assays to investigate the role of SHCBP1 in PC cell invasion. Notably, SHCBP1 knockout significantly decreased the number of PANC‐1 and CFPAC‐1 cells penetrating the lower chamber (Figure 2B,C). Additionally, we evaluated the impact of SHCBP1 knockout on the migration ability of pancreatic cells using a plate scratch assay. The results indicated that the SHCBP knockout exhibited lower healing ability compared with the control group (Figure 2D,E). Thus, we evaluated the expression of epithelial–mesenchymal transition (EMT)‐associated factors in SHCBP1 knockout PANC‐1 and CFPAC‐1 cells. The results revealed that knockout of SHCBP1 upregulated E‐cadherin expression and downregulated vimentin expression (Figure S1). These findings suggest that SHCBP1 plays a role in regulating the migration and invasion of PC cells.

FIGURE 2.

SHCBP1 knockout inhibited the proliferation, invasion and migration of PC cells. (A) SHCBP1 depletion in PANC‐1 and CFPAC‐1 cells was confirmed by western blotting. (B, C) Transwell invasion assay and statistical analysis detected the impact of SHCBP1 knockout on the invasion ability of PANC‐1 and CFPAC‐1 cell lines. (D, E) Cell scratch migration assay and statistical analysis detected the impact of SHCBP1 knockout on the migration ability of PANC‐1 and CFPAC‐1 cell lines. (F–J) MTT, colony formation, and 5‐ethynyl‐2′‐deoxyuridine (EdU) proliferation assays and statistical analysis detected the impact of SHCBP1 knockout on the proliferation ability of PANC‐1 and CFPAC‐1 PC cell lines. (K–M) SHCBP1 knockout inhibited tumor volume and tumor weight compared with the control group.

Furthermore, MTT, colony formation, and EdU proliferation assays were performed to investigate the role of SHCBP1 in PC cell proliferation. Our findings revealed a significant suppression in the proliferation and colony formation of SHCBP1 knockout cells compared with the control groups in both PANC‐1 and CFPAC‐1 cells (Figure 2F–J). To further explore the role of SHCBP1 in PC progression in vivo, we conducted a knockout of SHCBP1 in PANC02 cells (Figure S2A). Subsequently, we subcutaneously injected 1 × 10⁶ WT or SHCBP1^−/− PANC02 cells subcutaneously into female C57BL/6J mice and monitored changes in tumor volume and body weight. Notably, the SHCBP1 knockout group exhibited significant inhibition in both tumor volume and tumor weight compared with the control group (Figure 2K–M), without any significant impact on the body weight of the mice (Figure S2B).

In summary, these findings underscore the pivotal significance of SHCBP1 in regulating PC cell invasion, migration, proliferation, and tumor growth. This study provides valuable insights into the potential therapeutic implications of targeting SHCBP1 in the treatment of PC.

3.3. AZD5582 was identified as a novel SHCBP1 inhibitor

The above results highlight the important function and prognostic significance of SHCBP1 in tumorigenesis of PC, suggesting its potential as a therapeutic target. However, no targeted inhibitor of SHCBP1 in PC has been reported to date. To address this gap, we conducted structure‐based virtual screening. As the crystal structure of the SHCBP1 protein is not currently available in the Protein Data Bank database, the AF‐Q9BZQ2‐F1‐model_v4, which was predicted by AlphaFold was adopted as the receptor model for inhibitor docking. We conducted a comprehensive screening of potential SHCBP1 inhibitors using MOE software, cell activity detection, and microscale thermophoresis. Initially, we analyzed the binding pockets of SHCBP1 using the SiteFinder module of MOE software, and five binding pockets of SHCBP1 were identified (Figure 3A). Detailed information regarding these binding pockets, including their PLB score and associated amino acid residues, was displayed (Table 2). Among these pockets, we selected pocket 1 with the highest PLB score as the docking pocket for subsequent SHCBP1 inhibitor screening. A library containing 9202 compounds with broad bioactivities was screened with MOE (Figure 3B). To prioritize potential inhibitors, we implemented the drug similarity principle and excluded small molecules with molecular weights outside the range 300–1400. This filtration resulted in 5875 remaining molecules, which were then docked with the binding pocket 1 of SHCBP1 using MOE‐DOCK. The affinity (S‐score) of protein–ligand interaction was calculated by MOE (Figure 3C). Based on the docking score, the top 50 compounds with S‐score lower than 9.2 kcal/mol were selected for subsequent biological activity detection, and these molecule compounds with S‐score, and molecular weight were presented in table (Table 3). Furthermore, an MTT assay was conducted to evaluate the effects of the 50 small molecules on the activity of the PC cell line PANC‐1 at a concentration of 10 μM. Markedly, our study revealed that five small molecules, namely AZD5582, ABT‐263, Dronedarone HCl, Carfilzomib, and L755507, displayed marked inhibitory effects on PANC‐1 cells (Figure 3D). Among these, microscale thermophoresis was used to assess their affinity toward SHCBP1. Interestingly, the results indicated that only AZD5582 exhibited a high affinity for SHCBP1 (Kd = 1.212 μM) (Figure 3E,F). In addition, we applied Discovery Studio (TOPKAT) to predict the toxicological properties of AZD5582. The results showed that it had no mutagenicity and developmental toxicity potential, and its inhalational LC50 and Oral LD50 to rats were 1.87 and 2.46 g/kg, respectively (Figure 3G).

FIGURE 3.

AZD5582 was identified as a novel SHCBP1 inhibitor. (A) Five binding pockets of SHCBP1 were identified using the SiteFinder module of MOE software. (B) Virtual docking screening of SHCBP1 inhibitor from a library containing 9202 compounds. (C) Docking score of 5875 small molecules with SHCBP1 using virtual screening. (D) The effect of top 50 small molecules on the activity of the PC cell line PANC‐1. (E) Molecular structure of AZD5582 small molecule. (F) The affinity of SHCBP1 to AZD5582 was detected by microscale thermophoresis. (G) AZD5582 toxicity as predicted by TOPKAT. (H) Computational model and the interaction of AZD5582 with SHCBP1 complex. (I) The interacting amino acids and hydrogen bonds between the complexes.

TABLE 2.

Details of five small molecule binding pockets of SHCBP1.

Site	Size	PLB	Hyd	Side	Residues
1	122	4.93	33	64	GLU236 ASP237 ARG238 TYR309 TYR313 LYS419 GLY420 ASP421 ASP442 ALA443 VAL444 GLU445 GLY446 ILE449 THR466 THR467 THR470 ARG472 LYS488 GLY489 ALA490 GLU493 TYR495 LYS511 LEU515 LYS517
2	103	2.24	31	52	LEU32 GLY35 LEU36 GLY53 SER54 SER55 LEU56 GLN57 GLN72 THR73 ASN74 GLN75 LEU76 LEU77 GLU80 ARG83 ALA84 TYR88 GLY186 ASP189 GLN190 LEU193 HIS197
3	23	2.01	14	36	ARG207 TRP209 ASP210 GLU213 CYS223 ARG227 TYR247 LEU251 LYS296 LEU299 LYS300 GLU303
4	31	1.48	11	33	PHE59 PHE71 GLN72 THR73 ASN74 ARG81 TRP179 GLU196 ARG199 PHE200 GLN203 ASN204 HIS231
5	20	0.96	12	24	PRO240 SER241 GLY242 LEU243 GLY312 TYR313 ASN316 ASP421 THR422 ASP425 GLU445 HIS451

TABLE 3.

Details of the top 50 compounds in the docking score.

S. no.	Compound name	Binding energy (kcal/mol)	Molecular weight	CAS
1	Leuprolide acetate	−12.05	1269	74381‐53‐6
2	Terlipressin acetate	−11.87	1227	14636‐12‐5
3	Bradykinin	−11.85	1060	58‐82‐2
4	Venetoclax	−10.99	868.4	1257044‐40‐8
5	MC‐VC‐PAB‐PNP	−10.76	737.8	159857‐81‐5
6	ABT‐737	−10.71	813.4	852808‐04‐9
7	DAPTA	−10.43	856.9	106362‐34‐9
8	Vipivotide tetraxetan	−10.41	1042	1702967‐37‐0
9	AZD5582	−10.34	1015	1258392‐53‐8
10	Felypressin acetate	−10.24	1040	56‐59‐7
11	Octreotide acetate	−10.24	1019	79517‐01‐4
12	Parishin A	−10.22	996.9	62499‐28‐9
13	Cenicriviroc	−10.21	696.9	497223‐25‐3
14	MS4078	−10.14	914.5	2229036‐62‐6
15	Angiotensin (1–7)	−10.08	899	51833‐78‐4
16	ABT‐263	−9.996	974.6	923564‐51‐6
17	GSK1904529A	−9.98	852	1089283‐49‐7
18	BAPTA‐AM	−9.954	764.7	126150‐97‐8
19	Lypressin acetate	−9.922	1056	50‐57‐7
20	Lanreotide	−9.879	1096	108736‐35‐2
21	Elbasvir	−9.828	882	1370468‐36‐2
22	A‐1210477	−9.809	850	1668553‐26‐1
23	Z‐DEVD‐FMK	−9.793	667.7	210344‐95‐9
24	Tariquidar	−9.775	646.7	206873‐63‐4
25	Atracurium besylate	−9.769	1243	64228‐81‐5
26	Crocin I	−9.697	977	94238‐00‐3
27	Setmelanotide	−9.668	1117	920014‐72‐8
28	Dronedarone HCl	−9.653	593.2	141625‐93‐6
29	Thymopentin	−9.642	679.8	69558‐55‐0
30	UNC1999	−9.622	569.7	1431612‐23‐5
31	BIBR‐1048	−9.609	627.7	211915‐06‐9
32	Batefenterol	−9.557	740.2	743461‐65‐6
33	PD173074	−9.55	523.7	219580‐11‐7
34	Forsythoside B	−9.544	756.7	81525‐13‐5
35	GHRP‐2	−9.534	746.9	158861‐67‐7
36	MS140	−9.531	760.8	2229974‐83‐6
37	Revefenacin	−9.489	597.7	864750‐70‐9
38	Carfilzomib	−9.456	719.9	868540‐17‐4
39	Atosiban acetate	−9.425	994.2	90779‐69‐4
40	Sulbutiamine	−9.378	702.9	3286‐46‐2
41	Quercetin 3‐O‐β‐d‐glucose‐7‐O‐β‐d‐gentiobioside	−9.367	788.7	60778‐02‐1
42	RG7112	−9.365	727.8	939981‐39‐2
43	LLY‐507	−9.347	574.8	1793053‐37‐8
44	BRD7552	−9.345	711.6	1137359‐47‐7
45	L755507	−9.339	584.7	159182‐43‐1
46	MT‐802	−9.309	787.8	2231744‐29‐7
47	JNK‐IN‐8	−9.297	507.6	1410880‐22‐6
48	Troxerutin	−9.28	742.7	7085‐55‐4
49	Sennoside A	−9.231	862.7	81‐27‐6
50	ERK5‐IN‐1	−9.21	638.8	1435488‐37‐1

These findings suggest that AZD5582 has a strong affinity with SHCBP1, underscoring its potential value as a potent SHCBP1 inhibitor. The computational model and interaction of AZD5582 and SHCBP1 complex were further performed (Figure 3H), and the interacting amino acids and hydrogen bonds between the complexes were shown (Figure 3I).

3.4. SHCBP1 inhibitor AZD5582 inhibited PC growth in organoid model

To further investigate the inhibitory effect of the SHCBP1 inhibitor AZD5582 on PC, we conducted an MTT assay to determine the IC₅₀ values in PANC‐1 and CFPAC‐1 cell lines. Markedly, the results revealed that the IC₅₀ values were 1.21μM and 74.03 nM, respectively (Figure 4A,B). Subsequently, the inhibitory effect of AZD5582 on PC cells at the IC20 concentration was assessed through colony formation and EdU assay. The results demonstrated a significant inhibitory effect of AZD5582 on the proliferation of PC cells (Figure 4C–G). Additionally, we constructed two organoid models of spontaneous pancreatic tumors in mice and evaluated the effect of AZD5582 on the growth of PC organoids at various concentrations. The findings revealed that AZD5582 had a significant inhibitory effect on PC organoids at concentrations ranging from 5 to 10 μM (Figure 4H,I). These results further support the potential of AZD5582 as a promising therapeutic agent for the treatment of PC.

FIGURE 4.

SHCBP1 inhibitor AZD5582 inhibited PC growth in the organoid model. (A, B) Determination of AZD5582 IC50 values in PANC‐1 and CFPAC‐1 cells. (C–G) Colony formation assays and 5‐ethynyl‐2′‐deoxyuridine (EdU) proliferation assays detected the impact of AZD5582 on the proliferation ability of PANC‐1 and CFPAC‐1 cells. (H, I). The effect of AZD5582 on two PC organoids at various concentrations.

3.5. SHCBP1 inhibitor AZD5582 inhibited PC growth in a PDX model

Additionally, a PDX model of PC was developed to investigate the inhibitory effect of AZD5582 in vivo further. When tumor sizes reached 150 mm³, the mice were administered vehicle or the indicated concentrations of AZD5582 by i.v. 3 mg/kg once a week for 3 weeks (Figure 5A). The results demonstrated that AZD5582 significantly inhibited tumor growth in the PC model (Figure 5B–D), without causing any noticeable changes in the body weight of the mice (Figure S3). Additionally, we performed histological examinations using H&E staining and conducted immunohistochemical analyses of SHCBP1 and Ki‐67 positive cells in the mouse xenograft tumors. These results showed that AZD5582 effectively inhibited the expression of SHCBP1 (Figure 5E,F), and attenuated the proliferation rate of tumor cells (Figure 5G,H). Therefore, the above findings provide compelling evidence for the therapeutic potential of AZD5582 in inhibiting PC growth in vivo, supporting its further development as a promising treatment for PC.

FIGURE 5.

SHCBP1 inhibitor AZD5582 inhibited PC growth in the PDX model. (A) The administration diagram of AZD5582 in the PDX model. (B–D) AZD5582 significantly inhibited tumor volume and tumor weight compared with the control group. (E, F) Representative H&E and IHC staining of SHCBP1 in tumors of PDX mice treated with AZD5582. (G, H) Representative H&E and IHC staining of Ki67 in tumors of PDX mice treated with AZD5582.

3.6. SHCBP1 inhibitor AZD5582 promoted apoptosis of pancreatic cancer cells

To elucidate the intricate mechanisms underlying the inhibition of PC growth by the SHCBP1 inhibitor AZD5582, we performed a transcriptomic analysis of AZD5582‐treated PC cells. Using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database for pathway analysis of differentially expressed genes within the transcriptome, we identified 20 pathways that exhibited the lowest p‐values (Figure 6A). Notably, the SHCBP1 inhibitor AZD5582 exhibited the most profound impact on the apoptotic signaling pathway (Figure 6B). Furthermore, Gene Set Enrichment Analysis (GSEA) further supported these findings, revealing a significant enrichment of apoptosis‐related signals, with Normalized Enrichment Score (NES) values reaching 1.77 (Figure 6C). To further validate the consistency of the SHCBP1 function, we collected mRNA expression data of PC from TCGA database and conducted GSEA analysis. Markedly, the results showed a significant involvement of SHCBP1 in the apoptotic signaling pathway, with NES values reaching 1.97 (Figure 6D). In addition, R software GSVA package was used for single sample GSEA (ssGSEA), and subsequent Spearman correlation results showed that the SHCBP1 gene had a significant correlation with an apoptotic signaling pathway (Spearman ρ = 0.34, p = 2.88e−06) (Figure 6E). Whereafter, flow cytometry also revealed that AZD5582 effectively induced the apoptosis of PANC‐1 (Figure 6F,G). The above analysis results collectively indicated that the SHCBP1 inhibitor AZD5582 may exert its inhibitory effect on PC by mediating the apoptotic pathway. We also performed western blot experiments to detect the effect of AZD5582 on apoptotic pathway‐related markers. The findings revealed that AZD5582 significantly inhibited the expression of SHCBP1, concomitant with reduction of BCL‐2 and increase of the cleavage of Caspase 3 and Caspase 9 in PANC‐1 and CFPAC‐1 cells (Figure 6H). To further explore the potential pathways by which ADZ5582 induces caspase activation, we detected the expression level of MAPK/ERK, PI3K/AKT and TP53. Our results revealed that AZD5582 had a significant inhibitory effect on the PI3K/AKT signaling pathway and prevented TP53 degradation (Figure 6I). These findings suggested that SHCBP1 can regulate the apoptosis of PC, and its inhibitor AZD5582 exerted a significant suppression of the growth of PC.

FIGURE 6.

SHCBP1 inhibitor AZD5582 promoted the apoptosis of pancreatic cancer cells. (A) Transcriptome KEGG analysis revealed 20 pathways that AZD5582 had the most significant effect on PC cells. (B) AZD5582 exhibited the most pronounced impact on apoptosis signaling pathway of PC. (C) Transcriptome GSEA analysis revealed that the differentially expressed genes were significantly enriched in apoptotic signaling pathway following AZD5582 treatment (NES = 1.77, p < 0.001). (D) The GSEA analysis revealed that SHCBP1 mRNA expression was significantly enriched in apoptosis pathways (NES = 1.97, p < 0.01). (E) ssGSEA and Spearman correlation analysis revealed that the SHCBP1 gene had a significant correlation with apoptotic signaling pathway (Spearman ρ = 0.34, p = 2.88e−06). (F, G) Flow cytometry revealed that AZD5582 could significantly induce the apoptosis of PC cell PANC‐1. (H) Western blot detected the effect of SHCBP1 inhibitor AZD5582 on apoptotic pathway‐related markers. (I) Western blot detected the effect of SHCBP1 inhibitor AZD5582 on the PI3K/AKT, MAPK/ERK and TP53 signaling pathways.

4. DISCUSSION

PC is still one of the most lethal malignant tumors, causing high morbidity and mortality worldwide. ¹⁶ Therefore, it is crucial to comprehend the intricate mechanisms driving the initiation and progression of PC. Numerous studies have unequivocally demonstrated that SHCBP1, functioning as an oncogene implicated in tumorigenesis, serves as a significant prognostic indicator for cancer. ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² In this study, we showed that SHCBP1 was overexpressed in both PC cell lines and tissues with significant specificity. Knockout of SHCBP1 significantly inhibited the proliferation, migration and invasion of PC cells. Furthermore, we identified a novel targeted inhibitor of SHCBP1, AZD5582, which exhibited significant inhibitory effect in PC organoid and PDX.

Accumulating evidence has highlighted the significant overexpression of SHCBP1 in various tumors, including breast cancer, gastric cancer, prostate cancer, melanoma, and lung cancer. ⁸ , ⁹ , ¹⁰ , ¹¹ In the current study, we observed that SHCBP1 expression was significantly increased in PC patients and KPC spontaneous tumor mice, and it has a positive correlation with AJCC stage and CA199 value. These compelling results highlight the potential of SHCBP1 both as a valuable tumor marker and a promising therapeutic target for PC. Importantly, Yang et al.'s ¹² study also reported high expression of SHCBP1 in PC, associating it with tumor size, lymph node metastasis, portal vein invasion, and TNM staging. Therefore, SHCBP1 has significant specificity in PC, positioning it as a promising candidate for targeted tumor therapy. However, the precise role and underlying mechanism of SHCBP1 in PC remains unclear.

Previously researchers have reported that elevated expression of SHCBP1 promotes the proliferation, migration, and invasion of bladder cancer cells and non–small‐cell lung cancer (NSCLC) cells. ¹⁰ , ¹¹ In our study, we also found that the knockout of SHCBP1 significantly inhibited the proliferation of PC cell and affected the migration mainly by modulating the EMT signaling pathway. Moreover, knockout of SHCBP1 substantially suppressed tumor growth in vivo. These findings suggested that SHCBP1 possesses great potential as a therapeutic target for cancers and warrants further research and development. ¹⁷

Although SHCBP1 has been extensively studied in various tumors, there remains a scarcity of research focusing on SHCBP1 inhibitors. The only reported inhibitors of SHCBP1 have been identified in our previous studies, specifically TFBG, which targets the SHCBP1 and PLK1 complex. ⁹ However, there are certain limitations associated with the inhibitor TFBG. First, it does not directly target SHCBP1, but rather impacts the interaction between SHCBP1 and PLK1. Second, the structural instability of TFBG makes it susceptible to decomposition under normal conditions. As the three‐dimensional crystal structure of the SHCBP1 protein not been resolved, the utilization of AlphaFold becomes crucial. Its predictive accuracy is comparable with experimental methods such as X‐ray crystallography and cryo‐electron microscopy. ¹⁸ In this study, we used the predicted structure of SHCBP1 (AF‐Q9BZQ2‐F1‐model_v4) generated by AlphaFold for virtual docking screening of a library containing 9202 active small molecules. We successfully identified a novel SHCBP1 inhibitor AZD5582 that exerted a significant inhibitory effect on PC both in vivo and in vitro.

However, AZD5582 was originally designed and synthesized as an antagonist of IAPs, ¹⁹ which are indispensable regulators in the program of apoptosis. ²⁰ , ²¹ Additionally, the upregulation of IAP expression allows tumor cells to evade mitochondria‐dependent apoptosis pathways. ²² , ²³ Mechanically, AZD5582 induced apoptosis in tumor cells by disrupting the caspase‐9–XIAP interaction and inducing cIAP1 degradation. ¹⁵ However, the affinity value between AZD5582 and cIAP1 was not reported. In our study, we discovered that AZD5582 exhibits a strong affinity for SHCBP1. Treatment with AZD5582 significantly suppresses SHCBP1 expression in PANC‐1, CFPAC‐1 cells, and PDX tumors, underscoring the potent targeting and inhibitory efficacy of AZD5582 on SHCBP1. However, it remains unclear whether the antitumor effects of AZD5582 are primarily attributed to the inhibition of SHCBP1 and/ or IAP. To address this, we overexpressed SHCBP1 and IAP in PANC‐1 and CFPAC‐1 cells, respectively, then assessed the antitumor effects of ADZ5582 using cell activity assays. The results showed that both the upregulation of SHCBP1 and IAP partially impaired the antitumor effect of ADZ5582 (Figure S4). This suggested that ADZ5582 may be a multitargeted inhibitor, and its antitumor activity depended on the inhibition of both IAP and SHCBP1. However, elucidating whether the antitumor effects of AZD5582 are predominantly attributed to SHCBP1 or IAP, along with uncovering the specific underlying mechanisms, requires further exploration in future studies.

It is widely acknowledged that SHCBP1 serves a pivotal role in regulating tyrosine kinase receptors (TKRs) and its downstream pathways, including EGF/EGFR, FGF, NF‐κB, MAPK/ERK, PI3K/AKT, TGF‐β1/Smad and Wnt/β‐catenin. ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² Therefore, we analyzed the correlation of SHCBP1 expression with TKRs and their downstream signaling pathways using the Gene Expression Profiling Interactive Analysis (GEPIA) database. The Spearman correlation analysis revealed that SHCBP1 significantly correlated with TKRs and its downstream pathway PI3K/AKT and MAPK/ERK in PC (Figure S5). To investigate whether inhibition of SHCBP1 by AZD5582 can affect TKRs and their downstream signaling pathway activity, transcriptome analysis and the KEGG analysis were conducted. The results revealed that AZD5582 exhibited the most noteworthy influence on apoptosis signaling pathways. In addition, western blotting showed that AZD5582 induced the apoptosis of PC cells by inhibiting the activity of PI3K/AKT signaling and the degradation of TP53, and promoting the cleavage of Caspase 3 and Caspase 9. Interestingly, Liu et al. ¹¹ have underscored that SHCBP1 knockdown could induce apoptosis of cancer cells, which is primarily associated with a reduction in TP53 degradation, ultimately leading to the prolonged half‐life and increased stability of TP53. It has also been shown that SHCBP1 induces apoptosis mainly by affecting the nuclear translocation of β‐catenin and regulating the expression of caspase3, PUMA, and BCL‐2. ⁸ , ²⁴ , ²⁵ Therefore, the aforementioned findings highlight the crucial role of SHCBP1 in anti‐apoptotic signaling pathways within tumors. Moreover, its targeted inhibitor, AZD5582, demonstrates significant potential in promoting tumor cell apoptosis.

In summary, our study reveals the exceptional specificity of SHCBP1 in PC and its potential as a reliable tumor marker. Additionally, we have identified AZD5582 as a potent SHCBP1 inhibitor, demonstrating excellent targeting capabilities and significant therapeutic efficacy across various PC models. These findings underscore the tremendous potential of AZD5582 as a therapeutic agent for PC.

AUTHOR CONTRIBUTIONS

Zhijian Ma: Validation; writing – original draft; writing – review and editing. Qianlin Gu: Formal analysis; investigation; validation. Yiwei Dai: Data curation; investigation; resources. Qiaoyan Wang: Methodology; software; visualization. Wengui Shi: Conceptualization; investigation; writing – review and editing. Zuoyi Jiao: Conceptualization; funding acquisition; project administration.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

ETHICS STATEMENTS

Approval of the research protocol by an Institutional Reviewer Board: The Medical Ethics Review Board at the Lanzhou University Second Hospital.

Informed Consent: Informed consent was obtained from all participants.

Registry and the Registration No. of the study/trial: N/A.

Animal Studies: Animals were manipulated and housed according to protocols approved by the Animal Ethics Committee of Lanzhou University Second Hospital (Appl. No. D2023‐337).

Supporting information

Figures S1–S5.

Ma Z, Gu Q, Dai Y, Wang Q, Shi W, Jiao Z. Therapeutic potential of SHCBP1 inhibitor AZD5582 in pancreatic cancer treatment. Cancer Sci. 2024;115:820‐835. doi: 10.1111/cas.16059

DATA AVAILABILITY STATEMENT

All relevant data are within the manuscript and its Additional files.