Skip to main content
Download PDF

Advancing pancreatic cancer therapy by mesothelin-specific nanobody conjugation

Molecular Cancer volume 24, Article number: 124 (2025) Cite this article

Abstract

Pancreatic adenocarcinoma (PAAD) is highly challenging to treat due to its poor prognosis and limited effective treatment options. Liposomal nanotechnology has emerged as a promising drug delivery platform in oncology, but existing liposomal therapies face limitations such as systemic toxicity, insufficient tumor selectivity, and low target specificity. Mesothelin (MSLN), an antigen overexpressed in PAAD, has attracted attention as a potential target for precision therapy. Here, we present the development of an anti-MSLN nanobody (D3 Nb) with high binding affinity (KD = 2.2 nM) that can selectively bind to MSLN-positive cancer cells. Structural analysis revealed that hydrophobic and hydrogen bonds within the complementary determining region (CDR) of D3 Nb promote strong binding to MSLN, leading to significant inhibition of AKT/NF-κB signaling and downregulation of fibronectin 1 (FN1) and twist1, key drivers of PAAD oncogenicity. In vivo studies confirmed that D3 Nb alone inhibits tumor progression. Furthermore, selective delivery to MSLN-positive tumors in combination with gemcitabine-loaded liposomes (D3-LNP-GEM) significantly improved cytotoxicity and promoted tumor regression. These findings highlight the potential of the D3-LNP-GEM platform as a novel targeted therapy for MSLN-expressing malignancies, showing promising efficacy in preclinical models and paving the way for continued clinical evaluation.

Introduction

Pancreatic adenocarcinoma (PAAD), a deadly malignancy that is currently the seventh leading cause of cancer-related death worldwide, is increasing in incidence and is expected to become the second leading cause of cancer-related death by 2030. [1–2] As PAAD is asymptomatic until it is advanced, diagnosis is beneficial in only 20% of cases [3]. More than 50% of patientes are detected at stage IV, and the overall 5-year survival rate is only 6%. [4–5] Current treatment modalities, including surgery, chemotherapy, and radiation, offer limited benefit, and novel immunotherapies have yet to demonstrate substantial clinical efficacy [6,7,8]. This bleak outlook underscores the urgent need for innovative, targeted treatment strategies.

Nanotechnology-based liposomal drug delivery systems have transformed cancer treatment by improving the pharmacokinetics of chemotherapeutic agents, reducing systemic toxicity, and enhancing tumor-specific drug accumulation [9,10,11]. This approach reduces the systemic toxicity of drugs, fights resistance associated with conventional chemotherapeutic agents, and increases the effectiveness of treatment by increasing drug concentrations within the tumor. [12–13] For example, liposomal formulations of gemcitabine, a widely used chemotherapeutic agent, have demonstrated superior efficacy and reduced toxicity in preclinical models for several cancers, including pancreatic adenocarcinoma (PAAD). [14–15] Liposomes can also be conjugated with targeting ligands such as antibodies, peptides, or small molecules to achieve high specificity for cancer cells, minimizing off-target effects and further enhancing therapeutic efficacy. [16–17]

Mesothelin (MSLN), a cell-surface glycoprotein overexpressed in multiple solid tumors, including PAAD, ovarian cancer, and mesothelioma, has emerged as a promising target for cancer therapies [18,19,20]. MSLN plays a key role in cell adhesion, tumor invasion, and metastasis, making it a key factor in tumor progression and an ideal candidate for targeted therapy [21]. Strategies to target MSLNs using immunotoxins or monoclonal antibodies (mAbs) are currently in clinical trials, are currently being investigated in clinical trials, with drugs such as SS1P, amatuximab, and anetumumab labtansine showing early promise. [19, 21–22] Despite these advances, however, clinical outcomes have been limited by issues such as suboptimal efficacy, off-target toxicity, and the development of resistance. These limitations highlight the need for more sophisticated and precise mechanisms of MSLN targeting to increase therapeutic efficacy while minimizing side effects.

Nanobodies (Nbs), recombinant antibody fragments derived from the single antigen-binding domain of camelid heavy chain antibodies, are a next-generation approach for cancer targeting. Due to their unique structural and functional advantages, nanobodies are increasingly utilized in biopharmaceutical research and therapeutic development [23]. With a molecular weight of approximately 15 kDa, nanobodies exhibit nanomolar to picomolar binding affinities similar to conventional antibodies (average molecular weight of ~ 150 kDa), but with enhanced tissue penetration, improved stability, and more efficient production costs [24]. Futhermore, nanobodies also show enhanced biodistribution and the ability to efficiently bind to otherwise inaccessible epitopes, making them well suited for targeting tumor-associated antigens [25]. These properties position nanobodies as ideal candidates for targeted therapy in solid tumors characterized by deep tissue penetration challenges and stromal barriers, such as pancreatic adenocarcinoma (PAAD), where efficient tissue infiltration is critical for therapeutic success.

In this study, we present a novel therapeutic platform to target MSLN-expressing cancers by developing anti-MSLN nanobodies (D3 Nb) and conjugating them to gemcitabine-loaded liposomes (D3-LNP-GEM). This approach addresses the limitations of conventional liposomal therapies by leveraging the potent cytotoxicity of gemcitabine and the high tumor specificity of D3 Nb, resulting in enhanced therapeutic precision and reduced off-target toxicity. Preclinical studies have shown that D3-LNP-GEMs significantly increase intratumoral drug accumulation, effectively inhibit tumor progression and significantly prolong survival in a murine xenograft model of pancreatic adenocarcinoma (PAAD). These findings establish D3-LNP-GEMs as a promising candidate for the targeted therapy of MSLN-overexpressing malignancies and have the potential to be applied to a broader range of MSLN-expressing cancer.

Result

Generation and characterization of nanobodies binding to MSLN

Pan-cancer data from TCGA and GTEx were analyzed using GEPIA, revealing that MSLN is upregulated across multiple tumor types, including ovarian serous cystadenocarcinoma (OV) and pancreatic adenocarcinoma (PAAD) (Figure S1A). These results are consistent with public microarray datasets GES132956, GSE28735, GSE62452, and the Clinical Proteomics Tumor Analysis Consortium (CPTAC) database (Figure S1B, S1C). MSLN was highly expressed in pancreatic adenocarcinoma compared to normal pancreatic or non-pancreatic tumors. Furthermore, patients with high MSLN expression in PAAD had a poorer prognosis and shorter overall survival (p = 0.003) (Figure S1D). Immunohistochemistry analysis further validated the differential expression of MSLN at the protein level, showing significantly higher expression in tumor tissue compared to normal pancreatic tissue (Figure S1E). We evaluated MSLN expression in pancreatic cancer cell lines (AsPC-1, Capan-1, Capan-2, PANC-1, Mia-Paca-2) by flow cytometry and Western blot analysis, and found that MSLN expression was increased in AsPC-1 and Capan-2 cell lines (Figure S1F, S1G).

To identify high-affinity binders to MSLN, a comprehensive screening of a phage display nanobody library was conducted (Fig. 1A), resulting in the identification of 60 clones with high specificity for MSLN (Fig. 1B). Sequencing and secondary screening refined the selection to nine nanobodies (A6, B3, C3, C12, D3, E5, F12, G9 and H5), all of which showed binding affinities in the nanomolar range (Figure S2A). Flow cytometry confirmed specific binding to AsPc-1 and Capan-2 cell lines expressing MSLN (Figure S2B). Of these, three nanobodies (D3, E5, and H5) showed the most pronounced cell migration inhibition in vitro, making them strong candidates for further study (Fig. 1C).

Fig. 1
figure 1

Screening and characterization of MSLN-specific nanobodies. A. Schematic representation of the nanobody screening and identification process. B. ELISA results identified 60 phage clones that specifically bind to MSLN (highlighted in red). Of these, nine clones (A6, B3, C3, C12, D3, E5, F12, G9 and H5) were selected as potential candidates for further study based on their KD values. C. Transwell migration assay performed on AsPC-1 and Capan-2 cells treated with 10 µg mL-1 of MSLN-specific or control nanobodies (ctrl Nb, irrelevant non-binding control nanobody) for 24 h showed significant inhibition of cell migration. Data are presented as mean ± SD and were analyzed using a two-tailed Student’s t-test. *: vs. Ctrl Nb. *p < 0.05, **p < 0.01

Full size image

Effect of MSLN nanobodies on cell migration and invasion

The effects of three MSLN-targeting nanobodies (D3, E5, and H5) on wound healing, migration, and invasion were investigated in vitro. Treatment with nanobodies at a concentration of 10 µg ml-1 did not affect cell viability or proliferation within 24 h (data not shown). The results demonstrated significant inhibition of in vitro wound healing (Fig. 2A), migration (Fig. 2B), and invasion (Fig. 2C) in AsPC-1 and Capan-2 cell lines, both of which exhibit high MSLN expression. Among the nanobodies, D3 exhibited the most pronounced inhibition of MSLN-mediated processes, effectively blocking migration and invasion. Consequently, D3 was identified as the lead candidate due to its strong binding affinity for pancreatic cancer cells (Figure S2B), potent inhibition of cell migration, and favorable kinetic parameters. In contrast, none of the nanobodies, including D3, significantly inhibited wound healing in the Mia-PaCa-2 and Capan-1 cell lines (Figure S3A), nor did they inhibit migration in the Capan-1 and PANC-1 cell lines (Figure S3B), all of which express low levels of MSLN. These results support the conclusion that nanobodies, especially D3, selectively target cancer cells expressing MSLN, effectively inhibiting migration, while having no significant effect on cells with low MSLN expression.

Fig. 2
figure 2

Effects of MSLN nanobodies on wound healing, migration, and invasion of pancreatic cancer cells. A. The wound healing assay demonstrates reduced wound closure in cells treated with MSLN nanoparticles compared to control nanobody. B&C. Transwell migration (B) and Invasion (C) assays showed that MSLN nanobodies significantly inhibit migration and invasion in pancreatic cancer cells, visualized by 0.5% crystal violet. Data represented as mean values ± SD, determined by two-tailed Student’s t-test except for two-way ANOVA Tukey’s multiple comparisons for A, B, and C. *: vs. Ctrl Nb. *p < 0.05, **p < 0.01, and ***p < 0.001

Full size image

Characterization of D3 nanobody

The selected D3 Nb consists of 125 amino acids with a molecular weight of 15 kDa, confirmed by SDS-PAGE analysis (Fig. 3A). To ascertain the affinity constants of MSLN binding to purified D3 Nb, surface plasmon resonance (SPR) was utilized. Various concentrations of purified MSLN were prepared and injected over the surface after D3 Nb was immobilized on the surface of the Biacore Chip CM5. SPR results revealed that the equilibrium dissociation constant (KD) for D3 Nb against MSLN was 2.2 nM (Fig. 3B). To further understand its biochemical potency against MSLN, we conducted an in-depth analysis of the structural characteristics relevant to its binding affinity. Docking simulations were employed to predict the most likely binding configuration of D3 Nb with MSLN, as illustrated in Fig. 3C. Notably, the sidechains of Arg101 and Lys103 in the CDR of D3 Nb form hydrogen bonds with His405 and Glu400 of MSLN, respectively. The computed structure of the MSLN-D3 Nb complex reveals an extra intermolecular hydrogen bond formed between Arg30 of D3 Nb and Asp464 of MSLN. The binding affinity of D3 Nb to MSLN appears to be further promoted by the hydrophobic interactions of Leu29 and Ile31 of D3 Nb and Thr467 and Pro438 of MSLN. These hydrophobic interactions seem to play a significant role in stabilizing the MSLN-D3 Nb complex, particularly due to their proximity to the hydrogen bond formed between Arg30 and Asp464. Taken together, the nanomolar-level biochemical potency of D3 Nb against MSLN can be attributed to the synergistic effects of multiple hydrogen bonds and hydrophobic interactions, which together contribute to its strong binding affinity and biochemical effectiveness.

Fig. 3
figure 3

Characterization of D3 nanobodies. (A) The purity of the recombinant D3 nanobody protein was confirmed by SDS-PAGE. M indicates marker. (B) SPR analysis demonstrated that D3 Nb binds MSLN with high affinity, showing representative kinetic data fitted to a 1:1 binding model. (C) Molecular modeling depicts the probable binding poses of D3 Nb with MSLN, with intermolecular hydrogen bonds and van der Waals contacts indicated

Full size image

Regulation of AKT/NF- κB Pathways and EMT-related genes by MSLN targeting with D3 Nb

To investigate the specific binding, cellular uptake, and internalization of D3 Nb, AsPC-1 cells were treated with Cy5-labeled D3 Nb for 24 h. Fluorescence imaging (Fig. 4A) revealed significant accumulation of labeled nanobodies in both the cell membrane and cytoplasm, indicating efficient internalization and intracellular trafficking. These results suggest the potential of D3 Nb for intracellular targeting and regulation of downstream signaling pathways.

Fig. 4
figure 4

Effects of MSLN blockade on down-regulation of AKT/NFκB signalling and EMT properties in AsPC-1 cells. (A) Confocal microscopy showed that the cellular internalization of Cy5-labelled D3 Nb was enhanced compared to Ctrl Nb. AsPC-1 cells were further stained green with CMFDA as a cell tracer; scale bar represents 20 μm. (B) The activity of the AKT/NFκB pathway was assessed by measuring pAKT/AKT and pNFκB/NFκB in AsPC-1 cells treated with D3 Nb, and compared to Ctrl Nb. (C) The mRNA expression levels of EMT-related genes in AsPC-1 cells treated with D3 Nb for 24 h were analyzed by qPCR, and data were expressed as mean values ± SD. (D) Western blot analysis was used to determine the protein expression levels of key markers including E-cadherin, EpCAM, N-cadherin, TWIST1 and FN1 in AsPC-1 cells treated with D3 Nb for 24 h. (E) Confocal imaging confirmed that the expression of TWIST1 and FN1 was decreased after D3 Nb treatment. Nuclei were visualized by DAPI staining. Data were presented as mean values ± SD and were determined by a two-tailed Student’s t-test for D. All Nb concentrations are 10 µg mL-1 and scale bars represent 20 μm. *: vs. Ctrl Nb. *p < 0.05

Full size image

Previous studies have shown that MSLN promotes epithelial-mesenchymal transition (EMT) and tumorigenesis by activating PI3K/AKT and NF-κB signaling pathways [26–27] To explore this further, we examined changes in AKT and NF-κB signaling in MSLN-overexpressing cell lines, AsPC-1 and Capan-2, after D3 Nb treatment. The data showed a significant down-regulation of phospho-AKT and phospho-NF-κB levels in both cell lines after D3 Nb treatment (Fig. 4B, Figure S4A).

To elucidate the role of AKT/NF-κB signaling in the regulation of EMT, we performed qRT-PCR analysis to assess the expression of key genes involved in cell migration and EMT in D3 Nb-treated and control Nb-treated cells. The results showed that TWIST1 and FN1 expression was significantly reduced in D3 Nb-treated cells at both mRNA and protein levels (Fig. 4C and D). Immunofluorescence microscopy further confirmed the down-regulation of TWIST1 and FN1 in D3 Nb-treated AsPC-1 cells compared to control cells (Fig. 4E). Consistent results were also observed in the Capan-2 cell line overexpressing MSLN (Figure S4B, S4C, and S4D). Taken together, these results indicate that MSLN signaling inhibition by D3 Nb significantly regulates the expression of EMT-related genes, especially TWIST1 and FN1, through AKT/NF-κB pathway inhibition.

Effect of MSLN/TWIST/FN1 overexpression on tumor progression and survival in pancreatic cancer

To investigate the correlation between MSLN and the EMT-related genes TWIST1 and FN1, expression levels were analyzed using data from the TCGA database. The results demonstrated that TWIST1 and FN1 were significantly upregulated in pancreatic cancer tissues compared to normal pancreatic tissues (Fig. 5A). These findings align with public microarray datasets (GSE28735 and GSE62452), which also showed elevated expression of TWIST1 and FN1 in pancreatic adenocarcinoma compared to non-pancreatic tumors (Fig. 5B). Further analysis using the GEPIA database revealed a significant positive correlation between MSLN and TWIST1 (p < 0.00021, R = 0.27) as well as MSLN and FN1 (p < 0.0028, R = 0.22) (Fig. 5C).

Fig. 5
figure 5

TWIST1 and FN1 upregulation in PAAD predicts poor prognosis. A&B. Analysis of TCGA data (A) and GEO database (B) indicates upregulation of TWIST1 and FN1 in pancreatic adenocarcinoma tissues. GSE28735 (Pancreatic non-tumor tissue; n = 45, Pancreatic tumor tissue; n = 45, TWIST1 p < 0.0003, FN1 p < 0.0001), GSE62452 (Pancreatic non-tumor tissue; n = 61, Pancreatic tumor tissue; n = 69, TWIST1 p < 0.0023, FN1 p < 0.0001) C. Pearson’s correlation analysis revealed a positive correlation between MSLN/TWIST1 and MSLN/FN1 gene expression using the GEPIA tool. (Left) MSLN and TWIST1 were positively correlated (p = 0.00021; R = 0.27). (Right) MSLN and FN1 were positively correlated (p = 0.0028; R = 0.22) D. Immunohistochemistry confirmed high expression of MSLN, TWIST1, and FN1 in PAAD tissues. E. Kaplan-Meier analysis showed that elevated expression of MSLN, TWIST1, and FN1 is associated with poorer overall

Full size image

The clinical relevance of MSLN, TWIST1, and FN1 overexpression was further validated through immunohistochemical (IHC) staining of tissue microarrays from pancreatic cancer patients. MSLN and FN1 were significantly overexpressed in tumor tissues compared to normal tissues (Figure S1E, S5A). In tumor samples, high levels of MSLN, TWIST1, and FN1 were observed in the MSLN high-expression group, while lower levels were detected in the MSLN low-expression group (Fig. 5D). These observations were consistent with HPA tissue staining, which confirmed similar expression patterns for MSLN and FN1 in patient samples (Figure S5B).

Survival analysis using the TCGA database demonstrated that pancreatic adenocarcinoma patients with high TWIST1 or FN1 expression had a poorer prognosis and shorter overall survival (TWIST1; p = 0.07, FN1; p = 0.03) (Figure S5C). Further investigation revealed that patients in the MSLN/TWIST1 high-expression group had significantly lower survival compared to the low-expression group, with similar trends observed for MSLN/FN1 (Fig. 5E, Figure S5D). In particular, patients with high expression of all three genes had significantly lower overall survival compared to those with low expression (Fig. 5E, right).

Development and characterization of D3 Nb-conjugated LNP

LNP-GEM formulations were developed by modifying DPPC lipids followed by active loading of gemcitabine (GEM). They were then conjugated with thiolated D3 nanobodies (D3 Nb) to form D3-LNP-GEM complexes (Fig. 6A). The resulting formulations showed high GEM encapsulation efficiency (> 55%) and loading capacity (> 5.5%).

Fig. 6
figure 6

Characterization of LNP-GEM and D3-LNP-GEM. (A) Schematic representation of the preparation of LNP-GEM and D3-LNP-GEM. (B) Physicochemical analysis indicated slight increase in size, PDI, and surface charge for D3-LNP-GEM. (C) TEM images confirmed the morphological characteristics of liposomes. The scale bar represents 100 nm. (D) The in vitro drug release profiles showed sustained GEM release for 48 h at both pH5.2 and pH 7.4. Data represent mean ± standard deviation (n = 3). (E) In vitro cytotoxicity evaluation confirmed that LNP-GEMs and D3-LNP-GEMs were cytotoxic against AsPC-1. Data represent the mean ± standard deviation (n = 3). (F) Live/dead staining confirmed that D3-LNP-GEMs exhibited enhanced cytotoxicity against AsPC-1 compared to LNP-GEMs. Live cells are shown in green and dead cells in red

Full size image

Particle size, polydispersity index (PDI), and zeta potential measurements were performed to characterize both LNP-GEMs and D3-LNP-GEMs. A slight increase in particle size and PDI was observed for D3-LNP-GEMs (131.3 ± 2.1 nm, PDI = 0.244) compared to LNP-GEM (116.8 ± 2.9 nm, PDI = 0.188) (Fig. 6B). The zeta potential shifted from − 0.77 ± 3.3 mV for LNP-GEM to -0.23 ± 0.07 mV for D3-LNP-GEM, indicating successful surface coating with D3 Nb. Despite these minor changes, both formulations retained their favorable morphological properties and negative surface charge, contributing to enhancing the stability and cell interaction of the nanoparticles in cancer therapy. TEM images corroborated these results by showing well-defined nanoparticle morphology (Fig. 6C).

The in vitro release profiles of GEM were evaluated at pH 5.2 and pH 7.4 using the dialysis method. The release rate at pH 5.2 was about 5% higher than that at pH 7.4. LNP-GEMs showed slightly faster release kinetics at pH 7.4 compared to D3-LNP-GEMs, but no significant difference was observed at pH 5.2. Both formulations showed a cumulative release of 50–60% over 48 h at acidic pH (Fig. 6D), suggesting that D3 Nb enhances tumor targeting without significantly altering the GEM release profile in the tumor microenvironment.

Targeted cancer therapy using D3-LNP: enhanced cytotoxicity and lysosomal targeting in AsPC-1 cells

D3-mediated targeting ability of LNP was assessed by examining lysosomal co-localization. The results showed that D3-LNPs efficiently accumulated in lysosomes and then released into the cytoplasm. In addition, the red fluorescence intensity of cells treated with D3-LNP-rhodamine B was higher than that of cells treated with LNP-rhodamine B indicating an efficient active targeting mechanism (Fig. 7A). We then analyzed the cellular uptake of D3-LNP-rhodamine B in AsPC-1 cells. Flow cytometry analysis of AsPC-1 cells showed that D3-LNP-rhodamine B had a cellular uptake of about 25%, whereas LNP-rhodamine B had only about 3% (Fig. 7B). These results are consistent with the cell fluorescence images in Fig. 7A, confirming that D3-LNP-rhodamine B is internalized via MSLN receptor-mediated endocytosis and reaches the cytoplasm via the lysosomal pathway, indicating its potential as a targeted cancer therapeutic.

Fig. 7
figure 7

Cellular uptake and internalization of LNP and D3-LNP. A. Confocal imaging demonstrated colocalization of D3-LNP-Rhod B (red fluorescence) with lysosome in AsPC-1 cells. Scale bar indicated 20 μm. Data represent the mean ± standard deviation (n = 3). B. Flow cytometry analysis showed higher cellular uptake of D3-LNP compared to LNP. Data represent the mean ± standard deviation (n = 3). (C) In vitro cytotoxicity evaluation confirmed that LNP-GEM and D3-LNP-GEM were cytotoxic against AsPC-1. Data represent the mean ± standard deviation (n = 3). (D) Live/dead staining showed that D3-LNP-GEM exhibited enhanced cytotoxicity against AsPC-1 compared to LNP-GEM. Live cells are shown in green and dead cells in red. *p < 0.05, **p < 0.01, and ***p < 0.001

Full size image

Evaluation of the cytotoxic effects of LNP-GEM and D3-LNP-GEM in AsPC-1 cells revealed that both formulations showed dose-dependent cytotoxicity, with D3-LNP-GEM exhibiting greater cytotoxicity than LNP-GEM (Fig. 7C). This was also confirmed by live/dead cell staining at a concentration of 10 µg mL-1, which showed a significant increase in dead cells stained red in AsPc-1 cells treated with LNP-GEM or D3-LNP-GEM (Fig. 7D). The Capan-2 cell line also showed consistent results in cytotoxicity and live/dead cell staining (Figures S6A-B). These results highlight the enhanced therapeutic potential of D3-LNP-GEMs for targeted cancer therapy.

Biodistribution and targeting ability of D3-LNPs in the PAAD mouse model

The in vivo targeting ability of D3-LNPs was evaluated using fluorescence optical imaging with ICG as a tracer. Mice bearing AsPC-1 tumors (approximately 400–500 mm3) were intravenously injected with 5 mg kg-1 of LNP-ICG or D3-LNP-ICG. Whole-body fluorescence images were captured at 1, 24, and 48 h post-injection to track distribution kinetics (Fig. 8A). To further confirm biodistribution, we measured ex vivo fluorescence signals in resected tumors and vital organs 48 h after injection. In AsPC-1 tumors, fluorescence accumulation was observed starting at 1 h post-injection and persisted for up to 48 h. Among the organs examined, the highest fluorescence was detected in the liver at 48 h, followed by significant accumulation in AsPC-1 tumors. Notably, D3-LNP-ICG showed significantly higher fluorescence in AsPC-1 tumors compared to LNP-ICG, indicating enhanced tumor targeting.

To analyze the distribution kinetics of D3-LNP-rhodamine B in plasma, we intravenously injected it into nude mice and collected plasma samples at several time points after injection to measure drug concentrations. D3-LNP-rhodamine B was rapidly detected in the bloodstream, with maximum accumulation observed at 2 h (Figure S7A). Plasma concentrations decreased by 50% after 12 h and were barely detectable after 72 h (Figure S7B). These results demonstrate the efficient tumor targeting ability of D3-LNPs and a favorable pharmacokinetic profile for sustained drug delivery.

D3-LNP-GEM effectively inhibits tumor growth in a PAAD mouse model

The antitumor efficacy of D3 Nb and D3-LNP-GEM was evaluated in mice bearing AsPC-1 tumors (approximately 150 mm3). Mice were treated three times a week with Ctrl Nb (unrelated, unbound control nanobody), D3 Nb (5 mg kg-1), LNP-GEM, D3-LNP-GEM (GEM equivalent dose = 4 mg kg-1), or PBS. Mice treated with D3 Nb, LNP-GEM, or D3-LNP-GEM had significantly delayed tumor growth compared to the PBS group (Fig. 8B). LNP-GEM monotherapy led to a 41.3% reduction in tumor volume and a 59.2% reduction in tumor weight. D3 Nb monotherapy was more effective, reducing tumor volume by 56.1% and tumor weight by 71.4%, highlighting its strong potential as a standalone treatment. Most notably, D3-LNP-GEM exhibited the greatest antitumor efficacy, with tumor volume reduced by 67.9% and tumor weight by 80.4%, indicating its promise as a highly targeted and potent therapeutic strategy for pancreatic cancer (Fig. 8B and C).

Further analysis showed that D3-LNP-GEM-treated tumors had significantly lower microvessel density (Fig. 8D, top) and reduced tumor cell proliferation (Fig. 8D, middle) compared to tumors treated with Ctrl Nb or PBS. In addition, D3-LNP-GEM- and D3 Nb-treated tumors had increased numbers of apoptotic tumor cells (Fig. 8D, bottom). Consistent with the in vitro results, D3-LNP-GEM treatment reduced the expression of TWIST1 and FN1 in tumors, further supporting its ability to inhibit epithelial-mesenchymal transition (EMT) (Fig. 8E).

Fig. 8
figure 8

In vivo antitumor efficacy of LNP-GEM and D3-LNP-GEM. (A) The in vivo administration and imaging timeline shows the enhanced tumor targeting by D3-LNP-ICG. Tumor-bearing organs were harvested at 48 h and imaged using an IVIS system. (B) Tumor growth curves indicate a reduction in tumor size in mice treated with D3-LNP-GEM. (C) Endpoint tumor images and tumor weights indicate improved treatment outcomes in D3-LNP-GEM-treated mice. Data were expressed as mean ± SEM. Sample size is n = 5. (D) Immunochemistry analysis showed decreased Ki67, CD31, and increased TUNEL staining in D3-LNP-GEM-treated tumors, indicating inhibited proliferation, reduced angiogenesis, and enhanced apoptosis. Scale bar indicated 200 μm. (E) Western blot analysis confirmed that FN1 and TWIST1 protein levels were reduced in D3-LNP-GEM-treated tumor tissues. *: vs. PBS. *p < 0.05, **p < 0.01

Full size image

The antitumor efficacy of D3 Nb and D3-LNP-GEM was further evaluated in mice bearing Capan-2 tumors. Similar to findings in the AsPC-1 model, D3-LNP-GEM treatment resulted in significant reductions in tumor volume, cellular proliferation, angiogenesis, and expression of EMT-related markers (Figures S8A–E). These findings further confirm the high targeting specificity and potent therapeutic potential of D3-LNP-GEM in MSLN-overexpressing pancreatic tumors.

Biosafety was assessed by monitoring body weight and serum biomarkers throughout the treatment period. No significant differences in body weight were observed between the treatment groups (Figure S9A). Serum markers such as glutamic acid oxaloacetic transaminase (GOT), glutamic acid pyruvate transaminase (GPT), albumin (ALB), creatinine (CRE), and blood urea nitrogen (BUN) did not show significant differences between the Ctrl Nb, D3 Nb, LNP-GEM, D3-LNP-GEM, and PBS groups (Figure S9B). Histopathologic analysis (H&E staining) of major organs (lung, liver, kidney, spleen, heart, and pancreas) showed no adverse effects, confirming the safety of D3-LNP-GEM treatment (Figure S9C, D).

Discussion

Mesothelin (MSLN) is overexpressed in a variety of cancers and has limited expression in normal tissues, making it an attractive therapeutic target. Anti-MSLN antibodies are currently being investigated for use in anticancer therapy and as a noninvasive diagnostic tool. [28–29] MSLN-targeted therapies, including monoclonal antibodies, immunotoxins, and CAR T cells, are being actively investigated in clinical trials for several cancers, including pancreatic adenocarcinoma (PAAD) and mesothelioma [30]. Despite the promising potential of these approaches, therapeutic efficacy remains limited due to challenges such as poor tissue penetration and instability of large antibody molecules in vivo.

Nanobodies (Nbs) or single-domain antibodies (sdAbs) offer an innovative solution to overcome these limitations. Due to their small size, nanobodies have superior tissue penetration, pharmacokinetics, and in vivo stability compared to conventional antibodies. In this study, we developed an anti-MSLN nanobody, D3 Nb, which exhibited high binding affinity (KD = 2.2 nM) for MSLN-positive pancreatic cancer cells. Structural analysis by docking simulations revealed that the strong hydrophobic and hydrogen bonding interactions between D3 Nb and MSLN contributed to the stability of the complex, making D3 Nb a promising candidate as a targeted therapeutic agent. Notably, D3 Nb has demonstrated superior efficacy at much lower doses compared to existing anti-MSLN therapies; for example, amatuximab, a commonly studied anti-MSLN monoclonal antibody, typically requires a concentration of 100 µg mL-1, whereas D3 Nb achieved similar effects at a 10-fold lower concentration (10 µg mL-1) (Figs. 2 and 4) [31]. In vivo, D3 Nb and D3-LNP-GEM were effective at a dose of only 4 mg kg-1 compared to the 200 mg kg-1 required for amatuximab [32]. These results demonstrate the potential for nanobodies such as D3 Nb to achieve enhanced anti-cancer efficacy at much lower doses, minimizing off-target effects and improving patient outcomes.

A key driver of cancer progression and metastasis is epithelial-mesenchymal transition (EMT), a process regulated by transcription factors such as TWIST1, which controls the expression of genes such as E-cadherin and vimentin, and regulates EMT markers such as fibronectin (FN1) [31, 33,34,35]. Previous studies have shown a positive correlation between TWIST1 and FN1 in several cancers, including adenoid cystic carcinoma [36]. Extending these findings, this is the first study to demonstrate a strong correlation between MSLN, TWIST1, and FN1 expression in PAAD. Analysis using the GEPIA database revealed a highly significant correlation (p = 2e-45, R = 0.82) between these markers, providing new insights into the molecular mechanisms underlying PAAD progression and identifying new targets for diagnostic and therapeutic strategies.

Benrusif et al. previously demonstrated successful tumor targeting using 68Ga-labeled nanobodies as a diagnostic tool for MSLN-positive tumors [37]. In this study, they developed a novel D3 nanobody (D3 Nb) that not only enhanced targeting but also showed therapeutic efficacy when combined with gemcitabine-loaded liposomes (D3-LNP-GEM). The dual-targeting strategy, utilizing passive targeting via enhanced permeability and retention (EPR) and active targeting via D3 Nb, increased the accumulation of D3-LNP-GEMs in the tumor microenvironment, leading to superior therapeutic outcomes compared to LNP-GEM monotherapy. Nanobodies, such as D3 Nb, provide a key advantage over conventional antibodies by enabling a higher density of nanobodies to be conjugated to each liposome. This increased nanobody density enhances multivalent binding (avidity), leading to stronger interactions with MSLN-expressing tumor cells, which likely contributed to the improved targeting efficiency and therapeutic efficacy observed. [38–39] These findings demonstrate that nanobody-based platforms offer distinct benefits for targeted cancer therapy, particularly when combined with both passive and active targeting mechanisms to improve the delivery of cytotoxic agents like gemcitabine to tumors.

In addition, the biosafety of D3-LNP-GEM was thoroughly evaluated. Confirmed by histopathological analysis, no significant toxicity was observed in major organs, and serum biomarker levels and body weight remained stable in all treatment groups. These results demonstrate that D3-LNP-GEM have both high therapeutic efficacy and good biocompatibility, making them a promising candidate for targeted treatment of MSLN-positive cancers.

Materials and methods

Library construction

A humanized synthetic VHH library, with a diversity exceeding 1011 nanobody clones, was constructed and used to develop MSLN-specific nanobodies. In this library, the lengths of CDR1 and CDR2 were fixed at 9 and 6 amino acids, respectively, while CDR3 was randomly generated with variable lengths ranging from 6 to 25 amino acids. The hNbBcll10FGLA [40] nanobody was selected as the ‘universal VHH scaffold’ and the nanobody gene was inserted into a pkb vector (KRIBB vector) [41]. This vector includes a pelB leader, and the synthetic VHH genes were inserted using ApaI and HindIII-HF restriction enzymes (New England Biolabs).

The recombinant phagemid library was transformed into E. coli strain TG1 via electroporation under the following conditions: 2.5 kV voltage, 200 Ω resistance, and 25 µF capacitance. To verify the diversity of the nanobody sequences, next-generation sequencing (NGS) was performed, confirming that approximately 93.6% of the sequences were unique and non-redundant.

Screening of antibodies against MSLN antigens

Biopanning was performed to enrich for nanobody-labeled phages specific for MSLN antigen [41]. The nanobody library was incubated in 2xYT medium with ampicillin (100 µg mL-1) until an OD600 of 0.6. VCSM13 helper phage (1 × 10¹² PFU) was added, followed by incubation with kanamycin (70 µg mL-1) overnight at 37 °C. Phages were precipitated using PEG 8000/NaCl, resuspended in DPBS, and used to screen MSLN-coated immunotubes for 2 h at 37 °C. After washing with DPBST (0.05% Tween 20), MSLN-bound phages were eluted with glycine-HCl (pH 2.5) and neutralized with Tris-HCl (pH 9.0). The eluted phages infected E. coli TG1 cells (OD600 = 0.6) for 30 min. Phage amplification and titer determination were repeated in subsequent panning rounds.

After the third panning, individual clones were selected and induced with IPTG at OD600 = 0.6. After incubation at 28 °C for 12 h, the supernatant was screened for MSLN-specific nanobodies by indirect ELISA. Absorbance was measured at 450 nm using a microplate reader (Tecan for life sciences) and positive clones were sequenced.

Preparation of MSLN nanobodies

The VHH gene was cloned into the TGEX-scblue plasmid (Antibody Design Lab) with a C-terminal unpaired cysteine using SpeI restriction sites. Expi293F cells (1 × 10⁶ cells mL-1) were seeded into 1 L baffled Erlenmeyer flasks and incubated overnight at 37 °C with 5% CO2. Transient transfection of the Nbs expression plasmid was carried out using 293fectin reagent (Life Technologies) following the manufacturer’s protocol. After 120 h, the culture supernatants were harvested by centrifugation at 3000 × g for 10 min, filtered, and stored at 4 °C. Nanobodies were purified using protein A agarose resin (Thermo Fisher Scientific), eluted with 0.1 M glycine (pH 2.7), and immediately neutralized with 1 M Tris-HCl (pH 8.0). Purity was assessed by SDS-PAGE under reducing conditions.

Surface plasmon resonance (SPR) analysis

Surface plasmon resonance (SPR) analysis was performed using a Biacore T200 instrument (Cytiva, Marlborough, MA, USA) equipped with a CM5 sensor chip to determine the binding affinity of the purified nanobodies. HBS-EP buffer was used as the running buffer. Purified nanobodies were dissolved in 10 mM sodium acetate (pH 5.0) and covalently immobilized onto the CM5 chip via standard amine coupling. Serial dilutions of the target protein (0–500 nM) were injected onto the sensor surface at a flow rate of 30 µL min− 1 with an association phase of 300 s and a dissociation phase of 600 s. Binding kinetics were analyzed using the 1:1 binding model provided by the Biacore T200 Evaluation Software version 3.2.1.

Homology modeling of D3 Nb structure and docking simulations with MSLN

The X-ray structure of mesothelin (MSLN) in complex with a therapeutic antibody (PDB entry: 7UED) was used for docking simulations with D3 Nb [42]. Since there is no structural data for D3 Nb, the X-ray structure of KN044 Nb (PDB entry: 6RQM), specific for cytotoxic T-lymphocyte-associated protein 4, was used as a template to generate a three-dimensional (3D) structure through homology modeling. The D3 Nb structure was modeled based on 82% sequence identity with KN044 Nb, ensuring a high-quality homology model. The MODELLER program [43] was used to create the 3D structure, with structural optimization achieved via the conjugate gradient method and molecular dynamics simulations to minimize spatial restraint violations. Gap regions were constructed by connecting anchoring positions through a randomly distorted structure. This final D3 Nb structural model was used for docking simulations with MSLN.

Docking simulations between MSLN and D3 Nb were performed using the Rosetta Dock program [44], a Monte Carlo-based algorithm. Initially, D3 Nb was overlaid onto the therapeutic antibody in the MSLN-antibody complex, followed by systematic translation and rotation to identify the optimal orientation relative to the fixed MSLN structure. The search for optimal binding modes included the refinement of D3 Nb side-chain conformations using rotamer packing and minimization of rigid-body displacements through gradient-based methods. The binding energy function included terms for van der Waals interactions, electrostatics (with low weighting), implicit Gaussian solvation, orientation-dependent hydrogen bonding, and side-chain rotamer probabilities [45]. Among the 1000 binding configurations generated, the one with the lowest binding energy was selected as the final structural model for the D3 Nb-MSLN complex.

Migration and invasion assay

Human pancreatic cancer cell lines (AsPC-1, Capan-1, Capan-2, PANC-1, and Mia-Paca-2) were obtained from ATCC and cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 100 µg mL-1 penicillin, and 100 µg mL-1 streptomycin. Cells were maintained at 37 °C in a humidified CO2 incubator.

For the scratch wound healing assay, 1 × 10⁵ cells per well were seeded into culture insert wells (µ-dish, Ibidi, Fitchburg, WI, USA) and grown until fully confluent. After removing the inserts, cells were treated with 10 µg mL-1 of nanobodies (Nbs), including control (Ctrl), A6, B3, C3, C12, D3, E5, F12, G9, and H5. Images were captured at 0 and 24 h post-treatment and wound closure was calculated by measuring the reduction in scratched area. For the transwell migration assay, 1 × 10⁵ cells per well were seeded in transwell inserts (Corning) and treated with 10 µg µg mL-1 of Nbs for 24 h. Migrated cells were stained with 0.5% (w: v) crystal violet and counted manually. For invasion assays, 1 × 10⁵ cells per well were seeded in Matrigel-coated transwell inserts and treated with 10 µg mL-1 of Nbs for 36 h. Invaded cells were stained with 0.5% (w: v) crystal violet and manually counted to assess invasion efficiency.

Western blotting

Cells were lysed in RIPA buffer (Thermo Fisher Scientific) and proteins were separated by SDS-PAGE. Proteins were transferred to polyvinylidene fluoride (PVDF) membranes (Millipore). The membrane was blocked with 5% skim milk in TBS containing 0.05% Tween 20 for 1 h at 25 °C. After blocking, the membrane was incubated with primary antibody diluted in 3% skim milk for 1 h at 25 °C or 16 h at 4 °C, followed by incubation with HRP-conjugated secondary antibody in TBST for 1 h at 25 °C. Protein bands were visualized using a chemiluminescence kit (Intron Biotech).

Real-time quantitative reverse transcription-polymerase chain reaction (qRT-PCR) analysis

Total RNA was extracted from cells using the RNeasy Mini Kit (Qiagen). Complementary DNA (cDNA) synthesis was performed with the Verso cDNA kit (Thermo Scientific) using 50 ng of total RNA. The qPCR reactions were conducted in triplicate with Power SYBR Green PCR Master Mix (Applied Biosystems) on a QuantStudio 1 system (Applied Biosystems), employing specific primers listed in Table S1. Relative fold changes were calculated using the 2−ΔΔCT method, with results normalized to control samples.

Preparation of liposomes

Liposomes were prepared in a molar ratio of 75 using a lipid mixture consisting of dipalmitoylphosphatidylcholine (DPPC; Avanti Research, MO, USA), cholesterol (Sigma Aldrich, MO, USA), 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (DSPE-PEG2000; Avanti Research), and DSPE-PEG2000-maleimide (Avanti Research): 20:4. 5:0.5 (DPPC: Chol: DSPE-PEG2000). The lipids were dissolved in chloroform and the solvent was evaporated using a rotary evaporator to form a thin lipid film. This film was hydrated with 250 mM ammonium sulfate buffer to reach the desired lipid concentration. The liposome suspension was subjected to five freeze-thaw cycles using liquid nitrogen and 25 °C, and then extruded through 100 nm and 50 nm polycarbonate membranes for size uniformity. Extrusion was repeated 21 times.

For drug loading, the liposomes were incubated with GEM at 60 °C for 4 h to incorporate gemcitabine (GEM) into the liposomes at a level of 10 wt%. The GEM-loaded liposomes (LNP-GEM) were then cooled to 25 °C and purified using Amicon Ultra 30 kDa centrifugal filters. Quantification of GEM in liposomes was performed using high-performance liquid chromatography (HPLC) on an Agilent 1260 system equipped with a C18 column (5 μm, 250 mm × 4.6 mm, Agilent). GEMs were extracted from LNP-GEMs with a mixture of acetonitrile and water. The mobile phase consisted of water and acetonitrile (90:10, v/v) at a flow rate of 1 mL min-1 and was detected at 284 nm. The drug loading capacity and encapsulation efficiency were calculatedusing the followed equations: Drug loading capacity (%) = (Amount of GEM in liposomes / Total amount of liposomes) × 100; Encapsulation efficiency (%) = (Amount of GEM in liposomes / Total amount of drug added) × 100.

Similarly, rhodamine-B (Sigma-Aldrich) and indocyanine green (ICG; Sigma-Aldrich) were loaded at 10 wt% into liposomes. After purification, the drug loading capacity and encapsulation efficiency of rhodamine-B and ICG were measured by fluorescence measurements. The excitation and emission wavelengths were as follows Rhodamine-B (λex = 546 nm, λem = 567 nm) and ICG (λex = 789 nm, λem = 813 nm).

Conjugation of nanobodies to liposomes

D3 nanobodies (D3 Nb) were modified with N-succinimidyl-S-acetylthioacetate (SATA; Sigma Aldrich) at a molar ratio of 1:8 (D3 Nb to SATA) and incubated at 25 °C. A 0.1 M hydroxylamine solution (0.1 M NH₂OH, 0.1 M HEPES, 5 mM EDTA, pH 7.2-7. 5) was used to deacetylate the SATA-modified D3 Nb by introducing free thiol groups and allowing the reaction to proceed for 2 h. The deacetylated D3 Nb was then added to the liposomes and the mixture was incubated for 16 h at 4 °C. Unbound D3 Nb was removed using an Amicon Ultra 100 kDa centrifugal filter unit. A micro bicinchoninic acid (micro-BCA) assay was performed on diluted liposome samples (20- to 200-fold dilution in PBS) to confirm the conjugation of nanobodies to liposomes. To accurately quantify the number of nanobodies bound to liposomes, the lipid contribution to the micro-BCA assay was subtracted.

Characterization of liposomes

For particle size, polydispersity index (PDI) and surface charge measurements, liposome samples were diluted to 0.1 mg mL-1 in PBS and subjected to a zeta-sizer (Nano ZSP/Zen5600).

To assess liposome morphology, samples were diluted to 0.1 mg mL-1 in deionized water and placed on carbon-coated copper grids (Thermo Fisher Scientific). Samples were stained with 2% phosphotungstic acid (Sigma Aldrich) and analyzed using transmission electron microscopy (TEM; Tecnai G2 Spirit Twin).

The release kinetics of gemcitabine (GEM) from LNP-GEM and D3-LNP-GEM were evaluated using a dialysis device with a molecular weight cutoff of 3.5 kDa. Samples were incubated in 20 mL phosphate-buffered saline (PBS; pH 7.4) or sodium acetate buffer (pH 5.2) at 37 °C with continuous shaking at 350 rpm. At each time point, 1 mL of buffer was withdrawn and replaced with fresh buffer. GEM concentrations in the collected samples were analyzed using HPLC.

Immunohistochemistry staining

MSLN and FN1 protein expression in pancreatic cancer and normal tissues were verified using immunochemistry data obtained from the Human Protein Atlas, an online resource for genome-wide analysis (http://www.proteinatlas.org/) [46].

Cellular uptake and internalization

AsPC-1 cells (5 × 103 cells per well) were seeded into glass-bottom confocal dishes. After stabiliazation, the cells were incubated with Rhodamine B-loaded LNP (LNP-Rhod B) and D3-LNP-Rhod B (10 µg mL-1) for 6 h at 37 °C. Following incubaton, the cells were washed with PBS and fixed with with 4% paraformaldehyde. Fixed cells were then stained with LysoTracker Geen (Thermo Fisher) and 4’6-diamidino-2-phenylindole (DAPI; Invitrogen) according to manufacturer’s instructions. Fluorescent images were acquired using a Floview confocal microscopy (Olympus).

To quantify cellular internalization, AsPC-1 cells treated with LNP-Rhod B and D3-LNP-Rhod B for 6 h were analyzed using a flow cytometer (BD Accuri C6; BD Bioscience).

Cytotoxicity

AsPC-1 or Capan-2 cells (5 × 103 cells per well) were seeded in 96-well plates and treated with various concentration of LNP-GEM and D3-LNP-GEM, ranging from 0 to 100 µg mL-1. After 24 h of incubation, cell viability was assessed using a cell counting kit-8 (CCK-8; Abcam). To further visualize cytotoxicity, cells were stained with calcein-AM and propidium iodide (PI) solution using a live/dead staining kit (Thermo Fisher Scientific). Fluorescence images were acquired using a confocal microscope.

Immunocytochemistry staining

Cells were fixed with 4% paraformaldehyde at 37 °C for 15 min, followed by permeabilization with 0.1% (v: v) Triton X-100 at 4 °C for 15 min. After fixation, cells were blocked with 2% (w: v) BSA solution for 30 min at RT and incubated with primary antibodies, anti-TWIST1 (1:200; Cell signaling technology) and anti-fibronectin (FN1) (1:200; Santa Cruz Biotechnology) for 16 h at 4 °C. After incubation, cells were washed three times with PBS (10 min each) and incubated with secondary antibodies, anti-mouse IgG Alexa 488 and anti-rabbit IgG Alexa 594 (1:500; Molecular Probes) for 1 h at 25 °C. Stained cells were mounted using Vectashield mounting medium (Fluoroshield, Abcam) and fluorescence images were captured using a confocal microscope.

Biodistribution

Female BALB/c nude mice (6–8 weeks old) were purchased from SLC (Japan). Mice were injected subcutaneously with AsPC-1 cells (5 × 10⁶ cells in 100 µL PBS). When the tumor volume reached 400–500 mm³, mice were intravenously administered with ICG-labeled LNPs (LNP-ICG) and D3-LNP-ICG (5 mg kg− 1, n = 3). The in vivo distribution of D3-LNPs was monitored at 0, 24, and 48 h using an IVIS-200 system (PerkinElmer). Tissues were excised at 48 h for ex vivo analysis.

Study of anti-tumor efficacy

BALB/c mice were inoculated subcutaneously with 5 × 10⁶ AsPC-1 or Capan-2 cells and randomized into five groups: PBS, Ctrl Nb, D3 Nb, LNP-GEM, and D3-LNP-GEM (n = 5 per group). Mice were treated intravenously (Ctrl Nb and D3 Nb 5 mg kg− 1, LNP-GEM and D3-LNP-GEM 4 mg kg− 1) three times weekly for 3 weeks when tumor volume reached 100–200 mm³. During the treatment period, tumor size was measured and the volume of the tumor was calculated using the following formula Tumor volume = 0.5 × length × width². After 21 days, tumor tissues were harvested and weighed.

All animal experiments were performed in accordance with the guidelines of the Institutional Animal Care and Use Committee of the Korea Research Institute for Biotechnology (KRIBB) (KRIBB-AEC-23030).

Histological analysis

Tumor tissue was immersed in optimal cutting temperature (OCT) compound, sectioned at 10 μm thickness, and fixed in acetone for 10 min. Sections were blocked with anti-Ki67 antibody (1:200; Thermo Fisher Scientific) or anti-mouse CD31-PE antibody (1:200; BD Biosciences) in 2.5% goat serum in PBS and then stained with appropriate secondary antibodies, anti-mouse IgG-FITC (1:500). Cell death was detected using a TUNEL assay kit (Promega). After staining, tissues were mounted using ProLong™ Diamond Antifade Mountant with DAPI (Thermo Fisher Scientific). Fluorescence images were captured using a confocal microscope. For each tissue sample, three representative areas were captured. Vessel counts were performed using the “Count” function in Image J software (National Institutes of Health). The average vessel count from the three images was calculated for each tissue, and the mean values for each group were used to generate bar graphs. Ki67-positive and TUNEL-positive cells were quantified using the same method, and the results were averaged for each sample and visualized.

For human pancreatic cancer tissue arrays (purchased from US Biomax), slides were deparaffinized, antigen retrieved, permeabilized, and stained with anti-MSLN (1:500), anti-TWIST1 (1:20), or anti-FN1 (1:20) antibodies.

Statistical analysis

Data are presented as mean ± SEM and analyzed with Prism version 7.0 (GraphPad Software, Inc., San Diego, CA). Unpaired Student’s t-test was used for comparisons between two groups. One-way analysis of variance (ANOVA) with Tukey’s post hoc test was used to compare multiple groups. A p-value of less than 0.05 was considered significant.

Data availability

No datasets were generated or analysed during the current study.

References

  1. Wang S, Zheng Y, Yang F, Zhu L, Zhu XQ, Wang ZF, Wu XL, Zhou CH, Yan JY, Hu BY, Kong B, Fu DL, Bruns C, Zhao Y, Qin LX. Q Z Dong Signal Transduct Target Ther. 2021;6:249.

    Article  Google Scholar 

  2. Yeo D, Giardina C, Saxena P, Rasko JEJ. Mol Ther Oncolytics. 2022;24:561.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Khorana AA, Mangu PB, Berlin J, Engebretson A, Hong TS, Maitra A, Mohile SG, Mumber M, Schulick R, Shapiro M, Urba S, Zeh HJ, Katz MH. J Clin Oncol. 2016;34:2541.

    Article  PubMed  Google Scholar 

  4. Puckett Y, Garfield K. in StatPearls, Treasure Island (FL) 2024.

  5. Koo MM, Swann R, McPhail S, Abel GA, Elliss-Brookes L, Rubin GP, Lyratzopoulos G. Lancet Oncol. 2020;21:73.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Anderson EM, Thomassian S, Gong J, Hendifar A, Osipov A. Cancers (Basel) 2021, 13.

  7. Ju Y, Xu D, Liao MM, Sun Y, Bao WD, Yao F, Ma L. NPJ Precis Oncol. 2024;8:199.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Smith C, Zheng W, Dong J, Wang Y, Lai J, Liu X, Yin F. World J Gastroenterol. 2022;28:3297.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Giri PM, Banerjee A, Layek B. Cancers (Basel) 2023, 15.

  10. Chehelgerdi M, Chehelgerdi M, Allela OQB, Pecho RDC, Jayasankar N, Rao DP, Thamaraikani T, Vasanthan M, Viktor P, Lakshmaiya N, Saadh MJ, Amajd A, Abo-Zaid MA, Castillo-Acobo RY, Ismail AH, Amin AH, Akhavan-Sigari R. Mol Cancer. 2023;22:169.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Eldar-Boock A, Polyak D, Scomparin A, Satchi-Fainaro R. Curr Opin Biotechnol. 2013;24:682.

    Article  CAS  PubMed  Google Scholar 

  12. Yao Y, Zhou Y, Liu L, Xu Y, Chen Q, Wang Y, Wu S, Deng Y, Zhang J, Shao A. Front Mol Biosci. 2020;7:193.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Cheng Z, Li M, Dey R, Chen Y. J Hematol Oncol. 2021;14:85.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Graeser R, Bornmann C, Esser N, Ziroli V, Jantscheff P, Unger C, Hopt UT, Schaechtele C, von Dobschuetz E, Massing U. Pancreas 2009, 38, 330.

  15. Tamam H, Park J, Gadalla HH, Masters AR, Abdel-Aleem JA, Abdelrahman SI, Abdelrahman AA, Lyle LT, Yeo Y. Mol Pharm. 2019;16:2858.

    Article  CAS  PubMed  Google Scholar 

  16. Cho K, Wang X, Nie S, Chen ZG, Shin DM. Clin Cancer Res. 2008;14:1310.

    Article  CAS  PubMed  Google Scholar 

  17. Hong L, Li W, Li Y, Yin S. RSC Adv. 2023;13:21365.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Scales SJ, Gupta N, Pacheco G, Firestein R, French DM, Koeppen H, Rangell L, Barry-Hamilton V, Luis E, Chuh J, Zhang Y, Ingle GS, Fourie-O’Donohue A, Kozak KR, Ross S, Dennis MS, Spencer SD. Mol Cancer Ther. 2014;13:2630.

    Article  CAS  PubMed  Google Scholar 

  19. Lv J, Li P. Biomark Res. 2019;7:18.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Hassan R, Ho M. Eur J Cancer. 2008;44:46.

    Article  CAS  PubMed  Google Scholar 

  21. Baldo P, Cecco S. Onco Targets Ther. 2017;10:5337.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Morello A, Sadelain M, Adusumilli PS. Cancer Discov. 2016;6:133.

    Article  CAS  PubMed  Google Scholar 

  23. Muyldermans S. FEBS J 2021, 288, 2084.

  24. Desmyter A, Spinelli S, Roussel A, Cambillau C. Curr Opin Struct Biol. 2015;32:1.

    Article  CAS  PubMed  Google Scholar 

  25. Minatel VM, Prudencio CR, Barraviera B, Ferreira RS Jr. Front Immunol. 2023;14:1303353.

    Article  CAS  PubMed  Google Scholar 

  26. He X, Wang L, Riedel H, Wang K, Yang Y, Dinu CZ, Rojanasakul Y. Mol Cancer. 2017;16:63.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Bharadwaj U, Marin-Muller C, Li M, Chen C, Yao Q. Mol Cancer. 2011;10:106.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Lamberts LE, Menke-van der CW, van Houven EJ, ter Weele F, Bensch MM, Smeenk J, Voortman OS, Hoekstra SP, Williams BM, Fine D, Maslyar JR, de Jong JA, Gietema CP, Schroder AH, Bongaerts MN, Lub-de Hooge HM, Verheul SM, Sanabria Bohorquez AW. Glaudemans, E. G. de Vries. Clin Cancer Res. 2016;22:1642.

    Article  CAS  PubMed  Google Scholar 

  29. Kobayashi K, Sasaki T, Takenaka F, Yakushiji H, Fujii Y, Kishi Y, Kita S, Shen L, Kumon H, Matsuura E. J Immunol Res 2015, 2015, 268172.

  30. Faust JR, Hamill D, Kolb EA, Gopalakrishnapillai A, Barwe SP. Cancers (Basel) 2022, 14.

  31. Matsuzawa F, Kamachi H, Mizukami T, Einama T, Kawamata F, Fujii Y, Fukai M, Kobayashi N, Hatanaka Y, Taketomi A. BMC Cancer. 2021;21:200.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Mizukami T, Kamachi H, Fujii Y, Matsuzawa F, Einama T, Kawamata F, Kobayashi N, Hatanaka Y, Taketomi A. Oncotarget. 2018;9:33844.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Fujisaka Y, Kurata T, Tanaka K, Kudo T, Okamoto K, Tsurutani J, Kaneda H, Okamoto I, Namiki M, Kitamura C, Nakagawa K. Invest New Drugs. 2015;33:380.

    Article  CAS  PubMed  Google Scholar 

  34. Kang Y, Massague J. Cell. 2004;118:277.

    Article  CAS  PubMed  Google Scholar 

  35. Wang K, Seo BR, Fischbach C, Gourdon D. Cell Mol Bioeng. 2016;9:1.

    Article  PubMed  Google Scholar 

  36. Bulzico D, Pires BRB, PAS DEF, Neto LV, Abdelhay E. Anticancer Res. 2019;39:173.

    Article  CAS  PubMed  Google Scholar 

  37. Benloucif A, Meyer D, Balasse L, Goubard A, Danner L, Bouhlel A, Castellano R, Guillet B, Chames P, Kerfelec B. Front Immunol. 2023;14:1200652.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Schriek AI, van Haaren MM, Poniman M, Dekkers G, Bentlage AEH, Grobben M, Vidarsson G, Sanders RW, Verrips T, Geijtenbeek TBH, Heukers R, Kootstra NA, de Taeye SW, van Gils MJ. Front Immunol. 2022;13:893648.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Narbona J, Hernandez-Baraza L, Gordo RG, Sanz L, Lacadena J. Biomolecules 2023, 13.

  40. Vincke C, Loris R, Saerens D, Martinez-Rodriguez S, Muyldermans S, Conrath K. J Biol Chem. 2009;284:3273.

    Article  CAS  PubMed  Google Scholar 

  41. Guk K, Kim H, Lee M, Choi YA, Hwang SG, Han G, Kim HN, Kim H, Park H, Yong D, Kang T, Lim EK, Jung J. Nat Commun. 2020;11:3418.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Zhan J, Lin D, Watson N, Esser L, Tang WK, Zhang A, Liu X, Hassan R, Gleinich A, Shajahan A, Azadi P, Pastan I, Xia D. Cancer Res Commun. 2023;3:175.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Sali A, Blundell TL. J Mol Biol. 1993;234:779.

    Article  CAS  PubMed  Google Scholar 

  44. Gray JJ, Moughon S, Wang C, Schueler-Furman O, Kuhlman B, Rohl CA, Baker D. J Mol Biol. 2003;331:281.

    Article  CAS  PubMed  Google Scholar 

  45. Alford RF, Leaver-Fay A, Jeliazkov JR, O’Meara MJ, DiMaio FP, Park H, Shapovalov MV, Renfrew PD, Mulligan VK, Kappel K, Labonte JW, Pacella MS, Bonneau R, Bradley P, Dunbrack RL Jr., Das R, Baker D, Kuhlman B, Kortemme T, Gray JJ. J Chem Theory Comput. 2017;13:3031.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Thul PJ, Lindskog C. Protein Sci. 2018;27:233.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

We would like to thank Editage (www.editage.co.kr) for editing and reviewing this manuscript for English language.

Funding

This work was supported by NRF and NST grants funded by Korea government (MSIT) (2021M3H4A1A02051048, 2023R1A2C2005185, RS-2024-00348576, RS-2024-00438316, RS-2024-00459749, and RS-2024-00338397), KEIT grants funded by Korea government (MOTIE) (RS-2022-00154853, RS-2024-00403563, and RS-2024-00432382), KEITI grant funded by Korea government (ME) (2021003370003), IPET grant funded by Korea government (MAFRA) (RS-2024-00401639), Nanomedical Devices Development Program of National Nano Fab Center, and KRIBB Research Initiative Program (KGM1322511 and KGM1032511).

Author information

Author notes

  1. Soyeon Yi and Kyunghee Noha contributed equally to this work.

Authors and Affiliations

  1. Bionanotechnology Research Center, Korea Research Institute of Bioscience and Biotechnology (KRIBB), 125 Gwahak-ro, Yuseong-gu, Daejeon, 34141, Republic of Korea

    Soyeon Yi, Kyunghee Noh, Hyeran Kim, Eunkyeong Jung, Suhyeon Kim, Jieun Lee, Kyeonghye Guk, Jinsol Choi, Eun-Kyung Lim, Taejoon Kang & Juyeon Jung

  2. Department of Nanobiotechnology, KRIBB School of Biotechnology, UST, 217 Gajeong-ro, Yuseong-gu, Daejeon, 34113, Republic of Korea

    Kyunghee Noh, Eun-Kyung Lim & Juyeon Jung

  3. Department of Fundamental Pharmaceutical Sciences, Graduate School, Kyung Hee University, 26 Kyungheedae-Ro, Dongdaemun-Gu, Seoul, 02447, Republic of Korea

    Suhyeon Kim

  4. College of Pharmacy, Korea University, Sejong, 30019, Republic of Korea

    Jinsol Choi

  5. School of Pharmacy, Sungkyunkwan University, 2066 Seobu-ro, Jangan-gu, Suwon, 16419, Republic of Korea

    Eun-Kyung Lim, Taejoon Kang & Juyeon Jung

  6. Department of Health Science, the Graduate School of Dong-A University, Busan, 49315, Republic of Korea

    Seokho Kim

  7. Deparment of Bioscience and Biotechnology, Sejong University, 209 Neungdong-ro, Kwangjin- gu, Seoul, 05006, Republic of Korea

    Hwangseo Park

  8. Stem Cell Convergence Research Center, Korea Research Institute of Bioscience and Biotechnology (KRIBB), 125 Gwahak-ro, Daejeon, Republic of Korea

    Jung Hwa Lim & Cho-Rok Jung

  9. Department of Functional Genomics, Korea University of Science and Technology (UST), 217 Gajeong-ro, Daejeon, Republic of Korea

    Jung Hwa Lim & Cho-Rok Jung

Authors
  1. Soyeon Yi
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Kyunghee Noh
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Hyeran Kim
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Eunkyeong Jung
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Suhyeon Kim
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Jieun Lee
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Kyeonghye Guk
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Jinsol Choi
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Eun-Kyung Lim
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Seokho Kim
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Hwangseo Park
    View author publications

    You can also search for this author inPubMed Google Scholar

  12. Jung Hwa Lim
    View author publications

    You can also search for this author inPubMed Google Scholar

  13. Cho-Rok Jung
    View author publications

    You can also search for this author inPubMed Google Scholar

  14. Taejoon Kang
    View author publications

    You can also search for this author inPubMed Google Scholar

  15. Juyeon Jung
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

S.Y and K.N. contributed to experimental design, experiments and writing the original draft. H.K. E.J., S.K., J.L., K.G., and J.C. assisted with the experiments. H.P. contributed nanobody binding prediction. S.K., E.L., JH.L., C-R.J. and T.K. contributed to data analysis and project management. J.J. contributed to project management and writing manuscript.

Corresponding author

Correspondence to Juyeon Jung.

Ethics declarations

Ethics approval and consent to participate

All animal experiments were approved by the Animal Care and Use Committee of the KRIBB (KRIBB-AEC-23030) following the Guidelines for the Care and Use of Laboratory. All human cells and tissue samples from human subject were conducted in accordance with the KRIBB Institutional Review Board (P01-202408-02-007).

Consent for publication

All authors have consented to submit this article for publication.

Competing interests

The authors declare no competing interests.

Conflict of interest

The authors declare no conflicts of interest related to this work.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Yi, S., Noh, K., Kim, H. et al. Advancing pancreatic cancer therapy by mesothelin-specific nanobody conjugation. Mol Cancer 24, 124 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12943-025-02325-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12943-025-02325-7

Keywords

Molecular Cancer

ISSN: 1476-4598