Introduction
Solid tumors rely on abnormal vascular networks to obtain the energy required to sustain rapid proliferation, making the tumor vasculature an attractive therapeutic target [[1], [2], [3], [4]]. Vascular disrupting agents (VDAs), such as combretastatin A4 (CA4), selectively destroy immature tumor blood vessels and induce rapid ischemic necrosis [[5], [6], [7], [8], [9]]. Although VDAs cause substantial initial tumor destruction, the clinical benefits are limited due to intrinsic biological constraints, which hinder sustained therapeutic responses [[10], [11], [12]].
A major challenge with VDA therapy is the spatial heterogeneity of efficacy. Whilst immature blood vessels in the tumor core are highly susceptible, mature blood vessels in the tumor periphery remain substantially resistant, allowing peripheral cancer cells to survive and drive tumor regrowth and metastasis [13,14]. Furthermore, VDAs-induced vascular necrosis recruits immunosuppressive tumor-associated macrophages (TAMs), enhancing immune tolerance and inhibiting macrophage-dependent immunotherapies, such as CD47 blockade [[15], [16], [17]].
These findings illustrate the necessity for improved therapeutic strategies that not only expand the effective range of VDAs but also mitigate the adverse effects on tumor immune microenvironments [18].
A limitation of systemic administration of VDAs is the absence of intrinsic targeting capability, presenting considerable risks of off-target vascular damage and dose-related toxicity [[19], [20], [21], [22]]. To address the aforementioned challenges, rationally designed nano-frameworks present a promising approach for the combined delivery of multiple therapeutic agents, further enhanced by the incorporation of targeting ligands to improve tumor-specific accumulation [[23], [24], [25], [26]]. However, the overall efficacy of these frameworks remains limited by their heterogeneity and the generally weak enhanced permeability and retention (EPR) effect [27].
Recent systematic investigations indicate that intravenously administered nanomedicines, even with active targeting modifications, typically have a delivery efficiency of less than 1% to solid tumors [28,29]. Such inadequate accumulation within tumors thus constitutes a barrier to effective cancer treatment. Notably, the vascular remodeling triggered through VDAs-mediated mechanisms potentially offers a means of overcoming the delivery constraints.
The transient increase in vascular permeability and disruption of tumor vascular stability induced by VDAs potentially amplifies the EPR effect and creates a brief therapeutic window, facilitating the accumulation of nanomedicines and biologics (including immunomodulators and antibodies) within the tumors. Consequently, VDAs as therapeutic amplifiers represents an attractive yet insufficiently explored strategy for enhancing drug delivery and achieving synergistic combination therapy [30,31].
While previous studies have established that CA4-induced vascular disruption can transiently enhance nanoparticle delivery, these approaches typically rely on untargeted nanoparticles and the passive EPR effect, and the resulting enhancement diminishes upon repeated administration. In this work, we developed an Endoglin-targeted Nano-Framework to Obtain Repolarized Cancer-associated macrophages and Enhanced permeability (designated ENFORCE) for breast cancer immunotherapy by the combination of αCD47. ENFORCE incorporated an endoglin-targeting chimeric peptide (palmitic-K(palmitic)-AHKHVHHVPVRL) that enabled precise binding to tumor vasculature and facilitated the co-delivery of the vascular-disrupting agent CA4 and the immunomodulator R848 (Scheme 1A).
Upon binding to the vascular endothelium, ENFORCE released CA4 to disrupt vessels, thereby amplifying the EPR effect and promoting further ENFORCE accumulation within tumors (Scheme 1B). Meanwhile, CA4-mediated tumor damage recruited immunosuppressive macrophages, a process counteracted by co-delivered R848, which effectively reprogramed TAMs toward the pro-inflammatory M1 phenotype and alleviated vascular injury-induced immunosuppression. Furthermore, the high permeability induced by ENFORCE significantly enhanced the ability of αCD47 to penetrate tumors, overcoming a critical delivery barrier to CD47 blockade efficacy and thereby enhancing the systemic antitumor immune response.
In summary, this work establishes a promising targeted nano-framework that integrates vascular disruption with macrophage immunomodulation to concurrently strengthen drug delivery and potentiate immunotherapy.Scheme 1Chemical structure and the proposed mechanism of ENFORCE for enhancing breast cancer immunotherapy. (A) Amphiphilic chimeric peptide consists of an endoglin-targeting sequence (AHKHVHHVPVRL) flanked by two hydrophobic palmitic acid residues. The palmitic tails readily insert into the lipid bilayer, leaving the hydrophilic targeting sequence exposed on the surface.
CA4 and TLR7/8 agonist R848 are co-encapsulated within the hydrophobic core, forming a stable targeted liposomal formulation. (B) Endoglin-targeting peptides mediate the specific aggregation of ENFORCE within the tumor vasculature. Localized release of CA4 induces potent vascular injury and increases vascular permeability, amplifying EPR effect to enhances the penetration and retention of ENFORCE.
Concurrently, R848 reprograms TAMs toward an antitumor M1 phenotype, counteracting the immunosuppressive microenvironment induced by vascular rupture. This reshaped tumor microenvironment synergizes with αCD47 checkpoint blockade to induce systemic antitumor immune responses and eradicate metastatic tumors.Scheme 1 dummy alt text
Materials and methods
Materials
CA4 and R848 were purchased from MedChemExpress Co., Ltd. (Shanghai, China). Dioleoyl phosphatidylcholine (DOPC), DSPE-PEG2000−Mal, DSPE-PEG2000, cholesterol, and cardiolipin were obtained from HARVEYBIO Co., Ltd.
(Beijing, China). Dimethyl sulfoxide (anhydrous DMSO), trichloromethane, and methanol were acquired from Shanghai Chemical Co., Ltd. (Shanghai, China).
MTT and Annexin V-FITC apoptosis detection kits were provided by Beyotime Biotechnology Co., Ltd. Lipopolysaccharides (LPS) was purchased from PeproTech, Inc. (China).
Anti-endoglin (α-Endoglin) was purchased from Wuhan Sanying Biotechnology Co., Ltd (Wuhan, China). Anti-CD3-PerCP/Cy5.5, Anti-CD4-FITC, Anti-CD8-PE, Anti-CD86-PE, Anti-CD80-PB450-A, Anti-CD206-AF647, Anti-Foxp3-APC, and Anti-CD11c-FITC were obtained from BD Biosciences (USA).
Preparation of ENFORCE
An Endoglin-targeting peptide (palmitic-k (palmitic)-AHKHVHHVPVRL) was synthesized by solid-phase peptide synthesis. The molecular weight of the conjugated peptide was determined by Electrospray ionization mass spectrometry (ESI-MS) (Thermo, USA). The Endoglin-targeting peptide constituted a amphiphilic molecule with a hydrophobic insert and a hydrophilic targeting head group.
It was incorporated into the liposome membrane via hydrophobic interactions, with the targeting moiety exposed on the outer surface of the nanoparticles. Liposome nanoparticles were prepared using the thin-film hydration method. DOPC, cholesterol, cardiolipin, DSPE-PEG2000 and DSPE-PEG2000−Mal were dissolved in chloroform at a concentration of 50 g/l.
The endoglin-targeting peptide, CA4 and R848 were dissolved separately in 10 g/l methanol. Mix in a mass ratio of lipids:peptide:CA4:R848 = 20:1:0.5:0.5. After solvent evaporation with a rotary evaporator, add 1.0 ml ddH2O for hydration and stir overnight.
The lipid solution was then subjected to ultrasonic disruption. Finally, the solution was dialyzed for 4 h in ultrapure water (MWCO: 3500 Da) to remove excess components, yielding ENFORCE. Although CA4 and R848 shared the same feed mass ratio, differing physical and chemical properties would produce inconsistent encapsulation rates.
Similarly, empty liposomes (designated “Liposomes”) containing neither peptides nor active drugs, and liposomes loaded with CA4 and R848 separately (designated “Lipo-CAR”), were prepared through analogous methods for use in related experimental studies.
Characterization of ENFORCE
The morphology of the nanoliposomes was examined using transmission electron microscopy (TEM) (ZEISS, Germany). The particle size, polydispersity index (PDI), zeta potential, and stability of the nanoliposomes were measured using a Malvern Laser Particle Sizer (Malvern Panalytical Ltd., USA). The contents of CA4 and R848 within the nanoliposomes were quantified using high-performance liquid chromatography (HPLC) (Agilent Technologies, USA).
Cell culture
Cultured mouse mammary carcinoma cell line (4T1), mouse fibroblasts (3T3), human umbilical vein endothelial cells (HUVECs) and mouse monocyte macrophages (Raw264.7) in RPMI-1640 or DMEM medium supplemented with 10% fetal bovine serum (FBS) and 1 & antibiotics. The cells were maintained in an incubator at 37 °C and 5% CO2.
Endoglin expression assay
Western blot analysis was performed to detect endoglin expression in 4T1, 3T3 and HUVEC cells. Total protein was quantified using a BCA protein assay kit, separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene fluoride (PVDF) membrane, then incubated with 5% skim milk at room temperature for 1 h. The membrane overnight incubated at 4 °C with α-Endoglin (1:2000).
Secondary day, it further incubated with secondary antibody (1:2000) for 2 h. Visualization was performed using a gel imaging system (BIO-RAD, USA), and band intensity was quantified using Image J software for statistical analysis. Despite identical protein loading amounts, GAPDH levels varied among different cell types.
Endogenous proteins require normalization to GAPDH for relative comparison.
Cellular targeting
Confocal laser scanning microscope (CLSM) (Leica, Germany) and flow cytometry (Beckman, USA) were used to detect the cellular uptake behavior of cells. 4T1, 3T3, HUVEC cells were incubated with different concentrations of ENFORCE (CA4: 2, 4, 8 mg/l, containing Cy5.5), Lipo-CAR (containing Cy5.5) and Liposomes (containing Cy5.5) for 8 h in the dark. After washing three times with PBS, the cells were stained with Hoechst 33342 for 30 min and observed under CLSM.
Fluorescence intensity was then measured using flow cytometry. In addition, HUVECs were incubated with α-Endoglin for 4 h before ENFORCE administration, and cell uptake was then observed via CLSM.
MTT assay
4T1 cells and HUVECs were seeded into 96-well plates and incubated overnight. The cells were then treated with gradient concentrations of ENFORCE, CA4 and R848 for 24 h, respectively. Then 20 µl methyl thiazolyl tetrazolium (MTT) solution (5 mg/mL) was added to each well and incubated for 4 h.
The culture medium was then removed, and 150 µl of DMSO was added to dissolve the formazan crystals. The absorbance at 570 nm was measured using a multi-functional enzyme labeling instrument to calculate cell viability.
Cell apoptosis detection
4T1, 3T3, and HUVEC cells were seeded into 6-well plates and incubated overnight. The cells were subsequently treated with ENFORCE (CA4: 2 mg/l; R848: 0.9 mg/l), Lipo-CAR (CA4: 2 mg/l; R848: 0.9 mg/l), Liposomes, CA4 (2 mg/l), and R848 (0.9 mg/l) for 24 h, respectively. After treatment, the cells were stained with Annexin V-FITC/PI for 15 min.
Apoptosis was evaluated by flow cytometry analysis of intracellular fluorescence.
Scratch migration test
HUVECs were seeded into 6-well plates and incubated overnight. A scratch wound was created using the tip of a pipette. The cells were treated with ENFORCE (CA4: 0, 2, 4, 8 mg/l), Lipo-CAR (CA4: 8 mg/l; R848: 3.59 mg/l), Liposomes, CA4 (8 mg/l), and R848 (3.59 mg/l), respectively.
Cell migration was documented by capturing images at 0 and 24 h using an inverted microscope (Leica, Germany). Migration was quantified using Image J.
Transwell invasion assay
Matrigel was added to the upper chamber of transwell inserts and incubated at 37 °C for 1 h to allow gelation. Subsequently, 200 µl HUVECs suspension (2 × 104 cells) containing ENFORCE (CA4: 0, 2, 4, 8 mg/l), Lipo-CAR (CA4: 8 mg/l; R848: 3.59 mg/l), Liposomes, CA4 (8 mg/l), or R848 (3.59 mg/l) was seeded into the upper chamber. Culture medium (600 µl) containing 10% FBS was added to the lower chamber, and the cells were incubated for 24 h.
After incubation, the cells were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet. The membranes were photographed using an inverted microscope, and the images were analyzed using Image J.
Formation and destruction of capillary-like tube structures
Add 50 µl of matrix gel to a 96-well plate and cure at 37 °C for 1 h. HUVECs (1.8 × 104 cells) were seeded onto the matrix gel-coated culture plates and treated with different concentrations of ENFORCE (CA4: 8 mg/l; R848: 3.59 mg/l), CA4 (8 mg/l) and R848 (3.59 mg/l) for 4 h. Lumen formation was then observed under an inverted microscope, and the images were analyzed with Image J software.
Macrophage polarization assay
Raw264.7 cells were treated with IL-4 (0.1 mg/l) for 72 h, after which LPS (1 mg/l), ENFORCE (CA4: 8 mg/l, R848: 3.59 mg/l), CA4 (8 mg/l) and R848 (3.59 mg/l) were added separately, and the cells were incubated for a further 24 h. After cell collection, total protein was extracted and quantified using a BCA protein assay kit. The protein samples were separated by SDS-PAGE and transferred to a PVDF membrane.
Following transfer, the membrane was blocked with 5% skimmed milk for 1 h. After washing with tris buffered saline tween (TBST), the membrane was incubated overnight at 4 °C with antibodies (1:2000) against anti-iNOS, anti-Arg1, anti-CD80, and anti-CD206. On the following day, the membrane was incubated for 1 h with either anti-rabbit IgG (1:2000) or anti-mouse IgG (1:2000).
Visualization was performed using a gel imaging system (BIO-RAD, USA), and band intensity was quantified using Image J software for statistical analysis.
Fluorescence imaging in vivo
All animal experiments were conducted in strict accordance with the guidelines of the Institutional Animal Care and Use Committee (IACUC) of the Animal Experimentation Center of Guangzhou Medical University (GY2024-041), as well as the Regulations on the Management of Laboratory Animal Affairs. A mouse breast cancer tumor model was established by subcutaneous injection of 4T1 cells into the right hind limb of female BALB/c mice. One week later, liposomes, Lipo-CAR (containing 4.5 mg/kg CA4) and ENFORCE (containing 4.5 mg/kg CA4) were administered via the tail vein.
Fluorescence imaging was performed at 0.5, 2, 4, 8, 12 and 24 h post-injection. The mice were euthanized after 24 h, and fluorescence imaging was performed on the major organs. Tumor samples were collected for immunofluorescence staining for CD31, vWF, F4/80 and CD206.
Antitumor study in vivo
One week after subcutaneous inoculation of 4T1 cells into the right posterior back of BALB/c mice, the tumor-bearing mice were randomly divided into six groups. On Day 1, 4 and 7, saline, R848 (0.8 mg/kg), CA4 (1.8 mg/kg) and ENFORCE (containing 1.8 mg/kg CA4) were administered via the tail vein. The dosages of the aforementioned drugs were selected based on the maximum tolerated dose (MTD) identified in our preliminary screening and published data on CA4 nanomedicines as a benchmark for efficacy [9].
In preliminary toxicity studies, these doses did not cause significant weight loss or organ pathology, while matching the drug load in ENFORCE to ensure therapeutic exposure. αCD47 was administered intravenously on Day 2, 5 and 8. Tumor volume and body weight were measured every 2 d. Tumor volume was calculated using the formula: Tumor volume = (long axis) × (short axis)2 × 0.5.
After 21 d, the mice were euthanized, and images of the tumors, spleens, and lungs were captured. The weights of the tumors and spleens were recorded. Major organs (heart, liver, spleen, lung, kidney) and tumors were collected for H&E staining.
Blood samples were collected for biochemical analysis. Tumor sections were subjected to immunofluorescence staining for Ki-67, Tunel, CD8, CD4, Foxp3, CD86, CD80 and CD206. Spleen sections were used for immunofluorescence staining of CD8, CD4 and Foxp3.
To investigate the anti-metastatic effects of ENFORCE and ENFORCE + αCD47 in breast cancer, a 4T1 cell-based subcutaneous and tail vein injection model was established for lung metastasis. Following the treatment protocol for 21 d, the mice were euthanized, and the lungs were excised, photographed, and subsequently subjected to H&E staining.
Immune cell analysis in vivo
4T1 cells were subcutaneously injected into the right hind limb of BALB/c mice, which were then randomly assigned to six groups based on the treatment protocol described above. Following three treatment sessions, tumor and spleen tissue samples were collected for immune cell analysis. The tumor tissues were digested for 1 h in a tissue lysis buffer at 37 °C, followed by filtration to remove residual tissue fragments.
The obtained cells suspension were centrifuged, and collected for fluorescent antibody staining (including CD45, CD3, CD4, CD8, Foxp3, CD11C, CD80, CD86 and CD206), then analyzed by flow cytometry. Spleen tissues were homogenized and filtered before undergoing fluorescent antibody staining (CD45, CD3, CD4, CD8, Foxp3), followed by analysis via flow cytometry.
Statistical analysis
In each group for particle size, zeta potential, PDI, flow cytometry, western blot analyses, kit assay and in vivo blood samples analyses, 3 independent samples were measured. Five mice were compared in each group for in vivo antitumor studies. Statistical analysis and image processing were performed using GraphPad Prism 8.0.
The means of samples across multiple groups were compared using one-way ANOVA, followed by the Tukey-Kramer test for multiple comparisons. All data are presented as mean ± SD and statistical significance is defined as P < 0.05.
Results and discussion
Preparation and characterization of ENFORCE
The development of a tumor-targeting nanomedicine is pivotal for achieving site-specific drug delivery, thereby maximizing therapeutic efficacy while minimizing off-target toxicity [32,33]. Among the various available targeted ligands, endoglin has emerged as a promising molecular target due to its significant upregulation in rapidly proliferating endothelial cells in solid tumors [34,35]. Based on previous research [36], the amphiphilic chimeric peptide palmitic-k (palmitic)-AHKHVHHVPVRL was designed using the endoglin binding sequence.
Two hydrophobic palmitic acid moieties were incorporated into this structure to balance its amphiphilic nature and facilitate stable anchoring within the lipid bilayer, allowing effective integration into liposomal nanocarriers (Fig. S1A). ESI-MS confirmed the successful synthesis of the peptide, with a calculated molecular weight of 2032.6 and a corresponding [M + 2H]2+ signal at 1017.3 (Fig.
S1B).
Utilizing targeted modules, ENFORCE was constructed by incorporating amphiphilic peptides into conventional liposomes (Fig. 1A). In such a nano-framework, the hydrophobic palm-like tails readily insert into the lipid bilayer, with the hydrophilic targeting sequence retained on the outer surface, enabling endoglin-mediated vascular recognition [[37], [38], [39]].
The hydrophobic vascular disruptor CA4 and the immunomodulator R848 were co-encapsulated in liposomes, enabling the synergistic delivery of both drugs to the tumor microenvironment. TEM revealed that ENFORCE exhibits a uniform spherical morphology (Fig. 1B).
Dynamic light scattering measurements indicated an average hydrodynamic diameter of 123.63 ± 14.76 nm and a PDI of 0.16 ± 0.03, generally recognized as sitting within the acceptable window for passive tumor-internalization via the EPR effect (Fig. 1C). ENFORCE demonstrated excellent stability, with no significant changes in size or PDI observed over the course of 1 week (Fig.
1D& 1E). Empty liposomes (Liposomes) and non-targeted drug-loaded liposomes (Lipo-CAR) were used as controls. Lipo-CAR had a mean diameter of 136.10 ± 15.27 nm, whilst the Liposomes were slightly smaller at 114.87 ± 1.91 nm (Fig.
S2A&S2B). Both control formulations remained stable for at least 7 d (Fig. S2C&S2D).
Zeta potential analysis indicated that the negative surface charge of ENFORCE and Lipo-CAR generally enhances physiological stability by limiting non-specific protein adsorption and aggregation (Fig. 1F). HPLC confirmed the successful co-loading of the drugs, with concentrations of CA4 and R848 at 183.87 ± 1.64 mg/l and 82.59 ± 0.85 mg/l, respectively (Fig.
S2E&S2F and Table S1). In summary, the results demonstrate the successful construction of a structurally stable ENFORCE nanoframework, providing a solid foundation for subsequent biological evaluation.Fig. 1Characterization and cellular uptake of ENFORCE.
(A) Endoglin-targeting peptides assemble with phospholipids to form nanoparticle shells, enabling co-encapsulation of CA4 and R848 within a hydrophobic core; (B) Representative TEM images and (C) unimodal particle size distribution of ENFORCE. Scale bar: 300 nm; Changes in (D) particle size and (E) PDI of ENFORCE in PBS over 7 d at room temperature; (F) Zeta potential of several formulations; (G) CLSM images and (H) MFI of HUVECs after incubation with Cy5.5-labeled formulation for 8 h. Scale bar: 25 µm; (I) Flow cytometry histograms and (J) corresponding quantitative analysis of cellular uptake in HUVECs, 4T1, and 3T3 cells after treatment with varying concentrations (2, 4, 8 mg/l) of Cy5.5-labeled ENFORCE (n = 3).Fig 1 dummy alt text
Targeting specificity and anti-proliferative activity of ENFORCE
To assess the endothelial targeting capability of ENFORCE, the cellular uptake of different formulations in HUVECs were compared. As shown in Fig. 1G, CLSM imaging revealed all formulations were internalized by HUVECs to varying degrees compared with the untreated blank group.
However, the uptake efficiency of Liposomes and Lipo-CAR was suboptimal, with fluorescence signal intensity reaching only approximately 20%. In contrast, ENFORCE exhibited an uptake rate nearly twice as effective (with fluorescence signal intensity of approximately 40%), indicating the incorporation of the endoglin-targeting peptide significantly enhanced cellular internalization (Fig. 1H).
Additionally, HUVECs showed the highest endoglin expression, with ratios ∼1.4-fold and ∼2.4-fold above those of 3T3 and 4T1 cells, respectively (Fig. S3A–S3B). When ENFORCE was incubated with these three cell lines, flow cytometry analysis revealed a dose-dependent increase in fluorescence across all groups (Fig.
1I). However, HUVECs, which express the highest levels of endoglin, displayed substantially greater uptake than the other two cell lines. Quantitative analysis confirmed that, at the highest tested concentration, the fluorescence intensity in HUVECs was ∼3-fold and ∼1.8-fold higher than that in 3T3 and 4T1 cells, respectively (Fig.
1J). Notably, despite similar negative zeta potentials between ENFORCE and non-targeted Lipo-CAR (−17.50 ± 0.75 mV vs. −25.37 ± 0.06 mV, Table S1), ENFORCE uptake by HUVECs was roughly double that of Lipo-CAR. This difference ruled out non-specific electrostatic adsorption [40].
Instead, uptake across cell types tightly correlated with endoglin expression, pointing to the specific peptide-endoglin binding as the primary driver [41]. It was consistent with earlier reports on peptide-lipid systems, where sequence-specific receptor recognition, not charge effects, governed internalization [37,39]. To directly validate receptor mediation, HUVECs were pre-treated with an antibody (α-Endoglin) prior to incubation with ENFORCE (Fig.
S3C–S3D). This pretreatment significantly reduced the ENFORCE uptake, further confirming endoglin‑dependent internalization. These findings enhance ENFORCE’s affinity for endothelial cells that highly express endoglin, serving as a prerequisite for the effective disruption of blood vessels.
The antiproliferative effects of free CA4, free R848 and ENFORCE in HUVECs and 4T1 cells were assessed via MTT and apoptosis assays. As a typical vasodestructive agent, CA4 induces endothelial damage, leading to secondary tumor cell death [42]. Consistent with the mechanism, both ENFORCE and CA4 induced marked dose-dependent cytotoxicity in HUVECs (Fig.
2A), whereas only very mild cytotoxicity was observed in 4T1 cells (Fig. 2B). The observed differences are consistent with the biological roles of endothelial cells and tumor cells.
Endothelial cells express high levels of endoglin, whereas 4T1 cells exhibit lower endoglin expression and exhibit reduced sensitivity to the mechanism of action of CA4 [43]. ENFORCE induced nearly 30% apoptosis in HUVECs, significantly higher than the 15% apoptosis triggered by free CA4 (Fig. 2C& 2D).
This enhancement is attributed to peptide-mediated active targeting, facilitating more efficient internalization of ENFORCE [[44], [45], [46]]. The toxicity of free R848 alone on HUVECs was negligible, suggesting that the cytotoxic effect of ENFORCE stems primarily from CA4 rather than R848. In contrast, ENFORCE induced apoptosis in only approximately 2.5% of 4T1 cells, further supporting the notion that ENFORCE exhibits selective toxicity toward endothelial cells rather than cancer cells (Fig.
2C& 2E). This reflects the fact that the primary function of ENFORCE is to destabilize tumor blood vessels, rather than to directly induce apoptosis in tumor cells.Fig. 2In vitro cytotoxicity and targeted apoptotic induction by ENFORCE.
MTT assay showing dose-dependent cytotoxicity of free R848, free CA4 and ENFORCE in (A) HUVECs and (B) 4T1 cells. (C) Flow cytometry dot plots depicting apoptosis in HUVECs and 4T1 cells treated for 24 h with equal amounts of R848, CA4 or ENFORCE; Quantitative analysis of apoptosis rates in (D) HUVECs and (E) 4T1 cells, corresponding to the flow cytometry plots; (F) Representative flow cytometry dot plots and (G) statistical analysis of targeted apoptotic induction in HUVECs and 3T3 cells treated for 24 h with non-targeted Liposomes, non-targeted Lipo-CAR or targeted ENFORCE (n = 3).Fig 2 dummy alt text
To verify that the enhanced toxicity in HUVECs stemmed from peptide-mediated targeting, apoptosis assays were also performed on 3T3 cells (another low-endoglin control) (Fig. 2F). In HUVECs, ENFORCE induced approximately 20% apoptosis, significantly exceeding the levels observed in the Liposomes and Lipo-CAR groups (Fig.
2G). In contrast, the level of apoptosis induced by ENFORCE in 3T3 cells was only slightly higher than other groups, and the overall level of apoptosis remained far lower than that in HUVECs. This cell-type-specific difference strongly supports the proposition that the enhanced cytotoxicity of ENFORCE stems from its endoglin-targeted delivery mechanism.
These results indicate that ENFORCE not only exhibits good targeting in endoglin-expressing HUVECs but also effectively enhances the cytotoxicity of CA4, providing a basis for the vascular-destructive function of ENFORCE in tumor therapy.
Antivascular effects and immunomodulatory activities of ENFORCE in vitro
The anti-angiogenic activity of ENFORCE was evaluated through a series of assays targeting key steps in angiogenesis, including endothelial migration, invasion and tube formation. In the scratch assay, significant cell migration was observed in both the blank group and the R848-treated group after 24 h (Fig. 3A).
CA4 moderately inhibited migration, reduced to approximately 30%. In contrast, ENFORCE produced a markedly stronger inhibitory effect, limiting the migration rate to <20% (Fig. 3B).
Since R848 possesses immunomodulatory rather than cytotoxic properties, the anticipated impact appears negligible [47,48]. Compared with free CA4, the superior performance of ENFORCE indicates that peptide-mediated targeting enhances intracellular drug delivery and vascular disruption. A similar trend was observed in the matrix gel invasion assay.
Compared with the blank group, both CA4 and ENFORCE reduced endothelial cell invasion. However, ENFORCE achieved nearly twice the inhibitory effect of free CA4, reinforcing the ability of endoglin-targeted therapy to improve intracellular accumulation (Fig. 3C& 3D).
The effects on angiogenesis further corroborated this conclusion. The blank group formed an extensive and well-organized tubular network. R848 caused only moderate disruption, whereas CA4 significantly reduced the formation of cellular connections to just 0.16-fold compare with the blank group (Fig.
3E). It is worth noting that ENFORCE produced the most pronounced inhibition, reducing adhesion to 0.08-fold (Fig. 3F).
A direct comparison with non-targeted Lipo-CAR formulations confirmed that ENFORCE’s enhanced inhibitory effect on migration and invasion stems from endoglin receptor-mediated uptake, rather than being solely attributable to CA4 (Fig. S4). Thus, these findings confirm that ENFORCE effectively inhibits multiple angiogenic processes by combining targeted delivery with the efficient intracellular release of CA4.Fig.
3In vitro antivascular and immunomodulatory effects. (A) Representative comparative images and (B) quantitative analysis of scratch wound healing assays after 24 h treatment of HUVECs with Blank, free R848, free CA4, or equivalent doses of ENFORCE. Red lines indicate wound edges; (C) Representative images from the transwell invasion assay and (D) quantitative analysis of HUVECs invasion under each condition; (E) Representative images from in vitro angiogenesis experiments and (F) quantitative analysis of vascular junctions.
Scale bar: 200 µm; (G) Western blot bands of Raw264.7 cells and grayscale quantification of (H) Arg1, (I) CD206, (J) iNOS and (K) CD80 (n = 3).Fig 3 dummy alt text
Given that vascular disruption inadvertently promotes an immunosuppressive environment, the immunomodulatory capacity of ENFORCE was evaluated. Although it is currently widely assumed that CA4 disrupts tumor vasculature and releases recruitment signals, inducing the accumulation of immunosuppressive macrophages in the tumor microenvironment. CA4 also potentially intervenes directly in the polarization process, thereby exercising a regulatory role in macrophage polarization [49].
As shown in Fig. 3G–3I, treatment with free CA4 significantly up-regulated the M2-associated macrophage markers Arg1 and CD206, increasing their expression by approximately 3.8-fold and 2.4-fold, respectively. Conversely, R848 strongly promoted the polarization of macrophages toward the M1 phenotype, increasing the expression of iNOS and CD80 by approximately 7.7-fold and 3.4-fold, respectively, and simultaneously reducing M2 markers by 50% (Fig.
3J& 3K). It is worth noting that ENFORCE exhibited the strongest immunomodulatory effect across all treatment regimens. ENFORCE induced the expression of iNOS and CD80 at levels approximately 20-fold and 5-fold higher than those in the blank group, respectively.
Meanwhile, the expression of Arg1 and CD206 was suppressed to approximately 27% and 24% of the blank group’s levels, respectively. The CD80/CD206 ratio was approximately twice that observed with R848 monotherapy, reflecting a more effective shift toward the M1 phenotype. This enhancement appears to stem from ENFORCE’s ability to deliver R848 more effectively into cells and ensure its sustained release, prolonging the signaling for M1 macrophage polarization.
These results indicate that ENFORCE not only potentiates anti-angiogenic effects but also counteracts CA4-mediated immunosuppression, ultimately driving macrophages to adopt a potent anti-tumor M1 phenotype. Such synergistic anti-angiogenic and immunomodulatory activity addresses key limitations of conventional VDAs and demonstrates the therapeutic advantages of ENFORCE.
Tumor targeting ability and antitumor effect of ENFORCE in vivo
The biodistribution of Liposomes, Lipo-CAR, and ENFORCE was evaluated by real-time fluorescence imaging. After intravenous administration, signals were monitored at designated time points. As shown in Fig.
4A, all formulations exhibited detectable tumor signals within 0.5 h. However, the non-targeted Lipo-CAR group showed markedly stronger and more sustained accumulation than the Liposomes group, confirming that the released CA4 disrupted tumor vessels and enhanced permeability-mediated retention via the EPR effect. On this basis, ENFORCE exhibited even great accumulation, peaking at 2 h and remaining robust beyond 4 h.
Quantitative analysis at 24 h indicated that ENFORCE accumulation was approximately 10 times that of Lipo-CAR (Fig. 4B). To confirm tumor-specific retention, ex vivo fluorescence imaging was performed on resected tissues, including tumor, heart, liver, spleen, lung and kidney (Fig.
S5). The ENFORCE-treated tumor tissues displayed significantly higher fluorescence than any other organs. While a minor contribution from intercellular pathways activated during vascular stress could not be entirely excluded, the overall data revealed that ENFORCE achieved superior tumor targeting and prolonged retention, driven by the combined effect of CA4-induced vascular disruption and active endoglin binding, primarily through expansion of the paracellular space.
To evaluate subsequent anti-vascular and immunomodulatory effects, tumor tissue was collected 24 h after injection for further analysis. Immunofluorescence staining revealed significant differences in vascular integrity and immune cell infiltration between the groups (Fig. 4C).
Significant increases in the expression of von Willebrand factor (vWF) were observed following ENFORCE treatment, with significant aggregation within the vascular lumen, indicating enhanced vascular rupture and thrombus formation [50]. Compared with the lower intravascular vWF levels observed in the Liposomes and Lipo-CAR groups, active targeting by ENFORCE substantially amplified the vascular destructive effect, exceeding that mediated by passive EPR alone. At the same 24 h time point, a transient elevation of CD206 was observed, reflecting early monocyte recruitment in response to CA4-induced vascular injury, before R848-mediated repolarization to the M1 phenotype took effect.
Colocalization analysis of F4/80 and CD206 gave Pearson’s correlation coefficient (PCC) of 0.56 (Liposomes), 0.63 (Lipo-CAR), and 0.76 (ENFORCE), confirming early M2-type macrophage recruitment after vascular disruption (Fig. S6). Although the limited observation period was insufficient to observe the shift of macrophages from the suppressive M2 phenotype to the M1 phenotype, the changes in the immune microenvironment laid the groundwork for subsequent combination therapy.Fig.
4In vivo antitumor efficacy of ENFORCE. (A) In vivo fluorescence imaging of 4T1 tumor-bearing BALB/c mice at 0.5, 2, 4, 8, 12 and 24 h after intravenous injection of Cy5.5-labeled Liposomes, Lipo-CAR or ENFORCE. Color scale: Min=0.5e2, Max=3.4e2 p/sec/cm²/sr / (μW/cm²); (B) Quantitative analysis of average fluorescence intensity at tumor sites; (C) Immunofluorescence staining of tumor tissues from mice sacrificed 24 h post-injection.
Scale bar: 100 µm; (D) Tumor growth curves of 4T1 tumor-bearing mice treated with different regimens (saline, free R848, free CA4, ENFORCE, αCD47, ENFORCE + αCD47); (E) Representative photographs of tumors excised from representative mice in each treatment group at the study endpoint (Day 21); (F) Mean tumor weight at the endpoint for each group (n = 5); (G) Representative histological and immunohistochemical images of tumor sections from each group. Scale bar: 100 µm; (H) Percentage of Tunel-positive (apoptotic) area; (I) Percentage of Ki-67-positive (proliferating) nuclei; (J) CD31/vWF ratio, indicating vascular integrity/damage; (K) Relative vWF-positive area; (L) Relative CD47-positive staining intensity (n = 3).Fig 4 dummy alt text
To investigate whether ENFORCE enhances the penetration and retention of the αCD47 antibody through vascular modulation, a comprehensive treatment regimen was implemented. Saline, R848, CA4 and ENFORCE were administered via tail vein injection on Day 1, 4 and 7, respectively, while αCD47 was administered on Day 2, 5 and 8. As shown in Fig.
4D, tumors in the CA4 and αCD47 monotherapy groups exhibited progressive growth, with no significant inhibition observed. In contrast, the R848 treatment group demonstrated moderate tumor suppression. It is worth noting that ENFORCE demonstrated significantly enhanced anti-tumor activity, with a tumor growth inhibition rate of approximately 80%.
The most significant therapeutic effect was observed in the ENFORCE + αCD47 combination group. Tumor weights measured at the end of the study further corroborated our findings (Fig. 4E& 4F).
The results showed that, compared with the blank group, the mean tumor weight in the ENFORCE + αCD47 combination group was reduced by 95%. Mechanistically, CA4-mediated vascular disruption is known to increase antibody retention in tumors [[51], [52], [53]]. In this work, the substantially greater accumulation of ENFORCE over that of non-targeted Lipo-CAR indicated that CA4-enhanced permeability indeed improved nanoparticle retention.
This transient hyperpermeability, together with R848-mediated immune reprogramming, created a favorable environment for αCD47 delivery. The synergistic effect, comprising enhanced vascular leakiness, improved antibody bioavailability, and modulation of immunosuppressive macrophages, collectively overcame key barriers such as poor drug penetration, immune evasion, and a hostile tumor microenvironment.
Based on the observed antitumor effects, further histopathological analysis was conducted. As illustrated in Fig. 4G, Tunel and Ki-67 staining revealed significantly increased apoptosis and reduced proliferation in both ENFORCE-treated groups.
The apoptotic intensity in the ENFORCE group reached 218.24 ± 15.92, compared to 54.87 ± 8.42 in blank, while the Ki-67 positive rate decreased to 147.88 ± 12.02 from 191.81 ± 9.11 (Fig. 4H& 4I). The combination group showed enhanced effects, with 250.86 ± 5.34 apoptosis and only 85.79 ± 13.54 Ki-67 positivity.
Hematoxylin and eosin (H&E) staining confirmed extensive cellular damage in both the ENFORCE group and the combination group, including nuclear atrophy and cytoplasmic disruption. Immunofluorescence analysis demonstrated profound vascular disruption, which was further quantified by the CD31/vWF ratio (Fig. 4J).
In the ENFORCE group, the CD31/vWF ratio decreased to 1.35, compared to 2.79 in the blank group, indicating a severe loss of functional vasculature. It was evident that the vWF intensity in the ENFORCE group increased by 1.6-fold compared to the blank group, indicating that platelets recruited following vascular injury form localized thrombi (Fig. 4K).
The combination group showed more pronounced effects, with the CD31/vWF ratio further declining to 0.5, while vWF intensity rose 3.8-fold. These results confirmed that ENFORCE treatment not only reduced vascular density but also impaired vascular function, with the combination therapy producing the most substantial disruption. CD47 expression was significantly reduced in the combination group, achieving 60% suppression compared to blank (Fig.
4L). This surpassed the 20% reduction seen with αCD47 monotherapy, confirming that ENFORCE pretreatment enhanced immune checkpoint blockade. Although αCD47 was not directly labeled to track its distribution within tumors, the current data suggested that ENFORCE enhanced αCD47 delivery through CA4-mediated vascular remodeling, which accounted for the superior efficacy of the combined treatment.
Anti-metastasis effect of ENFORCE in vivo
A 4T1 breast cancer metastasis model was used to evaluate the anti-metastatic potential of ENFORCE. Spleen and lung tissues were collected for analysis in accordance with the pre-determined treatment protocol. As shown in Fig.
5A& 5B, the spleens in the blank group, the CA4 group and the αCD47 monotherapy group were markedly enlarged, with spleen weights exceeding five times that of the healthy control group. The degree of splenomegaly was significantly reduced in the R848 group, the ENFORCE group and the ENFORCE + αCD47 group. The spleen weight in mice from the ENFORCE + αCD47 group was closest to that of the healthy control group.
Histopathological analysis of spleen tissue revealed differences in microstructure (Fig. 5C). The boundary between the red pulp and white pulp was indistinct in the blank, CA4 and αCD47 groups, and the area of white pulp was reduced by approximately 60% compared with the healthy control group.
In contrast, the ENFORCE group and the combination therapy group maintained near-normal spleen structure, with a clear boundary between the red pulp and white pulp regions preserved at 70%−90% of normal levels. The normalization of microstructure suggests a restoration of immune function. At the same time, H&E staining confirmed extensive metastatic infiltration in the lungs of the blank group, whereas the ENFORCE + αCD47 group exhibited only minimal metastatic foci.
Furthermore, lung architecture was largely preserved, with cell density similar to that of healthy controls (Fig. 5D). These results indicate that ENFORCE not only inhibits the growth of the primary tumor but also suppresses tumor metastasis by restoring the function of the spleen and lungs.
The superior efficacy of the combination therapy demonstrates that combining vascular targeting with immune checkpoint inhibition is of great significance for the comprehensive control of tumor metastasis [54,55].Fig. 5Anti-metastasis effect of ENFORCE. (A) Representative macroscopic photographs of spleens excised from healthy mice and 4T1 tumor-bearing mice; (B) Quantitative analysis of spleen weights from each treatment group (n = 5); (C) Representative H&E stained histological sections of spleen tissues; (D) Representative gross morphology of lungs and corresponding H&E stained sections from each group.
Scale bar: 300 µm.Fig 5 dummy alt text
Immune activation behavior of ENFORCE in vivo
To further elucidate the immunological mechanisms underlying the antitumor efficacy of ENFORCE, comprehensive immunofluorescence and flow cytometry analyses were conducted to evaluate immune cells in tumor and spleen tissues following different treatments (gating details in Figs. S7 and S8). In tumor tissues, the infiltration and activation of T lymphocyte subsets were significantly modulated by the treatments.
As shown in Fig. 6A, stronger fluorescence signals for CD4 and CD8 were detected in the ENFORCE and combination therapy groups, while signals in monotherapy groups showed moderate enhancements. Conversely, the Foxp3 signal was markedly diminished in the ENFORCE + αCD47 group compared to other treatments.
The percentage of CD3+CD4+ T helper cells in the blank group was 3.9%, as compared to 10% in the R848 group (Figs. 6B and S9A). This proportion was increased to 20.6% in the ENFORCE group and further elevated to 32.6% in the ENFORCE + αCD47 combination group.
A more pronounced differential effect was observed for CD3+CD8+ cytotoxic T lymphocytes (Figs. 6C and S9B). The baseline level of 7.79% in the blank group was marginally increased to 15.7% by R848.
The ENFORCE group showed a robust increase to 30.24%, while the most substantial enhancement was achieved by the combination therapy, which increased the population of cytotoxic T lymphocytes to 38.9%, representing an approximately 5-fold increase compared to the blank. These findings indicated that ENFORCE effectively promoted the infiltration and activation of both helper and cytotoxic T cells, with the effect being synergistically enhanced by αCD47 co-administration. However, the effect on regulatory T cells (Tregs) was found to be opposite.
A moderate reduction in Tregs was observed across all monotherapies (Figs. 6D and S9C). The most significant suppression occurred in the ENFORCE + αCD47 group, reducing Tregs to 25% of blank levels.
To provide a more reliable assessment of intratumoral immunity, the ratio of CD8+ T cell and Treg was further calculated. This ratio was significantly higher in the combination group than in all others, confirming that the treatment effectively reverted the immune landscape from suppression toward permissiveness (Fig. S10A).Fig.
6Comprehensive immunophenotyping of tumor tissues. (A) Representative confocal immunofluorescence microscopy images of tumor tissue sections. Scale bar: 50 µm.
Flow cytometric quantification of (B) CD3+CD4+ T helper cells within the leukocyte gate, (C) CD3+CD8+ cytotoxic T lymphocytes, and (D) Foxp3+ Tregs; (E) Percentage of dendritic cells expressing the co-stimulatory maturation markers CD80 and CD86; (F) Percentage of TAMs expressing the M1 marker CD80; (G) Percentage of TAMs expressing the M2 marker CD206 (n = 5).Fig 6 dummy alt text
The impact of ENFORCE on innate immune cell populations was further investigated. Immunofluorescence staining revealed a significant increase in CD80+CD86+ signal intensity and a marked decrease in CD206+ signal intensity in the ENFORCE group and its combination group compared with the other treatment groups (Fig. 6A).
Flow cytometry showed that ENFORCE treatment significantly promoted the maturation of dendritic cells. The proportion of mature dendritic cells (CD80+CD86+) in the blank group was 34.8% (Figs. 6E and S9D).
This proportion was increased to 50.5% in the ENFORCE group and further elevated to 67.4% in the ENFORCE + αCD47 combination group. In contrast, only modest increases were observed in the R848 monotherapy. In parallel, macrophage polarization was significantly modulated by the treatments (Figs.
6F& 6G and S9E). Macrophages, as key innate immune effectors, can directly phagocytize tumor cells [56]. The proportion of M1 macrophages (F4/80+CD80+) was increased from 12% in the blank to 36% in the ENFORCE + αCD47 group.
Conversely, the proportion of M2 macrophages (F4/80+CD206+) was correspondingly decreased in the ENFORCE and combination treatment groups. This shift in macrophage polarization from the immunosuppressive M2 phenotype to the pro-inflammatory M1 phenotype was consistent with the previously observed immunomodulatory effects of R848, which were substantially enhanced by ENFORCE.
An analysis of immune cell populations within splenic tissue was conducted to investigate the systemic immunological effects of these therapeutic approaches. The proportion of CD3+CD4+ T helper cells in the blank group was measured at 22.6%. This percentage was increased to 36.6% in the ENFORCE + αCD47 combination group (Fig.
7A& 7B). Similarly, CD3+CD8+ cytotoxic T lymphocytes showed a substantial enhancement from 9.4% in blank to 16.4% in the combination therapy group (Fig. 7C).
Treg populations were significantly modulated by the treatments. Foxp3+ Tregs exhibited similar outcomes across the R848, ENFORCE monotherapy and combination therapy groups. Their mean fluorescence intensity (MFI) decreased from 23.4% in the blank group to ∼5%, indicating a substantial reduction in the population of this immunosuppressive cell subset (Fig.
7D). In line with the intratumoral data, the splenic CD8+ T cell/Treg ratio rose significantly in the ENFORCE + αCD47 group, further supporting the systemic nature of anti-tumor immune reprogramming (Fig. S10B).
In immunofluorescence staining of spleen tissue, stronger CD4 and CD8 signals were observed in the ENFORCE group and the combination therapy group, while Foxp3 expression was significantly suppressed (Fig. 7E). The differential effects observed between treatment groups emphasize ENFORCE’s superior immunomodulatory capacity.
The immune activation observed in spleen tissue was associated with anti-tumor efficacy, suggesting the occurrence of a systemic anti-tumor immune response. This compelling evidence indicates that ENFORCE, particularly when used in combination with αCD47, effectively overcomes tumor-induced immunosuppression. Our findings are consistent with recent liposome-based co-delivery studies that reprogram the tumor microenvironment via anti-angiogenic and immunomodulatory agents [57].
However, ENFORCE uniquely combines vascular endothelial targeting with injury-induced self-aggregation, and microenvironment remodeling, offering a more targeted and potent immunotherapeutic option for breast cancer.Fig. 7Systemic immune activation in spleen tissues. (A) Representative flow cytometry data illustrating immune cell populations in the spleen; Quantitative analysis of (B) CD3+CD4+helper T cells, (C) CD3+CD8+ cytotoxic T lymphocytes, and (D) Foxp3+ regulatory T cells (Tregs) (n = 5); (E) Representative immunofluorescence staining in spleen tissues.
Scale bar: 50 µm.Fig 7 dummy alt text
Biosafety evaluation in vivo
The biosafety of ENFORCE was systematically evaluated. Body weight was continuously monitored as an initial and general indicator of systemic toxicological responses. As shown in Fig.
8A, compared with the control group, there were no significant fluctuations in body weight in either the blank group or any treatment groups, indicating that none of the treatments induced significant systemic toxicity. To further assess potential organ-specific toxicity, comprehensive serum biochemical analyses were conducted. Liver function was assessed through the measurement of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels (Fig.
8B& 8C). Levels of both markers remained within the normal physiological range in all treatment groups. Similarly, renal function indicators, including blood urea nitrogen (BUN) and uric acid (UA), showed no significant alterations following treatment (Fig.
8D& 8E). Cardiac safety was also assessed by measuring creatine kinase (CK) and lactate dehydrogenase (LDH) levels. CK and LDH levels remained within safe ranges in all treatment groups (Fig.
8F& 8G). While current data indicated that the combination of ENFORCE and αCD47 did not cause significant weight loss or notable hematological changes, it is still necessary to conduct a comprehensive blood test. Since anti-CD47 therapy may induce mild neutropenia and short-term anemia, continuous complete blood count (CBC) monitoring will be implemented [12].
Even so, histopathological analysis further confirmed the favorable biosafety profile of ENFORCE. H&E staining was performed on major organs such as the heart, kidneys and liver. As shown in Fig.
8H, no significant pathological abnormalities such as tissue necrosis, inflammatory infiltration or structural degeneration were observed in the organs of any of the treatment groups. Importantly, the tissue morphology of mice treated with ENFORCE was comparable to that of healthy controls. It is worth noting that all biochemical parameters and histological features in the ENFORCE + αCD47 combination therapy group were similar to those in the control group, indicating enhanced anti-tumor efficacy without an increase in systemic or organ-specific toxicity.
Given the typical behavior of PEGylated liposomal carriers, ENFORCE is expected to undergo predominant hepatosplenic uptake, with lipid components hydrolyzed by phospholipases into endogenous metabolites and cleared via biliary and renal routes [58]. Overall, these safety assessments indicate that ENFORCE demonstrates acceptable biological safety, whether administered alone or in combination with αCD47, supporting the translational promise of this regimen.Fig. 8Biosafety analysis in vivo.
(A) Body weight changes of 4T1 tumor-bearing BALB/c mice over the 21-d treatment period. Serum biochemical analyses of (B) ALT, (C) AST, (D) BUN, (E) UA, (F) CK and (G) LDH after different treatments (n = 5); (H) H&E staining of major organs (heart, kidney, and liver) to assess histopathological changes following different treatments. Scale bar: 300 µm.Fig 8 dummy alt text
Conclusion
Our study has developed ENFORCE, an endoglin-targeted nano-framework that utilizes the vascular damage caused upon VDAs treatment for amplifying drug delivery, concurrently reprogramming the tumor immune microenvironment. ENFORCE actively anchors to the tumor vasculature upon insertion of an endoglin-targeting peptide into the liposome bilayer. The release of CA4 disrupts the integrity of the vascular endothelium and initiates a hyperpermeable window.
This vascular remodeling amplifies EPR-mediated nanomaterial accumulation, resulting in a 10-fold increase in tumor retention. In parallel, R848 drives the repolarization of immunosuppressive macrophages recruited by VDAs into an immunostimulatory phenotype. The therapeutic strategy also significantly enhances the intratumoral bioavailability of αCD47.
When combined with αCD47 blockade, systemically enhanced anti-tumor immunity is further amplified, delivering potent suppression of both primary and metastatic tumors. In summary, this study has established a promising targeted nano-framework that simultaneously overcomes the dual challenges of limited drug delivery and the treatment-induced immunosuppressive tumor microenvironment, presenting a potential strategy for enhanced clinical outcomes in breast cancer immunotherapy.
CRediT authorship contribution statement
Yi Cen: Conceptualization, Data curation, Investigation, Writing – original draft. Xiaocheng Ou: Data curation, Investigation. Ying Chen: Investigation.
Zhuofeng Li: Investigation. Xinxuan Li: Investigation. Baixue Yu: Investigation.
Xiayun Chen: Investigation. Lingwen Ding: Supervision. Hong Cheng: Supervision, Writing – review & editing.
Yuqing Wang: Supervision, Writing – review & editing. Shiying Li: Conceptualization, Project administration, Supervision, Writing – review & editing.
Conflicts of interest
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.