Synergistic anti-tumor effects of novel two-domain soluble Fms-like tyrosine kinase-1 and paclitaxel on three-dimensional breast cancer models: implications for targeted therapy Authors: Adel Zaid I Mutahar, Renu Dayal, Bharathi P Salimath Published: 2025 ABSTRACT AbstractBackgroundThree-dimensional (3D) multicellular spheroids mimic tumor architecture, diffusion limits, and drug resistance more effectively than two-dimensional (2D) monolayers. Triple-negative breast cancer (TNBC) remains highly aggressive, with limited benefits from anti-vascular endothelial growth factor (VEGF) therapy. We investigated the synergistic effects of a novel two-domain soluble Fms-like tyrosine kinase-1 (2d-sFlt-1) in enhancing paclitaxel efficacy compared with bevacizumab using 3D breast cancer models.Methods3D spheroids derived from triple-negative (MDA-MB-231 and MDA-MB-468) and luminal (MCF-7) breast cancer cell lines, as well as tumor–endothelial co-culture spheroids, were treated with paclitaxel alone or in combination with 2d-sFlt-1 or bevacizumab. Treatment efficacy was evaluated through drug-sensitivity, proliferation, viability, migration, and morphometric analyses, along with assessments of angiogenesis, epithelial-to-mesenchymal transition (EMT), and focal adhesion kinase (FAK) signaling. Statistical significance was determined using non-parametric tests or one- and two-way ANOVA with appropriate multiple comparisons.ResultsTransition from 2D monolayer to 3D spheroid culture markedly increased paclitaxel resistance, with IC₅₀ values elevated by roughly 25-fold. Co-treatment with 2d-sFlt-1 markedly enhanced paclitaxel efficacy, reducing the IC₅₀ values by approximately 10-fold in MDA-MB-231 spheroids and 6-fold in MDA-MB-468 spheroids, whereas bevacizumab produced only modest effects. Paclitaxel and 2d-sFlt-1 combination further suppressed spheroid growth and proliferation in both TNBC and luminal breast cancer models. In MDA-MB-231 spheroids, combination therapy reversed EMT by increasing E-cadherin while downregulating N-cadherin, SNAIL, and TWIST, and further reduced VEGF secretion and angiogenic tube formation. In tumor–endothelial co-cultures, FAK signaling was markedly reduced after combination treatment. Across all models and assays, paclitaxel and 2d-sFlt-1 combination consistently outperformed monotherapy and paclitaxel plus bevacizumab, underscoring its potential as a mechanistically synergistic strategy for overcoming chemoresistance in triple-negative breast cancer.Conclusions2d-sFlt-1 enhanced paclitaxel efficacy in physiologically relevant 3D breast cancer models. In MDA-MB-231 spheroids, the combination reduced chemoresistance and invasive growth through coordinated modulation of angiogenesis, EMT, and FAK signaling. In MDA-MB-468 and MCF-7 spheroids, it primarily increased cytotoxicity and growth inhibition, indicating a conserved functional benefit across subtypes. These findings provide a rationale for further mechanistic validation and preclinical evaluation of 2d-sFlt-1 plus paclitaxel to define dosing, safety, and translational feasibility in aggressive breast cancers.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12967-025-07585-x.Keywords: 3D-spheroids, Combinatorial therapy, Metastasis, Angiogenesis, Co-culture Background The global incidence of cancer continues to rise, posing significant challenges in oncology, particularly in ensuring timely and affordable access to effective therapies [1, 2]. Among various types of cancer, metastatic breast cancer, particularly triple-negative breast cancer (TNBC), remains one of the most aggressive and therapeutically challenging subtypes. TNBC is characterized by the absence of estrogen receptors (ER), progesterone receptors (PR), and human epidermal growth factor receptor 2 (Her2), which limits the effectiveness of targeted therapies and results in a poor prognosis for affected patients [3]. Despite advances in treatment strategies, metastatic TNBC remains a leading cause of cancer-related mortality worldwide [4], highlighting the urgent need for novel therapeutic approaches to treat this disease. Chemotherapy remains the cornerstone of TNBC treatment, with paclitaxel among the most commonly used chemotherapeutic agents. Paclitaxel disrupts microtubule dynamics, inhibits cell division, and promotes cell death [5, 6]. Although effective, the clinical utility of paclitaxel is limited by the development of drug resistance and systemic toxicity, thereby necessitating the exploration of alternative strategies to enhance its efficacy and minimize associated side effects [7]. Angiogenesis, the formation of new blood vessels, plays a critical role in tumor growth and metastasis, particularly in TNBC. The vascular endothelial growth factor (VEGF) pathway is a key regulator of angiogenesis, and its overexpression is often associated with poor prognosis in patients with TNBC [8, 9]. Consequently, targeting VEGF and its receptors has emerged as a promising strategy for inhibiting tumor vasculature and restricting tumor progression. Anti-VEGF therapies aim to block this pathway, reduce the blood supply to tumors, and limit their potential for growth and metastasis. Combining chemotherapy with anti-angiogenic agents, such as anti-VEGF therapies, can synergistically enhance therapeutic outcomes [10]. Preclinical and clinical studies have suggested that this dual strategy may overcome some limitations of monotherapy, reduce drug resistance, and improve patient outcomes [7, 11]. Soluble Fms-like tyrosine kinase 1 (sFlt-1), a soluble form of the VEGF receptor, has shown promise as an anti-angiogenic agent. As a decoy receptor, sFlt-1 binds to VEGF and prevents its interaction with endothelial cell surface receptors, thereby inhibiting angiogenesis [12]. Unlike traditional monoclonal anti-VEGF antibodies, sFlt-1 offers a more selective means of targeting VEGF signaling pathway, potentially leading to fewer adverse effects and enhanced therapeutic efficacy [13]. In this study, we developed a novel small molecule containing two domains of soluble sFlt-1 (2d-sFlt-1), which binds to VEGF receptors with high affinity, aiming to disrupt tumor vasculature and inhibit tumor growth when combined with paclitaxel. Although two-dimensional (2D) cell cultures are commonly used, they fail to accurately replicate the complex interactions between cancer cells and their microenvironment [14]. In contrast, three-dimensional (3D) spheroid models more closely mimic in vivo tumor architecture and provide a more physiologically relevant platform for evaluating therapeutic responses. In particular, 3D cultures facilitate the study of epithelial-to-mesenchymal transition (EMT), a process that enhances cell migration and invasion – hallmarks of metastatic cancers [15, 16]. EMT is characterized by the loss of epithelial markers, such as E-cadherin, and the gain of mesenchymal markers, such as vimentin [17, 18]. Given these advantages, 3D spheroids represent a more accurate model system for studying tumor invasiveness, metastasis, and combinatorial therapies. This study investigated the synergistic effects of paclitaxel and 2d-sFlt-1 on breast cancer using physiologically relevant 3D spheroid models. We elucidated the therapeutic impact of this combination in 3D breast cancer spheroids, with detailed mechanistic characterization performed in MDA-MB-231 cells and functional validation in additional breast cancer cell lines. By evaluating spheroid growth, viability, invasion, and angiogenic signaling, we demonstrated that 2d-sFlt-1 enhanced paclitaxel activity in a microenvironment-mimetic context. Together, these studies aim to establish a rationale for advancing this combination therapy into subsequent preclinical evaluation using advanced in vitro and in vivo models to assess its translational feasibility and guide future therapeutic strategies for aggressive breast cancer subtypes. Materials and methods Cell culture Breast cancer cell lines MDA-MB-231, MDA-MB-468 and MCF-7 were obtained from the National Center for Cell Science (NCCS; Pune, India). MDA-MB-231 cells were maintained in Leibovitz’s L-15 medium (11415-064; Gibco, Invitrogen, USA), whereas MDA-MB-468 and MCF-7 cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; 11965118, Gibco). All media were supplemented with 10% fetal bovine serum (FBS; 10270106, Gibco, Invitrogen, USA), 100 U/mL penicillin, 100 µg/mL streptomycin, and 500 µg/mL L-glutamine (15140-122; Gibco Life Technologies, Thermo Fisher, USA). MDA-MB-231 cells were cultured at 37 °C in a CO₂-free incubator, whereas MDA-MB-468 and MCF-7 cells were maintained at 37 °C in a humidified incubator with 5% CO₂. All cell lines were authenticated and confirmed to be free of mycoplasma contamination by the NCCS. Human Umbilical Vein Endothelial Cells (HUVECs, passage 2, CL002; Hi-Media, Mumbai, India) were cultured in HiEndoXL™ Endothelial Cell Expansion Medium (AL530; Hi-Media, India), supplemented with 2% FBS and 1% penicillin-streptomycin. HUVECs were detached using EnVzyme™ easy Endothelial Cell Dissociation Solution (TCL137; Hi-Media, India) and expanded under standard conditions. For experiments, cells were passaged at 80% confluence using 0.25% trypsin-EDTA (25300054; Gibco, Thermo Fisher, USA). Drugs, antibodies, and reagents Matrigel (Corning® Matrigel Matrix, 354248; Sigma-Aldrich, Germany) was used for the extracellular matrix-based assays. Methylcellulose (4000 cP, M0512; Sigma-Aldrich, Germany) was used to increase the viscosity of the medium and promote spheroid formation by preventing cell attachment to the culture surface. Hoechst 33342 (H3570; Molecular Probes, Thermo Fisher Scientific, USA), ethidium homodimer-1 (EthD-1, 46043; Sigma-Aldrich), resazurin sodium salt (R7017; Sigma-Aldrich), and mitomycin C (M4287; Sigma-Aldrich) were used for cytotoxicity and viability assays. Calcein-AM (C1430; Invitrogen, Thermo Fisher Scientific, USA) was used for live-cell staining. Doxorubicin (Zodox 50; Intas Pharmaceuticals, India), paclitaxel (Taxol®; Bristol-Myers Squibb BMS, Mumbai, India), and 5-fluorouracil (5-FU; Sanorva Biotech Pvt. Ltd. Mysore, India) were used as standard chemotherapeutic agents. Soluble Fms-like tyrosine kinase-1 (2d-sFlt-1), recombinant VEGF, and anti-VEGF antibodies were expressed and purified in-house, whereas bevacizumab was procured from Biocon (Bangalore, India). Ultra-Low Attachment (ULA) 96-well round-bottom plates (Corning® Costar®, CLS3474; Corning, USA) were used to generate tumor spheroids. Primary antibodies against vimentin (clone V9; sc-6260), E-cadherin (clone SPM471; sc-56527), and focal adhesion kinase (FAK, H-1; sc-1688) were obtained from Santa Cruz Biotechnology (Dallas, TX, USA), while Ki-67 antibody (Clone PRM113) was obtained from PathnSitu Biotechnologies (Pleasanton, CA, USA). All reagents used were of analytical grade. Generation of 3D tumor spheroids 3D tumor spheroids were generated from MDA-MB-231, MDA-MB-468, and MCF-7 breast cancer cell lines using ULA 96-well round-bottom plates. MDA-MB-231 cells were also seeded in 96-well flat-bottom plates pre-coated with 2% agarose (UltraPure™ Agarose, 16500500; Invitrogen). Single-cell suspensions were prepared in a complete growth medium (L-15 supplemented with 10% FBS and 1% penicillin-streptomycin) containing 0.25% methylcellulose to promote spheroid formation. To standardize the spheroid size, MDA-MB-231 cells were seeded at densities of 2000, 4000, 8000, 16,000 or 32,000 cells/well. The culture medium was replenished every three days with minimal disturbance to maintain spheroid architecture. Spheroid formation, morphology, and integrity were monitored daily using an inverted microscope (Zeiss, Germany), and spheroid dimensions were quantified using ZEN 2.5 imaging software (Zeiss, Germany). Based on these optimization experiments, spheroids generated from 2000 cells/well in ULA plates were selected as the standard model for all subsequent assays. Treatment of breast cancer cell lines using 2D monolayer and 3D spheroid models Monolayer (2D) and spheroid (3D) cultures of MDA-MB-231, MDA-MB-468 and MCF-7 cells were treated with chemotherapeutic agents, including paclitaxel (0.1, 1.0, 10, 100, and 300 nM), doxorubicin (0.1, 1.0, 10, 100, and 300 nM), or 5-fluorouracil (5-FU; 0.1, 1.0, 10, 100, and 300 µM), and anti-angiogenic agents (2d-sFlt-1 or bevacizumab; 0.1, 1.0, 10, and 100 ng/well). Combinatorial treatment was performed by co-treating spheroids with paclitaxel (0.1, 1.0, 10, and 100 nM) and either 2d-sFlt-1 or bevacizumab (100 ng/well). Cytotoxicity assay The cytotoxic effects of the drug treatments were assessed after 72 h using the trypan blue dye exclusion method. Monolayer cultures were harvested by trypsinization, and the spheroids were dissociated mechanically, followed by enzymatic digestion with trypsin. Viable cells were counted using a hemocytometer, and cell viability was expressed as a percentage relative to that of the untreated controls. Half-maximal inhibitory concentration (IC50) values were determined from dose-response curves using GraphPad Prism 8.0. All experiments were performed in triplicate (three independent experiments), and the data are represented as mean ± standard deviation (SD). Morphometric analysis of 3D spheroids Spheroid morphology was analyzed using AnaSP software (version 1.2) implemented in MATLAB R2019b (MathWorks, USA) [19]. Acquired images of the control and treated spheroids were subjected to 3D segmentation and virtual sectioning for quantitative assessment of the spheroids. The morphometric parameters extracted for each spheroid includes area, volume, solidity, sphericity, compactness, perimeter, length of the major diameter through the centroid (LMajDTC), and length of the minor diameter through the centroid (LMinDTC). Morphometric parameters were compared between spheroids treated with chemotherapy alone and those treated with a combination of paclitaxel with either 2d-sFlt-1 or bevacizumab. All experiments were performed in triplicate, and results are represented as the mean ± SD. Proliferation assay Spheroid proliferation was assessed on day 7 post-treatment using a resazurin-based cytotoxicity assay. Briefly, the spheroid culture medium was removed, and the wells were washed with sterile Dulbecco’s Phosphate-Buffered Saline (DPBS; no calcium, no magnesium, 1X, Gibco, USA). Resazurin working solution (1X) was added to each well and incubated in the dark for 4 h. Absorbance was recorded at 570 nm using a TECAN multiplate reader (Infinite 200 PRO, Tecan Austria GmbH, Austria) and the values were normalized to those of the untreated controls. Cell viability curves were generated using the GraphPad Prism 8. All experiments were performed in triplicate, and the data are expressed as the mean ± SD. Live/dead cell analysis Cell viability within the tumor spheroids was assessed using triple-fluorescence staining. Briefly, tumor spheroids grown in ULA 96-well plates were stained with a mixture of Hoechst 33342 (33 µM), calcein-AM (2 µM), and EthD-1 (3µM) for 30 min at 37 °C in a 5% CO2 incubator. Following incubation, the spheroids were gently washed with DPBS and imaged using a fluorescence microscope (Zeiss, Germany). Images were acquired under identical exposure settings, and the fluorescence intensity was quantified using ImageJ software (NIH, USA). Wound healing assay To evaluate the combinatorial effect of paclitaxel and 2d-sFlt-1 on MDA-MB-231 cells migration, a scratch assay was performed. MDA-MB-231 cells (4 × 10⁴ cells/well) were seeded in 24-well plates and allowed to reach confluence before treatment. The monolayers were serum-starved for 24 h, and scratches were created using sterile pipette tips. Detached cells were removed by washing with sterile PBS, and mitomycin C (10 ng/mL) was applied for 2 h to inhibit cell proliferation. Cells were then treated with 2d-sFlt-1 (100 ng/mL), paclitaxel (100 nM), or a combination of both in a serum-free medium. Wound closure was imaged at 0, 6, 12, and 24 h post-treatment using an inverted microscope, and migration distances were measured and quantified using ImageJ software. In vitro 3D invadopodia assay Spheroids of MDA-MB-231 cells were generated overnight and subsequently embedded in a complete medium containing the Matrigel matrix (1:1 ratio). The Matrigel-embedded spheroids were incubated at 37 °C under 100% humidity to facilitate interaction with the surrounding matrix. Following incubation, the spheroids were stained with calcein-AM (2µM) and imaged using a laser scanning confocal microscope (LSCM, Zeiss, Germany). Invadopodia formation was quantified as the sprouting area surrounding the spheroid perimeter, using ZEN 3.5 blue edition software (Zeiss, Germany). RNA extraction and quantitative real-time PCR Total RNA was extracted from the treated spheroids using the RNeasy Mini Kit (74104; Qiagen, USA) according to the manufacturer’s protocol. One microgram of RNA was reverse-transcribed into complementary DNA (cDNA) using a Quantitech Reverse Transcription Kit (205311; Qiagen, USA), following the manufacturer’s instructions. The resulting cDNA was diluted and used as a template for quantitative PCR (qPCR) as described previously [20]. Briefly, qPCR was performed using SsoFast Eva Green Supermix (1725200; Bio-Rad) with primers specific for EMT- related genes, including E-cadherin, N-cadherin, TWIST, and SNAIL. GAPDH served as an endogenous control. All reactions were performed in triplicate, and relative gene expression levels were calculated using ΔΔCT method. Data are presented as the mean ± SD from three independent experiments. Primer sequences are provided in the Supplementary Information (Supplementary Table 1). Immunofluorescence of 3D spheroids Spheroids were generated in ULA 96-well plates and processed for immunofluorescence (IF) on day-7 post-treatment as described previously [21]. Briefly, spheroids were fixed with 4% cold paraformaldehyde (PFA; cold) for 4 h and blocked with 3% bovine serum albumin (BSA, A9418; Sigma-Aldrich). Following blocking, spheroids were permeabilized with 0.01% Tween-20 and incubated overnight at 4 °C with primary antibodies against E-cadherin, vimentin, and Ki-67 (1:100 dilution). After washing, the spheroids were incubated with Alexa Fluor 488-conjugated secondary antibody for 2 h in the dark, followed by counterstaining with Hoechst 33342 for 20 min. Images were acquired using LSCM (Zeiss, Germany), and the fluorescence intensity was quantified using ImageJ software. Hematoxylin and eosin staining of 3D spheroids MDA-MB-231 tumor spheroids were generated in agarose-coated 96-well plates. The spheroids were rinsed with PBS and fixed in cold 4% PFA for 4 h at room temperature, followed by washing with PBS. Fixed spheroids were embedded in 2% ultra-pure agarose (Invitrogen, USA) and excess agarose was trimmed using a scalpel. The agarose blocks containing spheroids were dehydrated through a graded ethanol series (50%, 70%, 90%, 95%, and 100%) and embedded in paraffin wax. Paraffin blocks were sectioned at a thickness of 2 μm using a Leica microtome (Leica Biosystems, Germany). Sections from control and treated spheroids were mounted on SuperFrost®Plus slides (Menzel-Gläser, Germany), dried overnight, and baked at 60 °C for 1 h to ensure firm adherence. Slides were dewaxed in three changes of xylene (3 × 5 min each) and rehydrated through graded methanol washes (2 × 1 min). Standard hematoxylin and eosin (H&E) staining was performed using hematoxylin (3 min) and eosin (3 min). Immunohistochemistry of 3D tumor spheroids Tissues from the prepared paraffin blocks were sectioned, dewaxed, and rehydrated, followed by antigen retrieval in citrate buffer (10mM, pH. 6.0 at 121 °C for 3 min). Sections were then blocked with 3% BSA and incubated overnight at 4 °C with primary antibodies against E-cadherin, vimentin, and Ki-67 (1:100 dilution). The following day, these sections were incubated with horseradish peroxidase (HRP)-conjugated secondary antibody (1:2000 dilution) for 30 min at room temperature. Sections were washed 3 times for 5 min each with PBS containing 0.1% Tween-20 (PBST) after primary and secondary antibody incubations. Signal detection was performed using diaminobenzidine (DAB, PS013; PathnSitu Biotechnologies, USA) as the chromogen, followed by eosin counterstaining. Slides were rehydrated through graded ethanol, cleared in xylene, and mounted with dibutyl phthalate in xylene (DPX). The immunohistochemical slides were visualized under a phase-contrast microscope (Zeiss, Germany). Biological treatment of endothelial cell spheroids HUVECs were detached from monolayer cultures using EnVzyme™ Easy Endothelial Cell Dissociation Solution and resuspended in endothelial basal medium. Cells were seeded into ULA 96-well plates at a density of 2,000 cells/well to generate the spheroids. Spheroids were treated with 2d-sFlt-1 (20 ng/well), either alone or in combination with recombinant VEGF (20 ng/well), and cultured for 7 days. Spheroids were imaged on days 1, 3, and 7 post-treatments using an inverted microscope (Zeiss, Germany), and spheroid size was quantified using ZEN 2.5 imaging software. Co-culture of endothelial and tumor cell spheroids MDA-MB-231 tumor cells and HUVECs were detached from monolayer cultures and resuspended in a 1:1 ratio (1,000 cells each) in a mixed medium consisting of HiEndoXL Endothelial Cell Expansion Medium and DMEM (1:1), supplemented with 5% FBS and 0.25% methylcellulose. Endothelial-only, tumor-only, and tumor-endothelial co-culture spheroids were generated under these conditions. Spheroids were treated with or without 2d-sFlt-1 (50 ng/well) and subsequently processed for immunofluorescence staining against anti-FAK antibody on day 3 post-treatment, following the protocol optimized for 3D spheroids. Alexa Fluor 488–labeled spheroids were imaged using LSCM (Zeiss, Germany), and fluorescence intensity was quantified using ImageJ software. Tube formation assay Tube formation assay was performed as described previously [22]. Briefly, 50µL of Matrigel matrix was added to each well of a 96-well flat-bottom plate and allowed to polymerize at 37 °C, according to the manufacturer’s instructions. HUVECs were harvested, resuspended in endothelial cell basal medium, and seeded at a density of 1 × 104 cells/well. Recombinant VEGF (rVEGF; 20 ng/well) was added to induce capillary-like structure formation, and culture plates were incubated for 12 h at 37 °C. Following initial tube formation, HUVECs were treated with either 2d-sFlt-1 alone or 2d-sFlt-1 in combination with rVEGF and incubated for an additional 12 h under standard culture conditions (37 °C and 5% CO₂). Wells were gently washed with sterile PBS to remove debris and non-adherent cells, then stained with calcein-AM (2µM) for 30 min at 37 °C. Fluorescence images were captured at 10x magnification using a fluorescence microscope. Tube formation was quantified using ImageJ software by measuring the total tube length, number of tubes, nodes, and branch points. Data are expressed as mean ± SD. Statistical comparisons among treatment groups were performed using two-way ANOVA followed by Bonferroni post-hoc analysis using GraphPad Prism 8.0. Enzyme-linked immunosorbent assay (ELISA) for VEGF expression Conditioned medium from paclitaxel- and 2d-sFlt-1–treated spheroids was collected to measure VEGF levels using direct ELISA, as described previously [21]. Briefly, serial dilutions of recombinant VEGF (0.001- 5 ng) were immobilized on an ELISA plate as a standard. Likewise, 100 µl of conditioned medium from treated and untreated spheroids was added to the respective wells. Plates were incubated overnight at 4 °C, washed with PBS containing 0.05% Tween-20 (PBS-T) and blocked with 5% skimmed milk for 1 h at room temperature. The blocked wells were then incubated with rabbit polyclonal anti-VEGF primary antibody (in-house, 1:500) and subsequently with goat anti-rabbit alkaline phosphatase (ALP)-conjugated secondary antibody (AP132A; Sigma Aldrich, 1:1000 dilution). After washing, p-nitrophenyl phosphate (pNPP, N7653; Sigma-Aldrich) substrate was added and the absorbance was measured at 405 nm using a microplate reader (Tecan, Austria). Data were analysed using GraphPad Prism 8.0. Statistical analysis All experiments were performed in triplicate and independently repeated at least thrice. Data are presented as mean ± SD. Normality of the data was assessed using the Shapiro–Wilk test. Group comparisons were performed using one-way or two-way ANOVA, followed by Tukey’s test, Dunnett’s, or Bonferroni’s multiple comparison tests, as appropriate. IC₅₀ values were calculated from non-linear regression analysis, and qPCR data were analyzed using ΔΔCt method with GAPDH as the endogenous control. All statistical analyses were conducted using GraphPad Prism 8.0, with statistical significance defined as p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), or p < 0.0001 (****). Results Formation of 3D-tumor spheroids In vitro 3D tumor spheroids were efficiently established within 24 h after seeding varying cell numbers into ULA 96-well round-bottom plates and 2% agarose-coated flat-bottom 96-well plates. The size of the spheroids correlated directly with the number of cells seeded (Fig. 1A-C). Compared with ULA plates, spheroids cultured in agarose-coated plates were smaller, particularly at higher seeding densities, showing size reduction of 10.62% and 11.39% at 16,000 and 32,000 cells/well, respectively. Spheroids formed in the ULA plates developed well-defined structural zones (Fig. 1D), including a necrotic core that became increasingly prominent with spheroid size enlargement, whereas spheroids formed in 2% agarose did not exhibit necrosis. Monitoring growth kinetics of MDA-MB-231 spheroids over 35-days demonstrated progressive size expansion (Fig. 1E, F). Viability was assessed by Hoechst 33342, calcein-AM, and EthD-1 staining, confirming that higher seeding densities yielded spheroids with larger necrotic regions, as reflected by increased EthD-1 positivity. Specifically, spheroids seeded with 4,000 cells/well exhibited 20.34 ± 0.9% dead cells, compared with only 3.61 ± 0.53% at 2,000 cells/well (P < 0.001). Dead cells were predominantly localized to the spheroid core, consistent with limited oxygen diffusion and nutrient availability at higher cellular densities (Fig. 2A, B). Based on these observations, spheroids formed from 2,000 cells/well were selected as the optimized model for subsequent experiments to balance physiological relevance with minimal necrotic burden. Influence of MDA-MB-231 cell seeding density on spheroid size and necrosis. MDA-MB-231 cells harvested from the 2D culture were seeded at varying densities on either ULA 96-well plates or agarose-coated 96-well plates. (A) Representative images of spheroid growth on ULA plates. (B) Representative images of spheroid growth in 2% agarose-coated wells (50 µL/well). (C) Histogram of spheroid diameters (µm) at different seeding densities, data analyzed by one-way ANOVA with Tukey’s post hoc test. (D) Structural analysis of spheroids seeded at 16,000 cells/well, highlighting necrotic core, quiescent zone, and proliferative periphery. (E) Longitudinal monitoring of spheroid growth over 35 days on ULA plates. (F) Growth curve showing spheroid volume(µm³) at 2,000 cells/well over 35 days, analyzed by repeated-measures one-way ANOVA. Data are represented as mean ± SD of triplicate experiments. Scale bar indicates 5x magnification Cytotoxic effects of anti-cancer and anti-angiogenic agents in 2D versus 3D tumor models. (A) Live/Dead fluorescence imaging of MDA-MB-231 spheroids seeded at 2,000 or 4,000 cells/well using FITC, TRITC, and DAPI filters. (B) Quantitative analysis of fluorescence intensities (ImageJ) showing increased cell death with higher seeding density; data were analyzed using one-way ANOVA with Tukey’s post hoc-test. Spheroids seeded at 2,000 cells/well were used for subsequent assays. Cytotoxicity assays comparing 2D vs. 3D cultures treated with (C) paclitaxel at 0.1–300 nM, (D) 5-FU (0.1–300 µM), or (E) doxorubicin (0.1–300 µM). (F) Cytotoxicity of 2d-sFlt-1 vs. bevacizumab in 2D cultures. (G) Cytotoxicity of 2d-sFlt-1 and bevacizumab in 3D spheroid culture. (H) Cytotoxicity of paclitaxel alone or in combination with either 2d-sFlt-1 or bevacizumab (0.1 µg/well). IC50 values were calculated using GraphPad Prism 8. All experiments were performed in triplicate, and the data are shown as the mean ± SD. Scale bars are indicated Cytotoxicity of anti-cancer and anti-angiogenic agents in 2D versus 3D culture models The cytotoxic effects of standard chemotherapeutic agents- paclitaxel, 5-FU, and doxorubicin, were evaluated in MDA-MB-231 grown as 2D monolayer and 3D spheroids. Drug sensitivity was significantly attenuated in 3D spheroids with IC50 values increasing from 4.06 nM, 1.98 µM, and 11.58 µM in 2D cultures to 101.90 nM, 56.10 µM, and 69.37 µM in 3D cultures, respectively. These shifts represented approximately 25-fold, 28-fold, and 7-fold reductions in drug response in 3D spheroids compared to 2D monolayers (Fig. 2C-E; Supplementary Table 2). We next compared the cytotoxic activity of the anti-angiogenic agents, 2d-sFlt-1 and bevacizumab in both 2D and 3D culture models. Consistent with chemotherapy results, both agents exhibited reduced activity in 3D spheroids, with IC50 values increasing by 8.3-fold for 2d-sFlt-1 and 6-fold for bevacizumab compared to 2D cultures. Notably, 2d-sFlt-1 demonstrated superior potency compared to bevacizumab, exhibiting 3.5-fold more cytotoxicity in 2D cultures and 2.5-fold higher cytotoxicity in 3D spheroids (Fig. 2F, G). These findings were validated in MDA-MB-468 spheroids, in which paclitaxel IC50 values increased from 8.51 nM in 2D to 129.90 nM in 3D cultures (Supplementary Fig. 1A). Similarly, both 2d-sFlt-1 and bevacizumab exhibited significantly higher IC50 values in 3D compared to 2D cultures (2d-sFlt-1: 20.4 vs. 161.8 ng; bevacizumab: 43.8 vs. 206.3 ng, respectively; Supplementary Fig. 1B, C). Collectively, these data confirm that 3D spheroid models confer substantial chemoresistance to conventional 2D systems and underscore the enhanced efficacy of 2d-sFlt-1 over bevacizumab across both culture systems. 2d-sFlt-1 enhances paclitaxel cytotoxicity in 3D tumor spheroids Combining paclitaxel with 2d-sFlt-1 markedly increased cytotoxicity in MDA-MB-231 spheroids, reducing the IC50 value from 101.9 nM to 10.1 nM, representing an approximately 10-fold enhancement (Fig. 2H). This synergistic effect was validated in MDA-MB-468 spheroids, where the paclitaxel IC50 decreased from 129.9 nM to 20.8 nM when combined with 2d-sFlt-1, corresponding to an approximately 6-fold enhancement. In contrast, the combination of paclitaxel with bevacizumab did not show a significant improvement, with IC50 values shifting only from 129.9 nM to 121.9 nM (Supplementary Fig. 1D). These data indicate that 2d-sFlt-1 substantially enhances paclitaxel sensitivity in TNBC spheroids, suggesting its potency to overcome intrinsic chemoresistance mechanisms in physiologically relevant 3D tumor models and supporting its potential as a promising adjuvant to conventional chemotherapy. 2d-sFlt-1 surpasses bevacizumab in suppressing MDA-MB-231 spheroid growth 2d-sFlt-1 demonstrated enhanced anti-tumor activity compared to bevacizumab in MDA-MB-231 spheroids. By days 7–10, 2d-sFlt-1–treated spheroids exhibited marked morphological disruption and loss of cell–cell integrity compared to bevacizumab (Fig. 3A, B). Quantitative analysis showed a 62% reduction in spheroid diameter with 2d-sFlt-1 compared to 41% reduction with bevacizumab (p < 0.01, two-way ANOVA; Fig. 3C). Relative to standard chemotherapies, 2d-sFlt-1 achieved growth inhibition comparable to high-dose 5-FU, paclitaxel, and doxorubicin, that reduced spheroid size by 60–68% (ns, one-way ANOVA; Fig. 3D, E; Supplementary Table 3). Morphometric analysis further confirmed a significant reduction in spheroid area, volume, compactness, solidity, sphericity and perimeter following treatment with cytotoxic agents (5-FU and doxorubicin) and 2d-sFlt-1 (p < 0.05–0.001, Tukey’s post-hoc test; Supplementary Tables 4 and 5; Supplementary Figs. 2, 3, and 4). Collectively, these data demonstrate 2d-sFlt-1 as a more potent inhibitor of spheroid growth than bevacizumab and exhibiting an efficacy comparable to conventional chemotherapeutics in a 3D tumor model. Growth inhibition and morphological changes in 3D spheroids treated with 2d-sFlt-1, bevacizumab, or chemotherapeutic drugs. (A) Representative images of MDA-MB-231 spheroids treated with 2d-sFlt-1 compared to those treated with bevacizumab. (B) Alterations in the proliferative zone of spheroids post 2d-sFlt-1 treatment relative to those post bevacizumab treatment. (C) Quantitative measurement of spheroid size (ImageJ) in response to 2d-sFlt-1 and bevacizumab. (D) Representative images of spheroids treated with 5-FU, paclitaxel, and doxorubicin. (E) Quantitative analysis of spheroid volume following treatment with 5-FU (100, 300 µM), paclitaxel (0.1, 0.3 µM), or doxorubicin (0.1, 0.3 µM). Graphs were generated using GraphPad Prism (version 8). Significance was determined by one-way or two-way ANOVA (as appropriate) with Tukey’s post-hoc test (****p < 0.0001, **p < 0.01, and “ns” = not significant). Data are presented as mean ± SD. Scale bar, 10x magnification 2d-sFlt-1 enhances the anti-proliferative and cytotoxic effects of paclitaxel Combinatorial treatment with 2d-sFlt-1 and paclitaxel significantly inhibited the growth of MDA-MB-231, MDA-MB-468, and to a lesser extent, MCF-7 spheroids compared with paclitaxel alone. Representative images and morphometric segmentation of MDA-MB-231 spheroids using the AnaSP/ReViSP confirmed this synergistic response, with the combination achieving a 75% reduction in spheroid volume (p < 0.001) compared to 43% and 34% reductions with paclitaxel and 2d-sFlt-1 monotherapy, respectively (Fig. 4A-C). This was accompanied by a significant decrease in the spheroid compactness and area in the combination group (p < 0.001; Fig. 4D, E). In MCF-7 spheroids, 2d-sFlt-1 alone had no measurable effect, whereas paclitaxel induced characteristic central necrosis and accumulation of cellular debris (Fig. 4F). Consistent with the MDA-MB-231 response, combination treatment did not produce marked morphological differences on day 3 compared to paclitaxel alone; however, by day 7, the dual treatment showed a modest reduction in spheroid volume and increased debris accumulation, indicating a delayed yet incremental cytotoxic effect (Fig. 4G). In MDA-MB-468 spheroids, the combination of paclitaxel and 2d-sFlt-1 significantly reduced spheroid volume on days 3 and 7 (p < 0.001, two-way ANOVA; Fig. 5A, B). Live-dead staining confirmed this synergy: paclitaxel plus 2d-sFlt-1 increased cell death to 62% compared to 35% with paclitaxel alone (p < 0.001; Fig. 5C, D). In MDA-MB-231 spheroids, 2d-sFlt-1 also induced greater cytotoxicity than bevacizumab (63% vs. 47%; Fig. 6A, B). More importantly, the combination of paclitaxel and 2d-sFlt-1 resulted in 72% dead cells, markedly higher than the 32% observed with paclitaxel plus bevacizumab (p < 0.001; Fig. 6C, D). Collectively, these findings demonstrate that 2d-sFlt-1 not only surpasses bevacizumab in enhancing paclitaxel-mediated cytotoxicity but also broadens efficacy across breast cancer subtypes, supporting its promise as a potent combinatorial strategy for overcoming chemoresistance in physiologically relevant 3D tumor models. Effect of paclitaxel and 2d-sFlt-1 combinatorial treatment on 3D tumor spheroid growth and morphology. (A) Representative images of MDA-MB-231 spheroids on days 3 and 7 post-treatment with paclitaxel ± 2d-sFlt-1. (B) Binary masks, segmentation, and 3D projections of spheroids generated using the AnaSP/ReViSP software for morphological analysis. Quantitative measurements of (C) spheroid volume, (D) compactness, and (E) area were performed; data were analyzed using one-way ANOVA with Tukey’s post-hoc test. (F) Representative images of MCF-7 spheroids treated with paclitaxel ± 2d-sFlt-1 on days 3 and 7. (G) Graphical representation of MCF-7 spheroid size changes (ImageJ measurements) plotted using GraphPad Prism 8.0 software. A significant reduction in spheroid size was observed at both time points compared to control (two-way ANOVA, p < 0.001). Data presented as mean ± SD Combinatorial and synergistic treatments suppress the growth and viability of MDA-MB-468 tumor spheroids. MDA-MB-468 breast cancer cells were seeded in ULA 96-well plates (2,000 cells/well). (A) Representative bright-field images of spheroids on days 1, 5, and 7 post-treatments with 2d-sFlt-1, bevacizumab, paclitaxel, or combinations of paclitaxel + bevacizumab and paclitaxel + 2d-sFlt-1. Scale bar = 200 μm. (B) Quantification of spheroid volume changes measured by ImageJ and plotted using GraphPad Prism 10. Data are shown as mean ± SD (n = 3). Two-way ANOVA revealed significant treatment effects (p < 0.001). At day 7, paclitaxel + 2d-sFlt-1 (p < 0.001) significantly reduced spheroid volume compared to paclitaxel alone (p < 0.01) and with paclitaxel + bevacizumab (p < 0.05), indicating a stronger synergistic effect. Representative live/dead analysis of MDA-MB-468 spheroids at day 10. (C) MDA-MB-468 spheroid images post treatment with vehicle (PBS), 2d-sFlt-1 (0.1 µg), paclitaxel (0.1 µM), or paclitaxel + 2d-sFlt-1. Viability was assessed with calcein-AM (green, live), EthD-1 (red, dead), and Hoechst 33342 (blue, nuclei) and imaged with FITC, TRITC, and DAPI channels, respectively. Scale bar: 100 μm (10x). (D) Quantification of calcein-AM (live) and EthD-1 (dead) fluorescence intensities normalized to control (%). Bars represent the mean ± SD (n = 3). One-way ANOVA with Tukey’s post-hoc test showed significant reductions in viability for 2d-sFlt-1 (p < 0.01), paclitaxel (p < 0.001), and paclitaxel + 2d-sFlt-1 (p < 0.0001) compared to control. Importantly, the combination of paclitaxel + 2d-sFlt-1 (****) further reduced viability compared to paclitaxel alone (**), demonstrating synergistic sensitization. Asterisks denote: p < 0.05 (*), p < 0.01 (**), p < 0.001 (***), p < 0.0001 (****) Combinatorial effects of 2d-sFlt-1 and paclitaxel on the viability of MDA-MB-231 tumor spheroids. (A) Representative live/dead staining of MDA-MB-231 spheroids treated with vehicle (PBS), 2d-sFlt-1 (0.05–0.1 µg), or bevacizumab (0.05–0.1 µg) on day 10. calcein-AM (green; live), EthD-1 (red; dead), and Hoechst 33342 (blue; nuclei) were imaged using FITC, TRITC and DAPI filters, respectively. Scale bar = 100 μm (10x). (B) Quantitive analysis of live and dead cell fractions based on green and red fluorescence intensities (ImageJ). (C) Treatment regimens with paclitaxel (0.01 or 0.03 µM) alone or combined with 2d-sFlt-1 (0.1 µg) or bevacizumab (0.1 µg). (D) Histogram of fluorescence intensity measurements (ImageJ) for each treatment group. Data were analysed using one-way ANOVA with Tukey’s post hoc test. Results are shown as mean ± SD (n = 3) Combinatorial therapy inhibits tumor cell migration The combination of 2d-sFlt-1 and paclitaxel significantly inhibited MDA-MB-231 cell migration in both 2D and 3D models. In the wound healing assay, vehicle-treated cells demonstrated nearly complete gap closure within 24 h, whereas paclitaxel alone permitted 80% closure. In contrast, the combinatorial treatment markedly suppressed wound closure (p < 0.001), indicating a significant inhibition of migratory activity (Fig. 7A, B). In 3D Matrigel invasion assay, untreated MDA-MB-231 spheroids exhibited extensive invadopodia formation with a mean invasion distance of 309.23 ± 24.08 μm at 48 h post embedding in Matrigel. Combination therapy reduced invasion to 36.30 ± 6.55 μm (p < 0.001), representing a profound inhibition of invasive behavior and outperforming monotherapy (Fig. 7A). These results demonstrate that dual inhibition with 2d-sFlt-1 and paclitaxel disrupts both 2D motility and 3D invasion, supporting its potential to suppress metastatic dissemination in MDA-MB-231 tumors. In Matrigel migration assays MDA-MB-231 spheroids exhibited pronounced invasion (approximately 350 μm at 72 h), whereas MCF-7 spheroids showed no measurable invasion under identical conditions (Supplementary Fig. 5). These results further validated tumor subtype–specific differences in their invasive potential. Combinatorial treatment inhibits migration/invadopodia formation and modulates EMT gene expression in MDA-MB-231 spheroid cells. (A) Bright-field images from wound-healing assays at 0, 6, 12, and 24 h post-treatment, and confocal images of Matrigel-embedded 3D spheroids stained with calcein-AM at 72 h, showing reduced migration with paclitaxel + 2d-sFlt-1 compared to other groups. Scale bar: 10× magnification. (B) Quantification of wound distance and closure analyzed using ImageJ; data analyzed by repeated-measures two-way ANOVA (time x treatment) with Tukey’s post-hoc test. qPCR analysis of EMT markers showing upregulated E-cadherin expression (C) and downregulated N-cadherin (D), SNAIL (E), and TWIST (F) expression following combination treatment. Gene expression data were analyzed using one-way ANOVA with Tukey’s post-hoc test. All experiments were performed in triplicate; and the results are shown as mean ± SD. Combined treatment shifted EMT marker expression toward a less metastatic phenotype Combinatorial therapy reversed EMT in MDA-MB-231 spheroids EMT is a critical driver of TNBC invasion and metastasis. We evaluated the impact of combinatorial therapy on EMT by quantifying the expression of known EMT markers (E-cadherin, N-cadherin, SNAIL, and TWIST) in treated and untreated spheroids via qPCR. The combination of 2d-sFlt-1 and paclitaxel had a more pronounced effect on EMT modulation than paclitaxel or 2d-sFlt-1 alone. E-cadherin expression increased by approximately 2-folds compared with that in the untreated controls (p < 0.01; Fig. 7C), whereas N-cadherin expression was reduced by 4-folds (p < 0.001; Fig. 7D). The expression of the EMT transcription factors SNAIL and TWIST was also significantly downregulated by 1.5-fold and > 2-folds, respectively (p < 0.05–0.01; Fig. 7E and F). Collectively, these results show that 2d-sFlt-1 and paclitaxel combinatorial therapy not only suppresses mesenchymal gene expression but also restores epithelial identity, effectively shifting MDA-MB-231 spheroids towards a more epithelial phenotype. This provides a mechanistic basis for the observed anti-metastatic efficacy. Combinatorial therapy downregulates EMT markers and proliferation in MDA-MB-231 spheroids To validate the transcriptional EMT changes at the protein level, MDA-MB-231 tumor spheroids were analyzed by immunofluorescence (IF) and complementary assays. Untreated spheroids exhibited weak E-cadherin signals and strong vimentin staining, consistent with the mesenchymal phenotype (Fig. 8A, B). Treatment with either 2d-sFlt-1 or paclitaxel partially restored epithelial features, whereas their combination significantly enhanced E-cadherin expression and markedly reduced vimentin levels, indicating effective mesenchymal-to-epithelial reversion (Fig. 8C, D). Histopathological assessment further corroborated these findings. H&E staining revealed reduced nuclear density in spheroids treated with the combination compared to control or either paclitaxel or 2d-sFlt-1 alone (Fig. 8E). Immunohistochemistry (IHC) confirmed the downregulation of vimentin (Fig. 8F), whereas Ki-67 staining revealed a marked decrease in proliferating cells after combinatorial treatment (Fig. 8G). Quantitative resazurin-based proliferation assay showed that spheroids treated with 2d-sFlt-1 significantly reduced mitochondrial metabolic activity more effectively than bevacizumab (Fig. 8H). Interestingly, combinatorial treatment of paclitaxel and 2d-sFlt-1 reduced proliferation more efficiently than paclitaxel in combination with bevacizumab (Fig. 8I). Collectively, these results demonstrate that the combination of 2d-sFlt-1 and paclitaxel effectively downregulated EMT markers, inhibited cell proliferation, and suppressed the malignancy of MDA-MB-231 spheroids. Immunofluorescence and immunohistochemistry confirm the inhibition of EMT and proliferation markers post combinatorial therapy. Immunofluorescence staining of spheroids with antibodies against (A) E-cadherin (epithelial marker) and (B) Vimentin (mesenchymal marker). Quantitative analysis shows (C) increased E-cadherin and (D) reduced Vimentin signals in the combination group compared to single or vehicle treatments (one-way ANOVA, Tukey’s post-hoc test). (E) H&E staining of spheroids highlights morphological differences between combinatorial and single-agent treatments. (F) Immunohistochemistry (IHC) for vimentin demonstrated suppression of the metastatic phenotype in combinatorial-treated spheroids relative to controls. (G) IHC staining for Ki-67 shows reduced proliferative capacity following combination therapy. (H) Proliferation assay of spheroids treated with 2d-sFlt-1 and bevacizumab. (I) Quantification of proliferation by resazurin assay, plotted using GraphPad Prism 8, confirms the impact of combinatorial therapy (two-way ANOVA with multiple comparisons). Fluorescence images were captured using a confocal laser scanning microscope, and IHC slides imaged using bright-field microscopy. Scale bars = 100 μm. Data are presented as mean ± SD (n = 3). Statistical significance: ****P < 0.0001; **P < 0.01; ns = not significant 2d-sFlt-1 suppressed FAK expression and angiogenesis in 3D co-cultured spheroids FAK is a central regulator of tumor angiogenesis and metastasis. To determine the effect of 2d-sFlt-1 on FAK expression, we assessed FAK levels in MDA-MB-231 spheroids and tumor-endothelial co-cultured spheroids using immunofluorescence staining. FAK expression was elevated in MDA-MB-231 spheroids, and further increased in MDA-MB-231-HUVECs co-cultured spheroids (Fig. 9A). Treatment with 2d-sFlt-1 significantly suppressed FAK expression in co-cultured spheroids compared to that in untreated controls (p < 0.01, Student’s t-test), indicating potent inhibition of FAK signaling within the tumor-endothelial microenvironment. We further performed endothelial tube formation assay to examine the effect of 2d-sFlt-1 on formation of blood vessels. Endothelial cells cultured on Matrigel formed intense capillary-like networks in VEGF-treated wells. Exposure to 2d-sFlt-1 (100 ng) significantly disrupted VEGF-induced tube formation, resulting in the disruption of capillary-like networks (Fig. 9B). Quantitative ImageJ analysis revealed an 89% reduction in total tube length and an 87% reduction in branch points compared to VEGF controls (p < 0.001, one-way ANOVA with Tukey’s post-hoc test; Fig. 9C, D). Consistent with these findings, endothelial cell (EC) spheroid assays demonstrated that 2d-sFlt-1 significantly suppressed EC spheroid growth by 81% in the absence of VEGF (p < 0.001) and by 65% under VEGF stimulation on day 7 post-treatment (p < 0.01; Fig. 9E, F). To further assess the anti-angiogenic efficacy of the combinatorial therapy, VEGF secretion was quantified in conditioned medium from MDA-MB-231 spheroids by ELISA. The vehicle-treated spheroids secreted 3.6 ± 0.21 ng of VEGF. Paclitaxel alone did not reduce VEGF secretion (3.8 ± 0.27 ng; ns vs. control), whereas 2d-sFlt-1 monotherapy significantly lowered VEGF levels to 1.8 ± 0.18 ng (49% of control, p < 0.01). Notably, the combination of 2d-sFlt-1 and paclitaxel further suppressed VEGF release to 1.1 ± 0.12 ng, representing a 69% reduction relative to vehicle controls (p < 0.001 vs. vehicle and paclitaxel monotherapy; Fig. 9G). Collectively, these results demonstrate that 2d-sFlt-1 downregulates FAK expression and disrupts VEGF-induced angiogenesis, and when combined with paclitaxel, synergistically suppresses VEGF secretion. These findings highlight the potential of combinatorial therapy to target tumor–endothelial interactions and angiogenic signaling pathways in TNBC tumors. 2d-sFlt-1 reduces FAK expression in 3D co-cultured spheroids and disrupts VEGF-mediated endothelial tube formation. (A) Immunofluorescence staining of MDA-MB-231, HUVEC, and co-cultured spheroids with anti-FAK antibody (green, Alexa Fluor 488) and Hoechst 33342 (blue, nuclear staining). Untreated co-cultured spheroids displayed elevated FAK fluorescence, whereas 2d-sFlt-1–treated spheroids (100 ng) showed a reduced signal. Scale bar, 10 μm. (B) Tube formation assay of HUVECs on Matrigel under VEGF stimulation (0.02 µg), showing progressive disruption of tube networks with 2d-sFlt-1 treatment (0.02–0.1 µg). (C, D) Quantification of tube branches and total tube length using ImageJ software. (E) Representative images of HUVEC spheroids showing reduced growth with 2d-sFlt-1treatment. (F) Quantitative analysis of endothelial spheroid volume at days 3 and 7 demonstrated a sustained anti-proliferative effect (one-way ANOVA, Tukey’s post-hoc test, ****P < 0.0001). (G) Comparative effectiveness of 2d-sFlt-1 versus bevacizumab in enhancing paclitaxel (PAC) efficacy on MDA-MB-231 spheroids. The treatments included PAC, bevacizumab (Beva), 2d-sFlt-1, or combinations (PAC + Beva and PAC + 2d-sFlt-1). The controls included spheroids cultured without (CTR–) or with (CTR+) FBS. VEGF concentrations in the conditioned medium were quantified using ELISA with a standard curve generated from serial dilutions of recombinant VEGF. All experiments were performed in triplicate and the data shown as the mean ± SD (n = 3). Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test Discussion 3D spheroid models provide a physiologically relevant platform for screening anti-cancer and anti-angiogenic agents, as they better reproduce tumor architecture and microenvironmental gradients than conventional 2D monolayers [14, 23, 24]. Although 2D cultures remain simple and widely used, they fail to capture critical tumor–stroma interactions that influence drug sensitivity [25]. In contrast, 3D spheroids incorporate cell–cell and cell–extracellular matrix (ECM) interactions, which contribute to drug diffusion and resistance [15, 16, 26]. In our study, tumor cells grown as spheroids exhibited markedly higher chemoresistance than monolayers, with paclitaxel, 5-FU, and doxorubicin showing 25-, 28-, and 7-fold increases in IC50 values, respectively. The standardization of spheroid size remains essential for reproducible assays, as diffusion limits and hypoxic core formation vary with diameter [24, 26]. Live and dead staining with calcein-AM and EthD-1 further confirmed drug-induced cytotoxicity in these models. Consistent with our findings, several studies reported that cytotoxic drugs are less effective in 3D tumor models than in conventional 2D cultures. For example, paclitaxel exhibited reduced activity in breast, ovarian, and lung cancer spheroids, with IC50 values increasing up to 2–25-fold [27–29]. Similarly, 5-FU showed strong resistance phenotype in colorectal tumor spheroids, with reports of up to 60-fold higher IC50 values in 3D models [30, 31]. Doxorubicin also showed reduced efficacy in breast and colorectal cancer spheroids, underscoring the limited drug penetration and adaptive survival mechanisms that are unique to 3D systems [32–34]. Collectively, these studies highlighted the robustness of 3D spheroids as physiologically relevant platforms for modeling chemoresistance across cancer types. Using 3D models, this study demonstrated strong synergy between 2d-sFlt-1 and paclitaxel in triple-negative (MDA-MB-231 and MDA-MB-468) breast cancers. Consistent with prior reports of chemoresistance in 3D cultures [29, 35], we observed that the addition of 2d-sFlt-1 markedly reduced this resistance, lowering paclitaxel IC50 values by up to 18-fold and underscoring its potential as a potent adjuvant. 2d-sFlt-1 treatment significantly suppressed tumor growth and metastatic potential, consistent with previous studies reporting its ability to inhibit endothelial cell migration and sprouting [36]. Across multiple complementary assays, including viability, proliferation, live/dead staining, migration/invasion, and EMT markers analyses, 2d-sFlt-1 synergistically enhanced the cytotoxic effects of paclitaxel. In MDA-MB-231 spheroids, the combination therapy reversed EMT as evidenced by downregulating vimentin and N-cadherin and upregulating E-cadherin, consistent with previous reports of identifying EMT as a therapeutic target in TNBC [17, 37]. Mechanistically, we demonstrated that 2d-sFlt-1 impedes tumor–endothelial crosstalk, suppresses FAK expression, and disrupts angiogenic signaling pathways. This extends prior work on sFlt-1’s anti-angiogenic activity [22, 38, 39] and is consistent with reports linking elevated FAK expression to tumor survival and angiogenesis [40–42]. Unlike earlier studies that mainly focused on the role of FAK in viability, our findings emphasize that 2d-sFlt-1 downregulates FAK and interferes with pro-angiogenic cues in a 3D context, reinforcing its potential as a therapeutic partner for paclitaxel. The potent anti-angiogenic activity of 2d-sFlt-1 is consistent with prior evidence that soluble VEGF receptors act as effective traps for VEGF, placental growth factor (PlGF), and other pro-angiogenic ligands [43]. Our finding of reduced VEGF-induced tube formation and diminished VEGF secretion are in agreement with these reports, confirming the ability of 2d-sFlt-1 to disrupt vascular development. Our data also highlights the limitations of bevacizumab, which showed weaker effects on viability and metastatic potential, consistent with clinical trials questioning its benefits in breast cancer [44]. Although bevacizumab also neutralizes VEGF, its narrower specificity likely explains its reduced efficacy in our model. By intercepting a broader spectrum of pro-angiogenic factors, 2d-sFlt-1 offers a mechanistic advantage that aligns with recent advances in multi-targeted anti-angiogenic therapies. In summary, this study demonstrates that 3D spheroid models closely recapitulate key features of the tumor microenvironment and chemoresistance observed in vivo. The combination of 2d-sFlt-1 and paclitaxel enhanced therapeutic efficacy across triple-negative and luminal breast cancer spheroids by reducing chemoresistance, suppressing proliferation, and promoting cytotoxicity. Mechanistic investigations in MDA-MB-231 spheroids revealed that 2d-sFlt-1 potentiates paclitaxel efficacy through inhibition of angiogenic signaling, reversal of epithelial–mesenchymal transition, and downregulation of FAK expression. In MDA-MB-468 and MCF-7 spheroids, the combination primarily enhanced cytotoxicity and growth inhibition, indicating a conserved functional benefit across subtypes. Compared with bevacizumab, 2d-sFlt-1 exhibited superior anti-tumor activity, supporting its potential as a mechanistically distinct adjuvant to taxane-based chemotherapy in aggressive breast cancers. Conclusion Our findings demonstrated 2d-sFlt-1 as a potent enhancer of paclitaxel efficacy in physiologically relevant 3D breast cancer models. In MDA-MB-231 spheroids, combination therapy reduced chemoresistance and invasive growth through coordinated inhibition of angiogenesis, reversal of EMT, and downregulation of FAK signaling. In MDA-MB-468 and MCF-7 spheroids, the combination primarily increased cytotoxicity and growth inhibition, consistent with a subtype-independent functional synergy. These findings establish a solid rationale for further mechanistic validation in additional TNBC models and for preclinical evaluation of 2d-sFlt-1 plus paclitaxel in advanced ex vivo and in vivo systems to define dosing parameters, safety profiles, and translational feasibility. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements The authors acknowledge Institute of Excellence at the University of Mysore for providing access to the central core facility and confocal microscopy. AM acknowledge Abs Community College, Yemen, for providing PhD scholarship support. Abbreviations Analysis of variance Three Dimensional Two Dimensional Triple Negative Breast Cancer Human Umbilical Vein Endothelial Cells Two domain soluble Fms-Like Tyrosine Kinase − 1 Epithelial to mesenchymal transition Focal adhesion kinase Estrogen Receptor Progesterone Receptor Human epidermal growth factor receptor 2 Vascular Endothelial Growth Factor Placental Growth Factor Endothelial Cell Quantitative Polymerase Chain Reaction Enzyme-Linked Immunosorbent Assay Authors contribution Mutahar AZI., conducted critical experiments, analysed the data, wrote and finalized the manuscript, and discussed the results. Salimath BP., raised the research questions, planned and designed the experiments, guided and mentored the research. Dayal R., assisted with ELISA experiment, reviewed and edited the manuscript, Mutahar AZI., is the first corresponding author, and Salimath BP., is the second/senior corresponding author. Funding Not applicable. Data availability All data supporting the findings of this study are included in the manuscript and its supplementary materials. No additional datasets were generated or analyzed. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no conflict of financial interests or personal relationships. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Adel Zaid I. Mutahar, Email: amutahar@stanford.edu, Email: adelmutahar5@gmail.com Bharathi P. Salimath, Email: salimathuom@gmail.com References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement All data supporting the findings of this study are included in the manuscript and its supplementary materials. No additional datasets were generated or analyzed.