TGF-β inhibition combined with cytotoxic nanomedicine normalizes triple negative breast cancer microenvironment towards anti-tumor immunity

Tumor normalization strategies aim to improve tumor blood vessel functionality (i.e., perfusion) by reducing the hyper-permeability of tumor vessels or restoring compressed vessels. Despite progress in strategies to normalize the tumor microenvironment (TME), their combinatorial antitumor effects with nanomedicine and immunotherapy remain unexplored. Methods: Here, we re-purposed the TGF-β inhibitor tranilast, an approved anti-fibrotic and antihistamine drug, and combined it with Doxil nanomedicine to normalize the TME, increase perfusion and oxygenation, and enhance anti-tumor immunity. Specifically, we employed two triple-negative breast cancer (TNBC) mouse models to primarily evaluate the therapeutic and normalization effects of tranilast combined with doxorubicin and Doxil. We demonstrated the optimized normalization effects of tranilast combined with Doxil and extended our analysis to investigate the effect of TME normalization to the efficacy of immune checkpoint inhibitors. Results: Combination of tranilast with Doxil caused a pronounced reduction in extracellular matrix components and an increase in the intratumoral vessel diameter and pericyte coverage, indicators of TME normalization. These modifications resulted in a significant increase in tumor perfusion and oxygenation and enhanced treatment efficacy as indicated by the notable reduction in tumor size. Tranilast further normalized the immune TME by restoring the infiltration of T cells and increasing the fraction of T cells that migrate away from immunosuppressive cancer-associated fibroblasts. Furthermore, we found that combining tranilast with Doxil nanomedicine, significantly improved immunostimulatory M1 macrophage content in the tumorigenic tissue and improved the efficacy of the immune checkpoint blocking antibodies anti-PD-1/anti-CTLA-4. Conclusion: Combinatorial treatment of tranilast with Doxil optimizes TME normalization, improves immunostimulation and enhances the efficacy of immunotherapy.

Elastic modulus and hydraulic conductivity. Characterization of the mechanical properties and calculation of the elastic modulus were determined using an unconfined compression experimental protocol. Following excision of the primary tumor or the macroscopically metastatic nodules, specimens were loaded on a high precision mechanical testing system (Instron, 5944, Norwood, MA, USA) and compressed to a final strain of 30% with a strain rate of 0.1 mm/min. The dimensions of the primary tumor specimens were 3 × 3 × 2 mm (length × width × thickness), while metastatic nodules were tested as a whole owing to their small size. The elastic modulus was calculated from the slope of the stress-strain curve at the 25-30% strain range [3,7,9]. For the calculation of the hydraulic conductivity, stress relaxation experiments were performed in unconfined compression. Specimens underwent four cycles of testing for each of which a 5% compressive strain was applied for 1 minute, followed by a 10 min hold. Subsequently, a common biphasic model of soft tissue mechanics was employed [10], accounting for both the solid phase (cells and extracellular matrix) and the fluid phase (interstitial fluid) of the tumor. The hydraulic conductivity was calculated by fitting the model to the experimental data.
Tumor opening. For the tumor opening measurement, a cut was made along the tumor's longest axis (∼80% of its thickness). The tumor was then allowed to relax for 10 min to diminish the effect of resident transient and poroelastic response. The opening at the surface of tumor was then measured using a digital caliper [11,12].
Interstitial Fluid Pressure. Interstitial fluid pressure (IFP) was measured in vivo using the previously described wick-in-needle technique after mice were anesthetized with i.p. injection of Avertin (200mg/kg) and prior to tumor excision [3,7,9,13].
Tumor elasticity using AFM. Primary E0771 tumors were excised and immediately transferred into ice-cold PBS supplemented with a protease inhibitor cocktail (cOmplete Mini, Roce Dianostics GmbH, 1 tablet per 10 mL) [14,15]. Tumors were then sliced using a 0.5 mm prescored tissue cord matrice. Each specimen was immobilized on a plastic dish with a thin layer of two-component fast drying epoxy glue and stored at 4 °C to avoid tissue degradation [14,15]. AFM measurements were performed 1-72 h post tumor removal, so as to prevent any alterations in stiffness profiles [14,15]. AFM experiments were carried out on a commercial AFM (Molecular Imaging-Agilent PicoPlus AFM system) was used. Silicon nitride D cantilevers (MLCT Bruker Company with the half-open angle of the pyramidal face of θ ∼ 20°, tip radius: 20 nm, frequency in air: 15 kHz) were used in all experiment. The maximum applied loading force was set to 1.8 nN, the exact spring constant k of the cantilever was determined before each experiment using the thermal tune method and the deflection sensitivity was determined in fluid using petri dishes as an infinitely stiff reference material [16]. The collected force curves were analyzed by AtomicJ [16] so as to calculate the sample's Young's modulus using the Hertz model (the Poisson ratio, v, was set to 0.5).

Vascular perfusion.
To assess functional vasculature, mice were anesthetized with with i.p. injection of Avertin (200mg/kg) and slowly injected intracardially with 100 μl (4mg/kg) of biotinylated lycopersicon esculentum lectin (B-1175, Vector Labs) which was allowed to distribute throughout the body for 7 min [8]. Finally, mice were sacrificed via CO 2 inhalation and tumors/lungs were removed. Excised tumors/lungs were fixed in 4% PFA (Sigma) and processed as described above for IHC analysis. The number of blood vessels was measured from the positive staining of the endothelial marker CD31 (MEC13.3, BD Pharmingen 1:100) while the area fraction of perfused vessels was determined as the ratio of lectin and CD31 overlapping staining to CD31 positive staining. CD31 signal was detected with Alexa Fluor-647 Goat Anti-Rat IgG (H+L) (Invitrogen, A-21247, 1:400) secondary antibody and lectin staining with Streptavidin Alexa Fluor 488 conjugate (Invitrogen, S11223, 1:200 dilution).
Hypoxia studies. Mice bearing orthotopic E0771 breast tumors were injected with 60mg/kg of pimonidazole HCl at 60mg/kg 2 hr prior to tumor removal [8]. Primary tumors were then excised, fixed in 4% PFA, embedded in OCT and processed accordingly for IHC. Hypoxic regions were detected using the mouse anti-pimonidazole RED 549 conjugate antibody (HP7-100Kit, 1:100). Hypoxic area fraction across different treatment groups was normalized to DAPI staining.
T cell-CAF distance. Distances between CD3 + T cells and the corresponding nearest αSMA + area were calculated using custom MATLAB scripts with built-in image processing functions. CD3 + T cells were determined automatically based on signal intensity, size, and morphology thresholds and αSMA + area was similarly determined based on an intensity threshold. The thresholds were chosen manually by reviewing the attributes of all images, and the same thresholds were used for all images. Then, the centroids of CD3 + T cells were determined, and each cell's distance to the nearest αSMA + area was measured by finding the distance between the centroids and the nearest αSMA + pixels. Also, images were divided into two groups. One group contains images that represent αSMA-rich tissue and the other group contains images that are αSMA-poor. The average αSMA + area fraction was calculated and used as the parameter to determine which images displayed a high or low αSMA + area fraction. Quantification of αSMA positive staining was normalized to the total image area among different treatment groups.
Image acquisition. Images of immunelabeled tumor sections from the interior and periphery were acquired at 10x magnification using the Olympus BX53 fluorescence microscope. In order for the images to be comparable they were taken at the same settings and analyzed using a previously developed in-house code in MATLAB (MathWorks, Inc., Natick, MA, USA) [3,7,9,11].