Hypoxia Affects the Structure of Breast Cancer Cell-Derived Matrix to Support Angiogenic Responses of Endothelial Cells

Abstract

Hypoxia, a common feature of the tumor environment and participant in tumor progression, is known to alter gene and protein expression of several Extracellular Matrix (ECM) proteins, many of which have roles in angiogenesis. Previously, we reported that ECM deposited from co-cultures of Neonatal Fibroblasts (NuFF) with breast cancer cells, supported 3-dimensional vascular morphogenesis. Here, we sought to characterize the hypoxic ECM and to identify whether the deposited ECM induce angiogenic responses in Endothelial Cells (ECs). NuFF and MDA-MB-231 breast cancer cells were co-cultured, subjected to alternating cycles of 24 hours of 1% (hypoxia) and 21% (atmospheric) oxygen and de-cellularized for analyses of deposited ECM. We report differences in mRNA expression profiles of matrix proteins and crosslinking enzymes relevant to angiogenesis in hypoxia-exposed co-cultures. Interestingly, overt differences in the expression of ECM proteins were not detected in the de-cellularized ECM; however, up-regulation of the cell-binding fragment of fibronecin was observed in the conditioned media of hypoxic co-cultures. Ultrastructure analyses of the de-cellularized ECM revealed differences in fiber morphology with hypoxic fibers more compact and aligned, occupying a greater percent area and having larger diameter fibers than atmospheric ECM. Examining the effect of hypoxic ECM on angiogenic responses of ECs, morphological differences in Capillary-Like Structures (CLS) formed atop de-cellularized hypoxic and atmospheric ECM were not evident. Interestingly, we found that hypoxic ECM regulated the expression of angiogenic factors and matrix metalloproteinases in CLS. Overall, we report that in vitro, hypoxia does not alter the composition of the ECM deposited by co-cultures of NuFF/MDA-MB-231, but rather alters fiber morphology, and induces vascular expression of angiogenic growth factors and metalloproteinases. Taken together, these results have important implications for understanding how the hypoxic matrix may regulate angiogenesis in tumors.

Keywords: Angiogenesis; Extracellular matrix; Hypoxia.

Figure 1. Schematic of culture set up and hypoxia exposure

NuFF and MDA231 cells were co-cultured in a 1:1 ratio in atmospheric oxygen (21% O₂) or alternating 24 hour cycles of hypoxia (1% O₂) and atmospheric oxygen (21% O₂) for a period of 9 days. Subsequently, cells were lysed using a strong base and resulting ECM was investigated for changes in protein expression or was utilized as a substrate for HUVECs to form CLS.

Figure 2. Patterns of gene expression in hypoxia and atmospheric-exposed co-cultures differed during the culture period

(A) NuFF/MDA231 co-cultures were assessed for changes in the expression patterns of genes important for ECM assembly, stability and angiogenesis along the culture period. Differences were observed between time points for each gene tested. *p≤0.05, **p≤0.01, ***p≤0.001.

Figure 3. Hypoxia induced changes in ECM protein expression differed depending on the context of culture

(A) Analysis of collagen I and fibronectin protein expression in the conditioned media of hypoxia and atmospheric-exposed co-cultures. Ponceau S (Ponc-S) was utilized as a loading control (B) Western blots of collagen I, fibronectin, collagen IV and laminin in the de-cellularized hypoxic and atmospheric ECM. GAPDH was utilized as a loading control.

Figure 4. Hypoxia altered the morphology and characteristics of de-cellularized ECM

(A) ECM deposited from day 9 hypoxic and atmospheric-exposed co-cultures was observed using immunofluorescence for fibronectin, collagens I and IV, laminin and tenascin-C. (B) SEM images depict the ultra-structural morphology of hypoxic and atmospheric ECM in low (left) and high (right) magnifications. Scale bars are 5μM and 1μM for low and high magnification images, respectively. (C) Quantification of fibers for differences in (i) branches, (ii) area, and (iii) maximum diameters in hypoxic and atmospheric ECM. (D) Total ECM concentration in hypoxic and atmospheric ECM. *p≤0.05, **p≤0.01, ***p≤0.001.

Figure 5. Vascular morphogenesis and morphology were not altered on hypoxic ECM

(A) Immunofluorescence images of CLS formed from HUVECs cultured atop de-cellularized ECM from hypoxic and atmospheric-exposed NuFF/MDA231 co-cultures. Insets are high magnification of the boxed area. Scale bars are 50 μM for image insets. Phalloidin in green, CD31 in red and nuclei in blue. (B) Analyses of vascular morphological features including (i) vascular abundance, (ii) mean vascular branches, (iii) mean vascular diameters and (iv) maximum vascular diameters between structures grown on hypoxic and atmospheric ECM.

Figure 6. Pro-angiogenic factors and MMPs were differentially expressed in CLS grown atop hypoxic and atmospheric ECM

(A) qRT-PCR for (i) VEGFA, VEGFR2, Ang1, Ang 2 and (ii) MMPs 1,2, and 9 and MT1-MMPin CLS grown on hypoxic and atmospheric de-cellularized ECM. (B) (i) Western blot for MT1-MMP and (ii) zymography for MMP1 and 2 in CLS grown on hypoxic and atmospheric ECM. Serum free media (SFM) was used as a control. *p≤0.05, **p≤0.01, **p≤0.001.