Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2015 Oct;7(10):1120-34.
doi: 10.1039/c5ib00040h. Epub 2015 May 11.

Human Breast Cancer Invasion and Aggression Correlates With ECM Stiffening and Immune Cell Infiltration

Affiliations
Free PMC article

Human Breast Cancer Invasion and Aggression Correlates With ECM Stiffening and Immune Cell Infiltration

I Acerbi et al. Integr Biol (Camb). .
Free PMC article

Abstract

Tumors are stiff and data suggest that the extracellular matrix stiffening that correlates with experimental mammary malignancy drives tumor invasion and metastasis. Nevertheless, the relationship between tissue and extracellular matrix stiffness and human breast cancer progression and aggression remains unclear. We undertook a biophysical and biochemical assessment of stromal-epithelial interactions in noninvasive, invasive and normal adjacent human breast tissue and in breast cancers of increasingly aggressive subtype. Our analysis revealed that human breast cancer transformation is accompanied by an incremental increase in collagen deposition and a progressive linearization and thickening of interstitial collagen. The linearization of collagen was visualized as an overall increase in tissue birefringence and was most striking at the invasive front of the tumor where the stiffness of the stroma and cellular mechanosignaling were the highest. Amongst breast cancer subtypes we found that the stroma at the invasive region of the more aggressive Basal-like and Her2 tumor subtypes was the most heterogeneous and the stiffest when compared to the less aggressive luminal A and B subtypes. Intriguingly, we quantified the greatest number of infiltrating macrophages and the highest level of TGF beta signaling within the cells at the invasive front. We also established that stroma stiffness and the level of cellular TGF beta signaling positively correlated with each other and with the number of infiltrating tumor-activated macrophages, which was highest in the more aggressive tumor subtypes. These findings indicate that human breast cancer progression and aggression, collagen linearization and stromal stiffening are linked and implicate tissue inflammation and TGF beta.

Figures

Figure 1
Figure 1. Tumor progression correlates with significant ECM remodeling
(A) Immunohistological staining (H&E, trichrome, picrosirius red, and second harmonic generation) and imaging of human tumor samples featuring areas of normal, DCIS and invasive ductal carcinoma (IDC). (B) Quantitative analysis of trichrome and picrosirus red staining as a measure of collagen density. Bars represent the average of 20 patient samples and error bars represent standard deviation. (*** denotes P<0.05) (C) Quantitative analysis of collagen fiber length from second harmonic generation imaging represented as a histogram. Statistical significance was determined via a Wilcoxon Rank Sum Test, with DCIS and IDC showing statistically significant differences relative to normal (P<0.05).
Figure 2
Figure 2. Human tumor progression is accompanied by increased stromal density and enhanced mechanosignaling
(A) Q-POL and immunofluorescence imaging of human tumor samples. Q-POL images shown as a heat map of birefringence reflecting increased stromal density, alignment, and stiffness. Immunofluorescence imaging for mechanosignaling in the tumors includes activated beta 1 integrin, pFAK (Y397), and pMLC (S19). (B) Quantification of Q-POL and immunofluorescence imaging. Immunofluorescence quantified as average signal intensity per cell. Bars represent average of 20 patient samples and error bars represent standard deviation. (*** denotes P<0.05)
Figure 3
Figure 3. ECM remodeling at the tumor invasive front is correlated with increased ECM stiffness
(A) H&E staining and nuclear immunofluorescence images used to determine regions for atomic force microscopy (AFM) testing comparing tumor invasive front and adjacent healthy tissue. (B) AFM schematic (C) AFM force map result represented as a heat map. (D) Q-POL imaging comparing adjacent normal tissue and tumor invasive front measured via AFM. (E) Histograms of stiffness values from adjacent normal tissue and tumor invasive front. (F) Nuclear immunofluorescence images used to determine regions for atomic force microscopy (AFM) testing comparing tumor invasive front and tumor core. (G) AFM force map result represented as a heat map. (H) Q-POL imaging comparing tumor core and tumor invasive front measured via AFM. (I) Histograms of stiffness values from tumor core and tumor invasive front.
Figure 4
Figure 4. ECM remodeling and stiffening with tumor progression correlates with increased immune infiltrate
(A) Immunohistological staining to assess total immune infiltrate (CD45), macrophage infiltrate (CD68, CD163) and associated tumor cell signaling (pSMAD) comparing healthy, DCIS and IDC tissue. IDC tissue analysis was broken down further comparing the tumor invasive front and the tumor core. (B) Quantification of CD68 and CD163 infiltrate determined as a percentage of CD68+ relative to all cells and quantification of pSMAD signaling assessed as pSMAD staining intensity per pSMAD+ cell. Bars represent average of 20 patient samples and error bars represent standard deviation. (*** denotes P<.05)
Figure 5
Figure 5. Human breast cancer subtype influences ECM remodeling and mechanics associated with IDC
(A) Immunohistological staining (H&E, picrosirius red and second harmonic generation) and imaging of human tumor samples featuring areas of different subtypes (Luminal A, Luminal B, Her2+ and Basal). (B) Quantification of picrosirius red area and Q-POL imaging as a measure of collagen density, tissue birefringence and mechanics. (C) Quantification of AFM measurements of human tissue samples. The upper 10% of stiffness values are shown and normalized to measurements from prophylaxis tissue serving as a normal control. Bars represent an average of 5 patient samples per subtype and error bars represent standard deviation. (*** denotes P<0.05)
Figure 6
Figure 6. ECM remodeling associated with each subtype influences subsequent mechanosignaling
(A) Immunofluorescence imaging for mechanosignaling in the tumors including activated beta 1 integrin, pFAK (Y397), and pMLC (S19). (B) Quantification of immunofluorescence as average signal intensity per cell. (C) Immunohistological staining for YAP. Luminal A and Luminal B tumors assessed as one group due to lack of differences in YAP intensity. (D) Quantification of YAP staining as average percentage of nuclear YAP+ cells per patient within each subtype. Bars represent an average of 5 patient samples per subtype and error bars represent standard deviation. (*** denotes P<0.05)
Figure 7
Figure 7. AFM testing of human tissue reveals increased mechanical heterogeneity within the more aggressive subtypes
(A) Atomic force microscopy sample force map result represented as a heat map. (B) Histograms of AFM mechanical measurements for human prophylaxis tissue and human tumor tissue of each subtype. Statistical significance was determined via a Wilcoxon Rank Sum Test, with all subtypes showing statistically significant differences relative to prophylaxis tissue (p<0.05).
Figure 8
Figure 8. Tumor aggression and increased immune infiltrate correlates with ECM stiffness
(A) Immunohistological staining to assess total immune infiltrate (CD45), macrophage infiltrate (CD68), and associated tumor cell signaling (pSMAD) comparing breast cancer human subtype. IDC tissue analysis was broken down further comparing tumor invasive front and tumor core. (B) Quantification of CD68 infiltrate determined as a percentage of CD68+ relative to all cells and quantification of pSMAD signaling assessed as pSMAD staining intensity per pSMAD+ cell. (C) Correlation curves comparing ECM stiffness and macrophage infiltrate and macrophage infiltrate and pSMAD signaling. Curve fit via linear regression model. Bars represent an average of 5 patient samples per subtype and error bars represent standard deviation. (*** denotes P<0.05)

Similar articles

See all similar articles

Cited by 174 articles

See all "Cited by" articles

MeSH terms

Feedback