P2-01-05: Integrin avb6 Mediates HER2−Driven Invasion in Breast Cancer.

Abstract

Abstract We have shown strong expression of integrin αvβ6 reduces the 5 year survival of HER2−positive breast cancers from 66% (HR 1.84) for moderate/low αvβ6 expressors to 54% (HR 2.18) in cases with strong expression (2063 cases, unpublished). This is in direct comparison to HER2/αvβ6-double negative cases, where strong αvβ6 expression reduces survival from 86% (HR 1.00) to 77% (HR 1.20). The biological mechanism underlying these observations was investigated in two isogenic breast cancer models: MCF-7/neo-1 and MCF-7/HER2−18 (a gift from Prof. M-C. Hung, USA) and MCF10A and MCF10A.CA1a. Flow cytometry showed MCF-7/HER2−18 expressed high levels of both HER2 and αvβ6 whereas MCF-7/neo-1 expressed low levels of both receptors. MCF10A and MCF10A.CA1a both expressed high levels of αvβ6 whereas only MCF10A.CA1a expressed elevated levels of HER2. In charcoal-stripped (cs)-serum, comparing MCF-7/neo-1 and MCF-7/HER2−18, HRGβ1(1μM), which stimulates HER2/HER3 heterodimers, increased proliferation by 50.2%±9% (P=0.048) and 66.2%±5.5% (P=0.003), in MCF-7/neo-1 and MCF-7/HER2−18 cells respectively. In contrast, Herceptin reduced proliferation by 32.3%±13.4% (P=0.003) and 15.2%±3.4% (P=0.028), respectively. MCF10A and MCF10A.CA1a proliferation remained unchanged with HRGβ 1 treatment and antibody-blockade of αvβ6 did not affect proliferation of any cell line. (NB, in complete serum there was no effect on proliferation of any of the above treatments). Invasion through Matrigel of MCF-7/HER2−18 was inhibited by antibody blockade (10μg/ml) of αvβ6 (mAb 10D5; 38.6%±20.8%, P=0.005) or HER2 (Herceptin, 10μg/ml; 40.1%±28.6%, P=0.01). The same trend was observed in MCF10A.CA1a invasion (83%±30.2% (P=0.025) with 10D5 and 80.4%±8.7% (P=0.022) with Herceptin). Combination of both antibodies had no additional effect. siRNA knockdown of αvβ6 or HER2 in MCF-7/HER2−18 and MCF10A.CA1a cells also reduced invasion to a similar extent as the blocking antibodies. This suggests that HER2 driven breast carcinoma invasion is mediated by αvβ6. To investigate this further HER2/3 was stimulated with HRGβ1, which consistently increased invasion by 111.5%±35.4% (P=0.011) in MCF-7/HER2−18 cells and by 57%±34% (P=0.042) in MCF10A.CA1a cells; an increase that was abrogated by co-treatment with 10D5 or Herceptin. To determine the mechanism through which HER2 and αvβ6 co-operate we examined several signalling pathways. Analysis of total or activated Akt, ERKI/II, c-Jun or Src in the MCF-7 model showed no changes. However, elevated total and phospho-Stat3 in MCF-7/HER2−18 were observed and siRNA knockdown, or small-molecule inhibition, of Stat3 suppressed invasion of MCF-7/HER2−18 cells (54.5%±27.3% (P=0.008) and 55.3%±33.3% (P=0.01) respectively), possibly suggesting that activation of Stat3 may link αvβ6 and HER2 co-operative signalling in this model. Interestingly, Akt was constitutively phosphorylated in MCF10A.CA1a cells and, moreover, 10D5 reduced these levels suggesting αvβ6 may influence HER2 signalling via Akt in these cells. These data confirm HER2−driven invasion is αvβ6-mediated and provide a mechanistic explanation for our clinical observations. We suggest HER2 and αvβ6 should be considered as dual targets for future therapy of some breast cancers. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-01-05.