Several transgenic mice models solidly support the hypothesis that HER2 (ERBB2) overexpression or mutation promotes tumorigenesis. Recently, a HER2 splice variant lacking exon-16 (Δ16HER2) has been detected in human breast carcinomas. This alternative protein, a normal byproduct of HER2, has an increased transforming potency compared to wild-type (wt) HER2 receptors. To examine the ability of Δ16HER2 to transform mammary epithelium in vivo and to monitor Δ16HER2-driven tumorigenesis in live mice, we generated and characterized a mouse line that transgenically expresses both human Δ16HER2 and firefly luciferase under the transcriptional control of the MMTV promoter. All the transgenic females developed multifocal mammary tumors with a rapid onset and an average latency of 15.11 weeks. Immunohistochemical analysis revealed the concurrent expression of luciferase and the human Δ16HER2 oncogene only in the mammary gland and in strict correlation with tumor development. Transgenic Δ16HER2 expressed on the tumor cell plasma membrane from spontaneous mammary adenocarcinomas formed constitutively active homodimers able to activate the oncogenic signal transduction pathway mediated through Src kinase. These new transgenic animals demonstrate the ability of the human Δ16HER2 isoform to transform “per se” mammary epithelium in vivo. The high tumor incidence as well as the short latency strongly suggests that the Δ16HER2 splice variant represents the transforming form of the HER2 oncoprotein.

BackgroundMounting clinical and experimental evidence suggests that the shift of carcinomas towards a mesenchymal phenotype is a common paradigm for both resistance to therapy and tumor recurrence. However, the mesenchymalization of carcinomas has not yet entered clinical practice as a crucial diagnostic paradigm.Methodology/Principal FindingsBy integrating in silico and in vitro studies with our epithelial and mesenchymal tumor models, we compare herein crucial molecular pathways of previously described carcinoma-derived mesenchymal tumor cells (A17) with that of both carcinomas and other mesenchymal phenotypes, such as mesenchymal stem cells (MSCs), breast stroma, and various types of sarcomas. We identified three mesenchymal/stromal-signatures which A17 cells shares with MSCs and breast stroma. By using a recently developed computational approach with publicly available microarray data, we show that these signatures: 1) significantly relates to basal-like breast cancer subtypes; 2) significantly relates to bone metastasis; 3) are up-regulated after hormonal treatment; 4) predict resistance to neoadjuvant therapies.Conclusions/SignificanceOur results demonstrate that mesenchymalization is an intrinsic property of the most aggressive tumors and it relates to therapy resistance as well as bone metastasis.

The crucial role of HER2 in epithelial transformation and its selective overexpression on cancer tissues makes it an ideal target for cancer immunotherapies such as passive immunotherapy with Trastuzumab. There are, however, a number of concerns regarding the use of monoclonal antibodies which include resistance, repeated treatments, considerable costs, and side effects that make active immunotherapies against HER2 desirable alternative approaches. The efficacy of anti-HER2 DNA vaccination has been widely demonstrated in transgenic cancer-prone mice, which recapitulate several features of human breast cancers. Nonetheless, the rational design of a cancer vaccine able to trigger a long-lasting immunity, and thus prevent tumor recurrence in patients, would require the understanding of how tolerance and immunosuppression regulate antitumor immune responses and, at the same time, the identification of the most immunogenic portions of the target protein. We herein retrace the findings that led to our most promising DNA vaccines that, by encoding human/rat chimeric forms of HER2, are able to circumvent peripheral tolerance. Preclinical data obtained with these chimeric DNA vaccines have provided the rationale for their use in an ongoing Phase I clinical trial (EudraCT 2011-001104-34).

We developed a new phage-display based approach, the Large Fragment Phage Display (LFPD), that can be used for mapping conformational epitopes on target molecules of immunological interest. LFPD uses a simplified and more effective phage-display approach in which only a limited set of larger fragments (about 100 aa in length) are expressed on the phage surface. Using the human HER2 oncoprotein as a target, we identified novel B-cell conformational epitopes. The same homologous epitopes were also detected in rat HER2 and all corresponded to the epitopes predicted by computational analysis (PEPITO software), showing that LFPD gives reproducible and accurate results. Interestingly, these newly identified HER2 epitopes seem to be crucial for an effective immune response against HER2-overexpressing breast cancers and might help discriminating between metastatic breast cancer and early breast cancer patients. Overall, the results obtained in this study demonstrated the utility of LFPD and its potential application to the detection of conformational epitopes on many other molecules of interest, as well as, the development of new and potentially more effective B-cell conformational epitopes based vaccines.

The experimental data herein were obtained through grants from the Italian Association for Cancer Research (AIRC, IG11889) and the Italian Ministry for the Universities and Research. We thank Prof. Guido Forni for being a beacon in our scientific life.