Role of d16HER2 splice variant in HER2-positive breast cancer

Almost 90% of human primary breast cancers (BCs) express 4-9% of total wild-type (WT) HER2 as a splice variant lacking exon 16 (d16HER2). Consequent in-frame activating deletion of 2 cysteine residues causes an imbalanced conformation resulting in receptor constitutive homodimerization, enhanced signaling activity, transformation and altered trastuzumab (T) binding. We recently produced a human d16HER2 transgenic (tg) mouse characterized by a rapid multifocal mammary tumors onset and the expression of active d16HER2 dimers on tumor cells, whose downstream signaling was found coupled to multiple activated nodes that include Src kinase. In order to dissect d16HER2 role in the aggressiveness and in the susceptibility to anti-HER2 therapy, we focused on the generation of stable d16HER2-expressing mammary tumor cell lines to be used in different bioassays. An immunomagnetic purification procedure was applied to generate primary homogeneously d16HER2-expressing cell cultures. These purified tumor cell lines were analyzed by flow cytometry and immunofluorescence and their migration/invasion ability was assessed through Boyden chamber test. d16HER2 downstream signaling was evaluated by western blot and T, Lapatinib (L) and Dasatinib (D) drugs sensitivity was measured with WST-1 and soft-agar bioassays. As controls, we compared in vitro d16HER2-models oncogenic features to those of the human BC BT474, which also co-expresses a low amount of d16HER2 transcript (4%), and to a murine mammary carcinoma cell line (wtHER2), derived from a primary mammary tumor developed in human WT HER2 tg mice. We found that d16HER2 in vitro models expressed high levels of stable homodimers combined with the recruitment of activated Src, STAT3, MAPK and Akt, as in vivo primary mammary tumors. Both in bidimensional (2D) and matrigel-cultured tumor cells, we confirmed the T lower binding capability for d16HER2 than other anti-HER2 MAbs directed to different extracellular domain epitopes. d16HER2 tumor cells had an enhanced migratory and invasive capacity compared to wtHER2 and BT474 cells and, notably, were completely resistant to T treatment and less responsive to L. In virtue of a highly activated Src kinase expression in d16HER2-positive models, we tested in 2D the therapeutic effects of D and observed that d16HER2-cells were significantly more sensitive than wtHER2 and BT474 cells. Preliminary 3D-drug susceptibility assays showed that the sensitivity of d16HER2 cells increased for all the tested drugs, if assessed in a suitable environment such as soft-agar. Our findings further indicate that the constitutive expression of d16HER2 variant identifies an aggressive tumor phenotype and confirm, at least in vitro in 2D conditions, that this isoform is resistant to T and L, whereas is sensitive to D. Further analyses are ongoing to analyze in vivo drug sensitivity of d16 in comparison to WT HER2 model.