Fig 2: Circ-HER2 is enriched in part of TNBC. a The relative expression levels of circ-HER2 determined by q-PCR using junction specific primers in mammary normal cell lines (MCF10A) and different types of breast cancer cell lines as indicated. b The relative expression levels of circ-HER2 in 23 paired HER2 positive breast cancer specimens and the corresponding paired normal adjacent tissues was determined by using junction specific primers. c Left, the relative expression levels of circ-HER2 in 59 paired TNBC specimens and the corresponding paired normal adjacent tissues was determined by using junction specific primers; right, the ratio of circ-HER2 enrichment in 59 TNBC tissues. d Left, the relative expression levels of linear HER2 in 59 paired TNBC specimens and the corresponding paired normal adjacent tissues was determined by using linear specific primers. Right, the relative expression levels of linear HER2 in TNBC tissues from TCGA database. e Kaplan-Meier analysis of the correlation between circ-HER2 RNA expression and overall survival in the 59 TNBC patient’s cohort. Lines show the mean ± SD, ***, p < 0.001; **, p < 0.01; *p < 0.05. Data are representative from at least 2–3 experiments with similar results

Fig 3: Sema3d signals via ErbB2 in endothelial cells.(a) Western blot for phospho-ErbB2 (Y1248) from HUVECs incubated with increasing concentrations of Sema3d for 30 min. (b) Transwell migration assay showing the percentage HUVECs that migrated through a porous membrane in the presence of Sema3d or Sema3e when compared to vehicle control in the presence of a control siRNA (left) or ErbB2 siRNA (right). ***P<0.001, *P<0.05, NS, not significant (one-way ANOVA between groups P<0.001 for both groups; post-hoc multiple comparisons, Tukey's test) (c) Western blot for phospho-Akt (S473) and phospho-ErbB2 (Y1248) from HUVECs incubated with increasing concentrations of Sema3d for 5 min. (d) Western blot for ErbB2 and phospho-Akt (S473) of HUVECs in the presence of control or ErbB2 siRNA treated with Sema3d (10 nM) for 5 min. (e) Quantification of ErbB2 knockdown and Akt phosphorylation from (d). **P<0.01, *P<0.05 (Student's t-test, n=3).

Fig 4: Evaluation of tissues obtained from human patients with EpiView-D4.Brightfield cytology imaging of non-tumor human breast tissue specimens were sampled by FNA and imaged by standard benchtop microscopy (a) and brightfield imaging by EpiView (b). Left panels: LPF views by standard microscope equipped with a 10× objective lens (top) and native view on EpiView (bottom). Right panels: standard microscope image obtained by a 40× objective (top), showing the same ROI outlined with red dashed line in the 10× image, and a similar view obtained by digital zoom on the EpiView (bottom). Scale bars for LPF and HPF images are 0.2 mm and 30 µm, respectively. c, d Similar sequence of images as a and b but with a representative specimen with cytological evidence of tumor. e Results of HER2 credentialing by EpiView-D4 for 19 different patient breast tissue specimens: non-tumor, HER2-negative tumor, HER2-positive tumor (N = 7, 6, 6, respectively). Each data point represents the mean D4 fluorescence intensity (after background subtraction) of duplicate assays measured by the EpiView-D4 for an individual tissue specimen. The horizontal line represents the mean D4 signal intensity for each category. There was a statistically significant difference between groups as determined by one-way ANOVA (F (2, 16) = 9.264, p = 0.0021). Groups that are different by multiple comparison testing (Tukey post hoc test, p ≤ 0.05) are marked with different letters. HER2-positive tumors are defined as IHC 3+ or IHC 2+ with HER2 amplification by FISH. f EpiView-D4 readout vs ELISA for HER2 for all patient specimens. A strong, positive correlation between EpiView-D4 and ELISA across the 19 paired measurements is observed, which was statistically significant (Pearson r = 0.956, p < 0.0001, 95% CI: 0.886–0.983).

Fig 5: Immunohistochemical status of the candidate survival risk stratification markers secreted protein acidic and rich in cysteine (SPARC), Human Epidermal Growth Factor Receptor 2 (HER2),inhibin subunit beta A (INHBA), matrix metallopeptidase 11 (MMP11), tumor protein P53 (TP53), and platelet-derived growth factor receptor-beta (PDGFRB), and galectin-10 identified by quantitative polymerase chain reaction (qPCR) and proteomic analysis. Representative images of high and low immunostaining of each candidate marker are shown.