PMC:5007828 / 3874-10187 JSONTXT

{"project":"2_test","denotations":[{"id":"27526105-23459371-79084645","span":{"begin":2395,"end":2397},"obj":"23459371"},{"id":"27526105-23459371-79084646","span":{"begin":2559,"end":2561},"obj":"23459371"}],"text":"Results\n\np-Src expression in primary OCs\nWe examined the expression of Src by immunohistochemical staining in 381 primary epithelial OC tissues representing the major histological subtypes, using an antibody that specifically recognizes the Src protein when phosphorylated at tyrosine 416 (Src-pY416) (Figure 1a) (Supplementary Table 1). This modification is associated with enhanced biological activity of Src and represents a surrogate marker of active Src signaling. The immunohistochemical staining data, stratified by clinico-histopathological characteristics, are summarized in Table 1. Thirty-six percent of OCs expressed Src-pY416. There were no significant associations between Src-pY416 positivity and clinical stage or tumor grade. However, we found a significant association between Src-pY416 expression and histological subtype, with Src-pY416 positivity strongly associated with EAOCs (P=0.012) but not with the serous subtype (P=0.139). Src-pY416 positivity was the most common in clear cell OCs, where 24/41 cases (58.5%) were positive. In this study, Src-pY416 was only detected in 4/21 mucinous OCs.\n\np-Src expression is associated with worse patient outcomes\nSurvival data were available for 199/381 tumors. We tested for associations between active Src expression and survival using Kaplan–Meier survival analyses with log-rank testing. Samples with Src expression were first stratified into three groups based on the percentage or intensity of the staining. Src-pY416 expression was associated with worse overall survival based on staining intensity (log rank P=0.018), but was not significantly associated with survival based on the percentage of stained cells (log rank P=0.14), although the trend was the same (Figures 2a and b). When the results for staining intensity and percentage of positive cells were combined, the association with overall survival was stronger (log rank P=0.002) (Figure 2c). Median survival time was 498 days (16 months) less in tumors expressing Src compared with those that did not.\n\np-Src expression in cell line models of OC histotypes\nWe have previously described the histology of 31 OC cell lines grown as two-dimensional (2D) monolayers and as three-dimensional (3D) spheroid cultures, and found that 3D cultures more closely mimic the histological differentiation and marker expression of human primary tumors than their 2D counterparts.15 However, the majority of OC cell lines show poorly differentiated histologies when cultured in 3D, despite having originated from differentiated tumor specimens.15 In the present study, we included additional 35 cell lines to identify other good models of the main OC histotypes. In this panel of cell lines, we found that only 28.6% (2/7) of serous cell lines resembled human serous OCs when grown in 3D, but 76.5% of clear cell OC cell lines (13/17) showed clear cell differentiation when grown as 3D spheroids. These results indicate that OV-177 and SHIN-3 are good models of serous OC, and that HCH-1, JHOC-5, JHOC-7, JHOC-9, KOC-7c, OV207, OVAS, OVISE, OVMANA, OVTOKO, RMG-II, SMOV-2 and TU-OC-1 are good models of clear cell OC. These results can be found in full in Supplementary Table 2.\nWe proceeded by evaluating Src-pY416 expression in epithelial OC cell lines grown in 2D and 3D. Of 64 OC cell lines evaluated, 34 OC (53%) exhibited Src-pY416 positivity in 2D and/or 3D cultures. Cell lines with strong Src-pY416 staining included Hey.A8, HEY-C2, TYK-nu, HAC-2, SMOV-2 and OVISE (of which the latter three are derived from clear cell tumors) (Supplementary Table 3). The remaining 30/64 (48.5%) cell lines did not express active Src. Src activity in 2D and 3D cultured cells was highly correlated (R2=0.52), indicating that Src signaling is not altered by culture conditions.\n\nPhenotypic consequences of Src-pY416 inhibition in in vitro models of OC\nWe used 4-amino-5-(4-chlorophenyl)-7-(dimethylethyl)pyrazolo[3,4-d]pyrimidine (PP2) to inhibit Src in OC cell lines expressing Src-pY416. Addition of PP2 to the cell culture medium inhibits expression of Src-pY416 in OC cells in a dose-dependent manner (Supplementary Figure 1). We first evaluated the effects of PP2 on anchorage-independent growth in four cell lines that form colonies in semi-solid medium—HAC-2 and A2780.cp, which represent models of EAOCs and Hey.A8 and HEY-C2, which represent non-EAOCs. All four showed sensitivity to Src inhibition at a 50 μm PP2 dose (P\u003c0.05), while Hey.A8 also exhibited a dose-dependent sensitivity (correlation between dose and colony number, R2=0.8) (Figures 3a–e). A Src-pY416-negative cell line, TOV21G, also exhibited some sensitivity at 50 μm, suggesting that off-target effects can occur at this dose, although we note that inhibition of colony formation in TOV21G was less than in three of the four Src-pY416-positive lines. In HAC-2, Src inhibition was associated with reduction in colony numbers (Figure 3e) and a trend in reduced colony size (Supplementary Figure 2).\nAnchorage-dependent growth assays were also performed to test the sensitivity of OC cells to PP2 treatment when cultured on plastic in the presence of PP2 for 7 days. Cell lines positive for Src-pY416 (OVISE, SMOV-2 and HEY-C2) were more sensitive than the Src-pY416-negative TOV21G line (Figure 3g and Supplementary Table 4).\n\nInhibiting Src signaling in 3D models of OC\nIn vitro 3D multicellular aggregates (MCAs) mimic the OC aggregates found in the ascitic fluid of OC patients and so we tested whether inhibiting Src can impair the formation of 3D MCAs in Src-pY416-positive and -negative cell lines. Cell line models expressing high levels of active Src were Hey.A8 and HEY-C2 (high-grade OC models but subtype unknown), and SMOV-2 (clear cell OC) (Supplementary Tables 2 and 3). TOV21G was included as a clear cell model that does not express Src-pY416. Cell lines were established as MCAs in the presence or absence of PP2, and cultured for 7 days. In Src-pY416-positive cell lines (Hey.A8, HEY-C2 and SMOV-2), inhibition of Src induced a dose-dependent qualitative reduction in MCA size and number (Figure 4). MCAs were diffuse in PP2-treated Hey.A8 and SMOV-2 cells, indicating cell–cell contacts are disrupted when Src signaling is perturbed, while MCA formation was unaffected in Src-negative TOV21G cells (Figure 4).\n"}