Fig 1.
Hematoxylin/Eosin staining of paraffin sections of U87 tumor grafts 6 days after inoculation.
a: overview of explant/CAM interface (scale bar: 200μm). b: detailed view showing tumor cells (arrows) invading CAM parenchyma with hatched line indicating invasion front (scale bar: 25μm).
Fig 2.
Tumor growth as function of drug treatment.
a: Representative human U87 tumor explants subjected to treatments as indicated (from top left to bottom right): CTRL (PBS, 10μl topically applied around explant); Avastin® (AVA; i.v. injected: 10mg/kg); α-cyano-4-hydroxycinnamic acid (CHC, 60mg/kg; topically applied around explant); AVA+CHC (10mg/kg+60mg/kg). White arrows point to respective primary mass. b: Quantification of tumor growth kinetics of human U87 by image analysis between day 2 (considered 100%) and 6 of CAM inoculation Day 0 = explant day. Data are means ± SEM, n = 6. Statistical differences were tested with two-way ANOVA. *** = p<0.001. c + d + e: Growth quantification using D-loop PCR data obtained with human specific (c) and canine specific (d +e) primers normalized to D-loop amplicon abundance of chicken host tissue, as a function of treatment (AVA, AZD and AVA+ AZD), relative to control (CTRL) conditions. Tumor explants (c: human U87; d: canine 17CM98 and e: canine D17) were treated as indicated: control (PBS, 50μl i.v. injected, CTRL); Avastin® (i.v. injected, 10mg/kg, AVA); AZD3965 (i.v. injected, 2.5μM/egg, AZD); AVA+AZD (10mg/kg + 2.5μM/egg). Data in c), d) and e) are means ± SEM, n = 6. Statistical differences: one-way ANOVA, * = p < 0.05, ** = <0.02.
Fig 3.
Blood perfusion of tumor explants.
a + c + e: Representative perfusion images of tumor explants (a): U87; (c) 17CM98 and (e): D17, subjected to the indicated treatments. Minimal flow = blue; maximal flow = red; see color scale on the right of panel a. Note the greenish ring (“halo”) of angiogenic activity around the tumor grafts. b + d +f: Quantified flux of the tumor vicinity (b): U87; (d): 17CM98 and (f) D17, in response to indicated treatments. Data are means ± SEM, (b) n = 6, (d) n = 5; (f) n = 8. Statistical differences were tested with two-way ANOVA. * = p < 0.05, ** = <0.02. Scale bars = 1mm. AU = arbitrary units.
Table 1.
Scoring of cell line performance on the CAM.
Fig 4.
Assessment of local tumor hypoxia.
a: Representative brightfield image for illustration of different tumor regions. b: Representative images of pimonidazole (Pimo) stained (red) cross sections of U87 tumors after CTRL (PBS), AVA, CHC and AVA+CHC treatments. Note the designated tumor regions in the upper left Pimo-stained panel. Both the severity of tissue hypoxia (c), indicated by the intensity ratio of pimonidazole signals (maximal Pimo signal / average tissue background) and the extent of tissue hypoxia (d), as depicted by the percent area ratio (pimo-positive area / total tumor area), were quantified as function of treatment. Data are means ± SEM, n = 4. Statistical differences were tested with one-way ANOVA. * = p<0.05.
Fig 5.
Dissemination of tumor cells in response to treatment.
a: For quantification of cell dissemination genomic DNA (gDNA) and mitochondrial DNA (mtDNA) were extracted from MDA-MB231, 17CM98 and D17 tumor explants and nearby CAM tissue, divided into 4 zones (Zone 0, 1, 2, and 3; 2mm x 15mm stripes as indicated). b: Using human or chicken specific PCR primers against the hypervariable D-loop region of mtDNA (see S1 Table for sequences and fluorescent labels) pilot PCRs revealed species-specific amplification of the D-loop amplicon in comparison with positive control gDNA from human MDA-MB231 cells and chicken liver. Left gel: PCR using human specific D-loop primers; right gel: PCR using chicken specific D-loop primers. Loading of either gel: lane 1 = DNA Mw marker; lane 2 = DNA extraction (gDNA+mtDNA) from cultured MDA-MB231 cells; lane 3 = DNA extraction from chicken liver; lane 4 = DNA extraction from CAM-resident MDA-MB231 explant. c: Content of human DNA, normalized to content of chicken (CAM derived) DNA, was quantified for zones 0–4 of MDA-MB231 explants using ddPCR with human or chicken specific D-loop primers/probes, to estimate cell motility as function of treatment. For each treatment, zone 0 (explant periphery) was set to 100% of detected MDA-MB231 DNA signals. Data are means ± SEM, n = 3. Statistical differences were tested with Wilcoxon rank-sum test. * = p<0.05. d + e: Spreading behavior of 17CM98 and D17 cells, respectively, were quantified by qPCR with specific primers against mitochondrial D-Loop of canine (signal) and chicken (normalization) sequences (S2 Table). Handling of the samples was exactly as described in (a). d: canine 17CM98 explants showed increased spreading into zone 1 in response to AVA monotherapy and in the combination therapy. e: canine D17 showed overall a higher spreading compared to canine 17CM98. These explants demonstrated the superiority of the combination therapy in diminishing the spreading tendency of cells compared with CTRL and AVA-monotherapy. Data are means ± SEM, d) n = 12 and e) n = 6. Statistical analysis: Kruskal-Wallis test, ns.
Fig 6.
Quantitative PCR with human-specific primers against MCT1 and MCT4 as a function of treatment (AVA, AZD, AVA+AZD) of human U87 tumor explants.
MCT1 (a) and MCT4 (b) gene expression was normalized to β-actin abundance and relative to control (CTRL). Data are means ±SEM, n = 4. Statistical analysis: Kruskal-Wallis test, ns.