Fig 1.
Synthesis of fluorescently labeled gemcitabine (Gem).
A. Gemcitabine (Compound 1) was covalently labeled at the 4-amino group using the Atto680 fluorophore (red) through an inert linker (blue). The portion of the parent gemcitabine molecule most important for biological activity is labeled in green. TMSCl = Trimethylsilyl chloride; Pyr = pyridine; RT = room temperature; TsCL = p-Toluenesulfonyle chloride; MeOH = methanol; TEA = triethylamine; tBuOH = tert-butanol. LC-MS was used to verify the purity of fluorescently labeled Gem-Atto via B. area under the curve analysis of the absorbance at 254 nm and C. mass to charge (m/z) ratio in positive ion mode. Reaction yields were as follows: Compound 2 (83%); Compound 3 (48%); Gem-Atto (78%).
Fig 2.
The IF viability panel (Ki67, CC-3 and Beclin 1) was used to stain four 10 μm tissue sections per time point following tumor resection. A. Representative IF images of tumor tissue sections immediately after resection (Ctrl, top row) or following 72 hours in explant culture (TE, bottom row) are shown false colored as follows: DAPI (nuclear marker, blue), Cleaved Caspase-3 (CC-3, cytosolic staining marking apoptosis, green), Beclin 1 (cytosolic staining marking autophagy, white), Ki67 (nuclear stain marking proliferation, pink). Corresponding H&E images of the preserved tissue morphology are also shown (inset, upper right). Scale bar = 100μm. B. Tissue viability was quantified by assessing positive antibody staining and used to calculate the percentage of stained cells at each time point per condition. N = 3 athymic nude mice (two tumors/mouse) were used per CDX model to provide sufficient starting material for control and treated TE tissue samples.
Fig 3.
Gemcitabine (Gem) treatment equivalence in PDAC xenograft cell line models in vivo vs. ex vivo.
Mice bearing xenografts established from three PDAC cells lines were treated with 100 mg/kg gemcitabine IP on days 0, 3, 6 and 9 to model the clinical dose and administration schedule. Tumor volumes were quantified before and during treatment for both vehicle (Ctrl) and gemcitabine treated A. AsPC-1, B. PANC-1 and C. Capan-1 tumors as a measure of treatment response. The standard deviation in the tumor volume of untreated control samples is plotted as dashed gray lines. Ex vivo explants (TE) grown from untreated xenograft tumors were treated with gemcitabine (0.5–10 μM) in culture for 48 hours. All resected tumor tissue (in vivo or ex vivo) was fresh frozen following the final dose of gemcitabine. Serially sectioned tissues were stained with the IF viability panel to assess biological response, where at least two 10 μm sections per condition were quantified for the percentage of positive cells for D. AsPC-1, E. PANC-1 and F. Capan-1 CDX tumors treated in vivo or ex vivo in explant culture. G. Representative IF viability panel images from untreated and gemcitabine treated tissues are shown with staining false colored as DAPI (all nuclei, blue), CC-3 (apoptosis, green), Beclin 1 (autophagy, white) and Ki67 (proliferation, pink). Scale bar = 100 μm. N = 3 athymic nude mice were used per CDX model tumor generation (two tumors/mouse) for in vivo gemcitabine sensitivity assessment; n = 1 athymic mouse was used per cell line for TE tissue generation.
Fig 4.
Gemcitabine (Gem) treatment equivalence in a sensitive PDX xenograft model in vivo vs. ex vivo.
Mice bearing xenografts established from a gemcitabine sensitive patient tissue were treated with 100 mg/kg gemcitabine IP on days 0, 3, 6 and 9 to model clinical dose and administration schedule. A. Tumor volumes were quantified before and during treatment for both vehicle (Ctrl) and Gem treated PDX models. Ex vivo explants (TE) grown from untreated PDX tumors were treated with varied concentrations of gemcitabine (0.5–10 μM) for 48 hours. All resected tumor tissue was fresh frozen following the final dose of gemcitabine. B. Serially sectioned tissues were stained with the IF viability panel to assess biological response, where at least two 10 μm sections per condition were quantified for the percentage of positive cells. C. Representative IF viability panel images from untreated and gemcitabine treated tissues are shown with staining, false colored as DAPI (all nuclei, blue), CC-3 (apoptosis, green), Beclin 1 (autophagy, white) and Ki67 (proliferation, pink). Scale bar = 100 μm. N = 3 NSG mice were used for in vivo gemcitabine experiments; n = 1 NSG mouse was used for TE tumor generation.
Fig 5.
Small molecule fluorophore uptake and distribution in tumor explants (TE).
Resected xenograft tissue grown as explants were incubated for 48-hour with 500 nM of nonreactive fluorophore representing different chemical scaffolds: AF488, AF546, AF647 and Atto680. A. Representative images from fresh frozen samples, shown normalized between mouse models within each respective fluorophore channel, demonstrated relatively homogenous distribution of fluorescence signal throughout the explant (TE). Control, unstained tissue was imaged in each channel to quantify autofluorescence (Ctrl). Scale bar = 50 μm. B. A minimum of ten 10 μm sections per fluorophore were quantified for average fluorescence intensity for each respective channel to evaluate non-specific fluorophore uptake in explant culture, where non-specific fluorophore uptake was the lowest for Atto680. N = 3 athymic nude mice were used per CDX model (two tumors/mouse) for TE tissue generation. Likewise, n = 1 NSG mouse was used to provide sufficient TE tumor material for all PDX fluorophore uptake experiments.
Fig 6.
Gemcitabine-Atto680 (Gem-Atto) and parent fluorophore (Atto680) administration.
A. Tumor explant (TE) tissues were subjected to increasing concentrations of Atto680 (62.5-500nM) in optimized media for 48 hours, while mice were administered Atto680 (1.25–2.50mg) for 4 hours prior to tissue collection. Average tissue fluorescence was quantified and compared across model systems. B. Tumor explants and in vivo tissues were also subjected to various concentrations of the drug-fluorophore construct, Gem-Atto. C. Representative Capan-1 H&E and corresponding Gem-Atto fluorescence images for both control and treated tumor explant samples, illustrated localized uptake in necrotic regions. D. Representative tissue ROIs showed distinct fluorescence intensity and spatial distributions in the explants versus in vivo samples for the parent fluorophore (Atto680, pink) and Gem-Atto (gray). Scale bar = 50 μm. N = 6 athymic nude mice were used per CDX (two tumors/mouse) for in vivo Atto680 and Gem-Atto assessment, while n = 5 athymic mice were used per cell line for TE tissue generation. Likewise, N = 6 NSG PDX mice were used for in vivo Atto680 and Gem-Atto experiments, while n = 1 NSG mouse was used for PDX TE tissue generation.