Introduction

Targeted drug delivery to a specific cell population invariably results in differential exposure to therapeutic agents at the cellular level. As a consequence, targeted cells will be variably affected by the therapeutic agent. In addition, it is now evident that biological effects may also occur in untargeted cells, as they respond to signals sent by their targeted neighbors (1,2). These bystander effects can be detrimental or beneficial in nature and have attracted interest in immunology, pharmacology, gene therapy, and targeted radiation therapy. Among the bystander effects reported are cell killing, DNA damage, induction of micronuclei, apoptosis, mutations, neoplastic transformation, chromosome aberrations, altered proliferation, and propagation of stress responses (3).

Approaches have been developed that permit noninvasive detection of drug intake at the subtissue level (4). Moreover, nanoparticles that encapsulate a drug together with a fluorescent marker show promise in lowering drug-detection limits to the cellular level (5). To improve our understanding of the response of multicellular systems to targeted therapies, it is necessary to quantify the amount of drug delivered to individual cells and measure the corresponding biological responses (6,7). In addition, responses in neighboring bystander cells should also be monitored. The present work describes a method that meets these criteria and serves as an example of the scientific potential of such approaches.

Materials and Methods

Cultures

AG01522 primary human skin fibroblasts were obtained from the Coriell Cell Repository (Camden, NJ, USA), expanded to passage 10, and stored in liquid nitrogen. Cells were routinely cultured in Cellgro® minimum essential medium (MEM; Mediatech, Herndon, VA, USA) supplemented with uninactivated 12.5% v/v fetal bovine serum (FBS; Nova-Tech, Grand Island, NE, USA), 100 U/mL penicillin, 100 ng/mL streptomycin (Mediatech), and 2 mM L-glutamine (Invitrogen, Carlsbad, CA, USA). We refer to this complete medium preparation here as cMEM. For experiments, cells were grown as three-dimensional (3-D) cultures on Cytomatrix® carbon units (Cytomatrix, Chelmsford, MA, USA) and maintained in humidified air supplemented with 3% CO₂ (8). cMEM was replaced on alternate days. Under these 3-D culture conditions, the distribution throughout the phases of the cell cycle was approximately G0/G1: G2/M:S = 90%:5%:5%, the length of S phase was 6.0±0.5 h, and that of G2/M phase was 4.5±0.5h (flow cytometry determinations). G0/G1 phase length was not determined directly.

Cell Targeting and Dosimetry

Solutions of 10 µM BrdU (Sigma-Aldrich, St. Louis, MO, USA) were prepared in cMEM. 2′-deoxycytidine [5-³H(N); Moravek Biochemicals, Brea, CA, USA] was obtained at specific activity 855 GBq/mmol and diluted in cMEM to 185 kBq/mL. To radiolabel and thereby irradiate a small fraction of cells in the culture, the cultures were co-pulse-labeled for 3 h with ³H-deoxycytidine (³H-dC) and bromodeoxyuridine (BrdU). Cultures were then washed six times with cMEM, harvested from the Cytomatrix using 0.25% w/v trypsin/EDTA (Mediatech), seeded at low density in culture dishes containing cMEM with 1 µM iododeoxyurdine (IdU; Fluka, Sigma-Aldrich), and harvested at various times for assessment of cumulative labeling index (CLI) in unlabeled bystander cells as previously described (8).

After washing cells free of extracellular radioactivity, a known number of cells were subjected to liquid scintillation counting for determination of average cellular activity. Since only a minor fraction of the cultured cells contains ³H-dC, the measured average millibecquerel per cell was converted to average millibecquerel per labeled cell, upon dividing the former by the fraction of BrdU-positive cells, as estimated by flow cytometry on parallel cultures. Assuming spherical cell nuclei with a 3.5-µm radius (9), average cellular radiation dose rates were calculated from cellular activity measurements using the cellular S value 4.09×10⁻³ Gy Bq s⁻¹ (10).

To evaluate biological response of individual cells as function of radiation dose, the latter must be inferred from the intensity of BrdU immunofluorescence on a cell-by-cell basis. The pyrimidine nucleosides deoxythymidine (usually referred to as thymidine), deoxycytidine, and the thymidine analogue BrdU are known to be similarly incorporated into the DNA during S phase of the cell cycle. BrdU immunofluorescence intensity and incorporated ³H-dC activity both increase linearly as a function of concentration in the cMEM. Thus, BrdU immunofluorescence intensity can be correlated with intra-cellular ³H-dC activity, and therefore to dose rate in individual cells. This was achieved by means of the relationship _N=aF+b, where _N is the dose rate to the cell nucleus (in cGy/h), F is the fluorescence intensity of a single cell that incorporated BrdU, a is a fitted parameter, and b is a negative offset to correct for autofluorescence and unspecific fluorochrome binding (zero radiation dose corresponds to a non-zero fluorescence intensity). Representative values are a=0.053 and b=−1.6.