Together with long term lung colonization assays, the short term methodology we employed here to evaluate in vivo lung colonization by CTCs in distant organs clearly unveiled and differentiated the specific role of polyFN assembled on CTCs in colonizing the lungs, which then led to the extravasation and metastatic processes18,19,20,21. Although labeling cells with long term cell tracker CFSE allowed us to trace the intravenously injected CTCs for up to three days prior to the lung perfusion and retain sufficient green fluorescence for quantitative purposes, it was impossible to directly determine whether the tumor cells were located within the vessel lumen or had already extravasated from the blood vessels. To solve this problem, red fluorescent rhodamine-conjugated dextran which is able to non-specifically bind to lectins expressed on endothelia may be used to label the lung vasculature during lung perfusion28,29,30. Despite the fact that CFSE is a long term cell tracker, its fluorescence intensity is halved every cell division27, limiting the potential of using it in labeling techniques. To circumvent this problem, it should be ascertained that the labeling dosage of CFSE is sufficient for cells to remain detectable for the entire duration of in vivo experiments and that any treatment applied to the cells has no effect on cell proliferation. Because the comparison and quantification of the lung-colonizing shScr and shFN CTCs were done in separate mice, it might be argued that the differences between the two groups were due to individual variation. To exclude such variation, a mixture of two types of tumor cells labeled with distinct fluoresce dyes (e.g., CFSE/CM-diI) or stably transfected with GFP/dsRed may be concomitantly intravenously injected into the same animal in the lung colonization assay31,32,33. It is worth noting that cloning effects resulting from the cell cloning technology that is attempted to sort out stably fluorescent protein-transfected cell clones with sufficiently strong fluorescence34,35,36,37 may make the cloned cells unrepresentative to represent the entire heterogeneous populations38,39,40. If a stable transfection of any fluorescent protein into tumor cells is desired for the lung colonization assays, elevating the transfection efficiency of the tumor cells as a whole should better preserve the original characteristics than cloning the cell populations with sufficient fluorescence41,42.
The lung perfusion steps in the lung colonization assays are necessary in that some intravenously injected tumor cells, instead of specifically arresting to the lung capillary system, tend to be mechanically trapped and jammed within the capillary networks of lung tissues due to the averagely larger diameters of CTCs than the width of lung capillary lumens43,44. Quantification of the lung-colonizing tumor cells without perfusing the lungs may result in an overestimation by mistakenly counting the mechanically jammed cells45,46,47. Even if the lungs have been perfused, additional problems remain. For instance, it is still difficult to differentiate whether the lung-colonizing tumor cells are located on the lumenal endothelia or have already extravasated from the blood vessels. To resolve this issue, staining the blood vessels with Rhodamine-conjugated dextran as aforementioned may be useful30,48,49. Alternatively, if quantification of tumor extravasation is desired, calcium-chelator EDTA/EGTA or trypsin, a non-specific protease, can be used in the perfusion buffer to impede calcium-dependent and -independent tumor cell adhesion events50,51,52. Thus, the tumor cells remaining in the lung tissues may be considered as extravasated.
For quantification of the lung-colonizing tumor cells, the perfused lungs are often subjected to traditional confocal microscopy. However, the measurability of tissue depth is limited due in part to tissue autofluorescence and fluorescence scattering effects, rendering deeper tissues undetectable53,54. A two photon microscope with higher sensitivity than a traditional confocal microscope is suitable to resolve such problems in that the near-infrared radiation used in two-photon excitation with significantly less absorption by biological specimens than UV or blue-green light makes the technique more appropriate for imaging thick specimens55,56,57. Lung-colonizing tumor cells are counted by averaging tumor cell numbers in several representative confocal images, which do not include the entire number of lung-colonizing tumor cells in the whole lungs of an individual animal. To quantify the whole lungs in the lung colonization assays, tumor cells stably transfected with luciferase could be employed and the perfused lungs could then be subjected to IVIS imaging after intravenous inoculation of suspended tumor cells in the presence of luciferin46,58,59. Alternatively, the perfused lungs could be minced into pieces and subjected to enzymatic digestion with proteases, e.g., collagenase, releasing single cells into suspension60,61. The tumor cells with fluorescence dyes or stably transfected with fluorescence proteins could then be quantified with fluorescence-activated cell sorting (FACS) analysis56,62.