Isolation of mouse stromal cells associated with a human tumor using differential diphtheria toxin sensitivity

Abstract

Tumor vascularization is accompanied by the migration of stromal cells, including endothelial cells, smooth muscle cells, and fibroblasts, into the tumor. The biological contributions of stromal cells to tumor vascularization have not been well-defined, partly due to the difficulty of culturing stromal cells in the presence of large numbers of fast-growing tumor cells. To address this problem, a strategy was devised to kill tumor cells but not stromal cells. Advantage was taken of the observation that diphtheria toxin (DT) kills human but not rodent cells. Human melanoma (MMAN) tumor cells were injected subcutaneously into nude mice. The tumors were excised, homogenized, and treated with 50 ng/ml DT for 24 hours. Elimination of melanoma cells by DT treatment was demonstrated by lack of detectable levels of microphthalmia, a transcription factor that is a marker for melanoma cells. The murine stromal cells were viable and found to be mostly smooth muscle cells. These cells constituted about 1.5% of the MMAN tumor. RNase protection assays using a specific murine vascular endothelial growth factor probe confirmed the murine origin of the stromal cells. This method allows rapid isolation of stromal cells and should facilitate biochemical and genetic analysis of tumor-stromal interactions.

Figure 1.

Parental mouse 32D myeloid cells grown in suspension (open circles ), and 32D cells transfected with human HB-EGF cDNA (closed circles ) were treated with increasing concentrations of DT and incorporation of [³H]-leucine into protein was measured. Each point was performed in triplicate and the results were normalized to the cpm of incorporated [³H]-leucine obtained in the absence of DT.

Figure 2.

DT killing of human melanoma cells in culture. MMAN melanoma cells were incubated with increasing amounts of DT. Protein synthesis was determined as in Figure 1 ▶ .

Figure 3.

DT treatment of a human melanoma ablates microphthalmia protein levels. Melanoma tumors were cultured in the absence (lane 1) or presence (lane 2) of DT. A mouse B16 melanoma cell lysate was used as a positive control (lane 3). Western blot analysis of the lysates was carried out with anti-microphthalmia and anti-tubulin antibodies. The upper doublet represents microphthalmia protein (bracket), whereas the lowest band represents tubulin, which was used as a control for loading.

Figure 4.

Morphology and immunohistochemistry of tumor stromal cells. A: An MMAN xenograft was treated with collagenase and cells were plated in tissue culture flasks. B: The same culture in A observed 2 days after treatment with 100 μg/ml DT. C: Immunohistochemical staining of calponin, a smooth muscle marker. D: Immunofluorescent staining with diI-Ac-LDL, a marker of endothelial cells. The stromal cells were confluent at the time of photography, and the LDL-positive cells comprise 1 to 4% of the total cells.

Figure 5.

RNase protection assay using a murine-specific probe for the three VEGF mRNA isoforms. RNA from the indicated sources was hybridized to ³²P-labeled riboprobes for murine VEGF and mammalian 18S. Samples were either untreated (−) (lane 1) or treated with RNases (+) (lanes 2–6) and electrophoresed in a denaturing acrylamide gel. The full-length ³²P VEGF and 18S probes are indicated by arrows at left. The protected fragments (VEGF₁₈₈, VEGF₁₆₄, and VEGF₁₂₀) resulting from hybridization to each of the VEGF mRNA isoforms are indicated by arrows at right. Lanes 1–2: incubation of ³²P-labeled VEGF with control tRNA in the absence or presence of RNase. Lanes 3–6: The protected fragments from mouse liver, an MMAN tumor, cultured MMAN cells, and stromal cells derived after DT treatment of an MMAN tumor, respectively. The protected fragment in mouse liver (lane 3 ) between VEGF₁₈₈ and VEGF₁₆₄ is an artifact that results from internal cleavage of the VEGF₁₈₈ hybrid at a Poly-A-rich region.