In Vivo Bioluminescence Imaging to Study the Contribution of TNF-TNFR Interactions on Immune and Parenchymal Cells to Tumor Cell Progression in a Syngenic Mouse Model

Abstract 1110

The cytokine tumor necrosis factor-α (TNF) has pleiotropic functions both in normal physiology and disease. TNF and its relative lymphotoxin-α (LT) signal by activating two cell surface receptors TNFR1 and TNFR2. TNFR1 is expressed on most cells whereas TNFR2 is mainly expressed in cells of the hematopoietic system. TNF-TNFR interactions were shown to play a major role in graft-versus-leukemia effect and in the immunosurveillance of solid tumors.

To study the contribution of TNF-TNFR interactions on tumor cell progression we employed a syngenic B16 melanoma mouse model combined with in vivo bioluminescence imaging. Firefly luciferase-transgenic B16 melanoma cells were injected intravenously into syngenic albino C57BL/6 hosts. The host mice were either of wildtype, TNF, LT, TNFR1, TNFR2 knockout or TNFR1R2 double knockout genotype. The localization and expansion of the B16 cells was monitored by in vivo bioluminescence imaging for up to 14 days. On days 15, mice were sacrificed and internal organs were imaged ex vivo to further elucidate the organ-specific tumor burden.

B16 tumors were primarily found in the lungs of all genotypes. All female knockout genotypes displayed a higher lung tumor burden than wildtype mice. In male mice, only TNF knockout presented enhanced tumor cell signals.

Following ex vivo imaging we evaluated the pulmonary infiltration of NK1.1 or NKp46, CD8, CD4 and CD4/CD25/Foxp3 regulatory T cells by flow cytometry and immunofluorescence microscopy. Compared to wildtype mice, more regulatory T cells infiltrated the lungs of female TNFR1 knockout mice (200%). In LT knockout mice, very few NK cells (<20%) but more CD4⁺ cells (160%) infiltrated the lungs. Only subtle changes occurred in the other deficient mouse strains. However, these changes were independent of the presence of tumor cells and could also be found in normal knockout mice without B16 tumors. Within sections of tumor-bearing lungs, we found that TNF and all three TNFR knockouts exhibited less CD8⁺ cells within tumors than did wildtype or LT knockout mice. The number of CD8⁺ cells in normal lung tissue was not altered across the different genotypes. The deficit in NK cells of LT knockout mice was confirmed by histology.

The enhanced tumor progression in all knockout mice could be a secondary effect due to their altered immune phenotype rather than to the loss of TNF-TNFR interactions. To circumvent this potential experimental bias and to further assess the influence of the loss of expression of parts of the TNF/TNFR-system in immune cells only, we generated bone marrow chimeras by reconstituting lethally irradiated female wildtype mice with bone marrow derived from TNF, LT, TNFR1 or TNFR2 knockout mice. Tumor cell signals in these chimeric mice progressed more than in normal wildtype mice. In contrast to the first set of experiments with knockout mice, we found that mice reconstituted with either TNF or TNFR2 knockout bone marrow presented less tumor cell signal than did mice reconstituted with wildtype bone marrow. TNF-TNFR interactions between immune cells appear to exhibit pro-tumorigenic functions in our mouse model.

These results show that TNF-TNFR interactions are an important step in tumor cell progression and that the outcome of these interactions differs, depending on whether immune or parenchymal cells are deficient in TNF-TNFR signalling components.