Evaluation of TgH(CX3CR1-EGFP) mice implanted with mCherry-GL261 cells as an in vivo model for morphometrical analysis of glioma-microglia interaction

Abstract

Background: Glioblastoma multiforme is the most aggressive brain tumor. Microglia are prominent cells within glioma tissue and play important roles in tumor biology. This work presents an animal model designed for the study of microglial cell morphology in situ during gliomagenesis. It also allows a quantitative morphometrical analysis of microglial cells during their activation by glioma cells.

Methods: The animal model associates the following cell types: 1- mCherry red fluorescent GL261 glioma cells and; 2- EGFP fluorescent microglia, present in the TgH(CX3CR1-EGFP) mouse line. First, mCherry-GL261 glioma cells were implanted in the brain cortex of TgH(CX3CR1-EGFP) mice. Epifluorescence - and confocal laser-scanning microscopy were employed for analysis of fixed tissue sections, whereas two-photon laser-scanning microscopy (2P-LSM) was used to track tumor cells and microglia in the brain of living animals.

Results: Implanted mCherry-GL261 cells successfully developed brain tumors. They mimic the aggressive behavior found in human disease, with a rapid increase in size and the presence of secondary tumors apart from the injection site. As tumor grows, mCherry-GL261 cells progressively lost their original shape, adopting a heterogeneous and diffuse morphology at 14-18 d. Soma size increased from 10-52 μm. At this point, we focused on the kinetics of microglial access to glioma tissues. 2P-LSM revealed an intense microgliosis in brain areas already shortly after tumor implantation, i.e. at 30 min. By confocal microscopy, we found clusters of microglial cells around the tumor mass in the first 3 days. Then cells infiltrated the tumor area, where they remained during all the time points studied, from 6-18 days. Microglia in contact with glioma cells also present changes in cell morphology, from a ramified to an amoeboid shape. Cell bodies enlarged from 366 ± 0.0 μm(2), in quiescent microglia, to 1310 ± 146.0 μm(2), and the cell processes became shortened.

Conclusions: The GL261/CX3CR1 mouse model reported here is a valuable tool for imaging of microglial cells during glioma growth, either in fixed tissue sections or living animals. Remarkable advantages are the use of immunocompetent animals and the simplified imaging method without the need of immunohistochemical procedures.

Fig. 1

Epifluorescence imaging (a) 3, (b) 6 and (c) 18 dpi reveals tumor growth and metastasis. Coronal sections of a TgH(CX3CR1-EGFP) mouse brain implanted with mCherry-GL261 cells. a and d Glioma cells infiltrated the cortex and surrounding areas; red arrow (a) points to a secondary tumor in the left ventricle. Injection site three dpi of mCherry-GL261 cells, showing an intense microgliosis around the tumor tissue. b and e red arrows (b) point to secondary tumors in the right ventricle. Microglial infiltration in tumor core 6 days after injection, while microgliosis remained present in tumor border. c and f An enlarged and diffuse mass of glioma cells developed 18 dpi, including metastasis in the hypothalamus (red outlined arrow, c). Epifluorescence imaging in (a, b and c), confocal microscopy of the region indicated by white boxes in (a–c) in (d, e and f). Scale bars: (a–c), 500 μm; (d–f), 100 μm

Fig. 2

Analysis of microglial cells shows morphological changes depending on their location towards the tumor tissue. a Coronal section of a TgH(CX3CR1-EGFP) mouse brain implanted with mCherry-GL261 cells at 14 dpi. White boxes indicate the following regions in analysis: 1 − contralateral hemisphere (control non-implanted site); 2 − tumor border and; 3 − tumor core. b Quantitative analysis of microglia cell numbers in regions 1–3, showing an increased number of microglial cells in tumor regions 2 and 3 (border and core, respectively) in comparison with the control region 1 (p < 0,05). In addition the number increase also in the core compared to the border tumor region. c–e Higher magnification of the brain areas 1–3 (white squares) in (a); maximum intensity projections acquired by confocal microscopy. f–h Magnified views of single cells (white boxes in c–e) to show details in cell morphology. In the presence of glioma cells, microglial morphology changes from a ramified (c and f) to an amoeboid shape in the tumors border (d and g) and core (e and h). Scale bars: (a), 500 μm; (c–e), 20 μm; (f–h), 10 μm

Fig. 3

Quantitative morphometrical analysis of microglial activation during glioma growth reveals regional differences between core, border and control regions. Images from cortical regions implanted with glioma cells at 14 dpi. a Endogenous EGFP fluorescence in microglial cells (red arrowheads) in the tumor core in relation to (b) GL261 cells labeled with mCherry (white circle) were immunopositive for the microglial/macrophage marker Iba1 (c, red arrowheads). d–f) EGFP channel of representative slices showing morphological changes in microglia, from the non-inoculated region 1 (d) to the tumor regions 2 (e) and 3 (f). In (d), a set of parameters for morphometrical analysis of individual microglial cells is represented as follows. Blue filled in red outline: mean intensity value of Iba1 expression (g); End points, the ImageJ plugin processed Skeleton tags of all pixel/voxels in a skeleton image. All junctions were classified in different categories depending on their 26 neighbors. When they had less than 2 neighbors, they were counted as end-points voxels (h); Red contour: soma size measured in μm² (I); yellow contour: total perimeter length in μm (j). Data were expressed as mean ± SEM. (*p < 0.05 statistical significance, ANOVA one way followed by Tukey’s test). Scale bar: 20 μm

Fig. 4

Observation of microglial and tumor cells reveals morphological changes in the core during gliomagenesis. a–r Time-course of glioma growth in mouse brain implanted with mCherry-GL261 cells 3, 6, 9, 12, 14 and 18 days after mCherry-GL261 injection in adult TgH(CX3CR1-EGFP) mice shown as maximum intensity projections of the tumor core. Channels are represented by endogenous fluorescence of EGFP in microglia (left column), mCherry in GL261 glioma cells (middle column) and merge (right column). As shown in K and L (red arrow), a polykaryocyte cell appeared at 12 dpi. Scale bar: 20 μm

Fig. 5

Two-photon laser-scanning microscopy (2P-LSM) reveals cellular responses of microglia to tumor injections in vivo in TgH(CX3CR1-EGFP) mice. a–c Microgliosis around the tumor injection site (circle), 30 min after implantation. White triangles point to mCherry-GL261 cells infiltrating the brain cortex. d–f At 36 h, microglial cells presented either a typical amoeboid shape (orange squares, d) or a ramified morphology, with a small cell body and long processes (white squares, d). mCherry-GL261 cells were present in capillary borders, suggesting a route for cancer cell spreading (white arrows, e and f). g–i Microglia remained infiltrated and in close contact with tumor cells, suggesting a cell-cell communication during tumor growth. Scale bars: (a–f), 50 μm; (g–i), 20 μm