The Blood-Brain Barrier at Sites of Metastasis
The Blood-Brain Barrier at Sites of Metastasis
Animal procedures were carried out in accordance with the UK Animals (Scientific Procedures) Act 1986 and with local ethical committee approval. Female BALB/c mice (aged 6–8 weeks; 19.1g (±1.6); Charles River, Margate, UK) were injected in the left ventricle of the heart with 10000 4T1–green fluorescent protein (GFP) murine mammary carcinoma cells in 100 μL of phosphate-buffered saline using ultrasound guidance (VEVO 770; Visualsonics, Amsterdam, The Netherlands) under general anesthesia (1.5%–3.0% isoflurane in oxygen).
Brain tissue sections from mice injected with 4T1-GFP cells were assessed immunohistochemically for TNFR1 and TNFR2 expression, and colocalization with glucose transporter 1 (Glut-1) was determined using immunofluorescence. Six cases of human brain metastasis and one control case of noncancerous neuroinflammatory disease were examined immunohistochemically for TNFR1 and TNFR2 expression. See the Supplementary Methods (available online) for details of immunohistochemistry and immunofluorescence.
At 13 days postmetastasis induction in the 4T1-GFP model, mice were intravenously injected with 100 μL saline containing doses from 1 to 5 μg per mouse of either recombinant mouse TNF (Peprotec, London, UK) or recombinant mouse lymphotoxin (R&D Systems, Oxford, UK) or saline as control. Two to 24 hours later, mice were intravenously injected with of 100 μL of type II horseradish peroxidase (44kDa) (HRP; 300 units; SigmaAldrich, Dorset, UK) and killed by transcardial perfusion fixation. Brains were excised and prepared for histology. Slides were stained using a modified Hanker–Yates method and counterstained using cresyl violet. Brown staining indicating permeation of HRP was assessed blindly, and a percentage of positive metastases for each group was calculated. See Supplementary Methods (available online) for details of full experimental protocol and histology.
BALB/c mice injected with 4T1-GFP cells (n = 3 per group) were anesthetized and positioned in a quadrature birdcage coil. Respiration was monitored throughout, and body temperature was maintained at 37°C. MRI data were acquired using a 7 T horizontal-bore magnet with a Varian Inova spectrometer (Varian, Santa Clara, CA). Before and 2 hours after the 3-μg intravenous injection of TNF or LT, a set of 10 serial T1-weighted images were acquired using a spin-echo sequence both before and 5 minutes after intravenous injection of 30 μL of gadolinium-diethylenetriaminepentaacetic acid (Gd-DTPA, 590Da; Omniscan, GE Healthcare, Little Chalfont, UK) to assess BBB permeability. Regions of interest (ROIs) containing metastases (confirmed histologically) and contralateral same-slice nontumor ROIs were segmented using VnmrJ (Varian) on post-Gd images. The mean signal intensity values of the ROIs were quantitated before and after LT or TNF administration (ROI n = 14 total). See Supplementary Methods (available online) for details of sequence parameters.
Radiolabelled trastuzumab (In-BnDTPA-Tz, 145kDa) was synthesized as previously described. Mice injected with 4T1-GFP cells (n = 3 per group) were injected intravenously with 100 μL of saline containing 3 μg of TNF or LT or no cytokine together with 3.7 μg In-BnDTPA-Tz, 2 hours before single photon emission computed tomography (SPECT) under anesthesia. SPECT/computed tomography (CT) was performed using a nanoSPECT/CT scanner (Bioscan, Washington, DC). Data were reconstructed using InVivoScope (Bioscan) and analyzed using Inveon Research Viewer (Siemens, Erlangen, Germany) and normalized against a muscle ROI to compensate for intersubject variability in radioactivity dose. See Supplementary Methods (available online) for details.
Female SCID mice (aged 6–8 weeks) were intracardially injected (as above) with 10000 MDA231BR-GFP cells (subclone of a metastatic human breast carcinoma that preferentially metastasizes to the brain). Mice (n = 3) were intravenously injected with saline or 3 μg of TNF and perfusion-fixed 21 days after cell injection. This later time point was chosen because these tumors are slower growing than the 4T1-GFP metastases. Tissue sections were stained for TNFR1 and TNFR2 (as above).
BBB breakdown frequency in metastases was analyzed with one-way analysis of variance (ANOVA) followed by a Dunnet post hoc test or a Tukey post hoc test. The difference in ratio of signal intensity at sites shown to contain metastases vs equivalent regions in the contralateral hemisphere was assessed with the Wilcoxon signed rank test. Intracerebral radioactivity differences were analyzed with the Kruskal–Wallis test followed by a Dunn multiple comparison test. All statistical tests were two-tailed and considered statistically significant if P was less than or equal to 0.05.
Methods
Brain Metastasis Model
Animal procedures were carried out in accordance with the UK Animals (Scientific Procedures) Act 1986 and with local ethical committee approval. Female BALB/c mice (aged 6–8 weeks; 19.1g (±1.6); Charles River, Margate, UK) were injected in the left ventricle of the heart with 10000 4T1–green fluorescent protein (GFP) murine mammary carcinoma cells in 100 μL of phosphate-buffered saline using ultrasound guidance (VEVO 770; Visualsonics, Amsterdam, The Netherlands) under general anesthesia (1.5%–3.0% isoflurane in oxygen).
Immunohistochemistry for TNF Receptors
Brain tissue sections from mice injected with 4T1-GFP cells were assessed immunohistochemically for TNFR1 and TNFR2 expression, and colocalization with glucose transporter 1 (Glut-1) was determined using immunofluorescence. Six cases of human brain metastasis and one control case of noncancerous neuroinflammatory disease were examined immunohistochemically for TNFR1 and TNFR2 expression. See the Supplementary Methods (available online) for details of immunohistochemistry and immunofluorescence.
Histological Assessment of BBB Permeabilization by Cytokine Administration
At 13 days postmetastasis induction in the 4T1-GFP model, mice were intravenously injected with 100 μL saline containing doses from 1 to 5 μg per mouse of either recombinant mouse TNF (Peprotec, London, UK) or recombinant mouse lymphotoxin (R&D Systems, Oxford, UK) or saline as control. Two to 24 hours later, mice were intravenously injected with of 100 μL of type II horseradish peroxidase (44kDa) (HRP; 300 units; SigmaAldrich, Dorset, UK) and killed by transcardial perfusion fixation. Brains were excised and prepared for histology. Slides were stained using a modified Hanker–Yates method and counterstained using cresyl violet. Brown staining indicating permeation of HRP was assessed blindly, and a percentage of positive metastases for each group was calculated. See Supplementary Methods (available online) for details of full experimental protocol and histology.
In vivo Assessment of BBB Permeability: MRI
BALB/c mice injected with 4T1-GFP cells (n = 3 per group) were anesthetized and positioned in a quadrature birdcage coil. Respiration was monitored throughout, and body temperature was maintained at 37°C. MRI data were acquired using a 7 T horizontal-bore magnet with a Varian Inova spectrometer (Varian, Santa Clara, CA). Before and 2 hours after the 3-μg intravenous injection of TNF or LT, a set of 10 serial T1-weighted images were acquired using a spin-echo sequence both before and 5 minutes after intravenous injection of 30 μL of gadolinium-diethylenetriaminepentaacetic acid (Gd-DTPA, 590Da; Omniscan, GE Healthcare, Little Chalfont, UK) to assess BBB permeability. Regions of interest (ROIs) containing metastases (confirmed histologically) and contralateral same-slice nontumor ROIs were segmented using VnmrJ (Varian) on post-Gd images. The mean signal intensity values of the ROIs were quantitated before and after LT or TNF administration (ROI n = 14 total). See Supplementary Methods (available online) for details of sequence parameters.
In vivo Assessment of BBB Permeability: Single Photon Emission Computed Tomography/Computed Tomography
Radiolabelled trastuzumab (In-BnDTPA-Tz, 145kDa) was synthesized as previously described. Mice injected with 4T1-GFP cells (n = 3 per group) were injected intravenously with 100 μL of saline containing 3 μg of TNF or LT or no cytokine together with 3.7 μg In-BnDTPA-Tz, 2 hours before single photon emission computed tomography (SPECT) under anesthesia. SPECT/computed tomography (CT) was performed using a nanoSPECT/CT scanner (Bioscan, Washington, DC). Data were reconstructed using InVivoScope (Bioscan) and analyzed using Inveon Research Viewer (Siemens, Erlangen, Germany) and normalized against a muscle ROI to compensate for intersubject variability in radioactivity dose. See Supplementary Methods (available online) for details.
Validation in Human Breast Carcinoma Brain Metastasis Model
Female SCID mice (aged 6–8 weeks) were intracardially injected (as above) with 10000 MDA231BR-GFP cells (subclone of a metastatic human breast carcinoma that preferentially metastasizes to the brain). Mice (n = 3) were intravenously injected with saline or 3 μg of TNF and perfusion-fixed 21 days after cell injection. This later time point was chosen because these tumors are slower growing than the 4T1-GFP metastases. Tissue sections were stained for TNFR1 and TNFR2 (as above).
Statistical Analyses
BBB breakdown frequency in metastases was analyzed with one-way analysis of variance (ANOVA) followed by a Dunnet post hoc test or a Tukey post hoc test. The difference in ratio of signal intensity at sites shown to contain metastases vs equivalent regions in the contralateral hemisphere was assessed with the Wilcoxon signed rank test. Intracerebral radioactivity differences were analyzed with the Kruskal–Wallis test followed by a Dunn multiple comparison test. All statistical tests were two-tailed and considered statistically significant if P was less than or equal to 0.05.
Source...