Figure 1.
Bone Marrow Derived Cells Migrate Specifically to the Site of Cranial irradiation.
(A) H&E staining of a normal brain 7 days following 3*2Gy radiation illustrates the location from where immunofluorescence images in (B) were obtained, 2× magnification. Black arrowhead shows the positioning of the intracranial window (ICW) and so the radiation path. (B) The right irradiated hemisphere demonstrates a trajectory of bone marrow derived cell (BMDC) recruitment which follows a cranial-caudal direction, with maximal recruitment at the cortex gradually decreasing beneath the cortical surface, compared to the contralateral-left, non-irradiated hemisphere which shows no BMDC recruitment (Green: BMDC, blue: nuclei), 10× magnification. (C) 2 photon laser microscopy in-vivo images 7 days post 3*2Gy radiation, illustrate distinct BMDC recruitment to the site of cranial radiation, whilst non-irradiated control brains demonstrate minimal incorporation of BMDCs (Green: BMDC, red: blank, blue: CD31-APC), 5× magnification. (D) To confirm that recruitment of BMDCs visualized on 2 photon laser microscopy is not as a consequence of green auto-fluorescent signals, red fluorescent chimeric mice were used to demonstrate a similar pattern of BMDC recruitment following 3*2Gy radiation but demonstrate a negative green channel (Green: blank, red: BMDC, blue: CD31-APC), 10× magnification. (E) Quantification of BMDCs visualized per field of tissue volume for 2PLM, blue bars, and traditional histology, black line. Note, the 10-fold difference in the scale bars highlighting the statistically significant difference in sensitivity of the 2 methods. This figure highlights the distinct advantage of 2 photon laser microscopy in-vivo imaging over traditional 5 µm cross-sectional histological analysis, clearly allowing visualization of higher levels of BMDCs in the same region of the brain, comparing (B) to (C) which when quantified in context of tissue volume (E) shows 2PLM is 10-fold more sensitive than traditional histology.
Figure 2.
Bone Marrow Derived Cell Recruitment is Radiation Dose-Dependent.
(A,B) Increasing levels of BMDC recruitment can be observed with increasing radiation doses on 2 photon laser microscopy in-vivo imaging 7 days post radiation (Green: BMDC, blue: CD31-APC), 10× magnification. (A) Single dose increase from 2Gy to 15Gy, (B) fractionated doses increase from 3*2Gy to 3*5Gy, all demonstrate an increase in BMDC recruitment. (C) Bar graph representing quantification of the extent of BMDCs recruited to site of cranial radiation 7 days post, demonstrating a statistically significant increase in BMDC recruitment between 2Gy to 6Gy and 2Gy to 15Gy, both p<0.0001***. Fractionated radiation demonstrates lower rate of BMDC recruitment when compared to its single fraction equivalent, perhaps explained by radiobiology principle of repair that occurs with fractionated radiation. The increase in BMDC recruitment is noted at all time points post radiation (1 hour, 1 day, 21 day).
Figure 3.
Longitudinal Pattern of Bone Marrow Derived Cells Recruitment to Cranial Irradiation.
(A,B) The longitudinal fate and pattern of migration of BMDCs recruited to the site of cranial radiation is distinct from that seen in response to intracranial trauma. Using intra-vital EC staining antibody (CD31) it is clear that BMDCs do not contribute to vascular endothelial cells. (Green: BMDC, blue: CD31-APC), 5× magnification, unless stated. (A) BMDCs recruited to the site of 3*2Gy radiation in normal brain persist up to 1 month post radiation therapy, with a steady increase in extent of BMDCs present from day 1 up to day 21 as seen in the 2 photon laser microscopy images. BMDCs do not migrate or invade into the surrounding brain parenchyma with time, but stay within the radiation trajectory. (B) In control mice that received no radiation but received sham needle injection, there is evidence of BMDC recruitment to the site of injury within 24 hours of the sham injection but it rapidly diminishes before day 7 and is completely lost by day 21. (C) Quantification of the average BMDCs in high powered fields demonstrates the difference in patterns between RT and blunt traume. (D,E) AMD3100, an inhibitor of the Stromal-cell Derived Factor1/CXCR4 pathway, did not inhibit BMDC recruitment post radiation as shown in the 2 photon laser microscopy images (Green: BMDC, blue: CD31-APC), 5× magnification. (D) AMD3100 was administered concurrently with radiation and BMDC recruitment was unaffected from day 3 to day 10 post radiation. (E) Pretreatment with AMD3100 prior to radiation also did not show any significant difference in BMDC recruitment.
Figure 4.
Microvascular Alterations Post Radiation.
Immunohistochemical co-staining of CD31 and TUNEL confirms endothelial cell apoptosis occurs as early as 1 hour post radiation at site of cranial radiation. (A) Increasing radiation doses from 2Gy to 15Gy increases endothelial cell apoptosis as seen in immunohistochemistry sections (Brown: CD31, Pink: TUNEL), 40× magnification. (B) Graphical representation of endothelial apoptosis, CD31+ TUNEL+, illustrates that 15Gy significantly induces the most endothelial apoptosis when compared to 2Gy and 6Gy at 1 hour post radiation, (p = 0.0001***) and that endothelial cell apoptosis levels at 15Gy 1 hour post radiation are significantly increased compared to 1 day post radiation, (p = 0.0004**). (C) Quantification of apoptosis of total parenchymal cells shows that 15Gy significantly induces most apoptosis (p = 0.0075**) at 1 day post radiation, however significantly more apoptosis is induced 1 day post radiation than at 1 hour post radiation (p = 0.0224*). Compared with what is seen with endothelial cell apoptosis it is clear that maximal endothelial cell apoptosis occurs earlier than overall parenchymal cell apoptosis. (D) Further analysis of vessels structure at the site of radiation, 7 days post 3*2Gy, demonstrates significant increases in both the density, p<0.0001***, and diameter of vessels, p<0.0001***, when compared to non-irradiated controls. (E) CD31 immunohistochemistry sections confirm the change in vessel density and diameter 7 days following 3*2Gy radiation when compared to non-irradiated control tissue, 10× magnification.
Figure 5.
Characterization of Bone Marrow Derived Cells.
(A–H) Immunofluorescence analysis of brain sections for characterization of cell types that BMDCs potentially differentiate to at the site of cranial radiation 7 days post 6Gy radiation, in the staining a red fluorescent chimeric mouse was used. (Green: stain, Red: BMDC, Blue: nuclei), 40× magnification (A,B) No differentiation of BMDCs to vessel endothelial cells is evident, confirming what was seen on the 2 photon laser microscopy in-vivo imaging. (A) BMDCs do not co-localize with the endothelial cell marker CD31, using confocal immunofluoroscence, or (B) immunohistochemical overlay of CD31 (Brown: CD31), 10× magnification. (C) Few BMDCs co-localize with the pericytic differentiation marker smooth muscle actin, shown by white arrowhead. (D,E) Approximately 40% of recruited BMDCs stain positively for inflammatory markers: (D) MAC3, co-localization shown by white arrowhead, (E) CD11b, co-localization shown by white arrowhead. (F) 50% of recruited BMDCs found in the parenchyma co-localize with the microglial differentiation marker IBA1 shown here by white arrowhead. (G) IBA1+ BMDCs are also found in the perivascular space around vessel wall, shown by white arrowhead. (H) There was no astrocytic differentiation or glial scaring, as determined by the lack of GFAP co-localization.