In vivo imaging of graftversus-host disease and graftversus-leukemia - Immune Biology of Allogeneic Hematopoietic Stem Cell Transplantation - page 68

Fluorescence imaging of GVHD

Several significant advances in our understanding of GVHD pathogenesis
have resulted from both whole body and intravital fluorescence imaging. In
a study by Panoskaltsis-Mortari et al., a real-time spatial and temporal anal-
ysis of GVHD progression was conducted with eGFP cells from transgenic
mice with a fluorescence stereomicroscope ( Figure 4.1 B) [86] . While they
found the same general GVHD progression as in other studies [23] (i.e. infil-
tration to lymphoid tissues in the first 1-3 days, followed by migration to
target tissues; see the section “Bioluminescence imaging” above), they also
made several novel and interesting findings. First, they found that donor
splenocytes homing to lymph nodes shortly after infusion did not require
allogeneic stimulation or irradiation-induced injury. Previously, it was com-
monly thought that recipient conditioning (irradiation) created a “space”
in the lymphoid organs which could then be filled by the donor cells; it was
also believed that the release of inflammatory mediators after irradiation
would facilitate this process. However, the authors observed that trafficking
of immune cells and extravasation into lymph nodes occurred at similar
rates and quantities in non-irradiated mice in both allogeneic and synge-
neic environments. These data agree with the data from Miller et al., which
demonstrated that initial extravasation of T cells after infusion occurs ran-
domly [87] . While the only difference they found between allogeneic and
syngeneic splenocytes at 24 hours was increased allogeneic splenocytes
in the spleen, at later time points, allogeneic splenocytes demonstrated
far more proliferation and target tissue infiltration than their syngeneic
counterparts. By day 7 after transplantation, eGFP Tg cells were observed
in GVHD targets (i.e. liver, gut, skin and thymus) and could be visualized
through the skin. Additionally, several non-classical organs were targeted
including connective tissues, brain, kidney, tongue, gums and nasal cavi-
ties. These were surprising findings as to the extent of tissues which may be
targets of GVHD, some of which had previously been unrecognized.

69

Through the use of intravital microscopy, PI 3 Kγ expression on leukocytes
was found to be required for development of GVHD [88] . PI 3 Kγ has a crucial
role in leukocyte adhesion and rolling on microvasculature and leukocytes
which lack PI 3 Kγ and appear to have reduced adhesion and rolling (reviewed
in Castor et al. [88] ). Through intravital microscopy, it was observed that
cells derived from PI 3 Kγ −/− mice displayed reduced rolling and adhesion in
the gut microvasculature which was associated with reduced T-cell, DC and
macrophage infiltration, reduced injury to the intestine and reduced bacte-
rial translocation. The net result of transplanting PI 3 Kγ −/− splenocytes was a
significant decrease in the cytokines TNF-α and IFN-γ, a decrease in GVHD
score, and 100% survival of mice past day 60. Conversely, 100% mortality by
day 47 was observed in mice receiving WT splenocytes. The authors then
confirmed that these differences in GVHD were a result of reduced adhesion
and rolling, and not other changes, by evaluating the activation of CD4 + ,
CD8 + and B220 + cells 10 days after GVHD induction; no significant changes
were observed. Finally, it was demonstrated via use of a GFP-positive mas-
tocytoma line that the lack of PI 3 Kγ on donor cells did not interfere with the
GVT effect of the donor splenocytes. These results indicate that strategies
to block the action of PI 3 Kγ may have potential in regulating GVHD while
preserving GVL.