Biology Reference
In-Depth Information
in MHC-mismatched models, and aly/aly hosts in an MHC-matched model
[79]
. GVHD was intact, though in some cases reduced, in nonsplenecto-
mized aly/aly or LTα-deficient hosts
[79-81]
. Likewise GVHD was similar
or even increased in the setting of splenectomy alone, and of course it has
long been known that splenectomized humans can develop GVHD. Overall
these data suggest that spleen and LN are efficient sites of alloreactive T-cell
activation, but that T-cell activation may occur at other sites, perhaps in
bone marrow
[80]
. Consistent with this, residual GVHD in aly/aly splenec-
tomized mice was found to depend on intact hematopoietic cells
[79]
.
Priming independent of hematopoietic APCs
(revisited)
Recent studies by Toubai et al.
[82]
and Koyama et al.
[83]
revisited the pos-
sibility that nonhematopoietic host cells can initiate GVHD
[58]
when prim-
ing by hematopoietic cells is not possible. Toubai et al. induced GVHD with
CD8
+
Matahari TCR-transgenic CD8
+
T cells, which recognize a male-spe-
cific peptide derived from the Uty gene product
[84]
. Female B6 β2M
−/−
→
B6 male BM chimeras were reirradiated and transplanted with female B6
β2M
−/−
BM and Matahari T cells. GVHD occurred in these recipients but
was less severe than in mice in which hematopoietic cells of either the host
or the donor were competent to present host male antigens. Furthermore,
lethal GVHD in β2M
−/−
→ B6 recipients requires 10-fold more donor T cells
than in recipients with functional hematopoietic APCs. Koyama et al. used
models in which GVHD was mediated by TCR-transgenic or polyclonal
CD4
+
T cells in MHC-matched, miHA-mismatched or MHC-mismatched
strain pairings. They also took a BM chimera approach to create transplant
recipients wherein host hematopoietic APCs were replaced with APCs inca-
pable of presenting allogeneic MHCII:peptide complexes, combined with
MHCII-deficient donor BM. In sum, in all models tested, GVHD could be
induced in these chimeric recipients. In both studies, as was the case for
the aforementioned experiments by Kosaka, Sprent, and co-workers
[58]
,
GVHD could have been initiated by a small number of residual host hema-
topoietic cells or by nonhematopoietic cells. Koyama et al.
[83]
excluded a
role for residual recipient Langerhans cells, which have been shown to be
sufficient for skin GVHD in an MHC-mismatched model
[85]
.
182
The results from these elegant studies must be reconciled with prior data
that touch on the roles of hematopoietic versus nonhematopoietic APCs.
In polyclonal CD8-dependent models of GVHD and GVL, using a similar
bone marrow chimera approach, host hematopoietic APCs were shown to
be crucial
[68,69]
. Alloreactive NK cells mitigated GVHD in a CD4-dominant
MHC-mismatched model and this effect was abrogated when hematopoi-
etic cells were specifically resistant to NK-cell-mediated killing
[65]
. The
elimination of residual host Langerhans cells in donor → host chimeras
diminished cutaneous GVHD
[85]
. CD4-mediated GVHD has also been at
least partially dependent on CD28:CD80/CD86 costimulation, so nonhe-
matopoietic cells must express sufficient CD80/CD86 assuming this signal
is given
in cis
[71]
. Another question is how disease spreads if it is initiated
by only a few alloreactive clones activated by nonhematopoietic cells in tar-
get tissues such as skin and bowel. This would seem to require egress of
Search WWH ::
Custom Search