To evaluate the requirement for CD8(+) T cells in kidney transplant re
jection, we studied class I-deficient (class I-) mice that had receive
d vascularized renal allografts, Because of the absence of MHC class I
expression, these mice are grossly deficient in CD4(-)CDS(+)alpha bet
a TCR(+) cells. Despite the deficiency of CD8(+) T cells in naive clas
s I- mice, kidney allografts transplanted into class I- recipients dev
eloped significant reductions in renal blood flow and glomerular filtr
ation rate to levels comparable to allograft controls. This functional
deterioration was associated with histologic changes consistent with
cellular rejection. There were no significant differences in the patte
rn, severity, or phenotypic character of the cellular infiltrate in al
lografts transplanted into class I- recipients compared to controls. I
n fact, substantial numbers of CD8(+) T cells were present in these al
lografts, and the intensity and pattern of anti-CD8 staining was not d
ifferent from controls. Virtually all of the CD8(+) cells in the kidne
y grafts were class I- and CD4(-) and co-expressed CD8 alpha and beta
chains; the majority were alpha beta TCR(+). The CD8(+) infiltrating c
ells were cytotoxic to donor targets but also exhibited activity again
st class I+ cells bearing self-MHC. Despite the marked CD8(+) T cell i
nfiltration of grafts, CD8(+) T cells could not be detected by flow cy
tometry in freshly isolated splenocytes from the class I- recipients o
f allografts. High levels of circulating anti-class I antibodies were
present in the serum of class I- recipients of kidney allografts, and
these antibodies had unusual specificity in that they appeared to reco
gnize framework epitopes of MHC class I. Thus, class I- mice readily r
eject kidney allografts. Although the number of CD8(+) alloreactive pr
ecursors is substantially reduced in class mice, and their specificiti
es are atypical, the pattern and character of the intra-graft CD8(+) c
ellular response is not significantly altered. Thus, factors unrelated
to precursor frequency determine the dimension of the intra-graft CD8
(+) response. Such factors might include cellular and/or biochemical p
roperties of microenvironment within the graft.