PERIPHERAL TOLERANCE TO MAJOR HISTOCOMPATIBILITY COMPLEX ANTIGENS 
Richard A. Fiavell, Ph.D., Investigator 
Dr. Flavell's laboratory has used transgenic ani- 
mals to study immune function and immune toler- 
ance to the class II gene products of the murine 
major histocompatibility complex (MHC). Class II 
molecules of the MHC comprise two chains, called 
a and (3, which form la antigens and serve as a rec- 
ognition element for CD4^ T cells. In the mouse, 
both la antigens, I-A and I-E, are composed of an a- 
and a f^-chain. Strains of mice that lack I-E have 
served as a paradigm for studying I-E function in 
the mouse. 
I. Peripheral Tolerance to I-E. 
Considerable evidence suggests that tolerance to 
antigens on bone marrow-derived cells occurs pri- 
marily by deletion of reactive T cells during the de- 
velopment of those cells in the thymus. Mecha- 
nisms of tolerance to peripheral antigens, by 
contrast, have been largely obscure. If deletion of T 
cells is the mechanism, how then can the antigen 
expressed at a distal site from the thymus be trans- 
located to the thymus in sufficient quantity to cause 
deletion of those reactive cells? A number of alter- 
native hypotheses can be raised to explain toler- 
ance to peripheral antigens. To distinguish among 
these. Dr. Flavell's laboratory directed the expres- 
sion of the class II I-E protein to the P islet cells and 
the acinar cells of the pancreas, using the insulin 
and elastase promoters, respectively. Mice transge- 
nic for these constructs express these genes, and 
hence the I-E protein, in a tissue-specific manner; 
the resultant mice are tolerant to the I-E antigen in 
both cases. In collaboration with Drs. Ralph Brins- 
ter (University of Pennsylvania) and Richard Palmi- 
ter (HHMI, University of Washington), Dr. Fiavell 
and his co-workers have attempted to understand 
the mechanism of tolerance to the antigens ex- 
pressed. 
Tolerance does not seem to be a result of de- 
letion of I-E-reactive T cells, since both vpi7a* and 
Vp5^ T cells (both of which are deleted in con- 
trol I-E+ mice) are present at normal frequency 
in these transgenic mice. To test whether the po- 
tentially I-E-reactive T cells in these mice can 
respond to antigen. Dr. Fiavell and his co-workers 
stimulated these cells by crosslinking the T cell 
receptors (TCRs), using antibodies specific for I-E- 
reactive T cells (Vpi7a and Vp5). This procedure 
induces T cell proliferation in control SJL mice 
but not in insulin I-E transgenic mice. T cell prolif- 
eration of these populations was not observed in 
several experiments, suggesting that this popula- 
tion is in a state of paralysis. This clonal paralysis 
appears to be specific for the I-E-reactive T cell 
population and not general immune suppression. 
Thus the total T cell population of the transgenic 
mice is responsive to crosslinking of the receptors 
using anti-CD3, which is present on all T cells. 
Crosslinking of the receptors of the V(38 T cell pop- 
ulation, which is not I-E reactive, induces T cell 
proliferation in both transgenic and nontransgenic 
mice. 
These initial indications suggest that tolerance to 
the I-E molecule expressed in the periphery is the 
result of clonal paralysis of T cells. Similar phenom- 
ena of clonal paralysis have been observed pre- 
viously in vitro by several groups, notably by Dr. 
Ronald Schwartz and his colleagues. These data are 
interpreted to mean that normal T cell activation 
requires that the antigen MHC complex be recog- 
nized by the TCR in conjunction with a second sig- 
nal to be delivered by the antigen-presenting cell. 
Perhaps the clonal paralysis of I-E-reactive T cells 
results from recognition of the I-E protein on (3 
cells of the pancreas, which are incapable of pro- 
viding that second signal. It is not clear what the 
defect is in paralyzed T cells, nor what that second 
signal is. 
Tolerance in the elastase I-E mice seems to have 
some parallels to that described above but is not 
identical. As stated above, I-E-reactive cells appear 
not to be deleted, there is only partial clonal paral- 
ysis of the vp5 population, and there is little if any 
paralysis of the vpi7 population. When nontoler- 
ant T cells were adoptively transferred into irradi- 
ated or T cell-depleted elastase transgenic mice, 
the acinar tissue was rapidly destroyed, suggesting 
that the I-E molecule on the acinar tissue is clearly 
antigenic. However, when nontolerant T lympho- 
cytes were transferred into nonirradiated transgenic 
mice, this destruction did not occur. Moreover, 
nontolerant T cells injected into otherwise un- 
manipulated elastase I-E mice can be primed in situ 
to destroy the acinar tissue by injection into the 
same animal of I-E -I- spleen cells that are them- 
selves capable of antigen presentation. Priming the 
elastase transgenic mice with I-E -I- spleen antigen- 
presenting cells does not induce an autoimmune 
response to the I-E -I- acinar cells. 
Continued 
193 
