The Discrimination of Infectious Non-Self from 
Noninfectious Self 
Charles A. Janeway, Jr., M.D. — Investigator 
Dr. Janeway is also Professor of Immunobiology at Yale University School of Medicine. He is a graduate of 
Harvard College (B.A., chemistry) and of Harvard Medical School. He had research training during med- 
ical school at the National Institute for Medical Research in london, England; postdoctoral training in 
internal medicine at the Peter Bent Brigham Hospital in Boston; and immunology research experience at 
the NIH and the Biomedical Center in Uppsala, Sweden. In 1990 the students and faculty of Yale University 
Medical School selected Dr. Janeway as the Outstanding Teacher of Basic Biomedical Sciences. 
THE key function of the immune system is to 
recognize infectious agents as foreign and 
eliminate them from the body. The defense mech- 
anisms for the elimination phase can also damage 
the body's own tissues. This is normally avoided 
through a process of self/non-self discrimination 
mediated by highly specific receptors for anti- 
gen. Our studies have focused on rwo aspects of 
this process. First, we have explored the nature of 
receptors on T cells and of ligands on self cells in 
both normal immune responses and experimen- 
tal autoimmune diseases. Second, we have 
probed the mechanisms by which the immune 
system identifies a foreign material as infectious. 
A main objective is to identify evolutionarily 
stable microbial constituents that act on nonclon- 
ally distributed receptors and lead to the induc- 
tion of signals alerting the immune system to the 
presence of foreign matter. 
We have examined activation of normal T cells 
that requires both binding of ligand to their clon- 
ally distributed receptor and delivery of "co- 
stimulatory" signals through a still uncharacter- 
ized receptor-ligand mechanism. These studies 
have also revealed a diverse collection of micro- 
bial constituents that can induce the co-stimula- 
tor required for clonal expansion of T cells. By 
means of two signals — a specific ligand recog- 
nized by the clonally distributed receptor, and a 
nonspecific signal derived from a host cell upon 
activation by a stable microbial constituent — the 
immune system can discriminate infectious non- 
self from noninfectious self. Without the delivery 
of the co-stimulatory factor, no clonal expansion 
or differentiation to effector function occurs. 
Thus a major component of self/non-self discrimi- 
nation is mediated by the system of nonclonally 
distributed receptors for stable microbial 
constituents. 
Immune responses are triggered by specialized 
antigen-presenting cells whose distinguishing 
feature is the ability to express co-stimulatory 
factors. Immune responses, however, must be 
able to detect infection in normal tissue cells that 
do not express co-stimulatory signals. For this to 
occur, the delivery of effector responses must not 
be dependent upon co-stimulation — and is 
known not to be in the case of cytolytic T cells 
and helper T cells. A risk, however, is inherent in 
this aspect of immune effector function: the oc- 
currence of T cell responses to foreign microbial 
agents that cross-react with host cells. Such re- 
sponses might lead to autodestruction of host 
tissue. 
Avoidance of such destruction, however, might 
be effected through a third mechanism that we 
have recently revealed. Activated cells that en- 
counter ligand in the absence of co-stimulatory 
factors, as on the surface of host tissue cells, are 
induced to secrete effector lymphokines and then 
abruptly die. The mechanism of death involves 
interferon-7 produced by the effector T cell, and 
perhaps inhibited by the cytokine interleukin-4. 
This process of activation-induced cell death has 
been observed both in cultured cells and, re- 
cently in other laboratories, in vivo. 
If self/non-self discrimination is to be regu- 
lated not only at the level of receptor-ligand in- 
teraction but also by delivery of co-stimulatory 
signals, as required by all peripheral mechanisms 
of tolerance, then there must also be a require- 
ment for both the specific ligand and the co- 
stimulator to be delivered by a single cell. This 
would be necessary to avoid the activation of T 
cells that recognize ligand on the surface of a tis- 
sue cell but receive co-stimulation from resident 
tissue antigen-presenting cells. We have exam- 
ined this requirement for normal CD4 T cells and 
find that it does indeed apply. That is, delivery of 
the ligand for the T cell receptor by one cell and 
separate delivery of the co-stimulatory stimulus 
does not lead to T cell activation, clonal expan- 
sion, or the acquisition of effector function. This 
requirement allows antigen-presenting cells to 
be distributed throughout our tissues, where they 
can acquire local viruses and lead to the induc- 
tion of anti-viral immune responses without jeop- 
ardizing the critical parameter of self-tolerance. 
Given all of these restraints on self reactivity, 
one might ask how autoimmunity can ever occur? 
One mechanism that we have uncovered operates 
through the B lymphocyte. Activated T cells rec- 
ognizing antigen on the surface of a B lymphocyte 
221 
