transgenic mice and mice carrying a transgene that 
encodes a membrane-bound form of lysozyme 
(mHEL). 
In the double-transgenic offspring resulting from 
these matings, anti-HEL B cells encountered "self" 
mHEL on the surface of neighboring stromal and he- 
matopoietic ceils within the bone marrow, and as a 
result failed to appear as mature recirculating B 
cells in the peripheral lymphoid organs. The 
aborted development of self-reactive B cells in these 
mice contrasts with the normal development of 
self-reactive cells observed previously in double- 
transgenic mice expressing soluble lysozyme (sHEL; 
see below), implying that extensive receptor cross- 
linking by multivalent ligands such as cell surface 
molecules may be necessary for triggering B cell 
elimination. 
More recent studies by Suzanne Hartley and Dr. 
Michael Cooke have focused on defining cellular 
events leading to B cell elimination. In theory, the 
simplest explanation for the failure to detect mature 
B cells in these mice is that receptor crosslinking 
had directly induced a programmed cell death path- 
way. Analysis of bone marrow from the double- 
transgenic mice, however, revealed that the pool of 
immature self-reactive B cells that was first being 
exposed to mHEL was not in the process of cell 
death. FACS (fluorescence-activated cell sorting) 
and culture of these cells revealed that they had also 
not become committed to cell death, but could ma- 
ture and be triggered into antibody production if 
removed from further exposure to mHEL. Instead, 
the immediate consequence of continuous receptor 
crosslinking in vivo and in vitro was to arrest fur- 
ther development of the immature self-reactive B 
cells. 
The failure of the self-reactive B cells to mature 
and express a number of important receptor mole- 
cules, such as the lymph node homing receptor, 
complement receptors, and CD23, may in turn 
cause cell death by default. To test further the no- 
tion that cell death was secondary to arrested devel- 
opment, transgenic animals carrying the Ig trans- 
genes and a bcl-2 transgene were produced, using 
fcc/-2-transgenic mice provided by Drs. Alan Harris 
and Suzanne Cory (Walter and Eliza Hall Institute, 
Melbourne). The &c/- 2 transgene markedly delayed 
death and elimination of self-reactive B cells in 
mHEL-expressing mice but failed to perturb the 
more immediate process of arrested development. 
The significance of these findings is severalfold. 
First, the identification of proximal regulatory 
events in the elimination pathway should facilitate 
efforts to identify the underlying molecular mecha- 
nisms. Second, the potential reversibility of the 
early arrested phase of B cell elimination may con- 
tribute to escape of self-reactive B cells from elimi- 
nation in autoimmune diseases. Both issues are now 
being explored. 
Inactivation of Self-reactive B Cells 
In contrast to the fate of self-reactive B cells in 
double-transgenic mice expressing membrane- 
bound HEL, Dr. Goodnow and his colleagues pre- 
viously found that anti-HEL B cells of double- 
transgenic mice expressing a secreted form of HEL 
were not eliminated. In this case, the self-reactive B 
cells were nevertheless functionally silenced, in 
that they were incapable of mounting an effective 
antibody response in vivo. This phenomenon of 
functional silencing, or so-called clonal anergy, may 
represent a second, general mechanism for censor- 
ing self-reactive cells in both B and T cell reper- 
toires. 
Defining the basis for functional silencing in self- 
reactive B cells has formed a focus of research con- 
ducted by Drs. Sarah Bell and Michael Cooke. Stud- 
ies employing alloreactive helper T cells to promote 
B cell proliferation in vivo and in vitro have firmly 
established that the tolerant B cells are stimulated 
into clonal expansion much less efficiently than 
nontolerant controls. B cell proliferation under 
these conditions has been found to depend on lyso- 
zyme binding to the B cell antigen receptor in addi- 
tion to signals from helper T cells. 
In collaboration with Drs. Andrew Heath and 
Maureen Howard (DNAX Inc., Palo Alto), signaling 
through two of the most important molecules re- 
sponsible for receiving T cell help, namely the in- 
terleukin-4 receptor and CD40, has been shown to 
occur normally in the tolerant cells. By contrast, sig- 
naling through the B cell antigen receptor is mark- 
edly altered in the tolerant B cells, and this appears 
likely to account for their inefficient proliferation. 
Determining the molecular basis for the changes in 
antigen receptor signaling in tolerant B cells has 
therefore become a primary focus of this work. 
Development of New Mouse Models 
A number of important autoimmune diseases are 
caused by production of autoantibodies to tissue- 
specific receptors, such as the thyroid-stimulating 
hormone receptor or the acetylcholine receptor. To 
develop mouse models for studying B cell tolerance 
and autoimmunity to these and other types of tissue- 
specific self antigens, Dr. Goodnow and his col- 
leagues have begun producing new lines of HEL- 
transgenic mice. Srinivas Akkaraju has prepared a 
number of different gene constructs with different 
tissue-specific promoters and coding regions and 
332 
