molecules that interact with this inhibitor}' com- 
plex. Proteins of 57, 17, and 14 kDa were found by 
Jeff Friedman (Dr. Irving L. Weissman's laboratory, 
HHMI, Stanford University) that tightly and specifi- 
cally bound the inhibitory complex in vitro formed 
between cyclosporin A and its receptor cyclophilin. 
Dr. Jun Lui in Dr. Stuart Schreiber's laboratory (Har- 
vard University) was able to identify these proteins 
as calcineurin A, calcineurin B, and calmodulin. 
These proteins were known to act together to form a 
highly specific phosphatase. Direct biological evi- 
dence for the role of these proteins was obtained by 
Dr. Neil Clipstone in the Crabtree laboratory. Dr. 
Clipstone found that overexpression of calcineurin 
A and B made cells resistant to the drugs and, more 
significantly, rendered cells far more sensitive to ac- 
tivation than are normal cells. These studies both 
established calcineurin as the target of action of the 
drugs cyclosporin A and FK-506 and implicated 
calcineurin as a key signaling enzyme in T cell 
activation. 
The Regulated Nuclear Entry of NFATc 
Is an Element of the Signal 
Transfer Pathway 
The question that arises from the studies above is 
that of the relevant substrate for calcineurin. Dr. 
Mike Flanagan in the Crabtree laboratory began to 
explore this question by developing an in vitro 
transcription system that would faithfully mimic the 
complex requirements for T cell activation. He was 
successful in doing this in every respect except that 
the extracts of T cells treated with cyclosporin A did 
not show inhibition of in vitro transcription, even 
though the cells from which they were prepared 
demonstrated a complete block in transcription de- 
pendent on either the entire interleukin-2 (IL-2) en- 
hancer or a trimer of the NFAT (nuclear factor of 
activated T cells) site. Apparently something was oc- 
curring in vivo that could not be detected in vitro. 
Dr. Flanagan found that the drug cyclosporin A spe- 
cifically blocked the entry of a component of NFAT 
into the nucleus and thereby prevented IL-2 gene 
activation. Since nuclear and cytoplasmic compo- 
nents are mixed in the in vitro transcription ex- 
tracts, this explained the failure to see an effect of 
cyclosporin on in vitro transcription. These studies 
indicate that calcineurin is either directly or indi- 
rectly required for the entry of NFATc (cytoplasmic 
subunit of NFAT) into the nucleus. Dr. Jeff Northrop 
and Dr. Flanagan are presently purifying the cyto- 
solic component of NFATc to determine if it is a 
direct substrate for the phosphatase activity of cal- 
cineurin. 
Delineation of a Transcriptional Cascade 
Involved in Cell Type Specification 
In studies funded in part by the National Institutes 
of Health, Calvin Kuo, a graduate student in Dr. 
Crabtree's laboratory, has explored the question of 
how homeodomain-containing proteins function to 
specify cell type. Dr. Crabtree and his colleagues 
reasoned that these proteins must be responsive to 
signals from their neighbors and hence began look- 
ing at the events that regulate the expression of a 
homeodomain protein that was initially identified 
in the Crabtree laboratory. This protein, HNF-1 (he- 
patocyte nuclear factor- 1 ) , controls expression of a 
large group of genes expressed in endodermally de- 
rived tissues. HNF-1 is the only known homeodo- 
main protein that can form both homo- and hetero- 
dimers in solution and hence has the potential to 
formulate developmental codes based on regulated 
dimerization. Calvin Kuo found that HNF-1 was 
controlled by the transcription factor HNF-4 discov- 
ered by Dr. James Darnell and his colleagues (Rocke- 
feller University). HNF-4 is highly related to the 
predicted product of the Forkhead gene in Dro- 
sophila, mutations in which lead to abnormal head 
and endodermal development. HNF-4 induces 
HNF-1 expression and differentiation, as deter- 
mined by expression of proteins (such as albumin) 
characteristic of terminally differentiated cells 
when HNF-4 is transfected into undifferentiated 
cells. These studies establish a cascade governing 
terminal differentiation that the laboratory has 
termed the HNF- 1 HNF-4 hierarchy. 
Definition of a Transregulator 
of Homeodomain Protein Function 
Studies in Drosophila involving the exchange of 
domains within homeodomains have led to the con- 
clusion that the products of certain genes must act 
to modulate the function of the homeodomain. In an 
analogous experiment. Dr. Crabtree and his col- 
leagues found that HNF- 1 was inactive when trans- 
fected into some cell lines but in others could con- 
vey a high degree of transcriptional activation to its 
target promoters. These results implied that certain 
proteins were necessary for the function of HNF- 1 . 
Dirk Mendel in the laboratory set out to find them by 
looking for proteins that directly associate with the 
HNF-1 . By altering the standard method of oligonu- 
cleotide affinity chromatography, he was able to co- 
38 
