onstrates that the protease inhibitors are each encoded 
on separate exons, as are the serine/threonine-rich do- 
mains. The EGF repeats are flanked on the carboxyl 
terminus by introns. The region of similarity to laminin 
domain III is flanlced by introns on both ends and con- 
tains a single intervening sequence within the domain. 
This intron is in a similar position to that in authentic 
laminin, consistent with the hypothesis that this gene 
evolved from a pool of ancestral exons now found in a 
variety of other molecules. 
A variety of agrin proteins are synthesized from a 
single gene by alternative RNA splicing. Three re- 
gions of the protein make use of alternative splicing, 
resulting in 16 possible forms of the molecule. At 
amino acid position 1779, the alternate use of two 
exons gives rise to four possible splicing patterns. 
The result is a set of agrin molecules with 0,8, 11, 
or 19 (8 + 11) amino acid inserts at this position. As 
a first step in understanding the physiological roles 
of the alternate splicing patterns. Dr. Scheller and 
his colleagues expressed the four forms of agrin in 
CHO and COS cells. The transfected cells were co- 
cultured with a variety of muscle cells, including 
primary cells, C2-derived myotubes, or S27-derived 
myotubes. C2 myoblasts are cells that, when grown 
at the appropriate density in low serum, fuse to form 
myotubes. S27 cells are derivatives of C2 cells, se- 
lected because they do not synthesize proteogly- 
cans. The S27 cells generate a lower level of sponta- 
neous acetylcholine receptor (AChR) clusters than 
do the C2 cells. All forms of agrin are able to gener- 
ate clusters of AChR when cocultured with either 
primary or C2-derived myotubes. In contrast, only 
the forms of agrin containing the 8-amino acid in- 
sert (the 8- and 1 9-amino acid insert forms) were 
active in generating clusters on S27-derived myo- 
tubes. Fluorescence-activated cell-sorting (FACS) 
analysis clearly demonstrates that this is not due to 
different levels of agrin on the surface of the trans- 
fected cells. 
From these data, Dr. Scheller and his colleagues 
conclude that agrin may cluster receptors by two 
independent mechanisms — one that requires pro- 
teoglycans and another that is independent of pro- 
teoglycans but requires the 8-amino acid exon. The 
developmental expression of these splicing patterns 
is currently being investigated. It is possible that the 
introduction of different exon sequences at various 
times in development regulates different phases of 
development of the neuromuscular junction. 
Mechanisms of Synaptic Transmission 
When the action potential travels down the nerve 
and enters a release zone, changes in the membrane 
potential open channels that allow calcium to enter 
the cell. The calcium promotes transmitter release 
and membrane fusion. The membrane then recy- 
cles, forming new vesicles that are then replenished 
with chemical transmitter. This cycle might be con- 
sidered the fundamental process that underlies ner- 
vous system function, yet little is known about the 
molecular mechanisms involved. In an attempt to 
define the molecular mechanisms that regulate 
membrane flow in the nerve. Dr. Scheller and his 
colleagues have begun to characterize the proteins 
associated with the critical organelle in the process, 
the synaptic vesicle. 
Synaptotagmin, or p65, contains a membrane an- 
chor and two regions homologous to protein kinase 
C (PKC) . This homology is in the C2 region of the 
regulatory domain of PKC. This region is a feature of 
PKC isoforms that translocate to the membrane as 
part of their activation process and is found in other 
molecules, such as PLA2, which also translocate as 
part of their activation process. The laboratory pre- 
pared synaptosomes, solubilized the membranes in 
various detergents, and immunoprecipitated with 
p65 antibodies. A set of 35-kDa proteins are immu- 
noprecipitated under all conditions tested. Further 
characterization of these molecules, called syntax- 
ins, has resulted in the following model. Dr. 
Scheller and his colleagues propose that syntaxins 
are localized to the presynaptic plasma membrane 
via a carboxyl-terminal hydrophobic membrane an- 
chor. The data further suggest that the syntaxins as- 
sociate with both synaptic vesicle-anchored p65 
and the N-type calcium channel. These interactions 
are therefore proposed to be involved in docking 
synaptic vesicles at active zones. 
In the early 1980s, Dr. Regis Kelly and his co- 
workers isolated a set of monoclonal antibodies that 
specifically recognized proteins on synaptic vesi- 
cles. One of these antibodies recognizes an antigen, 
SV2, which is widely distributed throughout the 
nervous system and is highly conserved between 
species. To characterize SV2 further, Dr. Scheller 
and his colleagues obtained amino acid sequence 
and isolated cDNAs encoding the protein. The pre- 
dicted SV2 amino acid sequence has 1 2 hydropho- 
bic domains and significant amino acid sequence 
identity with bacterial sugar and drug transporters. 
The 12-membrane spanning organization is charac- 
teristic of transporters. The mammalian plasma 
membrane molecules use the Na^ gradient as an en- 
ergy source for the transport. In contrast, the bacte- 
rial transporters use an gradient, as do synaptic 
vesicles. The laboratory is currently characterizing 
other members of the SV2 family and investigating 
the function of the protein. 
The project described in this section is funded in 
434 
