exons at the amino terminal of APC; most known 
ctianges have occurred in exon 1 5 , the largest and 
most 3' of the coding regions. However, Lisa Spirio 
has identified a 4-bp deletion in the third exon of 
APC in a family segregating an attenuated form of 
polyposis. This is the most 5' mutation observed to 
date. Ms. Spirio has also identified a highly polymor- 
phic, dinucleotide-repeat marker system, <100 kb 
away from the APC gene, that is contributing to the 
precision of linkage studies and presymptomatic 
diagnoses in other families with atypical APC. 
Dr. Mark Leppert has shown that the phenotype of 
the attenuated form of this inherited disease may 
vary widely within a single kindred, from only one 
or a few polyps to many. Since the phenotype of 
individuals with few polyps overlaps with that of 
many people in the general population, mutations 
in the APC gene may well define some (perhaps a 
significant proportion of) individuals in the general 
population who are at high risk for colon cancer. 
Structure of the protein product of APC. Se- 
quencing studies have revealed that APC can be 
transcribed in two different ways — i.e., alterna- 
tively spliced — in normal human tissues. The mes- 
sages encode predicted polypeptides that are 2,742 
or 2,843 amino acids long, with no extended se- 
quence homologies to other known proteins. Clear 
functional motifs are absent. However, the first 900 
residues of the gene product contain proline-free 
blocks containing heptad repeats of hydrophobic 
residues — a pattern characteristic of proteins that 
form coiled coils. 
Experiments by Dr. Geoffrey Joslyn have shown 
that the first 55 amino acids of the APC protein, 
encoded within the first three exons, are sufficient 
to form the parallel, helical dimer expected of a 
coiled coil. To achieve this result, Dr. Joslyn chose a 
system in which fragments of APC were tested for 
their ability to encode a dimerization domain for the 
cl repressor of phage X in an Escherichia coli ex- 
pression system. The physical properties of a fusion 
peptide containing the 55-amino acid segment 
strongly favor a rwo-stranded coiled-coil structure 
for this domain. 
The significance of these findings to the abnormal 
function that leads to APC is speculative at present, 
but the position of the dimerization domain at the 
amino terminus of the gene might account for the 
dominant nature of the inherited disease by permit- 
ting a peptide product with a mutation farther 
downstream to dimerize with the product of the 
normal allele and "poison" the complex. 
Dr. Groden has initiated experiments designed to 
test the ability of nonmutant APC to revert colon 
carcinoma cells to normal morphology. Cell lines 
have been transfected with a full-length cDNA 
carried on a puromycin-selectable retroviral vector, 
and the outcome is now being characterized. 
Neurofibromatosis 1 
Ongoing investigations of the neurofibromatosis 
1 gene (NFl), a tumor suppressor that was identi- 
fied in Dr. White's laboratory two years ago, and its 
protein product, neurofibromin, are designed to 
clarify normal function and to explain how muta- 
tion leads to development of neurofibromas. Like 
APC, NFl is associated with an increased risk of ma- 
lignancy in carriers of a mutant allele. Gangfeng Xu 
is studying the promoter region upstream of NFl for 
clues to expression of this gene and has constructed 
a full-length cDNA that is available to other investi- 
gators for a variety of biochemical studies. 
Also in this laboratory, Ying Li and Dr. Richard 
Cawthon have demonstrated that mutations in NFl 
occurring in somatic cells can contribute to devel- 
opment of sporadic cancers, including tumors not 
associated with neurofibromatosis. (These NFl stud- 
ies have received partial support from the National 
Neurofibromatosis Foundation and the National In- 
stitutes of Health.) 
Studies of Other Diseases 
In collaboration with clinicians in Utah and else- 
where, studies are under way to map the causative 
genes in a variety of autosomal diseases. Several 
dozen new markers for chromosome 8 have been 
developed in an effort to map more precisely a gene 
on this chromosome that carries mutations responsi- 
ble for multiple exostoses. During the past year Dr. 
Louis Ptacek, in collaboration with the group under 
Dr. Leppen's direction, showed that mutations in an 
adult skeletal muscle sodium-channel gene are 
responsible for a disabling muscle weakness, 
hypercalcemic periodic paralysis. 
Genome Project 
Dr. White directs numerous activities of the Utah 
Genome Center, which is part of the Human Ge- 
nome Project supponed by the National Institutes of 
Health. His laboratory is developing highly polymor- 
phic new DNA markers that will permit construction 
of high-density linkage maps of human chromo- 
somes. Unlike traditional RFLPs (restriction frag- 
ment length polymorphisms), the new markers 
need not be maintained as clones, because they are 
reproduced by means of the PCR (polymerase chain 
reaction) and automated sequencing. These se- 
quence-tagged sites (STSs) contain tandemly re- 
GENETICS 285 
