CONTROL OF CELL SURFACE OLIGOSACCHARIDE EXPRESSION 
John B. Lowe, M.T)., Assistant Investigator 
The major goals of Dr. Lowe's research efforts are 
to understand both the mechanisms that regulate 
expression of cell surface oligosaccharide antigens 
and the functional consequences of these regula- 
tory events. During the past year these efforts have 
focused on exploiting gene transfer systems devel- 
oped in Dr. Lowe's laboratory for the isolation and 
molecular analysis of the glycosyltransferase genes 
that regulate expression of cell surface glyco- 
conjugates. 
Mammalian cells express a diverse array of oligo- 
saccharide molecules on their surfaces. These mole- 
cules consist of carbohydrate structures covalently 
linked to membrane-associated proteins and lipids. 
Many of these cell surface oligosaccharide struc- 
tures undergo striking changes during develop- 
ment and differentiation and in association with 
neoplastic transformation. A number of experimen- 
tal observations suggest that cell surface oligosac- 
charides may function as information-bearing mole- 
cules that mediate communication between cells 
and their environment during development and 
differentiation. The structure of these cell surface 
oligosaccharides is determined primarily by glyco- 
syltransferase enzymes. These enzymes act sequen- 
tially to synthesize the individual glycosidic bonds 
between the component monosaccharides within a 
final complex glycoconjugate. In general the oligo- 
saccharide structures displayed by a tissue are 
thought to be determined by tissue-specific glyco- 
syltransferase expression. However, the molecular 
mechanisms responsible for the regulation of ex- 
pression of these enzymes, and thus the expression 
of cell surface glycoconjugate structure, are un- 
defined. Moreover, in most instances the precise 
function(s) of the oligosaccharide structures deter- 
mined by these enzymes is also obscure. 
Dr. Lowe's laboratory is investigating representa- 
tive human blood group glycosyltransferases as 
models to understand 1) the structure and regula- 
tion of these mammalian enzymes and 2) the func- 
tion(s) of their oligosaccharide products during 
development and differentiation. These models in- 
clude fucosyltransferases whose expression is de- 
termined by the human H and Lewis blood group 
loci. Each of these loci determines the expression 
of a distinct fucosyltransferase. The H enzyme con- 
structs a fucosylated molecule, the H antigen, that 
serves as a precursor to the A and B blood group 
structures. By contrast, the Lewis enzyme con- 
structs a fucosylated molecule, similar to but 
distinct from the H antigen, that is not directly 
involved in A or B blood group biosynthesis. 
These systems provide convenient models for un- 
derstanding mammalian glycosyltransferase struc- 
ture and regulation. The genetics of these systems 
are well understood yet informative; interesting 
alleles exist at each of these loci that will serve to 
elucidate relationships between the substrate 
specificities and primary structures of glycosyl- 
transferases. Moreover, these enzymes and their 
structures are expressed in a tissue-specific and de- 
velopmentally regulated manner, and their expres- 
sion is often altered in association with neoplastic 
transformation. 
Cloned glycosyltransferase genes represent tools 
to investigate these processes. However, cloning 
of glycosyltransferase genes has proven difficult, 
because these enzymes are typically present in min- 
ute quantities and are often unstable. Moreover, 
attempts to generate antisera against them have 
often yielded reagents that primarily recognize the 
highly antigenic oligosaccharide moieties of these 
glycoproteins. Thus conventional cloning schemes 
requiring these reagents have not been generally 
successful, except when the glycosyltransferase was 
abundant or the enzyme had been successfully 
purified several hundred thousand-fold. Cloning 
strategies based on gene transfer methods were 
developed to circumvent these difficulties; these 
strategies use existing information about the 
substrate and acceptor properties of these en- 
zymes and take advantage of the multitude of anti- 
body and lectin reagents specific for the surface- 
expressed oligosaccharide products of these en- 
zymes. One of these approaches was used to isolate 
a human gene that encodes a fucosyltransferase 
with properties virtually identical to those of the 
blood group H fucosyltransferase. These studies 
have localized this gene to human chromosome 19, 
a finding consistent with linkage data that have as- 
signed the H locus to this chromosome. Sequence 
analysis of the gene's cognate cDNA indicates that 
the enzyme is a type II membrane glycoprotein, 
consisting of a short, cytoplasmic amino-terminal 
tail, a single membrane-spanning segment, and a 
large carboxyl-terminal catalytic domain that re- 
sides within the Golgi lumen. Multiple and alterna- 
tively spliced transcripts are generated from this 
gene in some cell lines, suggesting diverse possibili- 
Continued 
257 
