TRAFFICKING OF MAJOR HISTOCOMPATIBILITY COMPLEX CLASS I MOLECULES 
C. Geoffrey Davis, Ph.D., Assistant Investigator 
Studies in Dr. Davis's laboratory have focused on 
the trafficking of molecules of the human major his- 
tocompatibility complex (MHC), with emphasis on 
endocytosis. The primary approach has been to an- 
alyze the behavior of mutant and chimeric mole- 
cules expressed in cell lines. These analyses have 
also elucidated the role of structural features of 
other cell surface molecules in endocytosis and re- 
cycling. 
I. Endocytosis of MHC Class I Molecules. 
Endocytosis of class I molecules in activated T 
lymphocytes has been reported by other labora- 
tories. However, there have been no insights into 
its functional consequences. One reason for this 
has been the inability to dissect endocytosis from 
other processes associated with T cell activation. To 
circumvent this difficulty, Dr. Davis and his col- 
leagues have constructed chimeric molecules con- 
taining the extracellular domain of the human HLA 
molecule A2.1 and cytoplasmic domains from other 
molecules that endocytose either constitutively or 
in a regulated fashion. This promising approach has 
met with some difficulties. 
Most attention to date has been devoted to chi- 
meras in which either the cytoplasmic domain 
alone or the cytoplasmic and transmembrane do- 
mains together are derived from the low-density 
lipoprotein (LDL) receptor, a membrane protein 
that is endocytosed and recycled constitutively. Al- 
though these chimeras were readily expressed and 
behaved as anticipated when transfected into COS 
cells, repeated attempts at isolating stable trans- 
fectants (including the closely related CV-1 cells) 
expressing these molecules on their surface were 
unsuccessful. In several cases it was possible to im- 
munoprecipitate from transfected cells biosyntheti- 
cally labeled A2.1 heavy chains that had not associ- 
ated with P2 ""Microglobulin and appeared to be 
retained in the endoplasmic reticulum. These re- 
sults, together with the COS cell results, indicate 
that the chimeric constructs are capable of coding 
for transmembrane proteins but that the transport 
of these proteins out of the endoplasmic reticulum 
is blocked in many cell types. Expression of full- 
length A2.1 as well as an array of site-specific mu- 
tants in the same cell lines has encountered no 
such difficulties. 
In parallel. Dr. Juerg Baenziger has undertaken 
to develop a technique to expedite the expression 
of these constructs in a variety of cell lines. Accord- 
ingly, he has cloned three constructs, including the 
full-length A2.1 cDNA, a tail-minus A2.1, and an 
A2.1/LDL receptor chimera, into vaccinia expression 
vectors. Although infection of CV-1 cells with the 
recombinant vaccinia resulted in high levels of 
expression of all three constructs, all three were 
retained in the endoplasmic reticulum. This pat- 
tern was also observed after infection of two 
human cell lines. A third human cell line that 
lacks endogenous class I molecules allowed readily 
detectable surface expression of the three con- 
structs. 
Although the establishment of a model system 
for studying the function of class I molecule endo- 
cytosis has been challenging, the final product may 
provide important information on the synthesis and 
assembly of these molecules. The recent finding 
that peptide binding plays a role in the maturation 
of nascent class I molecules may be relevant to 
these studies, and this possibility is being explored. 
The question of whether these observations are 
specific to class I molecules that transport slowly is 
also being examined. 
II. Processing of MHC Class I Molecules. 
Molecules of the A2.1 haplotype are processed 
and transported at an extremely slow rate com- 
pared with most other cell surface molecules. 
The half-time for attaining resistance to endo- 
glycosidase H, an indicator of transport through the 
medial Golgi, is ~2 h in both JY lymphoblastoid 
cells and transfected CV-1 cells. The orientation 
of the cytoplasmic domain and the fact that this 
domain is displayed in many different forms in class 
I molecules as a result of alternate splicing led 
Dr. Davis to test the possibility that the cytoplasmic 
tail might contain retention signals that could 
be transferred to other molecules. Although the 
initial experiments clearly showed that splicing 
the A2.1 cytoplasmic domain to the LDL receptor 
molecule dramatically reduced its transport to a 
rate similar to that of native A2.1, this result was 
not reproduced in subsequent transfections. A nat- 
urally occurring shortened form of the A2.1 
cytoplasmic domain resulting from deletion of exon 
7 hinders transport of the LDL receptor signifi- 
cantly. 
Continued 
379 
