1. health and fitness
2. disease
3. aids and hiv

CHARACTERIZATION OF HUMAN MELANOTRANSFERRIN EXPRESSED IN
RECOMBINANT BACULOVIRUS INFECTED INSECT CELLS.
by
KATHERINE YUMIKO SHIMIZU
B.Sc.(4 yr.), The University of Winnipeg, 1990
A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
in
THE FACULTY OF GRADUATE STUDIES
(Department of Microbiology and the Biotechnology Laboratory)
We accept this thesis as conforming
to the required standard
THE UNIVERSITY OF BRITISH COLUMBIA
December 1993
© Katherine Yumiko Shimizu, 1993
In
presenting
degree
at
this
the
thesis
in
partial
fulfilment
University
of
British
Columbia,
freely available for
copying
of
department
publication
this
or
of
reference and study.
thesis
by
this
for
his
thesis
scholarly
or
for
her
of
gain
shall
that
agree
may
representatives.
financial
requirements
I agree
I further
purposes
the
It
not
that
be
the
by
understood
be
an
advanced
Library shall make
permission
granted
is
for
for
the
that
allowed without
it
extensive
head
of
my
copying
or
my
written
permission.
(Signature)
Department of
Microbiology
The University of British
Vancouver, Canada
Date
DE-6 (2/88)
December 15, 1993
Columbia
and
the
Biotechnology
Laboratory
ABSTRACT:
When producing recombinant mammalian proteins in an expression system, the
successful completion of posttranslational modifications is an area of concern. One
such modification is the attachment of a protein to a glycosyl-phosphatidylinositol
(GPI)-anchor in the membrane. In order to investigate this, the baculovirus/insect cell
system (Autographa californica multiple nuclear polyhedrosis virus/'Spodoptera frugiperda
Sf9 cells) was used to express the human melanoma-associated antigen, p97 or
melanotransferrin.
An unusual feature of this protein is its attachment to the cell
surface by a GPI-anchor. The expression of p97 at the surface of recombinant virus
infected Sf9 cells was shown by FACS analysis using monoclonal antibodies that
recognize two different epitopes. Immunoprecipitation of p97 from [35S]-methionine
metabolically labelled p97 B-2-I infected Sf9 cells revealed that recombinant p97 is
slightly smaller than p97 expressed by the SK-MEL-28 human melanoma cell line and
that a soluble form of p97 was present in the spent medium of the infected Sf9 cells.
While the GPI-anchored form of recombinant p97 partitioned into the detergent phase
upon Triton X-114 extraction, the form found in the spent medium partitioned into the
aqueous phase, suggesting that it may be analogous to the secreted form of p97
produced by SK-MEL-28 cells. The glycosylation of recombinant p97 from virus
infected Sf9 cells was also analyzed. Although an Endoglycosidase H-resistant form of
p97 was detected, it is likely that the processing of N-linked oligosaccharides to the
complex-type was incomplete. Finally, the GPI-llinkage of p97 expressed in Sf9 cells was
ii
demonstrated by phosphatidylinositol-specific phospholipase C sensitivity and Triton X114 extraction. The ability to express GPI-linked proteins in this system will be useful
for the bioengineering and commercial production of proteins.
Ill
TABLE OF CONTENTS:
ABSTRACT
ii
TABLE OF CONTENTS
iv
LISTOFFIGURES
vii
ACKNOWLEDGEMENT
ix
LISTOF ABBREVIATIONS
x
INTRODUCTION
1
Protein expression systems
1
The baculovirus/insect cell system
4
The tissue ditribution and structure of human melanotransferrin (p97)
Glycosyl-phosphatidylinositol(GPI)-anchors
7
9
The function of human melanotransferrin
16
Outline of the project
20
METHODS AND MATERIALS
22
Cells
22
Construction of pVL1393/p97plasmid
22
Transfection ofSf9 cells and isolation ofp97 recombinant virus
22
Analysis of p97 recombinant virus-infected cells by SDS-PAGE and Coomassie Blue
staining
25
Flow cytometry
26
Analyzing the expression of recombinant p97 during an infection period
iv
27
Immunoprecipitation of p97 from SK-MEL-28 and p97 B-2-1 infected Sf9 cell
tysates and spent medium
27
Pulse chase labelling and immunoprecipitation
29
Triton X-l 14 phase extraction
29
Endogfycosidase H digestion of recombinant p97
30
Determining sensitivity of recombinant p97 to PI-PLC treatment by flow
cytometry
30
PI-PLC treatment ofp97 that partitions into the detergent phase
31
RESULTS
33
Expression of human p97 in recombinant virus infected Sf9 cells
33
Recombinant p97 is expressed at the cell surface of recombinant virus infected Sf9
cells
37
p97 levels vary during the infection period
39
Recombinant p97 is slightly smaller than thep97 expressed in SK-MEL-28 cells...39
Pulse chase labelling and immunoprecipitation
44
A soluble form and a membrane-bound form of p97 are expressed in p97 B-2-1
infected Sf9 cells
44
Recombinant p97 is glycosylated and processed to an Endogfycosidase H-resistant
form
47
The two Endogfycosidase H-sensitive forms of recombinant p97 are found in the
aqueous and detergent phases of Triton X-l 14 partitioned Sf9 cell
fysate
Recombinant p97 is attached to the surface ofSf9 cells by a GPI-anchor
v
52
52
DISCUSSION
61
CONCLUDING STATEMENT
76
REFERENCES
77
vi
LIST OF FIGURES:
Figure 1: The amino acid sequence of human melanotransferrin
10,11
Figure 2: General structure of a GPI-anchor
13,14
Figure 3: A comparison of the structures of human melanotransferrin and human serum
transferrin
17,18
Figure 4: The pVL1393/p97 construct used to produce recombinant virus
23,24
Figure 5: Recombinant plaques can be recognized in a plaque assay by their
appearance
34
Figure 6: SDS-PAGE and Coomassie Brilliant Blue staining of recombinant p97....35,36
Figure 7: Flow cytometry analysis of p97 B-2-1 virus infected Sf9 cells
38
Figure 8: Surface expression of recombinant p97 in Sf9 cells infected at various
MOIs
40,41
Figure 9: Immunoprecipitation of p97 from SK-MEL-28 cells and from p97 B-2-1
infected Sf9 cells
42,43
Figure 10: Pulse chase analysis of recombinant p97 in infected Sf9 cell lysates and
corresponding supernatant
45,46
Figure 11: Triton X-114 phase separation of recombinant p97 present in p97 B-2-1
infected Sf9 cell lysates and corresponding supernatant
48,49
Figure 12: Analysis of recombinant p97 glycosylation in Sf9 cells by Endoglycosidase H
digestion
50,51
Figure 13: Endoglycosidase H digestion of Triton X-114 phase separated recombinant
vii
p97
53,54
Figure 14: Effect of bacterial PI-PLC on p97 cell surface expression of p97 B-2-1
infected Sf9 cells
56
Figure 15: Effect of bacterial PI-PLC on recombinant p97 found in the detergent phase
of p97 B-2-1 infected Sf9 cell lysate and corresponding supernatant
viii
57,58
ACKNOWLEDGEMENT:
I would like to thank the numerous people who contributed to the completion
of this thesis. In particular, I would like to thank Dr. Wilfred A. Jefferies for the
opportunity to work in his lab and for his assistance on the project. Special thanks to
Dr. Sylvia Rothenberger for the use of her pVL1393/p97 construct in this study, to Dr.
David Theilmann for the use of his facilities, for his technical advice, reviewing of the
manuscript and numerous conversations, and to Dr. Reinhard Gabathuler for his
technical advice, manuscript revisions, and difficult questions in preparation for the
defense. Thanks also to my committee members, Dr. Pauline Johnson and Dr. Fumio
Takei. I would like to thank Sandy Stewart for technical assistance, Dr. Malcolm
Kennard for information about protein engineering, and Mike Food for information
about p97 and rugby. Thanks to Dr. Jonas Eckstrand for his advice and vacationing
ideas. Also, many thanks to everyone in the lab, especially Ian Haidl for leeting me ask
him stupid questions, Cyprien Lomas for MAC computer advice and tips on what's
happenening in Vancouver, Roger Lippe for IBM computer advice and exposure to the
French language, and Forest, Joe, Renee, Alex, Daphne and Heather for keeping the
lab an "interesting" place.
My sincerest thanks to my parents for their support and telephone conversations
throughout this project and to Jan and Dave for everything, especially the family
humour. Ridiculously huge amounts of thank you's to Sandrina, Sean, Michelle, Aras,
and Christine for everything as always and to all of the members of Katari Taiko for
their inspiration. Many thanks to those in Winnipeg (Tani, Tannis, Jen, Beth, Christine,
Nageen and Genevieve) who keep me on the phone and help lenghthen my book list.
Thanks to Howard Damude, Michele Young, and Dave Nordquist for many laughs
during the early stages of this program. And of course, thanks to those who helped me
to procrastinate best, Andy "Weenie" Bennett for showing me the futility of life without
professional sports, Gareth Williams for keeping me informed about news in "The
Economist" and for going home early, Deb Tuyttens for keeping me sane when we're
trapped with the weenies, and especially Art Blundell for picking the V-man in the
hockey pool. Most of all, my deepest gratitude goes to Gregor Reid for his moral
support and advice, for suffering through the first draft of this manuscript, for his
company during breakfast at the Zen and the Funky Armadillo, and without whom
Scottish jokes wouldn't be nearly as funny as they are with him in mind ©.
ix
LIST OF ABBREVIATIONS:
The abbreviations used in this study, in alphabetical order, include: AChE,
acetylcholinesterase; ATCC, American Type Culture Collection; AcMNPV, Autographa
califomica multiple nuclear polyhedrosis virus; BSA, bovine serum albumin; DAF, decay
accelerating factor; DMEM, Dulbecco's Modified Eagle Medium; Endo H,
Endoglycosidase H; ECV, extracellular virus; FCS, fetal calf serum; FITC-conjugated
GAM IgG, Fluorescein isothiocyanate-conjugated goat anti-mouse immunoglobulin G;
FACS, fluorescence activated cell sorting; Gal, galactose; GIcNH2, glucosamine; Glc,
glucose; GPI, glycosyl-phosphatidylinositol; hpi, hours post infection; HPAP, human
placental alkaline phosphatase; kDa, kiloDaltons; MHC, Major Histocompatibility
Complex; Man, mannose; mAb, monoclonal antibody; MOI, multiplicity of infection;
GlcNAc, N-acetylglucosamine; OV, occluded virus; PBS, phosphate-buffered saline; PIPLC, phosphatidylinositol-specific phospholipase C; rER, rough endoplasmic reticulum;
SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; TX-114, Triton
X-114.
x
1
INTRODUCTION:
Protein expression systems. Protein production has become important not only in the
field of research, but also in the areas of medicine and agriculture. To produce large
amounts of a given protein rapidly and efficiently is the main goal. In mammalian cells,
most proteins are expressed in small amounts relative to the total cellular protein. In
order to study the structure and function of a protein, it is advantageous to have access
to a substantial quantity of a purified, biologically active form of that protein. For
example, X-ray crystallography and NMR studies, which are used to determine the
three-dimensional structure of a protein or protein complex, require large amounts of
the protein(s) in a soluble form. Proteins are also often used as reagents, vaccines,
therapeutic agents, and for the production of antibodies which can be used in the
diagnosis or treatment of diseases and for further studies of the protein produced
(Luckow, 1991). Much time and effort has been invested in the development of
different expression systems. The selection of a particular system is determined by
several parameters, such as the quantity of protein required, potential posttranslational
modification of the protein, and the cost of using that system.
The most commonly used protein expression system is the bacterial expression
system, in particular, the expression of heterologous proteins in Escherichia coli,
although other bacterial hosts can be used (Brent, 1993). The advantages of using E.
coli to express a protein of interest include the vast amount of knowledge already
acquired for this organism which allows greater manipulation of the system, the
2
simplicity and rapidity of the method, and the low cost of growing this organism.
However, an important drawback is that eukaryotic proteins may not be properly folded,
compartmentalized, or posttranslationally modified in the prokaryotic system. As a
result, large amounts of soluble protein aggregate inside the cell and form insoluble
inclusion bodies which must be denatured in order to harvest the protein (Schein, 1989;
Brent, 1993). As well, proteins that are meant to be secreted are largely retained within
the periplasmic space. This greatly reduces the efficiency of this system for the
production of soluble and secreted proteins (Brent, 1993).
Eukaryotic systems provide several advantages over their bacterial counterparts
for the production of mammalian proteins, including the likelihood that the
posttranslational modifications will be done correctly. Yeast expression systems are
similar to bacterial ones not only in terms of the basic technology, but also in terms of
the extensive knowledge on yeast biology, genetics and metabolism, the low cost of
growth medium, and the ability to grow huge quantities of yeast commercially (Collins,
1990). However, as in bacterial cells, soluble proteins often form insoluble aggregates
inside the yeast cell which make it difficult and expensive to harvest and purify
recombinant proteins (Collins, 1990). As well, glycosylation in yeast is different from
that in higher eukaryotic cells. The Glc3Man9GlcNAc2 oligosaccharide precursor
(Glc=glucose, Man=mannose, GlcNAc=N-acetylglucosamine) is transferred from a
lipid carrier to the asparagine residue of a polypeptide as it is in higher eukaryotic cells,
but after trimming of the core to Man8GlcNAc2, 50 to 150 mannose residues are added
to the core in the Golgi apparatus, which does not occur in higher eukaryotic cells
3
(Kornfeld and Kornfeld, 1985). As a result, when correct posttranslational processing
is required, this system is not used.
The mammalian expression systems have the significant advantage of producing
properly modified proteins that accumulate in their proper cellular location (Brent,
1993). There are three main systems used to express heterologous genes in mammalian
cells. The first involves the use of viruses, such as vaccinia, that infect mammalian cells.
By placing the gene of interest under the control of a viral control element, high
expression of the foreign protein can occur when the recombinant viruses infect
mammalian host cells (Kaufman, 1993a). A second system that is used extensively
requires African green monkey kidney cells transformed with origin-defective Simian
virus 40 (SV40), known as COS cells, that express high levels of large T antigen in the
absence of viral replication. COS cells can be transfected with a construct carrying the
gene of interest. Additionally, the SV40 origin of replication allows the amplification
of the construct by the large T antigen to > 100,000 copies per cell, resulting in high
expression of the protein of interest (Kaufman, 1993a). This method has the advantage
of being fairly rapid as it does not require the selection of clones. Both the vaccinia
virus and COS cell systems result in the transient expression of the desired protein
because in the first case, the virus is often cytopathic to the cells, and in the second,
cells carrying the foreign DNA often die or lose the episomes. A third method involves
the stable transfection of a mammalian cell line with a construct carrying the gene of
interest and a drug resistance marker. One system uses the dihydrofolate reductase
(DHFR) gene, which is required for the conversion of folate to tetrahydrofolate in
4
purine, nucleoside, and amino acid de novo synthesis. By transfecting a DHFR-negative
Chinese hamster ovary (CHO) cell line with a construct carrying the DHFR gene and
the gene for the protein of interest, transfectants with both genes integrated into the
genome can be selected in nucleoside-free medium. The enzymatic activity of DHFR
is inhibited by the drug, methotrexate.
By growing the transfected cells in the
increasing concentrations of methotrexate, the DHFR gene, and with it the gene of
interest, is amplified to compensate for the inhibitory effect of the drug. This results
in an increase in the gene copy number and thus, in the level of expression of the
desired protein (Kaufman, 1993b). However, culturing mammalian cells is expensive,
mainly due to the high cost of fetal calf serum (FCS), and the techniques required are
generally more difficult and time-consuming than those required for bacterial or yeast
expression systems. As well, purification of the desired protein may be complicated by
homologous proteins expressed by closely related mammalian cells.
The baculovirus/insect cell expression system. More recently, the baculovirus/insect cell
expression system has become very popular. Although it was originally chosen when a
protein could not be expressed successfully in other systems, it is now commonly the
system of choice because foreign proteins are almost always expressed at very high
levels. Some foreign proteins have been expressed in the baculovirus system at levels
20 to 250 times higher than those in mammalian systems (Luckow, 1991). The proteins
are expressed in the correct location in the cell, are soluble, and are usually correctly
folded. Furthermore, they undergo most of the posttranslational modifications that
occur in mammalian cells, including proteolytic cleavage, phosphorylation, glycosylation,
5
myristylation, palmitylation, farnesylation, carboxyl methylation, and assembly into
oligomeric complexes or disulfide-linked dimers or heterodimers (Miller, 1988; Luckow,
1991).
However, mammalian proteins are not always correctly modified and in
particular, often have a different glycosylation pattern from that of the corresponding
proteins expressed in mammalian cells (Luckow, 1991). In most studies, this does not
seem to have affected the biological activity of the glycoproteins examined (Luckow,
1991). While the system only allows for the transient expression of the protein of
interest due to cell death from the viral infection, the quantity of protein produced and
the relative ease with which recombinants are produced and screened compensate for
this potential drawback. In addition, because baculoviruses were originally produced
to be used as biological insecticides, much work has already gone into optimizing growth
conditions of insect cell lines, minimizing costs of medium, and developing a method
for semicontinuous production of insect cells and baculovirus. This has resulted in an
extremely efficient and relatively inexpensive expression system (Miller, 1988; Luckow,
1991). Finally, baculoviruses are not pathogenic to vertebrates, making this system
much less hazardous than some of the other systems.
The baculovirus strain used most often was isolated from Autographa californica
(the alfalfa looper) and is referred to as Autographa californica multiple nuclear
polyhedrosis virus (AcMNPV). AcMNPV has rod-shaped, enveloped virions that carry
the circular, double-stranded DNA genome of 128 kilobases (kb). The virus enters an
insect cell by receptor-mediated endocytosis (extracellular virus) or fusion of the
envelope with the insect cell plasma membrane (occluded virus) and the released
6
nucleocapsid enters the nucleus where the capsid is removed from the viral genome
(Volkman and Goldsmith, 1985; Blissard and Rohrmann, 1990). Two forms of virus are
produced. In the first phase of the infection, extracellular virus (ECV) buds from the
surface of the infected cell and can then infect adjacent cells. In cultures of insect cells,
the ECV are released into the medium which can be harvested as virus stock. This
occurs from 10 to 48 hours post infection (hpi). During the second phase of infection,
nucleocapsids in the nucleus are enveloped by an intranuclear process and are
assembled with the 29 kDa protein, polyhedrin, into a very large paracrystalline protein
matrix called viral occlusions or polyhedra (1 to 1.5 jim in diameter). Polyhedrin is
produced at very high levels and can comprise up to 50% of total cell protein under
optimum conditions. The occluded virus (OV) can be detected at 18 hpi and continue
to be synthesized for 4 to 5 days. Eventually, the infected cells lyse, which leads to the
death of the insect host, and the OV are released into the environment thereby allowing
for the transmission of the virus from one insect to another. Polyhedra protect the
virions from environmental stresses indefinitely, but can be solubilized by the high pH
of the insect midgut (pH 10.5) upon ingestion. This results in the release of infective
virions which infect midgut intestinal epithelial cells. The biphasic nature of an
AcMNPV infection lends itself well to the expression of foreign proteins in cultured
insect cells. Since polyhedra are not required for the infection of cultured insect cells,
the polyhedrin gene can be replaced by a foreign one (Miller, 1988; Luckow, 1991).
Integration of the foreign gene into the AcMNPV genome is achieved by
constructing a plasmid carrying the polyhedrin promoter upstream of the foreign gene
7
flanked by viral sequences which allows homologous recombination with genomic DNA.
The Sf9 cell line, a clonal line of the Spodoptera frugiperda (fall armyworm) IPLBSF21AE cell line, is most commonly used. When the insect cells are cotransfected with
wild-type AcMNPV genomic DNA and the construct DNA homologous recombination
at the site of the polyhedrin gene occurs to produce the viral genome carrying the
foreign gene under the control of the polyhedrin promoter. A mixture of recombinant
and wild-type viruses will be released into the medium of these transfected cells, which
can then be harvested. With several rounds of plaque assays, recombinant virus can be
purified from this transfection stock. The recombinant plaques can be visually detected
by examining them under a dissecting microscope (20x to 50x magnification). The wildtype plaques are composed of cells that have occlusions in the nuclei and appear
crystalline (see Results). The recombinant plaques are dull in appearance due to the
absence of polyhedrin expression in the infected cells and thus, failure to accumulate
occlusions.
The level of expression of a foreign gene appears to depend on the
particular gene, the 5'-untranslated sequences of the foreign gene included in the
construct, the construct itself, and the conditions under which the cells are grown.
However, it has been reported that production of a foreign protein can reach a level
similar to that of polyhedrin (Luckow, 1991).
Due to the advantages of the
baculovirus/insect cell system as described above, it was chosen to express human
melanotransferrin.
The tissue distribution and structure of human melanotransferrin. Melanotransferrin, or
p97, is a 97 kDa monomeric sialoglycoprotein that was originally described as a
8
melanoma-associated oncofetal antigen (Woodbury et al., 1980; Dippold et al., 1980;
Brown et al., 1980; Liao et al., 1982). It was characterized by a number of mAbs that
were created by immunizing mice with human melanoma cells. Various groups used
this approach and identified the same protein as gp95, gp87, gp90 and p97 (Woodbury
et al., 1980; Dippold et al., 1980; Brown et al., 1981b; Liao et al., 1982; Real et al.,
1985). It has been subsequently designated melanotransferrin because of its 37-39%
amino acid sequence identity to human serum transferrin, human lactoferrin, and
chicken transferrin (Brown et al., 1982). The expression of p97 by a wide variety of cell
lines and on normal adult and fetal tissue and tumor biopsy material has been examined
with varying results (Woodbury et al., 1980; Dippold et al., 1980; Woodbury et al., 1981;
Brown et al., 1981c; Brown et al., 1981a; Garrigues et al., 1982; Liao et al., 1982; Real
et al., 1988). However, these can be accounted for by the use of different mAbs and
the appearance of lower than detectable levels of p97. Thus, it appears that most
normal adult tissues express p97, although the levels of p97 are much lower than those
found for most melanoma cell lines and melanoma biopsy material. Certain fetal
tissues, including placenta, colon, umbilical cord, heart, liver sinusoidal lining cells, and
sweat gland ducts, express moderate to high levels of p97 with respect to most
melanomas and melanoma cell lines. The human melanoma cell line, SK-MEL-28,
expresses very high levels of p97 relative to all melanoma cell lines and melanoma tissue
samples tested (Woodbury et al., 1980; Dippold et al., 1980; Brown et al., 1981b). It
has been estimated that SK-MEL-28 cells express 300,000 to 400,000 p97 molecules/cell
(Brown et al., 1981b). For this reason, it has been used in many of the studies done
9
thus far and was chosen as the positive control cell line for this study.
The structure of p97 has also been analyzed (Fig. 1). From the cDNA sequence,
the size of the protein is predicted to be 80 kDa (738 amino acids) and an
unglycosylated form, which is made in a reticulocyte lysate system migrates at 84 kDa
(Rose et al., 1986). The fully glycosylated form has a molecular weight of 95,000-97,000
(Brown et al., 1981b). There are three potential N-glycosylation sites (Fig. 1, double
underline) and it has been shown that N-glycosylation and sialylation occur (Brown et
al., 1981b; Khosravi et al., 1985). Brown et al. (1985) have divided the structure of p97
into four domains. The first 19 amino acids comprise a signal peptide (Fig. 1, box),
correlating with its expression at the cell surface. The majority of the protein is divided
into two transferrin-like domains, the N-terminal domain (residues 20-361) and the Cterminal domain (residues 362-713), which have a 46% identity to each other (Brown
et al., 1985; Rose et al., 1986). Finally, there is a stretch of 25 hydrophobic amino acids
(residues 714-738) originally believed to be the membrane anchor (Fig. 1, single
underline). However, it was recently shown that the membrane-bound form of p97 is
attached to the cell surface by a glycosyl-phosphatidylinositol (GPI)-anchor (Alemany
et al., 1993; Food et al., 1993). In addition, a secreted form of p97 was found in the
spent medium of SK-MEL-28 cells (Food et al., 1993). This form had been detected
in earlier studies but was believed to be the result of dead cells or the result of shedding
from the surface of melanoma cells (Liao et al., 1985).
Glycosyl-phosphatidylinositol (GPI)-anchors. The covalent linkage of a surface protein
10
Figure 1. The amino acid sequence of human melanotransferrin.
The amino acid sequence of human melanotransferrin was deduced from the cDNA
(Rose et al., 1986). The first 19 amino acids are boxed to indicate the N-terminal signal
sequence. Amino acids 20 to 361 comprise the N-terminal transferrin-like domain while
amino acids 362 to 713 comprise the C-terminal transferrin-like domain.
domains are separated by a •.
These
Three potential N-linked glycosylation sites are
underlined with a double line. The last 25 amino acids, underlined with a single line,
are hydrophobic and serve as a signal for GPI-anchor attachment.
11
1 Met Arg Gly Pro Ser
16 |Thr Val Leu Gly |Gly
31 Pro Glu Gin His Lys
46 Ala Gly H e Gin Pro
61 Asp His Cys Val Gin
7 6 Thr Leu Asp Gly Gly
91 Leu Lys Pro Val Val
106 Ser Tyr Tyr Ala Val
121 H e Asp Thr Leu Lys
136 Arg Thr Val Gly Trp
151 Gly Arg Leu Ser Val
166 Asp Tyr Phe Gly Gly
181 Tyr Ser Glu Ser Leu
196 Glu Gly Val Cys Asp
211 Ser Gly Ala Phe Arg
22 6 Phe Val Lys His Ser
241 Leu Pro Ser Trp Gly
256 Leu Cys Arg Asp Gly
271 Cys His Leu Ala Arg
286 Asp Thr Asp Gly Gly
3 01 Arg Leu Phe Ser His
316 Glu Ala Tyr Gly Gin
331 Glu Leu Val Pro H e
346 His Glu Tyr Leu His
3 61 Arg«Leu Pro Pro Tyr
376 H e Gin Lys Cys Gly
391 Leu Lys Pro Glu H e
406 Cys Met Glu Arg H e
421 Ser Gly Glu Asp H e
436 Pro Ala Ala Gly Glu
451 Tyr Tyr Val Val Ala
466 Thr Leu Asp Glu Leu
481 Gly Ser Pro Ala Gly
496 Arg Gly Phe H e Arg
511 Ser Glu Phe Phe Asn
526 Asn Tyr Pro Ser Ser
541 Gly Arg Asn Lys Cys
556 Tyr Arg Gly Ala Phe
571 Ala Phe Val Arg His
586 Asn Ser Glu Pro Trp
601 Leu Leu Cys Pro Asn
616 Ala Cys Asn Leu Ala
631 Pro Asp Thr Asn H e
646 Gin Asp Leu Phe Gly
661 Phe Asp Ser Ser Asn
676 Ala Thr Val Arg Ala
691 Gly Trp Leu Gly Leu
706 Ser Gin Gin Cys Ser
721 Leu Leu Pro Leu Leu
73 6 Pro Ala Leu
15
10
5
Gly
Met
Cys
Ser
Leu
Ala
Gly
Ala
Gly
Asn
Met
Ser
Cys
Lys
Cys
Thr
Gin
Ser
Val
Leu
Glu
Lys
Ala
Ala
Leu
Asp
Gin
Gin
Tyr
His
Val
Arg
Trp
Pro
Ala
Leu
Val
Arg
Thr
Ala
Gly
Gin
Phe
Asp
Tyr
Val
Asp
Gly
Leu
Ala
Glu
Gly
Leu
He
He
Glu
Val
Val
Val
Gly
Cys
Arg
Ser
Leu
Val
Ala
Arg
Pro
He
Gly
Asp
Thr
Met
Arg
Met
Cys
Ala
Thr
Tyr
Val
Gly
Asp
Lys
Ser
Cys
Gly
Cys
Thr
Ala
Ala
He
Thr
Asp
His
Pro
Tyr
Ala
Pro
Leu
Val
Asn
Leu
Ala
Tyr
Val
Val
Lys
Pro
Cys
Val
Leu
Pro
Ala
Leu
Leu
Ala
Ala
Phe
Ser
Leu
Gin
Lys
Trp
Ala
Val
Glu
Ala
Ala
Arg
Lys
Val
Asp
Cys
Ala
Asn
Leu
Val
Glu
Arg
Pro
Val
His
Gly
Val
Val
Ala
Ala
Trp
Arg
Met
Cys
Ala
Glu
Tyr
Arg
Ser
Val
Asp
Pro
Cys
Leu
Glu
Glu
Leu
Asp
His
Arg
Ser
Leu
Thr
Gly
Cys
Val
Ser
Gin
Gly
Pro
Arg
Arg
Pro
Cys
Val
Leu
Ser
Val
Phe
Leu
Ala
Pro
Tyr
Asn
Gin
Gly
Ala
Ala
Leu
Leu Leu
Trp Cys
Ser, Glu
Val Arg
Gin Glu
Ala Gly
Asp Gin
Arg Ser
Cys His
Gly Tyr
Val Leu
Gly Ala
Arg Gly
Glu Arg
Gly Ala
Asn Thr
Ser Gin
Val Thr
Ala Val
Leu Leu
Phe Gin
Phe Lys
Tyr Glu
Leu Leu
Val Leu
Ala Phe
Ala Lys
Val Asp
Lys Lys
Glu Asp
Asp Ser
Ser Cys
Val Gly
Asp Val
Pro Val
Cys Val
Gin Glu
Glu Asn
Asp Asn
Arg Ser
Glu Val
His Ala
Gly Leu
Lys Asn
Asp Leu
Glu Lys
Ala Leu
Pro Ala
Ala Ala
Leu
Ala
Ala
Gly
Ala
Lys
Glu
Ser
Thr
Leu
Lys
Gly
Asp
Tyr
Gly
Asp
Asp
Glu
Val
Asn
Met
Asp
Ala
Cys
Ser
Arg
Ser
Ala
Tyr
Ser
Ser
His
Ala
Leu
Asn
Gly
Arg
Ala
Thr
Glu
Ser
Val
Leu
Gly
Leu
Thr
Glu
Pro
Arg
Ala
Thr
Phe
Thr
Asp
Glu
Val
His
Gly
Val
Ala
Glu
Ser
Tyr
Asp
Gly
Phe
Trp
Val
Glu
Phe
Ser
Trp
Asp
Thr
Arg
Pro
Val
Gly
Ser
His
Ala
Leu
Thr
Asn
Asp
Tyr
Gly
Asn
Asp
Gin
Met
Asp
Phe
Phe
Thr
Gly
Gly
Leu
Leu
Ser
Arg
Ser
Ala
His
Gly
Val
He
Glu
Val
Thr
Ser
Asp
Val
Lys
Glu
Arg
Arg
Gly
Ser
Thr
Leu
Pro
Pro
Gin
Gin
Thr
Leu
Asn
Ala
Gly
He
Ala
Pro
Glu
Tyr
Asp
Gly
Tyr
Phe
Val
Lys
Lys
Lys
Tyr
Met
Ala
Leu
Arg
Asp
Glu
Ala
He
Gly
Thr
Thr
Asn
Ser
Ser
Ser
Gly
Tyr
Ala
Thr
Leu
Gin
Ala
Gin
Ser
Ser
Gly
Asn
Glu
Arg
His
Leu
Val
Ser
Phe
Phe
Gin
Val
Lys
Gin
Gly
Val
His
Glu
Ala
Arg
Ala
Met
Asp
Arg
Ser
Pro
Pro
12
to a phospholipid embedded in the lipid bilayer has been reviewed (Low and Saltiel,
1988; Ferguson and Williams, 1988; Low, 1989). By studying the chemical composition
of GPI-anchors of various proteins, a general structure was compiled (Fig. 2) (Low and
Saltiel, 1988; Ferguson and Williams, 1988; Low, 1989). The a-carboxyl group of the
carboxy (C)-terminal amino acid of the protein is linked by an amide bond to a
phosphoethanolamine residue.
Depending on the protein, there are 1 to 3
phosphoethanolamine residues per anchor, each of which is linked by a phosphodiester
bond to a glycan which consists of variable numbers and types of sugars. The core
carbohydrates are mannose and glucosamine, but N-acerylglucosamine, galactose, and
galactosamine residues have been reported for some anchors. The glucosamine is
attached to carbon-6 of the inositol headgroup of a phosphatidylinositol that is a
component of the lipid bilayer (Fig. 2). Due to the rapid attachment of the anchor to
newly synthesized proteins (within 1 minute) in the rough endoplasmic reticulum (rER),
it is believed that the GPI-anchor is pre-assembled (Ferguson and Williams, 1988;
Masterson et al., 1989; Low, 1989). The pathway for assembly of the lipid precursor in
Trypanosoma brucei (Masterson et al., 1989) is thought to be similar in higher
eukaryotes due to the analysis of the GPI-anchor of rat brain Thy-1 (Homans et al.,
1988). It involves the initial addition of a GlcNAc residue from a UDP-GlcNAc donor
with subsequent deacetylation to form glucosaminyl phosphatidylinositol (Doering et al.,
1989), the addition of mannose residues from dolichol-phosphate-Man (Fatemi and
Tartakoff, 1986; Masterson et al., 1989; Menon et al., 1990; DeGasperi et al., 1990;
Takami et al., 1992) and finally, the addition of phosphoethanolamine from an unknown
13
Figure 2. General structure of a GPI-anchor.
The main features of a GPI anchor are illustrated with the cleavage site of bacterial PIPLC labelled. The glycan is variable depending upon the protein to which the anchor
is attached and upon the cell type and species. The GlcNH2 symbolizes a glucosamine
residue which is attached to the inositol ring by an al,6-glycosidic bond. This diagram
has been adapted from figures in Ferguson and Williams (1988) and in Kennard et al
(1993).
14
NH
Protein
Ethanolamine
Inositol
GLYCAN |-GlcNH 2
PI-PLC
•
CH 2 -CH—CH
2
i
c=o
Membrane
(CH2>n ( C H ^
CH,
CH
15
donor, possibly phosphatidylethanolamine (Doering et al., 1990). It has been suggested
that addition of galactose residues, and perhaps other types of sugar residues, to the
glycan moiety occurs after the formation of the core glycolipid (Masterson et al., 1989).
A C-terminal hydrophobic signal sequence for GPI-attachment has also been
identified (Caras et al., 1987; Low, 1989; Moran and Caras, 1991). The last 17 to 31
amino acids are cleaved from the C-terminus and the protein is attached to a GPI
anchor. The main components of the signal peptide are a terminal hydrophobic domain
(15 to 20 amino acids) and a cleavage/attachment site that is often 10 to 12 residues
amino (N)-terminal to the hydrophobic domain. The cleavage attachment site consists
of a small domain amino acid residue (serine, glycine, alanine, aspartic acid, asparagine,
and possibly cysteine) to which the GPI precursor is attached (Micanovic et al., 1990;
Moran et al., 1991) and two amino acids on the C-terminal side of the attachment
residue for which there are certain requirements (Gerber et al., 1992; Kodukula et al.,
1993). It has been suggested that a newly synthesized polypeptide may interact with the
membrane of the rER via its C-terminal hydrophobic domain and subsequently interact
with an enzyme or enzyme complex, perhaps a transamidase, which would cleave the
signal peptide and attach the polypeptide to the free amino group of the ethanolamine
residue of a pre-assembled GPI-anchor. It is possible to imagine that in the absence
of a GPI precursor, the cleaved protein is released into the rER lumen and secreted by
the cell.
There are a number of biochemical methods by which GPI-anchored, or
glipiated, proteins can be identified (Low and Saltiel, 1988; Ferguson and Williams,
16
1988; Low, 1989). The most common method used involves the release of the protein
from its glycolipid-anchor upon treatment with a bacterial phosphatidylinositol-specific
phospholipase C (PI-PLC). PI-PLC cleaves the anchor between the phosphate residue
and the glycerol of the phosphatidylinositol (Fig. 2). However, as some GPI-linked
proteins are PI-PLC resistant and a peripheral protein associated with a GPI-anchored
protein will give the same result as a protein that is directly linked to a GPI-anchor,
biosynthetic labelling with [3H]-ethanolamine, [3H]-fatty acids, or [3H]-inositol are also
used to detect and confirm a GPI-linkage. For example, the attachment of p97 to the
surface of SK-MEL-28 cells was demonstrated by p97 cell-surface release with PI-PLC
treatment as well as by [3H]-ethanolamine labelling (Alemany et al., 1993; Food et al.,
1993).
The junction of human melanotransferrin. Although the structure of p97 has been studied
in some detail, the function of the molecule has not been elucidated. Because of its
high sequence identity to the transferrin family of molecules (Fig. 3) and the discovery
that the p97 gene maps to the same region of human chromosome 3 as serum
transferrin and the transferrin receptor (Plowman et al., 1983; Seligman et al., 1986),
it was predicted that p97 would also have a role in iron (Fe+3) uptake.
Serum
transferrin binds two Fe +3 atoms, one per lobe which requires the initial binding of a
carbonate ion for each Fe +3 atom. The diferric transferrin subsequently binds to the
transferrin receptor on the surface of the cell and the diferric transferrin-transferrin
receptor complex is internalized by receptor-mediated endocytosis (Baker et al., 1987;
Klausner et al., 1983). Inside the endosome, transferrin undergoes a conformational
17
Figure 3. A comparison of the structures of human melanotransferrin and human serum
transferrin.
The proposed structures of human melanotransferrin (p97) and human serum
transferrin (Tf) are given above, including potential disulfide bridges (this figure
obtained from Dr. I. Hellstrom and Dr. K.E. Hellstrom).
Tf
P97
GMEVflHCATSDPEOHKCGMMSEAFn
QAAlLOVCHOASTCRVfLLiPOlCA
YPDKTVRHCAYSEHEATKC0SFRDHMK8
DAtTLOGCAlYEACKEHCLKPYYCEvSDOEY
NAAIARlfcOLYSAKKYCAYSPGDSPl*
C
E
TBN1GT©CSKVGKLT0ITYHSSRRYYAYA\<2)B
CV
rT
(D
P
f
OAYTLOAGLYY0AYLAPNNLKPYVAERjC8KEOP Q
CAFRiLAECAGDVAFVKHSTVLENTDCKTLP8HC0AU8flDFEULCR0GSRA
SBCLCTttteSKKCRLONMOSASOKKYYAYAYSF
.GEHSFLRaGENtLRFILCGDTOMBVVVAOAPVRALHCORHETY
?
R
C*
•
N
••
LP8KOCVGEG88
1
0
YSESLCRLCRC
S
I
TECACPYC
S
8
E
L
V
P
I
\nYMGCDVLKAYSOYFC5
f
Ep
T
T
L
k
L
Q
G
r
OSNGVCKNRCOE
Y
YPSSLIIALCVC
A
\
P
E
/
A
H
G
RUTVYEYGLYMKAOMRPPYKLF
CAFRCLVEKGOYAFYKHGTYPGNTllGKNPDPMAKNl.NEKOYELLCLOCTRK,,
M
JL
H
>
FKHF0S6HYHCQDULFK0.
A
N
KPNNVPVC
RPKDCDYLTAV8EFFN A
SCOGS8MCEI.AAYYOLCLHCRYT7
G
A
I
•nOCYATiacSKKGKLNOHTLOSABKKVVAVAFIJn^
GA
KNnoocrLooAKDLLCYVTFiNron5wMVAOpriOALN6AAr08v
e
A
m
TGDACPAC,
OAMSLOCGFVYJACKCALVPYLAEMINKSDNCEOTP,
'
KCALY:
R6AFRtLVENAG0VAFVRHTTVF0NTNGHN8EPMAAELR8EDYELLCPNGARA
Y„
R
LTs
CNIilKAlCOETT^ASVCEtKGYtV .
PSCFGA©C8RKCRLE0LtFAM8S0RRVYAYV«2BN
V
C
KDK«KEFQLF88PHC..
K
KHCALSrJHERLKCOEHTY
RHCYL8TPEI0KCC0HAYAFR
<p
G A
P
0AYTL»GE0IYTAGKKYGLVPAAGEMS\PED8 A
n
v
HEQAQNLL0HtUDEKGCM8®AVVTl98PYaAl.HCOKYEOY
C
E.R. E
„.G
M
CsX PRKPLEKAYANFF8
.EAQtREMCHapSKASVCOIEPKLRQ R
0
0
Y.
0
°-\
RNPDCULGKtIAHUYEHCLHAEYT
p"
P
Y
CAFKCLKDCAGOYAFYKH8TIFENLANKADRDaYELUCL0NTRK;
(p
©
M
\ f
IR
.C8CD1YN8CFLHa00RLtKHVCAEK0M8rTYY'<8NPATALHiNAYEEY
/..>.
— I
*
Y / N FCLFR8ETKDLLFR0_
Y /
D
/A
T
/ N .
/
K
KNNPECLNLGS
0
»8LC
«,,/«,„»««
I
/
/
/
C
K
C"
A.
H
E
o
CRFOEFFS
NR
^.LSSTsi
LSSTSCKRLNCVAKYYEECLYKEYT
A
P
GA
LL
S>Al.AARLlv
OO
19
change due to the lowered pH, resulting in the release of the Fe +3 atoms. The
apotransferrin-transferrin receptor complex is recycled back to the surface of the cell
where the empty apotransferrin is released (Klausner et al., 1983). It has been shown
that p97 also binds Fe +3 (Brown et al., 1982; Baker et al., 1992). By comparing the
three-dimensional (3D) crystal structure of human lactoferrin and rabbit serum
transferrin with the amino acid sequence of p97, Baker et al. (1987,1992) predicted and
then showed by iron binding studies, that each p97 molecule, unlike the other members
of the transferrin family, binds one Fe +3 atom rather than two. The C-terminal lobe of
p97 has amino acid substitutions for residues that are totally conserved in other
transferrins and that are predicted to be essential for Fe +3 and carbonate ion binding
(Baker et al., 1992). As a result, only the N-terminal lobe will bind an Fe +3 atom.
Additionally, amino acid sequence analysis suggests that p97 may also be able to bind
a Zn +2 atom and may even have peptidase activity (Garratt and Jhoti, 1992).
Functional studies using SK-MEL-28 cells have been performed by Richardson
and Baker (Richardson and Baker, 1990; Richardson and Baker, 1991a; Richardson and
Baker, 1991b; Richardson and Baker, 1992a; Richardson and Baker, 1992b). However,
their results are difficult to interpret with respect to the recent discovery that p97 is
GPI-anchored to the cell surface. Also, it has been reported that the surface expression
of the human transferrin receptor, which is known to have a role in iron internalization,
is up-regulated on rapidly proliferating cells (Seligman et al., 1986) and down-regulated
on cells in an iron overloaded environment (Seligman et al., 1986; Sciot et al., 1989).
In contrast, the expression of p97 did not change under either of these conditions,
20
indicating that the regulation of p97 differs from that of the transferrin receptor and
that the role of p97 in iron metabolism may differ from that of the transferrin receptor
(Seligman et al., 1986; Sciot et al., 1989). Thus, in light of the more recent discoveries,
additional functional studies of p97 are required.
Outline of the project. In order to further characterize the structure and function of p97,
as well as create polyclonal antisera required for further studies, including the search
for a mouse homologue, p97 was expressed in the AcMNPV/Sf9 insect cell system. The
surface expression and secretion of p97, the levels of p97 expression during an infection,
and posttranslational processing and transport were all examined. The possibility that
secreted and GPI-linked p97 are expressed from the cDNA in recombinant virus
infected Sf9 cells suggests that the two forms arise from a single mRNA product, not
from alternative splicing, and that the signals required to produce both forms are
recognized in Sf9 cells. Another important aspect of this study was the confirmation
that a GPI-anchored protein can be expressed in the baculovirus system. At the outset
of this project, only a few other glipiated proteins, such as hamster scrapie prion protein
(Scott et al., 1988) and Torpedo californica acetylcholinesterase (Radic et al., 1992), had
been expressed in this system, but the characterization of the GPI-anchor was not
performed in these studies. Therefore, in addition to producing a large amount of p97,
the expression of the GPI-anchor in the baculovirus system was examined. It has since
been shown that bacterial PI-PLC-sensitive, GPI-anchored forms of human class II
major histocompatibility complex (MHC) molecules, which were constructed by fusing
the extracellular domain of the DR4Dw4 a- and p- chains to a GPI signal sequence, are
21
expressed in Sf9 cells coinfected with recombinant viruses (Scheirle et al, 1992). The
study reported here reveals that different human C-terminal signal sequences can be
recognized by the intracellular machinery of Sf9 cells and result in GPI-linkage. The
ability to express glipiated proteins in this system will be of significant importance to the
bioengineering and commercial production of proteins.
22
MATERIALS AND METHODS:
Cells. SpodopterafrugiperdaIPLB-Sf21-AE clonal isolate 9 (designated Sf9) cells (a gift
from Dr. D. Theilmann of Agriculture Canada in Vancouver) were maintained in TC100 medium (Grace's medium supplemented with 3.3 g/L TC Yeastolate (Difco), 3.3
g/L Lactalbumin Hydrolysate (Difco), and 10% fetal calf serum (FCS), (Sigma)) as
described in Summers and Smith (1987). The human melanoma cell line SK-MEL-28
was obtained from the American Type Culture Collection (ATCC) and cultured in
Dulbecco's Modified Eagle Medium (DMEM, from Gibco) supplemented with 10%
FCS, 20 mM HEPES, 100 U/ml penicillin, 100 jig/ml streptomycin, 2 mM L-glutamine,
and 5 x 10"5 M 2-mercaptoethanol.
These cells were incubated at 37°C in a 5%
C0 2 /95% air, humidified environment. SK-MEL-28 cells were harvested from tissue
culture dishes using lx versene.
Construction of pVL1393lp97 plasmid. The human p97 cDNA was excised from the
pSV2p97a plasmid (provided by Dr. G.D. Plowman and Dr. K.E. Hellstrom of BristolMeyers Squibb, Seattle) (Estin et al., 1989), by a Hindlll and Nrul double digest. The
insert was isolated and made blunt ended using Klenow DNA polymerase and inserted
into the Smal site of the baculovirus transfer vector, pVL1393 (generously provided by
Dr. M. Summers) (Luckow and Summers, 1989) using standard recombinant DNA
techniques (Sambrook et al., 1989). This plasmid, called pVL1393/p97, was used to
insert the p97 cDNA into the baculovirus genome (Fig. 4).
Transfection ofSj9 cells and isolation ofp97 recombinant virus. Sf9 cells were co-
23
Figure 4. The pVL1393lp97 construct used to produce recombinant virus.
The p97 cDNA included in this construct encoded the p97 protein along with 32 bp of
the 5' untranslated region and 87 bp of the 3' untranslated region. The pVL1393
construct was cotransfected with wild-type AcMNPV genomic DNA into Sf9 cells as
described in the Materials and Methods in order to produce p97 recombinant virus.
24
1/9275
Bst EII
'960
PvuH
1350
Xhol
1900
pUC8 vector with
AmpR
Recombination
Sequences
Recombination
Sequences
polyhedrin promoter
Bam HI, Sma I, Xba I, Eco RI, Not I, Xma IH, Pst I, Bgl II
insert cloned into Sma I site
K Start
Stop
a
PQ
p97 cDNA
I
T^FT*^Hn^^r"T^n*"^T^'^rmT^nT'^TT^**T"T^rT^P
i
I
i
i
I
i
i
I
i
I
i
i
I
i
I
i
i
i
i
I
i
i
I
i
I
i
i
i
i i, 1i ,i 1i ,i 1i , 1r :y \
.i I .I I . iJ .i I . I i i i I i i i i 1
pVLl 393 vector
1
i Multiple cloning sites of pVLl 393
1
I 5' and 3' untranslated regions of p97 cDNA (incomplete)
E^i
Leader sequence/hydrophobic domain of p97
B3
Coding region of human p97 cDNA (2214 bp)
P4
H-(
25
transfected with a mixture of wild type AcMNPV genomic DNA (a generous gift of Dr.
D. Theilmann) and p97 construct according to Method I described by Summers and
Smith (1987). After a 7 day incubation at 27°C, this transfection mix was used to infect
cells in several rounds of plaque assays until recombinant virus was purified (Summers
and Smith, 1987). Occlusion negative recombinant plaques were detected by visual
screening and the purified recombinant virus was designated p97 B-2-1. Virus stock was
titered using the end-point dilution method (Summers and Smith, 1987).
Analysis of p97 recombinant virus-infected cells by SDS-PAGE and Coomassie Blue staining.
Sf9 cells were seeded in 25 cm2 flasks at a density of 4 x 106 cells. Cells were infected
with AcMNPV or p97 B-2-1 at a multiplicity of infection (MOI) of 1. 5 ml of TC100
was added to each flask and the cells were incubated at 27°C. At 72 hours post
infection (hpi), the cells were dislodged from the flask, washed with phosphate-buffered
saline (PBS), and resuspended in 200ul PBS and 200 u.1 disruption buffer (100 mM TrisHC1 pH 6.8,10% /3-mercaptoethanol, 0.2% bromophenol blue, 20% glycerol, 4% SDS).
After shearing the DNA with a lcc 26G3/8 Tuberculin syringe, a 35 \il aliquot of each
sample was electrophoresed through a 10% polyacrylamide gel. Wild type AcMNPVinfected cell lysates were run for comparison. In addition, 108 Sf9 cells grown in SF900
II serum-free insect cell medium (Gibco), were infected with p97 B-2-1 or another
recombinant virus (C) for 96 h. The supernatant was collected and 200 ml was passed
through a 0.22 \im filter and concentrated to 2.2 ml (p97) or 1.3 ml (C) using a
Centriprep-30 concentrator (Amicon) according to the instructions provided by the
manufacturer. A 30 u.1 aliquot of each supernatant, with 30 u.1 of disruption buffer, was
26
also loaded onto the gel. The gel was stained with 0.25% Coomassie Brilliant Blue,
50% methanol, 10% acetic acid and destained with 5% methanol, 7.5% acetic acid.
Flow cytometry. Two mouse monoclonal antibodies (mAbs) against human p97 were
used to detect the expression of p97 at the surface of recombinant virus-infected Sf9
cells. The mAb secreted by the L235 hybridoma cell line (ATCC), and the mAb
designated "C" secreted by the hybridoma 33B6E4 (kindly provided by Dr. S.K. Liao at
McMaster University in Hamilton) were used in the form of hybridoma culture
supernatant. The secondary antibody was Fluorescein isothiocyanate (FITC)-conjugated
goat anti-mouse (GAM) IgG (Jackson Immunoresearch Laboratories, Inc.) and was
used diluted 1:50 (vohvol) with FACS buffer (DMEM with 0.5% wt/vol bovine serum
albumin (BSA), 20 mM HEPES, and 20 mM NaN3 for SK-MEL-28 cells or TC100 with
20 mM NaN3 for infected Sf9 cells).
Sf9 cells were infected with wild type AcMNPV or p97 B-2-1 at an MOI of 1 and
harvested at 72 hpi. SK-MEL-28 cells were used as the positive control. For each
sample, 106 cells were washed twice with cold FACS buffer.
The cells were then
incubated with 100 u.1 of FACS buffer, L235 supernatant, or "C" supernatant at 4°C for
45 min and washed twice with FACS buffer.
The cells were stained with FITC-
conjugated GAM IgG at 4°C for 45 min, washed once with FACS buffer and once with
PBS with 20 mM sodium azide, and fixed in 1.5% (vol/vol) para-formaldehyde in PBS.
Fluorescence intensity was measured on a Becton-Dickinson FACScan flow cytometer.
The mean log fluorescence was converted to a linear value by the following formula:
linear mean fluorescence = 10<log mean fluorescence ^ channels) ( L i p p £
e t a j #> 1 9 9 1 )
27
Analyzing the expression of recombinant p97 during an infection period. To observe the
expression of p97 in recombinant virus infected Sf9 cells, Sf9 cells were infected at
MOIs of 0.1, 1.0, 10, and 100 and harvested at 24, 48, 72, and 96 hpi. At each time
point, cells were harvested and p97 expression was measured by FACS analysis with the
L235 mAb as described above. Uninfected Sf9 cells, AcMNPV infected Sf9 cells
(MOI=l), and SK-MEL-28 cells were stained at each time point to act as controls.
Immunoprecipitation of p97 from SK-MEL-28 and p97 B-2-1 infected SJ9 cell lysates and
spent media. SK-MEL-28 cells were grown to confluence in a 100 mm tissue culture
dish (approximately 8 x 106 cells per dish). The growth medium was removed and the
cells were pre-labelled for 1 h at 37°C in 5 ml methionine-free DMEM (Gibco)
supplemented with 20 mM HEPES and 2 mM L-glutamine. This medium was then
removed and replaced with 2.5 ml methionine-free DMEM containing 200 jiCi [35S]methionine/ml (specific activity >1000 Ci/mM for all batches) (Amersham Canada
Limited). The cells were pulsed for 30 minutes and chased with cold methionine for
8 h at 37°C. The spent medium was collected and the cells were incubated in 4 ml lysis
buffer on ice for 15 minutes. The cell lysate was divided into four samples, of which
two of these were used for the immunoprecipitation. Sf9 cells were infected with p97
B-2-1 at an MOI of 1 and harvested at 48 hpi. 4 x 106 cells were pre-labelled for 1 hour
at 27°C in 3 ml methionine-free DMEM. After removal of the pre-labelling medium,
the cells were pulsed with 0.2 mCi [35S]-methionine in 0.6 ml of methionine-free DMEM
for 30 minutes and chased in 2 ml TC100 for 8 h at 27°C. Aliquots of 1 ml were spun
at 11,000 rpm in a Canlab Biofuge B at 4°C for 5 minutes and the supernatants were
28
placed in separate tubes. The cells were lysed with 1 ml of lysis buffer (1% Nonidet P40 (NP-40), 150 mM NaCl, 50 mM Tris pH 7.4, 2 mM EDTA, 40 ug/ml phenylmethyl
sulfonylfluoride (PMSF)). These samples had been stored at -80°C for 1 month before
use in this experiment.
Cellular debris was removed from lysates and supernatants by centrifugation at
11,000 rpm in the Canlab Biofuge B for 20 minutes at 4°C. All samples were precleared with 2 u.1 normal rabbit serum/ml sample and 100 u.1 Protein A-sepharose
(Pharmacia). Samples were incubated with either 200 u.1 L235 supernatant and 50 u.1
Protein A-sepharose coated with rabbit anti-mouse IgG (Jackson) or with 50 u.1 rabbit
anti-mouse IgG coated Protein A-sepharose alone. The beads were washed twice with
Buffer B (0.2% NP-40, 150 mM NaCl, 2 mM EDTA, 10 mM Tris pH 7.4), once with
Buffer C (0.2% NP-40, 0.5 M NaCl, 2 mM EDTA, 10 mM Tris ph 7.4), and once with
Buffer D (10 mM Tris pH 7.4). After aspirating off Buffer D, the beads were either
used immediately or stored at -80°C. The beads were resuspended in 25 u.1 2x sample
buffer and boiled at 95°C for 4 minutes. The 2x sample buffer was made by combining
1 ml stock solution (0.2 M Tris pH 8.8,1 M sucrose, 5 mM EDTA, 0.01% bromophenol
blue), 0.1 ml 0.5 M dithiothreitol, and 0.2 ml 20% SDS. [14C]-methylated molecular
weight markers (Amersham Canada Limited) were also prepared. Finally, 5 u.1 of 0.5
M iodoacetamide was added to each sample before loading onto a 10% polyacrylamide
gel. After electrophoresis under reducing conditions, gels were fixed, incubated in an
enhancing solution for 15-30 min at room temperature, dried, and autoradiographed
using Kodak XAR film (Gabathuler et al., 1990; Food et al., 1993).
29
Pulse chase labelling and immunoprecipitation. Sf9 cells were infected with either wild
type AcMNPV or p97 B-2-1 virus at an MOI of 1 and harvested at 48 hpi. 1.4-1.8 x 107
cells were pre-labelled for 1 h at 27°C in 6 ml of methionine-free DMEM.
After
removal of the pre-labelling medium, the cells were pulsed with 0.5 mCi [35S]methionine in 3 ml methionine-free DMEM for 30 min at 27°C. The labelling medium
was then replaced with 6 ml TC100. The cells were chased for 24 h at 27°C. An
aliquot of 1 ml was removed at 0, 1, 2, 4, 8, 12, and 24 h chase time points. Each 1 ml
aliquot was spun at 11,000 rpm at 4°C for 5 minutes. The supernatant was placed in
a separate tube and the cells were lysed with 1 ml of lysis buffer at 4°C. Cell lysates
and supernatants were used immediately or stored at -80°C. Recombinant p97 was
immunoprecipitated using the L235 mAb as described above.
Triton X-114 phase extraction. Sf9 cells were infected at an MOI of 1 with wild type
AcMNPV or p97 B-2-1 and harvested at 48 hpi. Approximately 4 x 106 cells were
pulsed with 0.5 mCi [35S]-methionine in 1.5 ml met-free DMEM for 30 min at 27°C and
chased in TC100 for 8 h at 27°C. Aliquots of 1 ml were centrifuged at 4°C to separate
cells and supernatants and the cells were lysed in 1 ml lysis buffer containing 1% TX114 instead of 1% NP-40. TX-114 was obtained from Sigma and precondensated as
described (Bordier, 1981). The samples were cleared of nuclei and cellular debris by
centrifugation at 50,000 rpm for 1 h at 4°C in a Beckman TL100 centrifuge. 100 jil of
10% TX-114 was then added to the supernatants. All samples were incubated at 37°C
for 5 min and spun at 3000 g for 3 min at room temperature to obtain the phase
separation. The phases were placed in separate tubes and re-extracted to improve the
30
separation.
The aqueous phases from both extractions were pooled as were the
detergent phases. The volumes were equalized by adding lysis buffer without detergent
(50 mM Tris-HCl pH 7.4, 0.15 NaCl, 2 mM EDTA) to the detergent phases.
Subsequently, 1 ml aliquots of each phase from all samples were pre-cleared and used
for an immunoprecipitation experiment as described above.
Endoglycosidase H digestion of recombinant p97. Sf9 cells were infected with p97 B-2-1
virus at an MOI of 1. At 48 hpi, a pulse chase experiment was performed on two sets
of 1.2 x 107 cells. Recombinant p97 was immunoprecipitated from the cell lysate and
supernatant using L235 mAb for each chase time point and then 10 u.1 of 85 mM
sodium citrate pH 5.5 with either 2.5 u.1 Endoglycosidase H (Endo H) (1 mU/ul,
Boehringer Mannheim) or ddH 2 0 was added. All samples were incubated at 37°C for
12 h before another 10 u.1 of 85 mM sodium citrate pH 5.5 with 2.5 ul Endo H or
ddH 2 0 was added to the samples. After the second 12 h incubation at 37°C, the beads
were washed once with Buffer D and 25 ul 2x loading buffer was added to each tube.
The samples were prepared for SDS-PAGE as above and run on a 10% polyacrylamide
gel. The gel was fixed, treated with enhancing solution, dried and autoradiographed.
A similar experiment was performed on recombinant p97 immunoprecipitated
from labelled whole cell lysates and TX-114 partitioned lysates. These samples were
then used in an Endo H digest as described above.
Determining sensitivity of recombinant p97 to PI-PLC treatment by flow cytometry. The
control antibody for the PI-PLC treatment was the mAb AcVt (generously provided by
Dr. P. Faulkner of Queen's University in Kingston, Ontario) (Hohmann and Faulkner,
31
1983), in the form of hybridoma culture supernatant. AcVx reacts with the AcMNPV
protein gp64, also known as gp67, that is expressed at the plasma membrane of
AcMNPV-infected cells and becomes a part of the viral envelope as virus buds from the
cell surface (Volkman et al., 1984; Volkman and Goldsmith, 1985; Keddie and
Volkman, 1985; Whitford et al., 1989). Culture supernatant from Bacillus subtilis
(BG2320) transfected with the Bacillus thuringiensis PI-PLC gene, a gift from Dr. M. G.
Low of Columbia University in New York, was used for the PI-PLC treatment in the
FACS experiments. The PI-PLC treatment involved the incubation of cells with either
100 u.1 FACS buffer (control) or 100 u.1 PI-PLC supernatant (approximately 200-300
mU/ml, Dr. M. Kennard, personal communication) at 37°C for 60 min. Sf9 cells were
infected with AcMNPV or p97 B-2-1 at an MOI of 1 and harvested at 72 hpi. SKMEL-28 cells were used as a positive control. Samples of 106 cells were washed twice
with FACS buffer and then underwent PI-PLC or control treatment. The cells were
then washed with FACS buffer and labelled, fixed and analyzed by flow cytometry as
described above.
PI-PLC treatment ofp97 that partitions into the detergent phase. Sf9 cells were mfected
with p97 B-2-1 at an MOI of 1 and harvested at 48 hpi. 107 cells were pre-labelled,
pulsed with 1.0 mCi [35S]-methionine for 30 min, and chased in TClOO for 16 h as
described above. The cells and supernatant were separated and the cells were lysed in
1% TX-114 lysis buffer. The phases were partitioned as described above. Lysis buffer
without detergent was added to the detergent phases to make the final volume 1 ml.
Two 100 u.1 aliquots were removed from the detergent phase of the cell lysate and from
32
the detergent phase of the supernatant. To each aliquot, 900 u.1 of lysis buffer without
detergent was added in order to lower the detergent concentration. One lysate and one
supernatant aliquot were used as controls while 1 u.1 of purified Bacillus thuringiensis PIPLC (1.7 U) (a generous gift from Dr. M. G. Low) (Low et al., 1988), was added to the
other two aliquots. After incubating all samples at room temperature for 3 h, the TX114 concentration was raised to 1% and the phases were extracted once. The volumes
were equalized to 1 ml and p97 was immunoprecipitated with L235 mAb according to
the method above. This experiment was repeated for gp64 using the AcVx mAb in
order to ensure that all proteins were not affected by PI-PLC treatment. All samples
were electrophoresed through a 10% polyacrylamide gel as above.
33
RESULTS:
Expression of human p97 in recombinant virus infected Sf9 cells. Sf9 cells were cotransfected with wild type AcMNPV genomic DNA and the pVL1393/p97 construct.
A p97 recombinant virus was purified from the spent medium of these cells by plaque
purification. The recombinant (R) plaques were detected by their appearance which
is dull in comparison to the crystalline wild type (W) plaques (Fig. 5). The plaque
purified virus, designated p97 B-2-1, was used to infect Sf9 cells, which were lysed at 72
hpi and run on a 10% polyacrylamide gel and stained with Coomassie Blue (Fig. 6).
Supernatant from p97 B-2-1 infected Sf9 cells grown in serum-free medium was also run
on the gel. p97, migrating at approximately 85 kDa, was detected in the concentrated
supernatant (Fig. 6, Lane 4).
This protein was absent from the concentrated
supernatant from control (C) recombinant virus infected Sf9 cells (Fig. 6, Lane 5).
Comparison with bovine serum albumin concentration standards loaded on the gel
indicates that approximately 0.33 mg of p97 was harvested from 200 ml of the spent
medium of 108 cells, or 1.7 jig/ml of p97 in the original supernatant. However, p97
could not be detected in the cell lysate of p97 B-2-1 infected Sf9 cells by Coomassie
Blue staining (Fig. 6, Lane 3) as it migrated at the same rate as a Sf9 cell protein (Fig.
6, Lane 1). As a result, it was not possible to quantitate cell-associated p97 production.
The possibility that the cells had not become infected with p97 B-2-1 was ruled out by
the fact that the staining pattern of proteins was similar to that of wild type AcMNPV
infected cells (ie. other viral proteins were present in p97 B-2-1 infected cells), excluding
34
Figure 5. Recombinant plaques can be recognized in a plaque assay by their appearance,
Sf9 cells were cotransfected with a mixture of wild type AcMNPV genome and the
pVL1393/p97 vector as described in the Materials and Methods. The spent medium
from these cells after a 7 day incubation will contain a mixture of wild type and
recombinant virus. When this mixture is used to infect fresh Sf9 cells in a plaque assay,
the wild type (W) plaques appear crystalline while the recombinant (R) plaques appear
dull. As a result, the recombinant plaques can be isolated directly from the plate and
the recombinant virus can be purified by further rounds of plaque assays.
W
R
35
Figure 6. SDS-PAGE and Coomassie Brilliant Blue staining of recombinant p97.
Sf9 cells were mock infected (Lane 1) or infected with AcMNPV (Lane 2) or p97 B-2-1
(Lane 3) at an MOI of 1 and harvested at 72 hpi. 4 x 106 cells were lysed in a total
volume of 400 ul and a 35 ul aliquot was electrophoresed through a 10%
polyacrylamide gel under reducing conditions. Concentrated supernatants from p97 B2-1 and control recombinant virus (C) infected Sf9 cells were prepared as described in
the Materials and Methods. 30 ul aliquots of these samples were also analyzed (Lanes
4 and 5). M refers to the pre-stained molecular weight markers. The gel was stained
with Coomassie Brilliant Blue. Recombinant p97 and polyhedrin are marked with
arrows.
36
Cell Lysates
Supematants
p
Sf9
kDa
M
1
Ac
i
p97
p97
C
3
4
5
97.4
p97
69.0
46.0
polyhedrin
V
30.0
37
the very large amount of polyhedrin expressed by AcMNPV infected cells (Fig. 6, Lane
2)Recombinant p97 is expressed at the cell surface of recombinant virus infected Sf9 cells.
Flow cytometry analysis of p97 B-2-1 infected Sf9 cells revealed that recombinant p97
is expressed at the cell surface. SK-MEL-28 cells, a human melanoma cell line that
expresses p97, was used as a positive control. The presence of p97 was detected using
the L235 mAb (Real et al., 1985) and the "C" mAb that react with human p97 (Fig.
7a,b,c) (Food et al., 1993). Sf9 cells infected with p97 B-2-1 express high levels of p97
at the cell surface (Fig. 7g,h,i). Although the linear mean fluorescence is higher for p97
B-2-1 infected Sf9 cells (151.25 or 95.60) than for SK-MEL-28 cells (mean linear
fluorescence 91.40 or 49.58) when staining with L235 mAb or "C" mAb, respectively, the
population of p97 B-2-1 infected cells provided a greater range in levels of expression
(Fig. 7, compare width of h and b or i and c). From the forward scatter versus side
scatter profile from a FACS analysis, it was determined that virus infected Sf9 cells were
smaller than SK-MEL-28 cells (data not shown). Therefore, the higher intensity of
fluorescence staining by an anti-p97 mAb indicated that there were more p97 molecules
per cell on the surface of p97 B-2-1 infected Sf9 cells than on the surface of SK-MEL28 cells, which express 300,000 to 400,000 molecules per cell (Brown et al., 1981b).
Neither mock infected (data not shown) nor wild-type AcMNPV infected Sf9 cells are
stained by the L235 mAb or by the "C" mAb, indicating that they do not express any
surface proteins that cross reacts with these mAbs and that the increased fluorescence
of p97 B-2-1 infected Sf9 cells is due specifically to the presence of p97 (Fig. 7d,e,f).
38
Figure 7. Flow cytometry analysis ofp97 B-2-1 virus infected SJ9 cells.
SK-MEL-28 cells and Sf9 cells, infected with AcMNPV or p97 B-2-1 at an MOI of 1
and harvested at 48 hpi, were labelled with L235 mAb (b,e,h) or "C" mAb (c,f,i) and
subsequently, stained with FITC-conjugated GAM IgG.
The negative controls
indicating background levels of fluorescence were generated by no first antibody staining
of each cell type (a,d,g). The histograms are plotted on a log scale while the linear
mean fluorescence is indicated in the top right-hand corner of each profile.
AcMNPV
SK.MEL.28
Control
+L235
a
1.96
\b
91.40
-
51:
3w " nc
rT
A\
49.58
+"C"
t**^."
>••
!•>
I
Log of Fluorescence Intensity
p97 B-2-1
39
p97 levels vary during the infection period. In order to examine the cell surface expression
of recombinant p97 during an infection, FACS analysis was performed on cells infected
at various MOIs. The data shown in Figure 8 is representative of several independent
experiments. Generally, for MOIs of 1, 10 and 100, p97 surface expression peaked at
48 hpi and decreased as the infection progressed. Surface expression for cells infected
at an MOI of 0.1 did not peak until 72 hpi. The level of expression achieved in p97 B2-1 infected Sf9 cells after approximately 36 hpi was higher than that seen for SK-MEL28 cells while mock infected and AcMNPV infected Sf9 cells did not express any p97.
Because an MOI of 1 resulted in a high level of p97 expression in Sf9 cells without
using large amounts of virus stock while MOIs of 10 and 100 often resulted in a more
rapid decrease in p97 surface expression and cell viability, Sf9 cells were infected at an
MOI of 1 for the remaining experiments. Since the highest level of p97 surface
expression for an MOI of 1 was observed at 48 hpi, cells were harvested at this time for
subsequent experiments.
Recombinant p97 is slightly smaller than the p97 expressed in SK-MEL-28 cells. In order
to determine if the recombinant p97 expressed by infected Sf9 cells is identical to the
p97 expressed by SK-MEL-28 cells, an immunoprecipitation of p97 from both cell types
was performed. Recombinant p97 from infected Sf9 cell lysate or supernatant migrates
slightly faster than p97 immunoprecipitated from SK-MEL-28 cell lysate or supernatant
(Fig. 9, Lanes 1 to 4). There is a single major form of p97 in the SK-MEL-28 cell lysate
and supernatant (Fig. 9, Lanes 1 and 3) and a minor, slightly smaller form visible in the
cell lysate (Fig. 9, Lane 1). There is also a single major form of recombinant p97 in the
40
Figure 8. Surface expression of recombinant p97 in Sj9 cells infected at various MOIs.
Sf9 cells were infected with p97 B-2-1 at MOIs of 0.1, 1, 10, and 100 and harvested at
24, 48, 72 and 96 hpi. Surface expression of p97 was examined by FACS analysis using
the L235 mAb as the first antibody. Mock infected and AcMNPV infected (MOI= 1)
were used as negative controls while SK-MEL-28 cells were used as the positive control.
The Sf9 curve lies directly beneath the AcMNPV curve and thus, cannot be seen on the
graph. This is a figure depicts the results of a single trial.
Mean Linear Fluorescence
o
o
•
•
o
•
<*>
£ 8
—i
i
i
©
O
i
00
to
£-
-J
-Q - J
-J
ft
C/3 W
O O Q O
T IT 7 t
o o
Tt
•
oo
42
Figure 9. Immunoprecipitation ofp97from SK-MEL-28 cells and from p97 B-2-1 infected
SJ9 cells.
SK-MEL-28 cells (SK) were pulsed with [35S]-methionine for 30 minutes and chased in
normal cell culture medium for 8 h at 37°C. The cell lysate and supernatant were
incubated with L235 mAb to immunoprecipitate the labelled p97 (Lanes 1 and 3). Sf9
cells were infected with p97 B-2-1 at an MOI of 1 and harvested at 48 hpi (Bac). After
a 30 min labelling with [35S]-methionine and chase in TC100 for 8 h at 27°C, p97 was
immunoprecipitated from cell lysates and supernatants using the L235 mAb (Lanes 2
and 4). Each sample was also incubated with rabbit anti-mouse IgG coated Protein Asepharose alone as a control (Lanes 5 to 8). The p97 B-2-1 samples were used after a
month at -80°C. M refers to the [14C]-methylated molecular weight markers. Samples
were separated by SDS-PAGE (10% w/v) under reducing conditions.
The
autoradiogram represents a 5 day exposure to the dried gel about 2 months after the
p97 B-2-1 infected Sf9 cells were labelled and 1 month after the SK-MEL-28 cells were
labelled. The p97 from SK-MEL-28 cells and supernatant is marked with a single
asterix while recombinant p97 is marked with a double asterix.
43
L235
Control
n
Cell Lysates
i
kDa
200.0
M
Supematants
1
SK
1
Bac
2
~>
i
SK
3
Bac
4
r
Cell Lysates
r
i
SK
5
Bac
6
Supematants
r
SK
7
Bac
8
44
infected Sf9 cell lysate and supernatant which can be seen on a shorter exposure. The
autoradiogram had to be over-exposed with respect to the recombinant p97 in order to
visualize the p97 in the SK-MEL-28 supernatant. An approximately 110 kDa protein
found in the cell lysate and supernatant of recombinant virus infected insect cells may
have been immunoprecipitated due to interactions with either p97 itself or the p97/L235
mAb complex as it is not present in the control lanes (Fig. 9, Lanes 6 and 8). The
coated Protein A-sepharose seems to bring down some of the recombinant p97 in the
infected Sf9 cell lysate (Fig. 9, Lane 6), which is likely due to the large amount of p97
present (this has occurred for other recombinant proteins expressed in the baculovirus
system, data not shown).
Pulse chase labelling and immunoprecipitation. Recombinant p97 immunoprecipitated
from p97 B-2-1 infected Sf9 cells migrates as a single major form at approximately 85
kDa (Fig. 10, Lane 2). In addition, a single form of p97, migrating at the same rate as
the form found in the cell lysate, is detected in the medium of labelled p97 B-2-1
infected Sf9 cells (Fig. 10, Lanes 16, 18, 20, 22, 24). The continued detection of the
protein, from both the cell lysate and the supernatant, after a 24 h chase indicates that
it is relatively long-lived (Fig. 10, Lanes 12 and 24).
A soluble form and a membrane-bound form ofp97 are expressed in p97 B-2-1 infected Sf9
cells. It has been found that in SK-MEL-28 cells, two forms of p97 are expressed, a
soluble form and a membrane bound form (Food et al., 1993). These two forms can
be distinguished by differential partitioning in the detergent, TX-114. Due to their
amphiphilic nature, transmembrane proteins and GPI-anchored proteins, including p97,
45
Figure 10. Pulse chase analysis of recombinant p97 in infected SJ9 cell lysates and
corresponding supernatant.
Sf9 cells, infected with AcMNPV or p97 B-2-1 at an MOI of 1 and harvested at 48 hpi,
were pulsed with [35S]-methionine for 30 min and chased in TClOO at 27°C for various
lengths of time. p97 was immunoprecipitated from all samples using the L235 mAb.
The odd-numbered lanes are AcMNPV infected Sf9 cell lysates and supernatants at the
indicated chase times while the even-numbered lanes are p97 B-2-1 infected Sf9 cell
lysates and supernatants at the indicated chase times. M refers to the [14C]-methylated
molecular weight markers. Samples were separated by SDS-PAGE (10% w/v) under
reducing conditions. The autoradiogram represents an 8 h exposure to the dried gel.
Cell Lysates
ChaseTimes
Oh
•
IcDa
Ml
200
- -
974
lh
ii
2h
11
Supematants
4h 8h 24h
ii
ii
n
--
46.0
-"
30.0
-*•.
i
lh
ii
2h
n
4h
n
8h
ii
24h
ii
i
2 3 4 5 6 7 8 9 10 11 12 - - 13 14 15 16 17 18 19 20 21 22 23 24
-" m m m m m •
69.0
Oh
i
- . « • • *
ON
47
will partition into the detergent rich phase when extracted with TX-114 (Bordier, 1981;
Food et al., 1993). The form of p97 found in the supernatant could have arisen from
cleavage by PI-PLC or by the production of a secreted form. It will partition into the
aqueous phase when extracted with TX-114 due to its hydrophilic nature. To determine
if this was the case for the p97 expressed in recombinant virus infected Sf9 cells, [35S]methionine labelled p97 was immunoprecipitated from the lysate and spent medium of
infected cells and phase separated by TX-114 extraction as described in the Materials
and Methods. It was found that, as in SK-MEL-28 cells, recombinant p97 from the cell
lysate will partition into both the aqueous and detergent phases (Fig. 11, Lanes 5 and
6). This suggests that p97 expressed in Sf9 cells is attached by a GPI-anchor because
it is the hydrophobic nature of the GPI-anchor that causes the p97 to associate with the
detergent phase. Recombinant p97 found in the supernatant separates mainly into the
aqueous phase (Fig. 11, Lane 7). This suggests that the p97 found in the supernatant
is not merely from lysed cells but that it is a soluble form, either cleaved from the
surface of the cell or secreted into the medium. There is a small amount of p97 found
in the detergent phase of the supernatant which is likely due to the membrane-bound
form of p97 released from dead cells (Food et al., 1993).
Recombinant p97 is glycosylated and processed to an Endoglycosidase H-resistantform. The
glycosylation of p97 expressed in SK-MEL-28 cells has been studied (Brown et al.,
1981b; Khosravi et al., 1985; Food et al., 1993). To investigate the glycosylation of p97
expressed in Sf9 cells, recombinant p97 was digested with Endo H at various times
during a pulse chase experiment. Two Endo H-sensitive forms were visible (Fig. 12,
48
Figure 11. Triton X-114 phase separation of recombinant p97 present in p97 B-2-1 infected
Sf9 cell lysates and corresponding supernatant.
Sf9 cells were infected with AcMNPV or p97 B-2-1 at an MOI of 1 and harvested at
48 hpi. Infected cells were metabolically labelled with [35S]-methionine for 30 min and
chased for 16 h at 27°C. Cell lysates and supernatant were collected and all samples
were separated into the aqueous (A) and detergent (D) phases by TX-114 extraction
as described in the Materials and Methods. AcMNPV infected Sf9 cell lysate (Lanes
1 and 2) and corresponding supernatant (Lanes 3 and 4) as well as p97 B-2-1 infected
Sf9 cell lysate (Lanes 5 and 6) and corresponding supernatant (Lanes 7 and 8) were
incubated with the L235 mAb and rabbit anti-mouse IgG coated Protein A-sepharose.
Lanes 9 to 16 correspond to the same samples as Lanes 1 to 8 except that the samples
in Lanes 9 to 16 were immunoprecipitated with rabbit anti-mouse IgG coated Protein
A-sepharose alone as a control. M refers to the molecular weight markers. Samples
were analyzed by SDS-PAGE (10% w/v) under reducing conditions.
autoradiogram was developed after a 22.5 h exposure to the dried gel.
This
49
L235
AcMNPV
kDa
A D A D A D A D
Ml 2 3 4 5 6 7 8
200
97.4
69.0 -H
46.0
30.0
14.3
p97B-2-l
Control
AcMNPV
p97B-2-l
A D A D A D A D
9 10 11 12 13 14 1516
50
Figure 12. Analysis of recombinant p97 glycosylation in SJ9 cells by Endoglycosidase H
digestion.
Sf9 cells, infected with p97 B-2-1 at an MOI of 1 and harvested at 48 hpi, were pulsed
with [35S]-methionine for 30 min and chased for 0, 1, 2, 4, 8, and 24 h at 27°C. At each
time point, cell lysates and supernatants were harvested.
All samples were
immunoprecipitated with the p97 specific L235 mAb and incubated in the presence
(Lanes 1-6 and 13-18) or absence (Lanes 7-12 and 19-24) of Endo H as described in the
Materials and Methods. The samples were analyzed by SDS-PAGE (10% w/v) under
reducing conditions. The autoradiogram represents a 1.5 h exposure to the dried gel.
The two Endo H sensitive forms are indicated with arrows.
Cell Lysates
Supernatants
'
i
EndoH
+
I
i
II
i
1
.
•
+
II
!•
•
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Chase Times
kDa
Oh lh 2h 4h 8h 24h Oh lh 2h 4h 8h 24h Oh lh 2h 4h 8h 24h Oh lh 2h 4h 8h 24h
(Jy
52
Lane 1). Although there is a smear of proteins produced during the 1 h to 8 h chase
(Fig. 12, Lanes 2-6), these were eventually processed to a single higher molecular weight
form (Fig. 12, Lane 6). In the untreated samples from the cell lysate, there appears to
be a single form (Fig. 12, Lane 7). There also seems to be a single form of p97 in the
supernatant (Fig. 12, Lanes 20-24). This form is Endo H-resistant (Fig. 12, Lanes 1418) and it migrates at the same rate as the fully processed form in the cell lysate.
The two Endoglycosidase H-sensitive forms of recombinant p97 are found in the aqueous and
detergent phases of Triton X-l 14 partitioned Sf9 cell lysate. Because there were two Endo
H-sensitive forms, it was thought that perhaps one was the soluble form, while the
other was the membrane-bound form. To investigate this, p97 B-2-1 infected Sf9 whole
cell lysates and lysates that had been phase separated by TX-114 partitioning were used
to immunoprecipitate recombinant p97, which was then incubated with or without Endo
H. The single fully processed form of p97 was found in both the aqueous and detergent
phases (Fig. 13, Lanes 2 and 3) as were the two Endo H-sensitive forms (Fig. 13, Lanes
5 and 6).
Recombinant p97 is attached to the surface ofSf9 cells by a GPI-anchor. Sensitivity to PIPLC cleavage is a hallmark of GPI-anchored proteins. FACS analysis was used to show
that recombinant p97 expressed by infected Sf9 cells is cleaved from the surface with
bacterial PI-PLC. It has been shown that p97 is attached by a GPI-anchor to the
surface of SK-MEL-28 cells, Caco-2 cells (a human colorectal carcinoma cell line), and
fetal duodenum (Alemany et al, 1993; Food et al., 1993). The lipid anchor can be
cleaved by bacterial PI-PLC to release the protein into the medium (Food et al., 1993).
53
Figure 13. Endoglycosidase H digestion of Triton X-114 phase separated recombinant p97.
Sf9 cells, infected with p97 B-2-1 at an MOI of 1 for 48 h, were biosynthetically labelled
for 30 minutes with [35S]-methionine. Recombinant p97 was immunoprecipitated from
whole cell lysates (W) (Lanes 1 and 4) and from the aqueous (A) phase (Lanes 2 and
5) and detergent (D) phase (Lanes 3 and 6) of TX-114 partitioned cell lysates using the
L235 mAb. These samples were incubated with (Lanes 4 to 6) or without (Lanes 1 to
3) Endo H as described in the Materials and Methods. SDS-PAGE (10% w/v) analysis
under reducing conditions was performed on all samples. The autoradiogram was
developed after a 6 h exposure to the dried gel. The two Endo H sensitive forms of p97
are indicated with arrows.
54
+
EndoH
kDa
97.4
69.0
46.0
30.0
55
Bacterial culture supernatant containing PI-PLC was found to remove approximately
72% of the molecules of p97 on the surface of SK-MEL-28 cells (Fig. 14a,b,c) and
approximately 87% of p97 molecules from the surface of p97 B-2-1 infected Sf9 cells
(Fig. 14d,e,f). This suggested that human p97 expressed in the baculovirus system was
attached to the surface by a GPI-anchor. In order to ensure that this was a specific
effect, AcMNPV and p97 B-2-1 infected cells were treated with PI-PLC and labelled
with the AcVx mAb, which recognizes gp64, an integral membrane protein expressed
at the surface of infected cells. PI-PLC had no effect on the surface expression of gp64
in either cell type (Fig. 14g,h,i,j,k,l).
In order to confirm that the recombinant p97 is attached to the cell surface by
a GPI anchor and not simply associated with a glipiated protein, another biochemical
analysis of p97 was performed. p97 that is attached to its GPI-anchor will separate into
the detergent phase in a TX-114 extraction. When p97 is incubated with bacterial PIPLC, the lipid anchor is cleaved from the C-terminus and upon re-extraction with TX114, p97 will now separate into the aqueous phase as a result of the loss of the
hydrophobic tail (Food et al., 1993). When recombinant p97 was phase separated and
immunoprecipitated, the form that separated into the detergent phase was treated with
purified PI-PLC. It was found that PI-PLC treated recombinant p97 in the detergent
phase of the cell lysate will separate mainly into the aqueous phase upon re-extraction
with TX-114 (Fig. 15, Lane 5) whereas untreated recombinant p97 from the detergent
phase of the cell lysate will separate mainly into the detergent phase (Fig. 15, Lane 2).
The PI-PLC treatment of gp64, from the detergent phase of p97 B-2-1 infected Sf9 cell
56
Figure 14. Effect of bacterial PI-PLC on p97 cell surface expression of p97 B-2-1 infected
Sf9 cells.
SK-MEL-28 cells and Sf9 cells, infected with AcMNPV or p97 B-2-1 at an MOI of 1
for 48 h, were incubated with (c,f,i,l) or without (b,e,h,k) bacterial PI-PLC for 1 h at
37°C. Cells were then labelled with either L235 mAb (b,c,e,f) or AcV t mAb (h,i,k,l)
and subsequently, stained with FITC-conjugated GAM IgG. The negative controls
indicating background levels of fluorescence were generated by no first antibody staining
of each cell type (a,d,g,j). The histograms are plotted in a log scale while the linear
mean fluorescence is indicated in the top right-hand corner if each profile.
SK.MEL.28
+L235
Control
No PI-PLC
Treatment
JS
L
4.33
4.78
X2\
p97 B-2-1
+AcVl
AcMNPV
+AcVl
p97 B-2-1
+L235
:J
4.78
' f\
'
94.75
171.54
98.22
\
•i\
46.98
.k
-
'-.IS,-A
u
O
26.42
29.16
88.17
V^.
51.86
i
PI-PLC
Treatment
•*
• C . . iimj*
«»
•••
1
••»
M
"•*
| f'Mi^
••
••'
-| '
• ••
•%
'•'
,„.^mmt
'•*
Log of Fluorescence Intensity
, M,.
57
Figure 15. Effect of bacterial PI-PLC on recombinant p97 found in the detergent phase of
p97 B-2-1 infected Sf9 cell lysate and corresponding supernatant.
Sf9 cells, infected with p97 B-2-1 at an MOI of 1 for 48 h, were pulsed with [35S]methionine for 30 min and chased for 16 h at 27°C.
The cell lysates (Lanes
1,2,5,6,9,10,13,14) and supernatants (Lanes 3,4,7,8,11,12,15,16) were TX-114 extracted
and the detergent phase of each sample was incubated with (Lanes 5-8 and 13-16) or
without (Lanes 1-4 and 9-12) bacterial PI-PLC for 3 h at room temperature. The
samples were TX-114 extracted again and then incubated with the L235 mAb (Lanes
1-8) or the AcVx mAb (Lanes 9-16) to immunoprecipitate p97 or gp64, respectively.
Samples were analyzed by SDS-PAGE (10% w/v) under reducing conditions. The
autoradiogram was developed after a 4 day exposure to the dried gel.
L235
AcVl
I P
PI-PLC Treatment
T
kDa
r
A D A D A D A D A D A D A D A D
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
200
97.4
69,0
46.0
00
59
lysates, had no effect on its post-treatment phase extraction (Fig. 15, compare Lanes 9
and 10 with Lanes 12 and 13). Although this data clearly indicates that p97 expressed
at the surface of recombinant infected Sf9 cells is attached directly by a GPI-anchor and
not by association with another GPI-linked protein, the separation of p97 and gp64
from the detergent phase of p97 B-2-1 infected cell lysates into both the aqueous and
the detergent phase upon re-extraction complicates the analysis of the results (Fig. 15,
Lanes 1 and 2, Lanes 9 and 10, Lanes 13 and 14). This was not expected, but seems
to result from the amphiphilic nature of both GPI-linked and transmembrane proteins.
The transferrin receptor also separates into the aqueous and detergent phases upon TX114 extraction, although there is a larger quantity in the detergent phase (Food et al.,
1993). Despite this observation, it can be seen that PI-PLC treatment of recombinant
p97 had an effect on its distribution between the aqueous and detergent phases while
PI-PLC treatment of gp64 did not.
There is very little recombinant p97 in the detergent phase of the supernatant
(Fig. 11, Lane 8) and PI-PLC treatment had no effect on its phase partitioning after reextraction with TX-114 (Fig. 15, compare Lanes 3, 4, 7, and 8). The detection of p97
in the aqueous phase after re-extraction with TX-114 in the presence or absence of PIPLC may have been due to the improved partitioning of the soluble form of p97 (Fig.
15, Lanes 3 and 7). There is very little gp64 in the detergent phase of spent medium
(Fig. 15, Lanes 10, 11, 14, and 15). However, a longer exposure of the gel reveals a
small amount of gp64 that remains in the detergent phase of the supernatant after reextraction with TX-114, whether treated with PI-PLC or not (data not shown). This
60
likely results from dead cells that have lysed (Food et al., 1993). Again, PI-PLC
treatment did not cause the gp64 to partition into the aqueous phase, indicating that
gp64 is not glipiated and that the effect of PI-PLC on recombinant p97 is specific.
61
DISCUSSION:
There are many aspects of the structure and function of the human melanoma
cell marker, p97 or melanotransferrin, that are still unknown. The production of large
amounts of p97 would be useful for raising polyclonal antiserum which could be used
for the detection of p97 homologues in other organisms and p97-related molecules in
humans. It has been postulated that the secreted form of p97 binds to a receptor,
perhaps the transferrin receptor, on the surface of cells as another mechanism of iron
uptake (Food et al., 1993). Purified p97 that can be radiolabelled or biotinylated could
be used for such studies. Furthermore, the difference(s) in structure between the GPIlinked and secreted forms of p97 could be studied in more detail if substantial amounts
of both forms were available. With these ideas in mind, the expression of p97 in the
baculovirus/Sf9 cell system was undertaken. The isolation of a p97 recombinant virus
was successfully completed. Recombinant p97 was detected on the surface of infected
Sf9 cells by FACS analysis with mAbs against two different epitopes. When Sf9 cells
were infected with p97 B-2-1 recombinant virus at MOIs of 1, 10, and 100, optimal
levels of p97 surface expression occurred at 48 hpi. SDS-PAGE (10% w/v) analysis of
p97 immunoprecipitated from SK-MEL-28 cells and recombinant virus infected Sf9 cells
labelled with [35S]-methionine indicated that the recombinant p97 expressed by Sf9 cells
was slightly smaller than p97 expressed by SK-MEL-28 cells. Recombinant p97 could
also be detected in the culture medium of virus infected Sf9 cells and this form
migrated at the same rate as the cell associated form. The majority of the recombinant
62
p97 associated with the corresponding supernatant from these cells partitions into the
aqueous phase only, suggesting that this soluble p97 arose from PI-PLC cleavage or by
secretion. The glycosylation of recombinant p97 was analyzed by Endo H digestion
which revealed that at least the trimming of the oligosaccharide core occurs as the
protein is transported to the cell surface. Finally, the attachment of recombinant p97
to the surface of infected Sf9 cells by a GPI-anchor was confirmed by demonstrating
that it was sensitive to bacterial PI-PLC treatment.
It was estimated from Coomassie stained gels that the supernatant of p97 B-2-1
infected Sf9 cells had a p97 concentration of approximately 1.7 jig/ml, which is similar
to concentrations often obtained for secreted proteins (about 1-2 jig/ml) (Luckow,
1991). However, in the p97 B-2-1 infected Sf9 cell lysate, the band corresponding to
p97 was obscured by another protein with the same molecular weight that is also
expressed in mock-infected and AcMNPV infected Sf9 cells (Fig. 6, Lanes 1 and 2). As
a result, it was not possible to quantitate cell-associated p97 production, although from
radiolabelling experiments, it is known that a much larger amount of p97 was associated
with the cell compared to that found in the supernatant and compared to that found
associated with SK-MEL-28 cells.
The expression of recombinant p97 at the surface of p97 B-2-1 virus infected Sf9
cells indicated that the N-terminal signal sequence for rER translocation was recognized
in the insect cells, and that the protein had the correct conformation for transport to
the cell surface. This includes the 14 potential disulfide linkages that are believed to
be involved in establishing its tertiary structure (Rose et al., 1986). This result was
63
expected as most proteins expressed in this baculovirus system are found in their proper
cellular compartment and are usually in the correct conformation (Miller, 1988;
Luckow, 1991). It was also observed from the FACS analysis that the level of p97 on
the surface of recombinant virus infected Sf9 cells was more variable than on SK-MEL28 cells (Fig. 7). Heterogeneity in expression levels of the human transferrin receptor
on the surface of recombinant baculovirus infected Sf9 cells, indicated by
immunofluorescence microscopy, has also been reported (Domingo and Trowbridge,
1988). Interestingly, the GPI-anchored forms of HLA-DR4 molecules, produced by
fusing the extracellular domains of the DR4Dw4 a- and /2-chains to the 37 amino acid
C-terminal hydrophobic signal sequence of human decay accelerating factor (DAF), also
showed a wide variation of expression levels in virus infected Sf9 cells as shown by
FACS analysis (Scheirle et al., 1992). This heterogeneity may reflect the lytic nature
of the baculovirus life cycle which may cause cells near death to express lower levels of
p97 as a result of reduced transcription/translation. This hypothesis is supported by
data on the AcMNPV protein, gp64, as is discussed later.
In addition, variable
glycosylation, which often occurs for foreign proteins produced in the baculovirus system
(Miller, 1988; Jarvis and Summers, 1989; Davidson et al., 1990; Luckow, 1991), may
affect the transport and stability of p97 and result in varying levels of p97 at the cell
surface.
The fact that both the L235 mAb and the "C" mAb, which recognize different
epitopes (Kennard et al., 1993), react with p97 found at the surface of recombinant
virus infected Sf9 cells suggests that the recombinant p97 is properly folded. This
64
suggestion is strengthened by the finding that L235 mAb reacts more strongly with p97
on the surface of recombinant virus infected Sf9 cells than does the "C" mAb. This may
be due to the lower affinity of the "C" mAb for its epitope or to its specificity for a less
accessible epitope. These binding characteristics are the same as those observed with
the endogenous p97 expressed by SK-MEL-28 cells. Thus, the fact that recombinant
p97 reacts with both of these mAbs with an affinity similar to that of p97 expressed in
SK-MEL-28 cells suggests that p97 is synthesized correctly in Sf9 cells and that the main
intracellular machinery required to achieve this is conserved in insect cells.
In order to maximize protein production it was necessary to determine the
parameters of recombinant p97 expression in infected Sf9 cells. This would allow for
the harvesting of the protein at its peak level of expression while cell death, which
results from cell lysis, is at a minimum (Miller, 1988; Luckow, 1991). MOIs of 10 and
100 resulted in a more rapid decrease p97 surface expression, perhaps as a result of a
more rapid decrease in cell viability than MOIs of 1 or 0.1. This is supported by the
work of Licari and Bailey (1991), who analyzed the factors that would allow optimal /8galactosidase production in recombinant virus infected Sf9 cells and showed that as the
MOI was increased, the more rapidly cell viability decreased. For cells infected during
the early exponential phase of growth, the yield of /3-galactosidase activity did not vary
for a range of MOIs (0.1 to 100) and for an MOI of 10, the maximum synthesis rate
occurred at 50 hpi while intracellular concentrations were highest around 110 hpi (Licari
and Bailey, 1991). The maximal levels for the GPI-linked HLA-DR4 molecules,
measured by FACS analysis, were detected at 60 hpi (Scheirle et al., 1992). An MOI
65
of 1 was optimal for p97 expression since it resulted in high levels of cell surface p97
at 48 hpi, prior to the onset of most cell death. The decline in p97 surface expression
towards the end of the infection period may correlate with the progressive deterioration
of the cells from the cytopathic effects of the virus as mentioned above. It is likely that
time post-infection for maximal expression of different recombinant proteins will vary
depending on the protein and its cellular location since posttranslational modifications
and transport rates will also vary (Luckow, 1991).
The p97 found in the spent medium of SK-MEL-28 cells is the same size as the
form associated with the cell, as has been previously reported (Food et al., 1993). Food
et al. (1993) showed that p97 in the supernatant was a secreted form of the molecule
by surface biotinylating SK-MEL-28 cells and phase separating the lysate and
supernatant by TX-114 extraction. Labelled p97 was found in the detergent phase of
the cell lysate but could not be detected in the aqueous phase of the supernatant,
indicating that the p97 normally found in the spent medium of SK-MEL-28 cells is not
glipiated p97 that has been cleaved from the surface by an endogenous PI-PLC but
rather, is actively secreted. Further, evidence supporting this was the fact that [3H]ethanolamine labelled p97 was not found in the spent medium of SK-MEL-28 cells,
suggesting that the soluble form of p97 is not a GPI-linked form that had been PI-PLC
cleaved. Such a molecule would carry the remainder of the anchor, containing at least
one ethanolamine residue, and would thus result in the detection of [3H] in the medium
(Food et al., 1993).
Preliminary evidence suggests that the soluble form of p97 in recombinant virus
66
infected Sf9 cells shares the characteristics of secreted p97 from SK-MEL-28 cells. A
soluble form was also detected in the medium of recombinant virus infected Sf9 cells
as early as the 1 h chase time point. In order to confirm that the soluble form of
recombinant p97 was analogous to the secreted form of SK-MEL-28 cells and not
simply a result of lysed Sf9 cells, a TX-114 phase separation of p97 B-2-1 infected Sf9
cell lysates and corresponding supernatant was performed.
It was found that p97
immunoprecipitated from the cell lysate separated into both the aqueous and detergent
phases while p97 immunoprecipitated from the supernatant separated mainly into the
aqueous phase, with only trace amounts in the detergent phase (Fig. 11), suggesting that
most of the p97 in the supernatant lacked a GPI-anchor.
Since the secreted form is
found in SK-MEL-28 cells and transfected CHO cells (Food et al., 1993), it is possible
that an inherent characteristic of the p97 GPI signal causes a portion of the molecules
to bypass the anchor attachment step. It is also possible that since such a large number
of p97 molecules are being synthesized in the Sf9 cells, there are not enough GPI
precursors available for attachment. In any case, the C-terminal hydrophobic signal
sequence of a newly synthesized p97 polypeptide would be recognized and cleaved by
a transamidase or peptidase, but the protein would not be attached to a GPI-anchor.
Instead, the hydrophilic p97 would be released into the lumen of the rER and
transported out of the cell via a secretory pathway. This possibility is supported by the
finding that a C-terminal signal peptide cleaved form of Thy-1 is secreted by class E
mutant mouse lymphoma cells which cannot synthesize GPI-anchors (Fatemi and
Tartakoff, 1986). As well, Drosophila melanogaster acetylcholinesterase (AChE), which
67
is normally GPI-linked, is secreted from Xenopus oocytes injected with a cDNA
encoding the protein without its C-terminal hydrophobic peptide (Fournier et al., 1992).
Furthermore, the possibility that p97 in the supernatant is simply p97 with its Cterminal hydrophobic signal sequence but lacking its GPI-anchor is eliminated by the
fact that p97 in the supernatant is the same size as glipiated p97 associated with the Sf9
cells, and that p97 in the supernatant separates into the aqueous phase upon TX-114
extraction. As well, it has been shown that a fusion protein, consisting of human growth
hormone (normally secreted) and the GPI signal lacking a cleavage site from the Cterminus of DAF, is retained in a post-ER compartment inside the cell (Moran and
Caras, 1992).
However, it could not be determined from the phase separation
experiment whether this aqueous form of p97 was a result of endogenous PI-PLC
cleavage of GPI-linked p97 from the surface of recombinant virus infected Sf9 cells or
due to the secretion of p97 by these cells.
Experiments such as the surface
biotinylation/phase extraction or a [3H]-ethanolamine biosynthetic labelling would help
to solve this dilemma.
The size difference between p97 expressed in recombinant virus infected Sf9 cells
and that in SK-MEL-28 cells may have been due to differences in glycosylation, as is
often the case for foreign proteins expressed in the baculovirus system (Scott et al.,
1988; Domingo and Trowbridge, 1988; Jarvis and Summers, 1989; Webb et al., 1989;
Miller, 1988; Luckow, 1991; Radic et al., 1992). In order to analyze the glycosylation
and transport of p97 in Sf9 cells, an Endo H digestion was performed on
immunoprecipitated p97 from a pulse chase experiment. At the 0 h chase time, p97
68
from recombinant virus infected Sf9 cell lysates was found in two Endo H-sensitive
forms approximately 79 and 81 kDa in size (Fig. 12, Lane 1). After a 1 h chase, Endo
H-resistant forms of p97 are already present, indicating that transport of p97 through
the Golgi has already occurred. The highest molecular weight Endo H-resistant form
is present by the end of the 2 h chase, indicating that p97 is transported to the surface
within 2 h (Fig. 12, Lane 3). This is a more rapid rate of transport than that of p97 in
SK-MEL-28 cells, which takes 4 h to reach a fully processed form, but a slower rate of
transport than that of the transferrin receptor in SK-MEL-28 cells, which reaches a fully
processed form within 1 h (Food et al., 1993). The soluble form of p97 found in the
supernatant of p97 B-2-1 infected Sf9 cells is Endo H-resistant at all chase times,
including the 1 h chase time point when it is first detected (can be seen in Fig. 12, Lane
14 on a longer exposure), indicating that it is the fully processed form. This correlates
with either possibility that the soluble p97 is cleaved from the surface of recombinant
virus infected Sf9 cells or is secreted by these cells, as it is by SK-MEL-28 cells.
Although the secreted form of p97 in SK-MEL-28 cells is also Endo H-resistant, it is
not detected in the media of [35S]-methionine labelled cells until after a 4 h chase,
possibly due to slower transport of p97 in SK-MEL-28 cells than in p97 B-2-1 infected
Sf9 cells. Alternatively, the amount of secreted p97 in the medium at earlier chase time
points is undetectable.
In insect cells, it is believed that the same Glc3Man9GlcNAc2 (Glc=glucose,
Man=mannose, GlcNAc=N-acetylglucosamine) oligosaccharide core is transferred from
a dolichol pyrophosphate carrier to the asparagine residue of a polypeptide during its
69
translation in the rER as occurs in vertebrate cells (Quesada Allue, 1980; Butters et al.,
1981; Hsieh and Robbins, 1984; Miller, 1988; Luckow, 1991; Davidson et al., 1990).
Although it was originally believed that oligosaccharides on insect proteins and
recombinant proteins expressed in insect cells were only processed to a high-mannose
form and could not be processed to the complex-type oligosaccharides due to the
absence of glycosyltransferases, such as N-acetylglucosaminyl-, galactosaminyl-, and
sialyltransferases (Butters et al., 1981; Hsieh and Robbins, 1984; Domingo and
Trowbridge, 1988), recent evidence suggests that this is not the case, at least not in
lepidopteran cells (Jarvis and Summers, 1989; Davidson et al., 1990; Davidson and
Castellino, 1991b; Davidson and Castellino, 1991a). A subset of oligosaccharides linked
to recombinant human tissue plasminogen activator (t-PA) expressed in virus infected
Sf9 cells was shown to be Endo H-resistant, and thus, no longer a high-mannose form
(Jarvis and Summers, 1989).
More direct evidence comes from the study of
oligosaccharides attached to the single N-glycosylation site of recombinant human
plasminogen (HPg), which revealed that in virus infected S. frugiperda IPLB-SF-21AE
cells and Mamestra brassicae IZD-MBO503 cells, complex-type oligosaccharides
containing N-acerylglucosamine, galactose, and sialic acid in both cell types, and fucose,
in MBO503 cells only, were present in 40-60% of the oligosaccharides attached to
recombinant HPg (Davidson et al., 1990; Davidson and Castellino, 1991a).
Recombinant HPg expressed inManduca sexta (another lepidopteran insect) CM-1 cells
is also sialylated, again suggesting the presence of complex-type oligosaccharides
(Davidson and Castellino, 1991a).
70
One of the interesting findings of these studies is that one of the oligosaccharides
found attached to HPg in both SF-21AE cells and MBO503 cells, was a Man3GlcNAc2,
which is not a usual intermediate of the classical N-linked oligosaccharide processing
pathway in mammalian cells (Kornfeld and Kornfeld, 1985).
However, it is an
intermediate of an alternate glycosylation pathway proposed to involve the transfer of
Glc3Man5GlcNAc2 (Glcal -» 2Glcal -* 3Gla*l -• 3Manal -* 2Manal -• 2Manal -»•
3 (Manal -> 6) Man/31 -> 4GlcNAq81 -» 4GlcNAc) to asparagine residues (Kornfeld
et al., 1979; Chapman et al., 1980; Rearick et al., 1981; Yamashita et al., 1983).
Man3GlcNAc2 has been found attached to glycoproteins in vesicular stomatitis virus
infected class E Thy-1-negative mouse lymphoma cells (Kornfeld et al., 1979), to hen
ovomucoid (Yamashita et al., 1983), and to glycoproteins in Aedes albopictus mosquito
cells (Hsieh and Robbins, 1984). This pathway may also be important in terms of p97
processing in Sf9 cells because the Man3GlcNAc2 oligosaccharides would be Endo Hresistant, but would lack added complex carbohydrates which would result in its lower
molecular weight relative to p97 from SK-MEL-28 cells. This has also been suggested
as the explanation for the Endo H-resistant form of recombinant human t-PA expressed
in Sf9 cells, which is also smaller in size than its mammalian counterpart (Jarvis and
Summers, 1989). This prediction for recombinant p97 in Sf9 cells is supported by the
recent work by Dr. Reinhard Gabathuler and Michael Food (personal communication),
who have shown by two-dimensional gel electrophoresis that p97 from SK-MEL-28 cells
has 7 to 8 forms while p97 from recombinant virus infected Sf9 cells has only 3 to 4
forms. The most acidic forms are absent from virus infected Sf9 cells, indicating that
71
the recombinant p97 in Sf9 cells has not been sialylated.
The presence of two Endo H-sensitive forms of p97 in recombinant virus infected
Sf9 cells (Fig. 13, Lane 1) was thought to correlate with the secreted and GPI-anchored
forms of the protein. To determine if this was so, p97 immunoprecipitated from whole
cell lysates and from TX-114 partitioned cell lysates was incubated in the presence or
absence of Endo H. It was found that both of the bands found in Endo H-digested
samples were present in the aqueous and detergent phases of the lysates, dispelling the
theory that the two bands represented the two different forms of p97 and suggesting
instead that the bands represented differences in some other posttranslational
modification(s). However, it is unlikely that the two forms result from differences in
O-linked glycosylation because potential O-linked glycosylation sites have not been
identified for human p97 (Rose et al., 1986).
The membrane-bound form of p97 expressed by recombinant virus infected Sf9
cells is linked to the cell surface by a GPI-anchor as it is in SK-MEL-28 cells, Caco-2
cells and fetal duodenum (Alemany et al., 1993; Food et al., 1993). Bacterial PI-PLC
sensitivity of recombinant p97 on the surface of infected Sf9 cells was demonstrated
using FACS analysis. p97 B-2-1 infected Sf9 cells and SK-MEL-28 cells treated with
bacterial PI-PLC have low levels of p97 due to the cleavage of the p97 GPI-anchor
which releases the protein from the cell surface. Bacterial PI-PLC treatment of both
SK-MEL-28 cells and p97 B-2-1 infected Sf9 cells does not remove all of the p97
present (removal of 72% and 87% of p97, respectively). However, these experiments
were performed with bacterial PI-PLC supernatant (200-300 mU/ml) and other trials
72
using purified PI-PLC (8.5 U/ml) resulted in removal of approximately 95% of p97 from
the surface of both cell lines (data not shown). It has also been shown using FACS
analysis that Thy-1 on the surface of EL-4 cells (a mouse T cell line) is found in
bacterial PI-PLC-sensitive and -resistant forms (Low et al., 1988). All of the Thy-1
could not be removed from the surface of EL-4 cells using high concentrations of two
different PI-PLCs, one purified from Bacillus thuringiensis (>1.4 U/ml) and one purified
from Staphylococcus aureus (10 U/ml). Low et al. (1988) proposed some possible
explanations for the PI-PLC-resistant form(s) of Thy-1 on the surface of EL-4 cells that
may also apply to the p97 PI-PLC resistant form(s) on the surface of recombinant virus
infected Sf9 cells and on the surface of SK-MEL-28 cells. They suggest that the GPIanchor of some of the molecules may not be accessible to the enzyme due to the
interference of neighbouring proteins or lipids. An extension of this idea is that
perhaps steric hinderance from the GPI-linked protein itself prevents the interaction of
the PI-PLC with the anchor. Alternatively, variations in the structure of the GPIanchor may render it PI-PLC-resistant. Other GPI-linked proteins expressed in certain
cell types are also PI-PLC-resistant, including human and mouse erythrocyte AChE,
human erythrocyte DAF and lymphocyte function-associated antigen 3 (LFA-3), and
two Dictyostelium discoideum proteins, contact site A and Antigen 117 (Low et al., 1988;
Ferguson and Williams, 1988; Low, 1989). The GPI-anchor of human erythrocyte
AChE has a pahnitate residue attached to one of the hydroxyl groups of the inositol
ring (likely on Carbon-2), which is responsible for PI-PLC resistance (Low, 1989), and
the anchor of contact site A protein may contain ceramide, rather than alkyl- and
73
acylglycerols, which may account for this anchor's resistance to PI-PLC. Possibly
interference from sugar residues in the glycan moiety could also render certain anchors
PI-PLC-resistant. Another explanation is that the PI-PLC-resistant form of p97 that is
being detected is actually soluble p97, either PI-PLC released or the secreted form, that
has associated non-covalently with a "receptor" on the surface of Sf9 cells or SK-MEL28 cells. This possibility has been suggested for Thy-1 in EL-4 cells and heparan sulfate
proteoglycan in hepatocytes (Low et al., 1988). The possibility of a transmembrane
polypeptide-anchored form due to differential mRNA processing is eliminated by the
fact that in recombinant virus infected Sf9 cells, only the p97 cDNA is present.
PI-PLC removal of p97 from the surface of p97 B-2-1 infected Sf9 cells was
specific because surface levels of gp64, an integral membrane protein, on either
AcMNPV or p97 B-2-1 infected Sf9 cells were not affected by bacterial PI-PLC
treatment (Fig. 14).
It has been shown by immunofluorescence microscopy and
immunoelectron microscopy that gp64 localizes to the surface of AcMNPV infected Sf9
cells, allowing its detection using the AcVx mAb. FACS analysis of AcMNPV and p97
B-2-1 infected Sf9 cells using the AcVx mAb confirms these studies since only cell
surface proteins can be detected by this method. Finally, it was noted earlier that p97
B-2-1 infected Sf9 cells showed heterogeneity in terms of the amount of p97 on the cell
surface. The proposal that this was related to the variable states of "healthiness" of
virus infected cells at the time of and during the infection is supported by the fact that
gp64, a wild-type viral protein, also shows heterogeneous expression in AcMNPV and
p97 B-2-1 infected cells.
74
To ensure that recombinant p97 expressed in infected Sf9 cells was not merely
associated with a glipiated protein but was attached directly to a GPI-anchor, another
analysis was performed based on the premise that p97 still attached to a GPI-anchor
will separate into the detergent phase upon TX-114 extraction due to its amphiphilic
nature, while p97 that has been cleaved by PI-PLC will separate into the aqueous phase
due to the removal of the GPI-tail. When p97 from the detergent phase of p97 B-2-1
infected Sf9 cell lysates was treated with PI-PLC, it separated into the aqueous phase
upon re-extraction with TX-114, indicating that it is directly linked to a GPI-anchor.
The small amount of p97 remaining in the detergent phase after PI-PLC treatment
correlates with the PI-PLC-resistant p97 remaining on the surface of p97 B-2-1 infected
cells in the FACS experiment (see above).
The successful expression of a GPI-anchored protein in the baculovirus system
reveals that the cellular machinery involved in this posttranslational modification is
conserved in the Sf9 insect cell line. This finding is supported by the existence of a
GPI-linked AChE in Musca domestica (house fly) andDrosophila melanogaster (Fournier
et al., 1988). The characterization of the GPI-anchor in this system will allow the
baculovirus expression system to be combined with a powerful new approach to protein
purification. The C-terminal signal sequence for GPI-linkage from human DAF and
human placental alkaline phosphatase (HPAP or PLAP) have been used to produce
GPI-anchored fusion proteins (Caras et al., 1987; Lin et al., 1990; Moran and Caras,
1991; Scheirle et al., 1992). Kennard et al. have designed a method of harvesting
glipiated proteins whereby cells grown in suspension are separated from the growth
75
medium and treated with PI-PLC in a small volume of PBS. The protein of interest is
released into a small volume in which the only contaminants are other GPI-linked
proteins and the small amount of bacterial PI-PLC used, allowing one to recover the
desired protein at a purity of about 30% (compared to about 1% purity in serum-free
medium) and at a concentration of approximately 30 jig/ml (compared to 1-2 ng/ml for
many secreted proteins). Thus, by combining these methods, one could express very
large amounts of a glipiated fusion protein in recombinant baculovirus infected Sf9 cells
and harvest the glipiated protein by PI-PLC treatment of the cells in a small volume of
PBS, resulting in a high recovery of relatively pure protein.
76
CONCLUDING STATEMENT:
The production of GPI-anchored and potentially secreted p97 in the
baculovirus/insect cell system was successfully completed. There are few studies on the
expression of glipiated proteins in evolutionarily diverse systems. This study will add
to the growing knowledge in this area. It will be interesting to further analyze the
secreted form of p97 in the baculovirus system. As well, a more detailed study of the
glycosylation will determine if differences in this modification can account for the
differences in size between native and recombinant p97. Finally, the binding of iron,
and perhaps other metal ions, to recombinant p97 will have to be determined in the
search for a physiological role for this protein.
77
REFERENCES:
Alemany, R., Vila, M.R., Franci, C, Egea, G., Real, F.X., and Thomson, T.M. (1993).
Glycosyl phosphatidylinositol membrane anchoring of melanotransferrin (p97): apical
compartmentalization in intestinal epithelial cells. J. Cell Sci. 104, 1155-1162.
Baker, E.N., Rumball, S.V., and Anderson, B.F. (1987). Transferrins: insights into
structure and function from studies on lactoferrin. Trends Biochem. Sci. 12, 350-353.
Baker, E.N., Baker, H.M., Smith, C.A., Stebbins, M.R., Kahn, M., Hellstrom, K.E., and
Hellstrom, I. (1992). Human melanotransferrin (p97) has only one functional
iron-binding site. FEBS Lett. 298, 215-218.
Blissard, G.W. and Rohrmann, G.F. (1990). Baculovirus diversity and molecular biology.
Annu. Rev. Entomol. 35, 127-155.
Bordier, C. (1981). Phase separation of integral membrane proteins in Triton X-114
solution. J. Biol. Chem. 256, 1604-1607.
Brent, R. (1993). Protein expression. In Current Protocols in Molecular Biology. F.M.
Ausubel, R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A. Smith, and K.
Struhl, eds. (New York: Greene Publishing Associates, Inc. and John Wiley & Sons,
78
Inc.), pp. 16.0.5-16.0.6.
Brown, J.P., Wright, P.W., Hart, C.E., Woodbury, R.G., Hellstrom, K.E., and Hellstrom,
I. (1980). Protein antigens of normal and malignant human cells identified by
immunoprecipitation with monoclonal antibodies. J. Biol. Chem. 255, 4980-4983.
Brown, J.P., Hellstrom, K.E., and Hellstrom, I. (1981a). Use of monoclonal antibodies
for quantitative analysis of antigens in normal and neoplastic tissues. Clin. Chem. 27,
1592-1596.
Brown, J.P., Nishiyama, K., Hellstrom, I., and Hellstrom, K.E. (1981b). Structural
characterization of human melanoma-associated antigen p97 with monoclonal
antibodies. J. Immunol. 127, 539-546.
Brown, J.P., Woodbury, R.G., Hart, C.E., Hellstrom, I., and Hellstrom, K.E. (1981c).
Quantitative analysis of melanoma-associated antigen p97 in normal and neoplastic
tissues. Proc. Natl. Acad. Sci. USA 78, 539-543.
Brown, J.P., Hewick, R.M., Hellstrom, I., Hellstrom, K.E., Doolittle, R.F., and Dreyer,
W.J. (1982). Human melanoma-associated antigen p97 is structurally and functionally
related to transferrin. Nature 296, 171-173.
79
Brown, J.P., Rose, T.M., and Plowman, G.D. (1985). Human melanoma antigen p97,
a membrane-associated transferrin homologue. In Proteins of Iron Storage and
Transport. G. Spik, J. Montreuil, R.R. Crichton, and J. Mazurier, eds. (New York:
Elsevier Science Publishers B.V. (Biomedical Division)), pp. 39-46.
Butters, T.D., Hughes, R.C., and Vischer, P. (1981). Steps in the biosynthesis of
mosquito cells membrane glycoproteins and the effects of tunicamycin. Biochim.
Biophys. Acta 640, 672-686.
Caras, I.W., Weddell, G.N., Davitz, M.A., Nussenzweig, V., and Martin Jr., D.W.
(1987). Signal for attachment of a phospholipid membrane anchor in decay accelerating
factor. Science 238, 1280-1283.
Chapman, A., Fujimoto, K., and Kornfeld, S. (1980). The primary glycosylation defect
in class E Thy-1-negative mutant mouse lymphoma cells is an inability to synthesize
dolichol-P-mannose. J. Biol. Chem. 2551, 4441-4446.
Collins, S.H. (1990). Production of secreted proteins in yeast. In Protein Production by
Biotechnology. T.J.R. Harris, ed. (New York: Elsevier Science Publishing Co., Inc.), pp.
61-77.
Davidson, D.J., Fraser, M.J., and CasteUino, F.J. (1990). Oligosaccharide processing in
80
the expression of human plasminogen cDNA by Lepidopteran insect (Spodoptera
frugiperda) cells. Biochemistry 29, 5584-5590.
Davidson, D.J. and Castellino, FJ. (1991a). Structures of the asparagine-289-linked
oligosaccharides assembled on recombinant human plasminogen expressed in a
Mamestra brassicae cell line (IZD-MBO503). Biochemistry 30, 6689-6696.
Davidson, D.J. and Castellino, F.J. (1991b). Asparagine-linked oligosaccharide
processing in Lepidopteran insect cells. Temporal dependence of the nature of the
oligosaccharides assembled on asparagine-289 of recombinant human plasminogen
produced in baculovirus vector infected Spodoptera frugiperda (IPLB-SF-21AE) cells.
Biochemistry 30, 6161-611A.
DeGasperi, R., Thomas, L.J., Sugiyama, E., Chang, H.M., Beck, P.J., Orlean, P.,
Albright, C, Waneck, G., Sambrook, J.F., Warren, CD., and Yeh, E.T.H. (1990).
Correction of a defect in mammalian GPI anchor biosynthesis by a transfected yeast
gene. Science 250, 988-991.
Dippold, W.G., Lloyd, K.O., Li, L.T.C., Ikeda, H., Oettgen, H.F., and Old, L.J. (1980).
Cell surface antigens of human malignant melanoma: Definition of six antigenic systems
with mouse monoclonal antibodies. Proc. Natl. Acad. Sci. USA 77, 6114-6118.
81
Doering, T.L., Masterson, W.J., Englund, P.T., and Hart, G.W. (1989). Biosynthesis of
the glycosyl phosphatidylinositol membrane anchor of the trypanosome variant surface
glycoprotein.
Origin of the non-acetylated glucosamine. J. Biol. Chem. 264,
11168-11173.
Doering, T.L., Masterson, W.J., Hart, G.W., and Englund, P.T. (1990). Biosynthesis of
glycosyl phosphatidylinositol membrane anchors. J. Biol. Chem. 265, 611-614.
Domingo, D.L. and Trowbridge, I.S. (1988). Characterization of the human transferrin
receptor produced in a baculovirus expression system. J. Biol. Chem. 263,13386-13392.
Estin, CD., Stevenson, U., Kahn, M., Hellstrom, I., and Hellstrdm, K.E. (1989).
Transfected
mouse melanoma lines that express various levels of human
melanoma-associated antigen p97. J. Natl. Cancer Inst. 81, 445-448.
Fatemi, S.H. and Tartakoff, A.M. (1986). Hydrophilic anchor-deficient Thy-1 is secreted
by a class E mutant T lymphoma. Cell 46, 653-657.
Ferguson, M.A. and Williams, A.F. (1988). Cell-surface anchoring of proteins via
glycosyl-phosphatidylinositol structures. Annu. Rev. Biochem. 57, 285-320.
Food, M.R., Rothenberger, S., Gabathuler, R., Haidl, I.D., Reid, G., and Jefferies,
82
W.A. (1993). Transport and expression in human melanomas of a transferrin-like
glycosylphosphatidylinositol anchored protein. J. Biol. Chem. in press.
Fournier, D., Berge, J.-B., Cardoso de Almeida, M.-L., and Bordier, C. (1988).
Acetylcholinesterases from Musca domestica and Drosophila melanogaster brain are
linked to membranes by a glycophospholipid anchor sensitive to an endogenous
phospholipase. J. Neurochem. 50, 1158-1163.
Fournier, D., Mutero, A., and Rungger, D. (1992). Drosophila acetylcholinesterase.
Expression of a functional precursor in Xenopus oocytes. Eur. J. Biochem. 203, 513-519.
Gabathuler, R., Levy, F., and Kvist, S. (1990). Requirements for the association of
Adenovirus Type 2 E3/19K wild-type and mutant proteins with HLA antigens. J. Virol.
64, 3679-3685.
Garratt, R.C. and Jhoti, H. (1992). A molecular model for the tumour-associated
antigen, p97, suggests a Zn-binding function. FEBS Lett. 305, 55-61.
Garrigues, H.J., Tilgen, W., Hellstrom, I., Franke, W., and Hellstrom, K.E. (1982).
Detection of a human melanoma-associated antigen, p97, in histological sections of
primary human melanomas. Int. J. Cancer 29, 511-515.
83
Gerber, L.D., Kodukula, K., and Udenfriend, S. (1992). Phosphatidylinositol glycan
(PI-G) anchored membrane proteins. Amino acid requirements adjacent to the site of
cleavage and PI-G attachment in the COOH-terminal signal peptide. J. Biol. Chem. 267,
12168-12173.
Hohmann, A.W. and Faulkner, P. (1983). Monoclonal antibodies to baculovirus
structural proteins: Determination of specificities by Western blot analysis. Virology 125,
432-444.
Homans, S.W., Ferguson, M.A.J., Dwek, R.A., Rademacher, T.W., Anand, R., and
Williams, A.F. (1988). Complete structure of the glycosyl phosphatidylinositol
membrane anchor of rat brain Thy-1 glycoprotein. Nature 333, 269-272.
Hsieh, P. and Robbins, P.W. (1984). Regulation of asparagine-linked oligosaccharide
processing. Oligosaccharide processing in Aedes albopictus mosquito cells. J. Biol. Chem.
259, 2375-2382.
Jarvis, D.L. and Summers, M.D. (1989). Glycosylation and secretion of human tissue
plasminogen activator in recombinant baculovirus-infected insect cells. Mol. Cell. Biol.
9, 214-223.
Kaufman, R.J. (1993a). Expression of proteins in mammalian cells. Overview of protein
84
expression in mammalian cells. In Current Protocols in Molecular Biology. F.M.
Ausubel, R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A. Smith, and K.
Struhl, eds. (New York: Greene Publishing Associates, Inc. and John Wiley & Sons,
Inc.), pp. 16.12.1-16.12.6.
Kaufman, R.J. (1993b). Amplification using CHO cell expression vectors. In Current
Protocols in Molecular Biology. F.M. Ausubel, R. Brent, R.E. Kingston, D.D. Moore,
J.G. Seidman, J.A. Smith, and K. Struhl, eds. (New York: Greene Publishing Associates,
Inc. and John Wiley & Sons, Inc.), pp. 16.14.1-16.14.13.
Keddie, B.A. and Volkman, L.E. (1985). Infectivity difference between the two
phenotypes of Autographa californica nuclear polyhedrosis virus: Importance of the 64K
envelope glycoprotein. J. Gen. Virol. 66, 1195-1200.
Kennard, M.L., Food, M.R., Jefferies, W.A, and Piret, J.M. (1993). Controlled release
process to recover heterologous glycosylphosphatidylinositol membrane anchored
proteins from CHO cells. Biotechnol. Bioeng. 42, 480-486.
Khosravi, M.J., Dent, P.B., and Liao, S.-K. (1985). Structural characterization and
biosynthesis of gp87, a melanoma-associated oncofetal antigen defined by monoclonal
antibody 140.240. Int. J. Cancer 35, 73-80.
85
Klausner, R.D., Ashwell, G., van Renswoude, J., Harford, J.B., and Bridges, K.R.
(1983). Binding of apotransferrin to K562 cells: Explanation of the transferrin cycle.
Proc. Natl. Acad. Sci. USA 80, 2263-2266.
Kodukula, K., Gerber, L.D., Amthauer, R., Brink, L., and Udenfriend, S. (1993).
Biosynthesis of glycosylphosphatidylmositol (GPI)-anchored membrane proteins in intact
cells: Specific amino acid requirements adjacent to the site of cleavage and GPI
attachment. J. Cell Biol. 120, 657-664.
Kornfeld, R. and Kornfeld, S. (1985). Assembly of asparagine-linked oligosaccharides.
Annu. Rev. Biochem. 54, 631-664.
Kornfeld, S., Gregory, W., and Chapman, A. (1979). Class E Thy-1-negative mouse
lymphoma cells utilize an alternate pathway of oligosaccharide processing to synthesize
complex-type oligosaccharides. J. Biol. Chem. 254, 11649-11654.
Liao, S.-K., Clarke, B.J., Khosravi, M.J., Kwong, P.C., Brickenden, A , and Dent, P.B.
(1982). Human melanoma-specific oncofetal antigen defined by a mouse monoclonal
antibody. Int. J. Cancer 30, 573-580.
Liao, S.-K., Kwong, P.C., and Khosravi, M.J. (1985).
Immunopurification,
characterization, and nature of membrane association of human melanoma-associated
86
oncofetal antigen gp87 defined by monoclonal antibody 140.240. J. Cell. Biochem. 27,
303-316.
Licari, P. and Bailey, J.E. (1991). Factors influencing recombinant protein yields in an
insect cell-baculovirus expression system: Multiplicity of infection and intracellular
protein degradation. Biotechnol. Bioeng. 37, 238-246.
Lin, A.Y., Devaux, B., Green, A , Sagerstrom, C, Elliott, J.F., and Davis, M.M. (1990).
Expression of T cell antigen receptor heterodimers in a lipid-linked form. Science 249,
677-679.
Lippe, R., Luke, E., Kuah, Y.T., Lomas, C, and Jefferies, W.A. (1991). Adenovirus
infection inhibits the phosphorylation of Major Histocompatibility Complex Class I
proteins. J. Exp. Med. 174, 1159-1166.
Low, M.G., Stiernberg, J., Waneck, G.L., Flavell, R.A, and Kincade, P.W. (1988).
Cell-specific heterogeneity in sensitivity of phosphatidylinositol-anchored membrane
antigens to release by phospholipase C. J. Immunol. Methods 113, 101-111.
Low, M.G. (1989). The glycosyl-phosphatidylinositol anchor of membrane proteins.
Biochim. Biophys. Acta 988, 427-454.
87
Low, M.G.
and
Saltiel, A.R.
(1988). Structural
and
functional
roles of
glycosyl-phosphatidylinositol in membranes. Science 239, 268-275.
Luckow, V.A. (1991). Cloning and expression of heterologous genes in insect cells with
baculovirus vectors. In Recombinant DNA Technology and Applications. A. Prokop,
R.K. Bajpai, and C.S. Ho, eds. (New York: McGraw-Hill, Inc.), pp. 97-152.
Luckow, V.A. and Summers, M.D. (1989). High level expression of nonfused foreign
genes with Autographa catifornica nuclear polyhedrosis virus expression vectors. Virology
170, 31-39.
Masterson, W.J., Doering, T.L., Hart, G.W., and Englund, P.T. (1989). A novel pathway
for glycan assembly: Biosynthesis of the glycosyl-phosphatidylinositol anchor of the
trypanosome variant surface glycoprotein. Cell 56, 793-800.
Menon,
A.K.,
Mayor,
glycosyl-phosphatidylinositol
S.,
and
lipids
Schwarz,
in
R.T.
(1990).
Biosynthesis
of
Trypanosoma brucei: involvement
of
mannosyl-phosphoryldolichol as the mannose donor. EMBO J. 9, 4249-4258.
Micanovic, R., Gerber, L.D., Berger, J., Kodukula, K., and Udenfriend, S. (1990).
Selectivity of the cleavage/attachment site of phosphatidylinositol-glycan-anchored
membrane proteins determined by site-specific mutagenesis at Asp-484 of placental
88
alkaline phosphatase. Proc. Natl. Acad. Sci. USA 87, 157-161.
Miller, L.K. (1988). Baculoviruses as gene expression vectors. Annu. Rev. Microbiol. 42,
177-199.
Moran, P., Raab, H., Kohr, W.J., and Caras, I.W. (1991). Glycophospholipid membrane
anchor attachment. Molecular analysis of the cleavage/attachment site. J. Biol. Chem.
266, 1250-1257.
Moran, P. and Caras, I.W. (1991). Fusion of sequence elements from non-anchored
proteins to generate a fully functional signal for glycophosphatidylinositol membrane
anchor attachment. J. Cell Biol. 115, 1595-1600.
Moran, P. and Caras, I.W. (1992). Proteins containing an uncleaved signal for
glycophosphatidylinositol membrane anchor attachment are retained in a post-ER
compartment. J. Cell Biol. 119, 763-772.
Plowman, G.D., Brown, J.P., Enns, C.A., Schroder, J., Nikimnaa, B., Sussman, H.H.,
Hellstrom, K.E., and Hellstrom, I. (1983). Assignment of the gene for human
melanoma-associated antigen p97 to chromosome 3. Nature 303, 70-72.
Quesada Allue, L.A. (1980). The biosynthesis of glucose containing insect lipid linked
89
oligosaccharide and its possible role in glycoprotein assembly. Mol. Cell. Biochem. 33,
149-155.
Radic, Z., Gibney, G., Kawamoto, S., MacPhee-Quigley, K., Bongiorno, C, and Taylor,
P. (1992). Expression of recombinant acetylcholinesterase in a baculovirus system:
Kinetic properties of glutamate 199 mutants. Biochemistry 31, 9760-9767.
Real, F.X., Houghton, A.N., Albino, A.P., Cordon-Cardo, C, Melamed, M.R., Oettgen,
H.F., and Old, L.J. (1985). Surface antigens of melanomas and melanocytes defined by
mouse monoclonal antibodies: Specificity analysis and comparison of antigen expression
in cultured cells and tissues. Cancer Res. 45, 4401-4411.
Real, F.X., Furukawa, K.S., Mattes, M.J., Gusik, S.A., Cordon-Cardo, C, Oettgen, H.F.,
Old, L.J., and Lloyd, K.O. (1988). Class 1 (unique) tumor antigens of human melanoma:
Identification of unique and common epitopes on a 90-kDa glycoprotein. Proc. Natl.
Acad. Sci. USA 85, 3965-3969.
Rearick, J.I., Chapman, A., and Kornfeld, S. (1981). Glucose starvation alters
lipid-linked oligosaccharide biosynthesis in Chinese hamster ovary cells. J. Biol. Chem.
256, 6255-6261.
Richardson, D.R. and Baker, E. (1990). The uptake of iron and transferrin by the
90
human malignant melanoma cell. Biochim. Biophys. Acta 1053, 1-12.
Richardson, D.R. and Baker, E. (1991a). The release of iron and transferrin from the
human melanoma cell. Biochim. Biophys. Acta 1091, 294-302.
Richardson, D.R. and Baker, E. (1991b). The uptake of inorganic iron complexes by
human melanoma cells. Biochim. Biophys. Acta 1093, 20-28.
Richardson, D.R. and Baker, E. (1992a). The effect of desferrioxamine and ferric
ammonium citrate on the uptake of iron by the membrane iron-binding component of
human melanoma cells. Biochim. Biophys. Acta 1103, 275-280.
Richardson, D.R. and Baker, E. (1992b). Two mechanisms of iron uptake from
transferrin by melanoma cells. J. Biol. Chem. 267, 13972-13979.
Rose, T.M., Plowman, G.D., Teplow, D.B., Dreyer, W.J., Hellstrom, K.E., and Brown,
J.P. (1986). Primary structure of the human melanoma-associated antigen p97
(melanotransferrin) deduced from the mRNA sequence. Proc. Natl. Acad. Sci. USA 83,
1261-1265.
Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular Cloning. A Laboratory
Manual (Cold Spring Harbor: Cold Spring Harbor Laboratory Press).
91
Schein, C.H. (1989). Production of soluble recombinant proteins in bacteria.
Bio/Technology 7, 1141-1149.
Scheirle, A., Takacs, B., Kremer, L., Marin, F., and Sinigaglia, F. (1992). Peptide
binding to soluble HLA-DR4 molecules produced by insect cells. J. Immunol. 149,
1994-1999.
Sciot, R., De Vos, R., van Eyken, P., van der Steen, K., Moerman, P., and Desmet, V.J.
(1989). In situ localization of melanotransferrin (melanoma-associated antigen p97) in
human liver. A light- and electronmicroscopic immunohistochemical study. Liver 9,
110-119.
Scott, M.R.D., Butler, D.A., Bredesen, D.E., Walchli, M., Hsiao, K.K., and Prusiner,
S.B. (1988). Prion protein gene expression in cultured cells. Protein Eng. 2, 69-76.
Seligman, P.A., Butler, CD., Massey, E.J., Kaur, J.A., Brown, J.P., Plowman, G.D.,
Miller, Y., and Jones, C. (1986). The p97 antigen is mapped to the q24-qter region of
chromosome 3; the same region as the transferrin receptor. Am. J. Hum. Genet. 38,
540-548.
Summers, M.D. and Smith, G.E. (1987). A manual of methods for baculovirus vectors
92
and insect cell culture procedures. Tex. Agric. Exp. Stn. Bull. 1555,
Takami, N., Oda, K., and Ikehara, Y. (1992). Aberrant processing of alkaline
phosphatase precursor caused by blocking the synthesis of glycosylphosphatidylinositol.
J. Biol. Chem. 267, 1042-1047.
Volkman, L.E., Goldsmith, P.A., Hess, R.T., and Faulkner, P. (1984). Neutralization of
budded Autographa califomica NPV by a monoclonal antibody; Identification of the
target antigen. Virology 133, 354-362.
Volkman, L.E. and Goldsmith, P.A. (1985). Mechanism of neutralization of budded
Autographa califomica nuclear polyhedrosis virus by a monoclonal antibody: Inhibition
of entry by adsorptive endocytosis. Virology 143, 185-195.
Webb, N.R., Madoulet, C, Tosi, P.-F., Broussard, D.R., Sneed, L., Nicolau, C, and
Summers, M.D. (1989). Cell-surface expression and purification of human CD4
produced in baculovirus-infected insect cells. Proc. Natl. Acad. Sci. USA 86, 7731-7735.
Whitford, M., Stewart, S., Kuzio, J., and Faulkner, P. (1989). Identification and
sequence analysis of a gene encoding gp67, an abundant envelope glycoprotein of the
baculovirus Autographa califomica nuclear polyhedrosis virus. J. Virol. 63, 1393-1399.
93
Woodbury, R.G., Brown, J.P., Yeh, M., Hellstrom, I., and Hellstrom, K.E. (1980).
Identification of a cell surface protein, p97, in human melanomas and certain other
neoplasms. Proc. Natl. Acad. Sci. USA 77, 2183-2187.
Woodbury, R.G., Brown, J.P., Loop, S.M., Hellstrom, K.E., and Hellstrom, I. (1981).
Analysis of normal neoplastic human tissues for the tumor-associated protein p97. Int.
J. Cancer 27, 145-149.
Yamashita, K., Kamerling, J.P., and Kobata, A. (1983). Structural studies of the sugar
chains of hen ovomucoid. Evidence indicating that they are formed mainly by the
alternate biosynthetic pathway of asparagine-linked sugar chains. J. Biol. Chem. 258,
3099-3106.