Note: Descriptions are shown in the official language in which they were submitted.
WO 92/01715 j PCr/US91/0~082
-- 20875~
SOLUBLE CELL-SURFACE DIMERIC PROTEI~S
s
INTRODUCTION
Technical Field
The field of this invention is preparation of
soluble cell surface poly(sub-unit) proteins as soluble
entities.
BACKGROUND OF THE INVENT~ON
Cells are continuously interacting with their
environment, recei~ing cues concerning the state of the
environment and signals associated with the response of
the cell to the environment. In a multicellular host,
the cells interact with the environment, influencing the
environment by secreting various compositions, removing
- compositions from the en~ironment, and in some instances
. 20 moving away from or toward a particular environment.
Many different mechanisms are used for the cell to
receive a signal from the environment and translate that
signal into a response.
In many cases, the cell relies upon a surface
membrane protein, which may be a single chain or a
plurality of chains or in some instances, for the
purposes of transducing a signal, two or more
independent proteins may be associated. Proteins which
respond to hormones, cytokines, glucocorticoid steroids,
' 30 antigens, and other surface membrane receptors, are only
of the few of the surface membrane proteins present on a
mammalian cell.
For many purposes, it would be useful to have
- the surface membrane protein in a soluble form free of
3s the cell membrane or microsome. In this way, the nature
of binding of the protein to its ligand or other protein
could be studied. In addition, the soluble form of the
surface Dembrane protein may be used prophylactically or
therapeutically, as an aqonist o~ antagonist, to induce
ç
, ~ , ~ ............................ :
., ;. ~ . ~ ~ .. :
WO92/01715 2 0 ~ 7 5 4 5 PCT/US91/~082
or inhibit the`transduction of a signal.
A substantial proportion of the surface membrane
proteins are bound to the surface by means of a
hydrophobic domain present in the protein. By having a
s combination of a signal sequence at the N-~erminus and a
transmembrane integrator sequence or domain as an
internal portion of the protein, the transport of the
membrane protein through the membrane is arrested at the
transmembrane integrator sequence. Furthermore, for
many surface membrane proteins, the protein is poly(sub-
unit), having two or more sub-units, which may be
covalently or non-covalently bound together. The
association of the sub-units in the cell and their
transport to the surface is not understood. Therefore,
S in making changes in the protein, in order to provide it
in soluble form, it is not at all clear that
modifications of the protein will provide for proper
folding and proper association and processing, if the
protein is retained intracellularly or is secreted. It
is therefore of substantial interest to find methods for
producing proteins having a plurality of sub-units in a
form where they are properly folded and associated and
may be readily isolated in soluble form.
Relevant Literature
T cell Receptor (TCR) heterodimers are assembled
with the CD3 polypeptide to form complexes of at least
seven polypeptides before appearing in the surface
(Minamiel et al., ~o~. Natl. Acad. Sci. USA (1987)
84:2688). TCR (V)-Ig(C) hybrids expressed in myeloma
cell lines have been described (Gascoigne et al., ibid
(1987) 84:2936). Only the V~ (TCR) C~(Ig) chimeras were
assembled or secreted. Lipid-linked surface proteins
form a significant class of surface membrane proteins
3s (Ferguson and Williams, Ann. Rev. Biochem. (1988)
; 57:285). Caras et al., Science (1987) ~38:1280, has
shown that the carboxy-terminal thirty-seven imino acids
" . ,.,., ~
WO92/0171~ PCT~US91/05082
2087~
of decay accelerating factor could serve as the signal
sequence for the lipid-linked expression of a herpes
simplex virus membrane protein. The carboxyl terminus
of human placental alkaline phosphatase, a homodimer,
s has been shown to be a lipid-linked molecule
(~an et ~1., Proc. ~atl. Acad. Sci. USA (1985) 82:871S).
Attempts to use TCR-Ig chimeras have been reported by
Traunecker et al., Immunol. ~odav (1989) 10:29.
o SUMMARY OF THE ~NVENTION
Methods are provided for producing in soluble
form multi(sub-unit) surface membrane proteins. The
resulting proteins may be used in the study of their
interactions with their ligands, other surface membrane
ts proteins, or as agonists or antagonists for the
interaction of the naturally occurring surface membrane
protein. The technology is exemplified with the T cell
receptor.
:
DESCRIPTION OF T~E SPECIFIC EMBODIMENTS
The subject invention concerns solubilizing
surface membrane proteins of the mammalian hosts. The
surface membrane proteins are characterized by being
bound to the membrane by a hydrophobic sequence,
- 2s normally having charged or polar amino acids at its
boundaries. The proteins will also have a plurality of
sub-units, where one or more of the sub-units will
comprise the hydro-phobic sequence known as a
transmembrane integrator sequence or domain. In
~o addition, the surface me~brane protein will have
`~ translocation signals, such as signal sequences at the
- - N-terminus of the protein. The signal sequences will
direct the translocation of the sub-units with
~,. processing from the Golgi apparatus through the
`~ 3s membrane, where the signal seguence will normally be
-~ cleaved during the processing to produce a protein sub-
unit free of the signal sequence.
.~ ..
.~ .
.
. , .
, - ' ' ,,
' :', ' '-': - , -
.. ~
.
- . ~ : , . ---
,: ~ . . . . .
- - . .
.
~,
WO92/01715 PCT/US91/~082
20~7~45
The proteins of interest will generally have
from about 2 to 8, usually 2 to 6, more usually 2 to 4,
sub-units. Desirably, each of the sub-units will have
not more than 2, usually not more than l transmembrane
integrator sequence, so that the sub-unit passes through
the surface membrane only once. The size of the sub-
units is not critical to thiC invention and may vary
from at least about lOkDal (kiloDaltons) and is usually
not more than lOOOkDal, usually not more than 600kDal,
more usually not more than 200kDal. In many cases, the
sub-unit protein may be associated with other proteins
at the membrane surface, which proteins may be involved
with the formation and transport of the multi-sub-unit
proteins.
A wide variety of surface membrane proteins
fulfill the above requirements, particularly receptors,
enzymes, and the like. These proteins include T cell
receptors, surface immunoglobulins, major
histocompatibility complex antigens, both Class I and
II, hormone receptors, G proteins, etc.
In order to solubilize these proteins, they will
be modified by removing at least in part the
transintegrator membrane sequence of each subunit and
providing for a signal which results in the attachment
of a lipid to the sub-unit and transport of the sub-unit
to the surface membrane where it is non-covalently bound
to the surface membrane.
The particular signal sequence may be from any
convenient source which is functional in the expression
host, so that it may be endogenous or exogenous to the
expression host or common or foreign to the source of
the surface membrane protein. The signal sequence may
be associated with a surface membrane protein which has
a single unit or a multiplicity of sub-units, preferably
~5 from a protein which has a multiplicity of sub-units.
Illustrative signal sequences are derived from such
proteins as decay accelerating factor, placental
-
'
WO92~0171~ PCT/US91/05082
~7~
alkaline phosphatase, and the like. The sequence
providing for lipid attachment will generally include a
sequence which has from about l0 to 50, more usually
from about 15 to 30 amino acids which are cleaved from
s the carboxyl terminus of the sub-unit precursor and an
ethanolamine-carbshydrate-phosphatidylinositol linked to
the new carboxyl terminus. Once the precursor has been
processed in this way, it is then transferred to the
cell surface.
The cells may be isolated containing the
modified sub-units anchored to the surface membrane by
means of lipid and the various portions of the anchor
removed individually or together. Pronase cleaves at
the penultimate peptide of the amino acid bonded to the
C-terminal amino acid which is bonded to ethanolamine.
Nitrous acid may be used to cleave a glycosidic linkage
between an amino sugar, glucosamine and another sugar
inositol. Phosphatidylinositol-phospholipase C(PI-PLC)
may be used to cleave at the phosphate linkage between
~o the diacylglycerol and inositol phosphate.
Usually, the phosphatidyl inositol anchor signal
will only be partially removed, leaving from about l0 to
30 amino acids remaining at the carboxyl terminus of the
sub-unit of interest. The sequence provides a
convenient tag for isolation and binding. Where this
: additional sequence does not interfere with the purpose
of the membrane protein, it may be left with the
protein. Alternatively, by using carboxypeptidase, some
or all of the additional amino acids may be removed.
Various techniques exist for removing amino acids from
. the C- terminus of a protein individually. While this
is not convenient, in some individual cases this may
prove to be expedient. Preferably, a sequence may be
introduced between the phosphatidyl inositol anchor
3s sequence and the sub-unit sequence which provides a
signal for a peptidase enzyme.
A wide variety of sequences are available which
.... , : .. .
. ~
. : .
- . ., ', : . ~ .' . . ' :. .
.~ . . , - . , .
: : - . ~ : . .
WO92/0171~ PCT/US91/~082
~7;~
are specifically recognized by enzymes, where the
sequence is not encountered in the sub-unit sequence of
interest and the enzyme is not expected to be
encountered in the expression host. For example, the
s peptidase signal sequence of the alpha- or a-protein of
yeast, the Kex enzyme, or a sequence recognized by any
other convenient enzyme, may be used to provide for
removal of the peptidase signal and the phosphatidyl
inositol anchor sequence from the sub-unit.
o In providing for expression of the individual
sub-units, constructs will be prepared, where the sub-
unit is modified, as appropriate, by removal of at least
a portion of the transmembrane integrator sequence and
the cytoplasmic sequence. The sequences will be
replaced by the phosphatidyl inositol anchor sequence
from the appropriate protein. The exchange may be
achieved in a variety of ways. The sub-unit ~ene may be
cloned and by employing n vitro mutagenesis or primer
repair, the transmembrane integrator and cytoplasmic
sequences may be removed and in the case of n vitro
mutagenesis, replaced with the phosphatidyl inositol
. anchor sequence. Depending upon the size of the gene,
one may use the polymerase chain reaction, where the
primers comprise the phosphatidyl inositol anchor region
and a sequence of at least l0 nucleotides, preferably at
. least about 18 nucleotides, which are complementary to
the sequence of the sub-unit gene 5' prime of the
...... ....... .... transmemhrane integrator sequence. Other manipulations
: may include restriction at a site proximal, preferably
S'-proximal, to the transmembrane integrator sequence
and then insertion of a sequence comprising the
phosphatidyl inositol anchor sequence at the 3' terminus
~ of`the truncated sub-unit gene. After ligation, the
chimeric construct may then be cloned, the DNA isolated
3s and used for expression. Of course, the sequence which
includes the phosphatidyl inositol anchor sequence will
also include a peptid~se signal sequence at its 5'-
.
.: . . . . .
.
. - ' , ' :
. . , ~ '
~ WO92/01715 PCT/US91/~2
2~8~
terminus for cleavage at that site by t~e peptidase and
removal of the extraneous amino acids.
The gene for the sub-unit may be the genomic
gene or cDNA, preferably cDNA. The genomic gene, to the
s extent that it includes introns, will not be as
manageable during the various constructs and
manipulations. Where this is not a problem, the genomic
gene may find preference in providing for a higher level
of expression. Normally, the gene will be known and
sequenced, so as to allow for ready manipulation,
identification of restriction sites, and ease of
introduction and removal from vectors. The sub-unit
gene may be isolated, cloned in an appropriate cloning
vector and then manipulated by various techniques as
described above. See, Maniatis et al., A Laboratory
Manual, Cold Spring Harbor Laboratory, Cold Spring
Harbor, NY. The transme~brane integrator and
cytoplasmic sequences may be removed and replaced with
the phosphatidyl inositol anchor sequence as obtained
from an appropriate gene. By appropriate manipulations,
overhangs or blunt ends may be provided for ligation of
the two sequences in proper reading frame. The
construct may then be cloned and analyzed by any
convenient means, including restriction analysis,
~ Z5 sequencing, hybridization, or the like. Once the
desired construct has been prepared, it may then be used
in an appropriate expression vector in an appropriate
expression host.
A wide variety of hosts are available, but in
view of the desirability for processing and assembling
of the surface membrane protein, normally an appropriate
mammalian host will be employed for expression. The
vectors which are selected will normally include a
marker for identification of those hosts into which the
s construct has been introduced, where the host may be
identified by positive selection. Thus, the marker will
usually be protection from a biocide, particularly an
.. ... .. - ,,- - .. , . ,. .. . : .
.. ,.: . , . . ' ' . ~ .
' ' ' '~ ' : ' '' '
. ., . . , . : .:
, : . . : : :
,. ' ' " '' ' ~ - '' ~ '
. ~ . . . .
W092tO171~ PCT/US91/OS082
2 087~q~ 8
antibiotic, e.g. G4~8, or in appropriate cases, where
the gene provides prototrophy to an auxotrophic host.
For the most part, the vectors will not be stable in the
expression host, so that the gene will be integrated
s into the genome of the host. However, in some
instances, it may be desirable to retain the construct
as part of an episomal element. Viral replication
systems may be employed, suc~ as those of SV40,
papilloma virus, adenovirus, etc., where the viru6es
have been atten~ated, are capable of accommodating the
construct, and are usually not lytic. Various
constructs have been described in the literature, and
need not be exemplified here.
Various techniques may be employed for
s introducing the construct into the host, such as calcium
phosphate precipitated DNA, transfection, transduction,
electroporation, fusion, etc. After the cells have been
i transformed with the construct, the cells may be grown
on an appropriate selective medium for selection of
those hosts containing the marker.
Various expression hosts have been developed for
use and expression of mammalian proteins. These hosts
include chinese hamster ovary cells, COS cells, mouse
mammary kidney cells, HeLa cells, and the like. The
~` 25 cells may be grown in an appropriate medium, where the
desired proteins are translocated to the membrane. The
host cells may then be harvested, lysed, and the
microsomes treated with pronase and the resultinq
proteins isolated. The proteins may be further
processed to remove all or part of the lipid anchor
signal sequence. The desired surface membrane protein
may be purified by affinity chromatography,-
- electrophore6is, HPLC, or the like. Once the protein
has been isolated, it may be used in a variety of ways.
3s As already indicated, the soluble proteins may
; be used in the study of the interaction with other
surface membrané proteins or ligands. For example, the
, ~ : . ' ~ . . : .
:: . . .
: - ,
,. ' ~ - .
WO92/01715 PCT/US91/~82
208;7~54~
soluble T cell receptor may be used in studying the
interaction with major histocompatibility complex
antigens, Class I or Class II, or the like.
Alternatively, they may provide for cells with varying
s levels of the surface membrane protein bound to the
surface, by varying the periods of time that the cells
are contacted with pronase, phospholipase C, nitrous
acid, or the like. In this manner, the effect of the
number of the surface membrane protein molecules on the
surface in relation to other surface membrane proteins
may be investigated.
Besides usinq the soluble proteins to study
various interactions and gain an understanding of how
the surface membrane proteins fulfill its function, the
surface membrane proteins may also be used in culture
and n v vo in various hosts. The soluble surface
membrane proteins may be used as agonists or
antagonists. For example, the soluble T cell receptor
may be used to prevent interaction between T cells and
2~ other cells. When the other cells are B lymphocytes,
production of antibodies may be prevented by preventing
the binding between the T cell receptor and the MHC
Y antigen of the B lymphocyte. With other proteins, they
may be used to identify ligands, binding to other
2s surface membrane proteins, as agonists or antagonists,
affinity columns, or the like.
The subject soluble surface membrane proteins
may be formulated in any convenient-medium for
; administration to mammalian host. ~he proteins may be
administered in buffered or unbuffered solutions, such
} as saline, phosphate buffered saline, phosphate, aqueous
ethanol, or the like. The concentration of the protein
will generally vary from about O.OOl-So 5 mg/ml. Dosage
will vary widely depending upon the particular condition
,~ 35 being treated, the method of administration, the
frequency of administration, and the like. The amount
administered would generally range from about lpg to
.
:,. . , . . .. . . .. ~ , -
-
. , :: . :
: ':
';
W092/01715 PCT/US91/OS082
~87~5 ~
5mg/Xg host.
The following examples are offered by way ofillustration and not by way of limitation.
S EXPERIMENTAL
The PI-anchored TCRs were constructed by fusing
the PI-anchor signals (DAF or HPAP) to the fifth amino
acid residue of the TCR located 3' to the last cysteine
residue before the transmembrane domain. Full length
cDNAs encoding the 2B4 TCR ~ or ~ chain8 (Chien et ~1-,
Nature (1984) 309,322; Berkin et al., Nature (1985)
317,430) were inserted 5' to the DNA segment coding for
the last 37 amino acids of DAF (residues 311-347) (Caras
et al., Nature (1987) 325,545) or the last 47 amino
acids of HPAP (residues 467-51~) (Kam et al., PNAS USA
(1985) 82:8175); Micanovic et al., ibid (1988)85:1398)
cloned in plasmid bluescript pSK-. The in frame joining
between the TCX and the PI-a~chor signal was perfor~ed
by oligonucleotide directed in vitro deletional
mutagenesis (Xunkel et al., Methods in Enzymoloqy (1987)
154,367; Amersham Ha~dbooX RPN 2322 (1986)). In all
cases, synthetic 32-oligomers that span the junctional
point with 17 nucleotides complementary to the DAF or
~ HPAP sequences and 15 to the TCR sequences were used as
: 25 primers. After n vitro mutagenesis, only the first 227
and 265 amino acid residues of the TCR ~ ~ chains
respectively were retained in the chimeric fusion
protein. The truncated ~ and ~ chains were fused to
residue 311 of DAF and~or to residue 484(~PAP-L) or
495(HPAP-S) of HPAP. The reco~binant DNAs were
sequenced after mutagenesis. The TCR ~ and ~ DAF/HPAP
genes were then inserted in an expression vector (pSR~1
Neo is-a derivative of pcDL-SR~296 (Takeke et al., ~Ql~ -
-Cell_Bio. (1988) 8:466) to which a polylinker containing
the restriction sites of
5'XhoI/XbaI/S~lI/NaeI/EcoRI/EcoRV/~indIII/ClaI is placed
between the SR~ promoter and polyadenylation site for
. ~
.
- " . . .
.
:
- . ..
~- . ,
W092/017l5 PCT/US91/~082
` 2087~
11
convenient cDNA cloning. In addition, a neomycin
resistant gene under the control of the Sv40 early
promoter is inserted between the short arm between the
ampicillin gene and the SRQ promoter containing the
s neomycin gene generating constructs respectively called
~DAF, ~HPAP-L and ~HPAP-S. Both ~ and ~ fusion genes
were under the control of the SR~l promoter.
The following are the sequences for the PI
anchor signals:
LAPPAGTTD AAHPGRS WPAnTPLLAGTLLLLETATAp HPAP-S
CLEPYTACDLAPPAGTTD AAHPGRS WPALLPLLAGTLLnr~TATAP HPAP-L
PMXGSGTTSGTTRLLSGMTC FTLTGLLGTLVTMGLLT DAF
s Co~parison of CHO transfectants co-exDressina TCR ~ and
B chains on the cell surface with various PI-anchorina
si~nals.
C~O cells were transfected with ~DAF, ~BHPAP-L
or ~HPAP-S constructs using electroporation. The G418
resistant colonies were pooled and analyzed for surface
expression of the PI-anchored TCR ~ and B chains. Cells
were seauentially stained at 4 C for 1 hour in phenol
red free RPMI+ 5~ fetal calf serum with the hamster
~` anti-mouse KJ25 antibody (anti-VB3) (2 ~g/ml) and then
2s with the biotinylated A2B4.2 (anti-V~11). FITC
` conjugated goat anti-hamster immunoglobulin and
streptavidin were used as second receptors,
respectively. After staining, the cells were subjected
to FACS analysis and the brightest 5~ double positive
cells co-expressing TC~ ~ and ~ fusion proteins were
sorted. After two weeks in culture, the pooled cells
~; were sorted a second time for the highest 5% double
expressors and grown as massive culture for further
studies. The clones were analyzed again a few weeks
; 3s later for the stability for surface expression. The
data were plotted as log fluorescent intensity in an
arbitrary Unit. After long term culture, cells
..
,~ .
.
' , . ': ' ' ' ' '.~5~ ' ': '
''' ' , ' ' ' ' , ''' ' ~
. .
WO92~01715 PCT~US91/~082
20~7~5
12
transfected with ~DAF and ~HPAP-L both c~ntain a
significant population of dull double positive cells as
well as bright double positive cells. The ~BHPAP-S
transfected cells which seemed to have stably integrated
s the TCR fusion genes were used for subsequent studies.
81QCkina of KJ25 antibody binding to the PI-anchored TCR
by A2B4.2.
5x105 Jurkat ~- cells ex,pressing a functional
284 ~ TCR after transfection with wild type 2~4~ and
genes and CH0 transfected with the ~HPAP-S construct
were pretreated with or without antibody A2B4.2
(20/~g/ml) and then sequentially incubated with antibody-
, KJ25 (V~g/ml) and a FITC conjugated goat anti-hamster
immunoglobulin. Cells were analyzed on the FACS after
staining. The data were plotted as log fluorescent
~, intensity in arbitrary units. The presence of A2B4.2
abolishes the subsequent bindin~ of KJ 25 on both
` Jurkat ~- cells and CH0 ~BHPAP-S transfected cells. TheJurkat ~- population is non-clonal which explains the
presence of negative cells ~50%) after staining with KJ
25. CH0 cells have a higher fluorescence background
' than Jurkat cells as indicated by propidium staining.
2s Purification of a soluble TCR heterodimer.
A. Summar,v of soluble ~ÇR Durification scheme.
2X108 CH0 ~HPAP-S transfected cells were seeded
into the Cell PharmI Bioreactor (CD Medical). After one
week of culture (101 cells present in the Bioreactor),
the cells were treated with 10 units of PI-PLC at 37
degrees C for three hours (Tse çt 31., ,Science (1985)
' - ~30,1003; Hon and Kincarde, Nature (1985) 318,62). The
chimeric TCR was recovered in 200 ml of growth media
pulsed into the Bioreactor. The supernatant was
s filtered, adjusted to pH 8.0 and passed over an A2B4.2
cyanogen bromide affinity column (1 ml) pre-equilibrated
; with 3 volumes of O.IM phosphate/0.15M NaCl, pH 8.0
. '
.. . . . :
-
.
WO92/01715 2 0 ~ 5 ~ ~ PCT/US9l/05082
buffer. The column was then washed with 3 volumes of
the same buffer. The protein was eluted with 3 volumes
of 0.lM NaOAc/0.15M NaCl, pH 3.5 buffer in 0.5 ml
fractions and instantly neutralized with a solution of
s 2M Tris-HCl pH 8.5 (1/10 volume added). The first four
fractions which contain most of the soluble TCR were
pooled and applied to a RJ 25 affinity column. The
protein was then eluted at pH 4.0 with the buffer
described above.
to
B. Detection o soluble TCR on silver sta ned
olvacrvlamide gels.
Ten ~1 of each fraction eluted form the A2B4.2
and RJ 25 columns were electrophoresed on a 12.5% SDS-
polyacrylamide gel under reducing or non-reducing
conditions. ~he protein gel was stained with silver
nitrate (Morrisey, Anal; Biochem. (1981) 117,307). The
protein eluted from the A2B4.2 column contains both
monomers (~, 41 Kd) and dimers (~2 and ~, 70 Kd).
Under reducing conditions, the dimers were reduced to
the size of monomers. Both 2B4 ~ and ~ chains have the
same molecular weight (41 Kd) and therefore cannot be
j~ distinguished from each other on the gel. The ~
heterodimer is separated rom ~ monomer and ~2 dimer
Z5 ater passage over the KJ 25 column,- as conirmed in C.
.
C. Immunop~ec~Pitation of the purified soluble
~ TCR.
; one ~g of purified soluble TCR eluted from
either the A2B4.2 or the W 25 column was iodinated with
- Na 125I to the specific activity of 6x107 cpm/~g using
: Enzymobeads ~Bio-rad). 2X106 cpm^of iodinated TCR was
precipitated with either A2B4.2, XJ 25 or 14-4-4s
! antibody, respect~vely. Antibody 14-4-4s which
recognizes the E~ chain of MHC class II is-used as a
- negative control. The immunoprecipitates were subjected
to electrophoresis on a 1.25~ S~S-polyacrylamide gel
.: . :.. ,: .. .
., . . .: . :~ .
.. . .
... .
,
,: ' : : - : ~ -
- .: - , .
W092/0~71~ PCT/US91/~082
2087545
under reducing (C) and non-reducing (D) conditions. T~e
gels were dried and submitted to autoradiography.
Again, a band corresponding to a 70 Xd protein can be
visualized on the autoradiogram. The intensity of the
s signal in lane 1 is more intense than in lane 2 (A2B4.2
antibody lanes 1-3) while both signals in lanes 4 and s
(KJ 25 antibody, lanes 4-6) have identical intensities.
This result indicates the preæence of not only
heterodimers t~) but also homodimers (~2) after elution
from the A2B4.2 column (lane 1). The heterodimers are
eliminated after passage of the fractions over the KJ 25
column. The ratio of ~ versus ~ molecules then becomes
identical which proves the presence of heterodimers
exclusively after passage of the fractions over both
s columns.
From the results of silver staining and immuno-
precipitation, it is estimated that the ~ 2 heterodimer
represents approximately 60% of the total PI-PLC cleaved
products or 80% of the dimers.
To verify that the ~B 2 heterodimer is the
product from the PI-PLC Cleavage, the solu~le TCR was
immuno- precipitated with two rabbit antisera anti-CRD 1
and anti-CRD 2, which recognize the phosphoglycan
epitope exposed after PI-PLC treatment. The two
: 25 antisera showed lower affinity for ~ 2 heterodimer than
2 homodimer. This is consistent with observations
with other TCR systems.
Following the above described procedures with
some modi2ications, as indicated, the PI-anchored TCRs
were constructed by fusing the PI-anchor signals (DAF or
- HPAP) to the fifth amino acid residue of the TCR located
~ 3' to the last cysteine residue before the transmembrane
domain. Full len~th cDNAs encoding the 2B4 TCR ~ or ~
chains (Chien et al., Nature (1984) 309, 322; Becker et
al., Nature (1985) 317, 430) were inserted 5' to the DNA
segment coding 20r the last 37 amino acids of DAF
(residues 311-347) (Caras et al., Nature (1987) 325,
. : ~ . - .
- . - . .
,: , . , . . ~ : : . : -
. . . : : .
, . -
:
' ' ' ' ' ' . , : .
WO92/0171~ 2 0 8 7 ~ ~ ~ PCT/US91/~82
545) or the last 38 or 47 amino acids of HPA~ (residues
476-513 and 467-513, respectively) (Kam et al., PNAS USA
(1985) 82, 8715); Micanovic, i~i~ (1988) 85, 1398 cloned
in plasmid bluescript pSX-(Stratagene~. The in-frame
s joining between the TCR and the PI-anchor signal was
performed by oligonucleotide directed ~n vitro
deletional mutagenesis (Kunkel et al., ~eth. in
En~y~ol. (1987) 154, 367). After mutagenesis, the first
227 and 265 amino acid residues of the TC~ ~ and ~ -
o chains respectively were retained in the chimeric fusion
protein. The truncated ~ and ~ chains were fused to
residue 311 of DAF and/or to residue 484 (HPAP-L) or 495
(HPAP-S) of HPAP.
A comparison was made of CH0 transfectants
coexpressing TCR ~ and ~ chains on the cell surface with
various PI-anchoring signals. CH0 cells were
transfected with ~DAF, ~HPAP-L or Q~HPAP-S constructs
(inserted into pBJ1-Neo. (The cDNA expression vector,
pBJ1-Neo, is a derivative of pcDL-SR~296 (Takeke et al.,
Mol. Cell Biol. (1988) 8:466 in which a XhoI fragment
located between the SR~(HTLV-1) promoter and the SV40
polyadenylation site has been replaced for convenient
cDNA cloning by a polylinker that contains the following
restriction sites: 5' XhoI-XbaI-Sfi-I-NotI-EcoRI-EcoRV-
HindIII-ClaI-3'. In addition a neomycin gene
(resistance to antibiotic G418) was inserted in between
the ampicillin resistance gene and the SR~ promoter)) by
electroporation (Chu et ~1., Nucleic Acids Res. (1987)
15:1311). The G418 resistant colonies were pooled and
analysed for surface expression of the PI-anchored TCR
and ~ chains. Cells were sequentially stained at 4-C
for 1 hour in phenol red free RPMI + 5% fetal calf serum
with KJ25 (anti-V~3) and biotinylated A2B4.2 (anti V~).
Fluorescein isothiocyanate (FITC) conjugated goat anti-
3s hamster immunoglobulins and streptavidin-PE were used as
secondary antibodies.
SXlOs Jurkat ~-cells expressing a functional 2B4
r~ , . . .
,
.
' ' ' i ~, ' ' ~ "
' ~ ' '~ ' '
. ~ ~
~ ' ~
WO92/0171~ PCT/US91/0508'
2 0 g 7 ~
~ TCR after transfection with 2B4 ~ and ~ ge~es (Sarto
and Germain, Nature (1987), 329, 256) (A and B) and CHC
cells transfected with the ~HPAP-S construct ~C and D)
were pretreated with (B and D) or without (A and C)
s A2B4.2 (20 ~g/ml) and then KJ25 (2 ~g/ml) FITC-
conjugated goat anti-hamster immunoglobulins (A2B4.2 is
a mouse antibody and RJ25 is from a hamster, ~uch that
only the latter will stain with the FITC reagent).
Cells were analysed by flow cytometry after staining.
o The data were plotted as log fluorescent intensity in
arbitrary units. Approximately half of the 2B4 Q~
transfected Jurkat cells are negative for TCR, and half
are positive accounting for the two populations that
were visible in the profile. The Y axis shows propidium
iodine staining, a measure of cell viability, while the
X-axis shows the degree of XJ25 flourescence.
When 2B4 Jurkat cells are first treated with the
antibody to Va and then with anti-V~3, staining with the
latter is completely inhibited. The same inhibition
2D effect can be observed after pretreatment of ~HPAP-S,
~DAF and ~DAP-L transfectants. ~J25 staining of the
CHO cells is inhibited by 90% after pre-incubation with
A2B4.2, supporting the PI-anchored TCR ~-chains being
expressed on the cell surface as heterod~mers with the
` 25 chain and in a conformation indistinguishable from the
native structure, even in the absence of CD3.
The TCR heterodimer was purified as follows:
2xlO~ ~HPAP-S transfected CHO cells were seeded into a
Cell Pharm I Bioreactor (CD Medical). After two weeks
of culture (2xlOl0 cells present in the Bioreactor), the
cells were treated with approximately lO units of PI-P~C
at 37'C for three hours (Becker et al., Nature (1985),
;~ 317, 430). The chimeric TCR was recovered in 250 ml of
growth media pulsed into the Bioreactor. The
s supernatant was filtered, adjusted to pH 8.0 and passed
over an A2B4.2 (cyanogen bromide coupled sepharose)
affinity column (l ml) pre-equilibrated with O.l M
- . .
. ~ . . - ,
. ~ , .
,' : '` . .
" . , , . :
.. .. , ~ ' , , ~ .
W092/0171~ 2 ~ 8 7 ~ ~ ~ PCT/~S91/~08~
phosphate/0.15~ NaC1, pH 8.0 buffer. The column was
then washed with the same buffer. T~e protein was
eluted with O.lM NaOAc/0.15M NaCl, pH 3.5 buffer in 0.5
ml fractions and neutralized immediately with 2M Tris-
s HCl p~ 8.5 (1/10 volume added~. The first four
fractions which contain most of the soluble TCR were
pooled and applied to a KJ25 affinity column. The
protein was then eluted at pH 5.0 with O.lM NaOAc/l~
NaC1.
to Sample eluates from the A2B4.2 (lanes 1, 3 and
4) or A2B4.2 + W25 (lanes 2, 5 and 6) columns were
electrophoresed on a 10% SDS-polyacrylamide gel under
reducing (lanes 1 and 2) and non-reducing (lanes 3-6)
conditions. When run under non-reducing conditions, the
samples were either not boiled (lanes 3 and 5) or boiled
(lanes 4 and 6~ before being loaded on the gel. The gel
was stained with silver nitrate (Morrissey, Anal.
Biochem. (1981), 117, 307). The fraction eluted from
the A2B4.2 column, not boiled (lane 3) or boiled (lane
`~ 20 4) before being loaded on the gel, contains both
monomers (~, 41 Xd) and dimers (~Q and ~, 70 Kd).
After subsequent passage of this fraction over the XJ25
column and elution at pH 5.0, only dimers are detectable .
(lane 5, sample not boiled) showing a separation of
heterodimers from a monomers and aa dimers. If the
sample is boiled (lane 6), low quantities of monomers
become detectable. Under reducing conditions (lanes 1
and 2), dimers are reduced to the size of monomers.
The data demonstrate the presence of
heterodimers and a small amount of dimeric molecules
which are not disulfide-linked. Estimating from the
silver stained gel indicates about 40% monomers and 60%
dimers cleaved off the surface of the C~O cells of which
50% of the dimers are heterodimers (30% of the total).
; 35 It is evident from the above results, that in
accordance with the subject invention, soluble surface
membrane proteins having multiple sub-units can be
.
: ' ' ' ` .
,
-
~' ' , '
.
W092/01725 ~ 0~ 7 ~ ~ ~ PCT/US91/~82
18
obtained with appropriate conformation. The epitopes
exhibited by the proteins are bound by antibodies which
recognize the naturally occurring surface membrane
proteins. The surface membrane proteins may be used in
s a variety of ways, in studying the nature of their
action with other proteins, for use as agonist or
antagonists with the surface ~embrane protein, for
screening compounds which bind to the surface me~brane
protein, and the like.
All publications and patent applications
mentioned in this specification are indicative of the
level of skill of those skilled in the art to which this
invention pertains. All publications and patent
applications are herein incorporated by reference to the
same extent as if each individual publication or patent
application was specifically and individually indicated
to be incorporated by reference.
Although the foregoing invention has been
described in some detail by way of illustration and
example for purposes of clarity of understanding, it
will be readily apparent to those of ordinary skill in
the art in light in light of the teachings of this
invention that certain changes and modifications may be
made thereto without departing from the spirit or scope
of the appended claims.
'
.
.' ~ '
.
'
.. . : , '
'
. . ' . ' ' , , ,
'
,
