Note: Descriptions are shown in the official language in which they were submitted.
CA 02599944 2007-09-14
DEMANDES OU BREVETS VOLUMINEUX
LA PRESENTE PARTIE DE CETTE DEMANDE OU CE BREVETS
COMPREND PLUS D'UN TOME.
CECI EST LE TOME DE
NOTE: Pour les tomes additionels, veillez contacter le Bureau Canadien des
Brevets.
JUMBO APPLICATIONS / PATENTS
THIS SECTION OF THE APPLICATION / PATENT CONTAINS MORE
THAN ONE VOLUME.
THIS IS VOLUME 1 OF 2
NOTE: For additional volumes please contact the Canadian Patent Office.
CA 02599944 2007-09-14
=
WO 98/08949 PCT/US97/14159
=.4.
P-SELECIlr1 LIGAND PROTEINS
This
application also claims priority from Intemationat Application No.
PCT/i.1S93/110108, filed
October 22, 1993.
BACKGROUND OF THE INVENTION
The present invention relates to the field of anti-inflammatory substances
which act by
inhibiting leukocyte adhesion to endothelial cells. More pariicularly, the
present invention is
directed to novel ligands for the mammalian adhesion proteins known as
selectins.
During inflammation leukocytes adhere to the vascular endotheliurri' and enter
subendothelial tissue, an interaction which is mediated by specific binding of
the. selectin or LEC-
CAM class of proteins to ligands on target cells. Such selectin-tnediated
cellular adhesion also
occurs in thrombotic disorders and parasitic diseases and may be implicated in
metastatic spread
of tumor cells.
The selectin proteins are characterized by a N-terminal lectin-like domain, an
epidermal
growth factor-like domain, and regions of homology to complement binding
proteins. Thus far
three human selectin proteins have been identified, E-selectin (fotTnerly
Ei.AM-{), L-selectin
(formerly LAM-1) and P-selectin (formerly PADGEM or GMP- 140). E-selectin is
induced on
endothelial cells several hours after activation by cytokines, mediating the
calcium-dependent
interaction between neutrophils and the endothelium. L-selectin is the
lymphocyte homing
receptor, and P-selectin rapidly appears on the cell surface of platelets when
they are activated,
mediating calcium-dependent adhesion of neutrophils or monocytes to platelets.
P-selectin is also
found in the Weibel-Palade bodies of endothelial cells; upon its release from
these vesicles =P-
selectin mediates early binding of neutrophils to histamine-or thrombin-
stimulated endothelium.
Selectins are believed to mediate adhesion through specific interactions with
ligands
present on the surface of target cells. Generally the ligands of selectins are
comprised at least in
1
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
part of a carbohydrate moiety. For example, E-selectin binds to carbohydrates
having the terminal
structure
NeuAca(2,3) Gal(3(1,4) G1cNAc--R
I
Fuca(1,3)
and also to carbohydrates having the terminal structure
NeuAca(2,3) Galp(1,3) G1cNAcp(1,3)--R
I
Fuca(1,4)
where R =the remainder of the carbohydrate chain. These carbohydrates are
known blood group
antigens and are commonly referred to as sialyl Lewis' and sialyl Lewis'.
respectively. The
presence of the sialyl Lewisx antigen alone on the surface of an endothelial
cell may be sufficient
- to promote binding to an E-selectin expressing cell. E-selectin also binds
to carbohydrates having
the terminal structures
HSO3--Ga1(3(1,4)G1cNAc--R HS03--Ga1(3(1,3)G1cNAc--R
I and I
Fuca(1,3) Fuca(1,4)
As with E-selectin, each selectin appears to bind to a range of carbohydrates
with varying
affinities. The strength of the selectin mediated adhesive event (binding
affinity) may also depend
on the density of the carbohydrate and on the density of the selectin on the
cell surface.
P-selectin binds to carbohydrates containing the non-sialat$d form of the
Lewis' blood
group antigen and with higher affinity to sialyl Lewisx. P-selectin may also
recognize sulfatides,
which are heterogeneous 3-sulfated galactosyl ceramides, isolated from myeloid
and tumorcelis
by lipid extraction. However, the binding of cells bearing P-selectin to cells
bearing P-selectin
ligands is abolished when the ligand-bearing cells are treated with proteases,
indicating that the
P-selectin ligand may be a glycoprotein.
2 4
CA 02599944 2007-09-14
=
WO 98/08949 PCT/US97/14159
Two putative glycoprotein ligands for P-selectin have recently been
identified, one of
which has been partially purified, (Moore et al., J. Cell Biol. 118, 445-456
(1992)). However,
neither amino acid composition nor the amino acid sequence of these
glycoproteins are disclosed.
SUMMARY OF THE INVENTION
In one embodiment, the present invention provides a composition comprising an
isolated
DNA encoding a P-selectin ligand protein, said protein comprising the amino
acid sequence set
forth in SEQ ID NO:2 from amino acid I to amino acid 402. Also provided is a
composition
comprising an isolated DNA encoding a soluble P-selectin ligand protein, said
protein comprising
the amino acid sequence set forth in SEQ ID NO:2 from amino acid 1 to amino
acid 310. The
invention further provides a composition comprising an isolated DNA encoding a
mature P-
selectin ligand protein, said protein comprising the amino acid sequence set
forth in SEQ ID NO:2
from amino acid 42 to amino acid 402. In another embodiment, the invention
provides a
composition comprising an isolated DNA encoding a soluble mature P-selectin
ligand protein,
said protein comprising the amino acid sequence set forth in SEQ ID NO:2 from
amino acid 42
to amino acid 310. In another embodiment, the invention provides a composition
comprising an
isolated DNA encoding a P-selectin ligand protein, said protein comprising the
amino acid
sequence set forth iti SEQ ID NO:4. The invention further provides a
composition comprising
an expression vector comprising any one of the isolated DNAs of the invention,
said DNA being
operably linked to an expression control sequence; a host cell transformed
with the expression
vector containing any one of the DNAs described above; and a process for
producing the P-
selectin ligand protein, which comprises:
(a) culturing a host cell transformed with an expression vector containing any
one
of the DNAs of the invention in a suitable culture medium; and
(b) purifying the P-selectin ligand protein from the culture medium.
In another embodiment, the invention provides a composition comprising a
protein
comprising the amino acid sequence set forth in SEQ ID NO:2 from amino acid 21
to amino acid
402, said protein being substantially free from other mammalian proteins. The
invention further
comprises a soluble P-selectin ligand protein comprising the amino acid
sequence set forth in SEQ
ID NO:2 from amino acid 21 to amino acid 310, said protein being substantially
free from other
mammalian proteins. In another embodiment, the invention comprises a P-
selectin ligand protein
comprising the amino acid sequence set forth in SEQ ID NO:2 from amino acid I
to amino acid
402, said protein being substantially free from other mammalian proteins. The
invention also
provides a composition comprising a mature P-selectin ligand protein
comprising the amino acid
3 ~
CA 02599944 2007-09-14 -
= r
WO 98/08949 PCT/US97/14159
sequence set forth in SEQ ID NO:2 from amino acid 42 to amino acid 402. said
protein being
substantially free from other mammalian proteins. Further provided is a
composition comprising
a soluble mature P-selectin ligand protein comprising the amino acid sequence
set forth in SEQ
ID NO:2 from amino acid 42 to amino acid 310, said protein being substantially
free from other
mammalian proteins. In another embodiment, the invention provides a
composition comprising
a protein comprising the amino acid sequence set forth in SEQ ID NO:4.
In yet another embodiment, the invention provides compositions comprising
antibodies
specific for P-selectin ligand proteins.
In another embodiment, the invention provides a method of identifying an
inhibitor of
P-selectin-mediated intercellular adhesion which comprises:
(a) combining a P-selectin protein with a P-selectin ligand protein comprising
an
amino acid sequence selected from the group consisting of the amino acid
sequence set forth in
SEQ ID NO:2 from amino acid I to amino acid 402, the amino acid sequence set
forth in SEQ
ID NO:2 from amino acid 42 to amino acid 402, the amino acid sequence set
forth in SEQ ID
NO:2 from amino acid 42 to amino acid 310. and the amino acid sequence set
forth in SEQ ID
NO:4, said combination forming a first binding mixture;
(b) measuring the amount of binding between the P-selectin protein and the P-
selectin ligand protein in the first binding mixture;
(c) combining a compound with the P-selectin protein and the P-selectin ligand
protein to form a second binding mixture; -
(d) measuring the amount of binding in the second binding mixture; and
(e) comparing the amount of binding in the first binding mixture with the
amount
of binding in the second binding mixture;
wherein the compound is capable of inhibiting P-selectin-mediated
intercellular adhesion when
a decrease in the amount of binding of the second binding mixture occurs.
In another embodiment, the invention provides a method of identifying an
inhibitor of
E-selectin-mediated intercellular adhesion which comprises:
(a) combining a E-selectin protein with a P-selectin ligand protein comprising
an
anlino acid sequence selected from the group consisting of the amino acid
sequence set forth in
SEQ ID NO:2 from amino acid I to amino acid 402, the amino acid sequence set
forth in SEQ
ID NO:2 from amino acid 42 to amino acid 402, the amino acid sequence set
forth in SEQ ID
NO:2 from amino acid 42 to amino acid 310, and the amino acid sequence set
forth in SEQ ID
NO:4, said combination forming a first binding mixture;
4 4
CA 02599944 2007-09-14
t
WO 98/08949 PCT/US97/14159
(b) measuring the amount of binding between the E-selectin protein and the P-
selectin ligand protein in the first binding mixture;
(c) combining a compound with the E-selectin protein and the P-selectin ligand
protein to form a second binding mixture;
(d) measuring the amount of binding in the second bindinc, mixture; and
(e) comparing the amount of binding in the first binding mixture with the
amount
of binding in the second binding mixture;
wherein the compound is capable of inhibiting E-selectin-mediated
intercellular adhesion when
a decrease in the amount of binding of the second binding mixture occurs.
These methods may
also be used to iook for L-selectin inhibitors by substituting L-selectin for
E-selectin.
The invention also encompasses processes for producing P-selectin ligand
proteins which
comprise (a) co-transforming a host cell with a DNA encoding a P-selectin
ligand protein and a
DNA encoding a fucosyltransferase capable of synthesizing sialyl Lewis X(sLe')
or sialy] Lewis
A(sLe') (such as an (a1,3/a1,4) fucosyltransferase or an (al, 3)
fucosyltransferase), each of said
DNAs being operably linked to an expression control sequence; (b) culturing
the host cell in
suitable culture medium; and (c) purifying the P-selectin ligand protein from
the culture medium.
In certain other embodiments, the host cell is also co-transformed with a DNA
encoding a paired
basic amino acid converting enzyme and/or a DNA encoding a GIcNAc transferase
(preferably
a "core2 transferase"). In preferred embodiments, the P-selectin ligand
protein is a full-length or
soluble form.
In other embodiments, the present invention includes a P-selectin ligand
protein having
P-selectin ligand protein activity. In preferred embodiments, the ligand
protein is a protein
comprising the sequence from amino acid 42 to amino acid 60 of SEQ ID NO: 2.
consisting
essentially of the sequence from amino acid 42 to amino acid 60 of SEQ ID NO:
2, comprising
the sequence from amino acid 42 to amino acid 88 of SEQ ID NO: 2, consisting
essentially of the
sequence from amino acid 42 to amino acid 88 of SEQ ID NO: 2, consisting
essentially of the
sequence from amino acid 42 to amino acid i 18 of SEQ ID NO: 2, or consisting
essentially of the
sequence from amino acid 42 to amino acid 189 of SEQ ID NO: 2. In other
_prefetred
embodiments, at least one of the asparagine residues at positions 65, 111 and
292 of SEQ ID NO:
2 have been deleted or replaced. Certain preferred embodiments of the ligand
protein comprises
at least one of the tyrosine residues at positions 46, 48 and 51 of SEQ ID NO:
2. DNAs encoding
these P-selectin ligand proteins, host cells transformed with such DNAs,
process for producing
protein by culturing such host cells, pharmaceutical compositions comprising
the proteins,
methods of identifying selectin binding inhibitors using the proteins,
antibodies to the proteins
5 =
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
and methods of inhibiting selectin mediated binding using the proteins are
also encompassed by
the invention.
In yet other embodiments the invention provides an isolated DNA encoding a
fusion protein comprising (a) a first amino acid sequence comprising amino
acid 42 to
amino acid 60 of SEQ ID NO:2, and (b) a second amino acid sequence derived
from the
sequence of a protein other than P-selectin ligand. Preferably, an expression
control
sequence is operably linked to the nucleotide sequence. Host cells transformed
with such
DNAs are also provided. The invention also provides a process for producing a
fusion
protein, which comprises: (a) culturing the host cell under condition suitable
for
expression of the fusion protein; and (b) purifying the fusion protein from
the culture
medium. Fusion proteins produced according to such process are also provide. ~
In certain preferred embodiments, the first amino acid sequence of such fusion
protein comprises amino acid 42 to amino acid 402 of SEQ ID NO:2, amino acid
42 to
amino acid 310 of SEQ ID NO:2, amino acid 42 to amino acid 88 of SEQ ID NO:2,
amino acid 42 to amino acid 118 of SEQ ID NO:2, or amino acid 42 to amino acid
189
of SEQ ID NO:2.
In other preferred embodiments, the DNA comprises the nucleotide sequence of
SEQ ID NO:35 from nucleotide 123 to nucleotide 939, the nucleotide sequence of
SEQ
ID NO:35, the nucleotide sequence of SEQ ID NO:37 from nucleotide 123 to
nucleotide
807, the nucleotide sequence of SEQ ID NO:37, the nucleotide sequence of SEQ
ID
NO:39 from nucleotide 123 to nucleotide 1311, the nucleotide sequence of SEQ
ID
NO:39, the nucleotide sequence of SEQ ID NO:41 from nucleotide 123 to
nucleotide
792, or the nucleotide sequence of SEQ ID NO:41.
The present invention also provides a fusion protein comprising (a) a first
amino
acid sequence comprising amino acid 42 to amino acid 60 of SEQ ID NO:2, and
(b) a
second amino acid sequence derived from the sequence of a protein other than P-
selectin
ligand. Preferably, the first amino acid sequence comprises amino acid 42 to
amino acid
402 of SEQ ID NO:2, amino acid sequence comprises amino acid 42 to amino acid
310
of SEQ ID NO:2, amino acid 42 to amino acid 88 of SEQ ID NO:2, amino acid 42
to
amino acid 118 of SEQ ID NO:2, or amino acid 42 to amino acid 189 of SEQ ID
NO:2.
In certain particularly preferred embodiments, the fusion protein comprises
the
amino acid sequence of SEQ ID NO:36 from amino acid 42 to amino acid 313, the
6 .
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14155
amino acid sequence of SEQ ID NO:36, the amino acid sequence of SEQ ID NO:38
from amino acid 42 to amino acid 269, the amino acid sequence of SEQ ID NO:38.
the
amino acid sequence of SEQ ID NO:40 from amino acid 42 to amino acid 437. the
amino acid sequence of SEQ ID NO:40, the amino acid sequence of SEQ ID NO:42
from amino acid 42 to amino acid 264, or the amino acid sequence of SEQ ID
NO:42.
In other preferred embodiments, the second amino acid sequence is linked to
the
C-terminus or the N-terminus of the first amino acid sequence, with or without
being
linked by a linking sequence.
In yet other embodiments, the second amino acid sequence is derived from a
protein selected from the group consisting of an antibody, a cytokine, a
growth factor, a
differentiation factor, a hormone, an enzyme, a receptor or fragment thereof
and a ligand.
Preferably, the second amino acid sequence is derived from the sequence of an
antibody,
froin the Fc portion of an antibody, or is a mutation of a sequence derived
from an
antibody.
In yet further embodiments, the present invention provides for a composition
comprising (a) a first peptide comprising amino acid 42 to amino acid 60 of
SEQ ID
NO:2, and (b) a second peptide derived from the sequence of a protein other
than P-
selectin ligand, wherein the first peptide and the second peptide are
chemically linked by
a moiety other than a peptide bond. Any P-selectin ligand protein of the
invention can
be used in such a composition.
BRIEF DESCRIPTION OF THE FIGURES
Fig. I is a graph comparing the binding of P-selectin ligand proteins
expressed with and
without core2.
Fig. 2 is an autoradiograph of immunoprecipitations of P-selectin ligand
protein expressed
with and without core2.
Fig. 3 depicts the results of flow cytometry analysis of the binding of P-
selectin ligand
protein (expressed with and without core2) to P-selectin/IgG chimera (LEC-y1)
and anti-P-
selectin ligand protein monoclonal antibody (MAb 275).
Fig. 4 is an autoradiograph of proteins, including P-selectin ligand protein,
which bound
to P- and E-selectin/IgG chimeras.
Fig. 5 is a schematic representation of structural features of the full length
P-selectin
ligand protein of SEQ ID NO: 2.
7 .
CA 02599944 2007-09-14 -
WO 98/08949 PCT/US97/14159
Fig. 6 is a schematic representation of several P-selectin ligand protein
fragments
constructed for the purpose of examining the role of N-linked glycosylation
sites in binding of the
P-selectin ligand proteins to selectins.
Fig. 7 depicts the results of experiments to determine the role of N-linked
glycosylation
sites in binding of the P-selectin ligand proteins to selectins.
Fig. 8 is a schematic representation of several P-selectin ligand protein
fragments
constructed for the purpose of examining the role of sulfated tyrosine
residues in binding of the
P-selectin ligand proteins to selectins.
Figs. 9-11 depicts the results of experiments to determine the role of
sulfated tyrosine
residues in binding of the P-selectin ligand proteins to selectins.
Fig. 12 is a schematic representation of several P-selectin ligand protein
fragments
constructed for the purpose of examining the effects of various deletions on
the binding of the P-
selectin ligand proteins to selectins.
Fig. 13 is a schematic depiction of the quantitative plate binding assay of
Example 4(c).
Figs. 14-17 depict the results of experiments comparing the binding of various
deleted
and altered P-selectin ligand proteins to selectins.
Fig. 18 is a schematic representation of several P-selectin ligand protein
fragments
constructed for the purpose of examining the effects of alteration of tyrosine
residues in the
anionic region of the P-selectin ligand proteins on selectin binding.
Figs. 19-21 depict the results of experiments comparing the binding of various
deleted
and altered P-selectin ligand proteins to selectins.
Fig. 22 depicts a proposed model for binding of P-selectin ligand proteins to
P- and E-
selectin.
Figs. 23 and 24 depict the results of experiments comparing the binding of
various deleted
and altered P-selectin ligand proteins to selectins.
Some of the foregoing figures employ a convention for numbering amino acids
within
the depicted consttucts which is different that the residue numbering employed
in SEQ ID NO:2.
In the figures, residues are numbered using the first amino acid of soluble
mature P-selectin ligand
as a starting point. Hence, the residue numbers used in the figures are 41
less than those of SEQ
ID NO:2. For example, residue 19 in the figures corresponds to residue 60 in
SEQ ID NO:2.
Fig. 25 is an analysis of the expression products of CHO cells, already
expressing 314
fucosyltransferase and Core2 transferase, which were transfected with
psPSL.T7, OTM, 1316 or
psPSL.QC and amplified using methotrexate. Conditioned media was either
anaiyzed directly or
8 '
CA 02599944 2007-09-14
WO 98/08949 PCT/US97114159
first precipitated with LEC-y I and then anaiyzed by SDS-PAGE under non-
reducing and
reducing conditions.
Fig. 26. SDS-PAGE separation of myeloid cell membrane proteins affinity
captured by
P-and E-selectin. Membrane lysates were prepared from U937 cells metabolicaily
labeled with
'H-glucosamine and subjected to affinity precipitation with immobilized P-and
E-selectin and
control human IgG,. Eluted proteins were treated with ("reduced") or without
("non-reduced")
DTT prior to gel electrophoresis. Lanes: 1, affinity capture by human I-G,; 2,
affinity capture
by P-selectin; 3, affinity capture by E-selectin.
Fig. 27. Sequential affinity capture experiments. 3H-labeled U937 lysate
species were
affinity captured by P-or E-selectin, eluted, and then subjected to
immunoprecipitation with
anti-PSGL-1 antiserum Rb3443. Lanes: I and 2, control immunoprecipitations of
fresh myeloid
cell membrane lysates using pre-immune rabbit serum (lane 1) and Rb3443 (lane
2); 3-5,
immunoprecipitation with Rb3443 of myeloid cell membrane lysates previously
affinity captured
and eluted from P-selectin (lane 3), E-selectin (lane 4), and human IgG, (lane
5).
Fig. 28. Comparison of CD43 and PSGL-1 content of myeloid cell membrane
extracts.
Labeled U937 cell extracts were immunoprecipitated with anti-PSGL-1 rabbit
polyclonal antibody
Rb3443 or an anti-CD43 mouse MAb and then subjected to SDS-
PAGE/autoradiography. Lanes:
1, immunoprecipitation with control pre-immune rabbit serum; 2,
immunoprecipitation with
Rb3443; 3, immunoprecipitation with control isotype-matched mouse antibody; 4,
immunoprecipitation with anti-CD43 antibody.
Fig. 29. COS transfection experiments. COS M6 cells transfected with plasmids
encoding PSGL- I or CD43 as well as Fuc-TIII or fuc-TVII were metabolically
labeled with
35S-methionine, and membranes were prepared for affinity capture experiments
as described in
Materials and Methods. The cDNAs employed in the transfections are indicated
above the lanes.
Precipitations were performed using (A) E-selectin, (B) P-selectin, and (C)
anti-PSGL-l antiserum
Rb3443 and anti-CD43 MAb.
Fig. 30 summarizes the results of screening of various P-selecint ligand
proteins for
inhibition of P- and E-selectin binding (see Example 13).
DETAILED DESCRIPTION OF THE INVENTION
The present inventors have for the first time identified and isolated a novel
DNA which
encodes a protein which acts as a ligand for P-selectin on human endothelial
cells and platelets.
The sequence of the DNA is set forth in SEQ ID NO: 1. The complete amino acid
sequence of
the P-selectin ligand protein (i.e., the mature peptide plus the leader
sequence) is characterized by
the amino acid sequence set forth in SEQ ID NO:2 from amino acid 1 to amino
acid 402.
9 4
CA 02599944 2007-09-14
WO 98/08949 PCTlUS97114159
Hydrophobicity analysis and comparison with known cleavage patterns predict a
signal sequence
of 20 to 22 amino acids, i_e., amino acids I to 20 or amino acids I to 22 of
SEQ ID NO:2. The
P-selectin ligand protein contains a PACE (paired basic amino acid converting
enzyme) cleavage
site (-Arg-Asp-Arg-Arg-) at amino acids 38-41 of SEQ ID NO:2. The mature P-
selectin ligand
protein of the present invention is characterized by the amino acid sequence
set forth in SEQ ID
NO:2 from amino acid 42 to amino acid 402: A soluble form of the P-selectin
ligand protein is
characterized by containing amino acids 21 to 310 of SEQ ID NO:2. Another
soluble form of the
mature P-selectin ligand protein is characterized by the amino acid sequence
set forth in SEQ ID
NO:2 from amino acid 42 to amino acid 310. The soluble form of the P-selectin
ligand protein
is further charactetized by being soluble in aqueous solution at room
temperature. Of course, the
corresponding DNA sequences as set forth in SEQ ID NO: I -encoding these
proteins are also
included in the subject invention.
The P-selectin ligand of the invention is a glycoprotein which may contain one
or more
of the following tetminal carbohydrates:
NeuAca(2,3) Gal (3(1,4) G1cNAc-R
la(1,3)
Fuc
NeuAca(2,3) Gal P(1,3) G1cNAc-R
la(1, 4)
Fuc
Gal (3(1,4)G1cNAc-R
la(1,3)
Fuc
Gal (3(1,3)G1cNAc-R
I a(1,4)
Fuc
where R= the remainder of the carbohydrate chain, which is covalently attached
either directly
to the P-selectin ligand protein or to a lipid moiety which is covalently
attached to the P-selectin
ligand protein. The P-selectin ligand glycoprotein of the invention may
additionally be sulfated
or otherwise post-translationally modified. As expressed in COS and CHO cells,
full length P-
selectin ligand protein (amino acids I to 402 of SEQ ID NO:2 or amino acids 42
to 402 of SEQ
10 ~
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
ID NO:2) is a homodimeric protein having an apparent molecular of 220 kD as
shown by non-
reducing SDS-polyacrylamide gel electrophoresis.
The structure of the full-length P-selectin ligand protein is schematically
represented in
Fig. 5. Three regions of the P-selectin ligand protein of SEQ ID NO:2 are: an
extracellular
domain (from about amino acid 21 to 310 of SEQ ID NO:2), a transmembrane
domain (from
about amino acid 311 to 332 of SEQ ID NO:2), and an intracellular, cytoplasmic
domain (from
about amino acid 333 to 402 of SEQ ID NO:2). The extracellular domain contains
three
consensus tripeptide sites (Asn-X-Serrl'hr) of potential N-linked
glycosylation beginning at Asn
residues 65, 111, and 292. The extracellular domain further contains three
potential sites of
tyrosine sulfation at residues 46, 48, and 51. The region comprised of
residues 55-267 contains'
a high percentage of proline, serine, and threonine including a subdomain of
fifteen decameric
repeats of the ten amino acid consensus sequence Ala-Thr/Met-Glu-Ala-Gln-Thr-
Thr-X-Pro/Leu-
Ala/Thr, wherein X can be either Pro. Ala, Gln, Glu, or Arg. Regions such as
these are
characteristic of highly 0-glycosylated proteins.
COS or CHO cells co-transfected with a gene encoding the P-selectin ligand
protein and
a gene encoding fucosyltransferase (hereinafter FT), preferably an (aI,3/al,4)
fucosyltransferase
("3/4FT"), are capable of binding to CHO cells expressing P-selectin on their
surface, but are not
capable of binding to CHO cells which do not express P-selectin on their
surface. In order to bind
to P-selectin, either in purified form or expressed on the surface of CHO
cells, the gene encoding
the P-selectin ligand protein must be co-transfected with the gene encoding an
F'I', since
transfection of either gene in the absence of the other either abolishes or
substantially reduces the
P-selectin binding activity. The binding of the P-selectin ligand protein of
the invention to P-
selectin can be inhibited by EDTA or by a neutralizing monoclonal antibody
specific for P-
selectin. The binding of the P-selectin ligand protein of the invention to P-
selectin is not inhibited
by a non-neutralizing monoclonal antibody specific for P-selectin or by an
isotype control. These
results characterize the binding specificity of the P-selectin ligand protein
of the invention.
For the purposes of the present invention, a protein is defined as having "P-
selectin ligand
protein activity", i.e., variably referred to herein as a "P-selectin ligand
protein", or as a "P-selectin
ligand glycoprotein" or simply as a "P-selectin ligand", when it binds in a
calcium-dependent
manner to P-selectin which is present on the surface of cells as in the CHO-P-
selectin binding
assay of Example 4, or to P-selectin which is affixed to another surface, for
example, as the
chimeric P-selectin-IgGy 1 protein of Example 4 is affixed to Petri dishes.
The glycosylation state of the P-selectin ligand protein of the invention was
studied using
a chimeric. soluble form of the P-selectin ligand protein, described in detail
in Example 5(C) and
11 ~
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
designated sPSL.T7. The sPSL.T7 protein produced from COS cells co-transfected
with 3/4FT
is extensively modified by post-translational glycosylation. as described in
detail in Example 6(C).
Thus, it is believed that both N- and 0-linked oligosaccharide chains, at
least some of which are
sialated. are present on the P-selectin ligand protein of the invention.
The P-selectin ligand protein of the invention may also bind to E-selectin and
L-selectin.
Conditioned medium from COS cells which have been co-transfected with the DNA
encoding
sPSL.T7 or P-selectin ligand-Ig fusions and with the DNA encoding 3/4FT. when
coated on wells
of plastic microtiter plates, causes CHO cells which express E-selectin to
bind to the plates;
however CHO cells which do not express E-selectin do not bind to such plates.
The binding of
CHO cells which express E-selectin to microtiter plates coated with
conditioned medium from
COS cells which have been co-transfected with the DNA encoding sPSL.T7 and
with the DNA
encoding 3/4FT is abolished in the presence of EDTA or of a neutralizing
antibody specific for
E-selectin. Conditioned medium from COS cells transfected only with the
sPSL.T7 DNA does
not cause binding of CHO cells which express E-selectin when coated on wells
of microtiter
plates. For these reasons, the P-selectin ligand protein of the invention is
believed to be useful
as an inhibitor of E-selectin-mediated intercellular adhesion in addition to P-
selectin-mediated
intercellular adhesion.
Antibodies raised against COS-produced soluble P-selectin ligand protein are
immunoreactive with the major HL-60 glycoprotein that specifically binds P-
selectin as
determined by affinity capture using an immobilized Fc chimera of P-selectin.
U937 cells bear
a similar immunoreactive glycoprotein ligand. Thus, a single glycoprotein
species is observed
upon EDTA elution of immobilized P-selectin previously incubated with
detergent extracts of 3H-
glucosamine labeled U937 cells. This major species exhibits an apparent
molecular weight by
SDS-PAGE of 220 kD under non-reducing conditions and 100 kD under reducing
conditions.
As with the comparable species isolated from HL-60 cells, this U937 ligand is
immunoreactive
with a polyclonal antibody raised against COS recombinant P-selectin ligand
protein. In addition,
affinity capture of E-selectin ligands from U937 cell and cell membrane
preparations. using an
immobilized Fc chimera of E-selectin, yield a single major species with
identical mass and
electrophoretic behavior as the major U937 P-selectin ligand. Thus,-E- and P-
selectin recognize
the same major glycoprotein ligand in U937 cells, a glycoprotein ligand
immunoreactive with an
anti-P-seiectin ligand protein antibody and possessing the same apparent mass
and electrophoretic
behavior as full length, recombinant P-selectin ligand protein.
Fragments of the P-selectin ligand protein which are capable of interacting
with P-selectin
or which are capable of inhibiting P-selectin-mediated intercellular adhesion
are also encompassed
12 =
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
by the present invention. Such fragments comprise amino acids 21 to 54 of SEQ
ID NO:2, a
region of the P-selectin ligand protein having a low frequency of serine and
threonine residues;
amino acids 55 to 127 of SEQ ID NO:2, having a high frequency of proline,
serine, and threonine
in addition to two consensus sequences for asparagine-linked glycosylation
(Asn-X-Ser/Thr);
another larger fragment, amino acids 128 to 267 of SEQ ID NO:2, having both a
high frequency
of proline, serine, and threonine and containing fifteen repeats of the
following ten amino acid
consensus sequence: Ala-(Thr/Met)-Glu-Ala-Gln-Thr-Thr-(Pro/Arg/Gln/Ala/Glu)-
(Leu/Pro)-
(AlalThr) (smaller fragments within this large fragment may also retain the
capacity to interact
with P-selectin or act as inhibitors of P-selectin-mediated intercellular
adhesion); the region
containing a consensus sequence for asparagine-linked glycosylation and
comprising amino acids
268 to 308 of SEQ ID NO:2; the hydrophobic region of the protein represented
by amino acids
309 to 333 of SEQ ID NO:2; and the amphophilic region of the P-selectin ligand
protein from
amino acids 334 to 402 of SEQ ID NO:2. Additional fragments may comprise amino
acid 43 to
amino acid 56 of SEQ ID NO:2 or amino acid 42 to amino acid 60 of SEQ ID NO:2,
with one or
more sulfated or phosphorylated (Domcheck et al.. Biochemistry 31:9865-9870
(1992)) tyrosines
at amino acid 46, amino acid 48, and/or amino acid 51. Fragments of the P-
selectin ligand protein
may be in linear form or they may be cyclized using known methods, for
example, as described
in H.U. Saragovi, et al., Bio/Technology 10, 773-778 (1992) and in R.S.
McDowell, et al., J.
Amer. Chem. Soc. 114, 9245-9253 (1992), both of which are incorporated herein
by reference.
For the purposes of the present invention, all references to "P-selectin
ligand protein" herein
include fragments capable of binding to P-selectin.
Such fragments may be fused to carrier molecules such as immunoglobulins, to
increase
the valency of P-selectin ligand binding sites. For example, soluble forms of
the P-selectin ligand
protein such as the fragments from amino acid 42 to amino acid 295 or from
amino acid 42 to
am~no a dci 88 of SEQ ID NO:2 may be fused through "linker" sequences to the
Fc portion of an
immunoglobulin (native sequence or mutated sequences for conferring desirable
qualities (such
as longer half-life or reduced immunogenicity) to the resulting chimera). For
a bivalent form of
the P-selectin ligand protein, such a fusion could be to the Fc portion of an
IgG molecule as in
Example 5(D) and in SEQ ID NO:6. Other immunoglobulin isotypes may also be
used to
generate such fusions. For example, a P-selectin ligand protein - IgM fusion
would generate a
decavalent form of the P-selectin ligand protein of the invention.
Fusions of any of the P-selectin ligand proteins of the present invention to
amino acid
sequences derived from other proteins may also be constructed. Preferred P-
selectin ligand
proteins for such purpose include the fragments from amino acid 42 to amino
acid 295 or from
13 4
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
amino acid 42 to amino acid 88 of SEQ ID NO:2. Desirable fusion proteins may
incorporate
amino acid sequence from proteins having a biological activity different from
that of P-selectin
ligand, such as, for example, cytokines, growth and differentiation factors
(such as bone
morphogenetic proteins (e.g., BMPs), hormones, enzymes, receptor components or
fragments and
other ligands. Also, P-seiectin lioand protein can be chemically coupled to
other proteins or
pharmaceutical agents. In such usage, the P-selectin ligand protein, by virtue
of the abilit to
interact with selectin molecules, alters the phanmacokinetics and/or
biodistribution of the fused
or coupled agent thereby enhancing its therapeutic efficacy. For example,
fusion of a P-selectin
ligand protein sequence to a cytokine sequence can direct the cytokine's
activity to an area of
inflammation. In such instance, the P-selectin ligand protein protion of the
fusion protein will
bind to selectins expressed at the site of inflammation. This binding will
cause the cytokine
portion of the fusion protein to become localized and available to bind its
cognate receptor or any
proximal cell surface. Other ligands could similarly be used in such fusions
proteins to attract
cells expressing their corresponding receptors to a site of P-selectin
expression. Preferred
examples of such fusions are described in Example 15.
In any fusion protein incorporating a P-selectin ligand protein, the amino
acid sequence
derived from a protein or proteins other than P-selectin ligand can be linked
to either the C-
terminus or N-terminus of the P-selectin ligand-derived sequence. The linkage
may be direct (i.e.,
without an intervening linking sequence not derived from either protein) or
through a linking
sequence.
Methods of treating a mammalian subject using such fusion proteins are also
contemplated by the present invention. In such instances, the fusion protein
is used to treat a
condition which is affected by the protein to which the P-selectin ligand
protein is fused. For
exaample, a fusion of a P-selectin ligand protein to 1L-11 could be used to
localize the activity of
IL-I I to bone marrow endothelial cells which express selectins on their
surface. Once localized,
the IL-I I portion of the fusion protein will stimulate megakaryocyte
progenitors. Similarly, a
fusion of a P-selectin ligand protein to a BMP could be used to stimulate bone
or cartilage
formation in an area of injury. Injured tissues express P-selectin, which will
bind the fusion
protein. Once localized, the BMP portion of the fusion protein will stimulate
bone or cartilage
production in the area of injury.
As detailed in the Examples below, the P-selectin ligand protein of the
invention was
initiallv obtained using an expression cloning approach (Clark et al., U.S.
4,675,285). A cDNA
library was constructed from the human promyelocytic cell line HL-60 (S.J.
Collins, et al., Nature
270, 347-349 (1977), ATCC No. CCL 240). This library was cotransfected into
.COS cells with
=
14
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
a DNA encoding a 3/4FT, and the cotransfectants were screened for binding to a
chimeric
molecule consisting of the extracellular portion of P-selectin and the Fc
portion of a human IgGy 1
monoclonal antibody. Cotransfectants which bound to the chimeric P-selectin
were enriched for
cDNAs encoding the P-selectin ligand protein. This screening process was
repeated several times
to enrich the plasmid population further for cDNAs encoding the P-selectin
ligand protein. In a
second cloning stage, the enriched plasmid population was again cotransfected
into COS cells
with the 3/4FT gene and screened for binding to a fluorescently labeled CHO
cell line which
expressed P-selectin on the cell surface. A single cDNA clone was obtained
from this approach
and was designated pMT2]:PL85. The pMT21:PL85 plasmid was deposited with the
American
Type Culture Collection on October 16, 1992 and given the accession number
ATCC 69096.
One novel DNA of the present invention is set forth in SEQ ID NO:1. The DNA of
the
present invention may encode a variety of forms of the P-selectin ligand
protein. For example,
in one embodiment, the DNA of the invention encodes the entire P-selectin
ligand protein having
the amino acid sequence set forth in SEQ ID NO:2 from amino acid I to amino
acid 402. In
another embodiment, the DNA of the invention encodes a form of the P-selectin
ligand protein
which lacks the signal sequence and which is characterized by the amino acid
sequence set forth
in SEQ ID NO:2 from amino acid 21 to amino acid 402. In yet another
embodiment, the DNA
of the invention encodes the mature P-selectin ligand protein characterized by
the amino acid
sequence set forth in SEQ ID NO:2 from amino acid 42 to amino acid 402.
Another embodiment
of the DNA of the invention encodes a soluble form of the_P-selectin ligand
protein characterized
by the amino acid sequence set forth in SEQ ID NO:2 from amino acid I to amino
acid 310. The
DNA of the invention is also embodied in a DNA encoding a soluble form of the
mature P-
setectin ligand protein, said protein being characterized by the amino acid
sequence set forth in
SEQ ID NO:2 from amino acid 42 to amino acid 310. The DNA of the invention is
further
embodied in a DNA sequence encoding a soluble form of the P-selectin ligand
protein which
lacks the signal sequence, said protein being characterized by the amino acid
sequence set forth
in SEQ ID NO:2 from amino acid 21 to amino acid 310. The DNA of the present
invention is free
from association with other human DNAs and is thus characterized as an
isolated DNA. As
detailed above, DNAs which encode P-selectin ligand fragments which interact
with P-selectin
are also included in the present invention.
The expression of P-selectin ligand protein mRNA transcripts has been observed
in a
variety of human cell lines (HL-60, THP-1, U937) and in human monocytes and
polymorphonuclear leukocytes by Northern analysis using a P-selectin ligand
protein cDNA
probe. In all of these cell lines, a major transcript of 2.5 kb was observed.
A minor species of
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
approximately 4 kb was observed in the HL60 and U937 cell lines and in
polymorphonuclear
leukocytes. In contrast, no P-selectin ligand mRNA expression was detected in
the human
hepatoblastoma cell line HepG2.
The P-selectin ligand protein of the invention is encoded by a single copy
gene and is not
part of a multi-gene family, as determined by Southern blot analysis. The
genomic form of the
P-selectin ligand protein of the invention contains a large intron of
approximately 9 kb located
at nucleotide 54 in the 5' untranslated region. ln polymorphonuclear
leukocytes and monocytes.
the P-selectin ligand protein of the invention is encoded by the DNA sequence
set forth in SEQ
ID NO:3. In this embodiment, the P-selectin ligand protein contains sixteen
repeat regions. The
isolated DNA of the invention is correspondingly also embodied in the DNA
sequence set forth
in SEQ ID NO:3 and is contained on plasmid pPL85R16 which was deposited with
the American
Type Culture Collection on October 22, 1993 and given the Accession Number
ATCC 75577.
The invention also encompasses allelic variations of the isolated DNA as set
forth in SEQ
ID NO:1 or of the isolated DNA as set forth in SEQ ID NO:3, that is, naturally-
occurring
alternative fon-ns of the isolated DNA of SEQ ID NO: I or SEQ ID NO:3 which
also encode
proteins having P-selectin ligand activity. Also included in the invention are
isolated DNAs
which hybridize to the DNA set forth in SEQ ID NO:1 or to the DNA set forth in
SEQ ID NO:3
under stringent (e.g. 4xSSC at 65 C or 50% formamide and 4xSSC at 42 C), or
relaxed (4xSSC
at 50 C or 30-40% formamide at 42 C) conditions, and which have P-selectin
ligand protein
activity. Isolated DNA sequences which encode the P-selectin ligand protein
but which differ
from the DNA set forth in SEQ ID NO: I or from the DNA set forth in SEQ ID
NO:3 by virtue
of the degeneracy of the genetic code and which have P-selectin ligand protein
activity are also
encompassed by the present invention. Variations in the DNA as set forth in
SEQ ID NO:1 or in
the DNA as set forth in SEQ ID NO:3 which are caused by point mutations or by
induced
modifications which enhance the P-selectin ligand activity, half-life or
production level are also
included in the invention. For the purposes of the present invention all
references herein to the
"DNA of SEQ ID NO:1 " include, in addition to DNAs comprising the specific DNA
sequence
set forth in SEQ ID NO:1, DNAs encoding the mature P-selectin ligand protein
of SEQ ID NO:2;
DNAs encoding fragments of the P-selectin ligand protein of SEQ ID NO:2 which
are capable
of binding to P-selectin; DNAs encoding soluble forms of the P-selectin ligand
protein of SEQ
ID NO:2; allelic variations of the DNA sequence of SEQ ID NO:1; DNAs which
hybridize to the
DNA sequence of SEQ ID NO:1 and which encode proteins having P-selectin ligand
protein
activitv; DNAs which differ from the DNA of SEQ ID NO:1 by virtue of
degeneracy of the
genetic code; and the variations of the DNA sequence of SEQ ID NO:I set forth
above. Similarly,
16 4
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
all references to the "DNA of SEQ ID NO:3" include in addition to the specific
sequence set forth
in SEQ ID NO:3, DNAs encoding the mature P-selectin ligand protein of SEQ ID
NO:4: DNAs
encoding fragments of the P-selectin ligand protein of SEQ ID NO:4 which are
capable of binding
to P-selectin; DNAs encoding soluble forms of the P-selectin ligand protein of
SEQ ID NO:4;
allelic variations of the DNA of SEQ ID NO:3; DNAs which hybridize to the DNA
sequence of
SEQ ID NO:3 and which encode proteins having P-selectin ligand protein
activity; DNAs which
differ from the DNA of SEQ ID NO:3 by virtue of degeneracy of the genetic
code; and the
variations of the DNA of SEQ ID NO:3 set forth above.
A DNA encoding a soluble form of the P-selectin ligand protein may be prepared
by
expression of a modified DNA in which the regions encoding the transmembrane
and cytoplasmic
domains of the P-selectin ligand protein are deleted and/or a stop codon is
introduced 3' to the
codon for the amino acid at the carboxy terminus of the extracellular domain.
For example,
hydrophobicity analysis predicts that the P-selectin ligand protein set forth
in SEQ ID NO:2 has
a transmembrane domain comprised of amino acids 311 to 332 of SEQ ID NO:2 and
a
cytoplasmic domain comprised of amino acids 333 to 402 of SEQ ID NO:2. A
modified DNA
as described above may be made by standard molecular biology techniques,
including site-
directed mutagenesis methods which are known in the art or by the polymerase
chain reaction
using appropriate oligonucleotide primers. Methods for producing several DNAs
encoding
various soluble P-selectin ligand proteins are set forth in Example 5.
A DNA encoding other fragments and altered forms of P-selectin ligand protein
may be
prepared by expression of modified DNAs in which portions of the full-length
sequence have
been deleted or altered. Substantial deletions of the P-selectin ligand
protein sequence can be
made while retaining P-selectin ligand protein activity. For example, P-
selectin ligand proteins
comprising the sequence from amino acid 42 to amino acid 189 of SEQ ID NO: 2,
the sequence
from amino acid 42 to amino acid 118 of SEQ ID NO: 2, or the sequence from
amino acid 42 to
amino acid 89 of SEQ ID NO: 2 each retain the P-selectin protein binding
activity and the ability
to bind to E-selectin. P-selectih ligand proteins in which one or more N-
linked glycosylation sites
(such as those at amino acids 65, 111 and 292 of SEQ ID NO: 2) have been
changed to other
amino acids or deleted also retain P-selectin protein binding activity and the
ability to bind E-
selectin. P-selectin ligand proteins comprising from amino acid 42 to amino
acid 60 of SEQ ID
NO:2 (which includes a highly anionic region of the protein from amino acid 45
to amino acid
58 of SEQ ID NO:2) also retain P-selectin ligand protein activity; however, P-
selectin ligand
proteins limited to such sequence do not bind to E-selectin. Preferably, a P-
selectin ligand protein
retains at least one (more preferably at least two and most preferably all
three) of the tyrosine
17
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14139
residues found at amino acids 46, 48 and 51 of SEQ ID NO: 2. sulfation of
which may contribute
to P-selectin ligand.protein activity. Consauction of DNAs encoding these and
other active
fragments or altered fotms of P-selectin ligand protein may be accomplished in
aecordanee with
methods known to those skilled in the-art.
The isolated DNA of the invention may be operably linked to an expression
control
sequence such as the pMT2 or pED expression vectors disclosed in Kaufinan et
al., Nucleic Acids
Res. 19, 4485-4490 (1991), in order to produce the P-selectin ligand
t+ecombinantly. Many
suitable expression control sequences are known in the art. General methods of
expressing
recombinant proteins are also known and are exemplified in R. Kaufman. Methods
in
Enzymology 185, 537-566 (1990). As defined herein "operably linked" means
tnzymatically or
15, chemically ligated to form a covalent bond between the isolated DNA of the
invention and the
expression control sequence, in such a way that the P-selectin ligand protein
is expressed by a host
cell which has been transfonned {transfected) with the ligated DNA/expr+ession
control sequence.
Several endoproteolytic enzymes are known which cleave precursor peptides at
the
carboxyl side of paired amino acid sequences (e.g., -Lys-Arg- and -Arg-Arg-)
to yield mature
proteins. Such enzymes are generally known as paired basic amino acid
converting enzymes or
PACE, and'their use in recombinant production of mature peptides is
extensively disclosed in WO =
92/09698 and U.S. Patent No. 5,460,950.
.The PACE family of enzymes are known to increase the efficiency of
proteolytic
processing of precursor polypeptides in recombinant host cells. As mentioned
above, the P-
selectin ligand protein of the invention contains such a PACE cleavage site.
The solubie mature P-selectin ligand protein of the present invention may be
made by a
host cell which contains a DNA sequence encoding any soluble P-selectin ligand
protein as
described herein and a DNA sequence encoding PACE as described in WO 92/09598
and U.S.
Patent No. 5,460,950 or using the DNA sequence '
of SEQ ID NO:5. Such a host cell may contain the DNAs as the result of co-
transformation or
sequential transformation of separate expression vectors containing the
soluble P-selectiri ligand
protein DNA and the PACE DNA, respectively. A third DNA which encodes a 3/4fT
may also
be co-transformed with the DNAs encoding the P-selectin ligand protein and
PACE.
Alternatively, the host cell may contain the DNAs as the result of
transformation of a single
expression vector containing both soluble P-seiectin ligand protein DNA and
PACE DNA.
Construction of such expression vectors is within the level of ordinary skill
in molecular biology.
Methods for co-transformation and transformation are also known.
18
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
..~,
Many DNA sequences encoding PACE are known. For example, a DNA encoding one
form of PACE, known as furin, is di'sclosed in A.M.W. van den Ouweland et al.,
Nucl. Acids Res.
18, 664 (1990), A cDNA encoding a soluble form of PACE,
known as PACESOL, is set forth in SEQ ID NO:5. DNAs encoding other forms of
PACE also
exist, and any such PACE-encoding DNA may be used to produce the solubie
mature P-selectin
ligand protein of the invention, so long as the PACE is capable ofcleaving the
P-selectin ligand
protein at amino acids 38-41. Preferably, a DNA encoding a soluble form of
PACE is,used to
produce the soluble mature P-selectin,ligand protein of the present invention.
The DNAs encoding a soluble form of the P-selectin ligand protein and PACE,
separately
or together, may be operably linked to an expression control sequence such as
those contained in
the pMT2 or pED expression vectors discussed above, in order to produce the
PACE-cleaved
soluble P-selectin ligand recombinantly. Additional suitable expression
control sequences are
known in the art. Examples 3(C) and 3(D) below set forth methods for producing
the soluble
mature P-selectin ligand protein of the invention.
A number of types of cells may act as suitable host cells forexpression of the
P-selectin
ligand protein. Suitable host cells are capable of attaching carbohydrate side
chains characteristic
of functional P-selectin ligand protein. Such capability may arise by virtue
of the presence of a
suitable glycosylating enzyme within the host cell, whether naturally
occurring, induced by
chemical mutagenesis, or through transfection of the host cell with a suitable
expression plasmid
containing a DNA sequence encoding the glycosylating enzyme. Host cells
include, for example,
monkey COS cells, Chinese Hamster Ovary (CHO) cells, human kidney 293 cells.
human
epidermal A431 cells, human Colo205 cells, 3T3 cells, CV-1 cells, other
transformed primate cell
lines, normal diploid cells, cell strains derived from in vitro culture of
primary tissue. primary
explants. HeLa cells, mouse L cells, BHK, HL-60, U937, or HaKcells.
The P-selectin ligand protein may also be produced by operably linking the
isolated DNA
of the invention and one or more DNAs encoding suitable glycosylating enzymes
to suitable
control sequences in one or more insect expression vectors. and employing an
insect expression
system. Materials and methods for baculovirus/insect cell expression systems
are commercially
available in kit form from, e.g., Invitrogen, San Diego, Califorttia, U.S.A.
{the MaxBac(D i;it), and
such methods are well known in the art, as described in Summers and Smith,
Texas Aericultural
Experiment Station Bulletin No. 1555 (1987). Soluble forms
of the P-selectin ligand protein may also be produced in insect cells using
appropriate isolated
DNAs as described above. A DNA encoding a form of PACE may further be co-
expressed in an
insect host cell to produce a PACE-cleaved form of the P-seiectin ligand
protein. _
19 a
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
Altematively, it may be possible to produce the P-selectin ligand protein in
lower
eukaryotes such as yeast or in prokaryotes such as bacteria. Potentially
suitable yeast strains
include Saccharomyces cerevisiae, Schizosaccharomyces pombe, Kluyveromvices
strains, Candida,
or any yeast strain capable of expressing heterologous proteins. Potentially
suitable bacterial
strains include Escherichia coli, Bacillus subtilis, Salmonella evphimurium,
or any bacterial strain
capable of expressing heterologous proteins. If the P-selectin ligand protein
is made in yeast or
bacteria, it is necessary to attach the appropriate carbohydrates to the
appropriate sites on the
protein moiety covalently, in order to obtain the glycosylated P-selectin
ligand protein. Such
covalent attachments may be accomplished using known chemical or enzymatic
methods.
The P-selectin ligand protein of the invention may also be expressed as a
product of
transgenic animals, e.g., as a component of the milk of transgenic cows,
goats, pigs, or sheep
which are characterized by somatic or germ cells containing a DNA sequence
encoding the P-
selectin ligand protein.
The P-selectin binding activity of a P-selectin protein may be enhanced by co-
transformation of a host cell with a GIcNAc transferase, preferably UDP-
GIcNAc:Ga1 a 1-
3Ga1NAc-R(GlcNAc to Ga1NAc)p 1-6 GIcNAc transferase (EC 2.4.1.102), also known
as "core2
transferase."
0-linked glycans present on P-selectin ligand protein have been shown to be
important
for binding to P-selectin (D. Sako et aL, Cell Z. 1179-1186 (1993)). It has
been reported that
sialyl Lel on 0-linked glycans of myeloid cells are presented on complex,
branched structures
(Maemura, K. and Fukuda, M., J. Biol. Chem. 267, 24379-24386 (1992)). The
enzyme
responsible for generating such oligosaccharide structures is "core2". The
core2 enzyme activity
is found at very low levels in COS cells and at trace levels in CHO cells.
Host cells co-
transformed with DNAs encoding a P-selectin ligand protein, an (a1,3/a1,4)
fucosyltransferase
and core2 can produce P-selectin ligand protein exhibiting 20-30 fold enhanced
binding to P-
selectin.
In certain preferred embodiments, P-selectin ligand protein is produced by co-
transfecting
a host cell with DNAs encoding soluble P-selectin ligand protein, 3/4FT, core2
and PACE.
The P-selectin ligand protein of the invention may be prepared by culturing
transformed
host cells under culture conditions necessary to express a P-selectin binding
glycoprotein. The
resulting expressed glycoprotein may then be purified from culture medium or
cell extracts.
Soluble forms of the P-selectin ligand protein of the invention can be
purified by affinity
chromatography over Lentii lectin-Sepharose and subsequent elution with 0.5M
a-methyl-
mannoside. The eluted soluble P-selectin ligand protein can then be further
purifFed and
20 a
CA 02599944 2007-09-14
. ~ ,
WO 98/08949 PCT/US97/14159
concentrated by a 0-70% ammonium sulfate precipitation step. The protein is
then recovered,
resuspended. and further purified by size exclusion chromatography over a TSK
G4000SWxL.
Alternatively, full length forms of the P-selectin ligand protein of the
invention can be purified
by preparing a total membrane fraction from the expressing cell and extracting
the membranes
with a non-ionic detergent such as Triton X- 100. The detergent extract can
then be passed over
an affinity column comprised of immobilized P-selectin, and the P-selectin
ligand protein can be
eluted from the column with 10mM EDTA in a buffer containing 0.1 % detergent.
The material
eluted from the affinity column can then be dialyzed to remove EDTA and
further purified over
a Lentil lectin-Sepharose affinity column, again eluting with 0.5M a-methyl-
mannoside.
Alternatively, the P-selectin ligand protein of the invention is concentrated
using a
commercially available protein concentration filter, for example, an Amicon or
Millipore Pellicon
ultrafiltration unit. Following the concentration step, the concentrate can be
applied to a
purification matrix such as a gel filtration medium. Altematively. an anion
exchange resin can
be employed, for example, a matrix or substrate havincy pendant
diethylaminoethyl (DEAE)
groups. The matrices can be acrylamide, agarose, dextran, cellulose or other
types commoniy
employed in protein purification. Alternatively, a cation exchange step can be
employed.
Suitable cation exchangers include various insoluble matrices comprising
sulfopropyl or
carboxymethyl groups. Sulfopropyl groups are preferred (e.g., S-Sepharose
columns). The
purification of the P-selectin ligand protein from culture supematant may also
include one or more
column steps over such affinity resins as concanavalin A-agarose, heparin-
toyopearl0 or
Cibacrom blue 3GA Sepharose ; or by hydrophobic interaction chromatography
using such
resins as phenyl ether, butyl ether, or propyl ether: or by immunoaffinity
chromatography.
Finally, one or more reverse-phase high performance liquid chromatography (RP-
HPLC)
steps employing hydrophobic RP-HPLC media, e.g., silica gel having pendant
methyl or other
aliphatic groups, can be employed to further purify the P-selectin ligand
protein. Some or all of
the foregoing purification steps, in various combinations, can also be
empioved to provide a
substantially homogeneous isolated recombinant protein. The P-selectin ligand
protein thus
purified is substantially free of other mammalian proteins and is defined in
accordance with the
present invention as "isolated P-selectin ligand protein".
Isolated P-selectin ligand protein may be useful in treating conditions
characterized by
P-, E- or L-selectin mediated intercellular adhesion. Such conditions include,
without limitation,
myocardial infarction, bacterial or viral infection. metastatic conditions,
inflammatory disorders
such as arthritis, gout, uveitis, acute respiratory distress syndrome, asthma,
emphysema, delayed
type hypersensitivity reaction, systemic lupus ervthematosus, thermal injury
such as burns or
21
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
frostbite, autoimmune thyroiditis. experimental allergic encephalomyelitis.
multiple sclerosis,
multiple organ injury syndrome secondary to trauma, diabetes, Reynaud's
syndrome, neutrophilic
dermatosis (Sweet's syndrome), inflammatory bowel disease, Grave's disease.
alomerulonephritis,
gingivitis, periodontitis, hemolytic uremic syndrome, ulcerative colitis.
Crohn's disease,
necrotizing enterocolitis, granulocyte transfusion associated syndrome,
cytokine-induced toxicity,
and the like. Isolated P-selectin ligand protein may also be useful in organ
tranplantation, both
to prepare organs for transplantation and to quell organ transplant rejection.
Accordingly, P-
selectin ligand protein may be administered to a living or non-living organ
donor, prior to organ
removal. In addition, P-selectin ligand protein may be administered "ex-vivo"
to the donor organ
concomitantly with organ preservation solution, prior to, and/or subsequent to
surgical
anastomosis with the recipient. Isolated P-selectin ligand protein mav be used
to treat
hemodialysis and leukophoresis patients. Additionally, isolated P-selectin
lisand protein may be
used as an antimetastatic agent. Isolated P-selectin ligand protein may be
used itself as an
inhibitor of P-, E- or L-selectin-mediated intercellular adhesion or to design
inhibitors of P-, E-
or L-selectin-mediated intercellular adhesion. The present invention
encompasses both
pharmaceutical compositions containing isolated P-selectin ligand protein and
therapeutic
methods of treatment or use which employ isolated P-selectin ligand protein.
Isolated P-selectin ligand protein, purified from cells or recombinantly
produced, may be
used as a pharmaceutical composition when combined with a phanmaceutically
acceptable carrier.
Such a composition may contain, in addition to P-selectin ligand protein and
carrier, diluents,
fillers, salts, buffers, stabilizers, solubilizers, and other materials well
known in the art. The term
"pharmaceutically acceptable" means a non-toxic material that does not
interfere with the
effectiveness of the biological activity of the active ingredient(s). The
characteristics of the carrier
will depend on the route of administration. The pharmaceutical composition of
the invention may
also contain cytokines, lymphokines, or other hematopoietic factors such as M-
CSF, GM-CSF,
IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, II.-10, IL-11, IL-12, G-
CSF, Meg-CSF, stem
cell factor, and erythropoietin. The pharmaceutical composition may contain
thrombolytic or anti-
thrombotic factors such as plasminogen activator and Factor VIII. The
pharmaceutical
composition may further contain other anti-inflammatory agents. Such
additional factors and/or
agents may be included in the pharmaceutical composition to produce a
synergistic effect with
isolated P-selectin ligand protein, or to minimize side effects caused by the
isolated P-selectin
ligand protein. Conversely, isolated P-selectin ligand protein may be included
in formulations
of the panicular cytokine, lymphokine, other hematopoietic factor,
thrombolytic or anti-
thrombotic factor, or anti-inflammatory agent to minimize side effects of the
cytokine,
22 =
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
lymphokine, other hematopoietic factor, thrombolytic or anti-thrombotic
factor, or anti-
inflammatory agent.
The pharmaceutical composition of the invention may be in the form of a
liposome in
which isolated P-selectin ligand protein is combined, in addition to other
pharmaceutically
acceptable carriers, with amphipathic agents such as lipids which exist in
aggregated form as
] 0 micelles, insoluble monolayers, liquid crystals, or lame=llar layers which
in aqueous solution.
Suitable lipids for liposomal formulation include, without limitation,
rnonoglyc;rides,
diglycerides, sulfatides, lysolecithin, phospholipids, saponin, bile acids,
and the like. Preparation
of such liposomal formulations is within the level of skill in the art, as
disclosed, for example, in
U.S. Patent No. 4,235,871; U.S. Patent No. 4,501,728; U.S. Patent No.
4,837,028; and U.S.
Patent No. 4,737,323,
As used herein, the term "therapeutically effective amount" means the total
amount of
each active component of the pharmaceutical composition or method that is
sufficient to show a
meaningful patient benefit, i.e., healing of chronic conditions characterized
by P-selectin- or E-
selecti n -mediated cellular adhesion or increase in rate of healing of such
conditions. When
applied to an individual active ingredient, administered alone, the term
refers to that ingredient
alone. When applied to a combination, the term refers to combined amounts of
the active
ingredients that result in the therapeutic effect, whether administered in
combination, serially or
simultaneously.
In practicing the method of treatment or use of the present invention, a
therapeutically
effective amount of isolated P-selectin ligand protein is administered to a
mammal having a P-
selecti n -mediated disease state. Isolated P-selectin ligand protein may be
administered in
accordance with the method of the invention either alone or in combination
with other therapies
such as treatments employing receptor antagonists, ligand antagonists.
cytokines, lymphokines
or other hematopoietic factors. When co-administered with one or more
cytokines, lymphokines
or other hematopoietic factors, isolated P-selectin ligand protein may be
administered either
simultaneously with the cytokine(s), lymphokine(s), other hematopoietic
factor(s), thrombolytic
or anti-thrombotic factors, or sequentially. If administered sequentially, the
attending physician
will decide on the appropriate sequence of administering isolated P-selectin
ligand protein in
combination with cytokine(s), lymphokine(s), other hematopoietic factor(s),
thrombolytic or anti-
thrombotic factors.
Administration of isolated P-selectin ligand protein used in the
pharmaceutical
composition or to practice the method of the present invention can be carried
out in a variety of
23
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
conventional ways, such as oral ingestion, inhalation, or cutaneous,
subcutaneous, or intravenous
injection. Intravenous administration to the patient is preferred.
When a therapeutically effective amount of isolated P-selectin ligand protein
is
administered orally, isolated P-selectin ligand protein will be in the form of
a tablet, capsule,
powder, solution or elixir. When administered in tablet form, the
phannaceutical composition of
the invention may additionally contain a solid carrier such as a gelatin or an
adjuvant. The tablet,
capsule, and powder contain from about 5 to 95% isolated P-selectin ligand
protein, and
preferably from about 25 to 90% isolated P-selectin ligand protein. When
administered in liquid
form, a liquid carrier such as water, petroleum, oils of animal or plant
origin such as peanut oil,
mineral oil, soybean oil, or sesame oil, or synthetic oils may be added. The
liquid fotm of the
pharmaceutical composition may further contain physiological saline solution,
dextrose or other
saccharide solution, or glycols such as ethylene glycol, propylene glycol or
polyethylene glycol..
When administered in liquid form, the pharmaceutical composition contains from
about 0.5 to
90% by weight of isolated P-selectin ligand protein and preferably from about
I to 50% isolated
P-selectin ligand protein.
When a therapeutically effective amount of isolated P-selectin ligand protein
is
administered by intravenous, cutaneous or subcutaneous injection, isolated P-
seiectin ligand
protein will be in the form of a pyrogen-free, parenterally acceptable aqueous
solution. The
preparation of such parenterally acceptable protein solutions, having due
regard to pH, isotonicity,
stability, and the like, is within the skill in the art. A preferred
pharmaceutical composition for
intravenous, cutaneous, or subcutaneous injection should contain, in addition
to isolated P-selectin
ligand protein an isotonic vehicle such as Sodium Chloride Injection, Ringer's
Injection, Dextrose
Injection. Dextrose and Sodium Chloride Injection, Lactated Ringer's
Injection, or other vehicle
as known in the art. The pharmaceutical composition of the present invention
may also contain
stabilizers, preservatives, buffers, antioxidants, or other additive known to
those of skill in the art.
The amount of isolated P-selectin ligand protein in the pharmaceutical
composition of the
present invention will depend upon the nature and severity of the condition
being treated, and on
the nature of prior treatments which the patient has undergone. Ultimately,
the attending
physician will decide the amount of isolated P-selectin ligand protein with
which to treat each
individual patient. Initially, the attending physician will administer low
doses of isolated P-
selectin ligand protein and observe the patient's response. Larger doses of
isolated P-selectin
ligand protein may be administered until the optimal therapeutic effect is
obtained for the patient,
and at that point the dosage is not increased further. It is contemplated that
the various
24
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
pharmaceutical compositions used to practice the method of the present
invention should contain
about 0.1 g to about 100 mg of isolated P-selectin ligand protein per kg body
weight.
The duration of intravenous therapy using the pharmaceutical composition of
the present
invention will vary, depending on the severity of the disease being treated
and the condition and
potential idiosyncratic response of each individual patient. It is
contemplated that the duration
of each application of the isolated P-selectin ligand protein will be in the
range of 12 to 24 hours
of continuous intravenous administration. Ultimately the attending physician
will decide on the
appropriate duration of intravenous therapy using the pharmaceutical
composition of the present
invention.
Isolated P-selectin ligand protein of the invention may also be used to
immunize animals
to obtain polyclonal and monoclonal antibodies which specifically react with
the P-selectin ligand
protein and which may inhibit P-selectin-mediated cellular adhesion. Such
antibodies may be
obtained using the entire P-selectin ligand protein as an immunogen, or by
using fragments of P-
selectin ligand protein such as the soluble mature P-selectin ligand protein.
Smaller fragments
of the P-selectin ligand protein may also be used to immunize animals, such as
the fragments set
forth below: amino acid 42 to amino acid 56 of SEQ ID NO:2 and amino acid 127
to amino acid
138 of SEQ ID NO:2. An additional peptide immunogen comprises amino acid 238
to amino acid
248 of SEQ ID NO:2, with an alanine residue added to the amino terminus of the
peptide.
Another peptide immunogen comprises amino acid 43 to amino acid 56 of SEQ ID
NO:2 having
a sulfated tyrosine in any or all of positions 46, 48 or 51. The peptide
immunogens additionally
may contain a cysteine residue at the carboxyl terminus, and are conjugated to
a hapten such as
keyhole limpet hemocyanin (KLH). Additional peptide immunogens may be
generated by
replacine tyrosine residues with sulfated tyrosine residues. Methods for
synthesizing such
peptides are known in the art, for example, as in R.P. Merrifield,
J.Amer.Chem.Soc. 85, 2149-
2154 (1963); J.L. Krstenansky, er al., FEBS Lett. 211, 10 (1987).
Monoclonal antibodies binding to P-selectin ligand glycoprotein or to complex
carbohydrate moieties characteristic of the P-selectin ligand glycoprotein may
be useful diagnostic
agents for the immunodetection of inflammatory diseases and some forms of
cancer. Some
cancerous cells, such as small cell lung carcinomas, may express detectable
levels of the P-selectin
ligand protein. This abnormal expression of the P-selectin ligand protein by
cancer cells may play
a role in the metastasis of these cells.
Neutralizing monoclonal antibodies binding to P-selectin ligand giycoprotein
or to
complex carbohydrates characteristic of P-selectin ligand glycoprotein may
also be useful
therapeutics for both inflammatory diseases and also in the treatment of some
forms of cancer
25 =
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
where abnormal expression of P-selectin ligand protein is involved. These
neutralizing
monoclonal antibodies are capable of blocking the selectin mediated
intercellular adherence
function of the P-selectin ligand protein. By blocking the binding of P-
selectin ligand protein,
the adherence of leukocytes to sites of inappropriate inflammation is either
abolished or markedly
reduced. In the case of cancerous cells or leukemic cells, neutralizing
monoclonal antibodies
against P-selectin ligand protein may be useful in detecting and preventing
the metastatic spread
of the cancerous cells which may be mediated by the P-selectin ligand protein.
In addition, the
monoclonal antibodies bound to these cells may target the cancerous cells for
antibody-dependent
cell medicated cytoxicity (ADCC), thus helping to eliminate the cancerous
cells. Human
antibodies which react with the P-selectin ligand protein may be produced in
transgenic animals
which contain human immunoglobulin encoding genes in their getm lines. Example
7 below sets
forth production of a rabbit polyclonal antibody specific P-selectin ligand
protein fragments.
P-selectin ligand protein of the invention may also be used to screen for
agents which are
capable of binding to P-selectin ligand protein and thus may act as inhibitors
of P-selectin- or E-
selec ti n- mediated intercellular adhesion. Binding assays using a desired
binding protein,
immobilized or not, are well known in the art and may be used for this purpose
using the P-
selectin ligand protein of the invention. Appropriate screening assays may be
cell-based, as in
Examples 3 and 9 below. Alternatively, purified protein based screening assays
may be used to
identify such agents. For example, P-selectin ligand protein may be
immobilized in purified form
on a carrier and binding to purified P-selectin may be measured in the
presence and in the absence
of potential inhibiting agents. A suitable binding assay may alternatively
employ purified P-
selectin immobilized on a carrier, with a soluble form of P-selectin ligand
protein of the invention.
Any P-selectin ligand protein may be used in the screening assays described
above. For
example, the full-length P-selectin ligand protein set forth in SEQ ID NO:2
from amino acid I to
amino acid 402 may be used to screen for inhibitors; or the mature P-selectin
ligand protein set
forth in SEQ ID NO:2 from amino acid 42 to amino acid 402 may be used to
screen for inhibitors,
or the soluble mature P-selectin ligand protein set forth in SEQ ID NO:2 from
amino acid 42 to
amino acid 310 may be used to screen for inhibitors. Alternatively, the P-
selectin ligand protein
of SEQ ID NO:4 from amino acid I to amino acid 412, or a mature form of the P-
selectin ligand
protein as set forth in SEQ ID NO:4 from amino acid 42 to amino acid 412, or a
soluble mature
form of the P-selectin ligand protein set forth in SEQ ID NO:4 from amino acid
42 to amino acid
320 may be used to screen for inhibitors of intercellular adhesion in
accordance with the present
invention.
26
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
In such a screening assay, a first binding mixture is formed by combining P-
selectin or
E-selectin and P-selectin ligand protein, and the amount of binding in the
first binding mixture
(Ba) is measured. A second binding mixture is also formed by combining P-, E-
or L-selectin, P-
selectin ligand protein, and the compound or agent to be screened, and the
amount of binding in
the second binding mixture (B) is measured. The amounts of binding in the
first and second
binding mixtures are compared, for example, by performing a BBo calculation. A
compound or
agent is considered to be capable of inhibiting P-, E- or L-selectin mediated
intercellular adhesion
if a decrease in binding in the second binding mixture as compared to the
first binding mixture
is observed. The formulation and optimization of binding mixtures is within
the level of skill in
the art, such binding mixtures may also contain buffers and salts necessary to
enhance or to
optimize binding, and additional control assays may be included in the
screening assay of the
invention.
Compounds found to reduce by at least about 10%, preferably greater than about
50% or
more of the binding activity of P-selectin ligand protein to P-, E- or L-
selectin may thus be
identified and then secondarily screened in other seiectin binding assays,
including assays binding
to L-selectin and in vivo assays. By these means compounds having inhibitory
activity for
selectin-mediated intercellular adhesion which may be suitable as anti-
inflammatory agents may
be identified.
EXAMPLE I
CLONING OF THE P-SELECTIN LIGAND PROTEIN GENE
A. Construction of the HL60 cDNA librarv
An HL60 cDNA library was constructed for expression cloning the P-selectin
ligand.
PolyA' RNA was isolated from total RNA from the human promyelocytic cell line
HL60 (S.J.
Collins, et al., supra) using a Fast Track mRNA Isolation Kit (Invitrogen; San
Diego, CA).
Double stranded cDNA was synthesized from the polyA+ RNA fraction and blunt-
end ligated with
EcoRI adaptors (5'-AATTCCGTCGACTCTAGAG-3', SEQ ID NO:7; 5'-CTCTAGAGTCGACGG-3',
SEQ ID NO:8). The cDNA was ligated into the expression vector pMT21 (R.
Kaufman et al.,
J. Mol. Cell. Biol. 9, 946-958 (1989) that had been incubated sequentially
with EcoRI
endonuclease and calf intestinal alkaline phosphatase and gel purified. The
ligation product was
electroporated in 2 F-l aliquots into competent E. coli DH5a cells and grown
in I ml of SOB
medium (J. Sambrook et al., Molecular Cloning: A Laboratorv Manual, New York,
Cold Spring
Harbor Laboratory Press, p1.90 (1989)) which has been supplemented with 10 mM
MgCI,, 10
27
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
mM MgSO4, and 2% glycerol for one hour at 37 C. In order to divide the library
into smaller
subsets, an aliquot from each ml of bacterial suspension was plated onto agar
plates in the
presence of ampicillin, and the number of colonies per ml was calculated.
Assuming that each
colony represented one cDNA clone, 600,000 clones were generated and divided
into subsets of
approximately 16,000 clones per pool. Each of the 38 pools were grown
overnight in L-broth in
the presence of ampicillin and the plasmids were purified over a CsCI
gradient.
B. Screenine for the P-selectin ligand protein gene
In the first stage,the LEC-y I binding assay of Example 4(A) was utilized to
pan the HL60
cDNA library and thereby to enrich for the plasmid of interest. Six g of each
HL60 cDNA
library pool was co-transfected with 2 g of a 3/4ET gene (Example 2) into COS
cells.
Approximately 45 hours post-transfection, the COS cells were lifted from the
plates by incubating
the cells in 1 mM EGTA for 15 min. at 37 C, followed by scraping with cell
lifters. The cells
were washed twice in Hanks buffered saline solution containing 1 mM calcium
(HBSS). The cells
were resuspended in 4 ml of HBSS. The resuspended transfected COS cells were
screened using
the LEC-y I binding assay described in Example 4(A).
The plasmids from adherent COS cells were recovered from a Hirts extract [B.
Hirts, J.
- Mol. Biol., 26, 365-369 (1967)] and then electroporated into E_coli DH5a
cells for amplification.
The enriched population of plasmids was purified over a CsCi gradient and re-
transfected along
with the 3/4FT gene (Example 2) into COS cells. The transfection, screening,
and plasmid
amplification process was repeated for a total of three times before a pool
that bound to the LEC-
y1-coated plates was visually detected. The positive plasmid pool was
subsequently broken down
into subsets. This involved electroporating the Hirts extract from the
positive pool into E. coli
DH5a cells and quantitating colonies per ml as described above. Various pool
sizes were
producecTFy plating out a predetermined number of colonies on agar plates in
the presence of
ampicillin. Duplicate plates were prepared by performing nitrocellulose lifts
and storing the filters
on new agar plates. The duplicate plates served as reference plates for
selecting individual or
groups of colonies from any pool identified as being positive.
In the second stage of cloning, COS cells were co-transfected with the
sublibrary pools
and the 3/4FT gene by the same procedure used in the initial steps of
screening. Forty-eight hours
post-transfection, the transfected cells were screened using the fluorescent
CHO:P-selectin assay
of Example 4(B). Positive pools were further subdivided, as described above,
until finally
individual colonies were screened and positive clones identified. Using this
method, a single
positive clone, pMT21:PL85, was found to encode the P-selectin ligand protein.
The DNA
28 ;
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
sequence of the P-selectin ligand contained in pMT21:PL85 is set forth in SEQ
ID NO:1, and the
binding characteristics of the P-selectin ligand protein encoded by pMT21:PL85
are set forth in
Example 4(C) below.
EXAMPLE 2
CLONING THE a 1.3/1,4 FUCOSYLTRANSFERASE GENE
The a 1,3/1,4 fucosyltransferase gene (3/4FT) was cloned from total human
genomic
DNA (Clontech Laboratories) by means of PCR. The sense oligonucleotide primer
contained an
Xbal site and the 5' terminus of the gene (5'-
TAGCATACGCTCTAGAGCATGGATCCCCTGGGTGCAGCCAAGC-3', SEQ ID NO:9), and the
antisense oligonucleotide primer contained an EcoRI site and the 3' terminus
of the gene (5'-
CCGGAATTCTCAGGTGAACCAAGCCGC-3', SEQ ID NO: 10). The PCR product was
sequentially
digested with Xbal and EcoRI and purified by standard gel purification
methods. This gene was
then ligated with vector pMT3Sv2ADA (R. Kaufman, Methods in Enzymology, supra)
that had
also been sequentially digested with XbaI and EcoRl and purified by standard
gel purification
methods. Competent HB 101 cells (Biorad) were transformed with this ligation
product and then
plated on agar plates in the presence of ampicillin. Nitrocellulose filter
lifts of ampicillin-resistant
transformants were probed with a radiolabelled oligonucleotide (5'-
AAGTATCTGTCCAGGGCTTCCAGGT-3', SEQ ID NO: 11) complementary to the nucleotide
region
506-530 in the middle of the gene (J. Sambrook et al., supra).
Plasmid DNA minipreps were prepared from twelve positive clones. The purified
DNA
was then digested with EcoR1 and Xbal to identify the correct clone with the
proper size insert.
This clone (pEA.3/4FT) was then grown up large scale and the DNA isolated by
CsCI density
gradient banding (J. Sambrook et al., supra). DNA sequencing confirmed the
identity of the
3/4FT gene. The functionality of the gene was assessed in a cell-cell binding
assay as follows.
COS-1 monkey cells [(clone M6; M. Horwitz et al., Mol. Appl. Genet., 2:147-
149, (1983)) were
transfected with 3/4Ff using DEAE dextran followed by DMSO shock treatment and
chloroquine
incubation [L. Sompeyrac and K. Dana, Proc. Natl. Acad. Sci., 78:7575-7578
(1981); M. Lopata
et al., Nucleic Acids Res., 12:5707-5717, (1984); H. Luthman and G. Magnuson,
Nucleic Acids
Res., 11:1295-1308, (1983)]. The transfected COS cells were suspended and
quantitated for
binding to a CHO line expressing E-selectin [G. Larsen et a1., J. Biol. Chem.
267:11104-11110,
(1992)]. This assay confirmed that the COS cells transfected with 314FI' can
express the siaviated
Lewis' epitope on the cell surface.
29 4
CA 02599944 2007-09-14
WO 98/08949 PGT/US97/14159
EXAMPLE 3
EXPRESSION OF THE P-SELECTIN LIGAND PROTEIN
A. Expression of the P-selectin Lieand in LEC11 cells
Functional P-selectin ligand was expressed in the SLe'-positive Chinese
hamster ovary
(CHO) cell line LECI I (Campbell, C. and Stanley, P.Cell 35:303-309 (1983) as
follows:
approximately 8 pg of plasmid containing the P-selectin ligand gene
(pMT21:PL85, Example 1)
was transfected into LEC 11 cells. At 68 hours post-transfection, the cells
wer-e treated with 2.5
mM sodium butyrate for 4 hours. The cells were observed to induce P-selectin
adhesion, as
determined using the 6-CFD labeled CHO:P-selectin cell binding assay
(described in Example
4, section B). In contrast, neither LEC11 cells alone nor LECI I cells
transfected with a control
plasmid induced P-selectin adhesion.
B. Expression of Soluble P-Selectin LiQand in COS cells
COS cells were transfected with 8 pg pED.sPSL.T7 (see Example 5C) and 4 g
pEA.3/4
FT plasmid of Example 2, 8 pg pED.sPSL.T7 alone, or 8 pg plasmid vector
(pMT21) and 4 g
pEA.3/4 FT gene. Forty-five hr post-transfection, the cells were rinsed twice
in PBS and
incubated overnight at 37 C in serum-free DMEM minus phenol red (JRH
Biosciences)
supplemented with 2 mM L-glutamine, 100 U/mi penicillin and 100 g/mi
streptomycin.
Phenylmethylsulfonyl fluoride, aprotinin and NaN3 were added to final
concentrations of 1mM,
2 g/ml and 0.02%, respectively, and the conditioned medium was centrifuged to
remove all
debris.
For immunoprecipitation experiments, the labeled soluble P-selectin ligand
protein was
produced by co-transfecting COS cells with pED.sPSL.T7 and pEA.3/4 FT. At
forty-five hr post-
transfection, the COS cells were labeled with 250 Ci/ml'SS methionine (NEN)
for 5 hours and
the medium was collected. Expression of sPSL.T7 protein was confirmed by
immunoprecipitation with anti-T7 antibodies.
C. Expression of PACE-cleaved P-selectin ligand in COS Cells
COS cells were co-transfected with the pED.sPSL.T7 plasmid of Example 5(C),
the
pEA.3/4FT cDNA of Example 2, and a plasmid containing the PACE cDNA as set
forth in SEQ
ID NO:5. A parallel control co-transfection was done using only the
pED.sPSL.T7 plasmid and
the pEA.3/4FT plasmid. After 45 hours. conditioned medium from these
transfected COS cells
30 4
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
was coated onto plastic dishes and binding to CHO:P-selectin cells (Example 4)
was determined.
An approximately two-fold increase in bound CHO:P-selectin cells was observed
for dishes
coated with medium containing the P-selectin ligand co-expressed with PACE, as
compared with
medium containing P-selectin ligand which had not been co-expressed with PACE.
Amino acid
sequencing of the N-terminus of purified sPSL.T7 protein from the PACE co-
transfection showed
that all of the ligand had been cleaved at the PACE consensus site (amino
acids 38-41 of SEQ ID
NO: 1). Radiolabelling of co-transfected COS cells with'SS -methionine and
subsequent SDS-
poiyacrylamide gel electrophoresis and autoradiography showed that comparable
quantities of the
P-selectin ligand had been secreted in both co-transfections.
D. Expression of the P-selectin Ligand Protein in CHO Cells
A full-length form (amino acids 1-402) of the P-selectin ligand protein was
expressed in
the CHO(DUKX) cell line (Urlaub & Chasin, Proc. Natl. Acad. Sci. USA 77. 4216-
4220 (1980))
as follows: approximately 25 pg of the pMT21:PL85 plasmid and approximately 8
g of the
pED.3/4FT (produced by restriction of pEA.3/4FT with EcoRl and Xba1 and
insertion of the
resulting fragment into the pED plasmid) were co-transfected into CHO(DUKX)
cells using the
calcium phosphate method. Transfectants were selected for resistance to
methotrexate. After two
weeks, individual colonies were screened for SLe' expression by using a
conjugate of an anti SLe
antibody (CSLEX-1, U.S. 4,752,569) and sheep red blood cells (sRBC) prepared
by the chromic
chloride method (Goding, J. W., J. Immunol. Methods 10:61-66 (1976) as
follows: sRBC were
washed with 0. I5M NaCI until the wash became clear and then a 50% suspension
of sRBC was
prepared in 0.15M Nacl. One ml of 0.01 % chromic chloride solution was added
dropwise while
vortexing to 0.2 ml of a sRBC suspension containing 50 Ng of CSLEX- 1. After
incubating at
37 C for 30 minutes, 10 ml of phosphate buffered saline (PBS) solution was
added to the
reaction. The conjugate was washed once before resuspending into 10 ml of PBS.
The plates
containing transfectants were washed with PBS and then 3 ml of PBS and one ml
of the
sRBC/CSLEX-1 conjugate was added to each plate. Positive colonies were red on
a
transilluminator and were picked into alpha medium with 10% fetal bovine
serum. After two
weeks, colonies were subjected to stepwise amplification using methotrexate at
concentrations of
2, 10, 25, 100, 250 nM. The stable cell line obtained was designated CD-PSGL-1
(R3.4).
Expression of the P-selectin ligand protein was confirmed by
immunoprecipitation studies using
the polyclonal anti-P-selectin ligand protein antibody of Example 7(A). The
functionality of the
P-selectin ligand protein produced by the CD-PSGL-I (R3.4) cell line was
tested by assaying the
transfectants for binding to LEC-y I as in Example 4(A).
31
CA 02599944 2007-09-14 -
WO 98/08949 PCT/US97/14259
The sPSL.T7 protein was expressed in a stable CHO-PACE line which was alreadv
expressing the cDNA encoding PACE as set forth in SEQ ID NO:5 under adenosine
deaminase
selection (Kaufman, er at., PNAS (USA) 83:3136-3140 (1986)). The psPSL.T7 (25
pg) and
pED.3/4FT (8 g) plasmids were cotransfected into CHO-PACE cells using the
calcium
phosphate method. Transfectants were selected for resistance to methotrexate,
and individual
colonies which bound to the sRBC/CSLEX-1 conjugate were picked. After two
weeks in culture,
the colonies were subjected to stepwise amplification as described above. The
stable cell line
obtained was designated CP/PSL-T7 (R4. 1). Expression of sPSL.T7 protein was
confirmed by
standard immunoprecipitation methods using either a T7 specific monoclonal
antibody or the
LEC-y I chimera of Example 4(A). In a similar fashion, a stable cell line
expressing the mature
full length form (amino acids 42-402) of the P-selectin ligand protein was
obtained by co-
transfection of pMT21:PL85 and pED.3/4FT into the CHO-PACE line.
Stable cell lines expressing the sPSL.Q protein of Example 5(B) and the
sPSL.Fc protein
of Example 5(D) were constructed as follows: plasmids pED.sPSL.Q (25 pg) or
pED.sPSL.Fc
(25 pg) were cotransfected with approximately 25 pg of the pED.3/4FT plasmid
described above
and approximately 20 pg of a plasmid containing the PACE cDNA as set forth in
SEQ ID NO:5)
as well as the neomycin resistance gene into CHO(DUKX) cells using the calcium
phosphate
method. Transfectants were selected for resistance to methotrexate and the
G418 antibiotic.
Approximately two weeks later, individual colonies were screened for SLe'
expression using
sRBC/CSLEX-1 conjugate binding. The positive colonies were picked in G418
medium at 1
mg/ml concentration. After 2-3 weeks in culture, cells were amplified with
methotrexate in a
stepwise selection. The stable cell lines obtained were designated CD-sPSL.Q
(R8.2) and CD-
sPSL.Fc (R8.1), respectively. The expression of sPSL.Q and sPSL.Fc protein was
confirmed by
standard immunoprecipitation method using the anti P-selectin ligand protein
polyclonal antibody
of Example 7(A).
EXAMPLE 4
ASSAYS OF P-SELECTIN-MEDIATED INTERCELLULAR ADHESION
A. LEC-yl Bindin>? Assay
A DNA encoding a chimeric form of P-selectin conjugated to the Fc portion of a
human
IgGy I (LEC-y 1) was constructed using known methods (AruffQ et al. Cell 67,
35-44 { 1991)), and
32 4
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
stably transfected into dhfr" CHO cells (CHO DUKX) for high level production
of the chimeric
LEC-y I protein. which was purified for use in the binding assay set forth
below.
Petri dishes were coated first with a polyclonal anti-human IgGy I Fc antibody
and then
with LEC-y 1. This method orients the LEC-y I construct such that the P-
selectin portion of the
chimeric molecule is presented on the surface of the plates. Adhesion of HL60
cells to the
oriented LEC-y1 was quantitated in the presence and absence of calcium. HL60
adhesion was
shown to be calcium dependent, confirming that the chimeric molecule had
retained functional
binding of P-selectin to its ligand on HL60 cells. The binding of HL60 cells
to oriented LEC-y 1
was also shown to be blocked by a neutralizing monoclonal antibody to P-
selectin, demonstrating
the specificity of P-selectin binding.
B. Fluorescent CHO-P-selectin BindinQ Assay
The assay employed a fluorescently labeled CHO:P-selectin cell line (Larsen et
al., J.
Biol. Chem. 267, 11104-11110 (1992)) that can bind to and form clusters on the
surface of COS
cells that are co-transfected with the P-selectin ligand gene and the 3/4 FT
gene. The CHO:P-
selectin cells were suspended at 1.5 x 106 cells/mi in 1% fetal bovine serum
in DME medium and
labeled by adding 6-carboxyfluorescein diacetate (6-CFD) to a final
concentration of 100 ug/ml.
After incubation at 37 C for 15 minutes, the cells were washed in medium and
resuspended at I
x 105 cells/mi. Five ml of the labeled cells were added to each washed COS
transfectant-
containing plate to be assayed and incubated at room temperature for 10
minutes. Nonadherent
cells were removed by four washes with medium. The plates were then scanned by
fluorescence
microscopy for rosettes of adherent CHO:P-selectin cells.
C. Ouantitative adhesion assav usinQ radioactivelv
labeled CHO:P-selectin cells
COS cells were co-transfected with the pMT21:PL85 plasmid of Example I and the
pEA.3/4FT plasmid of Example 2 by the same procedure used in the initial
stages of screening.
As controls, COS cells were transfected with pMT21:PL85 alone, or with
pEA.3/4FT alone, or
with a similar plasmid containing no insert ("mock"). 24 hours post-
transfection, the transfected
cells were trypsinized and distributed into Costar 6-well tissue culture
plates. CHO:P-selectin
cells were labeled for 16 hours with 3H-thymidine using known methods and
preincubated at 0.5
x 106 cells/ml for 30 minutes at 4 C in a medium containing 1% BSA tcontrol);
a medium
containing 19o BSA, 5 mM EDTA and 5 mM EGTA; a medium containing 1% BSA and 10
g/ml of a neutralizing anti P-selectin monoclonal antibody; and a medium
containing 19c BSA
and a non-neutralizing anti-P-selectin monoclonal antibody. The preincubated
cells were then
33 =
CA 02599944 2007-09-14
WO 98/08949 PCTlUS97/14159
added to the wells containing the transfected COS cells. After a 10 minute
incubation, unbound
cells were removed by 4 changes of medium. The bound CHO:P-selectin cells were
released by
trypsinization and quantified by scintillation counting.
COS cells co-transfected with P-selectin ligand and the 3/4FT induced
approximately 5.4-
fold more binding of CHO:P-selectin cells relative to COS mock cells; assay in
the presence of
EGTA and EDTA reduced binding to the level of the mock transfected COS cells.
Likewise,
incubation with neutralizing anti-P-selectin antibody also eliminated specific
binding, whereas
non-neutralizing antibody had no effect. In contrast, the binding of CHO:P-
selectin to COS cells
transfected with P-selectin ligand alone was not statistically different than
binding to the mock-
transfected COS in both the presence or absence of EDTA and EGTA, or anti-P-
selectin
antibodies. The binding of CHO:P-selectin cells to COS cells transfected with
3/4 FT alone was
approximately 2-fold greater than to the mock-transfected COS, but was
unaffected by the
presence or absence of EDTA and EGTA.
EXAMPLE 5
CONSTRUCTION OF SOLUBLE P-SELECTIN LIGANDS
The EcoRl adaptors used to generate the cDNA library from HL60 cells in
Example I
contain an Xbal restriction site (TCTAGA) just 5' of the beginning of SEQ ID
NO: I as it is
located in the pMT21:PL85 plasmid. In order to generate soluble forms of the
PSL, the
pMT21:PL85 plasmid was restricted with XbaI and with HinclI (which cleaves
after nucleotide
944 of SEQ ID NO:1). The approximately 950 bp fragment thus generated,
containing all of the
encoded extracellular segment of the ligand up to and including the codon for
valine 295, was
isolated and used to generate DNAs encoding soluble forms of the P-selectin
ligand protein as set
forth in sections A though D below.
A. Construction of nsPSL.OC
The fragment was purified and ligated into mammalian expression vector pED
between
the XbaI and EcoRI sites, along with double stranded synthetic oligonucleotide
DNA that
recreated the codons from Asn 296 to Cys 310 and introduced a novel stop codon
immediately
following Cys 310. The sequence of the oligos is as follows:
5'-AACTACCCAGTGGGAGCACCAGACCACATCTCTGfiGAAGCAGTGCTAG (SEQ ID NO: 12)
5'-AATTCTAGCACTGCTTCACAGAGATGTGGTCTGGTGCTCCCACTUGGTAGTT (SEQ ID
NO: 13)
34 ,
CA 02599944 2007-09-14
WO 98/08949 PCT/US97J14159
The resulting plasmid was designated pED.sPSL.QC, and the protein expressed
from the plasmid
was designated sPSL.QC.
B. Construction of asPSL.O
The fragment was purified and ligated into the pED plasmid (Kaufman et al.,
1991)
between the XbaI and EcoRl sites, along with the double stranded synthetic
oligonucleotide DNA
that recreated the codons from Asn 296 to Gln 309 and introduced a novel stop
codon
immediately following Gln 309. The sequence of the oligos is as follows:
5'-AACTACCCAGTGGGAGCACCAGACCACATCTCTGTGAAGCAGTAG (SEQ ID NO: 14)
5'-AATTCTACTGCTTCACAGAGATGTGGTCTGGTGCTCCCACTGGGTAGTT (SEQ ID NO: 15)
The resulting plasmid was designated pED.sPSL.Q, and the protein expressed
from the plasmid
was designated sPSL.Q.
C. Construction of vsPSL.T7
Oligonucleotides encoding 14 amino acids including an epitope derived from the
phage
T7 major capsid protein were synthesized, creating a C-terminal fusion of the
epitope "tag" with
an additional 32 amino acids derived from the vector sequence. Two
oligonucleotides having the
sequences
5'-CTAGACCCGGGATGGCATCCATGACAGGAGGACAACAAATGGTAGGCCGTAG (SEQ ID NO:
16) and
5'-AATTCTACGGCCTACCCATTTGTTGTCCTCCTGTCATGGATGCCATCCCGGGT (SEQ ID
NO:17)
were dupiexed and ligated with the large Xba1-EcoRI fragment of mammalian
expression plasmid
pED. The resulting plasmid, pED.T7 was restricted with Xbal and Smal and
ligated to the 950
bp XbaJ-HinclI fragment described above, resulting in plasmid pED.sPSL.T7.
The protein resulting from expression of pED.sPSL.T7 was designated sPSL.T7.
D. Construction of Soluble P-selectin Ligand--IQGFc Chimera
The plasmid DNA encoding a soluble, extracellular form of the P-selectin
ligand protein
fused to the Fc portion of human immunoglobulin IgGi was constructed as
follows: the
mammalian expression vector pED.Fc contains sequences encoding the Fc region
of a human
IgGI with a novel linker sequence enabling the fusion of coding sequences
amino terminal to the
hinge region via a unique XbaI restriction site. A three fragment ligation was
performed: pED.Fc
was restricted with XbaI and gel purified in linear fotm. The 950 bp fragment
from pivIT21:PL85
35 4
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
described above comprised the second fragment. The third fragment consisted of
annealed
synthetic oligonucleotide DNAs having the foilowing sequence:
5' - CTGCGGCCGCAGT (SEQ ID NO: 18)
5' - CTAGACTGCGGCCGCAG (SEQ ID NO:19)
The ligation products were grown as plasmid DNAs and individual clones having
the correct
configuration were identified by DNA sequencing. The plasmid was designated
pED.PSL.Fc.
The DNA coding region of the resulting soluble P-selectin ligand /Fc fusion
protein is shown in
SEQ ID NO:6.
EXAMPLE 6
CHARACTERIZATION OF EXPRESSED P-SELECTIN LIGANDS
A. Binding Characterization of Full-Length P-selectin
Lieand Protein Exnressed on COS Cells
Co-transfection of COS cells with the pEA.3/4FT plasmid of Example 2 and the
pMT21:PL85 plasmid of Example I yields COS cells which specifically bind to
CHO:P-selectin
cells. This binding is observed only upon co-transfection of pEA.3/4FT and
pMT21:PL85: use
of either plasmid alone generates COS cells which do not bind to CHO:P-
selectin cells. No
binding is observed between the parental CHO(DUKX) cell line which does not
express P-selectin
and COS cells co-transfected with pEA.3/4FT and pMT21:PL85. The binding
between the co-
transfected COS cells and CHO:P-selectin cells is sensitive to chelators of
divalent ions such as
EDTA and EGTA, consistent with the Ca" dependency of P-selectin mediated
cellular adhesion.
A neutralizing anti-P-selectin monoclonal antibody blocked the binding between
the CHO:P-
selectin cells and the COS cells which had been co-transfected with pEA.3/4FT
and
pMT21:PL85, while a non-neutralizing anti-P-selectin monoclonal antibody had
no effect on the
binding. The antibody results indicate that the functional domain(s) of P-
selectin are required for
binding to P-selectin ligand protein expressed on the surface of COS cells.
B. Electrophoretic Characterization of Full-Leneth
P-selectin LiQand Exnressed in COS Cells -
Detergent extracts of co-transfected COS cells were prepared as follows: 45
hours post
co-transfection, approximately 1.5 x 10' cells were suspended in 5 ml of lysis
buffer ( l 0mM
Piperazine-N,N'-bis[2-ethanesulfonic acid] (PIPES) pH 7.5, 100 mM KCI, 3 mM
MgCIõ l mM
benzamidine, 0.5 g/ml leupeptin, 0.75 g/ml pepstatin, l mM ethyimaleimide,
and I Ng/m1
aprotinin) and lysed by sonication. Cellular debtis was removed by low speed
centrifigation (500
x g. 10 minutes), and a membrane fraction collected by ultracentrifugation
(100,000 x g, 60 min).
36 .
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
The high speed membrane pellet was resuspended in an extraction buffer (10 mM
3-(N-
Morpholinolpropanesulfonic acid] (MOPS) pH 7.5. 0.1 M NaCI, 0.02% NaN3, 1%
Thesit(b
(Sigma), 1 mM benzamidine, 0.5 Ng/ml leupeptin, 0.75 pg/mi pepstatin. 1 mM
ethylmaleimide,
and I Ng/ml aprotinin). Samples were then subjected to SDS polyacrylamide gel
electrophoresis
and transfer to nitrocellulose blots as follows: an aliquot of the detergent
extract was suspended
in 1% SDS loading buffer and heated for 5 minutes at 100 C before loading onto
an 8-16%
polyacrylamide gel (reduced) or a 6% gel (non-reduced) and electrophoresed in
the Laemmli
buffer system. Blots were prepared using Immobilon-P transfer membranes. The
blots were
immersed in 10 mM MOPS pH 7.5, 0.1 M NaCI, 0.02% NaN3, 1 mM MgC1õ I mM CaCIõ
and
10% non-fat milk overnight at 4 C. Blots were rinsed once in the above buffer,
minus the milk,
and incubated in blotting buffer (10 mMMOPS pH 7.5, 0.1 M NaCI, 1% bovine
serum albumin,
0.05% Thesit, l mM MgCIõ 1 mM CaCI,) for 30 minutes at room temperature.
The blots were then probed for the P-selectin ligand as follows: 50 ng of a P-
selectin/Fc
chimera was pre-incubated with 3 pCi of''-'I-Protein A in blotting buffer for
30 minutes at room
temperature. Additional excipients (e.g., EDTA, EGTA, monoclonal antibodies)
could be added
to the pre-incubation mixture at this point to evaluate their effects on
binding of the chimera to
the P-selectin ligand. The pre-incubated mixture was then incubated with the
blots (prepared as
above) for 60 minutes at room temperature. and the blots were subsequently
washed four times
with the same blotting buffer (without bovine serum albumin), air dried, and
autoradiographed
at -70 C.
Under non-reducing conditions, two bands were observed with this technique for
membrane extracts prepared from co-transfected COS cells. The major band
migrated with an
estimated molecular weight of approximately 220 kD, whereas the minor band
migrated with a
molecular weight of approximately 110 kD. Under reducing conditions, only a
single band was
observe wRth-a molecular weight of approximately 1 10 kD, indicating that
under non-reducing
conditions, the P-selectin ligand exists as a homodimer. The approximate
molecular weight of
the reduced monomer is greater than that predicted from the deduced amino acid
sequence of the
CDNA clone (45 kD), indicating that the expressed protein undergoes extensive
post-translational
modification (see Example 6(C)). The specificity of the P-selectin/Fc chimera
was confirmed by
the observation that a nonspecific IgG, probe yielded no bands on the blots.
Additionally, the
binding of the P-selectin/Fc chimera to the blots was abolished by EDTA, EGTA,
and a
neutralizing anti-P-selectin monoclonal antibody. Specific bands on the blots
were observed only
from membrane extracts of COS cells co-transfected with the pEA.3/4FT and
pMT21:PL85
37 ,
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
plasmids. Membrane extracts from control transfections (pEA.3/4FT or
pMT21:PL85 alone)
failed to yield observable bands on blots.
C. Givcosvlation of P-selectin Ligand Protein
The presence of covalently attached carbohydrate on recombinant P-selectin
ligand and
its role in binding to P-selectin was determined as follows: COS cells were co-
transfected with
pED.sPSL.T7 of Example 5(C) and the pEA.3/4FT plasmid of Example 2. After 48
hours, the
cells were pulsed with 3sS-methionine. 200 pl of 35S methionine-labeled
sPSL.T7 conditioned
medium was incubated with 5 pg LEC-y l in the presence of 2 mM CaCI, and I
mg/ml bovine
serum albumin (BSA). After rotating for 2 hours at 4 C, Protein A-Sepharose
beads (Pharmacia)
were added for I hour at 4 C, pelleted by centrifugation and washed twice in
Tris buffered saline
( 20 mM Tris-HC1, 150 mM NaC1 pH 7.5, hereinafter TBS) containing 2 mM CaCI,
and I mg/mi
BSA. The pellets were then resuspended and treated with neuraminidase
(Streptococcus
pneumoniae), 0-glycanase, and N-glycanase (all from Genzyme) as follows. All
glycosidase
digestions were done at 37 C ovemight. For neuraminidase digestion, the pellet
was resuspended
in 50 pl 2-(N-morpholino)-ethanesulfonic acid (MES) buffer, pH 6.5
(Calbiochem) and 0.1 %
SDS, heated at 95 C for 5 minutes, then pelleted. The supematant was modified
to contain 1.4%
n-Octyl B-D-glucopyranoside (OGP), 10mM calcium acetate, 20 mM sodium
cacodylate and 2.5
mM PMSF, final pH 7.0 Eight Nl neuraminidase was added for a final
concentration of I unit/ml.
For neuraminidase/O-glycanase digestion, the sample was prepared as above and
along with the
neuraminidase, the O-glycanase was also added to a final concentration of 0.1
unit/ml. For N-
glycanase digestion, the pellet was resuspended in 54 ul MES buffer and 19o
SDS, heated at 95 C
for 5 minutes. then pelleted. The supernatant was modified to contain 0.2 M
sodium phosphate,
3.5% OGP, and 2.5 mM PMSF, final pH 8.5. N-glycanase was added for a final
concentration
of 12 units/ml and incubated as above.
The effect of glycosidase treatment on sPSL.T7 was assessed in two ways. For
this, each
digested protein sample was divided into two equal fractions. One fraction was
precipitated with
the P-selectin polyclonal antibody of Example 7(A), to show the effect of
digestion on the
electrophoretic mobility. The other fraction was precipitated with the LEC-y I
chimera of
Example 4(A), to assess the remaining P-selectin ligand binding activity after
digestion. The
immunoprecipitationed samples were analyzed by SDS-polyacryiamide gel
electrophoresis under
reducing conditions and autoradiography.
In the absence of glycosidase treatment, autoradiography revealed comparable
bands
(with molecular weights of l 10 kD) for each precipitation. When the P-
selectin ligand protein
38 .
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
was treated with neuraminidase, anti-P-selectin ligand polyclonal antibody
precipitation revealed
a slight decrease in mobility, consistent with removal of sialic acid
residues. The amount of P-
selectin ligand protein precipitated by LEC-y I was significantly reduced
after neuraminidase
treatment, consistent with the role of sialic acid residues in the P-
selectin/P-selectin ligand
interaction. When the P-selectin ligand protein was treated with both
neuraminidase and 0-
glycanase, a substantial increase in electrophoretic mobility was observed
after precipitation with
the anti-P-selectin ligand polyclonal antibody, indicating that a number of O-
linked
oligosaccharide chains had been removed. However, removal of 0-linked
oligosaccharides from
the P-selectin ligand protein may not have been complete, since the
electrophoretic mobility did
not correspond to a protein with a molecular weight of 38 kD, as would be
predicted from the
amino acid sequence set forth in SEQ ID NO:1. The neuraminidaselO-glycanase
digested P-
selectin ligand protein bound to LEC-yl very poorly, further indicating the
role of
oligosaccharides in the P-selectin/P-selectin ligand inte;action. Treatment of
the purified P-
selectin ligand with N-glycanase resulted in a slight increase in
electrophoretic mobility,
demonstrating that some of the consensus sites for N-linked glycosylation are
occupied. The
amount of P-selectin ligand protein precipitated by LEC-y 1 was slightly
reduced, indicating that
N-linked glycosylation also contributes to the P-selectin/P-selectin ligand
interaction, though not
as dramatically as sialylation and 0-linked glycosylation.
EXAMPLE 7
POLYCLONAL ANTIBODIES SPECIFIC FOR P-SELECTIN LIGANDS
A. Polyclonal Rabbit anti-P-selectin LiQand Protein/Maltose Bindine Protein
Fusion Protein
The anti-P-selectin ligand polyclonal antibody was generated by immunizing
rabbits with
a fusion protein generated in E. coli. The fusion protein consisted of the
amino terminal one-third
of the P-selectin ligand (amino acids I to 110 of SEQ ID NO:1) fused in frame
to the maltose
binding protein (Maina, C. V. er aL, Gene 74, 365-373 (1988); Riggs, P., in
Current Protocols in
Molecular BioloQv, F. M. Ausebel et al., Eds., Greene Associates/Wiley
Interscience (New York,
1990) chapter 16.6). Under conditions employed herein, the fusion protein
antibody recognizes
the P-selectin ligand protein.
B. Polvclonal Rabbit Anti-sPSL.T7 Protein
A soluble form of the invention (sPSL.T7; see example 5(C)) was purified to
apparent
homogeneity according to the following scheme: COS cells were transfected with
three plasmids, --
39
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
one encoding each of the following: sPSL.T7 (Example 5(C)), 3/4FT (Example 2),
and a soluble
form of PACE (as set forth in SEQ ID NO:5). After 72 hours, the conditioned
medium was
collected and recombinant sPSL.T7 was purified as follows.
Conditioned medium was diluted two fold with 50 mM MOPS, 150 mM NaCI. 0.5 mM
CaCIZ and 0.5 mM MnCI , pH 7.2, and applied to a column of lentil lectin-
Sepharose 4B
equilibrated in the same buffer. After loading, the column was washed with the
same buffer until
the optical absorbance at 280 nm dropped to a stable baseline. The column was
then eluted with
the same buffer which had been adjusted to 0.5 M a-methyl-mannoside and 0.3 M
NaCI.
Recombinant sPSL.T7 was collected over 5-15 column volumes of this elution
buffer. The lentil
lectin eluate was then subjected to a 0-70% ammonium sulfate precipitation by
adding 472g of
ammonium sulfate per liter of column eluate at 4 C. After stirring for 30
minutes, the precipitate
was resuspended in a minimal volume of TBS (20 mM Tris-HCI, ] 50 mM NaCI, pH
7.5) and
applied to a TSK G4000SWXL gel filtration column equilibrated in TBS. The flow
rate on the
column was 0.5 ml/min and a guard column was employed. In aliquots of < 250
l, the
resuspended ammonium sulfate pellet was injected on the column and fractions
analyzed by SDS-
PAGE with Western analysis. Fractions containing sPLS.T7 were pooled and then
used for
immunizing rabbits.
- Antibodies to sPSL.T7 were generated in the standard fashion by antigen
priming and
subsequent boosting over a 3 month period. Specifically, primary immunization
was performed
by mixing 50 pg of sPSL.T7 (denatured by mixing in 0.1 % SDS and heating for
10 minutes at
100 C) with complete Freund's adjuvant and injected at five sites
subcutaneously. The second
(and all subsequent) boosts were performed by mixing 25 pg of sPSL.T7
(denatured by mixing
in 0.1% SDS and heating for 10 minutes at 100 C) [12.5 pg for the third and
subsequent boosts]
with incomplete Freund's adjuvant and injecting at two sites subcutaneously
(or later,
intramuscu ar y) every two weeks. Test bleeds were performed every two weeks
to monitor
antibody titer. When the antibody titer reached a suitable level, a larger
scale bleed was
performed and a total serum fraction prepared. This polyclonal antibody
preparation was used
to inhibit the specific binding of HL60 cells to CHO:P-selectin cells in a
manner similar to that
described in Example 4.
This assay employed fluorescently-labeled HL60 cells (labelled with BCECFAM;
2',7'-
bis-(2-carboxymethyl)-5-(and-6)-carboxyfluorescein, acetoxymethyl ester)
binding to CHO cells
plated on the bottom of microtiter plates. The labelled HL60 cells were pre-
incubated with either
sera containing polyclonal antibody or with pre-immune sera for 30 minutes at
4 C. The cells
were then washed and incubated with the CHO:P-selectin cells for 10 minutes.
The plates were
,
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
then washed and the fluorescence read with a fluorescence microtiter plate
reader. Using this
assay, a 1: 15 dilution of the anti-sPSL.T7 polyclonal serum resulted in
essentially complete
inhibition of HL60 cell binding to CHO:P-selectin. Demonstrable inhibition of
HL60 binding to
CHO:P-selectin was still observed at antiserum dilutions of 1:150. Pre-immune
serum had no
effect on HL60 cell binding to CHO:P-selectin.
EXAMPLES
COTRANSFORMATION WITH CORE2
A. Isolation of the cDNA encoding Core2 GIcNAc Transferase
The cDNA encoding core2 G1cNAc transferase was isolated by standard molecular
biology techniques. Two oligos were designed at the 5' and 3' end (including
translational
initiation and termination codon, respectively) based on the published human
core2 sequence
(Bierhuizen, M.F.A., Fukuda, M., Proc. Natl. Acad. Sci. 89, 9326-9330 (1992)).
The pools of an
HL60 cDNA library (Sako, D., Cel175, 1179-1186 (1993)) were used as template
to amplify the
core2 coding sequence by a standard PCR protocol. The PCR amplified fragment
was purified
and subcloned into pED vector. To isolate cDNA, the pools which gave a
positive signal in the
PCR reaction were transformed into E. cofi and plated. Transformants were
transferred onto
nitrocellulose filters and hybridized with a'ZP radiolabelled PCR fragment
according to standard
protocols. Positive clones were picked and purified by replating. The sequence
of the cDNA and
PCR clone was confirmed by dideoxy sequencing.
B. Generation of Stable PSGL-1 Chinese Hamster Ovarv Cell Lines Expressing
Core2
Enzyme
A cell line made in accordance with the methods of Example 3 expressing full-
length P-
selectin ligand protein and 3/4 fucosyltransferase was co-transfected with
core2 cDNA and a
neomycin resistance gene (pMT4Neo) by standard calcium phosphate methods.
After about two
weeks. stable G418-resistant transfectants were picked either as single
isolates or in a pool. These
transfectants were grown in I mg/ml G418 complete DMEM media and analyzed for
core2
enzyme activity (Higgins, E.A., et aL, J. Biol. Chem. 266, 6280-6290 (1991)).
Positive clones
or pools found positive for core2 activity were analyzed for P-selectin ligand
binding to P-selectin
by various methods. In a similar fashion, cell lines expressing either P-
selectin ligand protein or
soluble P-selectin ligand protein with both the 3/4 fucosyltransferase and
PACE enzymes (see
Example 3) were used to isolate stable cotransfectants of core2 as described
above.
C. Effects of Core2 on P-selectin Bindine Activitv
The effects of core2 on P-selectin binding activity was evaluated by three
different
methods:
41 =
CA 02599944 2007-09-14
WO 98108949 PCT/US97/14159
1. Bindins of mPSGL-1 transfectants to immobilized soluble P-selectin or P-
selectin/IeG chimera.
48- well plates were coated with 1 ug/ml anti human Fc antibody in 50 mM Tris
pH 9.5
at 4 C for five to six hours. After washing twice with HBSS buffer, P-
selectin/IgG chimera (0.1-
1 ug/m) conc., Example 5) was plated in HBSS buffer overnight at 4 C. The
plates were blocked
with BSA (3 mg/ml) at 4 C for three to four hours. In the case of soluble P-
selectin ligand
protein, the protein was coated directly onto plates in the same buffer. The
3H labelled CHO cells
were lifted with 2 mM EGTA, washed three times with PBS, and resuspended to a
final density
of l0 cells/ml. A 300 ul aliquot of this suspension was added to each well
(300,000 cell/well).
After incubating for 12 minutes at room temperature, wells were washed four
times with serum
free DMEM to remove unbound cells. Bound cells were lifted with 5 mM EGTA and
counted
in scintillation counter. U937 cells, used as a positive control for native P-
selectin ligand protein
binding, were pretreated with gamma globulin (5 mg/ml) to block endogenous Fc
receptor before
binding to P-selectin IgG chimera. Comparative binding data are shown in Fig.
1.
2. Immunoprecipitation of PSGL-1 with P-selectin/IgG Chimera.
Recombinant full-length or soluble P-selectin ligand protein prepared from
transformants,
with and without additional core2, was labelled with 35S-methionine and
subsequently
immunoprecipitated with either the anti P-selectin ligand protein polyclonal
antibody or P-selectin
ligand/IgG chimera as described previously in Examples 7 and 5; Sako, D., Cell
175, 1179-1186
(1993). Data are depicted in Fig. 2.
3. Flow Cvtometrv.
Stable murine P-selectin ligand protein transfectants (with and without core2)
were
analyzed by standard FACS techniques using either P-selectin/IgG chimera (LecY
1) (Example
5) or anti P-selectin ligand protein monoclonal antibody (MAb 275, raised
against a peptide
having the sequence from amino acid 42 to amino acid 56 of SEQ ID NO:2). Both
reagents were
preconjugated to FITC labelled Protein A. Cells were analyzed by FACS after
incubating with
this conjugate for 30 minutes at 4 C in the presence of 2 mM CaCl,. Data are
depicted in Fig.
3.
EXAMPLE 9
E-SELECTIN BINDING OF P-SELECTIN BINDING PROTEIN
E-Selectin/IgG chimera was made as described in Example 5 for the P-selectin
IgG
chimera using an E-selectin encoding DNA including amino acids -21 to 536 of
the sequence
reported in Bevilacqua et al., Science, 243:1160 (1989).
42
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
U937 cells (approximately 6.5 x 107) were recovered from tissue culture plates
and
divided equally into two 50 mL cultures (final concentration of 1.3 x 106
cells/mL) containing
fresh complete RPMI medium and 50 Ci/ml of 'H-glucosamine hydrochloride
(labels the
protein-linked carbohydrate.of glycoproteins [Varki, FASEB 5:226-235 (1991)].
After 48 hours
incubation, the cells from both cultures were recovered by centrifugation and
washed three times
with PBS. The pelleted cells were suspended in 2.5 'mL each of a lysis buffer
containing 1%
Triton X-100 and disrupted by probe sonication for two minutes. The detergent
lysates were
placed on ice for three hours and then resonicated for an additional two
minutes. The lysates were
centrifuged at 16,000 rpm for five minutes, the supematants were recovered and
each adjusted
to 12 mL with lysis buffer containing no detergent. To one of the two diluted
cell lysates was
added 100 uL of protein A sepharose precoupled with P-selectin/IgG chimera
(see Example 5)
and to the other was added 100 uL of protein A sepharose precoupled with E-
selectin/IgG
chimera. Both chimeric proteins were present at a density of approximately 2
mg protein/mL of
resin. Binding reactions were allowed to proceed ovemight at 4 degrees C with
end-over-end
mixing. On occasion, purified membranes from U937 cells served as the starting
material for the
detergent extraction of labeled proteins. In these cases, the detergent
extraction and affinity
precipitation steps were essentially identical to the above.
Following incubation, the two parallel reaction mixtures were each centrifuged
at 2,000 rpm and supernatants were discarded. The resin pellets were washed
four times with
buffer (10 mM MOPS, 100 mM NaCl, 1 mM CaCIõ 1 mM MgCI:, 0.02% NaN3, pH 7.5
with
Triton X-100 [0.25% for the first and second washes, 0.1 % for the third wash
and 0.019o for the
fourth wash]). A final I mL pre-elution wash of each resin pellet using buffer
containing 0.01 qfo
Triton X- 100 was conducted and these were retained for quantitation of
radioactive counts by
liquid scintillation counting (LSC). The resins were then eluted overnight at
4 degrees C with
end-over-end mixing in I mL each of buffer containing 0.01 % Triton X- 100 and
10 mM EDTA.
The supernatants were recovered by centrifugation and then quantitated by LSC.
Autoradiography of the materials released from the resins by EDTA was
performed by electrophoresis of samples (approximately 10,000 cpm samples
concentrated by
Centricon-10 units where needed) on 10% cross-linked SDS-PAGE gels, subsequent
treatment
of the gels with EN3HANCE (Dupont) as per the manufacturer's instructions
followed by drying
for two hours on a commercially available gel dryer (Bio-Rad). Exposure of the
dried gels to X-
ray film was conducted for a minimum of three days at -80 degrees C.
Elution of immobilized E- or P-selectin, previously exposed to detergent
extracts
of U937 cells and exhaustively washed, with EDTA yielded liberated. 3H-
glucosamine labeled
43
CA 02599944 2007-09-14 -
WO 98/08949 PCTIUS97/14159
proteins. The amount of radiolabel recovered from the EDTA eluates was at
least 10-fold hisher
than the counts observed in the final, pre-EDTA washes. This observation
suggests that both P-
and E-selectin chimeras affinity captured ligand(s) from U937 whole cell
lysates in an EDTA-
dependent manner and that captured ligands were subsequently released upon
treatment of the
resins with EDTA.
The evaluation of the proteins released by EDTA from the two chimeras was
performed by SDS-PAGE and autoradiography under reducing and non-reducing
conditions
(commercially available14C-labeled molecular weight standards were employed).
As shown by
the autoradiograph depicted in Fig. 4, the released counts from the whole cell
lysates treated with
the P-selectin chimera (lanes 2 and 10) and the E-selectin chimera (lanes 4
and 8) correlated to
a major species of 200 kD molecular weight, non-reduced (lanes 2 and 4), and
100 kD reduced
(lanes 8 and 10). In different experiments depicted in Fig. 4, where purified
membrane extracts
were used as the starting material in place of whole cells, both the E-
selectin chimera (lane 3, non-
reduced and lane 9, reduced) and the P-selectin chimera (not shown) gave
similar results. Other
experiments have demonstrated that the major U937 glycoprotein which binds to
P-selectin is
immunoreactive with Rb3026, a polyclonal antibody raised against recombinant
sPSGL1.T7.
Therefore, P- and E-selectin specifically recognize a single major
glycoprotein species with
identical properties in each case.
Example 10
Production and Analysis of Deleted or Altered
Forms of Soluble P-Selectin Ligand Protein
A. Generation of DNA Constructs
Truncated forms of the P-selectin ligand protein-IgG chimeras were generated
as follows.
Plasmid pED.PSL.Fc was restricted with Pstl and NotI and the 6kb fragment
comprising the Fc
portion and vector, pEDFc6kb, was gel purified. Plasmid constructs pED.149.Fc,
pED.47.Fc and
pED.19.Fc were created by standard PCR technique, using the following pairs of
oligonucieotide
primers:
"Upstream" primer for all constructs:
5'-CCAGGTCCAACTGCAGGTCGACTCTAGAGGGCAC TTCTTCTGGGCCCACG-3' (SEQ ID NO:20)
"Downstream" primer for 148Fc:
5'-TATTATCTGTGCGGCCGCCCTCCAGAACCCATGGCTGCTGGTTGCAGTGG-3'(SEQID NO:21)
"Downstream" primer for 47Fc:
5'-TATTATCTGT; CGGCCGCGCAGCAGGCTCCACAGTGGTAG-3' (SEQ ID NO:22)
"Downstream" primer for 19Fc:
44 =
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
5'-TATTATCTGTGCGGCCGCGGAGGCTCCGTTTCTGGCAG-3' (SEQ ID NO:23)
The template DNA for PCR reaction was pED.PSL.Fc. The PCR conditions were
94oC, l min.;
42oC, 1 min.; 720C. 3 min.; 25 cycies, using a Perkin-Elmer Thermocycler.
After completion
of the last cycle, the reaction was treated with Klenow enzyme at 25oC for 30
min.. extracted with
phenol chioroform, sodium acetate added to 0.3M, and the PCR product DNA was
precipitated
with 2.5 volumes of ethanol. The DNA pellet was rinsed with 70% ethanol and
residual ethanol
was evaporated. The resuspended DNA was digested with Pstl and Not1, gel
purified and ligated
with the pEDFc6kb fragment described above. Correct constructs were identified
by restriction
analysis and confirmed by DNA sequencing.
Plasmid pED.AY148.Fc, pED.H24.Q70.148.Fc were created by site directed
mutagenesis
(Maniatis et al., - 1989, Molecular Cloning: A Laboratory Manual, Cold Spring
Harbor
Laboratories) using pED. ] 48Fc as template and the following mutagenesis
oligonucleotides:
for AY ] 48: 5'-CGGAGACAGGCCACCGAATTCCTGCCAGAAACG-3' (SEQ ID NO:24)
for H24: 5'-CCTCCAGAAATGCTGAGGCACAGCACTGACACCACTCCTC-3' (SEQ ID NO:25)
for Q70: 5'- GAGCTGGCCAACATGGGGCAACTGTCCACGGATTCAGCAG-3'(SEQ ID NO:26)
Positive clones were identified by colony hybridization (Maniatis et al,
supra).
pED.FFFE.148.Fc was constructed by restricting pED.AY148.Fc with EcoRI and
ligating
the following duplexed oligonucleotides:
5'-AATTCGAGTTCCTAGATTTTG-3' (SEQ ID NO:27) and
5'-AATTCAAAATCTAGGAACTCG-3' (SEQ ID NO:28).
Constructs of the series pED.FYYD.19.Fc, pED.FFYD. I 9.Fc and pED.FFFD.19.Fc,
were
made by restricting pED.DY148.Fc with EcoRI and NotI and ligating the
following duplexed
oligonucleotides:
for pED.FYYD.19.Fc:
5'-AATTCGAGTACCTAGATTATGATTTCCTGCCAGAAACTGAGCCTCCGC-3' (SEQ ID NO:29)
and
5'-GGCCGCGGAGGCTCAGTTTCTGGCAGGAAATCATAATCTAGGTACTCG-3' (SEQ ID NO:30);
for pED.FFYD.19.Fc:
5'-AATTCGAGTTCCTAGATTATGATTTCCTGCCAGAAACTGAGCCTCCGC-3' (SEQ ID NO:31)
and
5'-GGCCGCGGAGGCTCAGTTTCTGGCAGGAAATCATAATCTAGGAACTCG-3' (SEQ ID NO:32);
for pED.FFFD.19.Fc:
5'-AATTCGAGTTCCTAGATTTCGATTTCCTGCCAGAAACTGAGCCTCCGC-3' (SEQ ID NO:33)
and
=
CA 02599944 2007-09-14 -
WO 98/08949 PCT/US97/14159
5'-GGCCGCGGAGGCTCAGTTTCTGGCAGGAAATCGAAATCTAGGAACTCG-3' (SEQ ID NO:34).
B. Plate Binding Assav for Analvsis of Deleted or Altered Forms of Soluble P-
Selectin
LiQand Protein
The individual plasmid DNAs encoding the various mutated fotms of solubie PSGL-
1/Fc
chimeras were co-transfected with pEA.3/4FT and PACE cDNA in COS cells as
described in
Example 3(c). 50m1s of serum free medium, collected 40-64 hours post
transfection from
approximately 10' COS cells, was purified on a column of 0.25m1 of protein A
sepharose
(Pharmacia) equilibrated with TBS supplemented with 2mM CaCI,. After washing
with 20m1s
of TBS/CaCIõ the bound material was eluted with 0.5m1s of 0.1 M acetic acid,
0.15M NaCI,
2mMCaC1,. The eluted material was neutralized with 1/20th volume 3M Tris pH
9Ø The
material was quantitated by measuring absorbance at 280nm and by comassie blue
staining of
PAGF./SDS/Laemmli gels.
In order to produce non-sulfated forms of soluble PSGL-1, COS cell
transfections of the
relevant Fc chimeras were performed as described above except that following
transfection the
cells were cultured in the presence of 50mM Chlorate (Sigma).
Quantitative adhesion of CHO:P-selectin, CHO:E-selectin and CHO:L-selectin
expressing
cells was performed as described in Example 4(c), with the following
modifications: COS cell and
antibodies were omitted. Instead, 48-well microtiter plates (Costar) were
coated for 16 hours at
4oC with varying quantities of protein A-purified soluble PSGL-1/Fc chimeras.
The unbound
material was removed and the coated wells were treated with Hank's buffered
saline (HBS) with
1 mg/ml BSA and 2mM CaCI, for 1 hour at 4oC. Tritium labeled CHO selectin
expressing cells
were added and binding quantitated as described in Example 4(c).
C. Effects of Alteration of N-Linked Glvcosylation Sites
Constructs expressing three P-selectin ligand-IgG chimeras were constructed to
examine
the effects of N-linked glycosylation sites on selectin binding. These
constructs had the following
characteristics:
148.Fc amino acids 42-189 of SEQ ID NO:2
Q70.148.Fc amino acids 42-189 of SEQ ID NO:2, with the asparagine residue at
position i 11 of SEQ ID NO:2 replaced with a glutamine residue
H24.Q70.148.Fc amino acids 42-189 of SEQ ID NO:2, with the asparagine
residue at position 65 of SEQ ID NO:2 with a histidine residue
and the asparagine residue at position 1 I 1 of SEQ ID NO:2
replaced with glutamine residue
These constructs are schematically represented in Fig. 6.
46
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
The binding of these constructs to protein A and P-selectin-IgG chimera (LEC-y
1) was
compared. The results of these experiments are shown in Fig. 7. Comparison of
lanes 4, 5 and
6 in the autoradiograph demonstrates that removal of one or both of the first
two N-linked
glycosylation sites in soluble P-selectin ligand protein does not
significantly effect its binding to
P-selectin.
D. Effects of Tyrosines
Constructs were made to examine the role of tyrosine in P-selectin ligand
protein binding
to selectins by alteration of the anionic region of the soluble protein. The
following constructs
were made:
AY148.Fc amino acids 42-189 of SEQ ID NO:2, with amino acids 46-52
deleted
FFFE.148.Fc amino acids 42-189 of SEQ ID NO:2, with the tyrosine residues at
positions 46, 48 and 51 replaced with phenylalanine residues and the
aspartic acid residue at position 52 replaced with a glutamic acid residue
These constructs are schematically represented in Fig. 8.
The degree and sites of sulfation of P-selectin ligand protein were examined
by
expressing relevant constructs in the presence of radioactively labelled
sulfate. The degree of
sulfation of 148.Fc and DY.148.Fc were compared to that of a P-selectin-IgG
chimera, which was
not sulfated. Results are depicted in Fig. 9. These data demonstrate that the
majority of sulfate
incorporation is into the anionic region of the P-selectin ligand protein.
Additional constructs were made to determine whether the sulfation of the
anionic region
occurred at the tyrosine residues. The following additional constructs were
made:
FYYD.19.Fc amino acids 42-60 of SEQ ID NO:2, with the tyrosine residue
at position 46 of SEQ ID NO:2 replaced with a phenylalanine
residue
FFYD.19.Fc amino acids 42-60 of SEQ ID NO:2, with the tyrosine residues
at positions 46 and 48 of SEQ ID NO:2 replaced with a
phenylalanine residues
FFFD.19.Fc amino acids 42-60 of SEQ ID N0:2, with the tyrosine residues
at positions 46, 48 and 51 of SEQ ID NO:2 replaced with
- 35 phenylalanine residues
These constructs are schematically represented in Fig. 9.
The degree of sulfation of these constructs was compared to 19.Fc
("YYYD.19:Fc").
Results are shown in Fig. 10. FYYD.I9.Fc showed significant sulfation while
FFFD.19.Fc was
47
CA 02599944 2007-09-14
/
WO 98/08949 PGT/US97/14159
substantially less sulfated. Thus, the tyrosine residues of the anionic region
of P-seiectin ligand
proteins are the major site of sulfation.
Removal of sulfate from P-selectin ligand protein substantially reduces its
bindino to P-
selectin. The binding of 148.Fc treated with chlorate to P-selectin was
examined. As shown in
Fig. 11-, inhibition of sulfation by chlorate treatment substantially reduced
the amount of P-
selectin ligand protein binding to P-selectin:
E. Effects of C-terminal Deletions
Several additional C-terminal deleted constructs were made as follows:
254.Fc amino acids 42-295 of SEQ ID NO:2
47.Fc amino acids 42-88 of SEQ ID NO:2
19.Fc amino acids 42-60 of SEQ ID NO:2
These constructs are schematically represented in Fig. 12.
The binding of 254.Fc. 148.Fc, 47.Fc and 19.Fc to P-selectin, E-selectin and L-
selectin
was tested. Figs. 23 and 24 compare the binding of these deletion chimeras to
selectins and
controls. Results are also summarized in Fig. 12.
F. BindinQ to P-selectin and E-selectin Exaressine Cells
Binding of various constructs described above to cells expressing P-selectin
and E-
selectin was compared using a quantitative plate binding assay of Example 4(c)
(which is
schematically described in Fig. 13).
Fig. 14 compares the binding of 148.Fc, AY.148.Fc, FFFE.148.Fc and human I-G I
to
P-selectin expressing CHO cells. Deletion of all of the tyrosine residues in
the anionic region in
DY.148.Fc eliminated binding. Changing the tyrosine residues to phenylalanine
residues in
FFFE.148.Fc substantially reduced binding as compared to 148.Fc. Thus, it was
demonstrated
that the presence of the full length anionic region is essential to P-selectin
binding and that P-
selectin binding is enhanced by sulfation in this region. Fig. 14 also reports
control experiments
demonstrating that 148.Fc, DY.148.Fc and FFFE.148.Fc do not bind to CHO cells
which do not
express selectin.
Fig. 15 compares the binding of 148.Fc, AY.148.Fe, FFFE.148.Fc and human IgGI
to
E-selectin expressing CHO cells. E-selectin binding was unaffected by the
deletions or alterations
of the native sequence. Thus, it was demonstrated that the anionic region is
not required for E-
selectin binding.
Fig. 16 summarizes the results of Figs. 14 and 15.
48
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
Fig. 17 compares the binding of 47.Fc to P- and E-selectin expressing CHO
cells. 47.Fc
demonstrated substantial binding to both selectins despite deletion of the N-
linked glycosylation
sites at positions 1 l 1 and 292 of SEQ ID NO:2.
Fig. 19 compares the binding of FYYD.19.Fc. FFFD.19.Fc, H24.Q70.148.Fc,
148.Fc,
and human IgG I to P-selectin expressing CHO cells. Replacement of all of the
tyrosine residues
in the anionic region in FFFD.19.Fc eliminated binding. Changing the tyrosine
residue at position
46 to a phenylalanine residue in FYYD.19.Fc substantially reduced binding as
compared to
148.Fc. Alteration of the N-linked glycosylation sites in H24.Q70.148.Fc did
not affect binding.
Thus, it was demonstrated that P-selectin binding is enhanced by sulfation in
the anionic region
and that N-linked glycosylation is not required for P-selectin binding. Fig.
19 also reports control
experiments demonstrating that FYYD.l9.Fc, FFFD.19.Fc, H24.Q70.148.Fc and
148.Fc do not
bind to CHO cells which do not express selectin more than human IgGI alone.
Fig. 20 compares the binding of FYYD.19.Fc, FFFD.19.Fc, H24.Q70.148.Fc,
148.Fc,
and human IgG I to E-selectin expressing CHO cells. Truncation of the ligand
protein to the
degree of FYYD.19.Fc and FFFD.19.Fc substantially reduced E-selectin binding.
Alteration of
the N-linked glycosylation sites in H24.Q70.148.Fc did not significantly
affect E-selectin binding.
Thus, it was demonstrated that P-selectin ligand proteins comprising amino
acids 42 to 60 of SEQ
ID N0:2 can selectively bind P-selectin, and, to a substantially less, extent
E-selectin.
Fig. 21 summarizes the results of Figs. 19 and 20.
G. Conclusions ReQardinQ P- and E-Selectin Binding
Although applicants do not which to be bound by any theory, these data allow
several
conclusion regarding the relationship between P-selectin binding and E-
selectin binding by P-
selectin ligand proteins. N-linked carbohydrates are not required for binding
of a P-selectin ligand
protein to either P- or E-selectin. P-selectin ligand proteins as small
comprising as little as amino
acids 42-60 of SEQ ID N0:2 are capable of binding to P-selectin, and, to a
substantially less,
extent E-selectin.
Fig. 22 depicts a proposed schematic model for binding of P-selectin ligand
proteins to
P- and E-selectin. 0-linked sLe' carbohydrate has been demonstrated to be
required for both P-
and E-selectin binding. Data presented herein demonstrate that sulfated
tyrosine residues are
implicated in P-selectin binding, but not E-selectin binding. Applicants' data
also suggests that
no N-linked glycosylation binding site is required.
49
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
Example 11
Examination of AQQreEation Phenomena and Dimer Formation
in Forms of PSGL-1
A panel of PSGL-1 mutants were constructed by site-directed mutagenesis and/or
PCR
amplification with primers that introduced a stopcodon. The template for all
mutagenesis
experiments was pPL85.R16 (ATCC 75577, deposited by applicants).
The first group of mutants (C3lOS and C327S) encode full-length PSGL-1.R16
with only
one amino acid change compared to wild-type PSGL-1.R16 (Cys to Ser at position
310 or 327,
respectively). COS cells, co-transfected with pEA.3/4FT and the mutants C310S
or C327S, were
labeled with 35S-methionine. Cell lysates were prepared and the mutant
proteins were
immunoprecipitated with the P-selectin ligand polyclonal antibody of Example
7(A) and analyzed
by SDS-polvacrylamide gel electrophoresis under non-reducing and reducing
conditions.
The mutant C327S as well as wild-type PSGL-1.R16 migrated as a homodimer under
non-reducino conditions and as a monomer under reducing conditions. In
contrast, the mutant
C310S migrated as a monomer both under non-reducing and reducing conditions,
indicating that
the cysteine at position 310 is required for dimer formation of PSGL-1.
Both mutants were also analyzed for their ability to bind to P-selectin.
Detergent extracts
of co-transfected COS cells were precipitated with the LEC-y I chimera of
Example 4(A). The
precipitates were analyzed by SDS-PAGE under non-reducing and reducing
conditions and by
autoradiography. Both PSGL-1.R 16 and C327S were efficiently precipitated by
LEC-g 1, whereas
C310S binding to LEC-y I was greatly reduced, indicating that the dimeric form
of PSGL-1 binds
P-selectin more tightly than the monomeric form.
The second set of mutants encode soluble forms of PSGL-1.R16 and are listed in
Table
1. The mutant ATM was generated by site-directed mutagenesis and has a
deletion of the
transmembrane domain (amino acids 313-333) followed by RLSRKA. The mutants L31
1, L312,
A313, 1314. L315, A318 and T322 were generated by site-directed mutagenesis or
PCR
amplication with PCR primers that introduced a stop codon in the desired
position. The name of
the mutant refers to the C-terminal amino acid of each truncated soluble form
of PSGL-1.R16.
The mutants were analyzed according to the following criteria:
1. Expression and secretion from transfected COS cells
2. Dimer versus monomer formation
3. Lack of aggregate formation
4. P-selectin binding (LEC-y 1 chimera)
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
The mutants OTM and 1316 fulfilled all four criteria. The shorter soluble
forms of PSGL-1. such
as sPSL.QC of Example 5(A), L311, L312, A313, I314 and L315 did not form
dimers as well and
the longer soluble forms of PSGL- I, such as sPSL.T7 of Example 5(C), A318 and
T322 formed
high molecular weight aggregates which were less desireable.
CHO cells, already expressing 3/4 fucosyltransferase and Core2 transferase,
were
transfected with psPSL.T7, ATM, 1316 or psPSL.QC and amplified using
methotrexate. Stable
clones were isolated and labeled with 'SS-methionine. Conditioned media was
either analyzed
directly or first precipitated with LEC-y I and then analyzed by SDS-PAGE
under non-reducing
and reducing conditions (Figure 25). The results indicated that OTM and 1316
were most efficient
in dimer formation and P-selectin binding.
Table I
Mutant Dimer High MW P-selectin
Formation Aggregates binding
PSL.QC + - +
L311 + - +
L312 - - -
A313 - - - -
1314
- - -
L315 + - +
1316 ++ - ++
A318 ++ + ++
T322 ++ + ++
OTM ++ - ++
Example 12
Specificitv of PSGL-1 Bindine to P- and E-Selectins
Materials. A chimeric protein comprising the extracellular domain of human E-
selectin
and the Fc portion of human IgG, was constructed analogously to the P-selectin
chimera, LEC-y 1,
described earlier. The soluble E-selectin chimera was expressed in baculovirus-
infected
51 4
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
Trichoplusia ni high five cells (Invitrogen) and purified to homogeneity by
Protein A Sepharose
chromatography. Plasmid vectors pEA.3/4FT, pPL85, pFCD43, and pEA.sPACE, for
COS
expression of a(1,3/1,4)-fucosyltransferase (Fuc-TIiI), PSGL-1, CD43
(leukosialin), and soluble
paired basic amino acid converting enzyme (PACE), respectively, have been
described herein and
in the literature (Sako et al. (1993) Cell 75, 1179-1186; Rehemtulla, A. &
Kaufman, R. J. (1992)
Curr. Opin. Biotechnol. 3, 560-565; Wasley et al. (1993) J. Biol. Chem. 268,
8458-8465).
Fuc-TVII cDNA (plasmid pMT.FT7) was cloned from an HL60 cDNA expression libmry
using
oligonucleotide probes derived from.the published sequence (Natsuka et al.
(1994) Journal of
Biological Chemistrv 269, 16789-16794; Sasaki et al. (1994) J. Biol. Chem.
269, 14730-14737).
A polyclonal neutralizing rabbit-antibody, Rb3443, was raised against a
peptide comprising the
15,, first 15 amino acids of the mature ~PACE-cleaved) N-terminus of PSGL-1.
Monoclonal
anti-CD43 antibodies from either Becton Dickinson or Biodesign Intemational
and isotype control
antibodies were coupled to a solid support consisting of Sepharose TM-4B with
a covalently
attached goat affinity-purified antibody to mouse IaG (Cappel, Organon Teknika
Corporation).
Affinity coupling of selectin chimeras and murine antibodies to Protein A
Sepharose 4 Fast Flow
(Pharmacia) and to the anti-mouse IgG resin, respectively, was carried out at
a ratio of 2 mg
protein/ml of resin. Antiserum Rb3443 was coupled to Protein A Sepharose at 1'
mUml resin.
Coupling efficiencies, indicated by micro-BCA assay (Pierce) of the post-
reacted supernatants,
were at least 95%. Aprotinin and pepstatin were from Boehringer Mannheim and
benzamidine,
leupeptin, and phenylmethylsulfonyl fluoride (PMSF) were from Sigma.
Labeling and Membrane Extraction of Mveloid Cells. U937 or HL60 cells g~rown
in
suspension to a density of -1.3 x 10 cells/mi were labeled in 50 ml of
RPMI1640 medium
supplemented with 10% fetal bovine serum and 2.5 mCi of'H-glucosamineHCl
(Dupont/NEN)
for 48 hr. Activities of greater than I cpm/cell were routinely obtained by
this technique. The
labeled cells were washed with PBS, resuspended in cell lysis buffer (10 mM
MOPS, 150 mM
NaCI, 4 mM CaCI, and 4 mM MgCIZ, pH 7.5 containing protease inhibitors 20
mg/ml aprotinin,
10 mM benzamidine, 20 mg/ml leupeptin, 8 mg/ml pepstatin, and 10 mM PMSF) and
subjected
to several cycles of probe sonication on ice. Nuclei and cell debris were
removed by low speed
centrifugation and the cell membranes recovered from the supernatant by
centrifugation at
100,000g RCF for I hr, washed by resuspension and high speed centrifugation in
cell lysis buffer
containing I M NaCI, and finally resuspended in 3 ml membrane solubilization
buffer (cell lysis
buffer containing l% Triton X- 100). Several cycles of sonication and
incubation on ice were
employed to solubilize the membrane fraction. Finally, a low speed
centrifugation step was
employed to remove insoluble membrane residue.
52
CA 02599944 2007-09-14
I õ =
WO 98/08949 PCT/US97/14159
Labeline and Membrane Extraction of Transfected COS Cells. COS M6 cells were
transfected using DEAE-dextran and chloroquine (25) employing 8}tg of plasmids
pPL85 or
pFCD43 and 4 pg of pEA.sPACE, as well as 4 pg of pEA.3/4FT or pMT.FT7. After
40-45 hr
recovery the transfected cells were starved in serum-and tnethionine-free DME
medium for 30
min and then fed [35S]-methionine in serum-free DME for 5 hr. The labeled
cells were washed,
incubated with EGTA to loosen them from the dish surface, scraped from the
dish, pelleted, and
suspended in cold IOmM PIPES buffer, pH 7.5, containing 100 mM KCI, 3 mM NaCI,-
3.5 mM
MgCIõ and protease inhibitors (see, above). Membrane extraction then was
carried out by
sonication, low speed centrifugation, high speed centrifugation, and
solubilization in membrane
lysis buffer as above, for labeled= myeloid cells.
15, Affinity Precipitations. Membrane extracts were diluted 1:4 or 1:5 with
cell lysis buffer
or with TBSC buffer (20 mM TrisHCl, 150 mM NaCl, 2 mM CaC1Z, pH 7.5)
supplemented with
5 mg/ml bovine serum albumin (approximately 99%, Sigma). Extracts thus diluted
to 0.2-0.25%
TritorriX-100 were incubated with human IgG,-Protein A Sepharose with end-over-
end mixing
at 4 C overnight. The ptecleared supematants then were reacted for 6-12 hrs at
4 C with Protein
A Sepharose precoupled with E-or P-selectin chimeras, control human IgG,,
Rb3443 or with
rabbit pre-immune serum or with anti-CD43 antibody or isotype control
precoupled to goat
anti-mouse IgG Sepharose. The resins were washed 5 or more times in buffer
containing 0.1-0.5 Io
Triton X-100 until the radioactivity of the wash supernatants was reduced to
background level.
Elution of proteins bound specifically to P-or E-selectin resins was
accomplished with 10 mM
EDTA or 5 mM EDTA/5 mM EGTA at room temperature or by boiling in SDS-PAGE
sample
buffer (Laemmli, U. K. (1970) Narure 227, 680-685), whereas elution of
proteins bound to
antibody resins was achieved exclusively by the latter means. For resolution
under reducing
conditions. dithiothreitol was added to the sample buffer to a final
concentration of 1,00 mM.
Samples thus prepared were resolved by SDS-PAGE on 7.5% gels, treated with
En3Hance
(Dupont), dried, and exposed to autoradiography film.
For sequential affinity capture experiments, membrane extracts were
precleared, affinity
precipitated with P-or E-selectin or human IgG,, and washed as above. Samples
then were eluted
twice from the resins with 5 mM EDTA in 10 mM MOPS, 150 mM NaC1, pH 7.5 for 1
hr at 4 C
with tumbling. The first and second eluates were combined and then
immunoprecipitated with
immobilized Rb3443 according to the protocols outlined above.
RESULTS:
Soluble E-and P-selectin chimeras were used, in parallel with control human
I2G, to
probe detergent-solubilized membrane extracts of 3H-glucosamine-labeled U937
cells as described
53 ~
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
under "Methods". Examination of eluates from the immobilized selectins by
SDS-PAGE/autoradiography (Fig. 26) revealed the presence in both P-and E-
selectin eluates of
a major protein species with identical electrophoretic properties: Mr 200-kDa
non-reduced with
conversion to a species of Mr 120-kDa following reduction (Fig. 26, lanes 2
and 3, respectively).
Occasionally, additional bands were observed in both E-and P-selectin eluates
presumably
corresponding to this major band and reflecting the presence of naturally
reduced material (the
120-kDa species in non-reduced samples) and incomplete reduction (the 200-kDa
species in
reduced samples). Additionally, a trace band of Mr 150-kDa in the E-selectin
eluate which was
unaffected by reduction with DTT was occasionally observed. No bands were
observed in control
experiments using immobilized human IgG, (Fig. 26, lane 1) or where elution of
selectin resins
was performed in the absence of EDTA or SDS (data not shown). Essentially
identical results
were obtained using HL-60 cells (data not shown). Hence, the nature of these
recognition events
is interpreted to be specific metal-dependant interactions of these proteins
with the respective
selectins, presumably via the lectin domains (Lasky, L. A. (1992) Science 258,
964-969;
Drickamer, K. (1988) J. Biol. Chem. 263, 9557).
The metal-dependant recognition and electrophoretic behavior of the major band
precipitated with E-selectin was consistent with the properties of the
previously identified
P-selectin counten:eceptor, P-selectin glycoprotein ligand or PSGL-1 (Moore et
al. (1994) J. Biol.
Chem. 269, 23318-23327; Moore et al. (1992) J. Cell Biol. 118, 445-456; Sako
et al.).
To assess whether this species was indeed PSGL-1, EDTA eluates of the both E-
and
P-selectin precipitates were subsequently reacted with the PSGL-1 specific
polyclonal antiserum
Rb3443. As shown in Fig. 27, the major band isolated by affinity capture with
either selectin was
immunoprecipitated using this antiserum (lanes 3 and 4. respectively). No
species were detected
after immun'oprecipitation of the control IgG, EDTA eluate (Fig. 27, lane 5).
Direct
immunoprecipations using fresh 3H-labeled U937 membrane extracts confirmed the
specificity
of Rb3443: precipitation with Rb3443 results in the recovery of a single band
with the
electrophoretic properties of PSGL-1 (Mr 200-kDa non-reduced, Mr 120-kDa
reduced; Fig. 27,
lane 2) whereas precipitation with pre-immune antiserum fails to capture any
material (Fig. 27,
lane 1). These results indicate that the major protein species specifically
captured from myeioid
cells by both E-and P-selectins is PSGL-1.
To further assess the specificity of E-selectin for PSGL-1, U937 membrane
lysates were
probed directly for the presence of CD43 (or leukosialin), an abundant cell
surface
sialoglycoprotein known to bear the major portion of myeloid cell SLe"
residues (Maemura, K.
& Fukuda, M. (1992) J. Biol. Chem. 267, 24379-24386). Thus, membrane extracts
of
54 =
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
'H-glucosamine-labeled U937 cells were probed with an anti-CD43 antibody in
parallel with the
PSGL-1 specific antiserum Rb3443 and control antibodies as described under
"Methods". From
identical quantities of membrane lysate, the CD43 antibody precipitated in
excess of 30-fold
greater radioactive counts than did the PSGL-1 antiserum. Evaluation of the
immunoprecipitates
by SDS-PAGE/autoradiography (Fig. 28) revealed a single specific band for each
antibody.
Rb3443 captured a single species with the electrophoretic characteristics of
PSGL-1 (Fig. 28,
Lane 2). In contrast, the CD43 antibody precipitated a species with an Mr 120-
kDa which was
insensitive to reduction (Fig. 28, Lane 4), consistent with the absence of
cysteine in CD43.
Immunoprecipitations with control antibodies (Fig. 28, lanes I and 3) proved
negative as
expected. There appears to be considerably greater quantities of CD43 than
PSGL-1 in U937
cells, consistent with the quantitation of these proteins in HL-60 cells
(Ushiyama et al. (1993) J.
Biol. Chein. 268, 15229-15237). Thus, the inability of E-selectin to
precipitate CD43 from
myeloid cells does not appear to be due to its absence in these cell lines.
While we cannot exclude
the possibility that E-selectin captures trace quantities of CD43 (ie., the
low-intensity Mr
120-kDa band in Fig. 26, non-reduced lane 3 which is also consistent with
monomeric PSGL-1),
PSGL-1 appears to be the major protein precipitated from myeloid membrane
extracts.
Recombinant PSGL-1 expressed in COS cells is best achieved with cotransfection
of the
PSGL-l cDNA with a cDNA encoding an a(1,3/1,4)fucosyltransferase (Fuc-TIII)
for P-selectin
binding (Sako et al.). Interestingly, initial efforts to demonstrate E-
selectin recognition of
recombinant PSGL- I failed: E-selectin was unable to capture the
counterreceptor from
cotransfected COS cell membrane lysates under conditions where P-selectin
capture was
successful. One interpretation of this result is that Fuc-TIII was able to
modify recombinant
PSGL-1 for recognition by P-selectin but was unable to replicate the
appropriate modification(s)
found in myeloid PSGL-1 necessary for E-selectin recognition. The recent
cloning of a myeloid
fucosyltransferase, Fuc-TVII (Natsuka et al. (1994) Journal of Biological
Chemistrv 269, 16789-
16794; Sasaki et al. (1994) J. Biol. Chem. 269, 14730-14737), that is also
capable of generating
SLexcarbohydrate structures, allowed evaluation of this interpretation.
COS cells were cotransfected with cDNAs encoding either PSGL-1 or CD43 and
either
Fuc-TIII or Fuc-TVII. Membrane lysates were prepared from the transfected COS
cells and these
were precipitated with either immobilized E-or P-selectin chimeras or with
antibodies to either
PSGL-1 or to CD43. The precipitated products were evaluated by SDS-
PAGE/autoradiography
following their release by EDTA/EGTA (for selectin mediated binding) or by
boiling in SDS (for
immunoprecipitations). The results are shown in Fig. 29.
CA 02599944 2007-09-14 -
WO 98/08949 PCT/US97/14159
As observed in Fig. 29A, E-selectin capture of COS-expressed PSGL-i was
dependant
upon the nature of the fucosyltransferase used in the transfection. In three
separate experiments,
Fuc-TVII, but not Fuc-TIII, supported PSGL-l precipitation by E-selectin. The
inability of
Fuc-TIII to confer E-selectin reactivity to PSGL-1 cannot be attributed to a
lack of PSGL-1
expression as the specific antiserum Rb3443 immunoprecipated significant and
comparable
quantities of PSGL-l from both Fuc-TIII and Fuc-TVII transfections (Fig 29C).
Furthermore,
P-selectin was capable of precipitating equivalent quantities of PSGL-l with
either
fucosyltransferase (Fig. 29B), demonstrating that Fuc-TIII and Fuc-TVU are
expressed and active
in these cotransfections.
Within the COS recombinant expression system as in myeloid cells, high-
affinity
E-selectin recognition was also dependent upon the presence of an appropriate
polypeptide.
Although the polypeptide length, apparent molecular weight, and high frequency
and specific
types of posttranslational modifications are similar in CD43 and PSGL-1
(Maemura, K. &
Fukuda, M. (1992) J. Biol. Chem. 267, 24379-24386), neither fucosyltransferase
was able to
confer high-affinity E-selectin (or P-selectin) recognition to recombinant
leukosialin in
cotransfected COS cells (Fig 29A). Immunoprecipitations with the anti-CD43
antibody indicate
that comparable quantities of leukosialin were expressed in both Fuc-TIII and
Fuc-TVII
cotransfections (Fig. 29C). The failure of E-selectin to capture CD43 was not
due to lack of
fucosyltransferase activity within the cotransfected COS cells. FACS analysis
of COS cells
transfected with either PSGL-1 or CD43 and either Fuc-TIII or Fuc-TVII all
show high levels of
reactivity with the SLeX specific antibody CSLEX-1. Therefore, these results
suggest that
high-affinity E-selectin recognition requires the presence of a specific
polypeptide(s) that is
appropriately modified by a specific fucosyltransferase.
Example 13
Inhibition of P-Selectin/PSGL-1 Binding
by PSGL-1 Derived Peptides
A number of peptides derived from the sequence of PSGL-1 (SEQ ID NO:2) were
tested
for their ability to inhibit P-selectin/PSGL-1 binding. The tested peptides
are listed in Fig. 30.
Inhibition was tested according to the following protocol. The wells of a 96
well plate
were coated overnight at 4 C with PSGL-1 in 50 l of 10 mM MOPS, 150 mM NaCI,
I mM
CaCIõ l mM MgCl2 at pH 7.5. After removal of the liquid from the wells, 150 l
of 10 mM
MOPS, 150 mM NaCI, I mM CaCIõ 1 mM MgC1210.05% tween-20, 0.05% gelatin at pH
7.5 was
added per well to block the unoccupied sites. After 1/2 hr. to 2 hr. the block
buffer was removed
56
CA 02599944 2007-09-14
WO 98/08949 PCT/U897/14159
from the wells and 100 1 of a complex of Lec-y 1(P-selectin-human IgG Fc
chimera)(2 g/ml),
biotinylated goat anti-human antibody, and streptavidin-conjugated alkaline
phosphatase (which
had been allowed to tumble at room temperature for 30 minutes to 1 hour) plus
any potential
inhibitors were added per well. Each plate was shaken and rapped to remove the
block from plate.
The inc-ubation proceeded for I hr at room temperature rotating in the dark.
The unbound
complex was washed off the plate with two-I 50 i portions of 10 mM MOPS, 150
mM NaCI,
I mM CaCIõ 1 mM MgCl,, 0.05% tween 20 followed by 150 Nl of I M
diethanolamine, 0.5 mM
MgC1,. The chrom6genic substrate for alkaline phosphatase, PNPP, in 10 mM
DEA/0.5 mM
MgCI, was added and the plate is then read at 405 nm.
The results of these assays, are reported in Fig. 30. The peptides comprising
amino acids
48-51 (in which the tyrosine residues, have been phosphorylated) and amino
acids 42-56 of SEQ
ID NO:2 provided particularly desirable results.
Example 14
Purification of a Soluble Form of PSGL-1
Substantial purification of a soluble form of P-selectin ligand protein has
been achieved
according to the protoci described below.
A soluble P-selectin ligand protein, 1316 (amino acid 42 to amino acid 316 of
SEQ ID
NO:2) was expressed in CHO cells as described herein. CHO cell conditioned
media was
concentrated with a Pellicon ultrafiltration membrane unit (Millipore) with
either 10,000
molecular weight cutoff (MWCO) or 30,000 MWCO to about 10 times the original
concentration.
The buffer was then exchanged into 25 mM Tris, 1 mM CaCIõ pH 7.4.
The buffer-exchanged concentrate was loaded onto a Toyopearl QAE 550C
(TosoHaas)
column. Altematively, the buffer exchange step can be omitted and the
concentrate can be diluted
one part concentrate to three parts 25 mM Tris, I mM CaCIõ pH 7.4, and then
loaded onto the
column. The column was washed with 5-10 column volumes (CV) of 25 mM Tris, 1
mM CaC1,,
pH 7.4 at 4 C.
The P-selectin ligand protein eluted from the column with a linear NaCI
gradient (0 M
NaCI to 1.0 M NaCI) in the 25 mM Tris, I mM CaCIõ pH 7.4 buffer in
approximately five column
volumes. Two peaks were eluted from the column. The second peak contained the
P-selectin
ligand protein and was collected in bulk.
The peak from the QAE column was concentrated with a tangential flow
ultrafiltration
membrane (Millipore) with a 30,000 MWCO and was then buffer exchanged into 25
mM Tris,
150 mM NaCI, i mM CaCI,, pH 7.4 at 4 C.
57 ,
CA 02599944 2007-09-14
WO 98/08949 PCT/US97/14159
The buffer exchanged concentrate was loaded onto a Jacalin Agarose column
overnight
at 4 C. The column was washed with the diafiltration buffer and the P-selectin
ligand protein was
eluted with a gradient of methyl a-D-galactopyranoside (0-100 mM or )-50 mM
methyl a-D-
galactopyranoside) at 20 C. Fractions from the Jacalin column were analyzed by
SDS-PAGE and
the purest fractions were pooled.
Example 15
P-Selectin Ligand Protein Fusions
Four fusions of a P-selectin ligand protein with a different amino acid
sequence were
constructed: 47.Fc, 47.AGP, 47.BMP and 47.IL11.
47.Fc: A cDNA was constructed encoding the signal peptide, PACE cleavage site
and
first 47 amino acids of the mature P-selectin ligand sequence fused to a
mutated Fc region of
human IgGI at His224 of the native Fc sequence. The sequence of the cDNA
construct is
reported as SEQ ID NO:35. The fusion point is a a novel NotI site at
nucleotide 261. The amino
acid sequence encoded by the cDNA construct is reported as SEQ ID NO:36. The
mature amino
acid sequence of the encoded fusion protein begins at amino acid 42 of SEQ ID
NO:36. The
mutations in the Fc portion were a change of Leu 234 and Gly237 of the native
Fc sequence to
Ala.
47.AGP: A cDNA was constructed encoding the signai peptide, PACE cleavage site
and
first 47 amino acids of the mature P-selectin ligand sequence fused to the
first leucine residue of
mature human AGP. The sequence of the cDNA construct is reported as SEQ ID
NO:37. The
fusion point is a a novel Notl site at nucleotide 261. The amino acid sequence
encoded by the
cDNA construct is reported as SEQ ID NO:38. The mature amino acid sequence of
the encoded
fusion protein begins at amino acid 42 of SEQ ID NO:38.
47.BMP: A cDNA was constructed encoding the signal peptide, PACE cleavage site
and
first 47 amino acids of the mature P-selectin ligand sequence fused to the
sequence of mature
human BMP-2 (with its first 8 amino acids deleted). The sequence of the cDNA
construct is
reported as SEQ ID NO:39. The fusion point is a a novel Notl site at
nucleotide 261. The amino
acid sequence encoded by the cDNA construct is reported as SEQ ID NO:40. The
mature amino
acid sequence of the encoded fusion protein begins at amino acid 42 of SEQ ID
NO:40.
47.IL11: A cDNA was constructed encoding the signal peptide, P?_CE cleavage
site and
first 47 amino acids of the mature P-selectin ligand sequence fused to mature
human IL-11. The
sequence of the cDNA construct is reported as SEQ ID NO:41. The fusion point
is a a novel NotI
site at nucleotide 261. The amino acid sequence encoded by the cDNA construct
is reported as
58 .
CA 02599944 2007-09-14
WO 98/08949 PCTIUS97/14159
$~Q ID NO:42. ne mature amino acid sequence of the encoded fusion protein
bevins at amino
acid 42 of SEQ ID NO:42.
59 .
CA 02599944 2007-09-14
DEMANDES OU BREVETS VOLUMINEUX
LA PRESENTE PARTIE DE CETTE DEMANDE OU CE BREVETS
COMPREND PLUS D'UN TOME.
CECI EST LE TOME DE
NOTE: Pour les tomes additionels, veillez contacter le Bureau Canadien des
Brevets_
JUMBO APPLICATIONS / PATENTS
THIS SECTION OF THE APPLICATION / PATENT CONTAINS MORE
THAN ONE VOLUME.
THIS IS VOLUME OF _2
NOTE: For additional volumes please contact the Canadian Patent Office.
