CA2191870A1 - Cross-reacting monoclonal antibodies specific for e-selectin and p-selectin - Google Patents

Abstract

The invention provides monoclonal antibodies that specifically bind to P-selectin and to E-selectin. Humanized monoclonal antibodies also are disclosed. Many of the antibodies block the functional interactions of P-selectin and E-selectin with their respective counterreceptors.

Description

~ W095/~324 2i91~7i) rc~

CROSS-REACTING MONOCLONAL ANTIBODIES
SPECIFIC FOR E-SELECTIN AND P-SELECTIN

RA(~- ~uNI~ OF THE INVENTION
The ability of cells to adhere to one another plays a critical role in development, normal physiology, and disease processes such as infl~ -tion. This ability is mediated by adhesion molecules, generally glycoproteins, expressed on cell membranes. Often, an adhesion molecule on one cell type will bind to another A~h~sion molecule expressed on a different cell type, forming a receptor counter-receptor pair. Three important classes of adhesion molecules are the integrins, selectins, and immunoglobulin (Ig) superfamily members (see Springer, Nature 346:425 (1990); Osborn, Cell 62:3 (1990);
Hynes, Cell 69:11 (1992). These molecules are vital to the interaction of leukocytes and platelets with themselves and with the extracellular matrix and vascular endothelium.
The selectin family of receptors are so named because of their lectin-like domain and the selective nature of their adhesive functions. There are three known selectins, L-selectin (also known as LECAN-l, Nel-14 or LAM-l or CD62L), E-selectin (also called ELAN-l or CD62E) and P-selectin (also known as CD62, CD62P, GNP140 or PADGEM). The selectins are highly homologous, containing a 120 amino acid (aa) N-terminal lectin domain, an EGF-like domain, a variable number of multiple short c--4~n~u~ repeat (SCR) domains homologous to those found in complement regulatory proteins, followed by a tr~n-- 'rane domain and short cytoplasmic tail. See Siegelman et al., Science 243:1165-1172 (1989); Lasky et al., Cell 56:1045-1055 (1989); Tedder et al., J. Exp. ~ed. 170:123-133 (1989); Johnson et al., Cell 56:1033-1044 (1989);
Bevilacqua et al., Proc. Natl. Acad. sci. USA 84:9238-9242 (1987), Bevilacqua et al., Science 243:1160-1165 (1989), Bevilacqua et al., J. Clin. Invest. 91:379-387 (1993), Camerini et al., Nature 280:496-498 (1989). The selectins have overlapping but distinct specificities for counterreceptors. See Bevilacqua et al., J. Clin. Invest.
91:3790-387 (1993); Feize, Current Opinion in Struct. Biol.
3:701-710 (1993); Berg et al., Biochem. Biophys. Res. Comm.
184:1048-1055 (1992); Foxall et al., J. Cell Biol. 117:895-902 (1992); Larsen et al., J. Biol. Chem. 267:11104-11110 (1992);
Polley et al., Proc. Natl. Acad. Sci. USA 88:6224-6228 (1991) (each of which is incorporated by reference in its entirety for all purposes).
P-selection is constitutively expressed by both platelets and endothelial cells where it is stored in .alpha.-granules or Weibel-Palade bodies for rapid (second to minutes) translocation to the cell surface upon activation by, for example, thrombin or histamine (McEver et al., J. Biol.
Chem. 250:9799-9804 (1984)). E-selectin is expressed by activated endothelial cells (e.g., after TNF-.alpha. or IL-1 stimulation for 6-8 hr). Its expression is controlled at the transcroptional level (Bevilacqua et al., 1987, supra; Belilacqua et al., 1989, supra). P-selectin and E-selectin both bind to neutrophils and monocytes (Larsen et al., Cell 59:305-312 (1989); Johnston et al., Cell 56:1033-1044 (1989); Bevilacqua et al., 1987, supra; Belivacqua et al., 1989, supra), as well as subsets of lymphocytes (Picker et al., Nature 349:796-799 (1991); Shimizu et al., Nature 349:799-802 (1991); Moore et al., BBRC 186:173-181 (1992)). L-selectin is constitutively expressed by leukocytes, and mediates lymphocyte adhesion to peripheral lymph node high endothelial venules (HEV) (Gallatin et al., Nature 304:30-34 (1983); Berg et al., Immunol. Rev.
108:5-18 (1989); Berg et al., J. Cell. Biol. 114:343-349 (1991)), and neutrophil adhesion to cytokine-activated endothelial cells (Hallman et al., Biochem. Biophys. Res.
Comm. 174:236-243 (1991); Smith et al., J. Clin. Invest.
87:609-618 (1991); Spertini et al., J. Immunol. 147:2565-2573 (1991)). L-selectin is a counter-receptor on neutrophils for both E-selectin and P-selectin (Kishimoto et al., Blood 78:805-811 (1990), Picker et al., Cell 66:921 (1991)), ~ wo g5/3432~ 2 i 9 1 ~ 7 ~ PCT~S95107302 although all three selectins probably have other counter-receptors as well.
E-selectin, P-selectin and L-selectin mediate leukocyte-endothelial cell and platelet-leukocyte adhesive interactions during inflammation (Bevilacqua et al., 1993, supra). All three selectins have been demonstrated to participate in an initial "rolling" interaction of leukocytes with activated endothelium (von Andrian et al., Proc. Natl.
Acad. Sci. USA 88:7538-7542 (1991); Ley et al., Blood 77:2553-2555 (1991); Abassi et al., J. Clin. Invest. 92:2719-2730 (1993); Dore et al., Blood 82:1308-1316 (1593); Jones et al., Piophys. J. 65:1560-1569 (1993); Mayadas et al., Cell 74:541-554 (1993)). This initial interaction precedes CD18-integrin-mediated adhesion and snhsD~lu~l~L migration of neuL~ùphils through the endothelium and into inflamed tissue sites (Lawrence et al., Cell 65:859-873 (1991); von Andrian et al., Am. J. Physiol. 263:H1034-H1044 (1992)). D~p~n~ing on the nature of inflammatory stimuli and time after initiation of infl; ~ory l~ul.se, either E-selectin or P-selectin may be functionally dominant in promoting neuLLu~hil-mediated tissue damage.
In principle, antiho~iec or other antagonists of the selectins could abort the adhesion process, thereby preventing neuL,u~hils from binding to the endothelium and from extravasating into tissues. A substantial number of antiho~ipe specific for one of the selectins have been reported. Some of these antiho~;Ds have been reported to block binding of selectins to count~ e~ors in vitro. Some of the antibodies have also been reported to block selectin-mediated interactions in animal models in vivo. For example,an~ihodi~c to E-selectin have been reported to protect against n~uLLùphil-mediated damage in an IgG complex model of lung injury in the rat (Nulligan et al., J. Clin. Invest. 88:1396 (1991)). Antibodies to P-selectin have been reported to protect against acute lung injury induced by i~lL-avenuua injection of cobra venom factor (Nulligan et al., J. Clin.
Invest. 90:1600-1607 (1992)), as well as in a rat model of systemic endotoxemia (~oughl ~n et al., J. Exp. Med. 179:329-WogS/3432~ 2191~ 7 ~ PCT~ISgSI07302 334 (1994)). Antiho~ to P-selectin have also been reported to be protective in a cat model of myocardial i~h~ and LepeLLusion injury tWeyrich et al., FASEB ~. 7:A785 (1gg3)).
Although some antibodies against E-selectin and p-5~lec~n have ~hown blocking activity, many, if not most, antiho~ies specific for E-selectin or P-selectin are nonblockin~ (see, e.g., Bevilacqua et al., 198g, supra; Erbe et al., J. Cell Biol. 119:215-227 (1992)). That is, these antibodies bind ts epitopes in the extracellular domains of E-s~lect;n or P-selectin that do not directly participate in counterreceptor binding or the sllhse~nt cellular adhesion process. The prevalence of nonblocking antibodies suggests that only small regions of the extracellular domain participate directly in binding or influence binding. Thus, de novo screening of ~tiho~ipc generated against E-selectin or P-selectin would be expected to generate mainly nonblocking antihO,~
Despite the large number of antibodies isolated to-date against the three selectins, there have been few reports of crossreacting antiho~ie~ that bind to more than one selectin. Crossreacting anti ho~ i ~c might be capable of aborting the inflammatory process at more than one level, thereby providing more broadly useful therapeutic agents for neu~L~hil-mediated inflammatory conditions than antiho~
specific for a single selectin. One antibody has been reported to ~Lu~Laact with human E-selectin and dog L-selectin but not with the two selectins from the same species (Ahassi et al., ~. Imm~nol. 147:2107-2115 (lg91)~. A
second antibody has been reported to crossreact ~ith human E-selectin and L-selectins (Jutila et al., J. Exp. Med.
175:1565-1573 (1992~; WO/9324614~. However, no antibody has been isolated that binds to both P-selectin and E-selectin, much less blocks the functions of both of these molecules.
Accordingly, there is a need for ant;ho~ that bind to both E-selectin and P-selectin, preferably so as to block the capacity of both of these molecules to participate in adhesion reactions with counterreceptors. The present invention fulfills this and other needs.

~ W095l~32~ ~1 9 1 ~ 7 ~

SUMMARY OF THE INVENTION
The invention provides monoclonal ant; ho~; Dc that have a binding eite that specifically binds to P-selectin and to E-selectin. For many such ant;ho~lDc, specific binding of the antibody to the P-selectin inhibits binding of the P-selectin to a countt~L~ or of P-selectin, and specific binding of the antibody to E-selectin inhibits binding of the E-selectin to a counte~.aceptor of E-selectin.
Countel,~c~ors of E-selectin and P-selectin are expressed on the surface of cells such as HL-60 cells and n~uLL~hils.
r ,l~ry an~iho~;~F are designated 57C.29, 2C9.11 and lD8.10.
Many of the antibodies of the invention compete with an DY li fied antibody for specific binding to P-selectin and to E-selectin. Some antibodies of the invention also specifically bind to L-selectin, whereas others do not. In one ~ the antibody rDcogni 7~q an epitope of E-selectin comprising amino acids Q21, R22, Y23, Tllg, and Al20 In another ~ , the ant;ho~;-oc bind to the same epitope of E-selectin and/or P-selectin as antibody 5C7.29. In addition to intact ant;ho~ies~ the invention also provides binding f.~ ~s such as Fab, Fab', F(ab')2, Fv or single-chain antibodies.
Some of the antibodies of the invention are non-human, e.g., mouse, whereas others are humanized or human ant;ho~ios. A hllr-n;7Dd antibody comprises a h~lr~niz-Dd heavy chain variable region and a hnr-ni7sd light chain variable region. The hl~r-nized light chain variable region can comprise c, l~ tarity detor~;ning regions te.g., CDRl, CDR2, CDR3) having amino acid se~lonrDc from the light chain of a mouse, antibody selected from the group consisting of 5C7.29, 2C9.11 and lD8.10, and having a variable region framework seguence substantially identical to a human light chain variable region r. ~Lk sDqUPnre. The hllr-n;70~ heavy chain variable region can comprise complementarity detDrmining ~ 35 regions (e.g., CDRl, CDR2 and CDR3) having amino acid SD~IDnrDC from the COLL~ 0n~1;ng mouse antibody heavy chain, and having a variable region framework sequence substantially identical to a human heavy chain variable region r, ..~Lk ~O95/34324 2 1 9 1 ~ 7 ~) PCT~S9~/0~30~ ~

se~l~nce. The antibodies o~tionally contain constant regions substantially identical.to hu~an ~un~LallL regions.
In particular ~ -~in-~Ls of the humanized ant;ho~ s of this invention, the h~qn;7pd light chain variable region has a sP~lon~e substantially identical to the mature sequence depicted in Figure 8A [SEQ ID NO:5] and the hllr-ni 7P~ heavy chain variable region has a se~l~n~e subs~nti~lly idDntjr~l to the mature se~ .n~-C depicted in Figure 8B [S~Q ID NO:8]. More particularly, this invention provides h~ ni70A ant;ho~lPc wherein (a~ the hllr-ni~ed light chain variable region has the se~uence: X1IX2X3TQSPSS
LSASVGDRVT ITCSASSSXllP YX12~wYQQ~ KAPKLLIYDT
SNX13X14XlsGVPX,IR X7fi~CGTXsx6 TX8TISSLQPE DX9ATYYCXl~;Xl7W
S~rlr~X1OG TRVEIK ~SEQ ID NO:9], wherein Xl = ~ or Q; X2 = Q
or V; X3 = M or ~; X4 = S or A; X5 = S or D: X6 = Y or F; X7 =
F or I: X8 = L or F; X9 = F, I or A; XlO = Q, G or S; Xll = V, I or L; Xl2 = M or L; X13 = any amino acid; Xl4 = any amino acid; Xls = S or T; Xl~ = Q, N or H; and Xl7 = Q, N or H; and (b) the hnr-nl7ed heavy chain variable region has the sequence: X3VQLVESGGG LVy~ LRL SCAASGFTFS SFGX7HWVRQA
PGRGLEWVXlF ISSGS2~ '1YY X8XgXloXllX12X13RFTI SRDNX4RNX5 L2MX25LRAED TAVYYCARPL PPFAYWGQGT LVTVSX6 [SEQ ID NO:10];
wherein, Xl = A or S; X2 = N or T; X3 = E, Q or D; X4 = S, A
or P: Xs = T or S; X6 = A or S; X7 = M, I, V or L; X8 = any amino acid; Xg = any amino acid; XlO = any amino acid; Xll = V, A, I, L, M or F; X12 = R, K or Q; and Xl3 = Gl A, D, T or S.
In certain e ~ Ls of the afo~ Lioned an~ihs~;ps~ the CDR regions of the light and heavy chain variable regions have the same amino acid seq~l~nre as the CDR se~Pn~Ps of Figure 8A
and 8B. That is, in the human light chain variable region, X1l = V; Xl2 - M; X13 = L; Xl4 = A; Xl5 = S; Xl6 = Q; and X17 =
Q; and in the heavy chain variable region, X7 = ~; X8 = A; X9 = D; X10 = T; Xl~ = V; Xl2 = R; and X13 = G. In another ,~ho~i ~, the variable light and heavy chain regions have the amino acid se~-Pnce depicted in Figures 8A and 8B.
In another aspect, the invention provides purified nucleic acid s~ -~ Pn~o~ling a light or heavy chain variable region of one of the monoclonal ~nt i ho~ i es ~; ~Cncspd above.

~ WO95J34324 21918 7 0 PCT~S95107302 The invention also provides stable cell lines capable of producing the anti ho~; e~ described above. The stable cell lines comprise nucleic acid s-, ~s respectively rnro~;ng the heavy chain and light chain of an antibody described above. The s-, L~ are operably linked to first and second promoters to allow expression of the heavy and light chains.
The invention further provides pharmaceutical compositions comprising the ~n~;ho~;P~ described above and methods of treatment using the same. The methods of treatment are particularly effective for ;nfl2 tory ~ 2~S including conditions such as ;~rh~ reperfusion injury, adult respiratory distress ~y"dL ~, sepsis, psoriasis and autoimmune disease.
In another aspect, the invention provides methods of generating an antibody capable of blocking E-selectin and/or P-selectin mediated functions. The method comprises ~u--~uLL~..tly or cu..se~u~ively immunizing a mammal with P-selectin and E-selectin. B-cells from the mammal are immortalized to generate immortalized cells producing ant;h~;r~. An immortalized cell is selected producing an antibody that specifically binds to E-selectin and to P-selectin.
The invention further provides methods of detecting E-selectin and P-selectin bearing cells in a biological sample suspected of containing the cells. The method comprises contacting the sample with an antibody as described above to form an immune complex with the E-selectin and/or P-selectin bearing cells. The presence of the immune complex is then detected to indicate the presence of the cells.

BRIEF DESCRIPTION OF THE DRAWINGS
Figures lA and lB: Crossreacting antibody 5C7.29 binds to naturally occurring human E-selectin. ~a) Binding of known anti-E-selectin antibody Hl8/7 to activated ~black histograms) and resting (grey histograms) HUVEC cells. (b) Binding of crossreacting antibody 5C7.29 to activated and wo ~3~32~ 2 1 9 1 ~ ~ O

recting HU~EC cellsf PAcs fluorescence intensity is indic~ted by the X axis.
Figures 2A and 2B: Crossreacting antibody 5C7.29 binds to naturally occurring P-so1er~iQ~ ~a~ Binding of known anti-P-selectin antibody WAPS 12.2 to platelets detected by staining with secr~ry antibody (black histogram), compared with staining with s~cnn~ry antibody alone ~control, grey histogram~. (b) Binding of 5C7.29 to platelets, shown similarly.
Figure 3: Crossreactivity of 5C7.29 resides in a single monoalonal antibody. 5C7.29 antibody was incubated with excess of (a, c) parent Ll-2 cells or (b, d) Ll-2P-8electin transfectants, and resulting a~e~llatants tested for reactivity with fresh samples of Ll-2P-~ele~t1~ (a, b) or L1-2E-selectin cells (c, d) by FACS analysis. This figure shows that Ll-2P-5electin depletes reactivity for E-selectin.
Figure 4: Monoclonal antibody 5C7.29 blocks binding of HL-60 (neuLLu~hil-like) cells to TNF-a-activated HUVEC
cells (expressing E-selectin). Average of four experiments.
Figure 5. Monoclonal antibody 5C7.29 blocks binding of ~L-60 cells to E-selectin tr~ncfpctAnt cells. Average of four experiments.
Figure 6. Monoclonal antiho~lpc 5C7.29, 2C9.11 and lD8.10 block binding of platelets to HL-60 cells as shown by platelet rosetting. The chart shows the percentage of HL-60 cells with > 2 platelets bound (rosetted). Average of three experiments.
Figures 7A-7B. S~lpnr~c of the cDNA (light chain -- SEQ ID N0:1; heavy chain -- SEQ ID N0:3) and translated amino acid sequences (light chain -- SEQ ID N0:2; heavy chain -- SEQ ID N0:4) of the light chain (A) and heavy chain (B) variable regions of the mouse 5C7.25 antibody. The first amino acid of each mature chain is indicated by a double underline. The three CDRs in each chain are underlined.
Figures 8A-8B. S~ crc of the synthetic DNA
(light chain -- SEQ ID N0:5: heavy chain -- SEQ ID N0:7) and translated a~ino acid sequences (light chain -- SEQ ID N0:6;
heavy chain -- SEQ ID N0:8~ of the light chain (A) and heavy ~ W0 9'~;134324 2 1 9 1 8 7 ~ P~ u.. ,~ u2 chain (B) variable regions of the humanized 5C7.29 antibody.
The first amino acid of each mature chain is indicated by a double underline. The three CDRs in each chain are underlined.
Figure 9. Schematic diagram of uu~ L~uLion of hn~qni~od 5C7.29 antibody variable region genes.
Figure 10. Hnrqni7s~ 5C7.29 antibody reactivity with E-selectin, P-selectin and L-selectin transfectants.
Ll-2 transfectant cell lines expressing the indicated selectin were analyzed for reactivity with hnr-nized 5C7.29 by flow cytometry.
Figures llA and llB. Competitive binding of mouse and htl~-n; 7~d 5C7.29 antibodies to cells expressing E-selectin tA) or P-selectin (B). Increasing concentrations of cold competitor antibody were incubated with the cells in the sence of ra~iolAh~lod tracer mouse 5C7.29 antibody, and the ratio of bound/free radioactivity was det~qrm;n~qd.
Figure 12. Inhibition of HL-60 cell adhesion to CHOE-8eleCtin cells by mouse and hnrqn;7ed 5C7.29 antibodies.
Fluorescently labelled HL-60 cells were incubated with CHOE-8eleCtin cells in the presence of the ant;hoAies at the indicated ~nnc~n~rations. After washing, adherent cells were counted micrnccopicAlly. The results from a representative experiment performed with each sample in quadruplicate t+/-standard deviation) are shown.
Figure 13. Inhibition of platelet rosetting toHL-60 cells by mouse and humanized 5C7.29 an~iho~;~C. Normal human platelets were incubated with HL-60 cells in the presence of the ant;ho~;~C at the indicated uu..ce.lLL~tions.
After fixation, the percent of HL-60 cells with greater than 2 platelets bound (rosetted) was det~rmin~. The results shown are from a representative experiment performed with each sample in triplicate (+~-standard deviation).

~lNlllONS
The term "substantial identity" or "substantial homology" means that two peptide sequences, when optimally aligned, such as by the programs GAP or BESTFIT using default _ _ _ _ _ _ _ _ _ _ _ _ _ ~09~3~324 2~9~7Q PCll~S95~07302 1'0 -:
gap weights, share at least 80 percent s~qu~nce identity, preferably 90 percent sD~IDnre identity, more preferably at least 95 percent se~u~..ce identlty or more (e.g., 99 percent sequence identity~. Preferably, residue positions which are not identical differ by conservative amino acid substitutions.
The term "substantially pure" or "isolated" means an ob~ect species is the pl~' ;n~nt species present ~i.e., on a molar basis it is more abundant than any other individual species in the composition), and preferably a subst~n~i~lly purified fraction is a composition wherein the object species comprises at least about 50 percent (on a molar basis) of all macromolecular species present. Generally, a substantially pure composition will comprise more than about 80 to 90 percent by weight of all macroDolecular species present in the composition. Most preferably, the object species is purified to essential hl , -ity ~contaminant species cannot be det~ct~ in the oomposition by conventional detection methods) wherein the composition consists eqeDntiAlly of a single macromolecular species.
I~T - l9blllin,~ antibody~' or "antibody peptide(s)" refers to an intact antibody, or a binding rL ~ thereof that competes with the intact antibody for specific binding. Binding fL~_ - Ls are ~L~du~d by ~ -n~t DNA terhni~l~s~ or by enzymatic or chemical cleavage of intact immunoglobulins. Binding fL~y~l~t~ include Fab, Fab', F~ab'~2, Fv and single-chain an~;ho~ =. An antibody other than a "bispecific" or "bifunctional" antibody is understcod to have each of its binding sites i~Dntic~l.
An antibody substantially inhibits ~h~cion of a receptor to a counterreceptor when an excess of antibody reduces the quantity of receptor bound to count~LL.c~or by at least about 20%, 40%, 60% or 80%, and more usually greater than about 85~ (as measured in an im ~itro competitive binding assay).
The term epitope includes any protein determinant capable of specific binding to an immunoglobulin or T-cell receptor. Epitopic determinants usually consist of chemically active surface groupings of molecules such as aminc acids or ~ wos5~4324 2 1 91~ ~ ~ pcT~ssslo73o2 sugar side chains and usually have specific three dimensional aLLu~LuLcl characteristics, as well as specific charge characteristics.
An antibody is said to specifically bind an antigen when the ~;~soci~tion ~ul.~Ldnt is s 1 ~M, preferably 5 100 nM
and most preferably 5 10 nM.
The term patient includes human and veterinary subjects.
The term P-selectin count~ c~tor denotes a protein other than an antibody that specifically binds to P-selectin at least in part by noncovalent bonds. Specific binding maintains cells respectively bearing receptor and counterreceptor in physical proximity and may also transduce a change in physical or functional phenoLy~e in either of the cells or both. Other selectin counterreceptors are analogously defined.

DESCRIPTION OF THE ~K~rrKh~ EMBODIMENT
I. Antibodies of the Invention The invention provides ant; ho~; ec that crossreact, i.e., specifically bind, with E-selectin and P-selectin.
Preferred an~;ho~;~C block the functions of both of these molecules.

A. General Characteristics of Antibodies The basic antibody structural unit is known to comprise a tetramer. Each tetramer is -se~ of two identical pairs of polypeptide chains, each pair having one "light" (about 25 kDa~ and one "heavy" chain (about 50-70 kDa). The amino-terminal portion of each chain includes a variable region of about 100 to 110 or more amino acids primarily responsible for antigen recognition. The carboxy-t~r~;n~l portion of each chain defines a constant region prinarily r~pon~;hle for effector function.
Light chains are classified as either kappa or lambda. Heavy chains are classified as gamma, mu, alpha, delta, or epsilon, and define the antibody's isotype as IgG, IgM, IgA, IgD and IgE, respectively. Within light and heavy w09sl~32~ 2 1~ 18 ~ ~ r. ~ ., /J~
12 ~
: ~' ' t ' ' chains, the variable and constant reqlons are joined by a "J"
region of about 12 or more amino acids, with the heavy chain also including a "D" region of about 10 more amino acids.
(See generally, FLndamental T - 707y (Paul, W., ed., 2nd ed.
Raven Press, N.Y., 1989), Ch. 7 (incvL~oLated by reference ln its entirety for all ~uL~oses).
The variable regions of each light/heavy chain pair for~ the antibody binding site. Thus, an intact antibody has two binding sites. Except in bifunctional or bi~eci f iC
antibodies, the twc binding sltes are the same. The chains all exhibit the same general structure of relatively conserved fL ~Lk regions (FR) joined by three hypervariable regions, also called complementarity deter~ining regions or CDRs. ~he CDRs from the two chains of each pair are aligned by the framework regions, enabling binding to a specific epitope.
From N-t~rmi n~ 1 to C-ter~inal, both light and heavy chains comprise the domains FRl, CDRl, FR2, CDR2, FR3, CDR3 and FR4.
The ACsi, ~L of amino acids to each domain is in accordance with the definitions of Kabat, Sequences of Proteins of T~ - 7~gical Interest (National Institutes of Health, Bethesda, MD, 1987 and 1991~/ or Chothia & Lesk, J. Mol. ~3iol.
196:901-917 ~1987): Chothia et al., Nature 342:878-883 (1g89).
A bispecific or bifunctional antibody i8 an artifici21 hybrid antibody having two different heavy/light chain pairs and two different binding sites. Bispecific antibodies can be pLvduced by a variety of methods including fusion of hybridomas or linking of Fab' rL ~ . See, e.g., Songsivilai & T~h- -nn, Cli~. Exp. I~munol. 79:315-321 ~1990);
Kostelny et al., J. Immunol. 148, 1547-1553 (1992).
Production of bispecific antibodies can be a relatively labor intensive process compared with production of conventional ant~ho~iPe and yields and degree of purity are generally lower for bispecific antibodies. Bispecific antibodies do not exist in the form of fL, ~5 having a single binding site (e.g., Fab, Fab' and Fv).

2191~7~
ss/34324 13 ~ 2 s. Bindinq S~ecific itY and Affinity The immunoglobulins (or antibodies) of the invention exhibit specific binding to both P-selectin and E-selectin.
That is, a single binding site on an antibody has affinity for both P-selectin and E-selectin. Thus, the antibodies bind to epitopes that are common to both molecules. The antibodies bind to the natural and/or recombinant human forms for P-selectin and E-selectin (see Johnston et al., 1989, supra;
Bevilacqua et al., 1989, supra) . Some antiho~i~C may alsc bind P-selectin and/or E-selectin from nA~ n species. Some of the ~ntiho~i~R also specifically bind to L-selectin (preferably human L-selectin (see Tedder, EPA 386,906 (1990)) whereas other antiho~i~c of the invention do not.
Surprisingly, the common epitopes bound by the crossreacting antiho~i~C of the invention are also epitopes important for both E-selectin and P-selectin to interact with their countelL._epLors on activated leukocytes, such as neutrophils.
Thus, most ~LUS~' eacting antibodies of the invention block the functional interactions of E-selectin or P-selectin and usually those of both of these molecules. Some crossreacting antibodies also block the functional interactions of L-selectin whereas others do not.
Blockage of P-selectin-mediated functions can be d LL~Led in vitro. In vitro assays measure the capacity of an antibody to inhibit binding of P-selectin to a counteILece~tor. Suitable sources of P-selectin for such assays are purified P-selectin (or an extracellular domain thereof), cells transfected with P-selectin, activated endothelial cells or platelets. Suitable sources of count~LLece~Lor are leukocytes, neuL~u~llils, monocytes, or HL-60 cells (ATCC CCL 240) and appropriate cell lines transfected with L-selectin. Ile LLu~hils can be isolated from whole blood (preferably human blood) by Ficoll-Hypaque gradient centrifugation. Neutrophils are usually pretreated with rabbit serum to block Fc receptors before adding to a binding assay. When both - ~nts in the binding assay are cellular, binding can be assayed microscopically or by flow cytometrY~ See Ki Ch;- Lo et al., supra. When one or both WO 9~3~32~ 2 1 9 1 ~7 ~ 14 P(~ISg~7302 ts is a purified protein, one ~ L is usually immobilized to a solid phase and the other labelled. sinding is then assayed from label bound to the solid phase. Usually, the antibody is preincubated with the source of P-selectin before adding the source of counterreceptor to the incubation mixture. Blocking activity is shown when an excess of antibody, i.e., 5-fold, 10-fold or up to 100-fold, substantially inhibits binding of P-selectin to its count~lLec~Lor. The precise degree of inhibition will depend on the assay used. In an assay that measures inhibition of platelet binding to ~L-60 cells, an excess of P-selectin blor~ing an~;ho~r~ typically exhibits at least 50, 60, 70, 80 or 90~ and usually about 80-90~ inhibition.
The binding specificity of many h1sr~ing antiho~;r~
of the in-~ention is further defined by their capacity to bind P-selectin in the complete or substantial absence of Ca++
(e.g., in the p~ r.~ of 2 m~ EDTA (a calcium chelator) and the absence of Ca++ in an i~ vi~ro assay). By O~IILL~D~, most blocking an~iho~ against P-selectin isolated to date reriuire ca++ for activity. See Geng et al., J. B~'ol. Chem.
266:22313-22318 (1991). Antiho~i~c requiring a Ca++ cofactor for hlork;ng activity may be less effective in in vivo conditions where levels of Ca++ are expected to fluctuate.
The capacity of the an~ho~q of the invention to block E-selectin-mediated functions can be ~ --L~ted by analogous i~ vitro assays to those employed to show blocking of P-selectin mediated fllnrtion~. Suitable sources of E-selectin are mammalian cell lines transfected with E-selectin, activated endothelial cells, as well as purified E-selectin (or extrRr~ r domains thereof). If the assay is performed using purified E-selectin, the E-selectin can be immobilized to a solid support. Suitable sources of ccunterreceptors to E-selectin are leukocytes, n~ut~phils, monocytes, and HL-60 cells and appropriate cell lines transfected with L-selectin. The desree of binding inhibition will again depend on the ~ - -nts in the assay. In an assay that measures binding between activated endothelial cells and HL-60 cells, the antibodies of the invention, when present in ~ W0~5/34324 2 1 3 1 ~ 7 ~ PcT~sg~lo73n2 excess, typically exhibit at least about 20, 40, 60, 80~
inhibition or more typically about 25-75% or 50% inhibition.
The capacity of antibodies to block L-selectin mediated functions can be d L.~ted in a variety of in vitro assays. See, e.g., cop~n~;ng applications 08/160,516, filed !l~v~b_r 30, 1993 and 08/160,074, filed ~lov '-r 30, 1993 (inco-yuL~ted by reference in their entirety for all purposes~. A simple visual assay for detecting such interaction has been described by K; Ch i Lo et al., supra.
Briefly, monolayers of human umbilical vein cells are stimulated with IL-l. rl_uL-u~hils, with or without pretreatment with the antibody under test, are added to the monolayer under defined conditions, and the number of adhering n~uL.u~hils is det~rmlnrd microscopically. In one method, the neutrophils are obtained from human leukocyte adhesion deficient patients. See Anderson et al., Ann. Rev. Med.
38:175 (1987). The ne~LLu~hils from such patients lack integrin receptors, whose binding to neuLLu~hils might obscure the effects of blocking L-selectin binding.
Preferred ant;ho~i~c selectively bind a functional epitope on P-selectin and E-seléctin molecules associated with a le~ol.se to tissue injury and inflammation. Binding of the antih~ R to a functional epitope on P-selectin and E-selectin effectively inhibits ~h~C;~ of leukocytes to the activated vascular endothelium and/or to activated platelets in vivo. Preferred antiho~i~c impair the adhesion of leukocytes to the activated vascular endothelium to prevent or inhibit an i nf 1; tory and/or thrombotic condition.
In vivo blocking efficacy can be ~ LL~ted in the same animal models that have been used to show efficacy for antiho~i~c specific for a single ~hpcion molecule. For example, Mulligan et al., 1991, 1992, supra, describe rat models to test the efficacy of antiho~;~C in protecting against lung injury: rotlghl~n et al., 1994, describe a rat model for testing the efficacy of antibodies in treatment of systemic endotoxemia; and Weyrich et al., supra , describe a cat model for testing the protective effect of antibodies in myocardial ;C~h~;a and reperfusion injury. Other animal 21gl87~
woss/3~32~ ~ ,~"~ 2 models for various inflammatory ~ic~ and disorder~ are described by Arfors et al., Blood 6~338 ~1987~ (skin lesions); Il -r et al., J."Exp. Med. 170:959 (1989) (brain edema and death produced by bacterial meningitis); Lindbom et al., Clin. I~munol. n ~th. 57:105 (1990) (tissue edema associated with delayed-type hypersensitivity reactions);
Wegner et al., SCience 247:456 (1990) (airway hy~e~Le~yunsiveness in allergic asthma); Goldman et al., FAS~3 J. 5:A509 (1991) (remote lung injury following aspiration~;
Gundel et al., ~. Clin. Invest. 88:1407 (1991) (late-phase br~nrhocnn~triction following antigen ~h~ nge~ t~hingc et al., ~ature 346:639 (199Oj (~i~heto~); Flavin et al., ~ranspla t, Proc. 23:533 (1991~ (cardiac allograft survival~;
Wegner et al., Am. Rev. Respir. Dis. 143:A544 (1991) (lung damage and dysfunction sec~n~Ary to oxygen toxicity); Cosimi et al., J. Immunol. 144:4604 (1990) (renal allograft rejection); Jasin et al., Arthritis Rheum. 33:534 (1990) (antigen-induced arthritis); Thomas et al., FASEB J. 5:A509 (1991) (vascular injury and death in endotoxic shock); Bucky et al., P~oc. ~m. Burn Assoc. 23:133 (1991) (burns); Hernandez et al., Am. ~. Physiol. 253:H6ss (1987) (permeability edema following i~} i~ L~Lfusion (IR) of intestine); Winguist et al., Circula~ion 82:III (1990); Ma et al., Cir. Res. 82:III
(1990) (myocardial damage following myocardial infarction);
Mileski et al., Surgery 108:206 (1990) (vascular and tissue damage following hemorrhagic shock and resuscitation); Clark et al., S~roke 22:877 (1991) (central nervous system damage following I/X of the spinal cord); Mileski et al., Proc. Am.
Burn Assoc. 22:164 (1990) (edema and tissue damage following frostbite and rewarming); Simpson et al., Circulation 81:226 (1990) (infarct size following I/R of myocardium). Preferred antibodies show efficacy in at least one and usually several of these inflammatory and thrombotic ~iCPAC~C and conditions.
~any of the blocking antibodies of the invention~5 show the same or similar binding specificity as one of the lAry antibodies designated 5C7.29, 2C9.11 and lD8.10.
That is, the ~ntihoAiefi compete with at least one of the exemplified antiho~iiPc for specific binding to E-selectin ~ W095l34324 2 1 9 1 ~ 7 Q

and~or P-selectin. The E-selectin and P-selectin used in the test is preferably human, and may be natural or recombinant.
Competition between antibodies is det~rminpd by an assay in which the immunoglobulin under test inhibits specific binding of a reference antibody (e.g., 5C7.29) to an antigenic d~t~rmin~nt on a P-selectin and/or E-selectin molecule.
11 uus types of competitive binding assays are known, for example: solid phase direct or indirect radioi --c5ay (RIA), solid phase direct or indirect enzyme ; --ccay (EIA), sandwich competition assay (see Stahli et al., Methods in Enzymology 9:242-253 (1983)); solid phase direct biotin-avidin EIA (see Kirkland et al., J. Immunol. 137:3614-3619 (1986)); solid phase direct labeled assay, solid phase direct labeled sandwich assay (see Harlow and Lane, "Antibodies, A
Laboratory Manual," Cold Spring Harbor Press (1988)); solid phase direct label RIA using I-125 label (see Morel et al., Molec. Immunol. 25(1):7-15 (1988)); solid phase direct biotin-avidin EIA (Cheung et al., Virology 176:546-552 (1990)); and direct labeled RIA (NoldGnhAupr et al., Scand. J. Immu~ol.
32:77-82 (1990)). Ty-pically, such an assay involves the use of purified P-selectin or E-selectin bound to a solid surface or cells bearing either of these, an nnlAh~lled test i ogl~h-llin and a labelled reference immunoglobulin.
Competitive inhibition is measured by det~rmining the amount of label bound to the solid surface or cells in the pL~sence of the test ; -~l~hnl;n. Usually the test i -~lObulin is present in excess. ~ntiho~iGc identified by competition assay (competing ant;ho~;es) include antiho~;~s binding to the same epitope as the reference antibody and antibodies binding to an adjacent epitope sufficiently proximal to the epitope bound by the reference antibody for steric hindrance to occur.
Usually, when a competing antibody is present in excess, it will inhibit specific binding of a reference antibody to P-selectin and/or E-selectin by at least 50 or 75~.
The antiho~i~c of the invention usually exhibit a specific binding af$inity for P-selectin and E-selectin of greater than or equal to about 106, 107, 108, 109, or 101~ M-l.
However, antibodies do not ~C~ccArily show the same specific 21~1~70 WO g513432~ PCr/l~S515~07302 binding affinity for each of these ligands. Usually the upper limit of binding affinity of the Antihofli~~ is within a factor of about three, five or ten of that of one of the ~ 1;fied an~;ho~ioc. Often the lower limit of binding affinity is also within a factor of about thr~e, five or ten of that of the exemplified antihofl~ The term "about" ~ --8~ the degree of experimental error that may typically occur in the mea~uL~ L of binding affinities.
A hybridoma producing the 5C7.29 antibody has been deposited with the American Type Culture Collection, 12301 Parklawn Dr., Rockville, Maryland under the Budapest Treaty on May 25, 1994 and given the Accession No. ATCC CRL 11640. The production of this antibody is described in Example l.

c. Production of Ant;hodies ~1~ 170nLu~al Antibodies Mouse, or other ~onhn~-n antiho~ios crossreactive with P-selectin and E-selectin can be obtained using a variety of immunization strategies. In some strategies, nonh~ n animals tusually n..,.l... ~-) ma als~, such as mice, are immunized with E-selectin and P-selectin antigens, either cu..uuLLently or cGnse~uLively. In other strategies, n~nh~ n animals are i ~70d with only one of these antigens.
Preferred i , are cells stably transfected with P-selectin or E-selectin and expressing these molecules on their cell surface. Other pre~erred i- , - include P-selectin and E-selectin proteins or epitopic fragments of P-selectin and E-selectin containing the 6egments of these molecules that bind to the exemplified crossreacting antibodies.
Mouse or non-human ~nt i hofl i ~ crossreactive with all three selectins, i.e., P-selectin, E-selectin, and L-selectin, can be generated by similar strategies. Briefly, mice are i i ~efl either simult~n~oncly or sequentially with cells stably transfected with either P-selectin, E-selectin, or L-selectin, or purified selectin proteins or epitopic fL~,_ ts thereof.

09~4324 ~19 18 7n PC~S~107302 Antibody-producing cells obtained from the i ;~ed animals are immortalized and selected for the production of an antibody which specifically binds to multiple selectins. See generally, Harlow & Lane, Antibodies, A Laboratory ~anual (C.S.H.P. NY, 1988) (incorporated by reference for all ~uL2,oses). The binding assays for the different selectins can be peLroL -~ separately or c~ncuLL~--tly. C~IICULL~ analysis is conveniently peLL~ -' by two-color FACS screening after incubation of hybridoma supernatants to cells transfected with selectins. For example, two populations of cells respectively expressing E-selectin and P-selectin are differentially labelled with a first label and tested for capacity to bind hybridomas supernatants. Binding is detected using an appropriate sec~n~2~qry antibody bearing a second label. This scheme is readily extendible to allow simultaneous detection of binding to all three selectins by differentially labelling three populations of cells respectively expressing E-selectin, P-selectin and L-selectin with different intensities of the first label. Alternatively, separate screening for E-selectin, P-selectin and, if desired, L-selectin binding, can be achieved by single color FACS analysis of supernatant binding to transfectant cells or by binding assay to immobilized E-selectin, P-selectin, or L-selectin.
Crossreacting ant;ho~2ioc are then further screened for their capacity to block functional properties of E-selectin, P-selectin and L-selectin using the in vitro and in vivo assays described above. Most anti ho~;rqc that crossreact with P-selectin or E-selectin also block the functional capacity of both of these molecules to interact with a counterreceptor.
(2~ H~lr~r2iz~qd Antibodies The invention provides hnr~nized antiho~;rqc having similar binding specificity and affinity to selected mouse or other n~nh~ n antibodies. ~ rqni~qd antibodies are formed by linking CDR regions (preferably CDR1, CDR2 and CDR3) of non-human antibodies to human fl ~Ik and constant regions by reco3binant DNA techniques. See Queen et al., Proc. Natl.
Acad. Sci. USA 86:10029-10033 (1989~ and W0 9o/07861 ~'0951343~ 2 i ~ 1 8 7 u I C ~ a~

(incu,uuL~ted by reference in their entirety for all ~uL~oses~. The humanized immunoglobulins have variable region fL 'OLk residues substantially from a human immunoglobulin (termed an acceptor immunoglobulin) and ~ mity detormininq regions ~ub~Ldntially from a mouse immunoglobulin described above, e.g., the 5C7~29 antibody (referred to as the donor immunoglobulin~. The constant region~s), if present, are also subst~nti~lly from a human immunoglobulin.
In principal, a rL - ~Lk Be~uence from any human antibody may serve as the template for CDR grafting. However, it has been ' L~ted that straight CDR rop~ onto such a rL ~Lk often leads to significant 1O5S of binding affinity to the antigen (Glaser et al., J. Immunol. lC9: 2606 (1992); Templst et al., Biotechnology 9: 266 (1992) Shalaby et al., J. Exp. Med. 17 217 (1992) ) ~ The more homologous a human antibody is to the original murine antibody, the less likely will c~-h~ning the murine CDRs with the human framework be to introduce distortions into the CDRs that could reduce affinity. ~ ole, homology (that is, percent cc~Qnre 20 identity) of at least 65% between the humanized antibody variable region framework and the donor antibody variable region rL JL~ iS preferred.
The heavy and light chain variable region framework residues can be derived from the same or different human 25 antibody se~an~Pc. However, a heavy chain and light chain framework se~uon~a~ chosen from the same human antibody reduce the possihility of i-- _tibility in assembly of the two chains. The human antibody Ee~ a~ can be the se~onc~c of naturally occurring human antiho~ipc or can be c~ncDnc~lc 30 ~o~lonro~ of several human antibodies. See Carter et al., W0 92/22653~ Certain amino acids from the human variable region rL ..JLk residues can be selected for substitution based on their possible influence on CDR conformation and/or binding to antigen. Investigation of such possible influences 35 is by modeling, examination of the characteristics of the amino acids at particular locations, or empirical observation of the effects of substitution or mutagenesis of particular amino acids.

~o g~,34324 2 1 9 1 ~ 7 '~ e~ 2 For example, when an amino acid differs between a murine 5C7.29 variable region framework residue and a salected human variable region rS . J1k residue, the human LL .~L~
amino acid should usually be substituted by the equivalent rL ,,~L~ amino acid from the mouse antibody when it is re~c~n~hly expected that the amino acid:
(1) contacts antigen directly, (2) is adjacent to a CDR region in the se~ nre, or (3) otherwise interacts with a CDR region (e.g., is within about 4-6 A of a CDR region).
Other candidates for substitution are acceptor human fL ~Lk amino acids that are unusual for a human i ,lobulin at that position. These amino acids can be substituted with amino acids from the equivalent position of the donor antibody or from the equivalent positions of more typical human ; -~lobulins. The variable region frameworks of hn~-n; 70d immunoglobulins usually show at least 85%
sP~anre identity to a human variable region fL W~L~
sequence or concal.~C of such sequancPc.
f3~ Human ~nt; h~ ies In another aspect of the invention, human ant;ho~;aC
cross-reactive with E-selectin and P-selectin are provided.
These antiho~;as are ploduced by a variety of terhn;quac described below. Some human Ant;ho~;es are selected by competitive binding experiments, or otherwise, to have the same epitope Crar;flrity as an exemplified mouse antibody, such as 5C7.29. Such antiho~;ac are particularly likely to share similar therapeutic properties.
a. Trioma ~ethodolooy The basic approach and an a lAry cell fusion partner, SPAZ-4, for use in this approach have been described by Oestberg et al., Hybridoma 2:361-367 (1983): Oestberg, U.S.
Patent No. 4,634,664; and Engleman et al., US Patent 4,634,666 (each of which is incuL~oL~ted by reference in its entirety for all purposes). The antibody-producing cell lines obtained by this method are called triomas, because they are ~ccan~a WO 95/3~32 1 2~19 1~ 7 0 PC~AIS9~073(12 from three cells--two human and one mouse. Initially, a mouse myeloDa line is fused with a human B-ly~ho~yLe to obtain a non-antibody-producing Y~n~g~n~i~ hybrid cell, such as the SPAZ-4 cell line described by Oestberg, supra. The x~n~g~n~ic cell is then fused with an ~ ; 79d human B-ly '- yLe to obtain an antibody-producing trioma cell line. Triomas have been found to produce antibody more stably than ordinary hybridomas m~de from human cells.
The B-l~ '-s~Les are obtained from the blood, spleen, lymph nodes or bone marrow of a human donor. In vivo ; ;7ation of a living human with E-selectin and/or P-selectin is usually undesirable because of the risk of initiating a harmful 1. ~Dnse. Thus, B-lymphocytes are usually ; ;~o~ in vitro with an E-selectin and/or P-selectin or an antigenic fragment of either of these, or a cell bearing either of these. Specific epitopic fL, Ls consisting essentially of the amino acid segments that bind to one of the exemplified murine ant; hofl; ~C are preferred for i~
vitro; ipation~ B-ly~ho~yLes are typically exposed to antigen for a period of 7-14 days in a media such as RPMI-1640 (see Englemanr supra) s~ppl~ ~ed with 10~ human serum.
The immunized B-ly~Lo~y~es are fused to a Y~n~g~n~;~ hybrid cell such as SPAZ-4 by well known methods.
For example, the cells are treated with 40-50~ polyethylene glycol o~ MW 1000-4000, at about 37 degrees, for about ~-10 min. Cells are saparated from the fusion mlxture and propagated in media selective for the desired hybrids (e.g., HAT or AH). Clones secreting antibodies having the required binding specificity are identlfied by assaying the trioma culture medium for the ability to bind to E-selectin and P-selectin u ing the same methods as fl;ccucced above for n~nhllr-n an~ 7 hofl; ~ . Triomas producing human antibodies having the desired specificity are subcloned by, e.g., the limiting dilution te~hniq~le and grown in vitro in culture Dedium.
Although triomas are genetically stable they may not produce anti~ofl;~c at very high levels. Expression levels can be increased by cloning antibody genes from the trioma into ~ wog~,34324 219 1 8 7 Q PCT~S95l07302 one or more expression vectors, and trans~orming the vector into a cell line such as the cell lines cl;cc~cs~cl, infra, for expression of recombinant or humanized i -71Obulins.

b. T~_ns~ ic Non-~nr~n Mammals Human ant;hs~;es crossreactive with P-selectin and E-selectin can also be pIc,~uced from non-human trAnegPn;~
mammals having trA-ncgenec ~n~o~ i n; at least a segment of the human immunoglobulin locus. Usually, the endogenous immunoglobulin locus of such transgenic mammals is functionally inactivated. Preferably, the segment of the human immunoglobulin locus includes unrearranged sequences of heavy and light chain ~ ents. Both inactivation of ~n~og~n~us immunoglobulin genes and introduction of exogenous immunoglobulin genes can be achieved by targeted homologous recombination, or by inLLudu~Lion of YAC chL, ~t ~. The transgenic mammals resulting from this process are capable of functionally rearranging the immunoglobulin _ -1 secr~nres, and expressing a repertoire of antibodies of various isotypes encoded by human immunoglobulin genes, without e~cpressing ~ndOCJ~n~InC immunoglobulin genes. The production and properties of mammals having these properties are described in detail by, e.g., Lonberg et al., WO93/12227 (1993); Kucherlapati, WO 91/10741 (1991) (each of which is incorporated by reference in its entirety for all purposes).
Transgenic mice are particularly suitable. Crossreacting P-selectin/E-selectin human antibodies are obtained by i i7;ng a t~Ancg~ nnnh~ n mammal, such as described by Lonberg or Kucherlapati, supra, according to the same strategy as ~iccnles~cl for a nonLlAi.scJ~Ilic n~Anhn~-n animal (section I.C.(l)). Monoclonal antibodies are prepared by, e.g., fusing B-cells from such mammals to suitable myeloma cell lines using convcnfi~AnAl Kohler-Milstein technology.

c. Phaqe Dis~lav Methods A further approach for obtaining human crossreacting antibodies to E-selectin and P-selectin is to screen a DNA library from human B cells as described by Dower W095~4324 2 ~ 7 0 PCT~595/07302 et al., ~0 91/17271 and McCafferty et al., wo 92~01047 (each of which is in~L~ted by reference in its entirety for all purposes). In these methods! Iibraries of phage are ~l~duced in which members display different an~iho~;pc on their outer surfaces. ~ntiho~ are usually displayed as Fv or Fab fL_~ Ls. Phage displaying antibodies are selected by affinity enrichment for binding to either P-selectin or E-selectin. Phage identified by the initial screen are then further s~L~ened for crossreaction with the other ligand.
In a variation of the phage-display method, human ant;ho~ having the binding spec;fi~ity of a selected murine antibody can be produced. See Winter, W0 92/20791. In this method, either the heavy or light chain variable region of the selected murine antibody ~e.g., 5C7.29~ is used as a starting material. If, for example, a light chain variable regicn is selected as the starting material, a phage library is ccnstructed in which members displays the same light chain variable region ti. e., the murine starting material) and a different heavy chain variable region. The heavy chain 2D variable regions are obtained from a library of rearranged human heavy chain variable regions. A phage showing strong specific binding for P-sel~ctin and E-s~3~c~;n (e.g., at least I08 and preferably at least lOg M-l) is selected. The human heavy chain variable region from this phage then serves as a starting material for col~LLu~Ling a further phage library.
In this library, each phage displays the same heavy chain variable region (i. e., the region identified from the first display library) and a different light chain variable region.
The light chain variable regions are obtained frcm a library of rea~ranged human variable light chain regions. Again, phage showing strong specific binding for P-selectin and E-selecti~ are selected. These phage display the variable regions of completely human ~nt;ho~;es that crossreact with E-selectin and P-selectin. These antibodies usually have the same or similar epltope specificity as the murine starting material (e.~., 5C7.29).

2191~7~
woss/~2~ ~ ~ PCT~S95/073n2 D. Bis~ecific- Antlbodies The invention also provides bispecific or bifunctional antibodies that have one binding site that specifically binds to P-selectin and E-selectin and a second binding site that specifically binds to a second moiety. In bispecific antibodies, one heavy and light chain pair is usually from a crossreacting antibody and the other pair from an antibody raised against another epitope. This results in the property of multi-functional vaiency, i.e., ability to bind at least two different epitopes simul~nPollclyr one of which is the epitope to which the anti P-selectin/E-selectin crossreacting antibody binds. The other epitope could be e.g., an epitope on L-selectin.
E. Other Therapeutic Aaents Having p~uduced an antibody having desirable properties, such as 5C7.29 and the other ,--- lified antihoAi~c, other nonantibody agents having similar binding specificity~and or affinity can be p~oduced by a variety of methods. For example, Fodor et al., US 5,143,854, discuss a terhni~lD termed VLSIPS~, in which a diverse collection of short peptides are for~ed at selected positions on a solid substrate. Such peptides could then be s~L~ened for binding to an epitopic rL L r~cor~ni~1 by 5C7.29, optionally in competition with the 5C7.29. Libraries of short peptides can also be produced using phage-display technology, see, e.g., Devlin WO91/18980. The libraries can be screened for binding to an epitopic rL L recognized by e.g., 5C7.29, optionally in competition with 5C7.29.

II. Nucleic Acids The genes ~nroAing the heavy and light chains of immunoglobulins ~L~duc~d by hybridoma or trioma cell lines secreting crossreacting antibodies are cloned according to methods described in Sambrook et al., Molecular Cloning: A
~aboratory Manual~ 2nd ed. (Cold Spring Harbor, NY, 1989~;
serger & Rimmel, Methods in Enzymology, Volume 152, Guide to Molecular Cloning Techniques (Academic Press, Inc., San Diego, CA, 1987); Co et al., J. Im~unol. 148:1149 (1992). For ~vos~3~324 2191870 26 P~'l'/LlS~51û73()2 example, genes P~co~lnq heavy and light chains are cloned from a hybridoma' 5 genomic DNA or cDNA produced by reverse tL~ns~Liption of RNA. Cloning ,IS accomplished by conventional t~hn i T'~~ including the use of PCR primers that hybridize to the sequences flanking or overlapping the genes, or 8- --ts of genes, to be cloned.
Typically, recombinant ounLL~u~Ls comprise DNA
5~ nno~; ng a complete human immunoglobulin heavy chain and/or a complete human i _lobulin light chain of an Jlobulin expressed by a hybridoma or trioma cell line.
Alternatively, DNA sesment6 ~nro~ i ng only a portion of the primary antibody genes are produced, which portions possess binding and/or effector activities. Other recombinant C~11L~LU~LS contain s~_ c of immunoglobulin genes fused to segments of other immunoglobulin genes, particularly s-of other human constant region cP~lon~c (heavy and/or light chain). ~uman constant region sequences can be selected from various reference sources, including those listed in Rabat et al., supra.
DNA s-, ~~ encoding crossreacting P-selectin/
E-selectin antibodies can be modified by recombinant DNA
techniques such as site-directed ~n~c~5 (see ~illman &
Smith, Gene 8:81-97 (1979~; Roberts et al., Nature, 328:731-7}4 (1987). Such modified segments will usually retain antigen binding capacity and/or effector function. Moreover, the modified segments are usually not so far changed from the original 5~1 PnC~c to prevent hybridization to these se~nr~C
under stringent conditions. The modified 5~, will usually encode an immunoglobulin showing substantial sequence identity to a reference immunoglobulin from which it was derived. Because, like many genes, immunoglobulin genes contain separate functional regions, each having one or more distinct biological activities, the genes may be fused to functional regions from other genes to produce fusion proteins te.g., immunotoxins) having novel properties or novel combinations of properties.
The recombinant polynucleotide constructs will typically include an expression control sequence operably ~ W095~4324 21~1 ~ 7 ~ PCT~lssslo73n2 linked to the coding ~e~l~on~oc, inr~ ;ng naturally-associated or heterologous p~ t~ regions. Preferably, the expression control se~lon~c will be eukaryotic promoter systems in vectors capable of transforming or transfecting eukaryotic host cells. Once the vector has been incuLyuL~ted into the appIup~iate host, the host is maintained under conditions suitable for high level expression of the nucleotide seyuoncoC~ and the collection and purification of the ~L~L.7-o-ing ant;ho~ies.
These expression vectors are typically replicable in the host organisms either as episomes or as an integral part of the host ~ l DNA. Commonly, expression vectors will contain selection markers, e.g., ampicillin-resistance or hy~L._i~in-resistance, to permit detection of those cells transformed with the desired DNA ce~lDnr~c.
E. coli is one prokaryotic host particularly useful for cloning the DNA se~lonces of the present invention.
Microbes, such as yeast are also useful for expression.
Saccharomyces is a preferred yeast host, with suitable vectors having expression control so~lon~oc~ an origin of replication, termination s~len~oc and the like as desired. Typical promoters include 3-ph~sphoglycerate kinase and other glycolytic enzymes. In~llrihl~ yeast promoters include, among others, promoters from alcohol dehydrogenase, iso~yLo~l.L~ - C, and enzymes roqp~ncihle for maltose and g~lactose utilization.
Mammalian cells are a preferred host for expressing nucleotide s~_ L~ ~nco~ing immunoglobulins or r. ~s thereof. See Winn~rk~r, From Genes to Clones, (VCH
Publishers, NY, 1987~. A numher of suitable host cell lines capable of secreting intact heterologous proteins have been developed in the art, and include CHO cell lines, various Cos cell lines, HeLa cells, L cells and myeloma cell lines.
Preferably, the cells are nonhn~-n. Expression vectors for these cells can include expression control sequences, such as an origin of replication, a promoter, an onh lnoor (Queen et al., Immunol. Rev. 89:49 (1986)), and ~ococs~ry processing information sites, such as ribosome binding sites, RNA splice sites, polyadenylation sites, and transcriptional terminator w0~s~34324 21918 7 ~ ~ pcTNsss~l73~l2 s~ . Preferred expression control se~l~nr~ are promoters derived from ~n~lo~J ~ c genes, cytomegalovirus, SV40, adenovirus, bovine papillomavirus, and the like. See Co et al., J. Immunol. 148:1149 (15,92~)~?.
The vectors c~nt~ini'n~ the DNA 9:, L~ of interest can be transferred into the host cell by well-known methods, ~p~n~ing on the type of c~llular host. For example, calcium chloride transfection is commonly utilized for prokaryotic cell~, whereas calcium phosphate treatment, electroporation, lipofection, biolistics or viral-based transfection may be used for other cellular hosts. Other methods used to transform ~ 11 ~n cells include the use of polybrene, protoplast fusion, li,- --, ele~Lu~oL~tion, and microinjection (see generally, Sambrook et al., supra).
lS Once expressed, crossreacting i ~;lobulins of the invention can be purified according to standard ~IUCedUL~S of the art, including HPLC purification, column chromatography, gel eleoLL~horesis and the like (see generally, Scopes, Protein Purification (Springer-Verlag, NY, 1982)).
III. E~ito~e MaD~in~
The P-selectin epitope(s) bound by the 5C7.29 or other crossreacting antibody can be determined by providing a family of rL t~ containing dirferent amino acid SE_ L~
from P-selectin. Each LL _ L typically comprises at least 4, 6, 8, 10, 20, 50 or 100 contiguous amino acids. The family of polypeptide f _ ~ Ls cover much or all of the amino acid sP~Pnre of the extracellular domain of a P-selectin polypeptide. Members of the family are tested individually for binding to e.g., the 5C7.29 antibody. The smallest ~L, t that can specifically bind to the antibody being tested contains the amino acid se~lPnre of the epitope rPcogni7ed by the antibody. The E-selectin epitope bound by the antibody is mapped by an analogous strategy using a family of peptides from E-selectin. The respective epitopes on P-selectin and E-selectin are expected to map to segments of these molecules showing a high degree of seguence identity.
The epitopic fragments are useful as i ~ for generating ~ ~095/34324 2 1 9 1 ~ 7 ~ Pc~T~g~/o73n2 further crossreacting antibodies. The epitopic fragments are also useful as therapeutic agents that agonize or an~g~nize the function of P-selectin or E-selectin.
Another method to map epitopes involves testing the ability of an antibody to bind to E-selectin or P-selectin to which random mutations have been introduced. This method is described in more detail in Example 9.

IV. Pharmaceutical Com~ositions The ph~ Lical compositions for use in the thar~p~u~ic methods ~iccllcs~d infra, typically comprise an active agent, such as crossreacting E-selectin/P-selectin antibody, dissolved in an acceptable carrier, preferably an aqueous carrier. Some compositions contain a cocktail of multiple active agents, for example, a crossreacting antibody and a thrombolytic agent. A variety of aqueous carriers can be used, e.g., water, buffered water, phosphate buffered saline (PBS~, 0.4% saline, 0.3% glycine, human albumin solution and the like. These solutions are sterile and generally free of particulate matter. The compositions may contain rh~rr--eutically acceptable auxiliary substances as required to approximate physiological conditions such as pH
adjusting and buffering agents, toxicity adjusting agents and the like, for example, sodium acetate, sodium chloride, potassium chloride, calcium chloride and sodium lactate. The c~.,cerLL~tion of antibody in these formulations can vary widely, i.e., from less than about 0.005~, usually at least about 1% to as much as 15 or 20% by weight and will be selected primarily based on fluid volumes, viscosities, and so forth, in accordance with the particular mode of administration selected.
Thus, a typical pharmaceutical composition for injection could be made up to contain 1 ml sterile buffered water, and 1-10 mg of i ngl obulin. A typical composition for intravenous infusion could be made up to contain 250 ml of sterile Ringer's solution, and 150 mg of antibody. Nethods for preparing parenterally administrable compositions are described in Remington 's ph~rm~ce~ltical Science (15th ed., WO g513.1324 21~ 1~ 7 0 r ~ 0 ~"~ ~
. ~

Nack Publishing Company, Easton, PA, 1980), which is in~uL~vL~ted by reference in its entirety for all ~L~03~9.
Therapeutic agents of the invention can be frozen or lyophilized for storage and reconstituted in a suita_le carrier prior to use. Lyophilization and reconstitution can lead to varying degrees of antibody activity loss (e.g., with conventional immune globulins, IgM antih~ c tend to have greater activity loss than IgG antiho~i~c). Dosages may have to be adjusted to t~.
V. ~.~eL~nPI~Lic ~ethods The antiho~i~c of the present invention are useful for t~eat ~ of infl. tory ~;~P~CPR and conditions, oCpP~i~lly those which are mediated by neuLlu~hils. The dual specificity of the antibodies leads to the inhibition of inflammatory events mediated by either P-selectin or E-selectin.
For example, the antibodies are suitable for therapeutic and prophylactic treatment of ic~h~ia-reperfu5ion injury caused by myocardial infarction, cerebral is~hP~c event (e.g., stroke), renal, hepatic or splenial infarction, brain surgery, lung injury, shock, cardiac surgery (e.g., coronary artery bypass), elective angioplasty, and the like.
other preferred applications are the treatment of sepsis, adult respiratory distress ~y--dL~ , and multiple organ failure. The ~ntiho~ies are also useful for treating injury due to trauma, burns, frostbite or damage to the spinal cord.
The antiho~ will also find use in treating aut~i -~iCPAC~ incln~in~ rheumatoid arthritis, systemic lupus erythematosus, multiple sclerosis, type I diabetes and uveitis, in treating inflammatory ~ica~cP~ of the skin such as psoriasis, and in treating meningitis and ~n~Prh~litis. The ~ntiho~i~q are also useful for treating allergic rhinitis, asthma and anaphylaxis. Other typical applications are the prevention and treatment of organ transplant rejection and graft-versus-host disease.
The ph~rr-~~utical compositions containing the antiho~i~C are particularly useful for parenteral ~l9lS70 w09sl34324 ~ r~

administration, i.e., subcutaneously, inLL ~ 7~rly or intravenously. The ant;ho~i~C of the invention may also be administered, typically for local application, by gavage or lavage, i..LL~elitoneal injection, ophthalmic ointment, topical ointment, intracranial injection (typically into a brain ventricle), intrapericardiac injection, or intrabursal injection.
The compositions containing the present antibodies or a cocktail thereof can be administered for prophylactic and/or therapeutic treatments. In therapeutic applications, compositions are administered to a patient already suffering from an inflammatory disease, in an amount sufficient to cure or at least partially arrest the disease and its complications. An amount adequate to accomplish this is defined as a "therapeutically effective dose." Amounts effective for this use will depend upon the severity of the disease and the general state of the patient's own immune system, but generally range from about 1 to about 200 mg of antibody per dose, with dosages of from 5 to 80 mg per patient being more commonly used. Dosing schedules will vary with the disease state and status of the patient, and will typically range from a single bolus dosage or continuous infusion to multiple administrations per day (e.g., every 4-6 hours), or as indicated by the treating physician and the patient's condition. In life-threatening or potentially life-threatening situations, it is possible and may be felt desirable by the treating physician to administer substantial ~Yc~cs~c of these antibodies.
In prophylactic applications, compositions containing the present an~ihs~l~C or a cocktail thereof are administered to a patient not already suffering from a particular disease to enhance the patient's resistance. Such an amount is defined to be a "prophylactically effective dose." In this use, the precise amounts again depend upon the patient's state of health and general level of immunity, but generally range from 1 to 80 mg per dose. Preferred prophylactic uses are for the prevention of adult respiratory distress ~y~ldL- - in patients already suffering from sepsis or WO g~ 324 2 1 9 1 8 7 ~ P~

trauma; prevention of organ transplant rejection; and prevention of Lr ~_L~ ~ion injury in patients suffering from icrh~mi~. In serlously ill patients, dosages of about 50 to 150 mg of hnr-ni~ or human i~unoglobulin per administration s are r-e~uent~y used, and large~r dosages may be indicated.
Single or multiple administrations of the composition~ can be carried out with dose levels and pattern being selected by the treating physician. In any event, the ph~rr--eutical formulations should provide a quantity of the antibody(ies) of this invention sufficient to treat the patient effectively.
The antiho~i~c can also be used in combination with other antiho~ios~ particularly antibodies reactive with different a~lesion molecules. For example, suitable antibodies include those specific for CDlla, CDllb, CD18, L-selectin, and ICAN-l. Other suitable an~iho~ipc are those specific for ly ~ in~c, such as IL-l, IL-2 and IFN-~, and their receptors. The antihoAi~C of the invention can also be administered in conjunction with chemotherapeutic agents.
Suitable agents include non-steroidal anti-infl; tory drugs and corticost.eroids, but ~ua additional agents (e.g., cyclosporin) can also be used.
In some therapeutic methods of iC~ho~ p_Lru_ion therapy, crossreacting antiho~ are used in comhination with thrombolytic agents. In previous methods, patients with myocardial infarction or unstable angina are often treated by opening the occluded coronary artery. Poop~ning of the obstructed coronary artery can be achieved by administration of thrombolytic agents which lyse the clot causing the obstruction, and which, thereby, restore coronary blood flow.
Reperfusion of the vessel can also be achieved by percutaneous transluminal coronary angioplasty (PTCA) by means of balloon dilation of the ob~LLu~ed and narrowed segment of the coronary artery. However, restoration of colonaLy blood flow leads to icrh~ -reperfusion injury in prior methods.
In the present methods, ic~h~ reperfusion injury is reduced or prevented by combination of a thrombolytic agent or of PTCA with crossreacting E-selectin/P-selectin ~ WO9~/34324 219 1~7 0 PCT~S9~07302 an~; ho~; ~c . Ant; ho~ i ~R are usually administered prophylactically before, or at the same time as, administration of thrombolytic agents or initiation of PTCA.
Further doses of antibody are then often administered during and after thrombolytic or angioplastic treatment. The interval between prophylactic administration of the antibodies and initiation of thrombolytic or angioplastic treatment is usually 5-60 mins, preferably 5-30 min, and most preferably 5-10 min. The ~ntiho~ies are administered parentally, preferably by i~ avenuus injection, in doses of 0.01-10 mg/kg body weight, preferably of 0.14 - 5 mg/kg and most preferably of 0.3 - 3 mg/kg. The ant; ho~ i Pc can be given as an il.LL~venvus bolus injection, e.g., over 1 - 5 min., as repeated injections of smaller doses, or as an intL~venous infusion. The bolus injection is P~pPri~lly useful for the prophylactic dose or in an emergency. Further doses of antiho~iPR can be repeated (e.g., every 4 - 24 hr) during and after thrombolytic or angioplastic treatment of acute myocardial infarction at the same proportions as described above to achieve optimal plasma levels of the antibody.
~ IL~ ' -lytic agents are drugs having the capacity, directly or indirectly, to stimulate dissolution of thrombi in vivo. Thrombolytic agents include tissue pl~m; n~g~n activator (see EP-B 0 053 619), activase, alteplase, duteplase, silteplase, streptokin~ce~ anistreplase, urokinase, heparin, warfarin and coumarin. Additional thrombolytic agents include saruplase and vampire bat pl~cmin~g~n activator. See Harris, Protein Engineering 6:449-458 (1987);
PCT/EP 90/00194; US Patent 4,970,159. ~ olytic agents are administered to a patient in an amount sufficient to partially disperse, or prevent the formation of, thrombi and their complications. An amount adeguate to accomplish this is defined as a ~therapeutically effective dose" or "efficacious dose." Amounts effective for this use will depend upon the severity of the condition, the general state of the patient, the route of administration and combination with other drugs.
Often, therapeutically effective doses of thrombolytic agents and administration regimens for such agents with crossreacting w09s~3~324 21~18 7 ~ r~ L~

an~;hs~iic to E-selectin and P-selectin are those approved by the FDA for in~er~n~nt uses of thrombolytic agents, e.g., lO0 mg of alteplase or 1.5 mil~ion'IU of strept~kin~e.
~, .
VI. ~ethods of Diaanosis The monoclonal antiho~i~c of the present invention are useful for ~aqn~sing the ~nfl; tory conditions ~iccllcs~ above and monitoring the treatment thereof. The ~n~iho~ies detect P-selectin and E-selectin in a tissue sample such as serum or endothelial cells, e.g., by ELISA or RIA.
The p-esence of either selectin is diagnostic of inflammation.
Selectin levels may be employed as a differentiation marker to identify and type cells of certain l1n~g~c and developmental origins.
In such pluced~Les, the antibody can be labelled directly (e.g., by r~io~rt1ve or fluuL~c~,,L label) and immune ~ lPY~c detected via the label. Usually, however, the antibody is 1lnl~h~lled and the desired antigen - -clonal antibody complex ls detected with an enGy conjugated antibody against the monoclonal antibody. Diagnosis can also be achieved by in vivo administration of a labelled crossreacting P-selectin/E-selectin antibody and detection by i~ vivo imaging. The c~r~ntration of antibody administered should be sufficient that the binding to cells having the target antigen is detectable compared to the ba~Luu.ld signal. The diagnostic reagent can be labelled with a radioisotope for camera imaging, or a paramagnetic isotope for magnetic I~con~nre or electron spin rPc~n~nre imaging.

VII. O~h~r ~ses The antibodies are also useful for affinity purification of selectins and cells expressing the same on their ~Yt~rn~1 surfaces. The antibodies can also be used to generate anti-idiotypic an~iho~C that mimic a selectin domain responsible for antibody binding. Anti-idiotypic antibodies are useful as competitive inhibitors of selectin binding. For example, an anti-idiotypic antibody to a crossreacting P-selectin, 219187~
wo~sl343~

E-selectin monoclonal antibody can be selected to compete with P-selectin and/or E-selectin for binding to their counterreceptors. The an~;hoai~s are also useful in screening for a therapeutic agent having the same binding specificity as a crossreacting antibody (see Section I. E).
The following examples are provided to illustrate but not to limit the invention:

FY~mnle 1: Pre~aration of Cells Transfected With Selectins L1-2 murine pre-B cell selectin transfectants are obtained by inserting the respective human selectin genes downstream of the LCNV promoter in pMRB101 or similar plasmid (pMRB101 is a derivative of EEb which contains the E. coli gpt gene. Mulligan et al., Proc. Natl. Acad. Sci. USA 78:2072-2076 (1981); Stephans et al., Nucleic Acids Research 17:7110 (1989)). Plasmid DNA is introduced into Ll-2 cells by standard methods, such as ele~LLopoLation, and the cells are selected for resistance to mycoph~n~lic acid. Cells expressing high levels of the appropriate selectin are further selected by "panning'l or fluo~-Ecence activated cell sorting ~hniques. See Lymphocytes, A Prac~ical Approach (G.C.B.
Klaus, IRL Press, Oxford, England, 1987).

ExamDle 2: Production of Crossreactinq Monoclonal Antibodies CLuss~eacting antibodies were ~Loduced using two different ;mmnni7ation procedures. In all of these procedures, the ;nocl11nm was 107 Ll-2 selectin transfectant cells (Berg et al., 1991, 1992, supra) in PBS per injection into mice. In one pL~cédu~e, Balb/c mice at 4-6 weeks of age (Simonson Labs, Gilroy, CA) were injected IP with Ll-2E-3eleCtin transfectants at day 0 and day 14, and L1-2P-selectin transfectants at day 46, followed by fusion of spleen cells on day 50. In a second pI~ced~L~, C57/Ld mice at 4-6 weeks of age (Jackson Labs, Bar Harbor, ME) were immunized in the footpad with hypotonically lysed L1-2E-9eleCtln cells on day 0, then with intact L1-2E-~eleCtin cells on days 3 and 6, and with L1-2P~9eleCtin cells on day 9- The draining lymph node lymphocytes were fused on day 12. In each ~loced~

~o~s/34324 219f~7~ 36 PcT~sS/~}73l~2 mouse B-cells were fused with P3X mouse myeloma cells using polyethylene glycol.
Hybridoma supernatants were s~L__ned for spec;f;c binding to both E- and P-sele~tin by two-color F~CS analysis.
Ll_2P-8e1ectin and Ll_2c2ntrol'tranSfectants were biotinylated by incubation with amino hexanoyl-biotin N ~ Lv~ suc~i n;m~
tZymed Labs, South San Francisco, CA) at 10 ~g/ml in PBS pH
8.0 for 25 min, at RT. After washing, 2 x 107 cells/ml were incubated with FITC-Z-Avidin (Zymed Labs, So. San Francisco, CA) diluted 1:150 for Ll-2P-~eleCtin cells and 1:1000 for Ll-2C~ntr~l cells in FAC5 Buffer 12~ BSA~PBS/10 mM NaN~ for 30 min at 4~C. After washing, cells were mixed with llnl~h~lled L1-2E-~electirA cells at a 1:1:1 ratio in FACS Buffer. 50 ~1 hybridoma supernatants were added to 200,000 mixed cells in 50 ~1 in 96-well plates and incubated for 1 hr on ice. After washing, secnn~ry agent was added, 50 ~1 of 1:500 Goat F(ab')2 anti-mouse IgG-PE ~o..juy~ted (TAG0, Burlingame, CA) for 30 min prior to washing and fixation. FACS analysis was p~Lf~ -~ on a Becton Dic~;n~on FACScan~ (San Jose, CA), according to standard procedures.
Supernatants c~ntA~n~n~ ~nt;ho~;es reacting with both P-selectin and E-selectin were identified by a shift in red fluuL~ccel.~e of the L1-2~-~e1eCtin transfectant (unlabelled with ~ITC) and the brightest FITC labelled cells (Ll-2P-8eleCtin transfectants). The control L1-2 cells (moderately labelled with FITC) did not show a shift in red fluoL~sc~nce, indicating that binding was srecifl~ for P-selectin and E-selectin. The yield of ~Lu~Laacting an~;hn~;es as a ratio of supernatants soLeened was 1/844 and 2/57 for the two ;~ln;z~tion schedules.
Supernatants showing binding to P-selectin and E-selectin transfectants were subc~oned by limiting dilution and grown in serum free medium containing residual amounts of FBS. Three E-/P-selectin cross-reacting antibodies, designated 5C7.2~, lD8.10 and 2C9.11, were purified from these supernatants on Protein A-Sepharose (Pierce) according to the re~- 'ed protocol. Two antiho~;~C reacting only with E-selectin, lE4 and 2D4, and an antibody reacting only with o g~34324 21918 7 ~ : pCT~usgsln73n2 P-selectin, 5F4, were identified by the same method. The isotypes of 5C7.29, lD8.10, 2C9.11, lE4, and 5F4 were det~rmined to be IgG1, and that of 2D4 was determined to be IgG2a using an Tnn~gPnPticS Inno-Lia mouse monoclonal antibody isotyping kit (Biosource International, Camarillo, CA).
The three E-/P-selectin crossreacting antiho~iPc were also tested for their ability to bind to the natural ligands, rather than the recombinant forms used in the initial screening assay_, by single color FACS analysis. The source of natural E-selectin used in these tests was TNF-a-activated human ll~hil;c~l vein endothelial cells (HUVEC). In activated form, ~UVEC cells express E-selectin, but do not express appreciable amounts of P-selectin. Fig. lb shows that the E-/P-cross-reactive antibody 5C7.29 reacts with TNF-a activated H WEC (shown by black histograms) but not unactivated HUVEC (grey histograms). Similar results were obtained for the two other cross-reacting antibodies 2C9.11 and lD8.10. The activated cells also reacted with the anti-E-selectin blocking antibody H18/7 tFig. la) (Becton Dir~inc~n (San Jose, CA)), but not with P-selectin-specific antibodies WAPS 12.2 and 5F4. (WAPS 12.2, a P-selectin blocking antibody, was provided by R. Aaron Warnock and Eugene C. Butcher (Stanford, CA).) The source of natural P-selectin used in these tests was thrombin-activated platelets. Fig. 2b shows that 5C7.29 binds to these cells as does the known P-selectin antibody WAPS 12.2 (Fig. 2a). Similar results were obtained with 2C9.11 and ID8.10. Platelets did not significantly react with anti-E-selectin antiho~ipc H18/7 or lE4.
The E-/P-selectin crossreacting antiho~iPc were further analyzed for binding to L1-2L-~e1eCtin transfe t t and with normal human lymphocytes. Specific binding was not observed, d~ LL-ting that the antibodies are specific for E- and P-selectins and do not bind to L-selectin.
To confirm that the crossreacting an~iho~ipc were truly monoclonal, preclearing experiments were performed. 10 ng antibody (a limiting amount) was incubated with a large 7 f Ll 2E-~eleCtin cells or L1-2P ~ cells fo Wl) gS/34324 2 1 ~ 1 8 7 ~ PC~TIU59~V07302 1 hr. The Cupernatant was then transferred to a second aliquot of Ll-2~-~eleCtin cells or Ll-2P-8eleCt~n cells fthe same cell type ~s before) and incub~ted for 1 hr. Supernatant was transferred to a third aliguot of cells of the same type as before for a further 1 hr incubation. Supernatant was then removed and ~Y~min~ for reactivity with Ll-2~-8eleCtin~
Ll-2F L l~t;~ or Ll-2 untransfected cells by one-color FACS
analysis.
Fig. 3 shows that preincubation of a solution of the 5C7.29 antibody with Ll-2P-5eleCtin ~L~IlsfeuLdnts eliminated ~ ~s~ nt reactivity for both P-selectin and E-selectin.
Similar results were found following preincubation with Ll-2E-8~lectln transfectants. These results would be obtained only if the antibody bound to both selectins, and not if the antibody were a mLxture of two different antibodies, one reactive with E-selectin and one reactive with P-selectin.
Therefore, the dual specificities of 5C7.29 reside in the same antibody. Similar results were obtained for the 2C9.11 and lD8.10 an~ ho~ i ~R .
EY~r~le 3: Inhibition of E-selectin-Mediated Functions The antibody 5C7.29 was tested for the ability to block E-selectin mediated fnnrtinn~. In one assay, the antibody was tested for inhibition of HL-60 binding to tumor necrosis factor-a fTNF-a~ activated human umbilical vein endothelial cells (HUVEC). This binding assay simulates the binding of neuLLophils to endothelial cells in an inflammatory le~u..se. The HL-60 cells are a promyelocytic cell line derived from a patient with acute promyelocytic lel~kPm;~.
Collins et al., ~ature 270, 347-349 (1977). The ~UVEC cells are endothelial cells that when activated with TNF-a for 4-6 hours express E-selectin, and not P-selectin.
HUVEC were obtained fron Clonetics (San Diego, CA) and cultured as suggested. Confluent cultures, up to passage 6, grown in 8 well plastic Lab Tek slides (Nunc, Naperville, IL) were activated for 4 hours with 1 ngJml ~NF-a fR&D Systems, M~nn~rolis, MN). HUVEC cultures were washed and incubated in 0.15 ml Assay Buffer (10% normal bovine ~ ~'O95134324 2 1 g 1 ~ 7 ~ PC~S~5107302 serum/ 10% normal rabbit serum/10 _M HEPES, pH 7.2/RPMI) containing antiho~ire at 17 ~g/ml (i.e., in excess) for 20 min.
HL-60 cells were fluorescently labelled with 6-carboxyfluorescein diacetate acetoxy-methyl ester (CFDA-AM, Molecular Probes, Eugene OR) (von Andrian et al., 1991, supra) by a 30 min incubation in 10 mg/ml RPMI/10 _M HEPES, pH 7.2, washed and roc~lcpon~d in Assay Buffer and incubated at RT for 20 min. The L-~ c~ d cells (6 x 105 cells in 0.15 ml) were then added to the HW EC cultures.
Slides were rotated at 50 rpm on a rotator (Innova 200, New Brunswick Inc.~ for 15 min at RT. The cover slips were removed and non-adherent HL-60 cells washed off by dipping slides in DMEM. Adherent cells were fixed by immersion in 1% paraformaldehyde-PBS. Slides were oYAmi no~
microscopically and the number of bound cells per field det~rminPd. Two treatments per slide (in quadruplicate) were analyzed.
Fig. 4 shows that the number of HL-60 cells binding to the activated HWEC was decreased 47~ by preincubation with 5C7.29. This compared favorably with blocking by the anti-E-selectin-specific antibody H18/7 (38%). Binding was not significantly reduced by a control antibody.
Because H W EC can also express P-selectin (although only at low levels under the present activation conditions), 5C7.29 was also tested for HL-60 binding to CHO cells transfected with E-selectin. CHO cells porr-nontly transfected with a truncated form of E-selectin c~n~ining the first four N-torm;n~l domains of E-selectin fused to the tr~- 'rane and cytopl~ c domain of another protein were ploduced according to standard methods. Expression was confirmed by reactivity with a control anti-E-selectin antibody (H18/7). Inhibition of binding between fluorescently labelled HL-60 and the transfected CHO cells was performed using the same assay as for the TNF-~-activated HW EC. 5C7.29 was found to block adhesion by 82% (Fig. 5). Similar results were observed with lD8.10, 2C9.11 and the E-selectin blocking :; ~
w09s/3~324 2 i~ PCT~S9~07302 antibody lE4. The non-blocking P-selectin specific control antibody 5F4 had no significant effect in this assay.
The cross-reacting a~n~iho~i~c also blocked norDal human peripheral blood n~ ~L~hil binding to TNF-a-activated HUVEC. At a final cvllceil~L~tion of 10 ~g/ml, 5C7.29 blocked 71 +/-13~, 2~9.11 blocked 62 +/-8~ and lD8 blocked 52 +/-10%
of n~uL ~hil binding to activated NUVEC, while the anti-E-selectin antiho~i~c lE4 and H18/7 tBevilacqua et al., 1987, su~ra~ blocked 68+/-4~ and 68 +/-15%, and a control mouse IgGl antibody did not block (-21% +/-11%), n-4. Fox these experi~ents, neutrophils were isolated from normal human blood by density gradient centrifugation and dextran sP~ tation by standard pL~ed~L~s (Curren~ Protocols in Immunol ogy, Coligan et al., eds., John Wiley and Sons, New York, 1992). Assays were peLroL ~ as for HL-60 cells except neutrophils were added to HUVEC at 7.5 X 104 in 0.15 nl.

le 4: Inhihition of P-selectin-Mediated Puncti~nc The antiho~;~R 5C7.29, 2C7.11 and lD8.10 were tested for their ability to block P-selectin-mediated functions.
Blocking was tested in a platelet-HL-60 rosette assay (Corral et al., 1990, supra~. The platelets provide a source of cells expressing P-selectin and the HL-60 cells simulate neuLL~hils. Normal huDan blood was collected with sodium citrate as anticoagulant and the platelet-rich plasma (PRP) prepared by centrifugation at 250g for 10 min. Platelets were isolated fro~ PRP by centrifugation at lOOOg for 20 min and ~ D~ at 3 x 10~/ml in PBS, pH 7.2. Mnnnclnn~l .
antiho~i~c (1 ~g in 20 ~1, i.e., an excess) were added to 20 ~1 platelets. In some experiments normal human throDbin (0.3 U/~l) was added to activate the platelets as descrioed by Corral et al., 1990, supra. After 45 min, 20 ~1 HL-60 cells (1O~/D1 in PBS) were added and further incubated for 45 min.
Bound platelets were fixed to HL-60 cells by addition of glutaraldehyde to 1.25~. At least 100 HL-60 cells for each sample were observed microscopically and the number of cells with bound platelets (>2 platelets per HL-60 cell) det~r~in~.

~ wO~.~t34324 2 1 9 1 ~ 7 ~ PCT~S9S/~7302 Fig. 6 shows that all three crossreacting antibodies block rosetting to about the same extent as the P-selectin specific blocking antibody WAPS 12.2. Similar blocking experiments can be performed using human peripheral blood neuLLophils in place of HL-60 cells. Neutrophils are prepared by the same method and used at the saDe conrentL~tion as described in Example 3.

Exam~le 5: Cloning and s~ qnring of mouse 5C7.29 heavy chain and liqht chain variable reaion cDNA
cDNAs for the heavy chain and light chain variable region genes of the mouse 5C7.29 antibody were cloned using anchored polymerase chain reactions as described (see Co et al., J. Immunol. 148: 1149 (1992)), using 3' primers that hybridized to the constant regions and contained HindIII
sites, and 5' primers that hybridized to the dG tails and contained EcoRI sites. The PCR amplified fL _ Ls were digested with EcoRI and HindIII and cloned into the pUC18 or pUCl9 vectors for sequencing. At least two gamma-l specific and two kappa specific clones were se~u~nced. The gamma-l clones and the kappa clones are respectively identical in se~uence. The variable region cDNA se~l~nr~c and the deduced amino acid sequences for the gamma-l and kappa chains are shown in Fig. 7A-7B [SEQ ID NOS:1-4].

Exam~le 6: Design of hllrqni7~qd 5C7.29 antibody variable domain Based on a ~ u-~re homology search against the NBRF
protein seg~l~nre database, the variable regions of light chain 5nhrlRcc I and heavy chain sl~hclAcc III show good homology to the mouse 5C7.29 antibody. In particular, the antibody III-3R
provides the best fL J~k homology with 5C7.29 and was chosen to provide the fL ..~L~ S~ nrqC for hnrqni7ation of 5C7.29. However, other members of the light chain 5~hcl~q~c I
and heavy chain s~hrlqcc III would also be especially suitable for use in providing the frameworks of the respective hll~-ni7ed 5C7.29 chains.
The computer program ENCAD ~M. Levitt et al., J.
Mol. Biol. 168: 595 (1983)) was used to ~uu~LLu~L a molecular ~O95l34324 2191~ 7 0 PCT~S95/07302 model o~ the 5C7.29 variable domain. The program ABMOD (8.T.
Zilber et al. Biochem. 29:10032-41) is also useful. The model was used to determLne the amino acids in the 5C7.29 framework that were close enough to the CDRs~to potentially interact with them. To design the hn~ni 7ed light and heavy chain 5C7.29 variable regions, the CDRs from the mouse 5C7.29 antibody were grafted into the CL ~rk seyuellces of the III-3R antibody. At ~L ~L~ positions where the model suggested contact with the CDRs, the amino acids from the nouse 5C7.29 antibody were chosen to replace the residues in the III-3R sey ~nre. For humanized 5C7.29, this was done at residues 69 and 70 in the light chain and at no residues in the heavy chain. Moreover, at some positions where the amino acid was unusual for human ~n~iho~i~e at that position, an amino acid Le~Lese-lting a cnncDnellc among the relevant human subclass was substituted for the III-3R CL ~ ~JL~ residue.
For hnr-nized 5C7.29, this was done at residues 61, 72, 8Z and 99 in the light chain and residues 1, 75 and 78 in the heavy chain.
The final se~ of the humanized 5C7.29 heavy and light chain variable region is shown in Figs. 8A-8B ~SEQ ID
NO5:5-8]. However, many of the potential CDR-contact residues are amenable to substitutions of other amino acids and may still allow the antibody to retain substantial affinity to the antigens. The following table lists a number of positions in the rL .L~ where alternative amino acids may be suitable (note LC = light chain, HC = heavy chain):

Position ~nr-nizDd 5C7.29 Alternatives LC-l D O

LC-5g ~ W09~ 2~ 2 1 9 1 8 7 ~ F~

Likewise, many of the tL ~olh residues not contacting the CDRs in the hnr-n;~o~ 5C7.29 heavy and light chains are also amenable to substitutions with amino acids from either the human III-3R antibody, or from the COLL~ .r1;ng position of other human antiho~ or from the mouse 5C7.29 or other mouse antibodies, while still preserving substantial affinity and non-i ~ icity of the humanized antibody. The following table lists a number of positions in the fL ~Lh where alternative amino acids may be suitable:

Position Humanized 5C7.29 Alternatives LC-82 F I,A
LC-99 Q G,S
HC-1 E Q,D
HC-75 S A,P

Finally, even certain residues in the CDRs may be substituted with other residues while the antibody may retain substantially the same affinity and specificity. 5 LL ~ LUL -function studies of antibody binding reveal that not all of the CDR amino acids participate equally in specifying affinity towards a given antigen (or set o~ related antigens). These studies enable prediction with some reliability of particular CDR positions least likely to change substantially the binding characteristics of an antibody. For example, Chothia and co-workers define structurally acceptable amino acids in CDR
positions (Chothia et al., J. Mol. Piol. 196: 902 (1987~;
Chothia et al., Nature 342: 877 (1989); and Tramontano et al., Proteins. Struct. Funct. Genet. 6: 382 (1989)), and many of these are not accessible to solvent (i.e, available to wo~s/343~4 219~ 870 Pcrnls~slo73o~ ~

participate in binding~, in the model of 5C7.29. Other workers have shown that residues 61-66 of CDR H2 may not participate in antigen binding (Ca~ter et al., Proc. Natl.
Acad. Sci. USA 89: 4285 (1992); ~siao et al., Protein Eng.

7:815 (1994)). Surveys of ant~ibody-antigen complex ~LLUCLULCS
support this notion (Glaser et al., ~. Immunol. 1~9: 2606 (1992); Padlan, Mol . I~mu~ol. 31: 169 11994~). Some of these CDR residues that may be changed in hllr-n;7~ 5C7.29 and their potential substitutions are listed in the following table:
CDR Position ~ ni ~e~ Alternatives 5C7.29 Ll 29 V I.L

L2 53 L any 54 A any S T
L3 88 Q N,H
89 Q N,H
H1 34 M I,V,L
H2 61 A any 62 D any 63 T any 64 V A,I,L,M,F
R K,Q
~ 66 G A,D,T,S

Selection of various combinations of alternative amino acids may be used to produce versions of ~llr-ni 7.
5C7.29 that have varying combinations of affinity, specificity, non-i- ja i~ity~ ease of manufacture and other desirable properties. The above exa~ples are offered by way of illustration, not of limitation.

Example 7: C~ Llu~Lion of humanized 5C7.29 For the construction of variable region genes for the humanized 5C7.29 antibody, nucleotide sequences were ~ W095/34324 21~18 7 (} PCT~S9~/07302 selected that encode the protein s~ nrDc of the humanized heavy and light chains, including the signal peptide, generally u~il i7;ng codons found in the mouse se~lpn~e.
Several degenerate codons were changed to create restriction sites or to remove undesirable ones. The nucleotide sol e!lces of the genes also included splice donor signals and an XbaI
site at each end. The nucleotide se~lu~ c and encoded light and heavy chain variable regions of the hn~-n;7ed 5C7.29 antibody are shown in Figs. 8A-8B [SEQ ID NOS:5-8].
Each gene was uus.aLLuuLed from eight overlapping synthetic oligonucleotides. Asse~bly and amplification of the genes were carried out in four steps as shown in Fig. 9:
(1) the four pairs of complementary ~liqnnll~leotides were annealed and extended with Klenow polymerase in separate reactions; (2) the resulting four double-stranded DNA
LL_, ' C were mixed in pairs, denatured, re-annealed and extended in two separate reactions using Klenow fragment;
(3) the resulting two double ~LLanded half-gene fL --ts were mixed, denaLuL~d, re-annealed and extended to create the full length double ~-Landed variable region genes; (4) the gene LL, LS were finally amplified, using Taq polymerase and two primers that hybridize to the 5' and the 3' end of the variable region genes and contain XbaI sites for cloning into the respective expression vectors, pVk and pVg4. Reactions were carried out under conditions well-known in the art.
The pVk vector for light chain expression and the pVgl vector for heavy chain expression have been previously described (see Co et al., J. Im~nol. 148: 1149 (1992)). To produce a hnr-n; 7~ 5C7.29 antibody of the IgG4 isotype, the heavy chain expression vector pVg4 has been constructed. To do so, the XbaI-BamHI f- t of pVgl containing the ~1 constant region was replaced with an approximately 2000 bp fragment of the human ~4 constant region gene (Ellison and Hood, Proc. Natl, Acad. Sci USA 79:1984 (1982)) that extended from the HindIII side preceding the CHl exon of the ~4 gene to 270 bp after the NsiI site following the CH4 exon of the gene, using methods well-known to those skilled in the art, including polymerase chain reaction.

woss~432~ 2 1 9 1 ~ 7 ~ P~ Ir~

The heavy chain and light chain pl~;rl~ were transfected into a mouse myeloma cell line Sp2/0-Agl4 (ATCC
CRL 1581). Transfection was by ele~L.u~oL~tion using a Gene Pulser apparatus (Bio-Rad) at 360 V and 25 uFD capacitance according to the manufacturer's instructions. Before ~ t transfection, the light~chain- and heavy chain-cnnt~ininq plasmids were linearized using PvuII, extracted with phenol-chloroform, and ethanol-precipitated. All transfections were done using 30-50 ~g plasmid DNA and about 107 cells in PBS.
The cell~ from each transfection were plated into 2 to 4 96-well tissue culture plates. After 48 hours, selective medium was applied.
Cells were selected for gpt expression in DME~ ~ 10 FBS + HT media supplement ~Sigma) + 1 ~g/ml mycophenolic acid.
Antibody-produciny clones were screened by assaying human antibody production in the culture supernatant by ELISA.
Antibody from the best-producing clones was purified by passing tissue culture supernatant over a column of protein A-Sepharose CL-4B (PhAr~ ). The bound antiho~;Pc were eluted with 0.2 N glycine-HCl, pH 3.0, and neutralized with 1 M Tris-HCl, pH ~Ø The buffer was ovrh~ngoA into phosphate buffered saline (Pi3S~ by passing over a PD10 column ~Phlrr~ ), or by dialysis. To obtain cells producing higher levels of antibody, the tr~nqfect~ clones may be cultured in increasing cvncenL~tions of methotrexate.

Exam~le 8: Pro~erties of humanized 5C7.29 To show that hnr-ni~fl 5C7.29 specifically binds to E-selectin and P-selectin, L1_2E-~electin and Ll_2P-selectin transfectants were incubated with humanized 5C7.29 or control antiho~ip~ for 1 hour. After washing, cells were incubated in a 1:400 dilution of pl.y~oeLy-hrin-conjugated anti-human Ig (Biosource, Camarillo, CA), washed, then analyzed for fluorescence by flow cytometry (FACS) as previously described (Berg et al., Bl~od 85: 31 (1995)). Ullr-ni7ed 5C7.29 reacts ith b th Ll 2E-selectin and L1-2P-seleCtin transfectantS~ but not L1_2L-seleCtin transfectants (Fi5- 10) ~ LL~ting that the h-lr-nization process did not result in loss of either w09~l3432~ 21 91 ~7n E-selectin or P-selectin binding or gain in the ability to bind L-selectin.
The affinity of the humanized 5C7.29 antibody for E-selectin and P-selectin was separately det~rmined by competition with the radio-iodinated mouse 5C7.29 antibody (Fig. 11). Purified mouse 5C7.29 antibody was labeled with Nal25I (Amersham, Arlington Heights, IL) using the lactoperoxidase ~LuueduLa to 4 mCi/mg of protein.
cHOE-selectin cells and 11_2P-selectin cells which were obtained by transfecting the E-selectin and P-selectin genes into the respective host cells CH0 and Ll-2 (Gallatin et al., Nature 304:30 (1983)) by methods well known in the art (see, e.g., Larsen et al., J. Biol. Chem. 267: 11104 (1992)), were used as sources for E-selectin and P-selectin. Increasing amounts of competitor antibody (mouse 5C7.29 or humanized 5C7.29) were added to 2 ng of radio-iodinated tracer mouse 5C7.29 antibody and incubated with 4 X 105 CHOE~seleCein cells or 11_2P-selectin cells in 0.2 ml of binding buffer (PBS with 2~ fetal calf serum, 0.1~ sodium azide) for 2 hours at 4- C with constant shaking. Cells were then washed and centrifuged, and their radioactivities r--- ed. The ratio of bound and free tracer antibody were calculated (Figs. llA and llB).
The binding affinities were calculated according to the methods of 8erzofsky and Berkower (J. A. Berzofsky and I.
J. Berkower, in Fundamental Il ~logy (ed. W.E. Paul), Raven Press (New York), p. 595 (1984)). The humanized 5C7.29 had an affinity of approximately 3 x 108 M 1 for E-selectin, with no significant difference from that of mouse 5C7.29, and an affinity of approximately 1.3 X 108 M~l for P-selectin, within about 3 to 4-fold of the mouse 5C7.29 antibody. This experiment also shows directly that humanized 5C7.29 has the ability to compete with the mouse 5C7.29 antibody for binding to both E-selectin and P-selectin. In another experiment, the affinities of mouse and humanized 5C7.29 for P-selectin were determined by the method of Scatchard (Berzofsky and Berkower, supra) to be approximately 1.9 x 108 M~1 and 6 x 1o8 M-1, respectively, in good agreement with the results of the competitive binding experiment.

w09~343~ 2 1 g 1 ~ 7 0 48 PCT~S95107302 To show that the hll7--ni~,7 5C7.29 antibody inhiblts binding of E-selectin to a counter-receptor for E-selectin, its ability to inhibit the binding of HL-60 cells to E-selectin transfectant cells was det~ 7 n~d . Assays o~ the ~.'h~7ion of HL-60 cells wi~h~CH0~-seleCtin cells were performed as previously described t~erg et al., Blood 85: 31 (1995), and supra) in the pl~sen~e of monoclonal antibodies at the indicated ~ncD l,~tions. Fig. 12 shows that hn~qnized 5C7.29 blocks binding of HL-60 cells to CHO~ eleCtin transfectants as well as mouse 5C7.29. For the representative experiment shown, two treatments per slide (each treatment in quadruplicate) were analyzed and the mean and standard deviations calculated. An isu-y~e ~-tched control antibody did not affect binding.
To show that the h-lr-ni~' 5C7.29 antibody inhibits binding of P-selectin to a counter-receptor for P-selectin, its ability to inhibit the binding of HL-60 cells and activated platelets was determined. Assays of the rosetting of activated platelets to the HL-60 cells were pelf~_ ~d as described (Berg et al., Blood 85:31 (1995) and supra) in the presence of monoclonal anti7Odies at the indicated conc~nLL~t~ons. Fig. 13 shows that hl7~-ni~ 5C7.29 blocks binding of platelets to HL-60 cells as well as mouse 5C7.29.
An isoLy~e --tched control antibody had no affect on binding in this assay. The rep~se--Lative experiment shown was p~Lfo ed in triplicate and the mean and standard deviations calculated.

~m~le 9: E~ito~e mao~oinG of 5C7.29 To determine the amino acids of E-selectin involved in the binding of 5C7.29 (the epitope), the following ~L~cedul~ was used. DNA ~nror7ing the lectin and EGF-like domains of human E-selectin were fused to a gene ~n~or7. i ng the human immunoglobulin lambda constant region (C~), which served as a tag. The chimeric DNA was inserted in a plasmid vector, which provided a lac promoter and pelB signal sequence for expression and secretion of the chimeric ~fusion) protein in E. col s . The E-selectin domains were randomly murag~ni~d by ~ w09~3~24 2191 ~ 7 ~ PCT~S9~/07302 eL.vr-p~one polymerase chain reaction (PCR~ utilizing AmpliTaq enzyme (Perkin Elmer) and Nn++, and the amino acid substitutions were determined by DNA s~ nrin7. E. col i strain TGl~recA was transformed with the wild-type and mutant plasmids, and chimeric proteins were ~va~b~ ssed by growing transformed E. co2i in 2YT broth. After 8 hours of induction with lmN IPTG, culture ~u~bLIla~nts containing the chimeric proteins were collected. All operations were performed according to methods well-known in the art of molecular biology.
Next, 96-well plates were coated with the 5C7.29 antibody (or control antibodies). After blocking, the plates were incubated with the E. col i supernatants and then with HRP-conjugated anti-human C; ant;hoAipc (Biosource, Camarillo, CA). After washing, bound enzyme was detected with TMB
substrate. Supernatants containing mutant E-selectin-CA
chimeric protein to which 5C7.29 could still bind gave a positive signal, while supernatants containing mutant E-selectin to which 5C7.29 could not bind gave a negative signal. The results are shown in the following table, where the symbol AXB means a mutant in which the Xth amino of E-selectin form the mature N-t~minnq is changed from the normal A to mutant B.

- Nutant Reactivity Q2lR

Y48H +
E92G +
N105S +
~lllE +

ulo9sl34324 2 ~ 9 ~ 8 7 ~ pCT~9S/07302 ~

Becaus- mutatlng amino acids Q21~ R22~ Y23~ Tl19 and A120 in E-selectin prevented binding of antibody 5c7.29, these amino acids must be in the epitope; of 5C7.29. The full amino acid s~ ro of E-selectin is given in Bevilacqua, s~pra and in ~nited States Patent 5,27i,263 (ELA~-l). (Another anti-E-selectin antibody was able to bind to these mutants, showing that they did not disrupt the overall ~LU~Ui~ of E-selectin~. Other E/P cross-reacting antibodies that show a different pattern of reactivity with these E-selectin mutants nu6t have a different epitope in E-selectin. The epitope of 5C7.29 in P-selectin may be determined by a sinilar pluced~l~
using P-selectin mutants, and may be similarly compared to the epitope of other E/P cross-reacting antihv~i-c~ The epitopes of 5C7.2g in E-selectin and P-selectin are preferred epitopes, because antihoSi~c such as 5c7.2s that bind to them may have high affinity and blocking activity.
While the foregoing invention has been described in some detail for ~UL~OSeS of clarity and understanding, it will be clear to one skilled in the art from a reading of this disclosure that various changes in form and detail can be made without departing from the true scope of the invention. All publications and patent ~-_ ts cited in this application are inc~L~oL~ted by reference in their entirety for all purposes to the same extent as if each individual publication or patent fl~ L were so individually denoted.

~ W 095/34324 2191~ 7 Q PCT~US95/07302 SEQUENCE L}STING

(1~ GENERAL INFORMATION:
(i) APPLICANT: 3erg, Ellen L.
(ii~ TITLE OF INVBNTION: Cross-Reacting Mr,nrrlrn~l 3nt1ho~i~c Specific for E-SeleCtin and P-Selectin ~iii) NU~3ER OF SEQUENCES: 10 ~iV) ~,:~Kt~ J~l~C!!; ADDRESS:
Al ~n~;FcRRR Townsend and Townsend Xhourie and Crew B 8TR:ET: One Uarket Plaza, Steuart Tower, Suite 2000 C CI r : San Francisco D .T~TE: California E rouNTRy: USA
F ~IP: 94105 ~v) CO~L'UTER READABLE FOKM:
(A~ MBDIUM TYPE: Floppy disk (B COMPUTER: IBM PC compatible (C OPERATING SYSTEM: PC-DOS/MS-DOS
(Dl SOFTWARE: PatentIn Release #1.0, Version #1.25 (vi) CURRENT APPLICATION DATA:
~A) APPLICATION NU~BER: WO
~B) FILING DATE:
(C) CLASSIFICATION:
(vii) PRIOR APPLICATION DATA-~A) APPLICATION NUMBER: 08/259,963 ~B) FILING DATB: 14-~UNE-94 ~viii) ATTORNEY/AGENT INFORMATION:
~A) NAME: Smith, William M.
~B) REGISTRATION NU~SER: 30,223 ~C) REFERENCE/DOCXET NU~13ER: 11823-005810PC
(ix) TRnRroMMTT~TcATIoN INFORMATION:
~A) TELEPHONE: 415-326-2400 (B) TELEFAX: 415-326-2422 W O9~l34324 219187 ~ 52 PCTnlS9Cm7302 ~2) INPORMATION FOR SEO ID NO:1:
(i) SEQ-JEiCE r~ TR~rCTICS:
A iE~GTH: 384 base pairs B '~E nucleic acid ~ ' CI -T~ r~cc: single . ;~ ' ID TO?OLOGY: lineer (ii) MOLECULE TYPE: cDNA

~ix) FEATURE:
~A) NAME/REY: CDS
~B) L~)CATION: 1..384 (xi) SEQUENCE DESCRIPTION: SE~ ID NO:1:

Met Asp Phe Gln Val Gln Ile Phe Ser Phe Leu Leu Ile Ser Ala Ser Val Ile Ile Ser Arg Gly Gln Ile Val Leu Thr Gln Ser Pro Ala Ile 2~ 25 30 Met Ser Al~ Ser Pro Gly Glu Lys Val Thr Met Thr Cys Ser Ala Ser Ser Ser Val Pro Tyr Met His Trp Tyr Gln Gln Lys Ser Gly Thr Ser Pro Lys Leu Trp Ile Tyr Asp Thr Ser Asn Leu Ala Ser Gly Val Pro Ala Arg Phe Ser Gly ser Gly Ser Gly Thr Ser Tyr Ser Leu Thr Ile 85 90 g5 Ser Ser Met Glu Ala Glu Asp Ala A1~ Thr Tyr Tyr Cys Gln Gln Trp AGT AGT GAC CCA Trrc ACG TTC GGC TCG GGG ACA AAG TTG GAA ATA AAG 384 Ser Ser Asp Pro Phe Thr Phe Gly Ser Gly Thr Lys Leu Glu Ile ~ys ~ W 095~4324 21918 7 q PCT~vS95~7302 (2) }NFORMATION FOR SEQ ID NO:2:
~i) SEQUENCE C~ARACTERISTICS:
(A) LBNGTH: 128 amino acids (B) TYPE- ~mino acid (D~ TOPOLOGY: linear (ii) MOLECULE TYPE: protein (xi) SEQUE~CE V~UKle~UN: SEQ ID NO:2:
Net Asp Phe Gln Val Gln Ile Phe Ser Phe Leu Leu Ile Ser Ala Ser Val Ile Ile Ser Arg Gly Gln Ile Val Leu Thr Gln Ser Pro Ala Ile Met Ser Ala Ser Pro Gly Glu Lys Val Thr Met Thr Cys Ser Ala Ser Ser Ser Val Pro Tyr Net His Trp Tyr Gln Gln Lys Ser Gly Thr Ser Pro Lys Leu Trp Ile Tyr Asp Thr Ser Asn Leu Ala Ser Gly Val Pro Ala Arg Phe Ser Gly Ser Gly Ser Gly Thr Ser Tyr Ser Leu Thr Ile Ser Ser Met Glu Ala Glu Asp Ala Ala Thr Tyr Tyr Cys Gln Gln Trp Ser Ser Asp Pro Phe Thr Phe Gly Ser Gly Thr Lys Leu Glu Ile Lys W095l34324 2 1 9 1 8 7 ~ ,2 l2) INFORMATION FOR SEQ ID NO:3:
li~ SEOUENCE cHARA~ s ~A .ENGTH: 405 base pairs ~~, ~B AYPE: nucleic acid ~' C ~ single~
~D ~:OPOLOGY: linear lii) MOLECULE TYPE: cDNA

lix) FEATURE:
IA~ NAME/~EY CDS
(B~ LOCATION 1..405 (xi) SEQUENCE ~hl511~N: SEQ ID NO:3:

Met Asp Ser Arg Leu Asn Leu Val Phe Leu Val Leu Ile Leu Lys Gly Val Gln Cys Asp Val Arg Leu Val Glu Ser Gly Gly Gly Leu Val Gln Pro Gly Gly Ser Arg Lys Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Ser Phe Gly Met His Trp Val Arg Gln Ala Pro Asp Lys Gly Leu Glu Trp Val Ala Phe Ile Ser Ser aly Ser Ser Thr Ile Tyr Tyr Ala 65 7C 75 ~0 Asp Thr Val Arg Gly Ar~ Phe Thr Ile Ser Arg Asp Ser Pro Lys Asn 85 g0 95 Thr Leu Phe Leu Gln Met Thr Ser Leu Arg Ser Glu Asp Thr Ale Met Tyr Tyr Cys Ala Arg Pro Leu Pro Pro Phe Ala Tyr Trp Gly Gln Gly Thr Leu Val Thr Val Ser Ala ~ wog5,34324 219 1 8 7 Q PCT/US9sm7302 (2) INFORMATION FOR SEQ ID NO:4:
~i) SEQUENCE CHARACTERISTICS-IA) LENGTH- 135 amino acids iB) TYPB: amino zcid (D) TOPOLOGY: linear (ii) MOLECULE TYPE: protein (xi) SEQUENCE ~SL~I~IsL~N: SEQ ID NO:4:
Met Asp Ser Arg Leu Asn Leu Val Phe Leu Val Leu Ile Leu Lys Gly Val Gln Cys Asp Vdl Arg Leu Val Glu Ser Gly Gly Gly Leu Val Gln Pro Gly Gly Ser Arg Lys Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Ser Phe Gly Met Hi8 Trp Val Arg Gln Ala Pro Asp Lys Gly Leu Glu Trp Val Ala Phe Ile Ser Ser Gly Ser Ser Thr Ile Tyr Tyr Ala Asp Thr Val Arg Gly Arg Phe Thr Ile Ser Arg Asp Ser Pro Lys Asn Thr Leu Phe Leu Gln Met Thr Ser Leu Arg Ser Glu Asp Thr Ala Met Tyr Tyr Cys Ala Arg Pro Leu Pro Pro Phe Ala Tyr Trp Gly Gln Gly Thr Leu Val Thr Val Ser Ala W 095~34324 2 1915 7 0 PCTAUS9'3~7302 ~2) INFOF~.sATION FOF SEQ ID NO:5:
(i) SE:flUE~SCE t~S~ t'T~T CTICS:
NGTRs 384 bzse pairs - 5 PB: nucleic acld ~ S
sl ~Tl~ar~qFsn~ss2cc: slngle P P~POLOGY: linear (ii1 MOLBCssLB TYPE: cDNA

(ix1 FEATURE:
(A) NAMB/?CEY: CDS
(?3~ LOCATIO~: 1..384 (xil SEQUENCE DESC~IPTION: SEQ ID NO:5:

Met Asp Phe Gln Val Gln Ile Phe Ser Phe Leu Leu Ile Ser Ala Ser GTC ATA ATA TCC As3A GGA GAT ATT CAA ATG ACC CAG TCT CCA TCT AGC 96 Vel Ile Ile Ser Arg Gly Aap Ile Gln Met Thr Gln Ser Prss Ser Ser Leu Ser Ald Ser Val Gly Asp Arg Yal Thr rle Thr Cya Ser Ala Ser Ser Ser Val Pro Tyr ~.et Eis Trp Tyr Gln Gln Lys Pro Gly Lya Ala Pro Lys Leu Leu Ile Tyr A~p Thr Ser Asn Leu Al~ Ser Gly Val Pro 65 70 '75 80 Ser Arg Phe Ser Gly Ser Gly Ser Gly Thr Ser Tyr Thr Leu Thr Ile 85 90 g5 AGC AGC CTG CAG CCT G~A GAT m GCC ACT TAT TAC TGC CAG CAG TGG 336 Ser Sor Leu Gln Pro Glu Asp Phe Ald Thr Tyr Tyr Cys Gln Gln Trp Ser Ser Asp Pro Phe Thr Phe Gly Gln Gly Thr Lys Val Glu Ile Lys ~ W O95/34324 2t~87~ pCTrL~S9510~302 (2~ INFORMATION FOR SEQ ID NO:6:
(i) SEQUENCB CHARACTERISTICS:
(A) LENGTH: 12~ amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear (ii) MOLECULE TYPE: protein (xi) SEQUENCE J~Kla-lON: SEQ ID NO:6:
Met Asp Phe Gln Val Gln Ile Phe Ser Phe Leu Leu Ile Ser Ala Ser 1 5 10 ~5 Val Ile Ile Ser Arg Gly Asp Ile Gln Met Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val Gly Asp Arg Val Thr Ile Thr Cy5 Ser Ala Ser ~0 ~5 Ser Ser Val Pro Tyr Met Hig Trp Tyr Gln Gln Lys Pro Gly Lys Ala Pro Lys Leu Leu Ile Tyr Asp Thr Ser Asn Leu Ala Ser Gly Val Pro 7s 80 Ser Arg Phe Ser Gly Ser Gly Ser Gly Thr Ser Tyr Thr Leu Thr Ile Ser Ser Leu Gln Pro Glu Asp Phe Ala Thr Tyr Tyr Cys Gln Gln Trp Ser Ser Asp Pro Phe Thr Phe Gly Gln Gly Thr Lys Val Glu Ile Lys W095/34324 2 19 18 7 0 5& PCTrUS95~7302 (2) INFORMATION FOR SB0 ID NO:7:
li) sEQrIENcE C~ARACTERISTICC~
A .ENGTH: 405 b2se pairs B Y PE: nucleic acid C I T~r~Rn~Rcq: single ~D 'OPOLOGY: linear tii~ MOLECULE TYPE: cD~A

(ix~ FEATURE:
~A) NAMEJREy: CDS
(B) LOCATION: 1..405 Ixi1 SEQUENCE DESCRIPTION: SEQ ID NO:7:

Met Asp Ser Arg Leu Asn Leu Val Phe Leu Val Leu Ile Leu Lys Gly Val Gln C'ys Glu Val Gln Leu Val Glu Ser Gly Gly Gly Leu Val Gln CCT GGA GGG TC'C CTT CGT CTC TCC TGT GCA GCC TCT GGA TTC ACT TTC 144 Pro Gly Gly Ser Leu Arg Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Ser Phe Gly Met His Trp Val Arg Gln Ala Pro Gly Lys Gly Leu GAG TGG GTC GCA TTC ATT AGC AGT GGC AGT AGT ACC ATC TAC TAT GCT 2~0 G1~ Trp Val Ala Phe Ile Ser Ser Gly Ser Ser Thr Ile Tyr Tyr Ala Asp Thr Val Arg Gly Arg Phe Thr Ile Ser Arg Asp Asn Ser Lys Asn Thr Leu Tyr Leu Gln Met Asn Ser Leu Arg Ala Glu Asp Thr Alu Val Tyr Tyr Cys Ala Arg Pro Leu Pro Pro Phe Ala Tyr l~p Gly Gln Gly Thr Leu Val Thr Val Ser Ala ~ WO ~5134324 2191 & 7 ~ PCTIUS95l07302 (2) INFORMATION FOR SEQ ID NO:8:
(i) SEQUENCE CHARACTERISTICS:
~A) LENGTH: 135 amino acids (B) TYPE- amino acid (D1 TOPOLOGY: linear (ii) MOLECULE TYPE: protein (xi) SEQUENCE DESCRIPTION: SEQ ID NO:8:
Met Asp Ser Arg Leu Asn Leu Val Phe Leu Val Leu Ile Leu Lys Gly Val Gln Cys Glu Val Gln Leu Val Glu Ser Gly Gly Gly Leu Val Gln Pro Gly Gly Ser Leu Arg Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Ser Phe Gly Met ~iB Trp Val Arg Gln Ala Pro Asp Lys Gly Leu Glu Trp Val Ala Phe Ile Ser Ser Gly Ser Ser Thr Ile Tyr Tyr Ala Asp Thr Val Arg Gly Arg Phe Thr Ile Ser Arg Asp Asn Ser Lys Asn Thr Leu Tyr Leu Gln Met Asn Ser Leu Arg Ala Glu Asp Thr Ala Val Tyr Tyr Cys Ala Arg Pro Leu Pro Pro Phe Ala Tyr Trp Gly Gln Gly 115 120 . 125 Thr Leu Val Thr Val Ser Ala W 09~13~32J 2 1 ~ 1 ~ 7 0 P~'T1US95107302 (2~ INFORMATION FOR SEO ID NO:9:
~i~ SE~UE~CE CHABACT RIcTICS:
A ENGTH: l06 amino acids B I - YPE ~ u~ino acid C ~r~r~rn~F.q.C: single o 1D 'OPOLOGY: linear ,~
(ii) MDl~EcurE TYPE: peptide ~i (ix) FBA)TUNE~E/~cEy Region (3) LOCATION: one-of(l) (D) OTHER INFORMATION: /note. ~Xaa is Asp or Gln.
(ix) FBATURE:
(A) NA~E/X~Y: Region (B) LOCATION- one-of(2) (D) OTHER INFORMATION: /note= ~Xaa is Gln or Val.
(ix) FBATURk (B) LOCAT}ON; one-of(4,32) tD) OTHBR INFORMATION: /note= ~Xaa is Met or Leu.
lix~ FBATURB:
(A) NAME/KEY: Region lB) LOCATION- one-ofl29) lD) OTHER INFORMATION: /note- ~Xaa i9 Val, Leu or Ile.
~iX) FB~A)TURNAME/KBy. Region (B) LOCATION one-of(53,54) (D) OT~ER INFORMATION: /note- rxaa is any amino acid.
(iX) FE~A)TUNAME~KBY- Region tB) LOCATION one-of(55) ~D) OT~ER INFORMATION: /note- rxaa is Ser or Thr.
( ix~
lA) NAMB/KBY: Region lB) LOCATION one-of(59) lD) OTHER INFORMATION: jnote. rXaa is Ser or Ala.
(ix) FBATURE:
(A) NAMB/KEY: Region (B) LOCATION: one-ofl6l) ~D) UrHER INFORMATION: /note- ~Xaa is Phe or Ile.

~ wo g5/34324 2 1 9 1 ~ 7 Q PCrrusg5/073O2 ~ix) FEATURE:
iA) NAME/XEY: Region (B) LOCATION: one-of(69) (D) OTHER lNru~ ù~: /note- ~Xaa is Ser or Asp.' (ix) FEATURE:
(A) NAME~KEY: Region (B) LOCATION: one-of(70) (D) OTHER l~ru.8_.TluN: /noteS ~X2a i5 Tyr or Phe.' (ix~ FEATURE:
(A) NAME/KEY: Region (B) LOCATION: one-of(72) (D) OTHER INFORMATION: /note- ~Xaa is Leu or Phe. n lix) FEATURE:
(A) NAME/KEY: Reglon (B) LOCATION: one-of(82) (D) OTHER INFORMATION: /note= rXaa is Phe or Ile.
(ix) PEATURE:
(A) NAME/KEY: Region (B) LOCATION: one-o~(88,89) (D) OTHER INFORMATION: /note- 'Xaa is Gln, Asn or His.' ~ix) FEATURE:
(A) NAME/KEY: Region (B) LOCATION: one-of(99) (D) OTHER INFORMATION: /note~ ~Xaa is Gln, Gly or Ser.' (xi) SEQUENCE DESCRIFTION: SEO ID NO:9:
Xaa Xaa Ile Xaa Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val Gly Asp Arg Val Thr Ile Thr Cy8 Ser Ala Ser Ser Ser Xa2 Pro Tyr Xa2 His Trp Tyr Gln Gln hys Pro Gly Ly8 A12 Pro Dys Leu Leu Ile Tyr Asp Thr Ser Asn X2a Xaa Xaa Gly Val Pro Xaa Arg Xaa Ser Gly Ser Gly Ser Gly Thr Xaa Xa2 Thr X2a Thr Ile Ser Ser Leu Gln Pro Glu Asp Xa2 A12 Thr Tyr Tyr Cys Xa2 Xa2 Trp Ser Ser Asp Pro Phe Thr Phe Gly Xa2 Gly Thr Lys Val Glu Ile Lys YVO9~34324 219 il S 7 n 62 Pt-r!~T'T9~/07302 ~21 INFORMATIOU FOR SEQ ID NO:10:
~i) S~OJ~IC~ CHARA~ l b L 1 ~:
A ~E~TGTH: 116 amino acid~
B TY 7E: amino acid C ;T~'~rnNRCC single D~ TO?OLOGY: linear "
(ii) MOLEt'LTLE TYPE: peptlde~

(ix) FEATURE:
(A) NAME/XEY: Region (B) LOCATION: one-of~l) ~D) C)THER INPORMATION: /note~ ~Xaa is Glu, Gln or Asp.r (ix) FEATURE:
(A) ~AME~~ Y: Region (B) LOCATION: one-of~34) ~D~ OTHER INFORMATION: /note- ~Xaa i8 Met, Ile, Val or Leu."
~ix) FEATURE:
(A) NAME/REY: Region (B~ LOCATION: one-of~49 ~D) t)THER INFORMATIO~: /note~ ~Xaa is Ala or Ser.
lix) FEATURE:
~A) NAME/XEY: Region (B) LOCATIaN: one-o~(61,62,63) ID) OTHER INFORMATION: /note~ ~Xaa is any amlno acid.' (ix) FEATURB:
(A) NAME/XEY: Region (B) LOCATION: one-o~(64) (D) OTHER INFOR~ATION: /note, ~Xaa is Val, Ala, Ile, Leu or Met.
(ix) FFATURE:
(A) NAME~REY: Region ~B) LOCATION: one-of~65) ~D) OTHER INFORMATION: /note= ~Xaa is Arg, Lys or Gln.
~iX) FEATURB:
~A~ I~UME/REY: Region tB) LOCATION: one-of(66~
~D) OTHER lN~ ~ lU~: /note= 'Xaa i9 Gly, Ala, Asp, Thr or Ser.
(ix) FEATLTRE:
~A) NAME/~EY: Region (B) LOCATION: one-of(75) ~D) OT~ER INFORMATION: /note, 'Xaa is Ser, Ala or Pro.

~ W O95/34324 2191~ 7 0 PCT~USg5/07302 (ix) FEATURE:
IA) NAME/XEY: Region (B) LOCATION: one-o~(78) (D) OTHER lNr'O~_.'llUN: /note= ~Xaa i9 Thr or Ser.
(ix) FEATURE:
(A) NAME/XEY: Region (B~ LOCATION: one-o~(84) ~D) OTHER INFO~ATION: /note- "Xaa is Asn or Thr. r (ix) FEATURE:
(A) h-AME~XEY Region (B) LOCATION one-o~(116) (D) OTHER INFORMATION: /note= ~Xaa i9 Ala or Ser.

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:10:
Xaa Yal Gln Leu Val Glu Ser Gly Gly Gly Leu Val Gln Pro Gly Gly Ser Leu Arg Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Ser Phe Gly Xaa His Trp Val Arg Gln Ala Pro Gly Lys Gly Leu Glu Trp Val Xaa Phe Ile Ser Ser Gly Ser Ser Thr Ile Tyr Tyr Xaa Xaa Xaa Xaa Xaa Xaa Arg Phe Ile Ile Ser Arg Asp Asn Xaa Lys Asn Xaa Leu Tyr Leu Gln Met Xaa Ser Leu Arg Ala Glu Asp Thr Ala Val Tyr Tyr Cys . g0 95 Ala Arg Pro Leu Pro Pro Phe Ala Tyr Trp Gly Gln Gly Thr Leu Val Thr Val Ser Xaa

Claims (33)

WHAT IS CLAIMED IS:

1. A monoclonal antibody having a binding site that specifically binds to P-selectin and to E-selectin, said antibody having an affinity for each of P-selectin and E-selectin of at least 10 8 M-1.

2. The antibody of claim 1, wherein the specific binding of the antibody to the P-selectin inhibits binding of the P-selectin to a counterreceptor of P-selectin; and the specific binding of the antibody to the E-selectin inhibits binding of the E-selectin to a counterreceptor of E-selectin.

3. The antibody of claim 2, wherein the counterreceptors are expressed on an HL-60 cell or a neutrophil.

4. The antibody of claim 2 that competes with antibody 5C7.29, ATCC accession number CRL 11640, for specific binding to P-selectin and to E-selectin.

5. The antibody of claim 2 that is a mouse antibody.

6. The antibody of claim 2 that is monoclonal antibody 5C7.29, ATCC accession number CRL 11640.

7. The antibody of claim 2 that is a Fab, Fab', F(ab')2, Fv fragment, or a single-chain antibody.

8. The antibody of claim 2 that is a human antibody.

9. The antibody of claim 1 that does not specifically bind to L-selectin.

10. The antibody of claim 1 that specifically binds to L-selectin.

11. The antibody of claim 1 that recognizes an epitope of E-selectin comprising amino acids Q21, R22, Y23, T119, and A120.

12. A humanized antibody that specifically binds to P-selectin and inhibits the binding of the P-selectin to a counterreceptor of P-selectin; and that specifically binds to E-selectin and inhibits the binding of the E-selectin to a counterreceptor of E-selectin, said antibody comprising a humanized light chain variable region and a humanized heavy chain variable region wherein (1) the humanized light chain variable region comprises complementarity determining regions having amino acid sequences from a non-human antibody light chain and comprises a variable region framework sequence substantially identical to a human light chain variable region framework sequence; and (2) the humanized heavy chain variable region comprises complementarity determining regions having amino acid sequences from a non-human antibody heavy chain, and comprises a variable region framework sequence substantially identical to a human heavy chain variable region framework sequence.

13. The humanized antibody of claim 12 wherein the humanized light chain variable region has a sequence substantially identical to the sequence:
DIQMTQSPSS LSASVGDRVT ITCSASSSVP YMHWYQQRPG
KAPKLLIYDT SNLASGVPSR FSGSGSGTSY TLTISSLQPE
DFATYYCQQW SSDPFTFGQG TKVEIK [SEQ ID NO:6]
and the humanized heavy chain variable region has a sequence substantially identical to the sequence:
EVQLVESGGG LVQPGGSLRL SCAASGFTFS SFGMHWVRQA
PGKGLEWVAF ISSGSSTIYY ADTVRGRFTI SRDNSKNTLY
LQMNSLRAED TAVYYCARPL PPFAYWGQGT LVTVSA
[SEQ ID NO:8].

14. The humanized antibody of claim 13 wherein (a) the humanized light chain variable region has the sequence:

DX9ATYYCX16X17W SSDPFTFGX10G TKVEIK [SEQ ID NO:9], wherein X1 = D or Q; X2 = Q or V; X3 = M or L; X4 = S or A; X5 = S or D;
X6 = Y or F; X7 = F or I; X8 = L or F; X9 = F, I or A; X10 = Q, G or S; X11 = V, I or L; X12 = M or L; X13 = any amino acid;
X14 = any amino acid; X15 = S or T; X16 = Q, N or H; and X17 =
Q, N or H; and (b) the humanized heavy chain variable region has the sequence: X3VQLVESGGG LVQPGGSLRL SCAASGFTFS

SRDNX4KNX5LY LQMX2SLRAED TAVYYCARPL PPFAYWGQGT LVTVSX6 [SEQ
ID NO:10]; wherein, X1 = A or S; X2 = N or T; X3 = E, Q or D;
X4 = S, A or P; X5 = T or S; X6 = A or S; X7 = M, I, V or L; X8 = any amino acid; X9 = any amino acid; X10 = any amino acid;
X11 = V, A, I, L, M or F; X12 = R, K or Q; and X13 = G, A, D, T
or S.

15. The humanized antibody of claim 13 wherein in the humanized light chain variable region, X11 = V; X12 = M;
X13 = L; X14 = A; X15 = S; X16 = Q; and X17 = Q; and wherein in the humanized heavy chain variable region, X7 = M; X8 = A; X9 = D; X10 = T; X11 = V; X12 = R; and X13 = G.

16. The humanized antibody of claim 13 wherein the humanized light chain variable region has the sequence:
DIQMTQSPSS LSASVGDRVT ITCSASSSVP YMHWYQQKPG
KAPKLLIYDT SNLASGVPSR FSGSGSGTSY TLTISSLQPE
DFATYYCQQW SSDPFTFGQG TKVEIK [SEQ ID NO:6]
and the humanized heavy chain variable region has the sequence:
EVQLVESGGG LVQPGGSLRL SCAASGFTFS SFGMHWVRQA
PGKGLEWVAF ISSGSSTIYY ADTVRGRFTI SRDNSKNTLY
LQMNSLRAED TAVYYCARPL PPFAYWGQGT LVTVSA
[SEQ ID NO:8].

17. The humanized antibody of claim 12 further comprising light chain and heavy chain constant regions substantially identical to human light chain and heavy chain constant regions.

18. A purified nucleic acid segment encoding a light or heavy chain variable region of the antibody of claim 2.

19. A purified nucleic acid segment encoding a light or heavy chain variable region of the antibody of claim 12.

20. The purified nucleic acid segment of claim 19 further comprising a light chain or heavy chain constant region substantially identical to a human light chain or heavy chain constant region.

21. A stable cell line comprising:
a nucleic acid segment encoding the heavy chain of the antibody of claim 2, the segment operably linked to a first promoter to allow expression of the heavy chain;
a second nucleic acid segment encoding the light chain of the antibody of claim 2, the second segment operably linked to a second promoter to allow expression of the light chain;
wherein the stable cell line can produce the antibody of claim 2.

22. A stable cell line comprising:
a nucleic acid segment encoding the heavy chain of the antibody of claim 12, the segment operably linked to a first promoter to allow expression of the heavy chain;
a second nucleic acid segment encoding the light chain of the antibody of claim 12, the second segment operably linked to a second promoter to allow expression of the light chain;
wherein the stable cell line can produce the antibody of claim 12.

23. A pharmaceutical composition comprising the monoclonal antibody of claim 2.

24. A pharmaceutical composition comprising the monoclonal antibody of claim 12.

25. A method of treating an inflammatory disease or condition, comprising administering to a human patient a therapeutically effective dose of the pharmaceutical composition of claim 12.

26. A method according to claim 25, wherein the inflammatory disease or condition is selected from the group consisting of ischemia-reperfusion injury, adult respiratory distress syndrome, trauma, stroke, sepsis, psoriasis, and autoimmune disease.

27. The method of claim 26 wherein the inflammatory disease or condition is ischemia-reperfusion injury after myocardial infarction or stroke.

28. The method of claim 27 further comprising the step of administering a therapeutically effective dose of a thrombolytic agent.

29. A method of generating an antibody capable of blocking E-selectin and P-selectin mediated functions, the method comprising:
immunizing a mammal with P-selectin;
immunizing the mammal with E-selectin;
immortalizing B-cells from the mammal to obtain immortalized cells producing antibodies; and selecting an immortalized cell producing an antibody that specifically binds to E-selectin and to P-selectin.

30. A method of detecting E-selectin and P-selectin bearing cells in a biological sample suspected of containing the cells, the method comprising:

contacting the sample with the antibody of claim 2 to form an immune complex with the E-selectin and P-selectin bearing cells: and detecting the presence of the immune complex to indicate the presence of the cells.

31. A method of detecting E-selectin and P-selectin bearing cells in a biological sample suspected of containing the cells, the method comprising:
contacting the sample with the antibody of claim 12 to form an immune complex with the E-selectin and P-selectin bearing cells; and detecting the presence of the immune complex to indicate the presence of the cells.

32. A monoclonal antibody that specifically binds to E-selectin and P-selectin, said antibody binding to the same epitope of E-selectin as antibody 5C7.29, ATCC accession number CRL 11640.

33. The monoclonal antibody of claim 32, said antibody further binding to the same epitope of P-selectin as antibody 5C7.29, ATCC accession number.