NZ264353A - Method of collecting or purifying leukocytes from a fluid sample, apparatus, immune response inhibitor test - Google Patents

Description

ly 2 6 4 3 5 3 Patents Form 5 } Priority Dste(s): Complete Specification Rled: WO,is/ov, CiLMloQ) ..Mmtiki Publication Date: P.O. Journal No: provisions of Bc§> li'bn 23 {";) Jr-S" Specifics. >j*i hcs bi;-n zrte-d^.c-^ .Stir .^<V' J********* • - fnftsis.

This application is divided out of N.Z. No. 242896 NEW ZEALAND Patents Act 1953 COMPLETE SPECIFICATION DEVICE AND METHOD FOR ANALYSIS OF BLOOD COMPONENTS AND IDENTIFYING INHIBITORS AND PROMOTERS OF THE INFLAMMATORY RESPONSE II We, CENTER FOR BLOOD RESEARCH, INC., a Research and Development corporation incorporated under the laws of the State of Massachusetts, United States of America of 800 Huntington Avenue, Boston, Massachusetts 02115, United States of America do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement:- - 1 - (Followed by 1A) 26 4 3 - l A — DEVICE AND METHOD FOR ANALYSIS OF BLOOD COMPONENTS AND IDENTIFYING INHIBITORS AND PROMOTERS OF THE INFLAMMATORY RESPONSE TC This invention was made with government 5 support under grant number CA 31799 from the National Institutes of Health. The government has certain rights in the invention. 1. INTRODUCTION 10 The present invention related to in vitro models of the in vivo rolling and arrest of leukocytes along the endothelial cell wall, an important part of the inflammatory response. The invention also relates to apparatuses and methods for collecting, purifying, 15 and analyzing blood and blood components, and methods for identifying inhibitors or promoters of components of the inflammatory response. 2. BACKGROUND OF THE INVENTION 20 2.1 LEUKOCYTE EXTRAVASATION The migration of leukocytes (white blood cells) out of the blood and into tissues (extravasation) is the central event in the inflammatory response. Leukocyte emigration is responsible for the successful host response to tissue injury and infection, but is also potentially harmful and contributes to the pathology of many diseases and inflammatory disorders.

Lymphocytes exit from the blood by selective 30 interaction with high endothelial venule (HEV) cells. In autoimmune disease and inflammation, most lymphocyte extravasation occurs through nonspecialized endothelium rather than HEV. Recirculating lymphocytes express homing receptors which interact in an organ-specific manner with HEV in peripheral lymph 26 4 3 nodes, mucosa-associated lymphoid tissues, and in inflamed joint tissue (Jalkanen et al., 1987, Ann. Rev. Med. 38:467-476). Related receptors are expressed on other leukocyte subsets f id.). The first 5 step in leukocyte migration into tissues is margination, when leukocytes leave the central stream of flowing blood cells in a postcapillary venule and roll along the endothelial lining of the vessel (Cohnheim, 1889, Lectures on General Pathology: A 10 Handbook for Practitioners and Students (London: The New Sydenham Society)). Leukocytic margination in postcapillary ventiles should be distinguished from the "marginafcing pool" of about 50% of leukocytes that may be in capillary beds in the lung or tissues and enter 15 the circulation in response to exercise or epinephrine. Postcapillary venules are major sites of leukocyte emigration in inflammation, and there are few or no marginating leukocytes in these venules in the healthy state (Fiebig et al., 1991, Int. J. 20 Microcirc. Clin. Exp. 10:127-144).

As observed more than 100 years ago using intravital microscopy (Cohnheim, 1889, Lectures on General Pathology: A Handbook for Practitioners and Students (London: The New Sydenham Society)), 25 leukocytes begin to interact with the vessel wall by rolling along the endothelium within minutes after injury to adjacent tissue. The rolling response is seen throughout Vertebrata, in cold-blooded animals such as amphibians as well as in mammals (Cohnheim, 30 1889, Lectures on General Pathology: A Handbook for Practitioners and Students (London: The New Sydenham Society)). The number of rolling cells increases dramatically during the course of an inflammatory reaction (Atherton and Born, 1972, J. Physiol. 35 233:157-165) and is important in the accumulation of 26 4 3 cells at the site (Fiebig et al., 1991, Int. J. Microcirc. Clin. Exp. 10:127-144). As the inflammatory reaction progresses, the endothelium becomes paved with leukocytes, and their rolling 5 decreases in velocity and is interrupted by halts until they come to a firm stop (Cohnheim, 1889, Lectures on General Pathology: A Handbook for Practitioners and Students (London: The New Sydenham Society)). Throughout this process the cells remain 10 round, but undergo a dramatic change in shape immediately upon initiation of emigration. A pseudopod is extended through the vessel at a junction between endothelial cells, and this often is accompanied by a flattening of the leukocyte against 15 the vessel wall (Marchesi, 1961, Q. J. Exp. Physiol^ 46:115-133). Transmigration continues as the pseudopod grows in ramifications and size until the entire cell body has emerged through a narrow gap between endothelial cells (Cohnheim, 1889, Lectures on 20 General Pathology: A Handbook for Practitioners and Students (London: The New Sydenham Society)). Cells appear to reach the point at which they emigrate by rolling; no active migration along the vessel wall is evident by intravital microscopy.

Both the rheology of blood and specific adhesive interactions may regulate the rolling response. Hydrodynamic studies of particles in suspension show that in Poiseuille flow, the larger particles are forced to the center of the stream, and 30 this effect is more pronounced as shear forces increase (Segre and Silberberg, 1962, J. Fluid Hech. 14:136-157). This effect has been confirmed for blood cells both in vivo and in vitro; the larger leukocytes are forced to the center of the stream in normal flow 35 (Goldsmith and Spain, 1984, Microvasc. Res. 27:204- — 4 264353 222; Nobis at al., 1985, Microvasc. Res. 29:295-304). In inflammation, vessels dilate and flow is slowed. Vascular permeability is increased, leading to plasma leakage and an increased hematocrit, and together with S slower flow, leads to erythrocyte rouleaux formation. A combination of these factors causes leukocytes to be displaced to the marginal region of flow near the vessel wall (Chien, 1982, Adv. Shock Res. 8:71-80). This makes contact of a circulating leukocyte with the 10 vessel wall more probable, but shear forces acting on the leukocyte at the vessel wall oppose adhesion to the endothelium. The velocity profile of a vessel shows no flow at the vessel wall and a parabolic increase toward the centerline. Because fluid 15 velocity increases with distance from the wall, cells near the wall have torque exerted on them and will tumble even if not in contact with the wall. However, the velocity at which cells tumble in a shear flow near to the vessel wall is much faster than observed 20 for rolling cells in inflammatory reactions, suggesting that adhesive interactions occur between the leukocyte and vessel endothelium (Atherton and Born, 1973, J. Physiol. 233:157-165).

More than 100 years after Cohnheim (1889, 25 Lectures on General Pathology: A Handbook for Practitioners and Students (London: The New Sydenham Society)) postulated molecular changes in vessel endothelium in inflamination, the molecular basis of leukocyte rolling remains unknown. However, three 30 families of adhesion receptors that participate in leukocyte interactions with endothelium have been defined: the integrin, immunoglobulin-related, and selectin molecules (reviewed in Springer, 1990, Nature 346:425-433)). 2 6 4 3 5 3 The iivtegrins LFA (lymphocyte function-associated antigen)-1 and Mac-l on the neutrophil bind to the Ig family member ICAM (intercellular adhesion molecule)-1 on endothelium (Smith et al., 1988, J.

Clin. Invest. 82:1746-1756; Smith et al., 1989, J. Clin. Invest. 83:2008-2017; Diamond et al., 1990, J. Cell Biol. 111:3129-3139). LFA-1 and not Mac-l binds to ICAM-2 (de Fougerolles et al., 1991 J. Exp. Med. 174:253-267; Diamond et al., 1990, J. Cell Biol. 10 111:3129-3139), an endothelial cell molecule that is more closely related to ICAM-1 than these molecules are to other Ig superfamily members (Staunton et al., 1989, Nature 339:61-64). Stimulation of neutrophils with chemoattractants is required to activate binding 15 of these integrins to ICAM-1 (Smith et al., 1989, J.' Clin. Invest. 83:2008-2017; Diamond et al., 1990, J. Cell Biol. 111:3129-3139). Stimulation of neutrophil integrin avidity is a rapid response occurring in minutes, does not require increased integrin surface 20 expression (Buyon et al., 1988, J. Immunol. 140:3156-3160; Philips et al., 1988, J. Clin. Invest. 82:495-501; Vedder and Harlan, 1988, J. Clin. Invest. 81:676-682; Lo et al., 1989, J. Exp. Med. 169:1779-1793), and appears analogous to an increase in avidity described 25 for LFA-1 on T lymphocytes in response to antigen receptor crosslinking (Dustin and Springer, 1989, J. Cell Biol. 107:321-331).

ICAM-1 induction is a second mechanism for regulating inflammatory cell interactions that occurs 30 on a time scale of hours and requires mRNA and protein synthesis (reviewed in Springer, 1990, Nature 346:425-433). ICAM-1 is expressed basally on endothelial cells but is greatly increased at inflammatory sites and by stimulation with lipopolysaccharide and 35 cytokines such as IL-1 and TNF. By contrast to 26 4 3 ICAM-1, ICAM-2 is expressed at higher surface density on resting endothelium but is not inducible (de Fougerolles et al., 1991, J. Exp. Med. 174:253-267).

LFA-1 and Mac-l together with pl50,95 5 comprise the leukocyte integrins, a subfamily of integrins that share a common ft subunit (CD18) and have distinct aL, aM and aX (CDlla, b and c) a subunits (reviewed in Larson and Springer, 1990, Immunol. Rev. 114:181-217; Springer, 1990, Nature 10 346:425-433). They are required for leukocyte emigration as demonstrated by an absence of neutrophil extravasation 1) in patients with mutations in the common (S subunit (leukocyte adhesion deficiency), and 2) after treatment of healthy neutrophils with a IS monoclonal antibody (mAb) to the common j3 subunit in vivo or in vitro (reviewed in (Anderson and Springer, 1987, Ann. Rev. Med. 38:175-194; Larson and Springer, 1990, Immunol. Rev. 114:181-217). Patient neutrophils, and healthy neutrophils treated with mAb 20 to the common /3 subunit or a combination of mAb to LFA-1 and Mac-l a subunits, are deficient in binding to endothelial cells in static adhesion assays (Buchanan et al., 1982, Blood 60:160-165; Harlan et al., 1985, Blood 66:167-178). However, when binding 25 of neutrophils in shear flow is measured, the leukocyte integrin-dependent component of binding is lost at a shear stress below the physiologic range (Lawrence et al., 1990, Blood 75:227-237).

Nonetheless, patient and CD18-treated cells that bind 30 to the endothelium through other adhesion mechanisms fail to undergo transendothelial migration, in agreement with the lack of neutrophil diapedesis in leukocyte adhesion deficiency (Smith et al., 1988, J. Clin. Invest. 82:1746-1756). 26 4 35 3 The integrin VLA-4, that contains the a4 (CD49d) subunit noncovalently associated with the /3l (CD29) subunit, is expressed by lymphocytes, monocytes, and neural crest-derived cells, and can S interact with vascular cell adhesion molecule-1 (VCAM-1) (Elices et al., 1990, Cell 60:577). Like ICAM-1 and ICAM-2, VCAM-1 is a member of the immunoglobulin (Ig) superfamily (Osborn et al., 1989, Cell 59:1203), but unlike the ICAMs, VCAM-1 is not 10 expressed by lymphocytes (Wellicome et al., 1990, J.

Immunol. 144:2558; Rice et al., 1990, J. Exp. Med. 171:1369). VCAM-1 expression is very low or absent on resting endothelial cells in culture but can be induced by cytokines such as TNF or IL-1 with kinetics 15 of induction similar but not identical to that of ICAM-1 (Wellicome et al., 1990, J. Immunol. 144:2558; Carlos fit al., 1990, Blood 76:965). Peak expression of VCAM-1 after continuous treatment of endothelial cells with TNF in culture occurs somewhat earlier than 20 the peak expression of ICAM-1, but both persist at levels substantially higher than basal expression for at least 48 hr (Carlos et al., 1990, Blood 76:965).

Unlike LFA-1, however, VLA-4 can also interact with fibronectin, binding to the alternatively spliced CS-1 25 . region located C-terminal to the RGD site of fibronectin recognized by the integrin VLA-5 (Guan and Hynes, 1990, Cell 60:53; Wayner et al., 1989, J. Cell Biol. 109:1321; Hemler, 1990, Annu. Rev. Immunol. 8:365). Two forms of VCAM-1 cDNA clones, which most 30 likely represent alternatively spliced products, have been reported (Osborn et al., 1989, Cell 59:1203; Polte et al., 1990, Nucl. Acids Res. 18:5901; Cybulsky et al., 1991, Am. J. Pathol. 138:815; Hession et al., 1991, J. Biol. Cheat. 266:6682). 264353 The selecting are the most recently recognized class of leukocyte adhesion molecules (reviewed in Springer, 1990, Nature 346:425-433).

They have an N-terminal lectin domain, one epidermal 5 growth factor-like module, and from two to nine short consensus repeats. By contrast to integrins and immunoglobulin (Ig) family members, selectins have been found to date only on circulating cells and the endothelium, suggesting that they may be specialized 10 for interactions within the vasculature. CD62 (PADGEM or GMP-140) is expressed in a granules of platelets and Weibel-Palade bodies of endothelial cells, and is mobilized to the plasma membranes of these cells after activation by mediators of inflammation and 15 hemostasia, allowing these cells to bind neutrophils and monocytes at the site of tissue injury (Larsen et al., 1989, Cell 59:305-312; Geng et al., 1990, Nature 343:757—760). CD62 is rapidly unregulated on the endothelial cell surface, suggesting that it may be 20 important early in inflammation (Hattori et al., 1989, J. Biol. Chem. 264:7768-7771; Geng et al., 1990, Nature 343:757-760). ELAM-1 is synthesized by endothelial cells in response to inflammatory agents and promotes adhesion of neutrophils, monocytes, and a 25 subpopulation of lymphocytes (Bevilacqua et al"., 1989, Science 243:1160-1165; Picker et al., 1991, Nature 349:796-798; Shimizu et al., 1991, Nature 349:799-802). The LAM-1 or LECAM-1 molecule is expressed on leukocytes and facilitates their binding to 30 endothelium during lymphocyte recirculation through peripheral lymph nodes and neutrophil emigration at inflammatory sites (Jutila et al., 1989, J. Immunol. 143:3318-3324; Spertini et al., 1991, Nature 349:691-694; Watson et al., 1991, Nature 349:164-167). 35 Carbohydrate ligands for selectins have recently been 264353 defined (reviewed in Springer and Lasky, 1991, Nature 349:196-197); that for CD62 has Lewis x as an important component (Larsen et al., 1990, Cell 63:467-474) and also appears to be sialylated (Moore et al., 5 1991, J. Cell Biol. 112:491-499). Neutrophils bear Lewis x both on glycolipids and at the termini of N-and 0-linked oligosaccharides (Symington et al., 1985, J. Immunol. 134:2498-2506; Fukuda et al., 1984, J. Biol. Chem. 259:10925-10935). Antibodies to selectins 10 and integrins additively inhibit neutrophil adhesion to endothelium, suggesting that they mediate distinct adhesion mechanisms (Luscinskas et al., 1989, J. Immunol. 142:2257-2263; Dobrina et al., 1989, Immunology 67:502-508; Smith et al., 1991, J. Clin. 15 Invest; 87:608-618; Hallmann et al., 1991, Biochem. Biophys. Res. Commun. 174:236-243). The molecular basis of rolling does not appear to involve the leukocyte integrins, based on the inability of Mab to the leukocyte integrin common CD18 ft subunit to 20 inhibit rolling in vivo (Arfors et al., 1987, Blood 69:338-340).

Chemoattractants bind to serpentine family receptors on the surface of a leukocyte. A highly selective class of chemoattractants described in the 25 last few years are small proteins of 70 to 80 amino acids that belong to a recently described protein family called the intercrines (Oppenheim et al., 1991, Ann. Rev. Immunol. 9:617-648). The chemoattractant receptors mediate pro-inflammatory and chemotactic 30 actions, and transduce ligand-mediated signals through interactions with G proteins (GTP-binding proteins). Actions mediated by chemoattractant receptors include stimulation of granule-enzyme release and superoxide anion production, upregulation of expression and 35 activity of the cell adhesion molecule Mac-l (CDIIb, - 2 6 4 3 5 3 CD18), increased expression of CR1, a decrease in cell surface glycoprotein 10QMEL-14 on neutrophils (Gerard and Gerard, 1991, Nature 349:6-14), and stimulation of neutrophil adherence to and emigration through 5 activated endothelial cells (Huber et al., 1991, Science 254:99). Interleukin (IL-8) can also act as an adhesion or migration inhibitor when added on the same side of activated endothelium as neutrophils (Huber et al., 1991, Science 254:99; Gimbrone et al., 10 1989, Science 246:1601). In vivo, these receptors may participate in anaphylactoid and septic shock (Gerard and Gerard, supra).

The best characterized chemoattractant receptor is the one which binds formylpeptides. cDNAs IS encoding receptors for four chemoattractants, formylpeptide [e.g. fNet-Leu-Fhe (fMLP)] (Boulay et al., 1990, Biochem. Biophys. Res. Commun. 168:1103-1109; Boulay et al., 1990, Biochemistry 29:11123-11133), C5a (Gerard and Gerard, 1991, Nature 349:614-20 617) platelet activating factor (PAF; Honda et al., 1991, Nature 349:342-346), and IL-8 (Holmes et al., 1991, Science 253:1278-1280) have been cloned.

Citation of a reference herein shall not be construed as an admission that such reference is prior 25 art to the present invention. 2.2. SEPARATION OF BLOOD AND BLOOD COMPONENTS Various practices have been developed for separating whole blood and blood components. Generally, whole blood comprises red blood cells (RBC), platelets, and white blood cells of various types known collectively as leukocytes. Red blood cells and platelets can be separated from whole blood by centrifugation, leaving the blood product plasma. 11 - 264 353 Removal of leukocytes to low levels In a blood sample Is desirable to prevent diseases caused by transfusion of blood that contains donor leukocytes. These diseases include viral infections 5 transmitted by leukocytes harboring viruses and Infections in which the transfused leukocytes elicit allo-antibodies that cause transfusion reactions. Moreover, enrichment of leukocytes is also desirable when different types of leukocytes are needed for a 10 variety of clinical and research purposes.

Nevertheless, currently used methods are problematic for depleting, extracting and identifying leukocyte populations in blood. Centrifugal methods for separating blood do not entirely separate out the 15 leukocytes; they may be present in substantial quantities in both the packed red blood cells and the platelet-concentrate fractions. Centrifugal methods are somewhat costly and the sterility of the product is such that it must be used within a short period of 20 time. A number of other devices have been proposed in which fibers are packed into housings and whole blood allowed to pass through them in order to remove a portion of the white cell content. Typically, these devices are based on size separation but the various 25 types of leukocytes are not widely divergent in size and most of these cells can deform so as to pass through much smaller openings than their normal size. Accordingly, removal of leukocytes can be accomplished by absorption in conjunction with filtration. 30 Attempts have been made to reduce leukocyte concentration in blood by exposure to a variety of surfaces, including polyamide, polyester, acrylics, cellulosics, cellulose acetate and siliconized glass wool. Devices of this type have been only partially 35 successful, primarily because many leukocytes become 12 26 4 35 activated by the act of absorbing or binding to the fibrous material. Even when activated, many of the leukocytes will not adhere to the fibrous material at physiologic flow rates.

A need exists to develop a quick and effective method for concentrating and purifying leukocytes that cam be used in research and clinical settings for identification, differentiation, and analysis of leukocyte populations. 3. SUMMARY OF THE INVENTION The present application is divided out of New Zealand specification No. 242896.

Both the present invention and that disclosed in NZ specification No. 242896 provide in vitro models of the in vivo rolling and arrest of leukocytes 15 along the endothelial cell wall, which are important steps in the migration of leukocytes out of the blood stream and into tissue, as part of the inflammatory response. The in vitro models of the present invention and that of NZ specification No. 242896 provide a physiologically relevant model of such 20 interactions, since the apparatuses of these inventions can reproduce in vitro the sequential molecular interactions that are steps in leukocyte accumulation at an inflammatory site in vitro; and, furthermore, function under physiologic flow conditions resulting in physiologic shear stresses similar to those -present in vivo at the sites of leukocyte transendothelial migration (extravasation).

Apparatuses embodying and methods relating to the in vitro models of the present invention and that of NZ specification No . 242896 are also provided. Apparatuses embodying the models of these inventions provide quick and effective methods for collecting, purifying, and analyzing leukocyte populations and subpopulations.

The present invention discloses apparatus for modelling leukocyte rolling comprising a solid phase surface with rolling mediator molecules present 26 4 35 thereon. Such rolling mediators sure, for example, selectins and selectin ligands which have binding partners expressed on leukocytes. NZ specification No. 242896 discloses apparatus for modelling leukocyte rolling followed S by adhesion/arrest comprising a solid phase surface with both rolling mediators and integrin binding partners present thereon. The apparatuses of the present invention and that of NZ specification No. 242896 can be used for collecting, concentrating, purifying, and 10 analyzing blood and blood components, in particular, leukocytes and subsets thereof. Therapeutic and diagnostic methods based on the foregoing are also provided.

The present invention and that of NZ specification No. 242896 further relate to methods for 15 identifying inhibitors or, alternatively, promoters' (agonists, functional components) of the processes of leukocyte rolling and adhesion, important components of the inflammatory response. Pharmaceutical compositions and kits are also provided. 4. DESCRIPTION OF THE FIGURES Figure 1. A Schematic Illustration of Leukocyte Rolling on a Substrate Containing a Selectin.

Figure 2. A Schematic Cross-Section of an Apparatus of the Invention.

Figure 3. A Schematic Perspective Illustration of an Embodiment of an Apparatus of the Invention. The apparatus shown is a parallel plate 30 flow chamber, after Lawrence et al. (1987, Blood 70:1284-1290), for measuring attachment of neutrophils under flow conditions. The chamber is shown upside down for illustration purposes; the glass slide with the artificial bilayer formed the lower parallel 35 plate. It was mounted on the stage of an inverted 264 microscope equipped with a video camera. The cell suspension was connected to the inlet manifold and a syringe pump to the outlet manifold.

Figure 4. Schematic of parabolic flow 5 profile in a blood vessel or parallel plate flow chamber.

Figure 5. Attachment of Neutrophils to Artificial Planar Bilayers during Flow. Resting or PMA-stimulated neutrophils were infused at varying 10 wall shear stresses through the parallel-plate flow chamber. A planar membrane containing CD62, ICAM-1, or both at the density indicated below was formed on one side of the chamber. After 3 minutes of continuous flow to equilibrate adherence and 15 deadherence to the bilayer, adherent neutrophils were quantitated. (A) and '(B) represent two different sets of experiments: day-to-day variation throughout this work was slight as exemplified by binding of unactivated neutrophils to CD62 at 200 sites per /m2 in 20 both panels. Data are averaged from four to six (A) and two (B) experiments. Bars show the standard error of the mean (SEM) of the experiments. At the next higher experimental point, at 7.3 dyn/cm2, binding was zero in all cases. (A) open circles: CD62 at 400 25 sites per /in2; closed circles: CD62 at 200 sites per Mm2; closed triangles: CD62 at 50 sites per /um2; closed squares: CD62 at 25 sites per /m2; open squares: ICAM-1 at 1000 sites per Mm2. Cells were either unstimulated (solid line) or PMA-stimulated 3° (broken line). (B) closed circles: CDE62 at 200 sites per /ra2; closed squares: ICAM-1 at 250 sites per Mm2; closed triangles: CD62 at 200 sites per fm2 plus ICAM-1 at 250 sites per jim2. Cells were either unstimulated (solid line) or PMA-stimulated (broken 35 line). - 26 4 3 Figure 6. Accumulation and Rolling of Neutrophils on Artificial Bilayers Containing CD62 Compared with Lack of Interaction with Bilayers Containing ICAM-1. (A) Neutrophils infused at a shear 9 stress of 1.8 dyn/cm2 accumulated and rolled on a bilayer containing 200 sites/Mm2 of CD62 (20 x objective). Rolling neutrophils are visualized as round distinct cells by the video camera because their rolling velocity (5.2 pm/a here) is slow compared to 10 the exposure time (30 frames per s) . By comparison, nonadherent neutrophils tumbling in the shear flow appear as blurred streaks. The streaks of cells closest to the bilayer are short because laminar flow is slowest here. (B) Same as (A), except on a bilayer 15 containing 1,000 sites per im7 of ICAM-1. No rolling cells have accumulated. Nonadherent cells appear as streaks, and those closest to the bilayer have a velocity of about 500 /xm/s. (C-F) video images at 2 s intervals of neutrophils rolling on a bilayer 20 containing 400 sites per /im2 of CD62 at a shear stress of 7.3 dyn/cm2 (40 x objective). Nuclei are resolved at this magnification, showing that the cells are rolling rather than sliding. Cells rolled at an average velocity of 8.2 nm/sj nonadherent cells 25 tumbled too fast (>1700 pm/s) to be visualized at this flow rate and magnification.

Figure 7. Rolling Velocity as a Function of CD62 Density and Shear Stress. CD62 and ICAM-1 were used at the density indicated below. Error bars 30 represent the. SEM based on measurements from independent experiments. Experimental points at 14.6 dyn/cm2 are not shown, but indicated by connecting lines. Closed circles: CD62 at 50 sites per /in2; open circles: CD62 at 200 sites per /xm2; closed squares: 26 4 35 3 CD62 at 200 sites per /ra2 plus ICAM-1 at 250 sites per /tun2; open triangles: CD62 at 400 sites per /im2.

Figure 8. Neutrophils in Contact with Artificial Membranes under Static Conditions.

Neutrophils, with or without stimulation with PMA, were allowed to adhere in the absence of flow to artificial bilayers containing 250 sites per /ra2 of ICAM-1 or 200 sites per fat? of CD62 for varying time periods, as indicated. (B)-(F) represent the same 10 field of cells at one minute intervals. Time points prior to 3 min are not shown because it takes 2 min for all cells to settle onto the bilayer.

Figure 9. Detachment Assay following Static Incubation of Neutrophils on Artificial Membranes 15 containing either CD62, ICAM-1, or a mixture of CD62 and ICAM-1. Neutrophils were injected through a port in the side of the flow chamber and allowed to settle onto artificial bilayers containing 200 sites per /id2 of CD62, 250 sites per /xm2 of ICAM-1, or both. For 20 some experiments, PMA was added to the neutrophil suspension before it was injected into the flow chamber (broken line). After 6 minutes of contact, shear stress was applied in staged increments. Neutrophils bound after 20 s at each shear stress 25 point are expressed as the percentage of neutrophils that settled onto the bilayer in the initial contact period. Solid lines represent binding of unstimulated neutrophils to the membrane. Broken lines represent binding of PMA-stimulated neutrophils to the membrane. 30 Error bars represent the SEM of three to five independent experiments (A) and two experiments (B). Experimental points at 36 dyn/cm2 are not shown but are indicated by connecting lines. (A) closed circles: CD62 at 200 sites per jun2; closed squares: ICAM-1 at 35 250 sites per /xm2. (B) closed circles: CD62 at 200 ■ , ; 26 4 35 3 sites per m*2; cloMd aquaru: ICAM-1 at 250 sites par M*2; closed triangles: CD62 at 200 sites per pa? plus ICAM-1 at 250 sites par Flgura 10. Stimulation with FMLP Arrasts Neutrophil Rolling on Artificial Membranes Containing both CD62 and ICAM-1. Neutrophils ware allowed to S adhere to artificial bilayers containing CO62 (200 sites per fa?) and ICAM-1 (250 sites per Mm2) during flow at a wall shear stress of 1.8 dyn/cm2 (closed circles in (A)) or 0.73 dyn/cm2 (open squares in (A)).

Rolling velocity -was constant for at least 5 min under these conditions. During the experiment, 10"* M FMLP was added to the medium being infused into the chamber and reached the cells rolling in the field of view 30 or 12 s later (narked 0 time) as verified in smother experiment with a dye solution. (A) Mean rolling velocity as a function of time after FMLP addition.

Bars show the SEM for two experiments. (B and c) Distribution of rolling velocities of adherent neutrophils with time after FMLP exposure at 0.73 and 1.8 dyn/cm2, respectively. (D) Photomicrographs of neutrophils before the infusion of FMLP (0 time) and after exposure to 10* M fMLP (5 min) on the same area of the planar membrane. fMLP addition induced arrest and spreading of the neutrophils. Flow was stopped for approximately 30 s to take the photographs in a separate experiment from the two experiments averaged for data in (A)-(C). (E) Shear resistance of r. ' rophil binding. Neutrophils were allowed to attach at 0.73 dyn/cm2 and shear resistance was determined 5 min after fMLP exposure.

. DETAILED DESCRIPTION OF THE INVENTION The present invention and that disclosed in the NZ specification No. 242896 are so closely related that the description from the parent application (NZ 242896) has been maintain in the present application. However it is clear to a person skilled in the art which parts of the description relate to each of the inventions.

The present invention provides in vitro models of leukocyte rolling-and adhesion, which are 18 - 26435; functional at physiologic shear stresses. The present invention is based in part on the discovery that under physiologic flow conditions, leukocyte rolling mediated by interactions between selectins and their 5 leukocyte binding partners is a prerequisite for chemoattractant-stimulated interaction of integrins on leukocytes with integrin binding partners, that results in arrest of the leukocytes. Thus, leukocyte extravasation appears to involve the following 10 sequential steps: (1) reversible adhesion of flowing leukocytes to the blood vessel wall and subsequent rolling, mediated by interactions between selectins and their carbohydrate ligands; (2) leukocyte activation mediated by binding of chemoattractants, 15 presumably diffusing out of the inflammatory site or expressed on the surface of the endothelium, to their receptors on the leukocyte cell surface; and (3) arrest and stable attachment of the rolling leukocyte mediated by binding of leukocyte integrins to 20 immunoglobulin (Ig) family member molecules on the endothelium.

Apparatuses embodying and methods relating to the in vitro models of the invention are also provided. In a specific embodiment, for modelling 25 leukocyte rolling, the apparatus of the invention comprises a solid phase surface with rolling mediator molecules present thereon. Such rolling mediators include but are not limited to selectins and selectin ligands, expressed on endothelium in vivo, which have 30 binding partners expressed on leukocytes. In another specific embodiment, for modelling leukocyte rolling followed by adhesion/arrest( the apparatus of the invention comprises a solid phase surface with both rolling mediators and integrin binding partners 35 present thereon. The apparatuses of the invention can # 26 4 35 3 be used for collecting, concentrating, purifying, and analyzing blood and blood components, in particular, leukocytes and subsets thereof. Therapeutic and diagnostic methods based on the foregoing are also 5 provided. The invention further relates to methods for identifying inhibitors or, alternatively, promoters (agonists, functional components) of the processes of leukocyte rolling and adhesion, important components of the inflammatory response. 10 Pharmaceutical compositions and kits are also provided. ^ In a specific embodiment of the present invention detailed in the examples sections infra. we show that a selectin is a rolling receptor. We 15 demonstrate that at physiologic shear stress, neutrophils bind to and roll on CD62 in artificial bilayers. We also demonstrate qualitative differences between selectin and integrin adhesion mechanisms, and cooperation between them. On bilayers containing both 20 CD62 and ICAM-1, the rolling interaction through CD62 is a prerequisite for chemoattractant-stimulated interaction of integrins on neutrophils with ICAM-l that arrests rolling and dramatically strengthens adhesion. This essentially reproduces in vitro the 25 steps of leukocyte accumulation at an inflammatory site in vivo. .1. THE IN VITRO MODELS OF THE INVENTION The present invention provides two general 30 types of apparatuses: one which provides an An vitro model of leukocyte reversible adhesion and subsequent rolling along an endothelial vessel wall in vivo (hereinafter "the rolling model"); and one which provides an in vitro model of leukocyte arrest and 35 stable attachment to the endothelium subsequent to 26 4 35 3 rolling and activation by chemoattractant binding (hereinafter "the arrest model") ; both of which models are functional at physiologic shear stresses.

The rolling model comprises a solid phase 5 surface with a rolling mediator present thereon. Interaction of the rolling mediator (preferably, a selectin or selectin ligand) with its binding partner on a leukocyte cell surface, at physiologic shear stresses, mediates reversible attachment and rolling 10 of the leukocyte on the solid phase surface of the rolling model. In a preferred aspect, the rolling model apparatus comprises, on a solid phase, planar lipid bilayers containing the selectin C062.

The term "rolling" as used herein with 15 reference to a leukocyte, refers to the literal rolling of leukocytes along a surface containing a rolling mediator, which rolling is induced by the interaction of the rolling mediator with the leukocyte in the presence of fluid drag forces arising from 20 relative movement between the surface containing the rolling mediator and a medium containing the leukocytes. This rolling by leukocytes is seen throughout the Vertebrata, in cold blooded animals such as amphibians as well as in mammals. The number 25 of rolling cells in vivo increases dramatically during the course of an inflammatory reaction and is important in the accumulation of cells at the site of injury.

The arrest model comprises a solid phase 30 surface with both a rolling mediator and an integrin binding partner present thereon, both with the same specificity for the leukocyte cell subset of interest (see infra1. In the presence of a chemoattractant with matching cellular specificity, and under 35 approximate physiologic shear conditions, interaction 26 4 35 3 of the rolling mediator with its binding partner on the leukocyte leads to rolling of the leukocyte along the solid phase surface, followed by activation due to chemoattractant binding, and stable attachment and 5 arrest of the leukocyte upon interaction of the leukocyte integrin with its binding partner on the solid phase surface, in a preferred aspect, the arrest model apparatus comprises, on a solid phase surface, planar lipid bilayers containing the selectin 10 CD62 and the integrin binding partner ICAM-1; and is preferably used with a formyl peptide such as N-formyl-methionyl leucyl phenylalanine (fMLP) as the activating chemoattractant. In the use of the arrest model, following arrest of the leukocytes on the solid 15 phase surface, the solid phase surface can be removed, and, e.g., the cells present thereon can be released or counted, and new solid phase surfaces inserted within the housing of the apparatus subsequent to use. Alternatively, attached cells can be dissociated from 20 the solid phase surface within the apparatus by exposing them to divalent cation chelating agents such as citrate, EDTA, EGTA, etc., and washed away from the solid phase surface; the surface can then be re-used.

The apparatus of the arrest model can also 25 be used as in vitro model of leukocyte rolling without subsequent arrest, by use of such device in the absence of the activating chemoattractant required for subsequent arrest.

In order to obtain arrest of the desired 30 leukocyte cell subset, a rolling mediator and an integrin binding partner must be present on the surface of the solid phase of the arrest model, and a chemoattractant must be used, that have respective cognate binding partners present on such leukocyte 35 cell subset. Similarly, in order to obtain rolling of 2 6 4 3 5 a desired leukocyte cell subset in either the apparatus of the rolling model or arrest model, the binding partner for the rolling mediator on the surface of the solid phase of the apparatus must be 5 present on such leukocyte cell subset. Leukocyte cell subsets include but are not limited to neutrophils, eosinophils, basophils, mast cells (collectively known as granulocytes or polymorphonuclear leukocytes), monocytes, macrophages, and lymphocytes (both T 10 lymphocytes and B lymphocytes). For purposes herein, platelets shall also be deemed included within leukocytes, unless clearly excluded by context or otherwise. Exemplary leukocyte cell subset specificities of interactions with various rolling 15 mediators, chemoattractants, and integrin binding partners are described infra. One or more types of rolling mediator molecules may be present on the solid phase surface(s) of the apparatuses of the invention. One or more types of integrin binding partners may be 20 present on the solid phase surface(s) of the arrest model apparatuses of the invention. One or more chemoattractants may be provided in the use of the arrest model to promote leukocyte arrest after rolling.

"Binding partner" of a molecule as used herein refers to a receptor or ligand interacting with such molecule, e.g., by binding or other type of noncovalent association.

Rolling of leukocytes on a solid surface 30 which contains one or more affixed rolling mediators, and arrest of leukocytes on such a surface which also contains one or more integrin binding partners, can occur at physiologic flow rates, and the methods of the invention are designed to produce physiologic flow 35 rates and induce physiologic shear stresses. It is # 26 4 35 3 % well known that In Newtonian fluids at Reynolds numbers less than 2000, flow is laminar. Furthermore, fluid velocity is zero at the vessel wall and increases parabolically toward the center of the 5 vessel. The change in velocity per change in radial displacement away from the vessel wall is called the shear rate and is highest at the wall. Shear stress scales linearly with the fluid forces acting on a cell under laminar flow conditions and is the product of 10 shear rate and the fluid viscosity.

Preferred shear stresses induced by methods of the invention bracket the range estimated to exist in vivo in post-capillary venules, namely, 0.5-30 dyn/cm2. Wall shear rates and stresses in the 15 apparatus of the invention can be calculated from chamber geometry, volumetric flow rate and viscosity (see, for example, Lawrence et al., 1990, Blood 75:227-237). Specifically, wall shear stress (T) can be calculated as follows: T - 3/*Q/2ba2 where T - wall shear stress, = coefficient of viscosity (cP), Q = volumetric flow rate (cm?/s) t ahalf channel height, and b « channel width.

The viscosity of water at 37#C (0.007 poise) can often be used as an approximation of the viscosity of the flow medium. The wall shear rate is given by T/fi. Shear stresses in vivo can be calculated from center-line velocity and vessel diameter using well known 3.P equations (see, e.g., Atherton and Born, 1972, J. Physiol. 222:447-474; Atherton and Born, 1973, J. Physiol. 233:157-165). Shear is defined according to the following: 26 4 35 3 dV » shear (s"1) dr shear x viscosity =■ shear stress (dynes/cm2) V » velocity; r — radius 5 Physiologic flow rates are those flow rates sufficient to induce shear stresses in the apparatus of the invention of between about 0.5 to about 30 dynes per square centimeter. Particularly preferred flow rates are sufficient to induce shear stresses of between 10 about 0.5 and about 4.0 dynes per square centimeter, with those in the range of about 0.5-2.0 dyn/cm2 most preferred.

The apparatuses of the invention and components thereof axe described in more detail in the subsections below. .2. ROLLING MEDIATOR COMPONENTS OF THE APPARATUSES OF THE INVENTION Any rolling mediator known in the art, with a binding partner present on a leukocyte, can be used in the practice of the invention, by providing it on the solid phases of the apparatuses of the invention. As used herein, "rolling mediator" shall mean any molecule capable of interacting with a leukocyte so as to mediate rolling of the leukocyte on the surface containing the rolling mediator. In a preferred aspect, the rolling mediator is a selectin or a binding partner (ligand) of a selectin. Such rolling mediators include but are not limited to those ascribed in Table l, which also discloses the presently known cell subset specificity of binding partners for such rolling mediators (see also Springer, 1990, Nature 346:425-433; Springer and Lasky, 1991, Nature 329:196-197; Butcher, 1991, Cell 67:1033-1036): - 25 26 4 353 TABLE 1 ROLLING MEDIATORS FOR USE IN THE PRESENT INVENTION Cell Subset That Binding Rolling Mediator Partner la Present On MECA-79 (lymph node All leukocytes, in addressin) particular L, N, M E-selectin (ELAM-1) N, M, smTL P-selectin (GMP-140, CD62, N, M PADGEM) N, neutrophils, M, monocytes; L, lymphocytes; sMT, j a subset of T lymphocytes and memory T lymphocytes Functional derivatives of the foregoing rolling mediators can also be used.

The binding partner for MECA-79 (lymph node addressin) (Berg et al., 1991, J. Cell Biol. 114:343) is the homing receptor selectin, also called LAM-1, LECAM-1, or L-selectin, which is expressed on all leukocytes and facilitates lymphocyte binding to endothelium during blood circulation through peripheral lymph nodes and lymphocyte and neutrophil binding to endothelium at inflammatory sites. The ELAM-1 glycoprotein is synthesized by endothelial cells in response to inflammatory agents and promotes adhesion of a variety of leukocytes. The granule membrane protein (CD62), also called PADGEM and GMP-140, is a granule-associated glycoprotein of platelets and endothelial cells that is brought to the cell surface after stimulation by thrombogenic agents, allowing platelets and endothelial cells to bind neutrophils and monocytes at the site of tissue injury. 264353 All known selectins have an N-terminal domain that is homologous to a variety of Ca+J-dependent animal lectins (thus the name selectin), one epidermal growth factor (EGF)-like module, and from 5 two to nine short consensus repeats. Molecules composed of short consensus repeats of the type found in selectins have random configurations as revealed by electron microscopy, suggesting a high degree of segmental flexibility. Location of the ligand for 10 selectin is at the termini of long carbohydrate structures on the leukocyte that also may confer flexibility. Thus, the high association and high dissociation rate constants of the selectin-ligand complex required for rolling may be facilitated by the 15 flexible nature of the selectin and its ligand.

The term "rolling mediators'* includes molecules that competitively block binding of neutrophils or other leukocytes to rolling mediators such as selectins. Selectin counter-structure ligands 20 have been identified (see Springer and Lasky, 1991, Nature 349:196-197) and certain molecules are now known to compete with selectins, especially the IAM-l glycoprotein, for binding sites and thus inhibit binding of selectins to neutrophils. These molecules 25 may thus contain binding sites that are identical, or closely related to, the neutrophil (or other leukocyte) binding site(s) on the selectin glycoprotein. These molecules include fucoidin, sulfatides, polyphosphomannose-ester (PPME) and 30 sulfated glucans and sulfated polysaccharides (e.g. - dextran sulfate, xylan sulfate) (see Skinner et al., 1991, J. Biol. Chem. 206:5371-74).

Rolling mediators can be obtained from any source known in the art, and are preferably purified 35 for use in the apparatuses of the invention. - 27 26 4 35 3 Purification can be carried out by standard methods commonly known in the art, including but not limited to chromatography (e.g., ion exchange, affinity, and sizing column chromatography), centrifugation, 5 differential solubility, or by any other standard technique for the purification of proteins.

For example, blood components containing selectins can be isolated from fresh blood using gel filtration (Larsen et al., 1989, Cell 59:305-312) or 10 dextran-sedimentation and density separation over Ficoll-Hypaque (Miller et al., 1987, J. Clin. Invest. 80: 535-544). Selectins can be purified from blood components by immunoaffinity chromatography, e.g., by binding to its binding partner or antibody). Briefly, 15 selectins are extracted from cell materials by . addition of detergent and, after centrifugation to remove organelles, the crude preparation can be applied to a chromatographic column containing anti-selectin antibody (see Larsen et al., supra). In a 20 particular embodiment relating to CD62 purification, CD62 expression on platelets provides a convenient source for biochemical isolation fid.). ELAM-1 can be purified as described (Lobb et al., 1991, J. Immunol. 147:124). MECA-79 can be purified as described (Berg 25 et al., 1991, J. Cell Biol. 114:343).

Recombinant DNA methods using well-known techniques can also be used to prepare purified selectins. Selectins on endothelium have been cloned and sequenced (see Springer, 1990, Nature 346:425). 30 Chemical synthesis may also be used, which can be - automated, e.g., by use of peptide synthesizers. .3. CHEMOATTRACTANTS FOR USE WITH THE ARREST MODEL APPARATUSES OF THE INVENTION Use of the arrest model apparatuses of the invention to achieve leukocyte arrest after rolling, - 28 26 4 3 5 3 comprises providing a chemoattractant to the rolling cell. Any chemoattractant having its receptor present on a leukocyte or leukocyte cell subset of interest, which functions in the arrest model to allow arrest of S the leukocyte or the subset thereof, can be used in the present invention. Such chemoattractants can include but are not limited to those presented in Table 2: TABLE 3 LEUKOCYTE CHEMOATTRACTANTS FOR USE IN THE PRESENT INVENTION Cell Subset That Chemoattractant Receptor Is Present On* Neut Neut Neut Neut Mon Mon, T subpop Mon Neut, Mon Neut, Mon, Eo, Mast, Baso Neut, Mon Neut, Mon, Eo, Baso, PI * Neut, Neutrophil; Mon, Monocyte; T subpop, T lymphocyte subpopulation'; Eo, Eosinophil; Mast, Mast cell; Baso, Basophil; PI, Platelet.

- Chemoattractant Intercrines IL-8 GRO/MGSA NAP—2 ENA-78 MCP-1/MCAF RANTES 1-309 Other Peptides formylated bacterial peptides (e.g., fMLP) C5a Lipid Derivatives Leukotriene B4 (LTB4) Platelet activating factor (PAF) In addition to those listed above, any chemoattractants known in the art can be assayed for # 264353 functional activity in and then used with the arrest model apparatuses of the invention. Such chemoattractants include but are not limited to the following: lymphokines (e.g., interleukin (IL)-1, IL-5 2, XL—4, etc.), collagen, fibrin fragments, oxidized lipid components from cell membranes, histamine (active on eosinophils) (Clark et al., 1975, J. Exp. Ned. 142:1462-1476), eosinophilotactic peptides (ECF-A) (active on eosinophils) (Goetzl and Austen, 10 1976, J. Exp. Med. 144:1424-1437), alveolar macrophage-secreted products (active on neutrophils) (Kazmierowski et al., 1977, J. Clin. Invest. 59:273-281; Hunninghake et al., 1978, Am. Rev. Respir.'Dis. 117:15-23), monocyte chemotactic protein CAP37 (PCT 15 Publication WO 91/00907, published January 24, 1991 by Larrick et al.), other monocyte chemotactic peptides (PCT Publication WO 90/08777, published August 9, 1990 by Yoshimura et al.), lymphocyte chemoattractant factor (LCF; active on CD4+ lymphocytes, monocytes, 20 and eosinophils) (Center and Cruikshank, 1982, J. Immunol. 128:2569-2574; Rand et al., 1991, J. Exp. Med. 173:1521-1528), casein (active on lymphocytes, monocytes, and polymorphonuclear leukocytes), cyclic GMP, 1,2-diacylglycerol (Wright et al., 1988, Proc. 25 Natl. Acad. Sci. USA 85:1869-1873), etc. (for lymphocyte chemoattractants, see Epps, 1982, Agents Actions 12 (Suppl.) 217-233; Berman et al., 1988, Immunol. Invest. 17 (8 & 9):625-677).

Chemoattractants for use should promote chemotaxis, 30 rather than chemokinesis.

The chemoattractants can be obtained from any source known in the art, and are preferably purified for use with the arrest model apparatuses of the invention. Purification can be by standard 35 methods known in the art, including but not limited to 2 6 4 3 5 3 chromatography (e.g., ion exchange, affinity, and sizing column chromatography), centrifugation, differential solubility, or by any other standard technique for the purification of proteins.

Chemoattractants can also be prepared by chemical synthetic techniques (e.g., by use of a peptide synthesizer), or by recombinant DNA methods by expression of a cloned chemoattractant nucleotide sequence. For purification procedures and amino acid 10 sequences of various intercrines, see Walz et al., 1991, J. Exp. Med. 1355—1362 (for a review, see oppenheim et al., 1991, Ann. Rev. Immunol. 9:617-648); regarding RANTES, see Schall, 1991, Cytokine 3:165). Various chemoattractants are commercially available. 15 For example, MCP-1 and RANTES can be purchased from Peprptech (Rocky Hill, NJ); FLMP, LTB4, and PAF are also commercially available. For purification and recombinant expression of C5a, see, e.g., Mand^cki et al, (1985, Proc. Natl. Acad. Sci. USA 82(11):3543-20 3547). .4. INTEGRIN BINDING PARTNERS IN THE ARREST MODEL APPARATUSES OF THE INVENTION Any molecule known in the art which is a binding partner of a molecule present on a leukocyte, and wftich mediates leukocyte arrest in the arrest model apparatuses of the invention when such molecule is presented on a solid phase surface of the arrest model, can be used therein. Such molecules are integrin binding partners, which include but are not limited to known Ig family members, normally present on endothelium, such as those listed in Table 3 (see Butcher, 1991, Cell 67:1033-1036): 26 4 353 TftBEE 3 INTEGRIN BINDING PARTNERS SUITABLE Integrin Bindina Partners Identity of Integrin Recocmized Leukocyte Subset on Which Intecrrin is Present** ICAM-1, ICAM-2 LFA-1 (aL/32) L > N, M ICAM-1, others Mac-l (aM/32) N, M, sT (to be identified) pl50, 95 (aX/32) N, M, ST VCAM-1 VLA-4 (a401) M, mL > VL fibronectin VLA-4 VLA-5 L, M sT, M laminin VLA-1 VLA-2 VLA-3 VLA-6 aL aT sL sT, M collagen VLA-1 VLA-2 aL aT ICAM, intercellular adhesion molecule; LFA-1, lymphocyte function-associated antigen-1; VCAM, vascular cell adhesion molecule; VIA, very late activation antigen.

N, neutrophils; M, monocytes; L, (B and T) lymphocytes; sT, subset of T lymphocytes; mL, memory T lymphocytes; vL, virgin T lymphocytes; aL, activated (B and T) lymphocytes; aT, activated T lymphocytes; sL, subset of (B and T) lymphocytes Functional derivatives of the foregoing integrin binding partners can also be used, as well as any other integrin binding partners known in the art.

Integrin binding partners are preferably purified prior to use in the arrest model apparatuses, 35 and can be obtained and purified by use of any method # 26 4 35 3 known in the art, e.g., by methods as described supra for the rolling mediators. In a particular embodiment relating to ICAM-1, ICAM-1 can be purified as described infra in Section 6.3.2 hereof.

Alternatively, cloned ICAM-1 (Staunton et al., 1988, Cell 52:925-933) can be expressed by recombinant methods known in the art. In another embodiment, a cloned nucleic acid encoding ICAM-2 or a functional derivative thereof, (see European Patent Application 10 Publication 387,688, published September 19, 1990 by Springer et al.) can be expressed to obtain ICAM-2. In another embodiment, VCAM-1 cDNA clones (Osborn et al., 1989, Cell 59:1203; Polte et al., 1990, Nucl. Acids Res. 18:5901) are expressed. .5. SOLID PHASES OF THE APPARATUSES OF THE INVENTION Many types of solid phases can be employed in the methods and apparatuses of the invention. Preferred solid phases are translucent materials such as glass, plastic, quartz, etc. Well-known solid phases include beads formed from glass, polystyrene, polypropylene, dextran, and other materials, tubes formed from or coated with such materials, or a fibrous support matrix made from one or more of these materials. In preferred embodiments, the solid phase includes planar shapes (e.g., plates, sheets) such as glass or plastic slides or coverslips. In a particularly preferred embodiment, the solid phase includes one or more solid or hollow glass fibers or tubes. These solid fibers or tubes can be packed within a housing or cartridge which preferably contains an inlet and an outlet so that blood containing leukocytes can flow through the cartridge in a unidirectional manner. In another embodiment, the solid phase is plastic. 26 4 35 3 One or more surfaces of the solid phase(s) can contain the rolling mediators, or rolling mediators and integrin binding partners, as the case may be.

The rolling mediator proteins and integrin binding partner proteins can be placed on the surface of the solid phase in any manner, e.g., by affixing or adsorbing then directly thereto, coating a solution or suspension containing the proteins on the surfaces, 10 etc., as long as the proteins are accessible to leukocytes flowing by the surfaces. In one embodiment, one or more of the desired rolling mediator and/or integrin binding proteins can be either covalently or non-covalently affixed directly 15 to the.solid phase by techniques such as covalent bonding via an amide or ester linkage or adsorption. In a particularly preferred method of binding* the rolling mediator and/or integrin binding partner is immobilized by incorporation into lipid bilayers. 20 Lipid bilayers can be prepared by any method known in the art, e.g., from liposomes. Preparation of liposomes can be achieved by any of several well known procedures (See, for example, Mimms et al., 1981, Biochemistry 20:833-840). In one embodiment, 25 one or more surfaces of the solid phase is coated with a liposome suspension. The liposome suspension is spread on glass substrates to form planar lipid bilayers containing the incorporated rolling mediators (see also Sections 6.3.3 and 6.3.4 intra), or rolling 30 mediators and integrin binding partners. In a preferred aspect, phosphatidylcholine is used in the preparation of liposomes. Other phospholipids can also be used, including but not limited to phosphatidylserine, phosphatidylinositol, and 35 phosphatidylethanolamine. Furthermore, the lipid 26 4 3 vesicles can also contain other lipid^soluble molecules such as cholesterol. By way of example but not limitation, a 7:2 ratio mixture of phosphatidylcholine to cholesterol can be used to make S lipid vesicles incorporating the rolling mediators and/or integrin binding partners of the invention (see e.g., Oustin and Springer, 1988, J. Cell Biol. 107:321). Lipid vesicles incorporating a rolling mediator, alone or in combination with an integrin 10 binding partner, can also be made as described in Smith et al. (1989, J. Clin. Invest. 83:2008-2017).

In another specific embodiment, the rolling mediators and/or integrin binding partners are affixed to a solid phase such as glass or plastic by direct 15 coating. For example and not by way of limitation, a rolling mediator or integrin binding partner in solution at a concentration in the range of 20-100 fig/ml in 1% octylglucoside detergent, can be diluted 1:10 in Tris-saline (pH 8.0), 2 mM MgCl2 at the time of 20 addition of the solution to a plastic or glass surface, followed by incubation for 16 h at 4°C to allow the protein to adsorb to the surface, followed by washing of the coated surface with 1% bovine serum albumin in Tris-saline (pH 8,0), 2 mM MgCl2 buffer 25 (see, Dustin and Springer, 1989, Nature 341:619).

In the practice of the present invention, rolling mediators and/or integrin binding partner proteins present in lipid-vesicle-coated solid phases are preferred over plastic surfaces coated directly 30 with such proteins, which are in turn preferred over glass surfaces coated directly with such proteins, since there is a higher background binding of leukocytes to the solid phase, when the solid phase is glass with directly adsorbed proteins, relative to 35 when the solid phase is plastic with directly absorbed 26 4 3 proteins, relative to when the solid phase is a surface coated with lipid vesicles containing such proteins. 5.6. THE APPARATUSES OF THE INVENTION One embodiment of the rolling model of the present invention is illustrated in Figure 1 which shows a solid phase 10 upon which is affixed one or more rolling mediators 12. A sample of blood 10 including components 14 (e.g., leukocytes or a subset thereof) which can bind with the rolling mediator and other components 16 (e.g., erythrocytes (red blood cells)) which cannot bind with the rolling mediator, are introduced under conditions providing relative 15 movement between the solid phase 10 and the sample, so that blood components 14 and 16 come into contact with the solid phase 10. Those components 14 that bind to the selectin will contact the solid phase and reversibly adhere thereto. Significantly, these 20 components 14 that bind with the rolling mediator will also begin to roll along solid phase 10 under physiologic flow conditions. The rate of rolling is very slow, on the order of microns per second. Components 16 that are not responsive to the rolling 25 mediator will continue to move rapidly across the solid phase without adhering to it, or tumble along the solid phase, as illustrated by the arrows shown associated with component 16. The rolling velocity of the adhering components 14 is several orders of 30 magnitude slower than the velocity of the freely moving components 16 and this serves as one basis for the separation and purification methods of the invention, as described infra. Components 16 that do not adhere to the solid phase 10 flow out and are 35 removed from the system. 264353 36 - In an embodiment of the invention directed to the arrest model, solid phase 10 also contains an integrin binding partner 17, in addition to the rolling mediator 12. In this embodiment, in the 5 presence of a chemoattractant, the components 14 that (i) bind * or otherwise interact) with the rolling mediator 12 and roll along solid phase 10, and (ii) express a receptor that binds the chemoattractant, and (iii) express an integrin that binds (or otherwise 10 interacts with) the integrin binding partner 17, will become arrested on the solid phase.

In a rolling model embodiment of the invention, an apparatus of the invention comprises (a) a solid phase having a plurality of rolling mediator 15 molecules on a surface of the solid phase; (b) inlet means for receiving a fluid sample and for permitting the sample to enter onto the surface of the solid phase; and (c) outlet means for permitting the fluid sample after it has flowed across at least a portion 20 of the surface of the solid phase to exit said surface. In an arrest model, the solid phase further comprises a plurality of integrin binding partners on its surface. The inlet and outlet means can be entrance and exit slots or holes, channels, tubes, 25 pipes, etc. The apparatuses can further comprise a means associated with the inlet means, for introducing the fluid sample into the inlet means, e.g., ports, injection systems, etc. The apparatuses can further comprise a means associated with the outlet means, for 30 removing a fluid sample from the outlet means, e.g., a pump. Other examples are described infra.

Figure 2 shows a schematic diagram of one embodiment directed to an apparatus of the rolling model of the invention. A housing 20 contains a 35 plurality of solid glass fibers 22 having affixed on - 37 264353 their peripheral surfaces 24 one or more rolling mediators 26. Solid glass fibers 22 can be formed by using well-known procedures. The thickness of the fibers should be as small as possible to maximize the 5 surface in contact with the sample, although the smallest possible fibers sure necessarily constrained by mechanical and manufacturing procedures. Blood 28 is introduced via inlet 30 into the housing 20 containing bundled fibers. The fibers may optionally 10 be separated from each other by a plurality of dividers or panels (not shown). The housing can be of any inert material including glass, plastic or other polymer such as polyte-trafluoroethylene (PTFE-Teflon®). Blood or other medium containing or 15 suspected of containing blood cells (e.g., whole blood, artificial blood, plasma, heparinized blood, or combinations or components of the foregoing, or other suspending vehicles for blood cells including buffers and the like) continually sweeps over the exterior 20 surfaces 24 of the fibers as the blood moves toward the outlet port 34. Blood can be fed under slight pressure so that the flow will exit the housing through outlet port 34. Alternatively, blood can be introduced into the housing at atmospheric or 25 subambient pressures in which the outlet port 34 is attached to a slight vacuum, the term "vacuum" meaning a pressure lower than the atmospheric pressure. Blood components that are capable of binding to the selectin(s) affixed to the bundled fibers will 30 reversibly adhere to them and roll. Blood components that do not adhere to rolling mediator(s) 26 on the bundled fibers are pulled across the exteriors of the fibers and exit the housing at port 34.

In another embodiment of the invention 35 directed to the arrest model, the peripheral surfaces 26 4 3 24 in Figure 2 also have affixed thereon one or more integrin binding partners, such that, in the presence of the appropriate chemoattractant, leukocytes which (i) express binding partners for the rolling mediator 5 26, (ii) express binding partners for the chemoattractant, and (iii) express the integrin recognizing the integrin binding partner, will stop and become arrested on the surfaces 24.

Relative movement of the solid phase and the 10 sample containing or suspected of containing blood cells is accomplished using a variety of methods. For example, bulk flow (i.e., flow of both fluid and particles in the fluid) of a sample can be accomplished by pumping the sample across one or more 15 surfaces of a solid phase containing rolling mediators (or both rolling mediators and integrin binding partners). A peristaltic pump or syringe pump is preferred for this purpose. Relative movement between sample and solid phase can also be provided by 20 capillary action which will draw the sample across one or more surfaces of a solid phase.

Other physical methods that do not rely on bulk flow can include electrophoretic methods in which certain blood cell subpopulations (e.g., T and B 25 cells, neutrophils and monocytes) which differ in their electrophoretic mobility can be exposed to an electric field. The leukocytes will be charged accordingly and will be induced to move within the fluid across an appropriate solid phase. 30 Other embodiments of the apparatuses of the invention are described infra, e.g., in Section 5.7.1. .7. METHODS OF BLOOD COLLECTION, PURIFICATION AND ANALYSIS The apparatuses of the invention can be used in various methods relating to the collection, - 39 ~ 2643 purification, concentration, and analysis of blood and blood cells.

In one embodiment, the apparatuses of the invention are used to effect the collection of 5 leukocytes or cell subset(s) thereof. In this embodiment employing a rolling model apparatus, the collection of the leukocytes is accomplished based on the cells' differential mobility (slower speed) in passage across the solid phase of the invention, due 10 to the cells' rolling arising from interaction with the rolling mediator molecules. That is, leukocytes which express a cell-surface binding partner for a rolling mediator (e.g., selectin) present on a solid phase of a rolling model apparatus can be collected 15 from a "medium flowing at physiologic flow rates through the rolling model apparatus, by removing the solid phase while such leukocytes are still reversibly adhering and rolling thereon. Alternatively, such cells can be collected by collecting medium after 20 passage over the solid phase at a time subsequent to passage of the non-rolling blood components; however, this method is not preferred, due to the long time period required for passage of the rolling leukocytes over the solid phase (see infra). The type(s) of 25 leukocyte that are collected depends on the cellular specificity of the binding partners for the rolling mediator(s) present on the solid phase (see Table 1, supra).

In an embodiment for collection of 30 leukocytes or cell subset(s) thereof employing an arrest model apparatus, a chemoattractant (or a plurality of chemoattractants) is introduced, preferably by addition to the medium flowing through the apparatus, without stopping flow. Rolling 35 leukocytes that are activated by the chemoattractant, 26 4 3 and that express an integrin on their cell surface which interacts with its binding partner on the solid phase of the arrest model apparatus, will become arrested on and strongly adherent to the solid phase.

The solid phase can then be removed, and the bound cells eluted for subsequent use (e.g., by exposure to chelating agents such as citrate, EDTA, EGTA, etc.). The type of leukocyte that is collected depends on the cellular specificity of the binding partners for the 10 rolling mediator(s), chemoattractant(s), and integrin binding partner(s) that are used (i.e., what type of leukocyte expresses binding partners for these molecules).

For example, in a specific embodiment 15 directed to a collection method of the invention employing a rolling model apparatus, a blood sample containing leukocytes and a solid phase containing at least one affixed rolling mediator protein are moved relative to each other under physiologic flow 20 conditions. Preferably, the rolling mediator is the selectin CD62. This selectin is specific for neutrophils and monocytes and will not bind to lymphocytes or red blood cells. Nevertheless, it will be understood that a different rolling mediator or 25 more than one rolling mediator can be affixed to the solid phase in order to bind with, and cause adherence of, a different type or a plurality of leukocyte subsets, depending on the cellular specificity of the rolling mediator's binding partner (see Table 1). 30 Moreover, as described supra, a plurality of rolling mediator analogs can also be affixed to the solid phase, such as sulfated glycans, fucoidin, or PPME (polyphosphomannose-ester).

The medium (e.g., blood sample) containing 35 leukocytes is allowed to flow over the substrate at # 264353 - ;i - flow rates sufficient to induce a shear stress preferably of between about 0.5 and about 2.0 dynes per square centimeter. At these flow rates, rolling will be induced in the leukocytes that come into 5 contact with the rolling mediator affixed to the solid phase. In vitro rolling velocities of leukocytes on CD62 are comparable to in vivo rolling velocities, as described below in the examples sjtions. Velocities generally range from under 2 to over 30 microns per 10 second, depending on the site density of selectin and the shear stress. These velocities are several orders of magnitude slow than the fluid velocities necessary to induce the physiologic shear stress. Because these rolling velocities are so small, the rolling 15 leukocytes will not traverse the solid phase of the. apparatus under most conditions. For example, leukocytes rolling between 2 and 30 microns per second will traverse a 10 centimeter long substrate in about 1 to 15 hours, for longer than most convenient 20 collection procedures.

After a sufficient number of leukocytes have been accumulated and reversibly adhered to the solid phase, the flow of blood is interrupted and the solid phase is washed to remove any non-adherent material. 25 A particularly preferred washing solution includes saline that contains calcium ions. Calcium ions are preferred because selectins have a N-terminal domain which is homologous to a variety of Ca2+-dependent animal lectins and, therefore, selectin binding is 30 calcium-dependent.

After non-adherent material has been washed from the surface of the solid phase, adhering, rolling, leukocytes can be eluted from the solid phase by using a chelating agent such as citrate or 35 ethylenediaminetetraacetic acid (EDTA) to bind calcium # 26 4 35 3 ions, thus enabling leukocytes to be released from the solid phase. In a preferred embodiment involving CD62 as the selectin, neutrophils and monocytes can be thus collected.

The above-described embodiment can also be adapted to collect cells using an arrest model apparatus, preferably containing CD62 as the rolling mediator and ICAM-1 as the integrin binding partner and with the use of fMLP as the chemoattractant, for 10 the collection of neutrophils and monocytes.

The above collection methods can also result in and be used for the concentration, purification, and/or quantification of the rolling or arrested leukocytes. Quantification can be by various methods 15 known in the art, including visualization of rolling cells, or staining methods, as described infra.

In the collection methods of the invention, it is not necessary that all the leukocytes within the flow medium that are of the cell subset(s) being 20 collected come into contact with the solid phase, and are actually collected thereby, although the higher the percentage that are thus collected, the more efficient the procedure. In one embodiment, the collection methods of the invention can be used to 25 extract substantially all of the leukocytes or a cell subset thereof from agiven volume of medium. Preferably, the total number of leukocytes remaining in the bulk sample is less than 1 x 106. The procedure and apparatus of this mode are identical to that 30 described above for collection, except that conditions are chosen so that essentially all of the leukocytes will come in contact with the solid phase. For example, in the flow apparatus of Figure 2, the distance between the fibers 22 can be adjusted by 35 altering the dimensions and spacing of the fibers. # 2 6 4 3 5 3 Blood components are then forced to pass through narrower gaps between fibers, which gaps preferably . • . are in the range of from about 10 to about 100 jim (somewhat larger than the width of a typical 5 leukocyte, i.e., 6-7 pim). Moreover, the flow rate can be controlled so that the shear stress is sufficient to allow all the leukocytes to come into contact with the rolling mediator on the solid phase. A preferred shear stress is in the range of from 0.5-2 dynes/cm2. 10 Furthermore, the site density or concentration of selectin on the solid phase can also be increased to accomplish the same ends. A site density of up to about 1,000 sites/inn2 can be used, with 100 sites//m2 the preferred site density. These and other design 15 manipulations are well within the knowledge of those skilled in the art.

It will be appreciated that extraction of substantially 100% of given leukocytes or a cell subset thereof in a blood sample or other medium will 20 also serve to concentrate and purify away from such leukocytes the remaining cells (e.g., erythrocytes) and other components that exit from the flow apparatus. Thus, the procedures of the invention can be used to both collect, concentrate and/or purify 25 leukocytes and to concentrate and separate away or purify the remaining components in the flow medium. Both the leukocytes and/or these remaining components can then be used for further testing and/or for transfusion procedures.

The collection and purification methods of the invention are useful in the clinical or laboratory setting. The volume and numbers of leukocytes per volume of blood can be determined by quantitation of the rolling or arrested leukocytes and by measuring 35 the volume of flowing blood. Alternatively, the 2 6 4 3 5 3 leukocytes can be eluted from a known blood volume using the apparatus of the invention and their total numbers separately determined using well known methods. .7.1. ANALYSIS AND QUANTITATION In one specific embodiment of an analysis aethod particularly useful in clinical and diagnostic tests (see Section 5.9 infra) and employing a rolling 10 model apparatus, glass slides containing planar bilayers having one or more rolling mediators immobilized therein are used in a parallel plate flow chamber 40, as illustrated in Figure 3. The chamber is shown upside down for illustration purposes; under 15 normal -conditions the chamber is placed on the stage of an inverted microscope. Glass slide 42 with an artificial lipid layer 44 containing one or more immobilized rolling mediators forms the lower parallel plate. A base 46, typically formed of polycarbonate, 20 contains both inlet manifold 48 and outlet manifold 50. These manifolds are in communication with a deck 52 consisting of an inert material such as glass or polycarbonate. The inlet and outlet manifolds 48, 50 are also in communication with respective entrance and 25 exit slots 54, 56 in the deck. Base 46 contains a recessed lip 58 on an upper surface 59 of the base into which is placed a compressible gasket 60 having an aperture 62, which aperture being sized and shaped to be congruent with the deck 52. Glass slide 42 or 30 other similar planar surface with an artificial bilayer containing one or more rolling mediators is placed on the gasket in communication with the gasket aperture. A vacuum source 64 is connected to the base 46 and this vacuum source applies enough force 35 compress the gasket 60 and slide 42 together and # 264 353 prevent their disengagement from the rest of the base assembly. A cell suspension is connected to the inlet manifold and preferably a syringe pump (not shown) is connected to the outlet manifold.

Another embodiment of an analysis method uses a solid substrate affixed with a plurality of different rolling mediators. Each rolling mediator can be confined to a distinct area or zone of the solid substrate. The rolling mediators are chosen so 10 that each zone will allow adhesion of a different leukocyte cell subset or subsets (e.g., monocyte, neutrophil, eosinophil, basophil, and the like). A blood sample is obtained from a patient and a small amount of blood or other medium containing blood cells 15 is allowed to contact the solid substrate. Leukocytes of various types will bind to the respective rolling mediators. Preferably, the flow can be generated by a syringe or syringe pump, as described above and allowed to flow across the parallel plate chamber of 20 Figure 3. Alternatively, the substrate, which can be of small size, can be moved through the blood sample to provide relative movement. This is most conveniently done by attaching the solid phase to an elongated member (e.g., a "dipstick") and moving the 25 substrate back and forth within the sample at controlled velocities. Different leukocytes flowing across the substrate will roll and reversibly adhere to their respective rolling mediators and will be physically separated on the substrate. The substrate 30 can be washed, as above, to remove unbound material and the adhering leukocytes stained using fluorescent or other labels well known in the art (e.g., Wright's Crimson).

A further embodiment of an analysis method 35 that utilizes a solid substrate affixed with a 46 - 2 6 4 3 5 3 plurality of different rolling mediators can include, as described above, a solid phase in which rolling mediators are confined to distinct areas or zones. In particular, the solid phase can include one or more 5 rolling mediators chosen so that substantially all of the leukocytes are concentrated in a particular zone of the solid phase. This initial concentration of leukocytes can take place within a so-called "starting zone." After cells are concentrated onto the starting 10 zone, they can be further separated and analyzed using, for example, subsequent electrophoresis onto a solid phase containing a plurality of different rolling mediators. The different leukocyte populations will then be separated on this so-called 15 "separation zone." Cells can then be counted and analyzed on the separation zone using a variety of conventional methods. For example, the separation zone can be scanned and the absorbance determined, which absorbance is a function of the leukocyte 20 population density. Further, the solid substrate can be viewed under a microscope and numbers of leukocytes, separated by leukocyte subset into distinct "zones", can be counted. The slide can be preserved and cells further examined by microscopy if 25 desired.

The foregoing methods of analysis can also be adapted for use with an arrest model apparatus, with provision of the appropriate chemoattractant(s) and integrin binding partner(s). 30 The apparatuses and methods of the invention can also be adapted to provide for visual analysis of leukocyte rolling velocities. In one embodiment, the parallel flow apparatus of Figure 3 is attached to a microscope stage, and the blood components flowing 35 across the surfaces are viewed through the microscope # 264353 objective by way of a video cassette recorder (VCR). Images are recorded on a time-lapse VCR at real time and then played back at slower speed. The location of cells can be marked at any given time and the location 5 of the individual cell is determined at some finite time afterwards.

Visualization of the rolling of different types of leukocytes is thus possible using a VCR in combination with the magnification power of the 10 microscope. The kinds and numbers of leukocytes that rollingly adhere to a substrate can be determined and their individual velocities can be recorded. Optionally, the analysis is completed by constructing a frequency histogram of the numbers of leukocytes (or 15 a cell'subset thereof) rolling at a particular velocity. This analysis also results in a determination of the rolling rate of a particular kind (subset) of leukocyte. Rolling leukocytes are visualized as round, distinct cells by the video 20 camera because their rolling velocity is on the order of microns per second, which roiling velocity is extremely slow compared to the exposure time of the VCR camera (typically 30 frames per second). By comparison, non-adherent blood components flowing and 25 tumbling through the apparatus tumble in the shear flow and appear as blurred streaks. Typically, nonadherent cells and those tumbling closest to the solid phase have a velocity of about 500 microns per second. Confirmation of rolling can be determined by 30 visualizing the nuclei of the leukocytes under magnification. Determination of the rotating position of the nuclei distinguishes between leukocytes that merely slide along the surface of the substrate and those that rollingly adhere. # -• 48 - 26 4 35 3 Leukocytes in a given medium, e.g., blood, can optionally be differentiated from erythrocytes prior to introducing the sample into the flow chamber by labelling the leukocytes, e.g., with acridine S orange or quinicrine dihydrochloride (see Nobis et al., 1985, Microvasc. Res. 29:295-304).

Alternatively, erythrocytes can be lysed prior to introduction of the labelled leukocytes into a flow chamber of the invention (e.g., by NHjCl lysis; see 10 e.g., 1991, Current Protocols in Immunology, ch. 3.1, J.E. Coligan et al., eds., John Wiley & Sons). The leukocytes are irradiated under ultraviolet (UV) light. The acridine orange or quinicrine dihydrochloride stain will absorb this radiation and 15 emit light at visible wavelengths (510-540 run) . The positions of leukocytes on the substrate are compared under UV irradiation in successive frames of the VCR recording to determine rolling velocities, and a histogram of velocities is constructed. Each peak in 20 the histogram is associated with a specific cell population that roHXis at a particular rate. Other methods of labelling Leukocytes or cell subsets thereof can also be used, e.g., by use of an antibody directed to a leukocyte cell surface antigen, which 25 antibody is labeled With a detectable marker.

In a further embodiment of an analysis method, a quartz cell with a hollow bore having an internal diameter between about 50 and about 500 microns is provided. The hollow bore is brought into 30 contact with rolling mediators so that one or more rolling mediators are affixed to the inner wall of the hollow bore. Leukocytes and erythrocytes can be physically separated by lysing the erythrocytes as described above, or leukocytes can be stained with a 35 fluorescent stain, thus differentially tagging them. 26 4 353 A blood sample containing leukocytes to be differentiated is allowed to flow through the chamber affixed with one or more rolling mediators/ and light of sufficient wavelength to excite the fluorescent S label of the leukocytes is directed at the chamber. Emission of light from the rolling cells is determined. The total time interval over which fluorescent energy is received from the excited label on the leukocyte will be a function of its rolling 10 velocity. This is because the light beam width can be made constant (preferably about 50-100 microns2) and the amount of time required for any individual leukocyte to traverse the beam is the beam width divided by the cell velocity. In particularly 15 preferred embodiments of this analysis method, laser light is used. Non-coherent light is also suitable.

Accordingly, one embodiment of a laser-based method comprises providing fluorescently tagged leukocytes and flowing the leukocytes through a tube, 20 the inner surfaces of which contain one or more rolling mediators. The sample is irradiated with laser light and the duration of fluorescent light emitted by each cell of the sample is measured as the cell flows in a direction perpendicular to the optical 25 axis of the laser beam. Leukocytes can be discriminated on the basis of the distribution of signals that are indicative of the intensities of the sensed fluorescent light. The value of emitted light data can be stored using A/D converters in a computer 30 system, or directly visualized with a video camera, such procedures being well known to those of ordinary skill in the art.

These cytometric methods can be calibrated by comparing the velocity of a given, but unknown type 35 of leukocyte, to the velocity of a known type of 26 4 35 3 leukocyte. The methods can also be used to produce a frequency histogram of leukocyte rolling velocities, which can be used in a diagnostic context, as described infra. .8. KITS The apparatuses for carrying out the methods of the invention can be conveniently arranged into kits for use in clinical laboratories and other 10 settings. For example, an apparatus can comprise a solid phase, optionally contained within a housing. This solid phase contains one or more rolling mediators affixed to it, as described above (and, in an arrest model, one or more integrin binding 15 partners). The housing can include a means for withdrawing the sample from an outlet end and a means for introducing sample into an inlet end. In one embodiment of a kit, the apparatus can contain a plurality of solid fibers upon whose cylindrical 20 surfaces are immobilized one or more rolling mediators (and, in an arrest model, one or more integrin binding partners). The fibers can be contained within a cartridge or housing with a means for introducing a fluid sample (e.g., of blood) into the housing and a 25 means for withdrawing the sample from the housing so that the sample flows over the external surfaces of the fibers. The size of fibers, their number, and the site density of rolling mediator can be adjusted to provide for a cartridge having different capacities 30 for accumulating leukocytes, or cell subsets thereof.

Apparatuses for use in diagnostic contexts can also be arranged into convenient kits. For example, a solid phase, such as a microscope slide or coverslip, can be provided that contains a plurality 35 of rolling mediators arranged in spatially distinct # 26 4 35 3 zones. The solid phase can be attached to a dipstick or other arrangement such as the parallel flow chamber described above, to provide for relative movement r0j between the solid phase and the sample. In a related embodiment, the solid phase can contain both rolling mediators and integrin binding partners in spatially distinct zones, for use with the addition of a chemoattractant, to promote affixing of the cells, in an arrest model.

Another embodiment of a kit of the invention includes, in one or more containers, the following components of an apparatus of the invention: (a) liposomes or other artificial lipid bilayer components or precursors (e.g., phospolipids such as 15 phosphatidylcholine); and (b) rolling mediator molecules. In a preferred aspect, such a kit comprises, in one or more containers, liposomes and CD62. In an embodiment relating to the arrest model, a kit of the invention comprises, in one or more 20 containers or precursors: (a) liposomes or other artificial lipid bilayer components or precursors; (b) rolling mediator molecules; and (c) integrin binding partners. Such a kit can further comprise a chemoattractant in a container. In a preferred 25 aspect, such a kit comprises, in one or more containers: liposomes, CD62, ICAM-1, and the chemoattractant fMLP. .9. DIAGNOSTIC AND THERAPEUTIC UTILITIES OF THE METHODS OF COLLECTION, 3V PURIFICATION, AND ANALYSIS OF BLOOD AND BLOOD COMPONENTS The methods of collection, purification, and analysis of blood and blood components provided by the present invention have diagnostic and therapeutic utilities. 264 353 The collection methods can be used to determine the number of leukocytes or cell subset thereof in a sample of blood or other fluid derived from a patient, or the number of leukocytes or cell 5 subset thereof which roll at particular velocities in such a sample. Such number can be compeared to the amount present in an equivalent sample from a normal or healthy subject, or a subject in remission from a disease or disorder, or the same patient at an earlier 10 time period, whereby increases or decreases relative to such amount indicate the presence or progression of a disease or disorder, the stage of the disease or disorder, or the response to therapy in such patient, thus providing methods of detection, diagnosis, 15 staging, and monitoring of treatment. Similarly, the methods of the invention can be used to determine a percentage, consisting of the number of leukocytes or cell subset thereof per number of one or more other cell types in the sample, or consisting of the 20 percentage of leukocytes rolling at a particular velocity, which percentage can be similarly compared to the percentage in an equivalent sample from subjects as described above, to detect, diagnosis, stage, and monitor treatment of diseases and 25 disorders. In one embodiment, the numbers of total leukocytes per volume of blood can be used diagnostically. The numbers of leukocytes from a patient with an unknown disease can be compared to the numbers of leukocytes in a blood sample that is 30 characteristic of a disease. Comparison of the two samples can determine the presence or absence of the disease in the patient from whom the original sample is derived.

The diseases or disorders suitable to 35 diagnosis, staging, and/or monitoring in the foregoing # 26 4 3 5 3 methods include those in which there is a disturbance in the normal amount of a leukocyte or cell subset thereof, so that the changed amount is characteristic of a diseased condition. Generally, the number of 5 circulating white blood cells (leukocytes) may be markedly decreased or increased in a variety of clinical disorders.

Disorders involving an increase in leukocytes or a subset thereof include but are not 10 limited to hematologic malignancies such as leukemias, lymphomas/ e.g., acute and chronic myeloid and lymphatic leukemias, including chronic myelogenous leukemia, adult T cell leukemia, non-Hodgkins lymphoma, chronic lymphatic leukemia, plasma cell 15 myeloma, etc. In addition, increased levels of eosinophils may occur in asthma or parasitic infections.An increase in the number of white blood cells in circulating blood is due mainly to granulocytosis. The most extreme and important 20 increases are, however, encountered in the various leukemias.

Diseases and disorders involving a decrease in leukocytes or a subset thereof include but are not limited to leukopenias such as neutropenia, Hodgkin's 25 disease, etc. An abnormally low white blood cell count (leukopenia) may occur because of decreased numbers of any one of the specific types of leukocytes within the circulating blood, but most often leukopenia involves the neutrophils. Low lymphocyte 30 . counts are much more common and are associated with specific clinical syndromes (e.g., Hodgkin's disease, non-lymphocidic leukemias). Other diseases or disorders involving decreased levels of T lymphocytes incl ie the immunodeficiency disorders congenital 35 thymic aplasia (DiGeorge syndrome) and severe combined 264353 immunodeficiency disease (in which both T and B cells may be completely absent), as well as AIDS (Acquired Immune Deficiency Syndrome; involving a decrease in CD4+ T cells). 3 Additionally, the histograms described in Section 5.7.1, jUlSEA/ can be used in a diagnostic, staging, or monitoring context. In a specific embodiment, blood samples taken from subjects known to have one or more of the foregoing diseases can be 10 analyzed using the methods of this invention to produce a velocity histogram, and the resulting histogram provides a rolling velocity "fingerprint" that is characteristic of the disease. A blood sample from a patient having em unknown disease is then 15 analyzed using the methods of this invention, and the velocity histogram is compared to this standard "fingerprint" to determine the presence or absence of the disease in the unknown blood sample.

Methods designed to differentiate leukocytes 20 based on rolling induced by interactions with different types of rolling mediators (optionally, in conjunction with arrest induced by chemoattractants and integrin binding partners), also provide a convenient method of determining numbers of different 25 leukocytes.

In addition, apparatuses and methods of the invention can be used in a therapeutic context to monitor the progression of a disease or the progression of therapy by taking frequency spectra or 30 . leukocyte rolling velocities from blood samples over a period of time.

In another embodiment, the methods of collection, purification, and analysis of the invention have utility for the diagnosis of diseases 35 and disorders involving a defect in leukocyte rolling, 2 6 4 3 5 3 chemoattractant activation, and/or arrest at the endothelial cell wall, by detecting a decrease in the percentage of leukocytes or a cell subset thereof from a patient which are able to roll, or roll and arrest, 5 in the apparatuses of the invention, relative to such percentage of leukocytes from a healthy patient. For example, leukocyte adhesion deficiency (Anderson and Springer, 1987, Ann. Rev. Med. 38:175-194) involves an inherited deficiency in the integrins LFA-1, Mac-l, 10 and pl50,95, resulting in deficient adherence of granulocytes, monocytes, and lymphoid cells, which can be thus detected in an arrest model apparatus (preferably using ICAM-1 as the integrin binding partner). In addition, diabetes mellitus, 15 granulocytasthenia, and recurrent pyogenic infections have been reported to involve cell adherence defects (see Gallin et al., 1980, Ann. Int. Med. 92:520-538). Genetic defects in leukocyte rolling mediator binding partners, chemoattractant receptors, and/or integrins 20 that interfere with the rolling and/or arrest processes are capable of detection by the methods of the invention. For example, patients whose leukocytes or a subset thereof lack the binding partner or binding partner determinant thereof necessary for 25 interaction with a rolling mediator can be diagnosed by the methods of the invention. In particular, patients whose leukocytes lack GDP-fucose and hence do not bear sialylated Lewis x and therefore cannot bind to E-selectin or CD62, can be diagnosed. 30 In another embodiment, leukocytes or a subset thereof from a subject, or blood components obtained after passage through an apparatus of the invention and thus deleted of such cells, can be analyzed in vitro by carrying out diagnostic tests 35 known in the art, e.g., analysis for expression of 26 4 3 certain cell-surface antigens (e.g., associated with a malignancy or with infection by a pathogenic microorganism), karotype analysis, etc.

In another embodiment, the invention can be S used ex vivo in a manner similar to a blood phoresis device. Blood is taken directly from a patient's blood vessel and is allowed to flow into the inlet port of the apparatus of Figure 2. Various valving arrangements that are well-known to those of ordinary 10 skill in the art may be attached to one or another end of the apparatus to provide for shunting of blood flow away from the apparatus during washing and elution of leukocytes.

In a specific embodiment, citrate IS anticoagulated whole blood derived from a patient in a clinical setting can be introduced into the apparatuses of the invention and analyzed as described above.

The collection and purification methods of 20 the invention also have therapeutic utility, by providing collected, purified, and/or concentrated blood components which can be administered to a patient, in a transfusion procedure. For example, leukocytes or a subset thereof isolated by the methods 25 of the invention can be administered to a patient suffering from decreased levels of such leukocytes. Alternatively, fluid passed through an apparatus of the invention and thereby substantially deleted or decreased in leukocytes or a subset thereof can be 30 administered to a patient in need of blood or blood components but for whom administration of such leukocytes is not desirable.

The therapeutic and diagnostic methods described herein have application to subjects that are 26 4 353 preferably mammals, including cows, dogs, pigs, and most preferably, humans. .10. IDENTIFICATION OF INHIBITORS OR PROMOTERS OF THE INFLAMMATORY RESPONSE The apparatuses and methods of the present invention can be used for the identification of inhibitors (e.g., antagonists) or promoters (agonists/functional components or enhancers) of the adhesion receptor-mediated migration of leukocytes through the endothelium (extravasation). Such inhibitors and promoters respectively inhibit and promote the inflammatory response, and thus have therapeutic utilities. The inhibitors and promoters are identified by detecting their abilities to respectively inhibit or promote the rolling of leukocytes or a cell subset thereof in a rolling model apparatus of the invention, or to respectively inhibit or promote the rolling and arrest of leukocytes or a cell subset thereof in an arrest model apparatus of the invention.

Since the present invention provides a simplified physiologically relevant model of the processes leading to leukocyte extravasation, the models of the invention can provide for the identification of inhibitors and promoters of the inflammatory response that are therapeutically useful in vivo. .10.1. IDENTIFICATION OF INHIBITORS The methods of the invention can be used to identify an inhibitor of the inflammatory response which acts by inhibiting one or more of the steps involved in leukocyte rolling (by use of a rolling model apparatus) or leukocyte rolling and arrest (by use of an arrest model apparatus) in an apparatus of 2 6 4 3 5 3 the invention. For example, inhibitors thus identified can inhibit a rolling mediator-binding partner interaction, a chemoattractant-chemoattractant receptor interaction, and/or an integrin-integrin 5 binding partner interaction. The inhibition may be competitive or non-competitive.

An inhibitor of the inflammatory response is thus identified as follows. In a rolling apparatus, the rolling mediator(s) incorporated onto the solid 10 phase surface (s) , and the flow medium containing leukocytes, are selected so as to provide for rolling of the leukocytes or one or more cell subsets thereof on the solid phase (s). The molecule to be tested is introduced into the flow medium prior to passage 15 though the apparatus, or preferably, during passage, when leukocytes have begun rolling. A decrease in rolling (e.g., as measured by a decrease in the percentage of leukocytes that are rolling, or a decrease in their velocity, or a decrease in the 20 number of rolling leukocytes per volume) in the presence of the molecule, relative to that observed in the absence of the molecule, indicates that the molecule is an inhibitor of leukocyte extravasation, a component of the inflammatory response. 25 For identification of an inhibitor, in an arrest model apparatus, the rolling mediator(s) and integrin binding partner(s) incorporated onto the solid phase surface(s), the chemoattractant(s) introduced into the flow medium, and the flow medium 30. containing leukocytes, are selected so as to provide for rolling and arrest of the leukocytes or one or more cell subsets thereof on the solid phase(s). The molecule to be tested is introduced into the flow medium prior to passage through the apparatus, or 35 during passage, when leukocytes have begun rolling. A 2 6 4 3 5 3 decrease in arrest of leukocytes (e.g., as measured by a decrease in the percentage of leukocytes that are arrested, or in the number of arrested leukocytes per volume) in the presence of the molecule, relative to 5 that observed in the absence of the molecule, indicates that the molecule is an inhibitor of leukocyte rolling, chemoattractant activation, and/or arrest, and thus an inhibitor of leukocyte extravasation, a component of the inflammatory 10 response.

Molecules to be tested for inhibitory activity can be any of interest, including but not limited to antibodies (preferably monoclonal, most preferably human or humanized monoclonal, or antigen-15 binding domains thereof) to any member of the following receptor-ligand pairs: rolling mediator-binding partner; chemoattractant-chemoattractant receptor; integrin-integrin binding partner, or a neutralizing epitope thereof, e.g., sialylated Lewis x 20 or GDP-fucose; peptide antagonists and peptidomimetics, etc.

A preferred specific embodiment for testing numerous compounds for inhibitory activity is as follows: As cells are continuously rolling on 25 E-selectin, P-selectin, or the ligand of L-selectin in a parallel plate flow chamber of a rolling model apparatus of the invention, a test compound is injected for some duration of time just upstream of the observation point. The flow brings the compound, 30 with some dilution factor (approximately 2-fold), into the area of observation. Its effect on the percentage of rollingly adherent cells, or their number per volume flow medium, or their velocity, is then measured. Injection is stopped for a sufficient 35 length of time to allow the compound to flow M 26 4 3 53 downstream, and for cells to accumulate from upstream and for rolling to reach equilibrium again, then a new compound is injected and the process is continuously repeated. With a cycle time of 2 minutes, it is 5 possible to screen 720 compounds per flow chamber per day. Multiple ports for simultaneous injection of compounds, with injection of buffer in intervening ports to keep the compounds separate in distinctive streams, may be used to increase throughput. The 10 ports are placed perpendicular to the direction of flow and thus the streams are parallel. Cells rolling in each stream may be visualized using multiple microscope objectives or by moving the same objective from one area to another. " .10.2. IDENTIFICATION OF PROMOTERS The promoters of the inflammatory response detected according to the present invention can act by increasing the efficiency of the adhesion process in 20 an apparatus of the invention, or by acting as a functional component thereof (e.g., a rolling mediator, chemoattractant, or integrin binding partner). Such a functional component is detected by its ability to promote rolling or arrest in a model 25 where this was previously lacking (e.g., due to lack of appropriate cellular specificity of a rolling mediator or integrin binding partner previously present in the apparatus, or lack of any rolling mediator or integrin binding partner). For example, 30 an arrest model apparatus can be used to identify a compound which is a chemoattractant or to identify an integrin binding partner, functional in leukocyte extravasation. A rolling model apparatus or an arrest model apparatus can be used to identify an endothelial # - 61 ~ 2 6 4 3 5 3 cell rolling mediator functional in leukocyte extravasation.

To identify a chemoattractant, an arrest model apparatus is used, in which the rolling mediator 5 and integrin binding partner present on a solid phase thereof have binding partners situated on the leukocytes present in the flow medium to be passed through the apparatus. A sample containing the molecule to be tested for chemoattractant activity is 10 introduced into the flow medium passing through the apparatus, and it is determined whether any leukocytes are arrested on the solid phase. Arrest of leukocytes in the presence of the molecule, but not in its absence, indicates that the tested molecule is a 15 chemoattractant, and that the leukocytes arrested oh the solid phase express cell surface receptors for the molecule (in addition to expressing receptors for the rolling mediator and integrin binding partner utilized), thereby activating integrin-mediated 20 arrest. In a preferred aspect, CD62 and ICAM-1 are used as the rolling mediator and integrin binding partner, respectively, to assay for chemoattractants with receptors on neutrophils and/or monocytes.

It should be noted that chemoattractants can 25 be used as either inhibitors or promoters of the inflammatory response depending on how they are administered. For example, a chemoattractant gradient directing leukocytes toward a specific tissue is expected to be pro-inflammatory at such tissue, 30 whereas general systemic administration of a chemoattractant is expected to be inhibitory to leukocyte extravasation, since the systemically administered chemoattractant would competitively inhibit leukocyte recognition of chemoattractant 35 gradients directing its migration toward tissues. * 26 4 35 3 To identify an integrin binding partner, an arrest model apparatus is used, in which the rolling mediators on a solid phase thereof have binding partners situated on the leukocytes present in the 5 flow medium to be passed through the apparatus. One or more chemoattractants are introduced into the flow medium. A chemoattractant is used which has a receptor on the same subset (s) of leukocytes that express the rolling mediator binding partner. A 10 compound or molecule to be tested for integrin binding partner function is affixed onto the surface of the solid phase. After the flow medium is passed through the apparatus, it is determined whether any leukocytes have arrested on the solid phase. Arrest of 15 leukocytes indicates that the molecule is an integrin binding partner which recognizes an integrin present on the same leukocyte subset(s) that express the rolling mediator binding partner and the chemoattractant receptor.

To identify a rolling mediator by use of a rolling model apparatus, the molecule to be tested for rolling mediator activity is incorporated onto the solid phase surface(s) of the rolling model apparatus, and flow medium containing leukocytes is passed 25 through. Rolling of the leukocytes along the solid phase indicates that the molecule has rolling mediator activity and that the leukocytes express a binding partner for the rolling mediator.

To identify a rolling mediator by use of an 30 arrest model apparatus, the molecule to be tested for rolling mediator activity is incorporated onto the solid phase surface(s) of the arrest model apparatus. Also incorporated onto the solid phase surface(s) is an integrin binding partner. A chemoattractant is 35 introduced into the flow medium passing through the # 26 4 35 3 apparatus. The flow medium contains leukocytes irhich express the integrin recognizing the integrin binding partner and express a receptor for the chemoattractant. After passage of the flow medium 5 tiirough the apparatus, it is determined whether any leukocytes have arrested on the solid phase. Arrest of leukocytes indicates that the test molecule has rolling mediator activity and that the leukocytes which express the integrin and the chemoattractant 10 receptor also express a binding partner for the test molecule.

A molecule can also be identified as a functional component in the processes of leukocyte rolling, or rolling and arrest, or as an enhancer 15 thereof, by the methods described supra in which an increase in number or percentage of cells rolling or arrested, is detected relative to the number or percentage of such cells in the absence of the test molecule. .11. THERAPEUTIC AND DIAGNOSTIC UTILITIES OF THE INHIBITORS AND PROMOTERS QF THE INFLAMMATORY RSSPQHSE The inhibitors and promoters of the invention have use therapeutically in diseases or disorders involving inflammation, and which involve extravasation of leukocytes. The invention provides methods of reducing inflammation, and of treating or preventing disorders associated therewith, by administration to a subject of an effective amount of the inhibitory compounds of the invention. In an alternative embodiment, the invention provides methods of stimulating the inflammatory response, and treating or preventing disorders associated with a deficit in the desired inflammatory response, by administration to a subject of an effective amount of the pro- # 26 4 353 inflammatory compounds (promoters) of the invention. The subject is preferably an animal, including but not limited to animals such as cows, pigs, chickens, etc., and is preferably a mammal, and most preferably human.

Disease and disorders which can be treated by administration of a therapeutically effective amount of the inhibitory compounds of the invention include but are not limited to the following: Inflammatory arthritis - e.g., rheumatoid 10 arthritis, seronegative spondeloarthritites (Behcets disease, Reiter's syndrome, etc.), juvenile rheumatoid arthritis, vasculitis, psoriatic arthritis, polydermatomyositis.

Systemic lupus erythematosus fSLEl. 15 Asthma.

Inflammatory dermatoses - e.g., psoriasis, dermatitis herpetiformis, eczema, necrotizing and cutaneous vasculitis, bullous diseases.

' Reperfuslon injury. W, Septic shock fSepsisl.

Adult respiratory distress syndrome YARDS1.

Tissue damage relating to tissue Other autoimmune disorders. In addition to 25 the autoimmune disorders SLE and rheumatoid arthritis, disorders such as glomerulonephritis, juvenile onset diabetes, multiple sclerosis, allergic conditions, autoimmune thyroiditis, allograft rejection (e.g., rejection of transplanted kidney, heart, or liver), 30 Crohn's disease, and graft-versus-host disease can be treated.

Thermal injury fburnl. The main complications due to burn are inflammatory in nature, including shock, and pulmonary edema. # 26 4 35 3 Cardiopulmonary bypass. Systemic inflammation has been associated with the use of pump-oxygenator systems in cardiopulmonary bypass and hemodialysis, which can%lead to organ dysfunction, 5 termed the post-pump syndrome or post-perfusion syndrome.

In addition, other diseases and clinical correlates of undesirable inflammatory responses can be treated with the inhibitors of the invention, 10 including but not limited to those associated with hemolytic anemia, hemodialysis, blood transfusion, certain hematologic malignancies, pneumonia, postishemic myocardial inflammation and necroslii, barotrauma (decompression sickness), ulcerative IS colitis, inflammatory bowel disease, atherosclerosis, cytokine-induced toxicity, necrotizig enterocolitis, granulocyte-transfusion-associated syndromes, Raynaud's syndrome, multiple organ injury syndromes secondary to septicemia or trauma, and acute purulent 20 meningitis or other central nervous system inflammatory disorders. In addition, the Inhibitors of the inflammatory response which bind to ICAM-1 can be used to treat or prevent viral infections such as rhinoviral infection, since the rhinovirus binds to --sv ICAM-1 on human cells and thereby initiates infection of the cells.

Diseases or disorders that can be treated by the pro-inflammatory compounds of the invention include but are not limited to immunosuppression .30 (e.g., due to AIDS, cancer chemotherapy, radiation therapy, corticosteroid therapy, or other therapy for autoimmune disease), and congenital immunodeficiencies. .11.1. DEMONSTRATION OF THERAPEUTIC UTILITY & \ 26 4 35 3 Compounds demonstrated to have the desired activity in the apparatuses of the invention can then be tested In vivo for the desired anti- or proinflammatory activity, as the case may be. For S example, such compounds can be tested in suitable animal model systems prior to testing in humans, including but not limited to rats, mice, chicken, cows, monkeys, rabbits, etc. suitable model systems are also used to demonstrate therapeutic utility (see 10 infra).

For In vivo testing, prior to administration to humans, any animal model system known in the art may be used. For example, several animal models are available to demonstrate the efficacy of anti-15 inflammatory compounds of the invention in the treatment of adult respiratory distress syndrome (ARDS). These include New Zealand white rabbits infused with activated complement (Nuytinck et al., 1986, Brit. J. Exp. Pathol. 67:537-548); cenxlean-20 induced acute pancreatitis in rats (Guice et al., 1988, Ann. Surg. 208:71-77); a porcine model produced by infusion of live Pseudomonas aeruginosa (Dehring et al., 1987, J. Trauma 27:615-625); cynomolgus monkeys (Macaca fascicularis) made acutely septic with 25 infusions of £. coli, resulting in severe sepsis and ARDS (Stevens et al., 1986, J. Clin. Invest. 77:1812-1816).

Two animal models of sepsis which can be used are a rat cecal ligation and puncture model (von 30 Allmeri et al., 1990, J. Surg. Res. 48:476-480) and a sheep common bile duct contamination model (Barke et al., 1990, Arch. Surg. 125:437-440).

A rabbit model of barotrauma is known (Ward et al., 1990, Undersea Biomed. Res. 17:51-66). # - 67 26 4 3 5 3 For animal models of thermal injury, see Bjornson et al., 1986, J. Infect. Ois. 153:1098-1107; Oldham et al., 1988, Surgery 104:272-279; Friedl et al., 1989, Am. J. Pathol. 135:203-217; Demling et al., 5 1989, surgery 106:52-59.

An animal model system for rheumatoid arthritis is that consisting of animals of the autoimmune MRL/1 mouse strain (Murphy, E.D. and Roths, J.B., 1978, in Genetic Control of Autoimmune Disease. 10 Rose, N.R., sfc al., eds., Elsevier/North-Holland, New York, pp. 207-219), that develop a spontaneous rheumatoid arthritis-like disease (Hang et al., 1982, J. Exp. Med. 155:1690-1701). 5.11.2. THERAPEUTIC ADMINISTRATION AND COMPOSITIONS Various delivery systems are known and can be used to administer the compounds of the invention, e.g., encapsulation in liposomes, microparticles, microcapsules, expression by recombinant cells, etc. 20 Other methods of introduction include but are not limited to intradermal, intramuscular, intraperitoneal, intravenous, subcutaneous, intranasal, and oral routes. The compounds may be administered by any convenient route, for example by 25 infusion or bolus injection, by absorption through epithelial or mucocutaneous linings (e.g., oral mucosa, rectal and intestinal mucosa, etc.) and may be administered together with other biologically active agents. Administration can be systemic or local, 30 e.g., direct injection at the inflamed joint of someone suffering from rheumatoid arthritis.

The present invention also provides pharmaceutical compositions. Such compositions comprise a therapeutically effective amount of a 35 compound of the invention, and a pharmaceutically - 68 ^643 acceptable carrier or excipient. Such a carrier includes but is not limited to saline, buffered saline, dextrose, water, glycerol, ethanol, and combinations thereof. The formulation should suit the 5 mode of administration.

The composition, if desired, can also contain minor amounts of wetting or emulsifying agents, or pH buffering agents. The composition can be a liquid solution, suspension, emulsion, tablet, 10 pill, capsule, sustained release formulation, or powder. The composition can be formulated as a suppository, with traditional binders and carriers such as triglycerides. Oral formulation can include standard carriers such as pharmaceutical grades of 15 mannitol, lactose, starch, magnesium stearate, sodium saccharine, cellulose, magnesium carbonate, etc.

In a preferred embodiment, the composition is formulated in accordance with routine procedures as a pharmaceutical composition adapted for intravenous 20 administration to human beings. Typically, compositions for intravenous administration are solutions in sterile isotonic aqueous buffer. Where necessary, the composition may also include a solubilizing agent and a local anesthetic such as 25 lignocaine to ease pain at the site of the injection. Generally, the ingredients are supplied either separately or mixed together in unit dosage form, for example, as a dry lyophilized powder or water free concentrate in a hermetically sealed container such as 30 an ampoule or sachette indicating the quantity of active agent. Where the composition is to be administered by infusion, it can be dispensed with an infusion bottle containing sterile pharmaceutical grade water or saline. Where the composition is 35 administered by injection, an ampoule of sterile water # 26 4 35 3 for injection or saline can be provided so that the ingredients may be mixed prior to administration.

The compounds of the invention can be formulated as neutral or salt forms. Pharmaceutically 5 acceptable salts include those formed with free amino groups such as those derived from hydrochloric, phosphoric, acetic, oxalic, tartaric acids, etc., and those formed with free carboxyl groups such as those derived from sodium, potassium, ammonium, calcium, 10 ferric hydroxides, isopropylamine, triethylamine, 2-ethylamino ethanol, histidine, procaine, etc.

The amount of the compound of the invention which will be effective in the treatment of a particular disorder or condition will depend on the 15 nature"of the disorder or condition, and can be determined by standard clinical techniques. The precise dose to be employed in the formulation will also depend on the route of administration, and the seriousness of the disease or disorder, and should be 20 decided according to the judgment of the practitioner and each patient's circumstances. However, suitable dosage ranges for intravenous administration are generally about 20-500 micrograms of active compound per kilogram body weight. Suitable dosage ranges for 25 intranasal administration are generally about 0.01 pg/kg body weight to 1 mg/kg body weight. Effective doses inay be extrapolated from dose-response curves derived from in vitro or animal model test systems.

Suppositories generally contain active 30 ingredient in the range of 0.5% to 10% by weight; oral formulations preferably contain 10% to 95% active ingredient.

The invention also provides a pharmaceutical pack or kit comprising one or more containers filled 35 with one or more of the ingredients of the 70 - 264353 pharmaceutical compositions of the invention. Associated with such container(s) can be a notice in the form prescribed by a governmental agency regulating the manufacture, use or sale of pharmaceuticals or biological products, which notice reflects approval by the agency of manufacture, use or sale for human administration. 6. LEUKOCYTES ROLL ON A SELECTIN AT PHYSIOLOGIC FLOW SATES: DISTINCTION FROM AND PREREQUISITE FOR ADHESION JGH INTEGRINS We show herein that rolling of leukocytes on vascular endothelial cells, an early event in inflammation, can be reproduced in vitro on artificial 15 lipid bilayers containing purified CD62, a selectin also named PADGEM and GMP-140 that is inducible on endothelial cells. Neutrophils rolled on this selectin under flow conditions similar to those found in postcapillary venules. Adhesion of resting or 20 activated neutrophils through the integrins LFA-l and m Mac-l to ICAM-1 in a lipid bilayer did not occur at physiologic shear stresses; however, static incubation of activated neutrophils allowed development of adhesion that is greater than 100-fold more shear 25 resistant than found on CD62. Addition of a chemoattractant to activate LFA-l and Mac-l resulted in the arrest of neutrophils rolling on bilayers containing both CD62 and ICAM-1. Thus, at physiologic shear stress, rolling on a selectin is a prerequisite 30 for activation-induced adhesion strengthening through integrins. 6.1. RESULTS 6.1.1. RECONSTITUTION OF CD62 AND 35 ICAM-1 IN &IPIP PHAVEE? - 71 264353 Immunoaffinity-purified CD62 and ICAH-1 in octyl-/?-D-glucopyranoside (OG), alone or together, were mixed with phosphatidylcholine, and liposomes were prepared by dialysis. Liposome suspensions were 5 placed on clean glass slides to form planar lipid bilayers with incorporated proteins (Watts et al., 1986, Nature 320:179-181). The incorporation of both CD62 and ICAM-1 into the membranes was quantitated by saturation binding with 12SI-mAb. Throughout the 10 experiments described below, neutrophil binding to planar bilayers containing CD62 and ICAM-1 was found to be highly specific. Neutrophils did not bind to planar membranes containing phosphatidylcholine alone. Neutrophil binding to planar membranes containing CD62 15 was 98% reversible by incubation with EOTA for 10 min. Binding of N-formyl methionyl leucyl phenylalanine (fMLP)-stimulated neutrophils to planar membranes containing ICAM-1 was inhibited 97% by a combination of ariti-LFA-1 (TSl/22) and anti-Mac-1 (LPM19c) a 20 subunit antibodies, in agreement with previous reports (Smith et al., 1989, J. Clin. Invest. 83:2008-2017; Diamond et al., 1990, J. Cell Biol. 111:3129-3139). TSl/22 or LPM19c alone did not inhibit fMLP-stimulated neutrophil binding to ICAM-1; thus, the results 25 reported below reflect interaction of both LFA-l and Mac-l on the neutrophil with ICAM-1 in the planar membrane. 6.1.2. THE SELECTIN CD62 IS DISTINCTIVE IN SUPPORT OF ADHESION AT VENULAR LEVELS OF SHEAR STRESS To measure attachment of neutrophils under flow conditions, glass slides containing planar bilayers were incorporated into a parallel plate flow chamber (Fig. 3) . Neutrophils (10*/ml) were infused at flow rates that were regulated to produce wall shear - 72 26 435 3 stresses bracketing the range estimated to exist in post-capillary venules. Wall shear rates and stresses in the flow chamber are calculated from chamber geometry and volumetric flow rate (Lawrence et al., 5 1990, Blood 75:227-237). Shear stresses in vivo can be calculated from centerline velocity and vessel diameter using the Hagen-Poiseuille equation. In both cylindrical and parallel plate geometries, the velocity profile of a Newtonian fluid is parabolic 10 (Fig. 4). The change in velocity per change in radial displacement is called shear and is highest at the wall. Shear stress, the product of shear and viscosity, better correlates with the forces acting on a cell under flow. Shear stresses of 1-10 dyn/cm2 have 15 been measured for postcapillary venules (Heisig, 1968, Adv. Microcirc. 1:89-94), and in the classic studies of Atherton and Born (1972, J. Physiol. 222:447-474), leukocytic rolling was observed at shear stresses that we calculate to be 1.5-4.0 dyn/cm2. 20 Flowing neutrophils readily bound to artificial bilayers containing CD62 at wall shear stresses within the physiologic range (Fig. 5A). Attachment was dependent on the density of CD62, with efficient attachment at 1.8 to 2.8 dyn/cm2 at 400 and 25 200 sites per im2, and significant attachment occurring at 1.8 dyn/cm2 at a density of 50 sites per fim2. By contrast, neutrophils did not adhere to artificial bilayers containing ICAM-1 under flow conditions, even at high ICAM-1 densities of 1,000 sites per nm2 (Fig. 30 5A)* Lower ICAM-1 densities than this support strong static binding of activated neutrophils (see below); however, neutrophils that were stimulated with PMA did not bind at physiologic shear stresses (>1 dyn/cm2), and bound only at low levels in stagnant flow at a 35 shear stress of 0.36 dyn/cm2 (Fig. 5A) . Flowing 26 4 3 5 3 lymphocytes that have been activated by cross-linking the T cell antigen receptor with mAb also 1 -xil to bind to ICAM—1 under flow conditions (not shown) but bind avidly through LFA-1 to ICAM-1 under static conditions 5 (Dustin and Springer, 1989, Nature 341:619-624). Lymphocytes also fail to bind through CD2 to LFA-3 (1,000 sites per jmz) under flow conditions (not shown), but do so at stasis (Chan et al., 1991, J.

Cell Biol. 115:245-255). The failure of several 10 integrin-Ig superfamily interactions and Ig-Ig superfamily (CD2-LFA-3) interactions to occur under flow conditions is distinct from the efficiency of interaction through the selectin CD62.

The efficiency of adhesion of unactivated 15 neutrophils to artificial bilayers containing a mixture of CD62 and ICAM-1 was indistinguishable from adhesion to CD62 alone (Fig. 5B). Thus, no cooperative interactions between these adhesion mechanisms occur on resting neutrophils. 20 Activation with PMA prior to infusion greatly decreased the efficiency of adhesion to CD62 (Fig. 5B). The amount of adhesion of PMA-stimulated neutrophils to the mixture of CD62 and ICAM-1 (Fig. 5B) was additive for the adhesion to CD62 alone (Fig. 25 5B) and ICAM-1 alone (Fig. 5B). 6.1.3. NEUTROPHILS ROLL ON CD62 Artificial bilayers containing CD62 were remarkable not only for their ability to support 30 adhesion under physiologic flow conditions, but also in the rolling mode of this adhesion. Neutrophils rolled on the CD62 substrate, driven by fluid drag forces, as readily recorded using a video camera (Fig. 6). Rolling neutrophils that accumulated on the CD62 35 substrate remained round and appeared as sharp images, 26 4 35 3 whereas nonadherent cells that tumbled past in the shear f low were blurred streaks (Fig. 6A). Only nonadherent, tumbling cells were visualized on ICAM—1 substrates (Fig. 6B); the length of the blur is 5 dependent on the distance from the wall because of the velocity profile (Fig. 4), with the slowest tumbling cells closest to the substrate.

Confirming that the cells roll rather than slide on the CD62 substrates, rotation of the 10 polymorphic nucleus of the neutrophil was observed using a 40 x objective and was clearly visible when videotapes were played back; the progress of rolling cells across a substrate at 2 s intervals is shown using a 40 x objective in Figures 6C-6F. At lower 15 shear stresses cells frequently wobbled as they rolled. Individual cells rolled with a relatively, but not completely, uniform velocity. Rolling was sometimes interrupted by brief pauses where the cells appeared to momentarily halt on the substrate, and 20 also by brief increases in velocity during which the cells appeared to detach from the substrate and then reattach several cell diameters downstream. Boiling cells were in equilibrium with nonadherent cells, as reflected by detachment of some rolling cells with 25 their loss from the field of view and attachment of other cells.

The overall similarity in rolling velocity among different cells led to the appearance that they rolled as a group over the substrate, as for the six 30 cells in Figures 6C-6F, but there was some variation in velocity; comparison of Figures 6C and 6F shows that the cell in the upper left in Figure 6C is the slowest of the group and the one 5 o'clock to it in Figure 6C is the fastest and has overtaken it by 35 Figure 6F. 264353 Disturbance of the laminar flow profile by attached cells was evident from observation of cells attaching to artificial bilayers of CD62 at a shear stress of 1.8 dyn/cm2 and above. Single cells bound in 5 a random distribution over the substrate. However, once one cell bound, binding of further cells immediately downstream was much more likely than in areas of the substrate with no attached cells. It appeared that the neutrophils were following locally 10 disturbed streamlines that increased the chance of collision with and subsequent attachment to the substrate. After several minutes, the cells became more evenly distributed. At shear stress levels of 0.73 dyn/cm2 and below, attachments were uniformly 15 distributed even at the earliest times.

Rolling velocity increased with increasing shear force (Pig. 7). The rolling velocity was proportional to the shear force at low shear stresses, and then began to plateau. This may reflect the 20 effect of torque acting on a rolling but deformable object, leading to a greater contact area and less fluid drag. Rolling velocity was dependent on the site density of CD62 (Fig. 7). The higher the CD62 density, the more slowly the neutrophils rolled at a 25 given shear stress. This is as predicted, because a higher number of receptor-ligand interactions will lead to a greater resistance to the fluid drag force. The slower rolling velocities at higher site density correlated with the increased effectiveness of 30 attachment. Inclusion of ICAM-1 in artificial bilayers containing CD62 did not alter rolling velocity (Fig. 7), correlating with its lack of effect on attachment.

Rolling did not appear to result in any 35 alteration in the cells or the substrate. The rolling 2 6 4 3 velocity of groups of cells followed along the substrate did not appear to change with time. The rolling velocity of cells on CD62 substrates was unaltered for at least 15 min at a single observation 5 condition, and the same substrate could be used for observations at many different shear stresses, without any alteration in rolling velocity measurements for the same shear stress replicated at the beginning and end of the set of observations. When the direction of 10 flow was reversed, the direction of rolling was reversed and the cells rolled back "over their tracks" with the same velocity as in the forward direction.

We compared the velocities of unstimulated neutrophils rolling over a CD62 substrate and tumbling 15 over an ICAM-1 substrate to the predicted velocity for a sphere of the same diameter, 7 tm (Table 4) . - 7* 264353 TABLE 4 Measured Neutrophil Rolling Velocities compared with Calculated Velocities for a Sphere in Shear Flow Meuured Velocities (pm/s) on Various Substrates' Calculated Velocities (pm/s) st Various Distances from Waif Shear Stveta (dyn/cm2) CD62 ICAM-1 ICAM-1 + PMA 500 nm nm 1 nm 0.18 Not dooe 52.4 ± 11 .2 0.0 54.9 28.5 21.4 0-36 1.9 + 0.7 (89*) 138.6+,45.8 0.0 110.6 57.5 43 0.73 3.4 ±.0.4(96%) 230.2 ±20.9 0.0 221 114 86 1.8 .7 .+ 0.5 (99*) 541 ±271 0.0 550 285 214 3.6 7.2 ±1.6 (100*) - 0.0 1104 573 429 7.3 - 11.4 + 3.0(99%) - 0.0 2206 1145 858 14.6 .9 ±1J (100%) - 0.0 4419 2296 1712 * Measured rolling velocities are from 2 or 4 s measurement periods on the indicated substrates (CD62 at 200 sites per /im2; ICAM-1 at 1000 sites per m® ); the _ percentage of cells that rolled in this time period on 2 CD62 is shown in parentheses. Velocities of resting neutrophils on ICAM—1 at shear stresses of 0.73 dyn/cm2 and above were measured from the leading edge of the streak on four consecutive frames. Asterisks indicate that cells moved too fast for velocity measurements. Measurements for PMA-stimulated neutrophils were made after cells were allowed to adhere to the substrate 25 under static conditions for 6 min. b Hydrodynamic calculations are for a 7 |im diameter hard sphere in a shear flow with no interaction with the wall, at the indicated distance from the wall. • 7 d " 26 4 35 3 The measurements on the ICAM-1 substrate are for the cells flowing closest to the substrate, as determined by the focal plane. The unstimulated neutrophils flowing over the ICAM-1 substrate moved at a velocity 100-fold greater than the 5 cells rolling on the CD62 substrate. Because of the torque exerted by shear flow, a spherical object will rotate at an angular velocity of one-half the shear rate, so its motion, particularly if close to the wall, is not qualitatively different from rolling along a flat surface. Predicted 10 velocities of a sphere 7 fm in diameter having no interaction with a smooth wall (Goldman et al., 1967, Chem. Engineer Sci. 22:653-660) at distances of l, 10 and 500 nm are more than an order of magnitude higher than the measured velocities for rolling on CD62 (Table 4). The rolling 15 velocities observed on CD62 thus require an adhesive interaction. By contrast, the measured velocities on the ICAM-1 substrates are in excellent agreement with the predictions for a sphere tumbling in shear flow, particularly with a separation of 500 nm. Distances on the 20 order of 1 and 10 nm are probably difficult to obtain without contact, since surface features such as ICAM-1 are on the order of this size; ICAM-1 is 18 nm long as measured in the electron microscope (Staunton et al., 1990, Cell 61:243-254). 6.1.4. APHEglW WPER STATIC CQNPITIQNS The ability of neutrophils to spread and develop adhesiveness on artificial bilayers under static conditions was examined for comparison to behavior under flow 30 conditions. Neutrophils were injected through a port and examined at different time points; 2 min of the total time was required before all the neutrophils had settled onto the bilayer. Resting neutrophils remained round on ICAM-1 bilayers, with no spreading after prolonged incubation (Fig. 35 8A). However, treatment with phorbol 12-myristate-13- 264353 acetate (PMA) or fMLP resulted in spreading of almost all neutrophils on ICAM-1 bilayers (Figs. 8B-8F).

Neutrophils on CD62 bilayers remained round even after prolonged contact of up to 12 min (Fig. 8G), in 5 agreement with observations on rolling cells tinder flow conditions. Neutrophil activation with PMA induces a bipolar shape change even when cells are held in suspension; PMA-stimulated neutrophils had a bipolar appearance on CD62 bilayers but no spreading was observed (Fig. 8H). 10 To examine adherence under static conditions, neutrophils were allowed to settle onto the planar membrane, and after 6 min of contact, controlled flow was used to create a detachment force. Binding to CD62 and ICAM-1 was compared, with and without PMA present. This assay 15 highlighted significant differences in patterns of adhesion strengthening between the CD18 and the CD62 pathways.

Unstimulated neutrophils did not adhere to artificial bilayers containing ICAM—1 (Fig. 9A). By contrast, PMA-stimulated neutrophils developed strong 20 adhesion to the bilayer during the 6 min contact period. These attachments were dramatically resistant to shear stress; almost all the cells remained attached at wall shear stresses as high as 36 dyn/cm2. The same results were observed after stimulation with 10*7 M fMLP. These strongly 25 adherent cells could not be induced to roll even at high shear stresses (Table 4). In the same assay, we found that binding was 97% inhibited by a combination of mAbs to the Mac-l and LFA-1 a subunits, but not by either mAb alone.

Thus, adhesion strengthening occurs through both Mac-l and 3-0. LFA-1. The differing abilities of ICAM-1 to mediate adhesion of PMA-stimulated neutrophils under flow as compared to static conditions were seen despite use of ICAM-1 at 1,000 sites per fxm2 in flow assays and at 250 sites per jim2 in static assays. 9.e 4 3 5 3 The formation of mechanically strong attachments through integrins following PMA or fMLP stimulation contrasted with the pattern observed for CD62* With the same contact time, neutrophils formed very reversible 5 adhesions to CD62 (Fig. 9A). The shear stresses required for detachment of neutrophils from CC62 were roughly comparable to the maximal shear stresses at which attachments could form (Fig. 6, above). More importantly, all the neutrophils rolled along the planar membrane 10 containing CD62, and at the same velocity as observed for neutrophils that adhered under flow conditions (data not shown). Thus, contact time did not result in adherence strengthening through the selectin CD62.

We used bilayers containing both CD62 and ICAM-1 15 to test for cooperativity in the sta-tic binding assay between the selectin and integrin/Ig gene family adhesion systems. No enhancement of resting neutrophil binding was observed on a bilayer containing both CD62 and ICAM-1 compared with CD62 alone (Fig. i*Z). The shear sensitivity 20 of binding to both types of planar membranes was equivalent, and rolling occurred as on the CD62 membrane alone. Thus, even under conditions where contact with CD 6 2 and ICAM-1 was maximized, CD62 did not stimulate adhesiveness of neutrophils for ICAM-1. The integrin/ig and selectin 25 pathways appear to function independently of each other; binding to the ligand for CD62 on neutrophils did not generate a signal that activated integrin avidity.

PMA activation of neutrophils resulted in strong attachments after 6 rain of contact with the planar membranes 30. containing both ICAM-1 and CD62 (Fig. 9B). Adherent cells had a morphology indistinguishable from that of PMA-activated neutrophils binding to ICAM-1 alone, and were equally shear resistant. On CD62 alone, PMA did not stimulate development of shear-resistant attachments, but 35 decreased shear resistance relative to unstimulated cells. 26 4 35 3 6.1.5. COOPERATION BETWEEN SELECTIN AND INTEGRIN ADHESION MECHANISMS ^ We tested for cooperation between the CD62 rolling and leukocyte integrin adhesion-strengthening mechanisms s under conditions that would approximate those at an inflammatory site by addition of fMLP to rolling neutrophils. Neutrophils were infused at two shear stresses in the physiologic range into a flow chamber with an artificial bilayer containing both CD62 and ICAM-1. After a 3 min equilibration period to allow accumulation of rolling neutrophils on the bilayer, 10*9 M fMLP v s added to the infusion medium and reached the rolling neutrophils 12 or 30 s later (0 time. Fig. 10A). Within 30 seconds of exposure, mean rolling velocities had begun to drop at both flow rates 15 (Fig. 10A) and continued to drop for 5 min. After 5 min, essentially all the neutrophils became arrested on the bilayer, as seen both from the average rolling velocity (Fig. 10A) and the velocity distributions (Figs. 10B and 10C). Arrest of round cells preceded spreading. The round, 2q rolling neutrophils at o time had almost completely spread after 5 min (Fig. 10D) and developed shear-resistant adhesion (Fig. 10E). Cells did not become arrested on CD62 alone, because even after contact of activated neutrophils with CD62 bilayers under static conditions, rolling occurred 25 after initiation of flow, and adhesion was not strengthened compared with unstimulated cells (Fig. 9B) . 6.2. DISCUSSION We demonstrate herein that leukocyte rolling, a 30 hallmark of the early stages of an inflammatory response, can be reconstituted in vitro on artificial lipid membranes containing an endothelial cell selectin, CD62. Neutrophils formed reversible rolling attachments to CD62 at physiologically relevant shear stresses. Another 35 endothelial cell adhesion molecule, ICAM-1, did not support rolling adhesions, and was found to be significantly less 82 - 2 6 4 3 5 3 effective than CD62 as a ligand for neutrophils under flow conditions. However, ICAM-1 was required for subsequent spreading and the development of a shear resistant attachment.

The selectin CD62 was distinctive from integrin and Ig family members both in ability to mediate rolling and to mediate adhesion during flow at physiologic shear stresses. When the avidity of neutrophil integrins for ICAM-1 was stimulated with PMA or fMLP, flowing neutrophils 10 did not bind to ICAM-1 significantly over a range of physiologic shear stresses at which CD62 was highly effective. Thia was despite the ability of activated neutrophils to form attachments in static assays to 4-fold lower densities of ICAM-1. These attachments to ICAM-1 were 15 greater'than 100-fold more shear resistant than attachments to CD62. Integrins on activated T cells also failed to bind to ICAM-1 and fibronectin at physiologic shear stresses. Furthermore, an Ig familylg family interaction between CD2 on T lymphocytes and LFA-3 in artificial bilayers failed to 20 occur at physiologic shear stresses. Even when adhesion of neutrophils or lymphocytes to ICAM-1 or lymphocytes to LFA-3 was initiated under static conditions, firm adhesion was unable to be converted to a rolling adhesion when shear flow was applied; cells either remained adherent or were 25 completely dislodged. The shear stress required to dislodge 50% of cells binding through CD2 to the transmembrane isoform of LFA-3 is 1.5 dyn/cm2 (Chan et al., 1991, J. Cell Biol. 115:245); at a comparable ligand density a shear stress of 1 dyn/cm2 dislodged 50% of neutrophils binding to 30 CD62. The comparability of these shear stresses suggests that rolling is not directly related to binding strength.

We saw significant binding of neutrophils to CD62 at site densities as low as 50 per /m2, which is a physiologically relevant density. CD62 is stored in 35 Weibel-Palade bodies of endothelial cells (Bonfanti et al., - 83 26 4 3 5 3 1989, Blood 73:1109-1112; McEver et al., 1989, J. Clin. Invest. 84:92—99), which in response to stimuli such as thrombin and histamine fuse with the plasma membrane and thereby upregulate expression of CD62. An increase from 20 5 to 50 sites per /ra2 within 5 min after stimulation and a decline to baseline levels by 30 min is seen for endothelium cultured in vitro (Hattori et al., 1989, J. Biol. Chem. 264:7768-7771). During passage of endothelial cells in vitro, there is a decline in CD62 and Weibel-Palade body 10 content and therefore densities of CD62 could be higher in vivo.

Our studies demonstrating that C062 is a receptor that mediates rolling suggest that its upregulation may be an important mechanism for regulating rolling in the 15 inflammatory response. Thrombin stimulation mediates a transient binding of neutrophils that is CD18 independent and likely due to CD62 expression (Zimmerman and Mclntyre, 1988, J. Clin. Invest. 81:531-537; Geng et al., 1990, Nature 343:757-760). Neutrophil binding to primary culture 20 endothelial cells at 2.0 dyn/cm2 wall shear stress is stimulated by thrombin and is characterized by the initiation of rolling detectable within less than a 1 min of thrombin exposure, suggesting the involvement of CD62 (M.B. Lawrence and L.V. Mclntire, unpublished data). Thus CD62 is 25 an excellent candidate for the initiation of neutrophil rolling during the early stages of an inflammatory response.

Rolling velocities on CD62 were comparable to in vivo rolling velocities. On bilayers containing CD62, velocities ranged from under 2 to over 30 fm/a, depending on 3.0 site density and the shear stress. Mean rolling velocities in vivo were 10 fin/s in mouse mesentery venules and 20 im/s in hamster cheek pouch venules (Atherton and Born, 1973, J. Physiol. 233:157-165). Rolling velocities on CD62 were proportional to the flow rate at low shear stresses, but 35 increases were not proportional at higher shear stresses. 84 - 264353 Atherton and Bom (1973, J. Physiol. 233:157-165) reported that rolling velocities in vivo were linearly dependent on shear stress (blood flow velocity) up to about 5 dyn/cm2 and plateaued above this point. The in vivo velocities at a 5 given shear stress (Atherton and Born, 1973, J. Physiol. 233:157-165) were about twice what we find on C062 at 50 sites per /on2; considering that other selectins nay contribute to rolling in vivo and their concentration is unknown, the correspondence between these values is 10 remarkable.

Neutrophils rolled on CD62 at shear stresses comparable to those found in post-capillary venules in vivo. Hall shear stresses at which rolling was observed in the measurements of Atherton and Born (1972, J. Physiol. 15 222:447-474; 1973, J. Physiol. 233:157-165) (assuming cylindrical geometry and a plasma viscosity of 2 centipoise) ranged from 1.5 to 4.0 dyn/cm2. These values are similar to the shear stresses at which neutrophils form rolling attachments to bilayers containing CD62, and also to 20 monolayers of endothelial cells stimulated with cytokines (Lawrence et al., 1990, Blood 75(l):227-237) or thrombin (M.B. Lawrence and L.V. Mclntire, unpublished data).

Other selectins have not been shown to mediate rolling; however, this family of molecules may be 25 specialized to mediate rolling adhesions, and rolling observed in vivo may involve contributions from all three. LECAM-1 contributes to interaction of unactivated neutrophils with cytokine stimulated endothelial cells at physiologic shear stresses (Smith et al., 1991, J. Clin. 30 invest. 87:609-618). On the basis of the ability of unactivated neutrophils to adhere under flow conditions, the Selectins so far studied appear to be structures that are capable of mediating the initial attachment of neutrophils to the vessel wall. 264353 The structure of selectins and their carbohydrate ligands appears ideally suited for their function in rolling and adhesion at high shear stresses. Kinetic considerations are very important here; a high on-rate for formation of the 5 selectin-ligand complex is required for efficient interactions of rapidly flowing cells with a substrate, and both a rapid on-rate and rapid off-rate are required for rolling. We found no change in the cells or the substrate after prolonged rolling, suggesting that dissociation at the 10 upstream edge of the cell is due to dissociation of the CD62-ligand complex rather than to loss of molecules from the cell or substrate.

Interactions between biological macromolecules are frequently diffusion limited; i.e., the rate at which they 15 can form specific noncovalent interactions is limited by their rates of diffusion. Diffusion of the extracellular ligand binding domain of an adhesion receptor has two components, one due to segmental flexibility of the tether by which it is attached to the membrane and the other due to 20 lateral diffusion in the membrane bilayer of the membrane anchor. With an appropriately high number of segments and high segmental flexibilitye the former type of diffusion can occur more rapidly than the latter, since proteins of adhesion receptor size have diffusion coefficients on the 25 order of 10'7 cm3/s (Tanford, 1961, Physical Chemistry of Macromolecules, New York: John Wiley and Sons, Inc.) whereas bilayer diffusion coefficients are on the order of 10"' cm2/s (Jacobsonetal.,1987, Ann. Rev. Physiol. 49:163-175). The volume of solvent above the cell siurface in which diffusion 30 can occur and the diffusion rate are related to the length of the membrane tether, its segmental flexibility, and the size of the segments. Molecules composed of short consensus repeats of the type found in selectins have random configurations as revealed by electron microscopy (Weisman 35 etal.,1990, Science 249:146-151), suggesting a high degree 26 4 35 of segmental flexibility. The (sialylated) Lewis x ligand of CD62 is found at the termini of long carbohydrate structures that are predicted to confer flexibility (Fukuda et al., 1984, J. Biol Chem. 259:10925-10935). The location 5 of the lectin binding domain and the Lewis x determinant at the termini of their respective structures maximizes diffusiveness due to segmental flexibility. The diffusion coefficient is inversely related to size; therefore, sialylated Lewis x, by virtue of its small size (molecular 10 weight 779) relative to protein adhesion ligands, allows a faster rate for diffusion-limited processes.

Adhesiveness through integrins and the selectin CD62 differ drastically in the time scale required for their development. Binding of flowing cells to a substrate places 15 stringent time constraints on adhesive bond formation that appear to prevent leukocyte integrin interactions with ICAM-1, even When neutrophils are activated. CD62 mediates adhesion on a time scale at which the leukocyte integrin interactions are ineffective. Bond formation through CD62 20 appears so rapid that little adhesion strengthening is apparent when binding under flow and static conditions are compared. Adhesion strengthening through integrins may take time to develop both because the globular putative ligand binding regions are quite large and are attached by two 25 stalks to the membrane (Nermut et al., 1988, EMBO J. 7:4092-4099), limiting diffusiveness, and because cooperative interactions between multiple integrins or with other molecules such as cytoskeletal components may be required. The latter may be reflected in the spreading that occurs . 30 during adhesion strengthening on ICAM-l but not on CD62.

Inhibition of neutrophil binding to CD62 under flow conditions by prior activation with PMA may reflect an effect of the change to bipolar change rather than of loss of ligand-binding sites. CD62 binds at saturation to 20,000 35 sites per cell with a similar affinity or avidity of 10* M"1 26 4 3 to both resting and PMA-stimulated neutrophils (Moore et al., 1991, J. Cell Biol. 112:491-499). When we bound PMA-stimulated and resting neutrophils to artificial bilayers containing CD62 at stasis and subjected them to 5 detachment with shear, PMA-stimulated cells bound less efficiently, but the decrease in efficiency compared to untreated cells was much less marked than for binding under conditions of shear. Activated neutrophils with their bipolar, elongated shapes would experience higher transient 10 torques than unstimulated, round cells in contact with a substrate. The shape change may therefore be an important factor that impedes adhesion in shear flow, and may help prevent activated leukocytes, if they fail to emigrate at an inflammatory site, from attaching and emigrating at an 15 uninvolved site downstream.

Our studies demonstrate how a "weak", rolling interaction can develop into what is termed a "firm" adhesion to the vessel wall that neutrophils develop during an inflammatory response (Pober and Cotran, 1990, 20 Transplantation 50:537-544). Since neutrophils do not migrate on the lumenal side of the vessel wall, rolling allows an unstimulated neutrophil to move to a site where it can undergo diapedesis. Transendothelial cell migration, however, first requires the arrest of the rolling 25 neutrophil. While it has been shown that transendothelial migration requires CD18 (Smith et al., 1988, J. Clin.

Invest. 82:1746-1756), how a neutrophil comes to a stop at an endothelial cell junction has been unknown. We demonstrated here that following activation of leukocyte 30 integrins with a chemoattractant, neutrophils rolling on a bilayer containing CD62 and ICAM-1 rapidly slowed and came to a stop. Activation of LFA-1 and Mac-l permitted interactions with ICAM-1 which then led to arrest and spreading. The firm adhesion observed in the 35 microcirculation is blocked by in vivo administration of mAb 26 4 35 3 to CD18 (Arfors et al., 1987, Blood 69:338-340), and therefore appears equivalent to the shear-resistant attachment neutrophils form on ICAM-1 bilayers. At physiologic shear stress, rolling on a selectin was a 5 prerequisite for neutrophil integrin interaction with ICAM-1. The rolling interaction may promote the integrin-ICAM-1 interaction both because it facilitates close physical interaction between integrins and ICAM-1, and because the neutrophil is rolling approximately two orders 10 of magnitude more slowly than a tumbling neutrophil near the wall, which would raise the chance of enough bonds being formed for the rolling adhesion to be converted to a stationary one. Our findings suggest that activated integrins contribute to leukocyte arrest at sites of 15 endothelial cell junctions and contribute to the mechanism for transendothelial migration. The neutrophil interaction with CD62 is highly reversible, and should not impede subsequent migration mediated by leukocyte integrins, since interactions through the leukocyte integrins are much 20 stronger.

The synergism between the selectin and leukocyte integrin/Ig pathways demonstrated in our flow system is qualitatively different from the additivity observed in static adhesion assays. Use of anti-ELAM-1 and anti-CD18 25 antibodies singly or in combination and anti-LECAM-1 and anti-CD18 antibodies singly or in combination had suggested that the selectin and integrin mechanisms, while distinct, functioned in an additive manner in neutrophil adhesion to cytokine-stimulated endothelial cells (Dobrina et al., 1989, 30 immunology 67:502-508; Luscinskas et al., 1989, J. Immunol. 142:2257-2263; Smith et al., 1991, J. Clin. Invest. 87:609-618; Hallmann et al., 1991, Biochem. Biophys. Res. Commun. 174:236-243). Similar studies on the relative contribution of CD62 and CD18 to binding of neutrophils to thrombin-35 stimulated endothelial cells have not been reported, but it 2 6 4 3 5 3 can be inferred from the existence of both a CD18-independent (presumably CD62) and CD18-dependent pathway in static assays (Zimmerman and Mclntyre, 1988, J. Clin. Invest. 81:531-537) that there would be additive 5 effects of blocking both adhesion mechanisms. Under flow conditions, however, we observed no additivity between CD62-and CD18-mediated mechanisms: no interaction was possible with ICAM-1 unless there was a rolling interaction with CD62. Our model of sequential interactions of rolling on 10 CD62 followed by adhesion strengthening on ICAM-1 predicts that blocking interaction through either integrins or through selectins alone should completely block formation of firm adhesion in vivo, and is in agreement with observations that mAb to either integrins or a selectin (LECAM—l/LAM-l) 15 can largely inhibit accumulation of leukocytes at inflammatory sites in vivo (Lewinsohn et al., 1987, J. Immunol. 138:4313-4321; Jutila et al., 1989, J. Immunol. 143:3318-3324; Arfors et al., 1987, Blood 69:338-340; Price et al., 1987, J. Immunol. 139:4174-4177). 20 The sequential steps of rolling on a selectin and adhesion strengthening through integrins elucidated here have important implications for the steps of leukocyte localization at inflammatory sites in vivo. Since rolling precedes firm sticking in vivo, and we have found that 25 rolling is mediated by a selectin and is a prerequisite for subsequent adhesion strengthening through an integrin in vitro, we suggest that in vivo the initial accumulation of rolling leukocytes at an inflamed site must be regulated by selectins and by changes in vessel tone that lower fluid 30 shear stresses to facilitate interactions through selectins. Small changes in vessel diameter and flow rate can result in significant reductions in wall shear stress, stimulation of endothelial cells can induce expression of CD62 and later ELAM-1, and might also induce on these cells the expression 35 of the ligand for LAM-1/LECAM-1. We found that rolling on 9 6 4 3 5 3 CD62 by itself does not activate integrins on the neutrophil, but does permit adhesion strengthening through integrins if the neutrophil is activated. Chemoattractants that activate the avidity of integrins on neutrophils and 5 monocytes are released at inflammatory sites. These include the complement component C5a, N-formylated bacterial peptides, platelet-derived growth factor, IL-8, and leukotriene B4. However, once these small molecules enter the bloodstream they are rapidly diluted and swept 10 downstream, leading to the commonly held opinion that they cannot be responsible for regulating adhesion of leukocytes in postcapillary venules, and may only be effective once a neutrophil has undergone diapedesis. However, rolling in vivo will bring leukocytes into close proximity with 15 chemoattractants that are diffusing from tissue through the junctions between endothelial cells, and to activating stimuli on the endothelial surface such as platelet-activating factor (Zimmerman et al., 1990, J. Cell Biol. 110:529-540), and rolling will greatly prolong the time 20 period over which leukocytes are exposed to these stimuli. This promotes the activation of integrin avidity on the leukocyte that we have shown slows rolling; activation of LAM-l/LECAM-1 on the leukocyte (Spertini et al., 1991, Nature 349:691-694) might have a similar effect. The slowed 25 rolling allows even more efficient exposure to chemostimulants and thereby sets in motion a positive feedback loop that results in arrest of the rolling leukocyte and finally spreading, integrin-mediated adhesion strengthening, and transendothelial migration. 30 In conclusion, we have shown that neutrophil rolling and subsequent arrest and adhesion strengthening at an inflammatory site in vivo can be accurately reproduced in vitro with a small number of purified components: a lipid bilayer, the selectin CD62, the Ig family member ICAM-1, and 35 the chemostimulant fMLP. We predict that other types of 91 ~ 26 4 3 5 selectins, integrin ligands, or chemostimulants will be found able to substitute for CD62, ICAM-1, or fMLP, respectively; nonetheless, the major point is that we have established the minimal molecular requirements for a complex 5 physiological process. The small number of required components on the vessel wall demonstrates the simple elegance of this process and suggests how it may be refined by substitution or addition of other selectins or integrin ligands. Our studies Illuminate the mechanisms by which 10 adhesion molecule antibodies or analogues profoundly inhibit inflammatory responses in vivo, and suggest that antagonists of selectin or integrins may have similar biological effects despite distinct mechanisms of action. - 6.3. EXPERIMENTAL PROCEDURES 6.3.1. MONOCLONAL ANTIBODIES Monoclonal antibodies used in these studies as purified IgG were AC1.2 (anti-CD62, IgGl) (Larsen et al., 1989, Cell 59:305-312), R6.5 (anti-ICAM-1, IgG2a) (Smith et 20 al., 1988, J. Clin. Invest. 82:1746-1756), and TSl/22 (anti-CDlla, IgGl) (Sanchez-Madrid et al., 1982, Proc. Natl. Acad. Sci. U.S.A. 79:7489-7493). They were used at 20 jig/ml for inhibition of neutrophil binding. LPM19c (anti-CDllb, IgG2a) (Uciechowski and Schiridt, 1989, in Leucocyte Typing 25 IV: White Cell Differentiation Antigens, W. Knapp, ed., Oxford: Oxford University Press, pp. 543-551) was used as a 1:200 dilution of ascites fluid. 6.3.2. PURIFICATION QF ICAM-l ANP CP 5 2 30 ICAM-1 was affinity-purified from the Epstein-Barr Virus-transformed B lymphoblastoid JY cell clone 33 (Hollander et al., 1988, J. Immunol. 141:4283-4290), as previously described (Marlin and Springer, 1987, Cell 51:813-819). Briefly, a Triton X-10Q lysate was passed over 35 a RRl/l Sepharose column, and the column was eluted with a 26 4 35 buffer containing 1% octyl-jS-D-glucopyranoside (OG) so that ICAM-1 could be incorporated into liposomes. CD62 was a generous gift of Drs. S. Sajer and B. Furie, and was purified as previously described (Larsen et al., 1989, Cell 5 59:305-312). 6.3.3. PREPARATION OF LIPOSOMES Liposomes were prepared by the method of OG dialysis (Mimms et al., 1981, Biochemistry 20:833-840) with 10 slight modification. Egg phosphatidylcholine (Avanti, AL) was diluted in chloroform and dried under an argon stream and then placed under a vacuum (30 /im Hg) for 2 hr to remove residual chloroform. The lipid film was redissolved at 0.4 MM in 250 Ml of 25 MM Tris-HCl (pH 8.0)/150 #iM NaCl (TS) , 2% 15 (w/v) OG, and was mixed with 250 til of detergent solution containing approximately 1-18 fig/ml CD62, 6-30 tig/ml ICAM-1, or both in TS, 1% 06, followed by three changes of dialysis against TS at 4aC over 36 hr. After the removal of OG by dialysis, the liposome suspension was stored at 4°C under 20 argon to minimize oxidation of lipids. 6.3.4. PREPARATION OF PLANAR BILAYERS Planar bilayers were formed by incubating drops of liposome suspension on glass coverslips or slides at 22°C 25 for 30 min. Prior to use, all glass surfaces were boiled in detergent (Linbro 7 x solution, Flow Lab, McLean, VA) for 15 min, rinsed extensively in deionized distilled water for at least 24 hr, and then stored in ethanol. For site number determinations, glass coverslips (5 mm diameter, no. 1 30 thickness; Bellco, Vineland, NJ) were attached to the bottom of 96-well microtiter plates (Linbro Titertek, Flow Lab, McLean, VA) by silicone household glue (General Electric Co., Water ford, NY), and liposome drops (20 fil) were placed on top. For laminar flow assays, glass slides (45 x 60 mm, 35 no. 2 thickness; Corning Glass Works, Corning, VA) were 264353 placed in 10 cm petri dishes. A liposome drop (20 nl) was placed in a demarcated area (1 cm diameter) and a glass coverslip (1 cm diameter; Bellco, Vineland, NJ) was then used to spread the liposome droplet. After the planar 5 bilayers were formed, excess liposomes were removed by several changes of binding mediun (DMEM, 10% FCS, 25 mM HEPES, [pH 7.4]). Planar bilayers were never exposed to air. 6.3.5. DETERMINATION OF SITE DENSITIES Liposomes were reconstituted with different quantities of either immunoaffinity-purified ICAM-l or CD62, and planar membranes were formed as described above. Monoclonal antibodies R6.5 to ICAM-1 (Smith et al., 1988, J. 15 Clin. Invest. 82:1746-1756) and AC1.2 to CD62 (Larsen et al., 1989, Cell 59:305-312) were iodinated to a known specific activity of about 70 /iCi/jug, and site densities of ICAM-1 and CD62 were determined by saturation binding, as previously described (Dustin and springer, 1988, J. Cell 20 Biol. 107:321-331). Site numbers assume binding of one IgG molecule per antigen molecule because saturation binding favors monomeric binding and because transmembrane proteins are immobile on glass-supported bilayers (McConnell et al., 1986, Biochim. Biophys. Acta 864:95-106) and at the highest 25 density were on average too far apart (32 nm) for bivalent binding. After initial measurements of bilayer incorporation, protein concentrations were adjusted to give round numbers of sites per /im2. The actual site densities were determined twice for each liposome preparation at each 30 density, in triplicate. The round values of sites per nm2 mentioned in results and actual values ± SD compared are as follows: ICAM-1,000: 950 ± 95; ICAM-1 250: 272 ± 16; CD62 400: 389 ± 62; CD62 200: 207 ±41; CD62 50: 64 ± 12; CD62 25: 34 ± 7. Incorporation of both CD62 and ICAM-1 did not 264353 affect the efficiency of incorporation compared to either alone. 6.3.6. ISOLATION OF POLYMORPHONUCLEAR LEUKOCYTES 5 Neutrophils were isolated from citrate anticoagulated whole blood following dextran-sedimentation and density separation over Ficoll-Hypaque (Miller et al., 1987, J. Clin. Invest. 80:535-544). Following isolation, neutrophils were stored in Hanks1 balanced salt solution 10 (HBSS; Gibco Laboratories, Grand Island, NY) supplemented with 10 mM HEPESat pH 7.3 and human serum albumin (0.1%) at room temperature for up to 6 hr. Before use in experiments, the neutrophils were washed into HBSS supplemented with 10 mM HEPES, 1.0 mM Mg2*, and 1.2 mM Ca2+ at pH 7.3, since CD18 15 interactions with ICAM-1 require divalent cations (Marlin and Springer, 1987, Cell 51:813-819), as does the CD62 interaction with its counterstructure (Geng et al., 1990, Nature 343:757-760). 6.3.7. LAMINAR FLPW A3SAY? A glass slide containing a planar bilayer was assembled in a parallel-plate laminar flow chamber (260 /im gap thickness) in which a uniform wall shear stress is generated. The flow chamber was mounted on the stage of an 25 inverted phase-contrast microscope (Diaphot-TMD, Nikon Inc., Garden City, NY). For continuous flow assays, neutrophils were resuspended at a concentration of 106/ml in HBSS supplemented with 10 mM HEPES, 1.2 mM Ca2+, and 1.0 mM Mg2+) and drawn through the chamber at controlled flow rates with 30 a syringe pump attached to the outlet. The wall shear stress was calculated from a momentum balance on a Newtonian fluid, assuming a viscosity of 1.0 centipoise. The flow rate was stepped down to allow measurements of cell binding at different shear stresses. Three minutes was allowed for 35 equilibration before the number of cells per unit area was 2 6 4 3 5 3 measured. Measurements on different areas of the bilayer were averaged, and no evidence for nonunifonnity was found. Attached neutrophils and their motion were observed with phase-contrast objectives and quantitated by analysis of S videotaped Images. For activation studies, neutrophils were treated with PMA (30 ng/ml, final concentration) for 5 min before perfusing the cell suspension through the flow chamber. io 6.3.8. DETACHMENT AS3AY5 For detachment assays, neutrophils (4 x l06/ml) were injected into the chamber through a port and allowed to settle. To determine the effect of CD18 upregulation, PMA was added to the neutrophils 1 minute before injection into 15 the flow chamber. All cells came in contact with the bilayer within 120 s, as indicated by their entry into the same focal plane. Controlled flow was applied following a 6 min incubation period. The initial shear force was 0.5 dyn/cm2; this force was increased every 20 s to a maximum of 20 36 dyn/cm2. All experiments were recorded on videotapes, and multiple fields of view were examined for each data point. 6.3.9. ANALYSIS QF FOTTPOPHIE pQLMNq Rolling velocities were measured for all cells in 25 two to five fields of view for each experiment at a given shear stress or ligand density. Results are presented as averages from experiments on different days. Velocities on CD62 were comparable whether cells were bound during shear flow or bound at stasis and then subjected to shear flow. 30 Images were recorded on a time-lapse videocassette recorder at real time and played back at six- or nine-fold slower speed. The tape was paused to mark the location of cells and the displacement of the center of individual cells was measured 2 to 4 s later. In experiments in which the effect 35 of chemotactic factors on neutrophil rolling was measured, 26 4 3 5 cells were first allowed to form rolling attachments to the bilayer at the indicated shear stress. Flow at the same rate was continued with medium lacking cells for 3 min, followed by addition of N-formyl methionyl leucyl 5 phenylalanine (fMLP) (1 x 10'® M, final concentration) to the perfusion media without stopping flow.

The present invention is not to be limited in scope by the specific embodiments described herein. Indeed, 10 various modifications of the invention in addition to those described herein will become apparent to those skilled in the art from the foregoing description and accompanying figures. Such modifications are intended to fall within the scope of the appended claims.

IS Various publications are cited herein, the disclosures of which are incorporated by reference in their entireties. 264353

Claims (62)

WHAT WE CLAIM IS:

1. An apparatus comprising: (a) a planar solid phase having a plurality of one or more substantially purified molecules contained on a surface of the 10 solid phase, said molecules being capable of mediating rolling of a leukocyte on said solid phase surface; (b) a base having (i) an inlet 15 manifold, (ii) an outlet manifold, and (iii; a recessed lip on its lover surface; (c) a deck positioned below the base, having (i) an entrance slot 20 communicating with the inlet manifold, and (ii) an exit slot communicating with the outlet manifold; (d) a longitudinal compressible gasket 25 positioned within the recessed lip of the base, its top surface contacting the base, and having an aperture of a size and Shape substantially congruent with the 30 deck; in which the planar solid phase is positioned on the bottom surface of the gasket, such that the solid phase surface with said molecules is communicating with the aperture of the gasket. 35 v.'/.. office M MAR 1996 - 98 264353

2. The apparatus of claim 1 further comprising a means for applying force whereby the gasket and solid phase are compressed together and their disengagement from the base is prevented.

3. The apparatus of claim 1 further comprising a pump connected to the outlet manifold.

4. The apparatus of claim 1 in which the solid phase comprises glass; and a planar lipid bilayer is present on said solid phase surface, said lipid bilayer containing said molecules.

5. The apparatus of claim 1 in which said molecules are CD62.

6. The apparatus of claim 4 in which said molecules are CD62. HZ. patent ofi- 1 4 MAR 1996 - 99 - 4353

7. The apparatus of claim 1, in which said molecules are selected from the group consisting of CD62, ELAM-1, and MECA-79, and any combination of the foregoing.

A method of collecting or purifying leukocytes from a fluid sample comprising: (a) contacting a surface of a solid phase with a fluid sample containing leukocytes, said surface containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a le* Jcocyte on said solid phase surface, in which the contacting is under conditions that provide relative Movement between the solid phase and the leukocytes in the sample, thereby allowing a plurality of leukocytes in the sample which express a binding partner to said molecules to roll on the surface of the solid phase; and (b) recovering the leukocytes that roll on the surface. HZ. ■' ' H MAR 1996 , WlCCtVEP - 100-

9. Tha method according to claim 8 in which the relative movement between the solid phase and the leukocytes is such that physiologic shear stress is achieved at the siurface of the solid phase.

10. a method of collecting or purifying leukocytes from a fluid sample comprising: (a) contacting a surface of a solid phase with a fluid sample containing leukocytes, said surface containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a leukocyte on said solid phase surface, in which the contacting is under conditions that provide relative movement between the solid phase and the leukocytes in the sample such that bulk flow is induced and shear stresses at the surface of the solid phase are in the range of from substantially 0.5 to substantially 4.0 dynes per square centimeter, thereby allowing a plurality of leukocytes in the sample which express a binding partner to said molecules to roll on the surface of the solid phase; and (b) recovering the leukocytes that roll on the surface.

11- The method according to claim 8 in which the leukocytes are recovered by recovering the solid phase and removing the leukocytes from the solid phase. N.2. PATENT.office 14 MAR 1996 264353 - 101 -

12- The method according to claim 8 or 10 in which said molecules are selected from the group consisting of CD62, ELAM-1, and MECA-79, and any combination of the foregoing.

13. The method according to claim 10 in which the solid phase is comprised of glass fibers, said molecules are CD62, and a planar lipid bilayer containing CD62 is present on the surface of the solid phase.

14- A method for analysis of a fluid sample containing leukocytes, comprising (a) contacting a surface of a solid phase with a fluid sample containing leukocytes, said surface containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a leukocyte on said solid phase surface, in which said contacting is under conditions that provide relative movement between the solid phase and the leukocytes in the sample, thereby allowing a plurality of leukocytes which express a binding partner to said molecules to roll along the surface of -the solid phase; and (b) detecting or quantifying the leukocytes that roll along the surface.

15. The method according to claim 14 in which the relative movement between the solid phase and the leukocytes is such that physiologic shea^ stress is achieved at the surface of the solid p patent PEROT 1 h. MAR 1998 r6cove& 264353 - 102

16* A method for analysis of a fluid sample containing leukocytes, comprising: (a) contacting a surface of a solid phase with a fluid sample 9 containing leukocytes, said solid phase containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a 10 leukocyte on said solid phase surface, in which said contacting is under conditions that provide relative movement between the solid phase and the leukocytes in 15 the sample such that bulk flow is induced and shear stresses at the surface of the solid phase are in the range of from substantially 0.5 to substantially 4.0 dynes per square centimeter, thereby 20 allowing a plurality of leukocytes which express a binding partner to said molecules to roll along the surface of the solid phase; and (b) detecting or quantifying the 25 leukocytes that roll along the surface.

17. The method according to claim 14 or 16 in which said molecules are selected from the group 30 . consisting of CD62, ELAM-1, and MECA-79, and any combination of the foregoing.

18. The method according to claim 16 in which the solid phase is comprised of glass, said 35 molecules are CD62, and a planar lipid bilayer 264353 -103 containing CD62 is present on the surface of the solid phase.

19. An in vitro method for diagnosing a disease or 5 disorder in a subject, comprising (a) contacting a surface of a solid phase with a fluid sample derived from a subject and containing leukocytes, said surface 10 containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a leukocyte on said solid phase 15 surface, in which said contacting is under conditions that provide relative movement between the solid phase and the leukocytes in the sample, thereby allowing a 20 plurality of leukocytes which express a binding partner to said molecules to roll along the surface of the solid phase; and (b) measuring the number or percentage 25 of the leukocytes or a subset thereof that roll on the surface of the solid phase; in which an increase or decrease in said nuaber or percentage relative to 30 the number or percentage of (b) leukocytes or the subset thereof that roll on the surface from an equivalent fluid sample from a healthy individual, indicates the 35 presence of the disease or disorder in the subject] n-z- patent office 1 4 MAR 1996 fWCBVEp 20 25 30 264353 . -104-

20. The method according to claim 19 in which the relative movement between the solid phase and the leukocytes is such that physiologic shear stress is achieved at the surface of the solid phase. 5

21. The method according to claim 19 in which said molecules are selected from the group consisting of CD62, ELAM-1, and MECA-79, and any combination of the foregoing. 10

22. The method according to claim 20 in which the solid phase is comprised of glass, said molecules are CD62, and a planar lipid bilayer containing CD62 is present on the surface of the solid phase.

23. The method according to claim 19 in which the disease or disorder is a leukemia and the number or percentage of rolling leukocytes or a subset thereof in the sample derived from the subject is increased.

24. The method according to claim 19 in which the disease or disorder is a neutropenia and the number or percentage of rolling leukocytes or a subset thereof in the sample derived from the subject is decreased.

25. a method for identifying a compound capable of inhibiting a component of the inflammatory response, comprising (a) contacting a surface of a solid phase with a fluid sample containing leukocytes and containing a test compound, said 26 4 353 - 105 - surface containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a 5 leukocyte on said solid phase surface, in which said contacting is under conditions that provide relative movement between the solid phase and the leukocytes in 10 the sample, thereby allowing a plurality of leukocytes which express a binding partner to said molecules to roll on the surface of the solid phase; and 15 (b) determining the number or percentage of leukocytes or a subset thereof rolling on the surface of the solid phase, in which a decrease in said number or 20 percentage relative to the number or percentage of leukocytes or the subset thereof which roll in the absence of the test compound indicates that the test compound 25 is capable of inhibiting the inflammatory response.

26. The method according to claim 25 in which the relative movement between the solid phase 30 . and the leukocytes is such that physiologic shear stress is achieved at the surface of the solid phase.

27. The method according to claim 26 in which said molecules are C062. 35 I N.Z. PAK. . H MAR 1996 264353 - 106 - 10 IS 20 25 30

28. The method according -to claim 26 in which said molecules are selected from the group consisting of CD62, ELAM-1, and MECA-79, and any combination of the foregoing.

29- The method according to claim 25 or 27 in which the solid phase is comprised of glass, and a planar lipid bilayer in which contains said molecules is present on the surface of the solid phase.

30. A method for identifying a compound capable of promoting a component of the inflammatory response, comprising: (a) contacting a surface of a solid phase with a fluid sample containing leukocytes and containing a test compound, said surface containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a leukocyte on said solid phase surface, in which said contacting is under conditions that provide relative movement between the solid phase and the leukocytes in the sample, thereby allowing a plurality of leukocytes which express a binding partner to said molecules to roll on the surface of the solid phase; and (b) determining the number or percentage of leukocytes or a subset thereof rolling on the surface of the solid phase, in which an increase in said number or percentage relative to the number or percentage of leukocytes or the subset thereof which roll in the absence of the test compound indicates that the test compound is capable of promoting the inflammatory response. *64353 -107 -

31. The method according to claim 30 in which the relative movement between the solid phase and the leukocytes is such that physiologic shear stress is achieved at the surface of the solid phase.

32- a method for identifying a molecule capable of mediating rolling of a leukocyte on a solid phase surface containing said molecule, comprising: (a) contacting a surface of a solid phase with a fluid sample containing leukocytes, said surface containing a plurality of a substantially purified test molecule, in which said contacting is under conditions that provide relative movement between the solid phase and the leukocytes in the sample; and (b) detecting any leukocytes that roll on the surface, in which rolling of leukocytes on the surface indicates that the test molecule is capable of mediating rolling of a leukocyte on a solid phase surface containing the test molecule.

33. The method according to claim 32 in which the relative movement between the solid phase and the .leukocytes is such that physiologic shear stress is achieved at the surface of the solid phase.

34. The method according to claim 32 in which the solid phase is comprised of glass,-and a lanar lipid bilayer which contains the test molecules present on the surface of the solid phase. 108 10 15 20

35. An apparatus comprising: (a) a solid phase having a plurality of one or more substantially purified molecules contained on a surface of the solid phase, said molecules being capable of mediating rolling of a leukocyte on said solid phase surface; (b) inlet means for receiving a fluid sample and for permitting the sample to enter onto the surface of the solid phase; (c) outlet means for permitting the fluid sample after it has flowed across at least a portion of the surface of the solid phase to exit said surface, (d) a housing containing said solid phase, inlet means, and outlet means; and (e) a video camera or microscope operatively linked for viewing the surface of the solid phase.

36. 30 An apparatus comprising: (a) a planar solid phase comprising glass; (b) a planar lipid bilayer present on a surface of the solid phase, said lipid bilayer containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a leukocyte; (c) inlet means for receiving a fluid sample and for permitting the sample to enter onto the lipid bilayer; and 1 J? 1996 'VGD 264353 -109-

37 • A solid phase having at least one surface containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a leukocyte on the solid phase surface, wherein the solid phase has a plurality of zones on said surface, each zone containing a different said molecule.

38- A solid phase having at least one surface containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a leukocyte on the solid phase surface, wherein the solid phase is a planar phase comprising glass, and a lipid bilayer is present on said surface, said lipid bilayer containing said molecules. 264353 - 110-

39. The solid phase of claim 37 and 38, further comprising a plurality of leukocytes adhering to said surface of the solid phase.

40. The solid phase of claim 37 or 38 wherein said molecules are selected from the group consisting of CD62, ELAM-1, and MECA-79, and any combination of the foregoing.

41. The solid phase of claim 37, wherein the solid phase comprises a plurality of glass fibers.

42. The solid phase of claim 37 in which the solid phase is plastic.

43. The solid phase of claim 37 or 38, wherein said molecules are immobilized.

44. The solid phase of claim 38 in which the solid phase is a microscope slide.

45. The solid phase of claim 38 wherein said molecules are CD62.

46. An apparatus comprising: (a) a solid phase having at least one surface containing a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a leukocyte on said solid phase surface; (b) an inlet for receiving a fluid sample and for permitting the sample to enter onto the surface of the solid phase- PATENT H MAR 1996 264353 - Ill - (c) an outlet for permitting the fluid sample after it has flowed across at least a portion of the surface of the solid phase to exit said surface; (d) a housing containing said solid phase, inlet, and outlet; (e) a means associated with the inlet, for introducing the fluid sample into the housing; and (f) a means associated with the outlet, for removing the fluid sample from the housing.

47. The apparatus of claim 46 in which the physiologic shear stress can be achieved at the surface of the solid phase by adjusting the bulk flow rate through the use of a peristaltic pump or syringe pump or by changing chamber geometry or by any other method known in the art.

48. The apparatus of claim 46 in which the shear stress is in the range of substantially 0.5 to substantially 4.0 dynes per square centimeter.

49 . The apparatus of claim 46 in which the means associated with the outlet is a pump.

50. The apparatus of claim 46 in which said molecules are selected from the group consisting of CD62, ELAM-1, and MECA-79, and any combination of the foregoing^

The apparatus of claim 46 in which said molecules are CD62. 26 4353 - 112-

52. The apparatus of claim 46 in which the solid phase is a planar phase comprising glass, and a planar lipid bilayer is present on said surface, said lipid bilayer containing said molecules.

53. A solid phase having at least one surface comprising: (a) a plurality of one or more substantially purified molecules, said molecules being capable of mediating rolling of a leukocyte on the solid phase surface; and (b) a plurality of leukocytes adhering to the solid phase surface.

54. An apparatus according to any one of claims 1,35,36 or 46 substantially as herein described or exemplified.

55. A method according to any one of claims 8 or 10 substantially as herein described or exemplified.

56. A method according to any one of claims 14 or 16 substantially as herein described or exemplified.

57. A method according to claim 19 substantially as herein described or exemplified.

58. A method according to claim 25 substantially as herein described or exemplified.

59. A method according to claim 30 substantially as herein described or exemplified.

60. A method according to claim 32 substantially as herein described or exemplified. J 264353 113-

61. A solid phase according to claim 37 or 38 substantially as herein described or exemplified.

62. An apparatus according to claim 46 substantially as herein described or exemplified. CENTER FOR BLOOD RESEARCH, INC By their Attorneys HENRY HUGHES Per: N.Z. PATENT OFFICE 1MAR 1996 RECEIVED