remove the cells from the surface will be greater if the cells have the ... interesting adhesive interactions in vivo are between a cell and a non-cellular sub-.
J. Cell Sci. 70, 93-110 (1984)
93
Printed in Great Britain © The Company of Biologists Limited 1984
ADHESION OF NEUTROPHIL LEUCOCYTES UNDER CONDITIONS OF FLOW J. V. FORRESTER* AND J. M. LACKIEf Departments of Ophthalmology and of Cell Biology, Glasgow University, Glasgow G12 8QQ, Scotland
SUMMARY
By passing a suspension of polymorphonuclear leucocytes through a parallel-plate chamber their adhesiveness can be assessed by scoring the number of cells trapped on the lower plate, and the fluid shear stress can be defined for a given flow rate. Since the adhesiveness of the cell at the instant of collision must exceed the distractive shear if the cell is to stop, the kinetics of cell accumulation provide a measure of the adhesiveness of the leucocytes and the adhesive interaction can be quantified. This measure of adhesion does not suffer from the complication that the force required to remove the cells from the surface will be greater if the cells have the opportunity to spread before the distractive force is applied. The assay is described in detail and the results of modifying the surface of the flow chamber and altering the composition of the suspension medium are used to illustrate the method. Plasma proteins generally seemed to reduce the adhesiveness of neutrophil leucocytes, whether they were present as a coat of adsorbed protein or in the suspension medium during perfusion. Neutrophil leucocytes, unless suspended in relatively high concentrations of plasma, were considerably more adhesive than other cells that have been tested in this assay system.
INTRODUCTION
Despite the importance of cell adhesion in many cellular activities, no wholly satisfactory method for measuring cellular adhesiveness is available. In many cases only qualitative comparisons of adhesion under different conditions are required, yet even this simpler problem poses difficulties. Before describing an assay system, which we feel has some particular merits, it is perhaps worth commenting briefly upon the deficiencies of more familiar assays, well recognized by those who use them, but which are often left unstated. Common assay systems are of two kinds, those relying upon cell—cell adhesion, in which the kinetics of aggregation are used as an indicator of individual cell adhesiveness, and those in which a population of cells is allowed to attach to a substratum and the percentage that resists a subsequent distraction procedure is taken as a measure of cell adhesiveness. Aggregation assays have several advantages but many of the most interesting adhesive interactions in vivo are between a cell and a non-cellular substratum, and cannot be tested directly. Subsidiary criticisms may also be made concerning the sensitivity of many aggregation systems to the viscosity of the medium, the adhesion of cells to the walls of the vessel and the formation of specialized junctional •Present address: Department of Ophthalmology, University of Aberdeen, Scotland, f Author to whom correspondence should be addressed.
94
jf. V. Forrester andjf. M. Lackie
complexes between cells following the initial adhesion. Although these subsidiary problems can be circumvented the basic problem remains, that it is probably inappropriate to equate cell-cell adhesion with cell-substratum adhesion. Most cell—substratum adhesion assays are variations on the general principle of applying a distractive force to cells that have been permitted to form an adhesive contact, and then measuring either the proportion of cells that resist distraction or the force required to remove all the cells. Several fundamental problems exist if it is the adhesiveness perse that is of interest. In almost all cases the adherent cells flatten and spread upon the test substratum. By spreading, the cell increases the area over which the adhesive interaction occurs, and thus the extent of spreading affects the total adhesive interaction; by lowering the profile of the cell the distractive force is reduced and the apparent adhesiveness is further increased. The spreading of a cell is not strictly analogous, except in very unusual circumstances, to the wetting of a surface by a fluid droplet; rather, spreading depends upon the active locomotory behaviour of the cell (Grinnell, 1978). Substances that interfere with the motile machinery will, therefore, affect the apparent adhesiveness, even though the adhesion per unit area remains unchanged. There is also confusion, rarely recognized, between the effect of altering the absolute strength of an adhesion and the effect of altering the life-span of an adhesion. Long-lasting adhesions will tend to encourage a very flattened morphology in a moving cell, because adhesions at the rear are relinquished only slowly as the cell moves forward; very transient adhesions will be characteristic of more rounded cells, and these cells will appear less adhesive in distraction assays, even though the absolute strength of each adhesive contact may be the same. Whilst the above discussion is by no means an exhaustive critique of aggregation and distraction assays, it does suggest that other assays should be considered (see Hubbe, 1981, for a much more detailed discussion). For the cells of the blood, escape into the tissues depends upon the formation of an adhesion to the blood-vessel wall that is adequate to immobilize the cell despite the constant distractive force exerted by the fluid medium. Such an adhesive interaction must precede flattening, although subsequent shape changes may stabilize the initial contact, and variation in the fluid shear force will alter the probability that cells halt upon the vessel wall. The assay described here is based upon that of Doroszewski and his co-workers (Doroszewski, Skierski & Przadka, 1977; Zachara & Doroszewski, 1978), although they used the assay in a rather different way in investigating the adhesion of leukaemic cells. The method involves the measurement of the kinetics of attachment of cells to the wall of a parallel-plateflowchamber, of the type described by Hochmuth, Mohandas & Blackshear (1973), as a cell suspension is passed through. The assay is particularly appropriate for cells of the blood, but has more general applicability for attempting to separate the 'adhesive' and 'spreading' components that contribute to the morphology of a cell on an inert substratum. It might be considered a method whereby the instantaneous adhesive capacity of a cell can be measured and, as we also show, the subsequent spreading behaviour of the cells can also be monitored. By measuring the value of the critical shear force required to counter the adhesive interaction the adhesion force can be calculated directly.
Leucocyte adhesion from
flow
95
Various modifications to the assay and the simple analytical method we describe should, we hope, recommend this as a system for studying cell adhesion in general and the interaction of blood cells with the vessel wall in particular. In describing the method we also present data on the interaction of neutrophil leucocytes with a variety of substrata, in some cases confirming observations with other assays, and extending previous work with plasma proteins and extracellular matrix components (Forrester, Lackie & Brown, 1983; Wilkinson, Lackie, Forrester & Dunn, 1984).
MATERIALS
AND
METHODS
Preparation of cells Human neutrophils were prepared as described previously (Forrester et al. 1983). Peripheral blood from healthy adult donors was collected in heparin without preservative (200 units/ml) and separated by dextran sedimentation and density gradient centrifugation on Ficoll-Hypaque (Pharmacia, Uppsala). The cells were washed twice in HEPES-buffered saline solution (HBSS). Contaminating red cells were removed after the first wash by hypotonic lysis. After the second wash the cells were passed through a Nitex gauze filter (mesh size 10 fun) (Plastok Associates, Birkenhead), which removes cell clymps and leaves a monodisperse cell suspension. Cells were counted and kept on ice until required, and were used within 15 min of washing. Cell suspensions contained >95 % neutrophils; cell viability was greater than 95 % as judged by Trypan Blue dye exclusion.
Adhesion assay system Aflowchamber was constructed between two microscope slides as described previously (Forrester et al. 1983); the system is similar to that described originally by Hochmuthe/ al. (1973) and adapted by Doroszewskie/a/. (1977). The two slides were separated by a gasket of Nescofilm (Nippon Shoji Kaisha Ltd, Osaka, Japan) approximately 150 /im thick in which a channel 4 mm wide and 40 mm long had been cut. The chamber was held in place by a metal and Perspex (lucite) clamping device with inlet and outlet ports that coincide with 4 mm holes drilled in the upper microscope slide (Fig. 1). Neutrophil suspensions were perfused through the chamber with a mechanical syringe drive. The flow rate could be varied by altering the speed of the plunger or by using syringes of different capacity. The assembled flow-chamber was placed on the stage of an inverted Leitz microscope with an air-curtain incubator (37 °C), and cells adhering to the lower surface of the flow channel were Flow-chamber
s^_»r- to videotape
' f
1
7
micro-
scope
/'z
=• o u t
1
slide
J_
slide
A
i! i b
O |
Plan
Fig. 1. Cross-section and plan view of the flow chamber. The inlet and outlet ports are drilled in a perspex block, which is clamped over the glass chamber, withsilicone-rubber gaskets between the perspex and the glass. In practice we have used the chamber on an inverted microscope with phase optics.
96
jf. V. Forrester andj. M. Lackie
viewed through a monochrome video-camera/monitor system, which was connected to a time-lapse video-recorder (National: NV-8030) with a time-date generator (FOR-A Ltd, Tokyo). By replaying the video-tape it was possible to count the numbers of adherent (stationary) cells at different times, the moving cells showing only as a blur. The results are expressed as the number of adherent cells per unit area with time and also as a 'collection efficiency', which takes account of the total number of cells that pass through the chamber during the course of the assay. The collection efficiency was calculated from the known cell-delivery rate and from a least-squares regression line fitted to the kinetic data for cell accumulation. Over the first 5 min the kinetics of accumulation were linear and the regression coefficient was usually >0-95. The assay system permits measurement of both the early adhesion events and the later spreading behaviour of the cells. The distractive force per cell can be calculated according to the formula: F = 32XR2X Tw , where R is the radius of the cell (taken to be 5 /xm) and Tw is the shear stress at the wall. The wall shear stress can be estimated from the expression Tw = 6 flv I'1, where fi is the viscosity (taken to be 10~3kg~' s" 1 ), v is the average velocity of fluid in the chamber and / is the depth of the chamber. This calculation is based on that given by Hubbe (1981).
Presentation of results Because only a limited number of experiments can be done on a single preparation of cells, before the cells have been held in vitro for too long to be used with confidence, we have chosen to present data only from those experiments in which the complete 'set' of experimental and control measurements were made. We have also chosen not to attempt to make comparisons between different preparations of cells, since the aim in this paper is to illustrate the method. In nearly all cases the results have been repeated several times on other occasions (and in many cases are broadly similar to the results obtained using other assay systems).
Reagents Proteins. Human serum albumin (HSA) was from Behringwerke (Marburg), transferrin, trypsin and superoxide dismutase from Sigma (London), fibrinogen from Calbiochem, thrombin from Sigma, plasminogen from Kabi (Sweden) and immunoglobulin G from Miles (Slough). Chemotactic factors. Casein was from Merck (Darmstadt) and formylmethionyl-leucylphenylalanine (fMet-Leu-Phe) from Miles (Slough). Antiproteases. Alpha-2-macroglobulin (ct-2-M) was prepared from human plasma as described previously (Forrester et al. 1983). The purity of the final product was checked using sodium dodecyl sulphate/polyacrylamide gel electrophoresis (SDS/PAGE) and by immunochemical analysis. Alpha-2-antiplasmin (a-2-AP) was a generous gift from Dr D. Collen, University of Louvain. Alpha-1-antitrypsin (a-l-AT) was purchased from Sigma (London). Extracellular-matrix components. Type I collagen was prepared from rat-tail tendons as described previously (Brown, 1982). Type II collagen was prepared from bovine vitreous humour by extraction of the solid material, obtained by ultracentrifugation of crudely macerated vitreous gel, with acetic acid. Purity of the protein was monitored by SDS/PAGE. A mixture of collagen types I and III was prepared as a crude salt (0-5 M-NaCl) extract of human placental tissue by the method of Glanville, Rauter & Fietzek (1979). Type IV collagen was purchased from Sigma (London). Fibronectin was prepared from calf serum by affinity chromatography on gelatinSepharose as described previously (Brown & Lackie, 1981).
RESULTS
Fluid dynamics of the flow chamber A fluid stream through the flow chamber was achieved with a constant pressure device that eliminates pulsatile flow. The velocity of the fluid stream could be varied by fixed doubling increments from 0-0875 /il/s to 2-8 /ul/s or 150/xm/s to 4800/im/s. Under these conditions, using a chamber of this design, flow should be laminar even
Leucocyte adhesion from
flow
97
at the highest velocities since the Reynolds number is well below unity (Doroszewski et al. 1977; Hochmuth et al. 1973); the distance required for the flow profile to stabilize will, under these conditions, be very short (Van Wagenen & Andrade, 1980). The values quoted for fluid velocity represent the average value but, since flow is laminar, the fluid velocity near to the walls of the chamber will approach zero. Estimates given by Zachara & Doroszewski (1979), based on the movement of small particles, indicate that the velocity of the fluid close to the wall is approximately one fifth of the magnitude of the average velocity. The velocity profile across the chamber is of considerable importance in trying to estimate the strength of the adhesive bond that resists the fluid distraction. If, however, we are concerned only with the delivery of cells into the chamber in unit time the average velocity is the essential parameter. Adhesion of cells underflow conditions Monodisperse suspensions of neutrophils were perfused through the chamber and the kinetics of attachment to the lower surface of the flow chamber were determined as described in Materials and Methods. The numbers of cells attached to clean glass increased linearly with time (Fig. 2) and the collection efficiency could be calculated knowing the cell concentration, the flow rate, and the ratio of the area under observation to the total area of the chamber. Values for collection efficiency are shown in Table 1. Cells that made contact with a clean glass substratum almost invariably formed an adhesion sufficiently strong to render them stationary. Initially cells on clean glass remained rounded, but after 20-30 s they adopted a more flattened morphology and showed continuous ruffling movements (Fig. 3). Increasing the fluid flow rate by a factor of 4 or 8 readily caused detachment of rounded cells but spread cells were more difficult to dislodge. The cell speed close to the wall and just prior to 100-.
c
50-
E 3
1
T 2
Time (min) Fig. 2. The kinetics of cell accumulation on clean glass (O) and on glass coated with HSA ( • ; 10 mg/ml in the coating medium). There is some deviation from linearity in the early stages, but generally the regression coefficient of a line fitted by least-squares is very close to unity.
J. V. Forrester and J. M. Lackie Table 1. Collection efficiency of human neutrophils on clean glass and HSA-coated glass at various times after starting Collection efficiency (%) Flow time (min) 1-0 2-0 5-0
A
f
Clean glass
HSA-coated
96-2 84-6 71-6
23-1 33-5 39-8
attachment varied from 6-30/xm/s. On clean glass, cells were rarely observed to roll along the chamber wall; contact seemed always to lead to an adhesion. Cell adhesion to glass coated with protein was, however, less likely to halt the cell (Fig. 2) and there was considerable variation depending upon the adsorbed protein. On a coat of HSA
Fig. 3. Four 'frames' from a time-lapse videotape record of cell accumulation in the flow chamber. The photographs are from the monitor screen and also show the time and date. Notice that some cells are changing shape and on the complete tape cells can be seen to be moving, often against the flow of medium. Flattening of the cells takes 2-3 min at the very least, and once a cell has flattened it is unlikely to detach.
Leucocyte adhesion from flow
99
(10 mg/ml), rolling cells were observed at slowflowrates but less frequently at speeds >500 ;Um/s. Cells spread well on HSA coats and not only showed signs of membrane ruffling but also appeared to move over the substratum despite the shearing stress caused by the fluid flow. Since neutrophil adhesion under physiological conditions will always be to surfaces with adsorbed protein we have used HSA-coated chambers in all the experiments except those in which other proteins were tested. Effects of varying fluid velocity on cell adhesion Increasing the velocity of the fluid cell suspension from 300-1200 jUm/s through the flow chamber produced a concomitant increase in the number of adherent cells per unit area (Fig. 4), but estimates of the collection efficiency, which take into account the delivery rate of cells to the chamber, indicated that the percentage of cells adhering was independent of the flow rate within this range (Table 2). For fluid flow rates of 300-1200 jUm/s the percentage of cells adherent after 5 min ranged between 18-7 % and 31-1% but within an experiment the differences were not statistically significant. These results suggest that the cell—substratum adhesion is sufficient to resist a 140 - i
Flow: 1200fflTi/s
600/im/s
Time (min)
Fig. 4. Accumulation of cells at different flow rates. The results from two experiments are shown to give an indication of the variability between chambers with a single batch of cells. At the fastest flow rate the delivery of cells is greater and the kinetics deviate from linearity after 5 min, possibly because the adherent cells begin to perturb theflowadjacent to the wall.
100
J. V. Forrester andj. M. Lackie
Table 2. Collection efficiency at different times as a function of the average flow rate Collection efficiency (%) A
Flow time (min)
300jUm/s
600 /an/s
1200jum/s
1-0 a b
21-1 33-7
20-5 31-1
191 61-0
2-0 a b
24-3 23-8
24-9 29-7
21-6 42-3
5-0 a b
18-7 25-3
27-5 28-8
23-1 31-1
Distraction force at 1200 ^m/s is 3-8 X 10"" N. Results from two experiments done on separate occasions with different cell suspensions are shown (a, b).
distractive force of approximately 4x 10~ n N. At flow rates of 2400 jum/s and above, cell attachment to the chamber was generally inhibited, which suggests that a shearing force of 7-7xlO~ u N exceeds the strength of the adhesive interaction. Very slow flow rates (below 150 jUm/s) permitted attachment, but because few cells were entering the chamber the number attaching was low and it was difficult to differentiate between attached cells and slow-moving cells when analysing the videotape record. The relatively high collection efficiency at a flow rate of 1200//m/s suggests that neutrophils are considerably more adhesive than the L-1210 cells used by Doroszewski et al. (1977), and BHK cells (unpublished data). Effect of cell concentration on adhesion Increasing the concentration of the cell suspension flowing through the chamber produced increased numbers of adherent cells (Fig. 5), but the collection efficiency remained constant (Table 3). In chambers coated with 1% HSA approximately 40—50 % of the cells entering the chamber adhered over the first 5 min, irrespective of the cell concentration. At later times the accumulation of cells on the chamber wall affected the local flow pattern and an increasing proportion of cells may attach to cells that had already been immobilized, or may attach in regions of 'slack water' downstream from aggregates, rather than to the protein-coated surface being tested. A practical point to be borne in mind is that with higher cell concentrations in the reservoir, the probability of aggregation is increased and aggregates may disturb the flow. Reproducibility of the assay The reproducibility of the assay is sensitive to several factors, and considerable care was taken to ensure that the cell suspension was of single cells and that the cells were kept on ice for a limited period before being used in the chamber. Some variability in the collection efficiency was observed during the first minute (Tables 1, 3),
Leucocyte adhesion from flow
101
140 - i
100-
SI T3
ro
60-
0-25x106 cells/ml
E Z
20-
Time (min) Fig. 5. Accumulation kinetics as the number of cells per ml is increased. The gradients of the lines are directly proportional to the delivery rate, the collection efficiency remains constant. Each line represents a single experiment; shaded areas indicate the range of adhesion values at each cell concentration.
Table 3. Collectionefficiency as the cell concentration is increased Cell concentration (106ml) Collection efficiency (%)
0-125
0-25
1 min a b
42 60
57 53
1-0 83 53
2 min a b
38 49
54 50
62 50
5 min a b
35 43
53 49
50 49
^
Results from experiments on two separate days are shown (a, b).
probably because during the initial stages the cell suspension was still warming up to 37 °C. The cell suspensions were agitated in a standard fashion immediately before being used and each sample was agitated only once, since this mechanical trauma affects the subsequent behaviour of the cells. Care obviously has to be taken to avoid air-locks or other obstructions in the flow channel.
102
J. V. Forrester and y. M. Lackie
Table 4. Collection efficiency as the composition of the suspending medium is altered Buffer conditions Collection efficiency (ft'•>) 1 min 2 min 5 min
A
HBSS
CMF
CMF + Ca2+
CMF + Mg2+
35-9 33-7 32-9
9-8 7-3 5-9
6-0 6-1 6-2
19-7 25-4 28-7
CMF, Ca 2+ , Mg^-free HBSS.
Under the standardized conditions used, human neutrophils passed through a chamber of clean glass coated with 1 % HSA consistently gave a collection efficiency of between 3 0 % and 5 0 % , a result that is in accord with previous estimates of adhesion to protein-coated glass under static conditions (Brown, 1982; Brown & Lackie, 1981; Forrester & Lackie, 1981).
The effects of divalent cations Divalent cations are essential for neutrophil adhesion, Mg2"1" being much more important than Ca2+ (Hoover et al. 1980; Atherton & Born, 1973). In order to demonstrate that the flow-chamber assay is sensitive to standard procedures that modify neutrophil adhesion, the effect of varying the divalent cation composition of Control
CMF+MQ :2+
Time (min) Fig. 6. Adhesion of neutrophils to HSA-coated glass with (A — A) and without (O—O) divalent cations in the medium. Addition of calcium ions (A — A) does not restore adhesion, whereas the addition of magnesium ions ( • — • ) brings adhesion to levels very similar to those with normal physiological saline. (CMF, Ca 2+ , Mg2+-free HBSS).
Leucocyte adhesion from
flow
103
Table 5. Collection efficiency as the amount of HSA used to coat the chamber is increased HSA concentration (mg/ml) Collection
efficiency (%)
A
r
0
0-01
-i
0-1
1-0
100
1 min
73-9
67-4
105-1
59-9
53-3
2min 5 min
70-8 68-9
67-2 67-2
102-7 101-4
57-9 56-7
36-3 26-1
The results are all from a single experiment although comparable results on more limited concentration ranges have been obtained on a number of occasions.
the suspending medium was tested. In the complete absence of divalent cations neutrophil adhesion was markedly reduced (Table 4, Fig. 6), and the addition of calcium alone did not restore basal levels of adhesion. The addition of magnesium ions alone was, however, almost sufficient to give normal adhesiveness. Effect of plasma proteins Neutrophil adhesion in vivo occurs in the presence of plasma, a complex mixture of proteins. Most of the experiments reported here have, therefore, been performed on a glass surface coated with one of the major plasma proteins, HSA. The concentration of HSA used to coat the chamber was important, and a minimum concentration of 1 mg/ml was required to reduce adhesion below that to uncoated glass (Table 5). The dose-dependence for the effect of HSA on adhesion seems to be rather complex, with a peak in collection efficiency when 0 - l mg/ml HSA is used to coat the glass. Previous experiments with HSA-coated surfaces using static assays have given comparably complex results; there is no obvious explanation, although the binding of native HSA to glass is very weak. Table 6. Effects of various plasma proteins on collection efficiency Collection efficiency (%) at 5 min Protein coat Control HSA, 10 mg/ml IgG, 5 mg/ml Transferrin, 2 mg/ml a-1-antitrypsin, 1 mg/ml a-2-macroglobulin, 0-5 mg/ml 1-5 mg/ml cr-2-antiplasmin, 100/ig/ml Lys plasminogen, 1 mg/ml
(a)
(b)
(c)
68-9 26-1 37-2 — — — — 36-3 43-3
71-6 39-7 — 2-3 — — 41 — —
70-7 — — — 14-5 30-1 — — —
Comparisons should, strictly, be made only within each of the three experiments (a, b, c) although there is quite good agreement between the control (no protein) collection efficiencies.
104
J. V. Forrester andj. M. Lackie
Several other plasma proteins were used to coat the glass chamber (Table 6). Adhesion to coats of alpha-2-macroglobulin (0 • 5 mg/ml in the coating medium), alphal-antitrypsin(l mg/ml), fibrinogen(3 mg/ml), alpha-2-antiplasmin(100/zg/ml) and immunoglobulin Gl (2-5 mg/ml) was little different from adhesion to HSA coats. Adhesion to alpha-2-macroglobulin coats prepared with concentrations of more than 1 mg/ml was, however, markedly reduced (Forrester et al. 1983). Cells were observed to contact the wall of the chamber and to roll along at around lOjitm/s, but adhesions did not seem to form, or were too weak to immobilize the cells. Glass coated with transferrin (2 mg/ml) also seemed to be a particularly non-adhesive surface for neutrophils (Table 6). A more physiologically realistic situation obtains when neutrophils are suspended in medium containing plasma proteins and are perfused through the chamber in the presence of these proteins. Under these circumstances, neutrophil adhesion was considerably lower than when the cells were suspended in protein-free medium (Fig. 7, Table 7). This effect was most marked in the presence of albumin, alpha-2macroglobulin, plasminogen and anti-plasmin, and was less marked with fibrinogen.
130 -i
100 -
E
50 -
Time (min)
Fig. 7. Adhesion to HSA-coated glass when cells are perfused through theflowchamber with plasma proteins in the suspension medium. ( • — • ) No protein in suspension medium; ( • — • ) IgG, 2-5 mg/ml; (O —O) fibrinogen, 3 mg/ml; (A —A) a-2antiplasmin, 100fig/ml; (A—A) HSA, 10mg/ml; (•—•) plasminogen, 1 mg/ml.
Leucocyte adhesion from flow
105
Table 7. Collection efficiency on HSA-coated glass with protein in the suspension medium Protein solution
Collection efficiency (%) (at 5 min) 34-8 8-4 25-2 2-9 11-2 30-5 1-2
Control HSA, 5 mg/ml Fibrinogen, 3 mg/ml Plasminogen, 1 mg/ml Antiplasmin, 100 /^g/ml IgG, 2-5 mg/ml a-2-macroglobulin, 1 mg/ml
IgG did riot appear to affect adhesion to HSA-coated glass. Cells suspended in autologous whole plasma (25 %) did not adhere either to clean glass or to HSA-coated glass, although the 'clean' glass surface would rapidly have become coated with protein from the suspending medium (Fig. 8, Table 8). The effect is to bring the adhesiveness of the neutrophils down to levels that are similar to those observed for 300-1
2
200-
c a)
