The effects of hyaluronate on rabbit neutrophil adhesion were studied using a variety of techniques. Exogenous hyaluronate inhibited neutrophil aggregation under conditions of both turbulent flow and constant shear rate. Hyaluronate also inhibited neutrophil adhesion to glass. Inhibition was dose-dependent above 100 μg ml−1 and a minimum molecular weight for hyaluronate of 1 × 104 was required. These effects were not simply the result of increased bulk viscosity of the hyaluronate-containing medium, nor did they appear to be mediated by putative cell-surface receptor mechanisms. Instead, physical factors such as steric hindrance and/or changes in the interfacial free-energy exchange at the cell surface due to the unusual hydrodynamic properties of the hyaluronate molecule were considered to be more important.

Since neutrophil migration in vivo occurs through hyaluronate-rich connective tissue matrices, the relevance of these findings for processes such as inflammation and wound healing is clear.

Cell migration in vivo occurs through tissues that are rich in glycosaminoglycans, particularly hyaluronate (HA) (for a review see Comper & Laurent, 1978). Hyaluronate contributes greatly to the viscoelasticity of connective tissues, and is likely to affect the movement of cells through tissues. The effects of this molecule on cell migration in vitro are therefore of relevance to processes such as inflammation and wound healing, which require invasion of tissues by cells.

HA occurs in adult tissues at concentrations between 0·05 and 3·0 mg ml−1 (Comper & Laurent, 1978), and in certain embryonic tissues at between 12 and 20 mg ml−1 (Pratt, Larsen & Johnston, 1975). It is an unbranched helical polymer composed of repeat disaccharide units of β-N-acetyl-D-glucosamine and β-D-glucuronic acid (Winter & Arnott, 1977), which in solution has the properties of a relatively stiff random coil (Scott & Tigwell, 1978) with a very large hydrodynamic volume, approximately 1000 times greater than its volume in the dry state (Ogston & Stanier, 1951 ; Laurent & Gergely, 1955). Physiological concentrations of HA inhibit neutrophil locomotion induced by chemotactic factors in vitro (Forrester & Wilkinson, 1981), partly by interfering with gradient formation and partly by inhibiting binding of chemoattractants to the cell. Both these effects are probably the result of the physical nature of the molecule. The same molecular properties of HA are likely to alter neutrophil adhesiveness, which may also contribute to changes in neutrophil motility.

Previous studies have shown that HA may inhibit (Underhill & Dorfmann, 1978), promote (Pessac & Defendi, 1972), or have no effect (Knox & Wells, 1979) on cell adhesion. On theoretical grounds alone, it is probable that increasing concentrations of HA would inhibit cell adhesion simply by increasing the viscosity of the extracellular medium. In the present study, we found that HA inhibited neutrophil adhesion, but there was no simple relationship between the degree of inhibition and the bulk viscosity of the medium. Inhibition of adhesiveness did not appear to be mediated by cell-surface receptors, but was probably the result of physical factors such as steric hindrance or changes in the interfacial free-energy exchange at the cell surface.

Cells

New Zealand white rabbits were injected intraperitoneally with 400 ml of sterile 150 mM-NaCl containing 0·1% oyster glycogen (Sigma Ltd) and the fluid was drained off 4 h later. The cells were stored in exudate fluid overnight. Before use, they were washed once in calcium- and magnesium-free balanced salt solution (CMF) buffered with 10 mM-HEPES (N-2- hydroxyethylpiperazine-N’-2-ethanesulphonic acid), pH 7·6, dispersed in CMF containing 1 mM-EDTA, passed through a 10-μm mesh nylon filter (Nitex; Begg and Cousland) to remove cell clumps and fibrin, and resuspended in HEPES-buffered balanced salts solution (HBSS) at a final concentration of 2 × 106 cells ml−1. Excess red cells were removed if necessary by brief (30 s) exposure of the cells to distilled water after the first wash. The monodisperse cell suspensions were checked visually in a haemocytometer for the absence of cell clumps, and for viability by dye exclusion.

Polysaccharides

The HA samples used in this study were the same as those described previously (Forrester & Wilkinson, 1981). Commercial hyaluronic acid (intrinsic viscosity [η], 59–67 ml g−1; mol. wt 1 × 104; Cleland & Wang, 1970) was from Miles Laboratories, Slough; it was prepared from human umbilical cord as a potassium salt and contained 2·3% protein and 3% chondroitin sulphate. Samples of high molecular weight hyaluronic acid were a gift from Dr E. A. Balazs, Columbia University, New York. They were prepared as sodium salts from rooster combs by repeated precipitation with cetylpyridinium chloride (Balazs, 1979) followed by heat treatment, and fractionation by ion-exchange chromatography on a (diethylamino)-ethyl-cellulose (DEAE-cellulose) column according to the method of Cleland, Cleland, Lipsky & Lyn (1968). Samples of various weight average molecular weights were eluted with 0-2 M-NaCl, exhaustively dialysed against several volumes of glass-distilled water to remove excess salts and other impurities, and lyophilized. The samples were stored at 4 °C in a desiccator before use. In this study samples with intrinsic viscosities of 600, 1600 and 3620 were used. These corresponded to weight average mol. wts of 2·9 × 105, 7·0 × 105 and 2·4 × 105, respectively. Protein concentrations of the samples were 3·0, 5·0 and 2·40%. Further samples of high molecular weight HA prepared from human umbilical cord were a generous gift from Pharmacia, Uppsala, Sweden (intrinsic viscosity, 2000; mol. wt 1 × 106). Tetrasaccharides of sodium HA were prepared by exhaustive digestion of HA solutions with testicular hyaluronidase (Worthington Biochemical, Freehold, N.J.) and elution of the digested material on a DEAE-cellulose column with 0·2 M-NaCl. Chondroitin 6-sulphate was obtained from Sigma Ltd, London. Dextrans (mol. wts 60–90 × 103 and 500 × 103) and dextran sulphate (500 × 103) were from Pharmacia AB. Heparin, 155 units/mg, was obtained as a dry powder from Sigma Ltd.

Other materials

D-glucuronic acid and N-acetyl-D-glucosamine were from Sigma. Ovine testicular hyaluronidase (sp. act. 300–800 units mg−1) was from Worthington. Crystalline (× 1) trypsin was from Sigma (proteolytic activity, 10000 BAEE units mg’1).

Aggregation assays

Neutrophil aggregation was assayed according to the method described previously (Lackie, 1977). Cells (1 × 106 polymorphonuclear leukocytes ml−1) were suspended in HBSS containing various concentrations of polysaccharides to a final volume of 1 ml in polystyrene vials (15 mm × 41 mm) and incubated at 37 °C for 60 min on a reciprocating shaker in a water bath with a 4 m−1 stroke, at 100 reciprocations min−1. Total particle number was estimated at time zero and at 60 min by withdrawing 0·5-ml aliquots from replicate vials, diluting the sample in 10 ml of 150 mM-NaCl and counting on a Coulter Counter model A with a 100-micron orifice tube, threshold setting 050 and aperture current switch setting 4. The results were expressed as the effect (f) of the various agents tested on neutrophil aggregation as :
formula

(Thome, Oliver & Lackie, 1977). Thus inhibition of aggregation is represented by a negative value and an increase in aggregation by a positive value for E.

Neutrophil aggregation was also assayed by the collision efficiency method for measuring cell adhesion (Curtis & Hocking, 1970). Aliquots (0·7 ml) containing 1 × 106 cells in polysaccharide solutions were placed in the wells of modified Couette viscometers, the outer cylinders of which were adapted to rotate at a constant shear rate of 10 s−1. The cell suspensions were incubated at 37 °C for 60 min. Aliquots were removed at various times and total particle counts were made with a haemocytometer. Total particle counts at time (t) were expressed as a percentage of the particle count at time zero.

Adhesion to substratum assay

Neutrophil adhesion to uncoated glass coverslips was assayed as described previously (Smith, Lackie & Wilkinson, 1979). Cell suspensions (1-ml aliquots) containing Na25lCrO4-labelled neutrophils (1 × 106 cells ml−1), in various concentrations of polysaccharides, were added to the wells of Linbro trays, which housed 13-mm diameter glass coverslips. The cells were incubated for 60 min at 37 °C. The coverslips were removed and washed by dipping, normal to the interface, 30 times in HBSS. The number of adherent cells was estimated by counting the radioactivity of the coverslips on a Wilj 2001 gamma counter. The results were expressed as the adhesion index (AI) as follows :
formula

Labelling of neutrophils

Neutrophils were labelled by suspending approximately 5 × 107 cells in 1 ml HBSS containing 150μCi of Na251CrO4 (100–300 mCi/mg Cr) (The Radiochemical Centre, Amersham). The cells were incubated for 45 min at 37 °C, washed 3 times in CMF, filtered through Nitex plastic mesh, and used immediately.

Viscosity measurements

The viscosity of the polysaccharide solutions was measured in a specially designed capillary microviscometer, similar to that described by McLean-Fletcher & Pollard (1980) and based on the principle of the ‘falling-ball’ viscometer (Van Wazer, Lyons, Kim & Cowell, 1963). Samples were drawn into capillary tubes (1·15 mm internal diam., 12·7 mm long) and warmed to 37 °C for 1 h. Viscosity was determined by measuring the time taken for a 0·78 mm diameter steel ball to fall through a fixed distance. An electronic stop-watch was triggered on and off by interrupting infrared beams between micro-emittors and sensors (Radiospares U.K. Ltd) placed at set distances along the length of the capillary tubes. Readings were highly reproducible to within 0·01 of a s, although considerable care was necessary to ensure that minute air bubbles were not trapped in the sample. According to Stokes Law, viscosity is inversely proportional to the velocity of the falling ball. Using a viscometer of similar dimensions, McLean-Fletcher & Pollard (1980) have shown that this relationship holds true over the viscosity range 1–12000 cP. The present viscometer was therefore calibrated in cP (1 cP = 10−3 kg m−1 s−1) using standard solutions of sucrose with known viscosities at various temperatures. Apparent viscosity readings for the polysaccharide solutions were extrapolated from the standard cune for sucrose.

Assay of proteolytic activity

Proteolytic activity of the hyaluronidase preparations was determined by the spectrophotometric assay of Schwert & Takenaka (1953). Samples contained 3 ml Tris buffer, 0·1 ml d-17-benzoyl-L-arginimic-ethylesterHCl and 0·01 ml of test solution. Activity was expressed as units/ml extrapolated from a standard curve for purified trypsin.

Interference reflexion microscopy

Cells that had adhered to glass in the presence of HA solutions were examined by interference reflexion microscopy to determine the size of the cell-substratum gap at the adhesion site (Curtis, 1964).

Aggregation of neutrophils in the presence of exogenous hyaluronate

Hyaluronate inhibited neutrophil aggregation under conditions of turbulent flow (shaker assay) (Fig. 1). The effect was dose- and molecular-weight-dependent. Consistently significant inhibition was observed at HA concentrations of 100 μg ml−1 and higher. In addition, a minimum molecular weight of 1 × 104 for HA was required before significant inhibition was observed (Table 1). Tetrasaccharides (mol. wt 1 × 108) and monosaccharides (TV-acetyl-D-glucosamine and β-D-glucuronic acid) had no reproducible effect on aggregation (data not shown). Conversely, increases in chain length of HA above mol. wts of 1 × 104 did not significantly increase the degree of inhibition of neutrophil aggregation at the minimum effective concentration of 100 μg ml−1 (Table 1), despite concomitant increases in the bulk viscosity of the HA solutions (Fig. 2). This suggested that the inhibitory effect of HA on neutrophil aggregation was not a direct consequence of the increased viscosity of the medium. To minimize the effects of bulk viscosity on the rate of shear, neutrophil aggregation was assayed using the method of Curtis & Hocking (1970), in which cells were allowed to aggregate in a modified Couette viscometer at a constant shear rate of 10 s−1 (see Materials and methods), in the presence and absence of HA. It has been shown that, provided the speed of rotation of the outer cylinder of the Couette apparatus is kept constant, this method of assaying cell aggregation kinetics is insensitive to differences in the bulk viscosities of samples, which can then be compared. The effect of HA (mol. wt 1 × 104; 1 mg ml−1) on neutrophil aggregation in this assay is shown in Fig. 3. Clearly, the rate of neutrophil aggregation is retarded in the presence of HA. This result also suggests that the inhibitory effect of HA on neutrophil aggregation was not merely a consequence of increases in the bulk viscosity of the sample.

Table 1.

Effect of hyaluronates of varying molecular weight on neutrophil aggregation (effect (E) * ± S.E.M.)

Effect of hyaluronates of varying molecular weight on neutrophil aggregation (effect (E) * ± S.E.M.)
Effect of hyaluronates of varying molecular weight on neutrophil aggregation (effect (E) * ± S.E.M.)
Fig. 1.

Effect of HA (mol. wt 1 × 104) on neutrophil aggregation in shaker assay. Curves show results of 2 separate experiments. Bars, ± S.E.M. Aggregation is measured as effect

(E)=log10controlexperimental×102(seematerialsandmethods)
⁠.

Fig. 1.

Effect of HA (mol. wt 1 × 104) on neutrophil aggregation in shaker assay. Curves show results of 2 separate experiments. Bars, ± S.E.M. Aggregation is measured as effect

(E)=log10controlexperimental×102(seematerialsandmethods)
⁠.

Fig. 2.

Viscosity curves (1 cP = 10−3 kg−1 s−1) for increasing concentration of HA. Molecular weights of various HA samples: •— •, 2·4×106; ○— ○, 1· 0 × 106; × — ×, 2·9 × 106; □— □, 1·0 × 104.

Fig. 2.

Viscosity curves (1 cP = 10−3 kg−1 s−1) for increasing concentration of HA. Molecular weights of various HA samples: •— •, 2·4×106; ○— ○, 1· 0 × 106; × — ×, 2·9 × 106; □— □, 1·0 × 104.

Fig. 3.

Neutrophil aggregation kinetics in Couette viscometer (Curtis & Hocking, 1970). Aggregation is measured as the reduction in particle count with time, × — ×, HA, 1·0 mg ml−1, mol. wt, 1 × 104; •— •, control.

Fig. 3.

Neutrophil aggregation kinetics in Couette viscometer (Curtis & Hocking, 1970). Aggregation is measured as the reduction in particle count with time, × — ×, HA, 1·0 mg ml−1, mol. wt, 1 × 104; •— •, control.

Effect of other polysaccharides on neutrophil aggregation

A variety of other polysaccharides was tested for effects on neutrophil aggregation using the shaker assay. Dextrans (mol. wt 500 × 103 and 60–90 ×103) at high concentration inhibited neutrophil aggregation, whereas dextran sulphate, heparin and chondroitin sulphate all caused increases in aggregation. At equivalent viscosities, the degree of inhibition of neutrophil aggregation by hyaluronates and dextrans of various molecular weights, and by chondroitin sulphate, was significantly different (Fig. 4A,B; Table 2). These results suggested strongly that factors other than bulk viscosity accounted for the inhibitory effect of HA on neutrophil aggregation. It was also unlikely that the high negative charge of the hyaluronate molecule was a factor in inhibiting aggregation since heparin and chondroitin sulphate, which have 4× and 2 × greater negative surface charge densities per sugar residue, respectively, than HA, both enhanced neutrophil aggregation at similar concentrations (Table 3). Other mechanisms for the effect of HA were therefore explored.

Table 2.

Effect of various polysaccharides at equivalent viscosity (1·3 cP) on neutrophil aggregation

Effect of various polysaccharides at equivalent viscosity (1·3 cP) on neutrophil aggregation
Effect of various polysaccharides at equivalent viscosity (1·3 cP) on neutrophil aggregation
Table 3.

Relationship between charge of polysaccharides and effect on neutrophil aggregation

Relationship between charge of polysaccharides and effect on neutrophil aggregation
Relationship between charge of polysaccharides and effect on neutrophil aggregation
Fig. 4.

A. Apparent viscosity curves for various polysaccharide solutions. □— □, HA, mol. wt, 1 × 104; •— •, DX 60–90 (dextran, mol. wt, 60–90000); × — ×, DX 500 dextran, mol. wt, 500000); ○— ○, chondroitin sulphate. B. Relationship between neutrophil aggregation and bulk viscosity of medium. Aggregation expressed as for Fig. 1; viscosity in centipoise (see legend to Fig. 2). Bars, ±S.E.M. ○— ○, HA, 1 × 104; × — ×, HA, 2·4 × 106; □— □, DX 500.

Fig. 4.

A. Apparent viscosity curves for various polysaccharide solutions. □— □, HA, mol. wt, 1 × 104; •— •, DX 60–90 (dextran, mol. wt, 60–90000); × — ×, DX 500 dextran, mol. wt, 500000); ○— ○, chondroitin sulphate. B. Relationship between neutrophil aggregation and bulk viscosity of medium. Aggregation expressed as for Fig. 1; viscosity in centipoise (see legend to Fig. 2). Bars, ±S.E.M. ○— ○, HA, 1 × 104; × — ×, HA, 2·4 × 106; □— □, DX 500.

Effect of hyaluronate on neutrophil surface membrane

It has been shown (Underhill & Toole, 1979; Angello & Hauschka, 1980) that some mammalian cells possess surface receptors for HA. Although no data are available for neutrophils, the possibility exists that HA may inhibit aggregation by blocking cellsurface receptors as suggested from cell-aggregation studies with transformed fibroblasts (Underhill & Dorfman, 1978). Neutrophils (2× 106 ml−1) were incubated with HA (mol. wt 1 × 104; 0·50 mg ml−1) for 30 min at 4 °C to permit binding of the polymer to the cell surface, washed once in HBSS, and tested for aggregation in the shaker assay. No difference was observed in aggregation for cells pretreated with HA as compared to similarly treated control cells (E = 1·83 + 2·14). This suggested that either neutrophils did not bind HA in significant amounts or that, if significant cellsurface binding of HA had occurred, it had no role in neutrophil aggregation. The role of presumptive endogenous cell-surface HA in neutrophil aggregation was studied by pretreating cells with hyaluronidase. Cells (2 × 108 ml−1) were incubated with ovine testicular hyaluronidase at various concentrations for 30 min at 37 °C, washed twice with HBSS, redispersed by agitation in a rotomixer for 10 s at medium power, and incubated at 4 °C for 30 min before use. Absence of aggregated cells within control and test samples was checked by haemocytometry. Pretreatment of cells with hyaluronidase (10–100 μg ml−1) produced a small reduction in aggregating ability (Fig. 5). This effect was not related to proteolytic activity of the hyaluronidase sample, which contained 0·3–0·6 mg ml−1 protein with trypsin-like activity when tested by the method of Schwert & Takenaka (1953) (see Materials and methods).

Fig. 5.

Effect of hyaluronidase on neutrophil aggregation. Cells were pretreated in ovine testicular hyaluronidase for 30 min at 40 °C, washed twice, redispersed and tested for aggregation. Aggregation expressed as for Fig. 1. Bars, ± S.E.M.

Fig. 5.

Effect of hyaluronidase on neutrophil aggregation. Cells were pretreated in ovine testicular hyaluronidase for 30 min at 40 °C, washed twice, redispersed and tested for aggregation. Aggregation expressed as for Fig. 1. Bars, ± S.E.M.

Equivalent concentrations of pure crystalline trypsin produced an increase in neutrophil aggregation (E = 7·0 ±1·8). The inhibitory effect of hyaluronidase, although small, suggests a role for cell-surface HA in promoting neutrophil aggregation. This effect, however, may be distinct from the inhibitory effect of exogenous HA in neutrophil aggregation (see Discussion).

Effect of hyaluronate on cell-to-substratum adhesions

Neutrophils labelled with 51Cr were suspended in solutions of HA and allowed to adhere to glass coverslips for 2 h at 37 °C. Adhesion was measured as the total radioactivity of the coverslips after non-adherent cells had been removed by interface washing in a standardized manner as described in Materials and methods. HA at concentrations >0·5 mg ml−1 reduced neutrophil-to-glass adhesion. Lower concentrations of HA did not affect adhesion to glass, although significant inhibition of neutrophil aggregation did occur with concentrations of HA below 0·5 mg ml−1 (see above). Cell adhesion to glass was also more dependent on molecular size of HA (Fig. 6) unlike neutrophil aggregation. These differences, however, probably reflect the different methods of assessing adhesion rather than any underlying mechanism in the adhesion process. Tetrasaccharides and monosaccharides of HA had a negligible effect on neutrophil adhesion to glass (data not shown).

Fig. 6.

Effect of hyaluronate on adhesion of 51Cr-labelled neutrophils to glass cover-slips.

Adhesion index = [radioactivity (c.p.m) of coverslips incubated with cells in  polysaccharideradioactivity (c.p.m) of coverslips incubated with cells in absence  of polysaccharide]

Mol. wts of various HA samples are shown: ○, 1 × 103; □, 1 × 104; ×, 1×10• 2·4 × 106. Bars, ± S.E.M.

Fig. 6.

Effect of hyaluronate on adhesion of 51Cr-labelled neutrophils to glass cover-slips.

Adhesion index = [radioactivity (c.p.m) of coverslips incubated with cells in  polysaccharideradioactivity (c.p.m) of coverslips incubated with cells in absence  of polysaccharide]

Mol. wts of various HA samples are shown: ○, 1 × 103; □, 1 × 104; ×, 1×10• 2·4 × 106. Bars, ± S.E.M.

In a further experiment, the rate of settling of monodisperse cell suspensions was observed by direct visual counting of the number of cells present at the bottom of wells in Linbro trays over a period of 180 min. The results were expressed as the number of cells per unit area (defined as the area within the squares outlined by a microscope eye-piece grid at a magnification of × 40). In the absence of HA no further increase in cell numbers per unit area was observed after 80 min, indicating that all cells had settled to the bottom of the culture dish. In the presence of HA (0·1 mg ml−1) the rate of cell settling was slightly reduced. With higher concentrations of HA, not only was the cell settling rate reduced but the total number of cells reaching the bottom of the dish was reduced, indicating that cells were retained in suspension (Fig. 7). This effect on cell settling probably accounted for the reduced cell-to-substratum adhesion of neutrophils (Fig. 6) observed with concentrations of hyaluronate >0·5 mg ml−1. However, although the viscosity of the hyaluronate was considered a likely factor in inhibiting cells settling on glass, equivisçous solutions of chondroitin sulphate failed to inhibit neutrophil adhesions to glass and indeed produced a slight enhancement (Table 4).

Table 4.

Effect of hyaluronate and chondroitin at equal viscosity on the adhesion of neutrophils to glass

Effect of hyaluronate and chondroitin at equal viscosity on the adhesion of neutrophils to glass
Effect of hyaluronate and chondroitin at equal viscosity on the adhesion of neutrophils to glass
Fig. 7.

Effect of hyaluronate on the rate of settling of monodisperse neutrophil suspensions on glass. Unit area = area within the squares outlined by a microscope eyepiece grid at a magnification of x 40. × — ×, Control; •— •, HA, 0·1 mg ml−1; ○— ○, HA, 0·3 mg ml−1; □— □, HA, 0·5 mg ml−1.

Fig. 7.

Effect of hyaluronate on the rate of settling of monodisperse neutrophil suspensions on glass. Unit area = area within the squares outlined by a microscope eyepiece grid at a magnification of x 40. × — ×, Control; •— •, HA, 0·1 mg ml−1; ○— ○, HA, 0·3 mg ml−1; □— □, HA, 0·5 mg ml−1.

Interference reflexion microscopy

Although the total number of cells adhering to glass substrata was reduced in the presence of HA, examination of adherent cells by interference reflexion microscopy showed no difference in the appearance of the contact areas in HA-incubated and control cells (Fig. 8). This suggested that within the limits of sensitivity of the technique, the adhesion sites of neutrophils to glass were unaffected by exogenous HA where contacts between the glass and the cells were permitted.

Fig. 8.

Interference reflexion microscopy of adherent polymorphonuclear leukocytes in the absence (A) and presence (B) of hyaluronate.

Fig. 8.

Interference reflexion microscopy of adherent polymorphonuclear leukocytes in the absence (A) and presence (B) of hyaluronate.

This study has shown that neutrophil adhesion, both to solid substrata and to autologous cells, is inhibited by HA in vitro. The effects were dose-dependent and in both assays a minimum molecular weight of 1 × 104 was required before inhibition of adhesion by HA was observed. Effective inhibitory concentrations were lower in cell aggregation studies (100 μgml−1) than in cell-to-substratum studies (500 μg ml−1), but this was probably due to differences in the sensitivity of the 2 assays used.

Inhibition of neutrophil adhesion by HA could not be explained satisfactorily by a receptor-mediated mechanism. Cells preincubated in hyaluronate showed no reduction in adhesiveness after a single wash and the lack of effect of hyaluronate monosaccharides and tetrasaccharides indicated that any supposed binding site on the neutrophil required at least hexamers of HA, possibly with a specific tertiary conformation. Hyaluronidase-induced inhibition of cell adhesion has been cited as evidence for the participation of a HA receptor in cell adhesiveness (Underhill & Dorfmann, 1978) and cell-surface binding of labelled HA has been shown for SV-3T3 cells (Underhill & Toole, 1979) and muscle fibroblasts (Angello & Hauschka, 1980). It is uncertain, however, whether these results indicate the presence of a true HA receptor or whether they reflect the known binding of HA to cell-surface proteins such as fibronectin (Perkins, Ji & Hynes, 1979). Indeed, the kinetics of binding for HA are considerably different from those of other bound agents, e.g. hormones (Underhill & Toole, 1979). On the other hand membrane-bound cell-surface HA may be involved in cell adhesion, through interaction between heterologous glycosaminoglycans in the substrate adhesion sites (Rollins & Culp, 1979) as can occur between glycosaminoglycan-derivatized plastic beads (Turley & Roth, 1980). Treatment of cells with crude hyaluronidase would abrogate such interactions, thus reducing cell aggregation, but specific HA cell-surface receptors need not be involved.

HA-induced inhibition of neutrophil adhesion also showed little relation to the charge of the polymer. Heparin, chondroitin sulphate and dextran sulphate, all of which have considerably greater charge densities, caused a variable increase in aggregation in contrast to the inhibitory effect shown by HA. In addition, uncharged dextran at the same concentration had no effect on adhesion.

The relationship between the viscosity of HA solutions and neutrophil adhesion was more difficult to define. Increases in bulk viscosity of the medium were associated with reduced neutrophil adhesion, within the same viscosity range as has been shown for agarose-induced inhibition of macrophage migration and adhesion (Folger et al. 197.8). Conversely, equiviscous solutions of dextran and chondroitin sulphates did not inhibit adhesion to the same degree. In addition, when neutrophil aggregation was tested under constant shear-rate conditions in a modified Couette viscometer (Curtis, 1973), HA still inhibited cell adhesion. However, alterations in the viscosity of the extracellular medium may not only lead to changes in the shear rate, but may influence other factors involved in cell adhesion. Recent studies by Maroudas (1979), on the low adhesiveness of fluid phospholipid substrata, have emphasized the importance of hydrodynamic factors and relative Brownian motion of molecules at the adhesive interface. During cell adhesion, energy is required to displace extracellular fluid from the gap between approaching cells, and this potential energy barrier is raised when adhesion takes place in fluids of increased viscosity (Curtis, 1973). No adequate system has been devised to measure this hydrodynamic force, but theoretically at least such considerations provide an explanation for the inhibitory action of viscous solutions of hyaluronate on cell aggregation; i.e. through alteration in the interfacial free-energy exchange. Folger et al. (1978) hinted at this mechanism when they stated that viscous solutions of agarose reduced macrophage migration not by increasing the viscous shear or ‘drag’ on moving cells, but by retarding the formation of adhesions between the cells and the substratum.

Cell adhesiveness in the presence of large polymers will depend greatly on the structure and physical properties of the individual polymer. For instance, viscous solutions of branched polymers such as dextran and Ficoll are much more homogeneous than unbranched polymers such as HA or methylcellulose, which form quasi-rigid porous gels (Berg & Turner, 1979). Dextran solutions undergo considerable osmotic compression at high concentrations (Ogston & Preston, 1979) unlike HA, which has a very large hydrodynamic volume (1000 times greater than its size in the dry state; Ogston & Stainer, 1951) and resists molecular compression at high concentration, tending instead to form a continuous network of relatively stiff random coils (Comper & Laurent, 1978). Such molecular configurations are poorly reflected in bulk viscosity measurements, which may partly explain the disparity in the effects of HA and other polymers at equal bulk viscosity on cell aggregation.

The very large hydrodynamic volume of hyaluronate may not only have effects on the interfacial free-energy exchange as discussed above, but may alter adhesion between cells by simple steric effects. An HA polymer of molecular weight 1·2 × 105 has dimensions of 5 nm × 200 nm while an entangled coil of mol. wt 2 × 108 averages 200 nm × 300 nm. Hydrated specific volumes vary with molecular weight from 50 to 5000 ml g−1 (Balazs, personal communication). Clearly such large molecules will exert considerable steric hindrance, perhaps even on particles as large as cells, thereby inhibiting contact between cells. The molecular domains of such polymers may be very large, particularly if the molecule is hydrated and extended, as is the case with HA, rather than globular (Flory, 1953). Edwards (1978) has suggested that, in mixed solutions of 2 different polymers, excluded-volume effects tend to produce phase separation of the polymers. If cells coated with glycoprotein can be regarded as huge glycoprotein molecules, excluded-volume effects of HA on cells should promote rather than inhibit cell aggregation, since the cells would separate out in one phase and HA in the other. Indeed, such a mechanism has been proposed for chondroitin sulphate-induced enhancement of cellular aggregation (Morris, 1979). The opposite effect would occur, however, if the HA was adsorbed to the cell surface. The polymer would then preferentially adopt a continuous phase in the intercellular space, keeping the cells apart.

Endogenous cell-surface HA may also have a role in weakening cell-substratum adhesions. Rollins & Culp (1979) have shown that a direct correlation exists between increasing concentrations of HA in the substrate-adhesion sites of mammalian fibroblasts and the ease of cell detachment. Similar results have been obtained by Kraemer & Barnhart (1978), while the low adhesiveness of certain detachment variants of Chinese hamster fibroblasts is associated with increased quantities of newly synthesized extracellular hyaluronate (Barnhart, Cox & Kraemer, 1979).

HA-induced inhibition of cell adhesiveness may have considerable significance for cell function in the tissues. Inhibition of neutrophil migration (Forrester & Wilkinson, 1981), and macrophage phagocytosis (Forrester & Balazs, 1980) and migration (Balazs & Darzynkiewicz, 1973) by HA are probably secondary to reduced cell adhesiveness, at least in part. Similar mechanisms may operate in HA-induced inhibition of targetcell killing by cytolytic immune cells in vitro (McBride & Bard, 1979). Tissues rich in HA, such as the vitreous gel and cartilage, are known to resist invasion by inflammatory cells (Forrester & Grierson, 1979). In addition, certain tumour cells may protect themselves from attack by producing a dense halo of HA-rich material around the cell (Glimelius, Norling, Westermark & Wasteson, 1979). Indeed, a correlation between the amount of extracellular HA and the malignancy of a variety of tumour cells has already been shown (Atherly, Barnhart & Kraemer, 1977; Satoh, Duff, Rapp & Davidson, 1973).

The implications for processes such as wound healing and invasiveness are clear. We suggested previously that extracellular HA may have a role in the negative feedback of the inflammatory response (Forrester & Wilkinson, 1981). In the woundhealing process, tissue concentrations of HA are low during the early phase of inflammatory cell migration, but rise pari passu with the decline of cell influx during the later stages of healing (Shetlar et al. 1978). It is possible that this decline is a consequence of the increased production of HA by cells in the wound.

The authors wish to acknowledge the excellent technical assistance of Gordon Campbell in the construction of the viscometer.

Angello
,
J. C.
&
Hauschka
,
S. D.
(
1980
).
Hyaluronate-cell interaction
.
Expl Cell Res
.
125
,
380
400
.
Atherly
,
A. G.
,
Barnhart
,
B. J.
&
Kraemer
,
P. M.
(
1977
).
Growth and biochemical characteristics of a detachment variant of CHO cells
.
J. cell. Physiol
.
90
,
375
386
.
Balazs
,
E. A.
(
1979
).
Ultrapure hyaluronic acid and the use thereof
.
U.S. Patent No
.
4
,
141
,973.
Balazs
,
E. A.
&
Darzynkiewicz
,
Z.
(
1973
).
The effect of hyaluronic acid on fibroblasts, mononuclear phagocytes, and lymphocytes
.
In The Biology of the Fibroblast
(ed.
E.
Kulonen
&
J.
Pikkarainen
), pp.
237
252
.
London, New York
:
Academic Press
.
Barnhart
,
B. J.
,
Cox
,
S. H.
&
Kraemer
,
P. M.
(
1979
).
Detachment variants of Chinese hamster cells
.
Expl Cell Res
.
119
,
327
332
.
Berg
,
H. A.
&
Turner
,
C.
(
1979
).
Movement of microorganisms in viscous environments
.
Nature, Land
.
278
,
349
351
.
Cleland
,
R. L.
,
Cleland
,
M. C.
,
Lipsky
,
J. J.
&
Lyn
,
V. F.
(
1968
).
Ionic polysaccharides. I. Absorption and fractionation of polyelectrolytes on (diethylamino)ethyl cellulose
.
J. Am. chem. Soc
.
90
,
5140
5153
.
Cleland
,
R. L.
&
Wang
,
J. P.
(
1970
).
Ionic polysaccharides. III. Dilute solution properties of hyaluronic acid fractions
.
Biopolymers
9
,
799
810
.
Comper
,
W. D.
&
Laurent
,
T. C.
(
1978
).
Physiological function of connective tissue polysaccharides
.
Physiol. Rev
.
58
,
255
315
.
Curtis
,
A. S. G.
(
1964
).
The adhesion of cells to glass; a study by interference reflexion microscopy
.
J. Cell Biol
.
19
,
199
215
.
Curtis
,
A. S. G.
(
1973
).
Cell adhesion
.
Prog. Biophys. molec. Biol
.
27
,
315
386
.
Curtis
,
A. S. G.
&
Hocking
,
L.
(
1970
).
Collision efficiency of equal spherical particles in a shear flow
.
Trans. Faraday Soc
.
66
,
1381
1390
.
Edwards
,
P. A. W.
(
1978
).
Differential cell adhesion may result from nonspecific interactions between cell surface glycoproteins
.
Nature, Lond
.
27
,
248
249
.
Flory
,
P. J.
(
1953
).
Principles of Polymer Chemistry
, pp.
329
341
.
Ithaca, New York
:
Cornell University Press
.
Folger
,
R. L.
,
Weiss
,
L.
,
Glaves
,
D.
,
Subjeck
,
J. R.
&
Harlos
,
J. P.
(
1978
).
Translational movement of macrophages through media of different viscosities
.
J. Cell Sci
.
31
,
245
257
.
Forrester
,
J. V.
&
Balazs
,
E. A.
(
1980
).
Inhibition of phagocytosis by high molecular weight hyaluronate
.
Immunology
40
,
435
446
.
Forrester
,
J. V.
&
Grierson
,
I.
(
1979
).
The cellular response to blood in the vitreous
.
J. Path
.
129
,
43
52
.
Forrester
,
J. V.
&
Wilkinson
,
P. C.
(
1981
).
Inhibition of leukocyte locomotion by hyaluronic acid
.
J. Cell Sci
.
48
,
315
331
.
Glimelius
,
B.
,
Norling
,
B.
,
Westermark
,
B.
&
Wasteson
,
A.
(
1979
).
A comparative study of glycosaminoglycans in cultures of human normal and malignant glial cells
.
J. cell. Physiol
.
98
,
527
538
.
Knox
,
P.
&
Wells
,
P.
(
1979
).
Cell adhesion and proteoglycans. I. The effect of exogeneous proteoglycans on the attachment of chick embryo fibroblasts to tissue culture plastic and collagen
.
J. Cell Sci
.
40
,
77
89
.
Kraemer
,
P. M.
&
Barnhart
,
B. J.
(
1978
).
Elevated cell-surface hyaluronate in substrate attached cells
.
Expl Cell Res
.
114
,
153
157
.
Lackie
,
J. M.
(
1977
).
The aggregation of rabbit polymorphonuclear leukocytes (PMN)
.
Inflammation
2
,
1
15
.
Laurent
,
T. C.
&
Gergely
,
J.
(
1955
).
Light scattering studies on hyaluronic acid
.
J. biol. Chem
.
212
,
325
333
.
Mcbride
,
W. H.
&
Bard
,
J. L.
(
1979
).
Hyaluronidase-sensitive haloes around adherent cells. Their role in blocking lymphocyte mediated cytolysis
.
J. exp. Med
.
149
,
507
516
.
Mclean-Fletcher
,
S. D.
&
Pollard
,
T. D.
(
1980
).
Viscometric analysis of the gelation of Acathamoeba extracts and purification of two gelation factors
.
J. cell Biol
.
85
,
414
428
.
Maroudas
,
N. G.
(
1979
).
On the low adhesiveness of fluid phospholipid substrata
.
J. theor. Biol
.
79
,
101
116
.
Morris
,
J. E.
(
1979
).
Steric exclusion of cells
.
Expl Cell Res
.
120
,
141
153
.
Ogston
,
A. G.
&
Preston
,
B. N.
(
1979
).
The molecular compression of dextran
.
Biochem. J
.
183
,
1
9
Ogston
,
A. G.
&
Stainer
,
J. E.
(
1951
).
The dimensions of the particle of hyaluronic acid complex in synovial fluid
.
Biochem. J
.
49
,
585
599
.
Perkins
,
M. E.
,
Ji
,
T. H.
&
Hynes
,
R. O.
(
1979
).
Cross-linking of fibronectin to sulphated proteoglycans at the cell surface
.
Cell
16
,
941
952
.
Pessac
,
B.
&
Defendi
,
V.
(
1972
).
Cell aggregation: role of acid mucopolysaccharides
.
Science, N.Y
.
175
,
898
900
.
Pratt
,
R. M.
,
Larsen
,
M. A.
&
Johnston
,
M. C.
(
1975
).
Migration of neural crest cells in a cell free hyaluronate-rich matrix
.
Devi Biol
.
44
,
298
305
.
Rollins
,
B. J.
&
Culp
,
L. A.
(
1979
).
Preliminary characterisation of the proteoglycans in the substrate adhesion sites of normal and virus-transformed murine cells
.
Biochemistry
18
,
5621
5629
.
Satoh
,
H. A.
,
Duff
,
R.
,
Rapp
,
F.
&
Davidson
,
E. A.
(
1973
).
Production of mucopolysaccharides by normal and transformed cells
.
Proc. natn. Acad. Sci. U.S.A
.
70
,
54
56
.
Schwert
,
G. W.
&
Takenaka
,
Y.
(
1953
).
A spectrophotometric determination of trypsin and chymotrypsin
.
Biochim. biophys. Acta
16
,
570
575
.
Scott
,
J. E.
&
Tigwell
,
M. J.
(
1978
).
Periodate oxidation and the shape of glycoaminoglycans in solution
.
Biochem. J
.
173
,
103
114
.
Shetlar
,
M. R.
,
Davitt
,
W. F.
,
Shetlar
,
L. F.
,
Posett
,
R. L.
,
Cross
,
M. F.
&
Lautsch
,
E. V.
(
1978
).
Glycosaminoglycan changes in healing myocardial infarction
.
Proc. Soc. exp. Biol. Med
.
158
,
210
214
.
Smith
,
R. P. C.
,
Lackie
,
J. M.
&
Wilkinson
,
P. C.
(
1979
).
The effects of chemotactic factors on the adhesiveness of rabbit neutrophil granulocytes
.
Expl Cell Res
.
122
,
169
177
.
Thorne
,
K. J. L
,
Oliver
,
R. C.
&
Lackie
,
J. M.
(
1977
).
Changes in the surface properties of rabbit polymorphonuclear leucocytes induced by bacteria and bacterial endotoxin
.
J. Cell Sci
.
27
,
213
225
.
Turley
,
E. A.
&
Roth
,
S.
(
1980
).
Interactions between the carbohydrate chains of hyaluronate and chondroitin sulphate
.
Nature, Land
.
283
,
268
271
.
Underhill
,
C.
&
Dorfmann
,
A.
(
1978
).
The role of hyaluronic acid in intercellular adhesion of cultured mouse cells
.
Expl Cell Res
.
117
,
155
164
.
Underhill
,
C. B.
&
Toole
,
B. P.
(
1979
).
Binding of hyaluronate to the surface of cultured cells
.
J. Cell Biol
.
82
,
475
.
Van Wazer
,
J. R.
,
Lyons
,
J. W.
,
Kim
,
K. Y.
&
Cowell
,
R. E.
(
1963
).
Viscosity and Flow Measurement
.
London
:
Interscience
.
Winter
,
W. T.
&
Arnott
,
S.
(
1977
).
Hyaluronic acid: the role of divalent cations in conformation and packing
.
J. molec. Biol
.
117
,
761
784
.