Characterization of the Murine Macrophage Receptor for Group B Streptococci

Zbl. Bakt. 278, 541-552 (1993) © Gustav Fischer Verlag, Stuttgart· Jena . New York

Characterization of the Murine Macrophage Receptor for Group B Streptococci ANNE R. SLOAN l and THOMAS G. PISTOLE * Department of Microbiology, University of New Hampshire, Durham, NH 03824, USA 1 Present address: Department of Biochemistry and Molecular Biology, University of New Hampshire, Durham, NH 03824, USA

With 2 Figures· Received June 24, 1992 . Revision received October 23, 1992 . Accepted November 28, 1992

Summary The macrophage has been shown to bind potentially pathogenic bacteria in the absence of serum components or opsonins but the mechanism is poorly understood. The rich array of sugars on the surface of group B streptococci plus the presence of membrane-associated lectin receptors on the macrophage suggests that this is a likely means for bacterial recognition by these host defense cells. Inhibition studies with free sugars and neoglycoconjugates of bovine serum albumin, however, failed to confirm this hypothesis. Furthermore, neuraminidase-treatment to expose galactose residues and the use of isogenic bacterial strains having no capsule or no capsular sialic acid yielded no confirmation of lectinmediated recognition. The trypsin-sensitive receptor exhibited temperature dependence and a requirement for divalent cations distinct from that reported for the lectin-like galactose receptor. The activity of this streptococcal binding receptor was inhibited by 2-deoxy-Dglucose but not by neutrophil elastase. Pre-exposure of macrophages to bound fibronectin and treatment with phorbol ester each enhanced bacterial binding. These data fail to support a role for the galactose lectin and provide preliminary evidence for involvement of the leukocyte integrins in macrophage recognition of group B streptococci. Zusammenfassung Der Makrophage bindet bekanntermagen potentiell pathogene Bakterien auch wenn Serumbestandteile oder Opsonine fehlen, wobei der zugrunde liegende Mechanismus kaum bekannt ist. Die Anordnung der Zucker auf der Oberfliiche von Gruppe B Streptokokken und die membrangebundenen Lectinrezeptoren auf dem Makrophagen liegen vermuten, dag es sich urn einen moglichen Mechanismus der Erkennung von Bakterien durch diese Abwehrzellen des Wirtes handeln konnte. Hemmstudien mit freien Zuckern und Neoglycokonjugaten aus Rinderserumalbumin konnten diese Hypothese jedoch nicht bestiitigen.

* Corresponding author

542

A. R. Sloan and T. G. Pistole

Auch Behandlung mit Neuraminidase zur Freilegung von Galactoseresten sowie der Einsatz isogener Bakterienstiimme ohne Kapsel oder Kapsel-Sialinsiiure lieferten keine Bestiitigung fur eine Lektin-vermittelte Erkennung. Der Trypsin-sensitive Rezeptor erwies sich als abhiingig von Temperatur und zweiwertigen Ionen, anders als es fur den Lectin-iihnlichen Rezeptor berichtet wurde. Dieser Streptokokken-bindende Rezeptor wurde durch 2-Desoxy-D-glucose gehemmt, aber nicht durch neutrophile Elastase. Vorinkubation der Makrophagen mit gebundenem Fibronectin und Phorbolester-Behandlung verstiirkten die Bindung der Bakterien. Nach dies en Daten spielt Galactose-Lectin keine Rolle, dagegen ergibt sich ein Hinweis auf die Beteiligung von Leukozyten-Integrinen bei der Erkennung von Gruppe B Streptokokken durch Makrophagen.

Introduction Phagocytic cells in higher animals are capable of discriminating between foreign material and self components. They are also able to distinguish between self and alterations to self, as in effete or damaged cellular components or aberrant self constituents, such as tumor cells (17). The recognition abilities that underlie this dis~ criminating capacity must have evolved with the emergence of primitive species. Some form of self-recognition mechanism is likely to have formed the basis of the abilities of colonial marine species such as Porifera to exclude unrelated organisms from inclusion in their colonies (31). In addition the cells lining the cavity of Porifera are capable of capturing microorganisms present in the water drawn into the cavity. Phagocytosis also plays a role in the metamorphosis of insects in removing dead cells and disintegrated tissue (10). The role of invertebrate amebocytes in recognizing and engulfing foreign material is likely to be a development of these earlier, more primitive recognition mechanisms and begins to involve cooperation with soluble factors present in the hemolymph, acting as opsonins (21). In vertebrates further specialization has occurred, involving cooperation between phagocytes and components of the immune system including antibodies, complement and lymphoid cells. The more primitive forms of phagocyte recognition appear, however, to have been conserved and phagocytic cells retain their ability to recognize and bind foreign particulate material without involvement of exogenous opsonins (27). The mechanisms that underlie these highly important recognition abilities of phagocytes are poorly understood. Group B streptococci (GBS), for example, can bind directly to mammalian macrophages (M0) in the absence of opsonins (9). These organisms are a major cause of neonatal meningitis and septicemia (23) and in post-partum infections. Increasingly they are being recognized as an importnt pathogen in non-pregnant adult humans as well (24). Phagocytosis of these organisms followed by intracellular killing is considered to be the major host defense in mammals, and serum opsonins, e.g., anti-capsule antibody and complement, have been shown to playa major role in effecting endocytosis (3). Nearly two-thirds of all neonatal infections with this organism are caused by serotype III (19). Nearly all of the remaining clinical isolates belong to capsular types la, Ib, or II, although at least two additional types (IV and V) have recently been described (32). Each of these capsular types has a backbone composed of a repeating unit of 2 to 5 monosaccharides and at least one side chain in which ~-D-galactose is found in either a terminal position or penultimate to sialic acid (32). As noted, GBS can bind directly to M0 in the absence of serum opsonins (9). Non-

Macrophage Receptor for Croup B Streptococci

543

opsonin-mediated phagocytosis of other microorganisms can be achieved by membrane-associated lectins on the macrophage reacting with complementary sugar residues on the microbial surface (18, 25). Our original hypothesis, then, was that macrophage-derived lectins mediate the recognition and binding of GBS. Our data do not support this idea, but instead suggest involvement of the leukocyte integrins in this process. Materials and Methods Chemicals. All reagents were obtained from Sigma Chemical Co. (St.Louis, MO, USA) unless otherwise noted. Bacteria. A prototypic strain of CBS, originally obtained from the late Dr. Rebecca Lancefield, Rockefeller University, was provided by Dr. Dennis Kasper, Harvard Medical School (Boston, MA, USA). This strain is designated 18RS21 (type II) (15). Two type III CBS, strains M732 and COH31, originally clinical isolates from infants with meningitis (15), were also provided by D. Kasper. All strains were obtained as frozen suspensions of pure culures, which were stored at -70°C. Each strain was streaked onto a 5% blood agar plate and incubated overnight at 37°C. A colony from each strain was used to inoculate 10 ml of Todd-Hewitt broth (Difco Laboratories, Detroit MI, USA) and this was incubated overnight to stationary phase. A 5-ml volume of each strain was used to inoculate one liter of Todd-Hewitt broth and this was incubated at 37°C to log phase under static conditions. The bacteria were harvested by centrifugation at 10000 X g for 10 min and were washed three times in Dulbecco's phos~hate-buffered saline (DPBS). Stock suspensions of CBS were adjusted to a density of 10 1 /ml, as determined by direct counts in a Petroff-Hausser chamber (C. Hausser and Son, Philadelphia, PA, USA) and were stored at -70°C in DPBS containing 8% dimethylsulfoxide (DMSO). Before use in each assay frozen aliquots from the original batch culture of each strain were thawed and washed with DPBS to insure continuity between experimental studies performed on different days. Neuraminidase treatment of CBS. Desialylation of CBS strains was done by adding 0.15 units of neuraminidase from Vibrio cholerae, (Sigma, Type II, specific activity: 10 units/mg protein using NAN-lactose) to 3 ml of bacterial suspension at 101o/ml in 0.05 M sodium acetate buffer, pH 5.5, containing 154 mM NaCI and 9 mM CaCh (20). After one hour at 37°C with constant end-over-end rotation, the bacteria were washed three times with DPBS and stored in 8% DMSO as described above. Nondesialylated bacteria were treated likewise but:without the addition of neuraminidase. isogenic strains of type III CBS COH31. Capsular mutants of CBS type III strain COH31 obtained from D. Kasper (Harvard) were derived by Tn916 transposon mutagenesis. The mutant strain designated COH31-15 does not react with type III CBS antiserum from rabbits and has no visible extracellular capsule material when visualized by electron microscopy (16). Another mutant strain designated CO H31-21 lacks terminal sialic acid on the side chain, leaving ~-D-galactose as the terminal sugar. Interaction of CBS with Ricinus communis agglutinin. The availability of galactose residues on the surface of CBS was verified by the aggregation of a 1 % suspension of each strain by the galactose-specific lectin from Ricinus communis (castor bean) at a concentration of 0.2 !tg/m!. D( + )-Calactose, D( + )-glucose and methyl ~-D-galactopyranoside were tested at concentrations of 10 mg/ml, 1.0 mg/ml and 0.1 mg/ml for their ability to inhibit the aggregated CBS. Detection of f3-glucan receptor on Me. ~-Clucan from S. cerevisiae inhibits binding through the ~-glucan receptor (7). We have exploited this mechanism for Me binding to a surface structure of yeast as a positive control in our inhibition studies. A suspension of zymosan A from Saccharomyces cerevisiae, measuring 3-4!tm in diameter, was prepared by sonication using a pulsed cycle for 30 seconds, at output setting 4, 50% duty cycle (Heat Systems Ultrasonics, Inc. Plainview, NY, USA) in DPBS. Soluble u-mannan and ~-glucan

544

A. R. Sloan and T. G. Pistole

from S. cerevisiae and ~-glucan from barley were tested for their ability to inhibit binding of zymosan to Mo. The lowest concentration of each capable of yieldin~ 50% inhibition of zymosan attachment to Mo at particle concentrations of 10 9 /ml and 10 Iml was determined by the visual assay. Macrophages. Peritoneal exudate Mo from 8 to 12 week-old female BALB/c mice were elicited by intraperitoneal injections of 2 ml of Brewer thioglycollate (Difco Laboratories). After 3 days the Mo were harvested by lavage using 10 ml of RPMI medium (Gibco Laboratories, Grand Island, NY, USA). The Mo were washed twice by centrifugation at 100 x g, using cold RPMI and resuspended in RPMI at a concentration of 10 6 cells per m!. Antisera to GBS. Group-specific and type-specific antisera to GBS, prepared in rabbits, was kindly provided by D. Kasper. Binding assays. Adherence of GBS to Mo was determined either by direct microscopic observation, based on a modification of (11) or by an ELISA-based system, developed in our laboratory (26). Inhibition studies. Mo mono layers were pre-incubated with monosaccharides in DPBS for 30 min and then for an additional 1 h in the presence of bacteria. The following monosaccharides were used at concentrations of 2, 20 and 200 mM: D( +)-galactose, methyl a-D-galactoside, methyl ~-D-galactoside, D( +)-glucose, methyl a-D-glucoside, methyl ~-D-glucoside, D( + )-mannose, methyl a-D-mannoside, methyl a-L-fucose, Nacetyl-D-glucosamine (GluNAc) and N-acetyl-D-galactosamine (GaINAc). Neoglycoconjugates prepared by E·Y Laboratories, San Mateo, CA, USA, from bovine serum albumin (BSA), containing 30 to 40 mole saccharide per mole BSA, were tested for their ability to inhibit binding of GBS to Mo. Glucosylated-, galactosylated- and mannosylated-BSA were incubated with the Mo mono layers as described for the monosaccharide solutions at concentrations of 20, 200 and 2000 [!g/ml. The concentrations of the monosaccharide were equivalent to approximately 11, 110 and 1100 [!M, respectively. Un conjugated BSA, Fraction V, was included as a negative control at the same protein concentrations. Treatment of Mo monolayers with trypsin. The procedure for trypsin treatment of Mo mono layers was a modification of that devised by Czop et al. (6). Mo were layered with 0.25 ml analytical grade, bovine pancreatic trypsin (Boehringer Mannheim, Indianapolis, IN, USA) having a specific activity, determined by the manufacturer, of 40 U/mg lyophilized material (25°C; benzoyl-L-arginine ethyl ester as substrate). The trypsin was diluted in RPMI containing an additional 5 mM MgCI 2 , over the range of 4.0 [!g/ml to 4.0 mg/ml (0.172 [!M to 172 [!M), incubated at 37"C for 30 min in the CO 2 chamber, and rinsed twice with RPM!. Each monolayer was then layered with 0.25 ml aprotinin (Trasylol TM, 200 UI mg lyophilizate trypsin inhibitor units, Boehringer Mannheim) at varying molar concentrations selected to be equivalent to the previous trypsin concentration, incubated for 15 min at 37"C, rinsed four times with RPMI, and overlaid with wild-type COH31 and the two isogenic strains of GBS for detection of adherence. Treatment of Mo mono layers with aprotinin alone, at any of the concentrations tested, followed by washing, did not inhibit adherence. Treatment of Mo monolayers with neutrophil elastase. Lyophilized human neutrophil elastase (Sigma) with a specific activity of 110 U/mg protein was dissolved in sterile distilled water at 1 mg/ml and stored in small aliquots at -70°C. Mo monolayers were treated with 0.25 ml of neutrophil elastase at final concentrations of 30, 10 and 3 [!g/ml in DPBS for 1 h at 37°C, using a modification of the method of Tosi et a!. (30). The elastase was removed by washing the chambers individually four times with DPBS followed by the addition of GBS (strain COH31) for assay of adherence. Treatment of macrophage monolayers with 2-deoxy-D-glucose. We used a modification of the method of Sung and Silverstein (28) to examine the effect of 2-deoxy-D-glucose treatment of Mo on their ability to bind GBS (strain COH31). Briefly, Mo mono layers were incubated with 500, 50, 5.0 and 0.5 mM 2-deoxy-D-glucose for 2 h at 37"C, followed by the addition of GBS to the monolayers for 1 h at 37"C, without removal of 2-deoxy-Dglucose, for assay of adherence.

Macrophage Receptor for Croup B Streptococci

545

Role of temperature and divalent cations on the adhernce of CBS to M0. M0 monolayers were incubated with CBS strain COH31 at 37°C, as described in the adherence assay, and at the additional temperatures of 20°C and 4°C to determine the effect of temperature on CBS adherence to M0. The effect of 0.5 mM MgCl z and CaCtz on this adherence was tested using a modification of the method of Wright and fong (35). M0 were washed with Ca2+ - and Mgz+ -free DPBS, then incubated for 3 h in the chamber-slide to prepare mono layers deficit in divalent cations. CBS strain COH31 was added to the M0 as described for the adherence assay in DPBS containing both Ca2+ and Mgz+, DPBS containing either Ca z+ or Mgz+, or DPBS without divalent cations. M0 were exposed to these same conditions without CBS to observe morphological changes in M0 adherence to the chamber-slide during the incubation period with bacteria. Treatment of M0 with phorbol ester and ligation of the M0 fibronectin (FN) receptor on the adherence of CBS by M0 monolayers. Preparation of M0 monolayers was performed as

described previously using eight-chamber tissue culture slides (Lab-Tek) having a plastic surface which enhances protein coating. Prior to addition of the M0 the plastic slide surfaces were coated with mouse serum albumin (1 mg/ml) or FN from rat plasma (0.1 mg/ml) by a 60 min incubation at 20°C (34). The surfaces were washed three times with DPBS, then incubated for 3 h at 3rc with the cells (0.25 mllchamber). Where indicated, phorbol dibutyrate (PDB) at 500 ng/ml was included during the plating of the cells. CBS strain COH31 was added to the M0 in the presence or absence of PDB (500 ng/ml), as previously described for the adherence assay. M0 monolayers were exposed to these same sets of conditions without CBS to observe morphological changes in macrophage adherence to the chamber-slide.

Results

Effect of sugars as inhibitors of CBS adherence to macrophages To examine our hypothesis that lectins on the M0 are involved in this binding, we studied the effect of added monosaccharides and BSA-glycoconjugates on bacterial adherence to these cells. Using the visual assay we found no significant inhibition of binding of either type II or type III CBS by any of the other sugars tested at concentrations ::::: 200 mM (Table 1) or the neoglycoconjugates at concentrations::::: 2 mg/ml (Table 2). Specific inhibition of binding of S. cerevisiae zymosan particles to M0 by yeast ~-glucan but not barley ~-glucan or yeast u-mannan (data not shown) was used to verify that the visual assay can quantify receptor inhibition.

Effect of desialylation of capsular CBS polysaccharide CBS types II and III were treated with neuraminidase from V. cholerae to remove sialic acid. We used the lectin from Ricinus communis (castor bean) to confirm that the galactose residue was exposed following neuraminidase treatment (data not shown). For each strain tested, the dose-dependent binding of neuraminidase-treated and untreated CBS to M0 was similar (Fig. 1 A-C).

Comparison of type III isogenic strains for adherence to Me A further study comparing macrophage adherence of wild-type COH31 and of isogenic strains of CBS type III, derived by Tn916 transposon mutagenesis and having no capsular sialic acid (COH31-21) or no capsule expressed (COH31-15), was performed by the visual assay. Table 3 depicts data from three experiments in which the

546

A. R. Sloan and T. G. Pistole

Table 1. Effect of monosaccharides as inhibitors of GBS adherence to Mo. Inhibitors were used at concentrations of 200 mM. Results are reported as percentage binding ± s.d. (N = 8). Analysis of variance for all systems (excluding PBS controls) indicated no s~atistically significant difference in the effects of the test sugars on bacterial adherence (F = 2.11; P = 0.079) . Inhibitor

PBS D( + )glucose Methyl-a-glucose Methyl-~-glucose

D( + )galactose Methyl-a-galactose Methy l-~-galactose L(-)fucose D( +)mannose Methyl-a-mannose N-acetylglucosamine N-acetylgalactosamine

Strain of GBS 18RS21

M732

COH31

100± 11 62±9 69± 13 59±7 69± 12 64±8 64± 19 64±6 76± 14 71±12 69± 8 66± 8

100± 8 55 ± 13 45 ± 12 58 ± 10 58 ±8 55±14 55 ± 11 58 ± 11 63±9 84± 14 77±7 68 ±6

100 ± 12 64±12 60±6 76±8 68 ± 15 76± 12 68±6 76±9 80±0 90± 12 96± 10 96±10

Table 2. Effect of BSA-glycoconjugates as inhibitors of GBS adherence to Mo. Inhibitors were used at concentrations of 2 mg/m!. Results were reported as percentage binding ± s.d. (N = 8). Analysis of variance for all systems (excluding PBS controls) indicated no statistically significant difference in the effects of the test glycoconjugates on bacterial adherence (F = 1.18; P = 0.38) Strain of GBS

Inhibitor

PBS Glucose-BSA Galactose-BSA Mannose-BSA Glucose/galactose-BSA Glucose/mannose- BSA Galactose-mannose-BSA

18RS21

M732

COH31

100±8 112± 18 115 ±20 112±9 112± 12 106 ± 11 115 ±9

100 ± 10 100±8 100± 10 94± 12 115 ± 13 106± 15 94±7

100± 15 96±9 115±6 104 ± 12 133 ± 15 136±21 133 ± 16

binding of all isogenic strains was similar, about 35% and 12% for bacterial concentrations of 109/ml and lOs/ml, respectively. Galactose (200 mM) had no inhibitory effect on the binding of any of the GBS isogenic strains to Mo.

Trypsin sensitivity of the macrophage capacity to bind CBS The effect of pretreating Mo with trypsin on the capacity of these cells to bind GBS was studied to determine whether the binding was sensitive to this treatment and, hence, likely due to a protein on the Mo membrane. Adherence of each isogenic strain

547

Macrophage Receptor for Group B Streptococci A

~~

B

0.8

~

0.6

C

ec::

::!! 0.6

18RS2118RS21 +

~ 0.4

~

.e

105 10 6 Bacteria/well

0.2

DlB

60

107

0.5 0.4

.
«

".

:630

M732M732+

0.2

:; 20

10

COH31COH31+

10 7

treatment control

~ 10

0.1 0.0 4

-

c::

1.::40

::!! 0.3

~

10 6 Bacteria/ well

~50

~

.e51

0.4

!

02 0.0 4 10

-

0.8

~

::!!

j

1.0

IS>

10 5 10 6 Bacteria/well

:::E

107

0

~

.1 ~,

F treatment control

10 100 (Ilmole/lJ

1000

lB 60 (.;ISO

~

-g"

40

treatment control

'<:30

-=

20

.~ 10

110 100 (j.lmole/l)

~

1000

0 ..........................................................

~

~,

.1

1 ~,

10 100 (j.lmole /lJ

1000

Fig. 1. A-C. Effect of neuraminidase treatment on adherence of three GBS strains to M0, as measured by ELISA. In each case (+) indicates treated system and (-), the untreated control. D-F. Effect of trypsin treatment of M0 on GBS binding, as determined by direct observation. (D) strain COH31, (E) strain C0H31/15, (F) strain COH31/21.

Table 3. Effect of D( + )-galactose on the adherence of GBS type III isogenic strains to M0. Where indicated (+), 200 mM of D( + )-galactose was added. Results are expressed as percentage binding ± s.d. (N = 6) GBS strain

D(+)-gal added

Bacteria/ml

109

10 8

+

30±2 33±4

10±2 12±1

+

35±5 30±4

9±3 12±1

+

37±3 39±6

13 ± 1 14±3

COH 31 COH 31-15 COH 31-21

A. R. Sloan and T. G. Pistole

548

was relatively little affected by 4.0 ftg/ml of this enzyme, whereas GBS binding fell to less than 40% after treatment with 40 ftg/ml for each strain (Fig. 1 D-F). Increasing the dose of trypsin further diminished the capacity of Mo to bind bacteria; and at trypsin concentrations of 4 mg/ml, fewer than 10% of Mo were still able to bind GBS. Effect of neutrophil elastase on macrophage binding to CBS

Neutrophil elastase has been shown to cleave the C3b receptor, CR1, but not the C3bi receptor, CR3 (30). To determine its effect on GBS adherence, Mo monolayers were treated with increasing concentrations of neutrophil elastase over the range of 3.0 to 30 ftg/ml and washed exhaustively with DPBS to remove the enzyme prior to the addition of wild-type COH31 GBS. Treatment of Mo at each concentration of elastase had no effect on the percentage of Mo binding of five or more bacteria, as determined by counting 1000 Mo, when compared to the bacterial binding by untreated Mo (data not shown). The percentage of Mo binding five or more bacteria remained at about 55%. Effect of 2-deoxy-D-glucose treatment of Mo on the binding of CBS

This treatment has been shown to inhibit both Fc and complement-receptor mediated binding by mouse peritoneal Mo (28). In this study Mo mono layers were treated with 2-deoxy-D-glucose over the range of 0.5 to 500 mM for 2 h prior to the addition of bacteria as well as during the period of bacterial binding. At the lowest concentration of 2-deoxy-D-glucose, 0.5 mM, there was a decrease in binding to 38% Mo, followed by an increasing reduction in binding of GBS by Mo at the higher concentrations of 2-deoxy-D-glucose (Fig. 2 A). Temperature and divalent cation dependence of CBS binding

Lectin-mediated bacterial binding (31; I. Ofek, personal communication) and binding due to the FcR (35) are unaffected by temperature over the range of 4-37°C, whereas adherence mediated by leukocyte integrins is temperature-dependent (35). A

A

B

60

]40

.

20 10

o

0.5 5 50 500 2-iieoxy-D-glucose (~g/rnl)

80

60

~

40

~

0

~

o

100

.<:

"Cl

-"30

~~

~

i

.~ 50

20

None P only P + B B only Treatment

Fig. 2. A. Effect of pretreatment of Mo with the phagocytosis inhibitor, 2-deoxy-D-glucose, on adherence of GBS to these cells, as determined by direct observation. B. Effect of phorbol ester on binding of GBS to murine Mo. Key: none = no phorbol ester present; P only = phorbol ester present only during the plating of Mo; unbound ester removed prior to addition of bacteria; P & B = phorbol ester present during both plating of Mo and incubation of Mo with bacteria; B only = phorbol ester present only during incubation of bacteria with Mo.

Macrophage Receptor for Group B Streptococci

549

study of the effect of temperature on the binding of GBS (strain COH31) by M0 monolayers revealed dramatic differences. Nearly 60% of the M0 bound five or more bacteria at 37 °C, whereas binding was lowered three-fold to 22 % at 20 °C and was almost absent (3%) at 4°C, Lectin- and FeR-mediated binding requires both Ca2+ and Mg2+ (31, 35), whereas adherence due to leukocyte integrins occurs in the presence of either cation (35). There was an absolute requirement for either Ca2+ or Mg2+ to support adherence of GBS to the M0 monolayers. Divalent cation-free assay systems exhibited very low level binding (3 ± 1 %), while those in which either cation was present showed a significantly higher binding activity (Mg: 35 ± 1 %; Ca: 40 ± 2%) The presence of both cations, at 0.25 mM each, increased the percentage of M0 binding bacteria to 57 ± 3%.

Effect of phorbol ester andlor ligation of fibronectin (FN) receptors on CBS binding In this study we investigated the effect of phorbol dibutyrate (PDB) on the ability of M0 to bind GBS. In addition, we studied the effect of ligation of the M0 FN receptor, by itself or on combination with the presence of phorbol ester, on the binding of bacteria. Both conditions have been shown to enhance leukocyte integrin-mediated binding, but not that due to lectin or the FeR (33). Fig. 2B depicts the effects of phorbol ester on the binding of GBS (strain COH31) to M0. The addition of the phorbol ester to the M0 monolayer increased the bacterial binding values from about 55% in the untreated systems to about 95% at maximum exposure to this chemical. M0 monolayers prepared on a fibronectin surface exhibited maximal binding (98-100%) whether or not the cells were exposed to the phorbol ester (data not shown).

Discussion Evidence from in vitro studies indicate that phagocytosis of bacteria occurs in opsonin-free media (1, 14, 22, 31), while findings from in vivo studies show clearance of bacteria from the serum opsonin-deficient lungs by alveolar macrophages (12) or from the blood by the reticuloendothelial system of animals depleted of complement (5). The molecular basis for non-opsonic recognition between bacteria and phagocytes is largely unknown. During the last decade, considerable evidence has accumulated showing that specific recognition by phagocytes may be accomplished by the interaction of carbohydratebinding proteins, e.g., lectins, in the cytoplasmic membrane of this defense cell with complementary sugars on the target cell (27). This type of recognition, which also leads to phagocytosis, has been termed lectinophagocytosis (18). M0 have several cell-associated lectins that have been implicated in microbial recognition (13). A mannosespecific receptor on human M0 has been shown to recognize mannose-bearing pathogens and to mediate their internalization (8). The mannosyl/fucosyl receptor recognizes Leishmania promastigotes and synthesis of reactive oxygen metabolites is triggered by adherence to this receptor (4). Human monocytes bear ~-glucan receptors that react with zymosan particles (7), thus providing the cells with the ability to detect microorganisms bearing this glycan in the absence of serum factors. The clearance of Klebsiella pneumoniae from the lung may be due to a mannosyllN-acetyl-D-glucosaminyl receptor found on alveolar macrophages (2).

550

A. R. Sloan and T. G. Pistole

In this study we attempted to show involvement of the galactose receptor for the non-opsonin-mediated adherence of GBS to peritoneal M0. Our results, however, do not support this hypothesis. First, none of the monosaccharides or BSA-glycoconjugates acted as a specific inhibitor of type II and III GBS adherence to M0. Second, neuraminidase-treated GBS, with exposed galactose on their surface, bound to peritoneal M0, but the binding could not be reduced by pre-exposure of the M0 to galactose. Third, isogenic strains of type III GBS having no capsule or no sialic acid expressed, leaving galactose residues exposed, exhibited the same degree of attachment as wild-type type III to M0. Blood clearance of neuraminidase-treated type I GBS, which express galactosyl residues on their surfaces, has been shown to be strongly inhibited by galactosyl- but not by mannosyl- or fucosyl-BSA (20). In the same study the blood clearance of type II GBS could be inhibited by neither galactosyl- or mannosyl-BSA alone, whereas the presence of both neoglycoconjugates in the injected suspension significantly inhibited the blood clearance of streptococci. No such synergism was obtained in our experiments with type II and III GBS adherence to murine peritoneal M0. These results suggest that lectins expressed on the surface of phagocytic cells in the liver are not present on peritoneal M0. However, in the same study (20) the adherence of type II GBS to M0 was inhibited by galactosy-, mannosyl- and glucosyl-BSA. As with blood clearance, the combination of galactosyl- and glucosyl- or mannosyl-BSA significantly increased the inhibition of attachment. Differences in our results and those of Perry et al. (20) may be due to the stringency of conditions in our studies for detecting inhibition of adherence of GBS to M0. The trypsin sensitivity of the M0 mechanism involved in the recognition of GBS is compatible with the function of a membrane protein. This recognition mechanism appears to be distinct from known M0 receptors such as the Fc receptor and the C3b receptor, CRl, whose activities are resistant to trypsin digestion (6). Other properties of this streptococcal receptor, including resistance to inactivation by elastase, divalent cation requirement, sensitivity to inhibition by 2-deoxy-D-glucose, and enhancement by fibronectin and phorbol ester, are shared by members of the ~2 integrin family (see Table 4). This hypothesis is supported by additional studies in our laboratory (Sloan, A and T. Pistole, in preparation) and in others (1).

Table 4. Properties of known bacterial receptors on M0. Data on GBS receptor derived from data presented in this paper; those for other receptors from literature sources noted Properties

Temperature dependence Trypsin sensitivity Elastase resistance Ca2 + and Mg2+ requirement Ca2 + or Mg2+ requirement 2-deoxy-D-glucose inhibition FN-substrate enhancement Phorbol ester enhancement

Receptors GBS

Integrin

Lectin

Fc

+ + +

+ (35) + (6) + (30)

- (35)

+ + + +

+ (35) + (28) + (33) + (33)

- (35) - (31) ? + (31) - (31)

- (35)

- (28)

- (33) - (33)

- (6)

- (29) + (35) - (35)

+

(28)

- (33) - (33)

Macrophage Receptor for Group B Streptococci

551

Acknowledgments. We thank D. Kasper for his generous gift of bacterial strains and antisera and I. Ofek for helpful suggestions. This research was supported in part by Public Health Service Grant AI27930 from the United States National Institutes of Health and Biomedical Support Grants 2 S07 RR07108-15 and 2 S07 RR07108-17.

References

1. Antal, J. M., J. V. Cunningham, and K. J. Goodrum: Opsonin-independent phagocytosis of group B streptococci: role of complement receptor type 3. Infect. Immun. 60 (1992) 1114-1121 2. Athamna, A. and I. Ofek: Enzyme-linked immunosorbent assay for quantitation of attachment and ingestion stages of bacterial phagocytosis. J. Clin. Microbio!. 26 (1988) 62-66 3. Baker, C. J.: Immunization to prevent group B streptococcal disease: victories and vexations. J. Infect. Dis. 161 (1990) 917-921 4. Blackwell, J. M., R. A. B. Ezekowitz, M. B. Roberts, J. Y. Channon, R. B. Sim, and S. Gordon: Macrophage complement and lectin-like receptors bind Leishmania in the absence of serum. J. Exp. Med. 162 (1985) 324-331 5. Brown, E. J., S. W. Hosea, and M. M. Frank: The role of complement in the localization of pneumococci in the splanchnic reticulo-endothelial system during experimental bacteremia. J. Immuno!. 126 (1981) 2230-2235 6. Czop, J. K., D. T. Fearon, and K. F. Austen: Opsonin-independent phagocytosis of activators of the alternative complement pathway by human monocytes. J. Immuno!. 120 (1978) 1132-1138 7. Czop, J. K., N. M. Valiante, and M. J. Janusz: pp. 287-296. In: Biochemistry. of Acute Allergic Reactions: Fifth International Symposium. A. R. Liss, New York (1989) 8. Ezekowitz, R. A. B., K. Sastry, P. Bailly, and A. Warner: Molecular characterization of the human macrophage mannose receptor: Demonstration of multiple carbohydrate recognition-like domains and phagocytosis of yeasts in Cos-l cells. J. Exp. Med. 172 (1990) 1785-1794 9. Fearon, D. T. and K. F. Austen: The alternative pathway of complement: a system for host resistance to microbial infection. N. Eng!. J. Med. 303 (1980) 259-263 10. Gehring, W. and R. Northiger: The imaginal disks of Drosophila, pp. 211-290. In: S. Counce and C. H. Waddington, (eds.), Developmental Systems: Insects, voL 2. Academic Press, New York (1973) 11. Glass, E., J. Stewart, and D. M. Weir: Presence of bacterial binding 'lectin-like' receptors on phagocytes. Immunology 44 (1981) 529-534 12. Goldstein, E., W. Lippert, and D. Wadshauer: Pulmonary alveolar macrophage defend against bacterial infection of the lung. J. Clin. Invest. 54 (1974) 519-528 13. Gordon, S., V. H. Perry, S. Rabinowitz, L.-P. Chung, and H. Rosen: Plasma membrane receptors of the mononuclear phagocyte system. J. Cell Sci. Supp!. 9 (1988) 1-26 14. Griffin, F. M.: Mononuclear cell phagocytic mechanisms and host defense, pp. 31-53. In: J. I. Gallin and A. J. Fauci (eds.), Host defense mechanisms, vo!' 1. Raven Press, Publishers, New York (1982) 15. Kasper, D. L., C. J. Baker, B. Galdes, E. Katzenellenbogen, and H. J. Jennings: Immunochemical analysis and immunogenicity of the type II group B streptococcal capsular polysaccharide. J. Clin. Invest. 72 (1983) 260-269 16. Kasper, D. L., M. R. Wessels, C. E. Rubens, N. J. Levy, V. Pozsgay, and H. J. Jennillgs: Capsular polysaccharide of type III group B Streptococcus: Role in virulence and interactions with the immune system, pp. 171-186. In: Bacteria-Host Cell Interaction. Alan R. Liss, Inc., New York (1988) 17. Klein, J.: Chap. 1, Immunology: What is it all about?, pp. 3-19. In: Immunology. Blackwell Scientific, Oxford (1982)

552

A. R. Sloan and T. G. Pistole

18. Ofek, 1. and N. Sharon: Lectinophagocytosis: a molecular mechanism of recognition between cell surface sugars and lectins in the phagocytosis of bacteria. Infect. Immun. 56 (1988) 539-547 19. Paoletti, L. c., M. R. Wessels, F. Michon, H. J. Jennings, and D. L. Kasper: Group B Streptococcus type III glycoconjugate vaccines. Trends Glycosci. Glycotechnol. 4 (1992) 269-278 20. Perry, A., Y. Keisari, and I. Ofek: Liver cell and macrophage surface lectins as determinants of recognition in blood clearance and cellular attachment of bacteria. FEMS Microbiol. Lett. 27 (1985) 345-350 21. Pistole, T. G.: Limulus lectins: analogues of vertebrate immunoglobulins, pp. 283-288. In: J. Bonaventura, C. Bonaventura, and S. Tesh (eds.), Physiology and biology of the horseshoe crab. A. R. Liss, New York (1982) 22. Rollag, H. and T. Hovig: Phagocytosis of non-opsonized Escherichia coli by mouse peritoneal macrophages: an electron microscopy study. Zbl. Bakt. Hyg. A 257 (1984) 93-107 23. Schuchat, A., M. Oxtoby, S. Cochi, R. K. Sikes, A. Hightower, B. Plilaytis, and C. v. Broome: Population-based risk factors for neonatal group B streptococcal disease: Results of cohort study in metropolitan Atlanta. J. Infect. Dis. 162 (1990) 672-677 24. Schwartz, B., A. Schuchat, M. J. Oxtoby, S. L. Cochi, A. Hightower, and C. V. Broome: Invasive group B streptococcal disease in adults. A population-based study in metropolitan Atlanta. J. Amer. Med. Ass. 266 (1991) 1112-1114 25. Sharon, N.: Surface carbohydrates and surface lectins as recognition determinants in phagocytosis. Immun. Today 5 (1984) 143-147 26. Sloan, A. R. and T. G. Pistole: A quantitative method for measuring the adherence of group B streptococci to murine peritoneal exudate macrophages. J. Immunol. Meth., 154 (1992) 217-223 27. Stossel, T. P.: Phagocytosis: recognition and ingestion. Semin. Haemat. 12 (1975) 83-92 28. Sung, S.-S. J. and S. C. Silverstein: Role of 2-deoxy-D-glucose in the inhibition of phagocytosis by mouse peritoneal macrophage. Biochim. Biophys. Acta 845 (1985) 204-215 29. Tosi, M. and M. Berger: Functional differences between the 40 kDa and 50 to 70 kDa IgG Fc receptors on human neutrophils revealed by elastase treatment and anti-receptor antibodies. J. Immunol. 141 (1988) 2097-2103 30. Tosi, M. F., H. Zakem, and M. Berger: Neutrophil elastase cleaves C3bi on opsonized Pseudomonas as well as CRl on neutrophils to create a functionally important opsonin receptor mismatch. J. Clin. Invest. 86 (1990) 300-308 31. Weir, D. M. and H. M. Ogmundsd6ttir: Non-specific recognition mechanisms by mononuclear phagocytes. Clin. expo Immunol. 30 (1977) 323-329 32. Wessels, M. R., J. L. DiFabio, V.-J. Benedi, D. L. Kasper, F. Michon, J.-R. Brisson, J. Jelinkova and H. J. Jennings: Structural determination and immunochemical characterization of the type V group B Streptococcus capsular polysaccharide. J. BioI. Chern. 266 (1991) 6714-6719 33. Wright, S. D. and S. C. Silverstein: Receptors for C3b and C3bi promote phagocytosis but not the release of toxic oxygen from human phagocytes. J. Exp. Med. 159 (1983) 405-416 34. Wright, S. D., M. R. Light, L. S. Craigmyle, and S. C. Silverstein: Communication between receptors for different ligands on a single cell: ligation of fibronectin receptors induces a reversible alteration in the function of complement receptors on cultured human monocytes. J. Cell. BioI. 99 (1984) 336-339 35. Wright, S. D. and M. T. C. Jong: Adhesion-promoting receptors on human macrophages recognize E. coli by binding to lipopolysaccharide. J. Exp. Med. 164 (1986) 1876-1888

Dr. Thomas G. Pistole, Department of Microbiology, University of New Hampshire, Durham, NH 03824-3544, USA