Characterization of the Pseudomonas aeruginosa Pilus Adhesin ...

4 downloads 0 Views 2MB Size Report
Mar 28, 1989 - Following the colonization of the respiratory mucosal sur- ..... 70:677-680. 7. Hazlett, L. D., M. Moon, and R. S. Berk. 1986. In vivo identification ...
Vol. 57, No. 12

INFECTION AND IMMUNITY, Dec. 1989, p. 3720-3726

0019-9567/89/123720-07$02.00/0 Copyright © 1989, American Society for Microbiology

Characterization of the Pseudomonas aeruginosa Pilus Adhesin: Confirmation that the Pilin Structural Protein Subunit Contains a Human Epithelial Cell-Binding Domain R. T. IRVIN,l 2t* P. DOIG,' K. K. LEE,3 P. A. SASTRY,3 W. PARANCHYCH,3 T. TODD,4 AND R. S. HODGES3 Departments of Botany and Microbiology, Erindale College, University of Toronto, Mississauga, Ontario L5L IC61; Department of Microbiology2 and Division of Thoracic Surgery, Department of Surgery,4 Toronto General Hospital, 200 Elizabeth Street, Toronto, Ontario M5G 2C4; and Department of Biochemistry, University ofAlberta, Edmonton, Alberta T6G 2H73; Canada Received 28 March 1989/Accepted 4 September 1989

Previous studies have suggested that the Pseudomonas aeruginosa PAK pilus adhesin moiety resides in an epithelial cell-binding domain located in the C-terminal region of the PAK pilin structural protein. Synthetic peptides Acl7red (a synthetic peptide with a sequence identical to that of PAK pilin residues 128 to 144, with the Cys-129 and Cys-142 residues being in the reduced state) and Acl7ox (a synthetic peptide with a sequence identical to that of PAK pilin residues 128 to 144, with a formed disulfide bridge between the amino acid residues Cys-129 and Cys-142), which should contain the epithelial cell-binding domain, were synthesized. Acl7red and Acl7ox both bound to buccal epithelial cells (BECs) and to ciliated tracheal epithelial cells (TECs). Acl7ox had a Km of 6.40 ,uM for binding to BECs, while Acl7red had a Km of 9.87 I,M. Acl7red bound to the same receptor sites that purified pili did and competitively inhibited the binding of purified PAK pili to BECs. BEC glycoproteins with molecular masses of 82, 55 to 51, and 40 kilodaltons immobilized on nitrocellulose exhibited periodate-sensitive receptor activity for Acl7red; similar activity has been found for PAK pili. Acl7red, Acl7ox, and PAK pili bound to the cilia and luminal portions of the cytoplasmic membrane of human TECs, the same regions to which P. aeruginosa whole cells bind. PAK pilin has an epithelial cell-binding domain that resides in the C-terminal region of the protein. rial binding, possibly because of the two hydrophilic domains within the peptides, one domain interacting with the pilin subunit and one interacting with the BEC receptor. The peptide would thus act as a high-affinity bridge that would enhance binding. Additionally, monoclonal antibodies specific for the C-terminal end of pilin inhibit P. aeruginosa adherence to BECs, while monoclonal antibodies specific for other regions of the pilin protein do not affect adherence (P. Doig, P. A. Sastry, R. S. Hodges, K. K. Lee, W. Paranchych, and R. T. Irvin, Infect. Immun., in press). While this evidence suggests that the P. aeruginosa pilin structural protein has an epithelial cell-binding domain that functions as the pilus adhesin, it is not conclusive direct proof. We have thus investigated the ability of synthetic peptides with an amino acid sequence identical to the C-terminal region of PAK pilin (i.e., residues 128 to 144, inclusive) to function as an epithelial cell-binding domain and to bind to human respiratory epithelial cells. This peptide contains only one hydrophilic domain, unlike C-terminal peptides used previously. We have established that this synthetic peptide binds specifically to BECs and inhibits pilus binding by competing for the same receptor sites that native PAK pili do. The synthetic peptide was also observed to bind to the same locations on human tracheal epithelial cells (TECs) that native PAK pili do. These studies have confirmed that PAK pilin contains an epithelial cell-binding domain in the C-terminal region of the protein.

Pseudomonas aeruginosa respiratory infections continue a significant source of morbidity and mortality for intubated intensive care unit patients (6, 10) and cystic fibrosis patients (26). P. aeruginosa infections are thought to be initiated upon adherence of the organism to, and subsequent colonization of, the respiratory epithelium (11, 12, 29). Following the colonization of the respiratory mucosal surface, a descending-infection mechanism leads to the development of a pulmonary infection (25, 27). P. aeruginosa can use both pili and alginate (the principle component of the P. aeruginosa capsule) as adhesins to mediate attachment to human respiratory epithelial cells (3, 4, 17, 23, 24, 30). The pilus adhesin probably mediates the initial adherence of P. aeruginosa to the respiratory epithelial surface, because the pilus adhesin is considerably more effective than the alginate adhesin in mediating adherence (3, 4, 19, 20). Characterization of the P. aeruginosa pilus adhesin may thus lead to a better understanding of the initiation of P. aeruginosa respiratory infections. The exact nature of the P. aeruginosa pilus adhesin is not known, although the available evidence suggests that, unlike the P pilus and type I pilus of Escherichia coli (1, 13, 15, 16), the pilin structural subunit functions as the actual adhesin. P. aeruginosa PAK pilin appears to contain an epithelial cell-binding domain because native peptides derived from the C-terminal end of the PAK pilin structural protein and synthetic peptides with an identical sequence significantly affect P. aeruginosa binding to human buccal epithelial cells (BECs) (4, 22). These C-terminal peptides enhanced bacteto constitute

MATERIALS AND METHODS Corresponding author. t Present address: Department of Medical Microbiology, University of Alberta, Edmonton, Alberta T6G 2H7, Canada. *

Bacterial strain, culture conditions, and pilus purification. P. aeruginosa PAK and the culture conditions used have 3720

VOL. 57, 1989

PAK PILIN EPITHELIAL CELL-BINDING DOMAIN

been previously described (2, 21). Pili from PAK were purified as previously described (21). Peptide synthesis. Synthetic peptides Acl7red (a synthetic peptide with a sequence identical to that of PAK pilin residues 128 to 144, with the Cys-129 and Cys-142 residues being in the reduced state) and Acl7ox (a synthetic peptide with a sequence identical to that of PAK pilin residues 128 to 144, with a formed disulfide bridge between the amino acid residues Cys-129 and Cys-142) were prepared as previously described (4). Ac17 cysteine residues were oxidized to form the intrachain disulfide bridge (Acl7ox) by the method of Hodges et al. (9), and the formation of the disulfide bridge was verified on the basis of circular dichroism spectra by the method of Matsoukas et al. (18). BECs. BECs were collected with wooden applicator sticks from healthy, nonsmoking male volunteers (n = 10). BECs were removed from the applicator sticks by gentle agitation in phosphate-buffered saline (PBS), pH 7.2. The BECs were washed three times (2,000 x g for 10 min at 4°C) with PBS and then passed through a 70-,um-pore-size nylon mesh. The cell concentration for the BECs was determined with a hemacytometer and was adjusted to 2.0 x 105/ml. TEC preparation. Human ciliated TECs were obtained from patients in the surgical intensive care unit at Toronto General Hospital by bronchoscopic brushing of the bronchial mucosa by the method of Franklin et al. (5). All procedures were approved by the Toronto General Hospital Ethics Committee, and informed consent was obtained from the relatives of the patients. TECs were fixed with 1% formaldehyde in PBS for 1 h at room temperature. The cells were then washed twice with PBS and resuspended in PBS. TECs were counted with a hemacytometer, the TEC concentration was adjusted to 105/ml, and the formaldehyde-fixed cells were then used directly. Ac17 binding to BECs. An immunoassay was performed to assess the binding of Acl7red and Acl7ox to BECs. BECs (0.2 ml at 2.0 x 105/ml) were added to an equal volume of synthetic peptide (0 to 120 nmol/ml) in PBS and incubated at 37°C and were then agitated at 300 rpm. After 1 h, BECs were collected by centrifugation (13,000 x g for 2 min at room temperature) and washed five times with PBS. Mono-

clonal antibody PK99H (0.2 ml of a 10-' dilution [in PBS] of purified immunoglobulin G [IgG] with a titer of 106) was added to the BEC pellet and incubated, as described above, for 1 h. The BECs were then collected by centrifugation (13,000 x g for 2 min at room temperature) and washed five times with PBS. Goat anti-mouse heavy and light chain [IgG(H+L)] immunoglobulin G-peroxidase conjugate (Jackson Laboratories) was added to the BEC pellet (0.2 ml of a 1:10,000 dilution in PBS), and the mixture was incubated, as described above, for 1 h. The BECs were collected by centrifugation (13,000 x g for 2 min at room temperature) and washed five times with PBS. The pellet was then suspended in 0.2 ml of 1 mM 2,2'-azino-di-(3-ethylbenzthiazoline sulfonic acid) in citrate buffer (pH 4.2) plus 0.03% (wt/vol) peroxide and transferred to a clean tube. The horseradish peroxidase enzyme reaction was stopped by the addition of 0.2 ml of 4 mM NaN3, and the optical density at 405 nm was determined after removal of the BECs by centrifugation. The BEC concentration in each tube was

3721

determined with a hemacytometer at the end of the assay, prior to the removal of BECs by centrifugation. Ac17 inhibition of pilus binding to BECs. An immunoassay was performed to assess the effect of Acl7red on pilus binding to BECs. BECs (0.2 ml at 2.0 x 105/ml) and synthetic peptide Acl7red (0.1 ml such that a final concentration of 0, 40, 80, or 120 nmol of synthetic peptide per ml was obtained) were preincubated for 30 min at room temperature. Pili (0.1 ml at 0 to 100 ,ug/ml) were then added to the BECs with various concentrations of synthetic peptide (0, 40, 80, or 120 nmol/ml). The mixtures were then incubated for 2 h at 37°C while being agitated at 300 rpm. BECs and bound pili were then collected by centrifugation (13,000 x g for 2 min) and washed five times with PBS to remove unbound pili. Monoclonal antibody PK3B (0.1 ml of a 10-4 dilution in PBS) (this antibody recognizes PAK pili but does not react with synthetic peptide Acl7red [Doig et al., in press]) was then added to the BECs with bound pili and incubated for 1 h, as described above. The remainder of the immunoassay was the same as that described for pilus binding to BECs. Pilus and Ac17 binding to TECs. TECs (0.1 ml at 105/ml) were mixed with an equal volume of PAK pili (345 ,ug/ml),

0.25

0.2 to

5* 0.15

0

4C

0.05

/

/.II

O0 T

0

5

10

15

20

25

30

35

nmol/ml FIG. 1. Binding of synthetic peptide Acl7red (-) and synthetic peptide Acl7ox (+) to human BECs. Binding of the synthetic peptides to BECs was determined by a whole-cell ELISA utilizing the monoclonal antibody PK99H (which binds to both Acl7ox and Acl7red) to quantitate the amount of synthetic peptide bound to the surface of BECs.

3722

INFECT. IMMUN.

IRVIN ET AL.

TABLE 1. Summary of binding parameters calculated from modified Lineweaver-Burk plots for PAK pilus binding to BECs in the presence of various concentrations of synthetic peptide Acl7red Concn of Acl7red

Kma

Vmax

200 rb

(nmollml)

92 0.88 0.97 0.99 0.96

66

Units for the Km for pilus binding to BECs are in milligrams per liter. r is the correlation coefficient calculated for the least-squares regression line of the data. c Significantly different (P s 0.005) from the control (O nmol of Acl7red added per ml) by using a two-tailed Student t test.

45

16.4 14.6 20.7 18.8

0 40 80 120

± 0.7

± 1.0 ± 1.OC ± 0.8c

1.098 1.090 0.979 0.898

± 0.142 ± 0.140 ± 0.115 ± 0.137

a

b

Acl7red (10 nmol/ml), Acl7ox (50 nmol/ml), or PBS. The mixture was incubated at 37°C for 1 h and agitated at 300 rpm. TECs were then collected by centrifugation (6,000 x g for 1 min at room temperature) and washed three times with PBS. Anti-pilus monoclonal antibody PK3B (0.1 ml of a 10-' dilution of purified IgG with a titer of 108 in PBS) or monoclonal antibody PK99H (0.1 ml of a 10-3 dilution of purified IgG with a titer of 106 in PBS) was added to TECs incubated with pili or with synthetic peptides Acl7red and Acl7ox, respectively. (PK3B reacts with PAK pili without affecting pilus-binding activity, but PK3B does not react with either Acl7red or Acl7ox.) Control preparations included monoclonal antibodies PK3B and PK99H incubated with TECs in PBS without pili or synthetic peptides. TECs were then collected by centrifugation and washed three times with PBS. Rabbit anti-mouse IgG, IgM(H+L) affinitypurified IgG conjugated to fluorescein isothiocyanate (Cedarlane Laboratories) in PBS (0.1 ml of a 10-2 dilution) was added to the washed TEC preparations, incubated for 30 min at 37°C, and agitated at 300 rpm. The TECs were washed three times, as described above, and resuspended in 0.1 ml of PBS. Wet mounts were prepared and examined by epifluorescence and phase-contrast microscopy by using a Leitz Laborlux microscope equipped with an MPS4 camera system. Photographs were recorded on Kodak T-Max film. Ac17 binding to BEC blots. The discontinuous sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-

3

v-

-0.05

0

0.05

0.1

1/(ug/ml PAK pili) FIG. 2. Modified Lineweaver-Burk plot of the binding of PAK pili to human BECs in the presence of 0 (,- ), 40 (x), 80 (A), and 120 (O) nmol of synthetic peptide Acl7red per ml.

32

123 4 56 7 8 9 10S FIG. 3. Binding of PAK pili and synthetic peptide Acl7red to blotted BEC proteins on nitrocellulose. Binding of synthetic peptide to the immobilized proteins was assessed (following blocking of the nitrocellulose with bovine serum albumin) by using monoclonal antibody PK99H (or monoclonal antibody PK3B for PAK pili), followed by standard immunoblotting methods. PAK pili at 150 ,ug/ml (lane 1) or Acl7red at 20 (lane 2), 10 (lane 3), 5 (lane 4), o4 0 (lane 5) nmol/ml was incubated with the blotted BEC protein. BEC proteins were oxidized by exposure to 30 mM periodate and then reduced with borohydride before incubation with 20 nmol of Acl7red per ml (lane 6) or buffer (lane 7, as a control for lane 6). Acl7red at 20 nmol/ml was initially reacted with 100 ,ug of Fab fragments of monoclonal antibody PK99H per ml (which binds to synthetic peptide Acl7red) (lane 8) or reacted with 100 ,ug of Fab fragments of monoclonal antibody PK41C per ml, which does not bind to synthetic peptide Acl7red (lane 9). Amido black-stained BEC proteins are shown in lane 10. Standard molecular weight markers stained with amido black are shown in lane S.

PAGE) described by Laemmli and Favre (14) was used to assess Ac17 binding. SDS-PAGE of BECs was performed by using 8% acrylamide gels. BECs (2 x 105/ml) were solubilized at 100°C for 15 min in 2% (wt/vol) sodium dodecyl sulfate-5% (vol/vol) P-mercaptoethanol-10% (vol/vol) glycerol in 0.625 mM Tris buffer (pH 6.8). Solubilized BECs (25 pul) were loaded on the gel and electrophoresed at 20 mA per gel (constant current). Electrophoretically separated material was transfened to nitrocellulose (Schleicher & Schuell, Inc.) by the electrophoretic transfer method of Towbin et al. (28). After transfer, nitrocellulose blots were blocked with 3% (wt/vol) bovine serum albumin-0.25% (wt/vol) gelatin0.1% (vol/vol) normal rabbit serum-0.05% (vol/vol) Nonidet P-40-5 mM EDTA-150 mM sodium chloride in 50 mM Tris buffer (pH 7.5) at 37°C for at least 3 h. Before use, blots were rinsed with PBS. Blots were then incubated with Acl7red (0 to 20 nmol/ml) at 37°C with shaking at 100 rpm. After 2 h, blots were washed three times with 50 mM Tris (pH 7.5) containing 0.85% (wt/vol) sodium chloride and 0.1% (wt/vol) Tween 20 (TTBS) (10 min per wash). Murine monoclonal antibody PK99H (10-4 dilution in TTBS) (this monoclonal antibody recognizes the Ac17 syn-

PAK PILIN EPITHELIAL CELL-BINDING DOMAIN

VOL. 57, 1989

A

C

3723

mm

-

D

FIG. 4. Indirect immunofluorescence localization of PAK pilus binding to fractionated ciliated TECs. (A) Phase-contrast micrograph of TEC with bound PAK pili. (B) Immunofluorescence micrograph of PAK pili bound primarily to the cilia and luminal portion of the cytoplasmic membrane of the same ciliated TECs visualized by phase-contrast microscopy in panel A. (C and D) Phase-contrast image and immunofluorescence image, respectively, of a control TEC exposed to monoclonal antibody PK3B and fluorescein isothiocyanate-conjugated anti-mouse IgG but not exposed to PAK pili.

thetic peptide) was incubated with the blot at 37°C for 1 h at 100 rpm. The blot was then washed three times with TTBS. A goat anti-mouse IgG(H+L) immunoglobulin-alkaline phosphatase conjugate (Jackson Laboratories) in TTBS was added and incubated for 1 h as described above. The blot was washed three times with TTBS and once with Trisbuffered saline. A substrate solution (NBT/BCIP), consisting of 0.33 mg of Nitro Blue Tetrazolium chloride per ml, 0.165 mg of 5-bromo-4-chloro-3-indolyl-phosphate per ml, 100 mM sodium chloride, and 5 mM magnesium chloride in 100 mM Tris buffer (pH 9.5), was added, and color development was stopped by rinsing the blot in distilled water. The binding of synthetic peptide Acl7ox to blotted BEC protein was not examined at this time. Synthetic peptide Acl7ox was not used, in part because of the limited availability of the synthetic peptide and in part because of the limited affinity of monoclonal antibody PK99H for the Acl7ox peptide, compared with its affinity for the Acl7red peptide (Doig et al., in press). While monoclonal antibody PK99H was suitable for immunofluorescence studies with Acl7ox, the higher background levels associated with blots precluded examination of Acl7ox binding to blotted BEC proteins at this time. Periodate oxidation of BEC blots. Periodate oxidation (30 mM periodate) and subsequent potassium borohydride reduction of BEC blots were performed by the method of

Woodward et al. (31). Oxidation-reduction of the BEC glycoproteins was performed on preblocked blots. Blots were assessed for synthetic peptide binding as described above. RESULTS Binding of Acl7red and Acl7ox to BECs. Synthetic peptides Acl7ox and Acl7red both bound to BECs in a concentration-dependent manner (Fig. 1). Saturation of binding of the synthetic peptide Acl7ox to BECs appeared to occur at a concentration of -30 nmol/ml, but the binding of the Acl7red synthetic peptide to BECs did not saturate under our experimental conditions (Fig. 1). Modified LineweaverBurk plots of synthetic peptide binding to BECs indicated that the apparent Km of binding to BECs (i.e., the concentration at which 50% of the receptor sites would be occupied) for Acl7red was 9.87 ,uM, while the Km of binding to BECs for Acl7ox was 6.40 puM. Synthetic peptide inhibition of pilus binding. Preliminary studies indicated that both Acl7red and Acl7ox significantly inhibited the binding of PAK pili to BECs in binding studies (data not presented). Because of the limited availability of synthetic peptide Acl7ox, the nature of the inhibition of PAK pilus binding to BECs was investigated with the Acl7red synthetic peptide. The inhibition of PAK pilus

3724

..:5KIwtB-XiE