INFECTION AND IMMUNITY, Apr. 1991, p. 1223-1230 0019-9567/91/041223-08$02.00/0 Copyright © 1991, American Society for Microbiology

Vol. 59, No. 4

Binding Kinetics of Clostridium difficile Toxins A and B to Intestinal Brush Border Membranes from Infant and Adult Hamsters RIAL D. ROLFE Department of Microbiology, School of Medicine, Texas Tech University Health Sciences Center, Lubbock, Texas 79430 Received 6 September 1990/Accepted 3 January 1991

This study was undertaken to determine if the relative resistance of neonates and infants to Clostridium

difficile-associated intestinal disease can be related to age-dependent differences in intestinal receptors for C. difficile toxins A and B. Brush border membranes (BBMs) from the small intestines of adult and infant hamsters were examined for their ability to bind radiolabeled toxins A and B. ['251]toxin A bound to both infant and adult hamster BBMs at physiological temperature, whereas ('25lJtoxin B did not bind to the BBMs under any of the conditions examined. The number of ['251]toxin A molecules bound at saturation was approximately 4 x 1010 per ,ug of membrane protein for adult BBMs and 1 x 101l per ,ug of membrane protein for infant BBMs. Scatchard plot analysis suggested the presence of a single class of toxin A binding sites on both infant and adult hamster BBMs. Maximal binding capacity and Kd values were 0.63 pmol/mg of protein and 66.7 nM, respectively, for the infant BBMs, and 0.24 pmol/mg of protein and 27 nM, respectively, for the adult BBMs. Sodium dodecyl sulfate-polyacrylamide gel electrophoretic analyses of extracted BBM proteins revealed differences in the proteins of infant and adult BBMs. However, there were not any detectable differences in the protein bands which bound [1251Jtoxin A between infant and adult hamsters. The results from these investigations indicate that differences in the binding kinetics of toxins A and/or B to infant and adult hamster BBMs do not account for the observed differences in their susceptibility to C. difficile-associated intestinal disease.

The pathogenic role of Clostridium difficile in the etiology of antimicrobial agent-associated intestinal disease is well established. C. difficile is the cause of essentially all cases of pseudomembranous colitis, a severe and life-threatening disease, and approximately 20% of cases of less severe intestinal disease including antimicrobial agent-associated nonspecific colitis or diarrhea without colitis (4, 14). The underlying basis for these intestinal diseases is a disruption of the normal intestinal microbial flora which affords toxigenic C. difficile the opportunity to multiply and obtain high numbers in the gastrointestinal tract (47). C. difficile produces two biochemically and immunochemically distinct toxins which appear to play important roles in the pathogenesis of this microorganism (9, 23, 26, 28). These toxins are referred to as toxin A (enterotoxin) and toxin B (cytotoxin). During the last several years it has been documented that C. difficile intestinal colonization and toxin production occur in both the presence and absence of disease (38). Asymptomatic intestinal colonization is particularly evident in human infants, where greater than 50% of infants less than 1 year of age are colonized asymptomatically with toxigenic C. difficile (38). Intestinal carrier rates for C. difficile decline to approximately 30% during the second year of life, a carrier rate still significantly higher than the average 2% reported to occur in the healthy adult population (38). Concentrations of C. difficile and toxins A and B in the feces of healthy infants are frequently similar to concentrations found in the intestinal tracts of adults with clinical manifestations of C. difficileassociated intestinal disease (33, 38). The factors which permit asymptomatic intestinal colonization of newborns and infants by toxigenic C. difficile remain to be elucidated, although a number of mechanisms have been proposed. For example, in many animal species, the small intestine epithelial cells undergo marked develop-

mental changes during postnatal development. Postnatal age has a significant influence on the appearance of mucosal receptor proteins for bacterial toxins as well as the types of intestinal diseases which occur (6, 22, 30, 38). However, it is not known if developmental changes occur in the intestinal tract of humans that influence the degree of binding of C. difficile toxins A and B. The investigations in this study were designed to determine if the relative resistance of neonates and infants to C. difficile-associated intestinal disease can be related to agedependent differences in intestinal receptors for C. difficile toxins A and B. Brush border membranes (BBMs) from the small intestines of adult and infant hamsters were examined for their ability to bind radiolabeled toxins A and B. The hamster model of C. difficile-associated diarrhea has been recognized on several grounds as the most appropriate animal model for diarrhea caused by toxigenic C. difficile (2, 35). Adult hamsters treated with antimicrobial agents consistently develop an intestinal disease similar to pseudomembranous colitis in humans (35). On the other hand, infant hamsters have an age-dependent susceptibility to asymptomatic enteric C. difficile colonization similar to the restricted age distribution of human newborn asymptomatic colonization (40, 41). The similarities between symptomatic and asymptomatic C. difficile intestinal colonization of hamsters and humans suggest that the results obtained with the hamster model will directly aid in our understanding of C. difficile-associated disease in humans. MATERIALS AND METHODS Purification of C. difficile toxins A and B. C. difficile toxins A and B were purified from brain heart infusion (BHI; BBL Microbiology Systems, Cockeysville, Md.) dialysis bag culture filtrates of C. difficile VPI 10463 (Department of Anaer1223

1224

ROLFE

obic Microbiology, Virginia Polytechnic Institute and State University, Blacksburg) by using previously established techniques (17, 18, 28, 42). Briefly, 72-h dialysis bag culture filtrates of this highly toxigenic strain of C. difficile were subjected to positive-pressure ultrafiltration through an XM100 (100,000-molecular-weight cutoff; Amicon Corp., Lexington, Mass.) membrane filter. The retentate was then fractionated on a Sepharose CL-6B (Pharmacia Fine Chemicals, Uppsala, Sweden) column, and the fractions displaying cytotoxicity to HeLa tissue culture cells were pooled. Toxins A and B present in the pooled fractions were separated by DEAE-Sepharose CL-6B ion-exchange chromatography with step gradient elutions at 0.25 and 0.45 M NaCl, respectively. The separated toxins were further purified by DEAE-Sepharose CL-6B ion-exchange chromatography and NaCl linear gradients. In the last step, toxin A was further purified on a thyroglobulin affinity column (20). The homogeneity of purified preparations of C. difficile toxins A and B were characterized as described previously by the following techniques: polyacrylamide gel electrophoresis (PAGE) in 7.5% native gels and Western blot (immunoblot) analyses, crossed immunoelectrophoresis, cytotoxin neutralization, and enzyme-linked immunosorbent assays with affinity-purified rabbit antibodies to toxins A and B (18, 42). Preparation of radiolabeled toxins A and B. As previously described, purified C. difficile toxins A and B were radioiodinated enzymatically by using Enzymobead lactoperoxidase-glucose oxidase (Bio-Rad Laboratories, Richmond, Calif.) (42). Approximately 500 pug of protein was labeled with 2 mCi of reductant and carrier-free sodium 1251 (ICN Radiochemicals, Irvine, Calif.; specific activity, 17 Ci/mg). The iodination was performed at room temperature for 30 min. Upon completion of the reaction, the mixture was centrifuged and the supernatant was applied to a Sephadex G-25 (Pharmacia Fine Chemicals) column to separate free 1251 from protein-bound 1251. Each fraction of the column effluent was monitored for gamma emissions and assayed for cytotoxicity to HeLa cells. The fractions containing labeled toxins were pooled and stored at -70°C. Maintenance of hamsters. Adult and infant golden Syrian hamsters (Sasco, Inc., Omaha, Nebr.) were used throughout the course of this investigation. Female hamsters were bred by housing one female and one male hamster together for 7 days after which the females were caged individually. Adult (>3 months of age) and 7-day-old infant hamsters were housed in conventional rooms and maintained ad libitum on Purina Laboratory rodent chow 5001 (Ralston Purina Co., Richmond, Ind.) and water. Before introduction into the study, fecal specimens from all hamsters were cultured on C. difficile selective medium (15). None of the hamsters used in these investigations were colonized with C. difficile. Preparation of hamster BBMs. BBMs were prepared from the small intestines of 7-day-old infant and adult hamsters by a modification of previously described methods (3, 16). To obtain a sufficient yield of brush borders from 7-day-old infant hamsters, seven hamsters (male and female) from an individual litter were killed at the same time and their mucosal scrapings were combined and used as one independent sample. Individual male adult hamsters from different litters were sacrificed and used as independent samples. Animals were anesthetized with intraperitoneal sodium pentobarbital and killed by cervical dislocation. The entire small intestine of each animal was removed, and the lumen was immediately irrigated with cold 0.9% NaCl. All further procedures were performed at 4°C. The small intestine

INFECT. IMMUN.

beyond the ligament of Treitz, excluding the terminal lumen, was used for preparation of BBMs. The intestine of each animal was everted over a glass rod, and the mucosa was obtained by scraping the intestine with a glass microscope slide. The musocal scrapings of each animal were suspended in HMBA buffer (HEPES [N-2-hydroxyethylpiperazine-N'2-ethanesulfonic acid], 10 mM; butylamine, 7 mM; adjusted to pH 7.4 with 0.5 M maleic acid) containing 5 mM neutralized EGTA [ethylene glycol-bis(3-aminoethyl ether)-N,N, N',N'-tetraacetic acid] and 10 mM sorbitol. The mixture was homogenized in a Waring blender for 25 s, and the homogenate was passed through a nylon mesh membrane (7-mm mesh size). Solid sorbitol was added to the filtrate to bring the final sorbitol concentration to 500 mM. The filtrate was then homogenized in a Waring blender for 4 min, followed by the addition of solid MgSO4 to the homogenate so that the final concentration was 20 mM. After a 15-min incubation period at 4°C, the suspension was centrifuged at 3,000 x g for 15 min, and the pellet, which contained cellular debris and organelles, was discarded. The supernatant was recentrifuged at 28,000 x g for 30 min, and the recovered pellet, containing BBMs, was suspended in HMBA buffer containing 500 mM sorbitol and stored in small aliquots at -70°C. Characterization of BBMs. The purity of the BBMs was evaluated morphologically by examination under phasecontrast microscopy and biochemically by comparing the specific activities of sucrase, lactase, and alkaline phosphatase with those in the initial homogenate. Sucrase and lactase activities were determined by the measurement of liberated glucose during hydrolysis of sucrose or lactose, respectively, by using a glucose oxidase reagent (7). Alkaline phosphatase was assayed as described by Ray (37). Binding of radiolabeled toxins A and B to hamster BBMs. The binding of radiolabeled C. difficile toxins A and B to infant and adult hamster BBMs was assayed indirectly by determining the amount of residual, unbound radiolabeled toxin present after incubations were completed. Binding assays were performed in 1.5-ml polypropylene microcentrifuge tubes which were treated overnight at 4°C with 1 ml of MEM (Eagle minimum essential medium with L-glutamine; Whittaker Bioproducts, Inc., Walkersville, Md.) containing 2% bovine serum albumin. Toxin and BBM preparations were quickly thawed and diluted to give the desired protein concentrations in MEM containing 0.2% bovine serum albumin. Unless otherwise stated, the binding assays were performed by incubating 7.5 ng of [125I]toxin A or ['25I]toxin B with 15 ,ug of BBM protein in a total volume of 0.3 ml of MEM containing 0.2% bovine serum albumin. Reaction mixtures, prepared in triplicate, were incubated at 4°C for 2 h. At the end of the incubation period, separation of free and bound radiolabeled toxin was performed by centrifugation of the BBMs at 16,000 x g for 5 min. Unbound toxin in 200 pl of the supernatant was counted in a gamma counter (GammaTrac 193; Tm Analytical, Elk Grove, Ill.). Preliminary investigations demonstrated that measuring unbound toxin in the supernatant after centrifugation could be used to accurately quantitate the amount of toxin bound to BBMs. For each experiment, controls were done in which BBMs were replaced by the same volume of buffer; these controls were processed as described for samples and counted to determine the nonspecific binding of radiolabeled toxin to the microcentrifuge tubes. Under these conditions, binding of the labeled toxins to the microcentrifuge tubes was less than 1% of the added radioactivity. Specific binding of [125I]toxin A or [125I]toxin B was determined by simultaneous addition of a 1,000-fold excess

VOL. 59, 1991

of unlabeled toxin to duplicate reaction mixtures prepared as described above. Residual counts under these conditions were subtracted from counts obtained in the absence of unlabeled toxin. Nonspecific toxin binding to infant and adult BBMs consistently accounted for less than 10% of BBM bound radioactivity. All toxin binding data are presented as specifically bound radioactivity. SDS-PAGE of BBM proteins. BBM preparations (10 ,ug of protein) from infant and adult hamsters were diluted 1:1 (vol/vol) with Tris buffer (0.25 M, pH 6.8) containing 4% (wt/vol) sodium dodecyl sulfate (SDS) and 10% (vol/vol) 2-mercaptoethanol. The samples were heated at 100°C for 5 min and then centrifuged at 16,000 x g for 1 min. The supernatants were then subjected to SDS-PAGE by utilizing a 7.5% lower gel and an upper stacking gel essentially as described by Laemmli (21). On each gel, the same amount of BBM protein from animals of different ages was used. After electrophoresis, the gels were fixed and stained with Coomassie brilliant blue R-250 as previously described (39). Autoradiography of toxin-BBM protein interactions. BBM proteins, extracted and separated by SDS-PAGE as described above, were transferred electrophoretically to nitrocellulose membranes essentially by the procedure of Towbin (44). After transfer, the membranes were washed in phosphate-buffered saline (PBS; 0.01 M, pH 7.2) for 10 min and then incubated in a blocking solution (PBS containing 3% [wt/vol] bovine serum albumin and 1% [vol/vol] newborn calf serum) overnight at 4°C. The membrane was then incubated with shaking for 2 h with a 1-,ug/ml solution (in PBS) of [125I]toxin A or ['25I]toxin B. The membranes were then washed three times with PBS and dried at room temperature. Autoradiography was performed by exposing the nitrocellulose membranes to X-OMAT AR film (Eastman Kodak Co., Rochester, N.Y.) at -70°C for 18 to 24 h with an intensifying screen. Protein determination. Protein concentrations of BBMs and toxin preparations were determined by the procedure of Lowry et al. (24) with bovine serum albumin as the standard. Statistical analysis. The Student's t test for differences between independent means was used to evaluate statistical differences between toxins binding to infant and adult BBMs. RESULTS Purified toxin characteristics. As previously reported, at the final stages of the purification process, toxins A and B were biochemically and immunologically distinct (17, 18). The enterotoxic, cytotoxic, and lethal activities of the purified toxins were within the range of values reported previously for toxins A and B (26, 27, 42). Radiolabeled toxin characteristics. ['25Iltoxin A and [125I] toxin B, when passed over a Sephadex G-25 column, came off in the void volume as a single peak. Greater than 90% of the total radioactivity associated with both [1251]toxin A and [125I]toxin B was precipitated with cold 10% trichloroacetic acid. The average specific activities and labeling efficiencies for each toxin were, respectively, 6.5 x 106 cpm/,Lg and 82% for [125 ]toxin A and 3.2 x 106 cpm/,ug and 86% for [1251] toxin B. As previously reported, both toxins retain their physicochemical, immunochemical, and biological integrity during the radioiodination procedure (42). Nondenaturing PAGE and autoradiography demonstrated that [125 ]toxin A and [1251]toxin B comigrated with their respective unlabeled toxins. Both radiolabeled toxins retained their immunological reactivity as demonstrated by cytotoxicity neutraliza-

BINDING OF C. DIFFICILE TOXINS TO BBMs

1225

TABLE 1. Enzymatic activities of hamster small intestine BBMs Enzymatic activity

Enzyme

Aklaline phosphatase Sucrase Lactase

(IU/mg of BBM protein) of: Adult BBMs

Infant BBMs

66.2 + 17.5 25.4 + 3.4 39.1 + 4.1

84.8 + 23.0 10.2 + 1.9 311.5 + 16.3

tion, Western blot analyses, and crossed immunoelectrophoresis by using monospecific rabbit antisera to each unlabeled toxin. Immunodiffusion with monospecific rabbit antisera to toxins A and B showed that both radiolabeled toxins displayed reactions of identity with their respective unlabeled toxins. [125I]toxin A and ['25I]toxin B maintained greater than 90% of their biological activity as demonstrated by cytotoxicity to HeLa tissue culture cells, mouse lethality (intraperitoneal inoculation), hamster lethality (intraperitoneal inoculation), and induction of hamster ileocecitis (intracecal inoculation). In addition, the cold hemagglutination activity of [1251]toxin A against rabbit erythrocytes did not appreciably change as a result of iodination (19). Characteristics of hamster BBMs. Table 1 shows the alkaline phosphatase-, sucrase-, and lactase-specific activities in the final infant and adult BBM preparations. In all cases, these marker enzymes were enriched at least 10-fold, compared with the initial crude homogenate. Enrichments of these enzymatic activities are in close agreement with previous reports (7, 8, 13). Effect of incubation time on toxin binding. The binding of [1251]toxin A and [125I]toxin B to infant and adult hamster BBMs was analyzed by examining the levels of radioactivity associated with BBMs in suspension. Adult and infant BBMs were incubated at 40C with [125I]toxin A or [125I]toxin B. At timed intervals after the addition of the labeled toxins, the amount of bound toxin was determined. Figure 1 shows that binding of [1251]toxin A to adult and infant BBMs was very rapid, achieving steady state within 30 min. A significant amount of [1251I]toxin A was bound immediately after the addition of [1251]toxin A. [125I]toxin B did not bind to infant or adult BBMs at any of the time intervals examined. Effect of toxin concentration on binding to BBMs. Binding of [1251]toxin A and [125I]toxin B was also examined by incubating increasing concentrations of labeled toxin with a constant amount of adult or infant BBMs. A linear relationship between binding and [1251]toxin A concentration occurred up to a toxin concentration of approximately 5,000 ng/ml (Fig. 2). Above this concentration, [125I]toxin A binding reached a plateau, indicating a saturation of the BBM binding sites. Binding of [1251]toxin A to adult and infant BBMs was specific as shown by the virtually complete competitive inhibition of [1251]toxin A binding with a 1,000fold excess of unlabeled toxin A. In accordance with earlier findings, [1251]toxin B did not bind to either infant or adult BBMs at any of the toxin concentrations examined. Consistent with this finding was the observation that unlabeled toxin B did not competitively inhibit the binding of [12511 toxin A to either infant and adult BBMs. The number of [125I]toxin A molecules bound at saturation was approximately 4 x 1010 per ,ug of membrane protein for adult BBMs and 1 x 1011 per jig of membrane protein for infant BBMs. These calculations were based on a molecular weight of 308,000 for toxin A (10).

1226

INFECT. IMMUN.

ROLFE 70

50

z

0

40

60

m

I

.C 0

z

-a

Oz

30-

LUG.L

C.

.

L

.J

50

z

L Z

0.scz ulE

20

Xa.

Z) E

-a 40

.cn

w

10-

-J

c

0

20

40

60

80

100

120

o0

INFANT TIME (minutes)

ADULT

FIG. 1. Time course of [125I]toxin A binding to infant (A) and adult (0) hamster BBMs. Reaction mixtures contained BBMs (15 ,ug of membrane protein) and 7.5 ng of [1251]toxin A in 0.3 ml of MEM containing 0.2% bovine serum albumin. Each point represents the mean of three experiments in which BBM preparations from different hamsters were used for each experiment. Standard deviations were less than 10% of the mean. ['25I]toxin B (7.5 ng) did not bind to infant and adult BBMs at any of the incubation times.

FIG. 3. Effect of temperature on the binding of [125I]toxin A to infant and adult hamster BBMs. BBMs from infant and adult hamsters (15 p.g of membrane protein) were incubated with [I251]toxin A (7.5 ng) at 4 or 37°C for 2 h in 0.3 ml of MEM containing 0.2% bovine serum albumin. Each point represents the mean of three experiments in which BBM preparations from different hamsters were used for each experiment. Standard deviations were less than 10% of the mean. [1251]toxin B did not bind to BBMs at either temperature.

Effect of incubation temperature on toxin binding. Figure 3 shows that infant and adult hamster BBMs bound ['251]toxin A at both 4 and 37°C. However, binding of ['25I]toxin A to both BBM preparations was enhanced at 4°C. Timed incu-

bation experiments demonstrated that [125I]toxin A binding to adult and infant BBMs was consistently lower at 37°C than at 4°C at all time intervals examined (2 to 120 min). However, similar to what was observed at 4°C (Fig. 1), [1251]

a

100000

a0.

z Oz

JiD I1000000

0

EL- z 4x 100000

m

u)

1000 10000

Z E x

1000o

01-

100

0

0

0

5000

200

400 10000

600

800

1 oo0

15000

1251-TOXIN A (ng) FIG. 2. Effect of toxin concentration on the binding of [l251]toxin A to infant (A) and adult (0) hamster BBMs. Increasing amounts of labeled toxin (1 to 15,000 ng) were added to 15 ,ug of BBM protein and incubated for 2 h at 4°C in MEM containing 0.2% bovine serum albumin. Each point represents the mean of three experiments in which BBM preparations from different hamsters were used for each experiment. Standard deviations were less than 10% of the mean. [1251]toxin B did not bind to BBMs at any toxin concentration. Inset figure shows the effect of low [1251I]toxin A concentrations on binding to BBMs.

VOL. 59, 1991

BINDING OF C. DIFFICILE TOXINS TO BBMs

1227

0.2

CL

60

-

z 0

1--

0.1 0

.J Z

0 cc -0 40 .1

0

0~~~~~~

0.0' Toxin A (pmol)

FIG. 4. Competitive inhibition of [1251]toxin A binding to adult (0) and infant (A) BBMs by increasing concentrations of native toxin A. BBMs from infant and adult hamsters (15 ,ug of membrane protein) were incubated with a constant amount of [1251]toxin A (7.5

ng) and increasing concentrations of native toxin A in 0.3 ml of MEM containing 0.2%t bovine serum albumin. Each point represents the mean of three experiments in which BBM preparations from different hamsters were used for each experiment. Standard deviations were less than 10%t of the mean.

toxin A binding to adult and infant BBMs at 37°C was very rapid, achieving a steady state within 30 min. In addition, the level of [1251]toxin A binding at 37°C remained constant between 30[125-]toxin and 120 mm. B did not bind to infant or adult BBMs at either temperature examined. Competitive toxin binding and Scatchard plot analysis. For competitive binding studies, BBMs were preincubated with specific amounts of unlabeled toxin A for 2 h at 4°C. [12511 toxin A (7.5 ng) was then added and incubated for an additional 2 h at 4°C, and the amount of ['251]toxin A bound was determined after centrifugation. Preliminary investigations demonstrated that a steady state of bound [1251]toxin A was achieved within 30 min of addition of ['251]toxin A. The binding of [1251]toxin A to adult and infant BBMs in the presence of increasing concentrations of unlabeled toxin A is shown in Fig. 4. Scatchard analysis of the data was performed by using the least-squares-fit computer program LIGAND as described by Munson and Rodbard (31) and modified (Elsevier-Biosoft, Cambridge, United Kingdom) for use on a Macintosh computer (12, 43). Figure 5 is a representative Scatchard plot generated by the LIGAND computer program comparing [1251]toxin A binding to infant BBMs with ['251]toxin A binding to adult hamster BBMs. LIGAND statistical analysis of the Scatchard plots of C. diffcile [1251]toxin A binding to infant and adult BBMs indicated that the data were consistent with a single class of receptor sites on both infant and adult hamster BBMs. However, LIGAND did reveal differences in the maximal binding capacities (Bmax) and apparent dissociation constants (Kd) for infant and adult hamster BBMs. Bmax and Kd values were 0.63 pmollmg of protein and 66.7 nM, respectively, for infant BBMs and 0.24 pmol/mg protein and 27 nM, respectively, for adult BBMs. Similar results were obtained if labeled and unlabeled toxins A were added simultaneously to infant and adult BBMs.

0

2

4

6

8

TOXIN A BOUND (nmol)

FIG. 5. Scatchard plot of ['25I]toxin A binding to infant (A) and adult (0) hamster BBMs generated by the LIGAND program.

Effect of heat on toxin binding. Heating [125I]toxin A at either 60 or 100°C for 5 min significantly inhibited its binding to adult and infant BBMs by greater than 90% (Table 2). On the other hand, heating infant or adult BBMs at either 100 or 600C for 5 min inhibited the binding of [1251I]toxin A to infant and adult BBMs less than 20% (Table 2). Blocking of binding with antibody. Affinity-purified rabbit antibodies against C. difficile toxins A and B were prepared and characterized as previously described (17). These antibodies were then examined for their influence on the binding of [1251I]toxin A to adult and infant BBMs. Antibody specific for toxin A inhibited the binding of [1251I]toxin A to BBMs by greater than 80% (Table 2). On the other hand, antibody specific for toxin B inhibited the binding of [1251I]toxin A to hamster BBMs less than 40% for adult BBMs and less than 20% for infant BBMs. To eliminate the possibility that rabbit gamma globulin competed for toxin A binding sites on intestinal membranes, BBMs preincubated with gamma globulin from nonimmunized animals were mixed with [1251] toxin A. No inhibition of [1251]toxin A binding was observed. SDS-PAGE and autoradiography of BBM proteins. BBM proteins were dissolved in SDS, separated by SDS-PAGE, and transferred to nitrocellulose sheets, which were overlaid with [125I]toxin A. Figure 6 shows that the protein composition of the BBMs was different for infant and adult hamsters. Autoradiography demonstrated at least three protein bands in infant and adult hamster BBMs which specifically bind [1251]toxin A. However, there was not any evidence that the BBM proteins which bound [1251I]toxin A differed between infant and adult hamsters. [125I]toxin A binding to the proteins on nitrocellulose could be completely inhibited by prior incubation of the nitrocellulose membrane with unlabeled toxin A (but not unlabeled toxin B) or prior treatment of [125I]toxin A with heat (60 or 100°C for 5 min) or monospecific rabbit antibody directed against toxin A. DISCUSSION There is evidence to support postnatal development of intestinal receptors for C. difficile toxins A and B. It has

1228

INFECT. IMMUN.

ROLFE

TABLE 2. Effect of heat treatment and antibody on the binding of ['251]toxin A to adult and infant BBMs [1251]toxin A specificially bound (% of control)'

[251I]toxin A heat treatment"

BBM

Antibody treatmentd

BBM heat treatment'

Room temp

600C

1000C

Room temp

600C

1000C

Anti-toxA

Anti-toxB

100 100

6 4

6 5

100 100

100 86

81 86

18 7

63 83

Adult Infant

"Numbers represent the mean of three experiments in which BBM preparations from different hamsters were used in each experiment. Standard deviations were less than 10% of the mean. b ['251]toxin A was incubated at room temperature, 60°C, or 1000C for 5 min, cooled to 4°C, and reacted with BBMs. As a control, [1251I]toxin A kept at 4°C was reacted with BBMs (100% binding). ' Infant and adult BBMs were incubated at room temperature, 60°C, or 100°C for 5 min, cooled to 4°C, and then reacted with BBMs. As a control, infant and adult BBMs kept at 4°C were reacted with ['251]toxin A (100% binding). d Mixtures of ['25Iltoxin A and antibody specific for toxin A (anti-toxA) or toxin B (anti-toxB) were preincubated for 1 h at 37°C and then reacted with BBMs. As a control, normal rabbit serum was used instead of toxin-specific antiserum (100% binding).

been reported that fetal intestinal mucosa cells are more resistant in vitro to C. difficile toxins A and B than adult intestinal mucosa cells (5). It is not known if the absence of toxin receptors accounts for the observed differences in susceptibility, but it demonstrates that resistance to toxins A and B may be due to the nature of the intestinal cells and not to factors in the lumen of the intestine. Eglow et al. (11) reported that newborn rabbit ileal BBMs do not possess measurable specific binding sites for labeled toxin A but that A

B

D

C

2 0C0.Q00

-1 it,2O0

-

Binding kinetics of Clostridium difficile toxins A and B to intestinal brush border membranes from infant and adult hamsters.

This study was undertaken to determine if the relative resistance of neonates and infants to Clostridium difficile-associated intestinal disease can b...
2MB Sizes 0 Downloads 0 Views