ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, June 1992, p. ... combination with 10 antipseudomonal antibiotics against intraperitoneal infection was ...
Vol. 36, No. 6
ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, June 1992, p. 1290-1295
Effects of a Human Antiflagellar Monoclonal Antibody in Combination with Antibiotics on Pseudomonas aeruginosa Infection IKUKO UEZUMI,l* MASAZUMI TERASHIMA,l TSUNEO KOHZUKI,l MASUHIRO KATO,' KENJI IRIE,1 HIROSHI OCHI,2 AND HIROSHI NOGUCHI2 Research Laboratories, Sumitomo Pharmaceuticals Co., Ltd., Osaka 554,1 and Biotechnolog Laboratory, Takarazuka Research Center, Sumitomo Chemical Co. Ltd., Takarazuka, Hyogo 665, Japan Received 13 September 1991/Accepted 1 April 1992
The in vivo activity of human immunoglobulin M monoclonal antibody IN-2A8, which is specific for flagellum type b ofPseudomonas aeruginosa, was evaluated in comparison to anti-O antigen (serotype B) MAb KO-2F2 and in combination with antibiotics. IN-2A8 showed stronger activity than KO-2F2 against subcutaneous infection in burned mice, while it was much less active against intraperitoneal infection in normal mice. In a burn infection model, IN-2A8 inhibited the increase of bacteria in skin lesions weakly and that in blood significantly, suggesting that it strongly suppressed bacterial spread to blood. The activity of IN-2A8 in combination with 10 antipseudomonal antibiotics against intraperitoneal infection was examined. Clear additive effect was observed with a combination of either carbapenem or aminoglycoside antibiotics in terms of mouse survival. The administration of an antibiotic, imipenem-cilastatin, simultaneously with or before that of IN-2A8 gave a combined effect, but the reverse order did not. The combination of IN-2A8 with imipenem-cilastatin decreased numbers of viable bacteria in the peritoneal cavity and blood and kept them low for a longer time than did either treatment alone. These results suggest that an antiflageUar monoclonal antibody would be effective against systemic infection in combination with some kinds of antibiotics.
of in vitro activities of two types of human immunoglobulin M (IgM) MAbs (18), IN-2A8, which recognized b-type flagella of P. aeruginosa, and KO-2F2, which was directed against 0 polysaccharide of Homma's serotype B strains. In this study, we first examined the activities of IN-2A8 in a bum infection model to assess the in vivo activity of antiflagellar MAb compared with that of anti-O antigen MAb. Then we investigated the activity of IN-2A8 against an i.p. infection in combination with antibiotics or without antibiotics in order to assess its usefulness against systemic infection.
Pseudomonas aeruginosa is recognized as a serious opportunistic pathogen in compromised hosts. Antibiotic therapy against P. aeruginosa infection has met with only limited success, because the causative strains have become increasingly resistant (4, 13, 20). Meanwhile, the potential of active or passive immunization to enhance protective activity against P. aeruginosa infection has been suggested in several reports (1, 6, 15, 19, 21, 23, 30). The lipopolysaccharide (8) and flagella (10, 16) of P. aeruginosa have been shown to be associated with its virulence and have been studied for their effectiveness as protective antigens (2, 6, 7, 11, 14, 23). Monoclonal antibodies (MAbs) specific for 0 polysaccharides of lipopolysaccharide have been reported to be highly protective (22, 24, 26, 28). However, these antibodies have narrow efficacy spectra, because there are 13 to 17 serotypes of lipopolysaccharide and each MAb is reactive only to the strains of a specific serotype (26). On the other hand, flagella of P. aeruginosa are known to be classified as only two serotypes, designated a and b (3). Passive protection with antiflagellar antisera (9) or MAbs (25) against P. aeruginosa in mice challenged subcutaneously (s.c.) after they were burned has been demonstrated, although no studies compared their activities with those of other types of MAbs in one experiment. This model is suggested to be that of a local infection resulting in systemic sepsis, because the appearance of the bacteria in the blood is the result of rapid growth in the skin lesion. However, there are no reports describing the protective activity of antiflagellar MAbs in an intraperitoneal (i.p.) infection, which is considered to be a model of acute systemic sepsis, as the bacterial growth in blood is as fast as that in the peritoneum. We already reported the establishment and the comparison *
MATERIALS AND METHODS MAbs. The production of human IgM MAbs IN-2A8 and KO-2F2 against P. aeruginosa was described by Ochi et al. (18). IN-2A8 is specifically reactive to b-type flagella, and KO-2F2 is specific for the 0 antigen of Homma's serotype B strains. They were chromatographically purified and used for the experiments. Bacterial strains. P. aeruginosa SP10052 was a clinical isolate of serotype B and was reactive to the two MAbs described. The serotype of the bacterium was determined by a serotype grouping kit Mei assay (Meiji Seika Co., Tokyo, Japan) after the bacterium was cultured on heart infusion agar (Nissui Pharmaceuticals, Tokyo, Japan) at 37°C for 18 h. It was shown to produce exotoxin A, elastase, and protease. It was grown as described above and kept frozen at -80°C in water-40% glycerol. Antibiotics. Antibiotics used were imipenem-cilastatin (IPM-CS) (Tienam; Banyu Pharmaceutical, Tokyo, Japan); SM-7338 (Meropenem), synthesized in the Research Laboratory of Sumitomo Pharmaceuticals, Osaka, Japan; tobramycin (Shionogi Pharmaceuticals, Osaka, Japan); gentamicin (Essex Japan, Osaka, Japan); ceftazidime (Tanabe
Corresponding author. 1290
VOL. 36, 1992
A HUMAN MAb AGAINST FLAGELLA OF P. AERUGINOSA
Pharmaceuticals, Osaka, Japan); cefsulodin (Takeda Chemical, Osaka, Japan); cefpiramide (Sumitomo Pharmaceuticals, Osaka, Japan); aztreonam (Eisai Pharmaceutical, Tokyo, Japan); carmonam (Takeda Chemical); and piperacillin (Toyama Chemical, Tokyo, Japan). Protection experiments with mice. Four-week-old male ICR mice (Japan SLC Inc., Shizuoka, Japan) (body weight, 22 to 24 g) were used in the following experimental infection models. P. aeruginosa SP10052 (serotype B) was cultured on heart infusion agar at 37°C for 18 h. (i) Burn infection model. Burned mice were prepared according to the method of Stieritz and Holder (27), with some modification. After anesthesia, an 8-cm2 burn was created on the shaved backs of mice by pressing an ethanol flame on a glass filter for 10 s. Immediately after burning, 0.3 ml of saline was injected i.p. for fluid replacement, and the bacterial suspension in 0.2 ml of saline was inoculated s.c. at the burned site. One hour later, 0.2 ml of MAb with or without an antibiotic in phosphate-buffered saline (PBS) was intravenously injected. The survival of animals was observed for 7 days. The 50% lethal dose (LD50) of challenged bacteria was calculated by probit analysis. The significant difference was indicated by nonoverlapping 95% confidence intervals. (ii) Systemic-infection model. Bacterial suspension was serially diluted in saline and mixed with bacteriological mucin (Difco Laboratories, Detroit, Mich.). Two hundred microliters of the bacterial suspension containing 5% mucin was injected into the peritoneums of the mice. One hour later, 0.2 ml of MAb with or without an antibiotic in PBS was injected i.p. The negative-control group received neither MAb nor an antibiotic. The observation and the analysis of the results were done in a manner similar to that for the burn infection model. Bacterial counts in blood and tissues. At time intervals after the inoculation of P. aeruginosa, animals were sacrificed. Blood was collected by cardiac puncture, and 1 ml of peritoneal fluid was collected after the puncture was washed with 2 ml of saline containing 0.1% bovine serum albumin. In the case of burned mice, full-thickness samples of burned skin were also removed, weighed, and homogenized in sterile saline with sea sand; the homogenized samples were then centrifuged at 200 x g for 1 min to remove debris, and the supernatant was collected. Numbers of viable bacteria in blood (CFU per milliliter), in the peritoneal cavity (CFU per peritoneal cavity), and in skin (CFU per gram of tissue) were counted by using 10-fold serial dilution plates in duplicate on heart infusion agar. The results were expressed as means ± standard deviations or as each value for four to five animals sacrificed at each time.
TABLE 1. Protective activity of antiflagellar MAb (IN-2A8) and anti-O antigen MAb (KO-2F2) against experimental infection with P. aeruginosa in mice Infection system
Dose (,ug/ mouse)
LD,5( (CFU/ mouse)
s.c. after burn"
IN-2A8 KO-2F2 None
6.5 x 105 2.1 x 104 2.8 x 102
0.1 1 10 0.1 1
1.6 1.5 1.5 2.6 4.6 4.0
activity,' 2,300 1 75 1
x x x x x
4.0 3.8 3.8 6.5 11.5 1 _J
105 105 105 105
a Ratio of for the untreated group. for the treated LD5(0different b Significantly (P < 0.05) between the bracketed groups. ' MAbs were intravenously administered 1 h after s.c. challenge with P. aeruginosa SP10052 to 10 burned mice per group. d MAbs were i.p. administered 1 h after i.p. challenge with P. aeruginosa SP10052 in 5% mucin to 10 normal mice per group.
shown in the previous study (18) was reproduced. Meanwhile, in the case of i.p. infection in normal mice, KO-2F2 was highly protective when administered i.p. at a dose of 0.1 or 1.0 jig per mouse (Table 1). The LD50 for the groups treated with IN-2A8 increased about four times, but not significantly, indicating that MAb IN-2A8 was only weakly protective against this type of infection. Bacterial count in the burn infection model. In order to answer the question of how IN-2A8 exerts protective activity, we examined the numbers of P. aeruginosa cells in various tissues after s.c. infection in burned mice (Fig. 1). After inoculation of 2 x 105 CFU, rapid proliferation of bacteria was observed in the skin lesions near the inoculation sites of untreated mice. Bacteria were detectable in the blood 8 h after infection, while those in the skin numbered to 106 CFU/g of tissue. A rapid increase was observed in the A
4 a 0
O0 _j 4
RESULTS Protection against experimental infection in mice. We compared the protective activity of antiflagellar human IgM MAb IN-2A8 with that of another type of MAb, KO-2F2, which was directed against the 0 polysaccharide of P. aeruginosa serotype B strains, by using a clinical isolate, SP10052, as the challenge bacterium in two experimental infection models with mice. In cases of s.c. infection after burn, the LD50 for the groups treated with 1 ,ug of IN-2A8 or KO-2F2 per mouse was about 2,300 or 75 times higher, respectively, than that for the untreated group (Table 1). The protective activity of IN-2A8 was significantly higher than that of KO-2F2. The high protective activity of IN-2A8 against burn infection
Time after infection (h) FIG. 1. Bacterial cell count in skin (A) and blood (B) of mice treated with MAb IN-2A8 after s.c. infection with P. aeruginosa after the mice were burned. Mice were challenged s.c. with P.
aeruginosa SP10052 (4.0 x 105 CFU per mouse) after being burned,
and IN-2A8 (1 jig per mouse) was administered intravenously 1 h after infection. Groups of four mice were sacrificed at indicated times after infection, and the numbers of viable bacteria in skin and blood were counted. Each mark indicates mean + standard deviation (A) or the value for an individual mouse (B). Symbols: 0, IN-2A8-treated group; *, untreated group. The countable limits of bacteria were 103 CFU/g in skin and 102 CFU/g in blood.
UEZUMI ET AL.
ANTIMICROB. AGENTS CHEMOTHER. _
(A) Challenge doses of 40-50 LD50 3
o°0 IPM/CS 7338
Antibiotic alone IN-2A8 alone
TABLE 2. Protective activity of IN-2A8 in combination with IPM-CS against i.p. infection in micea Treatment
Dose (Ig/ mouse)
IN-2A8 + IPM-CS
1 + 100 1 + 500 100 500 1
9.6 x 107 4.4 x 107 1.1 x 107 2.4 x 107