Involvement of Bradykinin Generation in Intravascular Dissemination of Vibrio vulnificus and Prevention of Invasion by a Bradykinin Antagonist

Abstract

Involvement of bradykinin generation in bacterial invasion was examined by using a gram-negative bacillus, Vibrio vulnificus, which is known to invade the blood circulatory system and cause septicemia. V. vulnificus was injected intraperitoneally (i.p.) into mice with or without bradykinin or a bradykinin (B2 receptor) antagonist. Dissemination of V. vulnificus from peritoneal septic foci to the circulating blood was assessed by counting of viable bacteria in venous blood by use of the colony-forming assay. Intravascular dissemination of V. vulnificus in mice was significantly potentiated by simultaneous injection with bradykinin but was markedly reduced by coadministration with the B2 antagonist d-Arg,[Hyp³, Thi^5,8, d-Phe⁷]-bradykinin. Furthermore, V. vulnificus lethality was significantly increased when bradykinin was administered simultaneously with the bacillus, whereas it was definitely suppressed by treatment with d-Arg,[Hyp³, Thi^5,8, d-Phe⁷]-bradykinin. Similarly, ovomacroglobulin, a potent inhibitor of the V. vulnificus protease, showed a strong suppressive effect on the V. vulnificus septicemia. We also confirmed appreciable bradykinin production in the primary septic foci in the mouse peritoneal cavity after i.p. inoculation with V. vulnificus. It is thus concluded that bradykinin generation in infectious foci is critically involved in facilitation of intravascular dissemination of V. vulnificus.

Septicemia, a serious medical condition in which the host’s immune system overreacts to an infection (4), is one of the most critical consequences of systemic bacterial infection (27). Most frequently, gram-negative bacilli gain access to the blood from extravascular septic foci via the lymphatics or by direct invasion of small blood vessels within a local site of infection (19). Among a number of gram-negative bacilli, Vibrio vulnificus exhibits a strong tropism for blood vessels and often spreads intravascularly, causing septicemia (3).

The pathogenic properties of the extracellular protease and of cytolysin and phospholipase A₂ from V. vulnificus are well documented (5, 12, 20, 30). Accumulating data indicate that extracellular bacterial proteases play an important role in the pathogenesis of various bacterial infections (13, 14, 31). It has been suggested that bacterial proteases may facilitate bacterial invasion of the vascular system through tissue destruction because of their potent proteolytic action against various extracellular matrices (17, 23, 26). However, the mechanism of bacterial invasion of the vasculature and entry into the circulatory system is not fully understood.

Previous studies showed that a number of microbial proteases from pathogenic bacteria and fungi activate the bradykinin-generating cascade, including Hageman factor, prekallikrein, and high-molecular-weight kininogen (8–11, 15, 18, 21). Thus, bradykinin generated by stimulation with these bacterial proteases during the infection acts as a universal mediator in inflammatory reactions, e.g., pain, edema formation, and modulation of vascular tone (13, 14, 31).

In the present experiments, we examined the role of bradykinin in triggering septicemia caused by V. vulnificus. Our results indicate that bradykinin generation in primary septic foci facilitates intravascular dissemination of the bacteria and septicemia. In this context, the effect of a bradykinin antagonist (specific for the B2 receptor) on prevention of vascular bacterial invasion is described here for the first time.

A strain of V. vulnificus (KVV 9207) used throughout these experiments was isolated from a patient who had severe necrotizing fasciitis and septicemia in Kumamoto, Japan, in 1990. V. vulnificus was cultured for 6 h at 37°C in brain heart infusion broth (Difco, Detroit, Mich.) supplemented with 2% NaCl. The bacteria were collected by centrifugation (20,000 × g for 30 min at 4°C) and were washed three times in 0.01 M phosphate-buffered 0.15 M saline (PBS) (pH 7.4). The bacteria suspended in PBS were then injected intraperitoneally (i.p.) into male ddY mice (specific pathogen free, 6 weeks old; Japan SLC, Shizuoka, Japan) to allow intravascular dissemination and to produce a model for lethal V. vulnificus septicemia. To examine the role of bradykinin in the pathogenesis of V. vulnificus septicemia, the male ddY mice were inoculated with V. vulnificus with or without bradykinin (Peptide Institute, Osaka, Japan) at a dose of 20 μg/mouse and a bradykinin (B2 receptor) antagonist, d-Arg,[Hyp³, Thi^5,8, d-Phe⁷]-bradykinin (Sigma Chemical, St. Louis, Mo.) at a dose of 200 μg/mouse. Similarly, ovomacroglobulin (OVM) (Japan Immunoresearch Laboratories, Takasaki, Japan), a potent inhibitor of V. vulnificus, was tested for its effect on the V. vulnificus septicemia in mice.

First, the effects of bradykinin and the bradykinin antagonist on the lethality of the infection were investigated by determining the survival rate of the mice. Second, intravascular dissemination of V. vulnificus was assessed by counting viable bacteria in the blood. Specifically, 1 h after injection of bacteria (10⁷ CFU/mouse) with or without bradykinin or the bradykinin antagonist, a blood sample was taken by cardiac puncture, and the number of viable bacteria in the blood was quantified by use of a colony-forming assay with brain heart infusion agar supplemented with 2% NaCl for selected growth of V. vulnificus.

Intravascular bacterial dissemination as assessed by viable V. vulnificus counts was found to be remarkably enhanced by bradykinin treatment (P = 0.011) (Fig. 1A). Furthermore, treatment with bradykinin at a dose of 20 μg/mouse resulted in a significant decrease in the survival rate of mice inoculated with V. vulnificus at a dose of 10⁶ CFU (P < 0.0001) (Fig. 1B). These results clearly indicate that bradykinin contributes to invasion of the circulatory system by V. vulnificus.

FIG. 1.

Effect of bradykinin (BK) on intravascular dissemination of V. vulnificus in mice (A) and the survival rate of mice given V. vulnificus (B). (A) V. vulnificus (10⁷ CFU/mouse) was injected i.p. into mice with or without bradykinin (20 μg). One hour after injection of the bacteria, the number of viable bacteria in the blood was quantified by use of the colony-forming assay (n, 5 for each group). ∗, P < 0.05 compared with the bradykinin-treated group (two-tailed t test for unpaired data). (B) V. vulnificus (10⁶ CFU/mouse) was administered i.p. to the mice with or without bradykinin (20 μg). A significant difference in survival rate was found for the infected control group (•) and the infected group treated with bradykinin (▪). ∗∗, P < 0.01 (Fisher’s exact probability test; n, 20 for each group). See the text for details.

V. vulnificus was given i.p. to mice with or without the bradykinin antagonist d-Arg,[Hyp³, Thi^5,8, d-Phe⁷]-bradykinin (200 μg/mouse). As shown in Fig. 2A, intravascular dissemination of V. vulnificus was markedly inhibited by treatment with the bradykinin antagonist (P = 0.021). The survival rate of mice injected with V. vulnificus (2 × 10⁶ CFU) was also much improved by administration of d-Arg,[Hyp³, Thi^5,8, d-Phe⁷]-bradykinin (P = 0.025) (Fig. 2B). The experiments described above were repeated twice with 10 mice for each experimental group to examine the effect of bradykinin and its antagonist on the survival rate of mice injected with V. vulnificus. Because very similar results were observed in these two experiments for each agent, the data obtained under the same experimental conditions were combined and analyzed statistically. These results suggest that the bradykinin antagonist leads to the reduction of the lethal effect of V. vulnificus.

FIG. 2.

Effect of a bradykinin antagonist on intravascular dissemination of V. vulnificus in mice (A) and survival rate of V. vulnificus-infected mice (B). (A) V. vulnificus (10⁷ CFU/mouse) was injected i.p. into mice with or without 200 μg of the bradykinin antagonist d-Arg,[Hyp³, Thi^5,8, d-Phe⁷]-bradykinin. The experiments were performed as described in the legend to Fig. 1A. ∗, P < 0.05 (two-tailed t test for unpaired data; n, 5). (B) Mice received i.p. V. vulnificus (2 × 10⁶ CFU/mouse) and simultaneously received i.p. d-Arg,[Hyp³, Thi^5,8, d-Phe⁷]-bradykinin at a dose of 200 μg. A significant difference in survival rate was found for the infected control group (•) and the infected group treated with the bradykinin antagonist (▪). ∗, P < 0.05 (Fisher’s exact probability test; n, 20 for each group). See the text for details.

Generation of bradykinin in the peritoneal cavity was examined by using a highly sensitive enzyme immunoassay for kinins (MARKIT A; Dainippon Pharmaceuticals, Osaka, Japan), as described previously (16), after i.p. administration of the bacteria to mice. One hour after i.p. injection of V. vulnificus (10⁷ CFU/mouse), peritoneal lavage was performed with 5 ml of PBS containing 1 mM EDTA and a mixture of kininase inhibitors captopril (Sankyo, Tokyo, Japan), SQ20881 (Glu-Trp-Pro-Arg-Pro-Glu-Ile-Pro-OH; a gift from Squibb Institute, Princeton, N.J.), and Plummer’s inhibitor (Calbiochem, La Jolla, Calif.) (each at 100 μg/ml). The peritoneal lavage fluid was deproteinized by adding trichloroacetic acid (final concentration, 6%) to the samples, which were centrifuged at 10,000 × g for 10 min. The resulting supernatants were then subjected to quantitation of kinins by the enzyme immunoassay MARKIT A. Both bradykinin and kallidin (Lys-bradykinin), but not des-Arg-bradykinin, are known to show quite similar immunoaffinity to the antikinin antibody used in this assay kit; thus, only the B2 receptor agonists were quantitated in our assay.

Two groups of five mice each were treated with either V. vulnificus (10⁷ CFU) or saline. The amounts of bradykinin generated in the peritoneal cavities were 10.9 ± 0.9 ng in mice treated with V. vulnificus and <0.5 ng in those treated with saline alone. An appreciable amount of kinins was found in the peritoneal cavity, the primary septic focus in the infection with V. vulnificus, by the enzyme immunoassay for bradykinin (detection limit, <0.1 ng/ml). Bradykinin generation was almost completely abrogated by i.p. injection of OVM, which has broad and potent antiprotease activity, together with V. vulnificus (<0.5 ng). Also, it is of considerable importance that V. vulnificus septicemia in mice was appreciably ameliorated by the simultaneous administration with OVM (1.0 mg), as evidenced by significant reductions of both intravenous dissemination of the bacteria (P < 0.0001) and its lethality (P < 0.001) (Fig. 3). In this experiment, the effect of OVM was investigated twice (n = 10 for each group), and very similar results were obtained.

FIG. 3.

Effect of OVM on (A) intravascular dissemination of V. vulnificus in mice and (B) survival rate of V. vulnificus-infected mice. (A) V. vulnificus (10⁷ CFU/mouse) was injected i.p. into mice with or without 1.0 mg of OVM. The experiments were performed as described in the legend to Fig. 1A. ∗∗, P < 0.01 (two-tailed t test for unpaired data; n, 5). (B) Mice received i.p. V. vulnificus (2 × 10⁶ CFU/mouse) and i.p. OVM (1 mg/mouse) simultaneously. A significant difference in survival rate was found for the infected control group (•) and the infected group treated with OVM (▪). ∗∗, P < 0.01 (Fisher’s exact probability test; n, 20 for each group). See the text for details.

In contrast, OVM has no appreciable antibacterial effect in vitro, although it suppresses strongly the metalloprotease excreted by V. vulnificus. Specifically, when V. vulnificus was cultured in a gelatin-glucose medium (2% gelatin, 2% glucose, and 2% NaCl in 20 mM sodium phosphate buffer [pH 7.4]) in the presence or absence of OVM, no significant growth-inhibitory activity was observed with OVM at a concentration of 500 μg/ml (data not shown).

It is thus suggested that kinins generated by bacterial proteases in the peritoneal cavity (21) may exacerbate V. vulnificus septicemia in mice.

In recent years, much attention has been given to the important roles of extracellular bacterial proteases in the pathogenesis of bacterial infection (7–15, 17, 18, 20–26, 28, 29, 31). In particular, we have been studying the pathogenic potential of a series of bacterial proteases, with a focus on their potent stimulation of bradykinin generation (9–11, 13–15, 18, 21, 28, 31).

We previously reported that extracellular proteases from various microbes, including V. vulnificus, activate the bradykinin-generating cascade at several steps in vitro and cause increased vascular permeability in vivo. V. vulnificus protease proteolytically activates Hageman factor and acts directly on high-molecular-weight kininogens, followed by generation of bradykinin (20, 21). Thus, generation of kinins triggered by bacterial proteases in septic foci may be one of the most important pathological reactions in the host infected with pathogenic bacteria.

The results obtained in the present experiments provide evidence that bradykinin generation facilitates movement of bacteria into the vascular system and septicemia. It is of considerable importance that one of the bradykinin B2 receptor antagonists tested showed significant suppression of intravascular V. vulnificus dissemination. This indicates that the kinin-generating cascade may be involved in the mechanism of intravascular bacterial invasion of the systemic blood circulation. Moreover, a novel pharmacological action of bradykinin, i.e., enhancement of invasion by bacterial cells of the host’s tissue and the systemic blood-borne dissemination, now becomes apparent. This effect will have general clinical implications for the pathogenesis of septicemia caused by various bacteria other than V. vulnificus. In fact, we recently reported a similar result suggesting the possible role of bradykinin in promoting intravascular bacterial invasion in Pseudomonas aeruginosa infection in mice (28).

Bradykinin is a principal inflammatory mediator; it causes one of the most fundamental reactions, i.e., fluid accumulation and edema formation resulting from enhanced vascular permeability (1). Thus, increased intravascular dissemination of V. vulnificus may be explained by the potent action of bradykinin in causing the intercellular junctions of endothelium at postcapillary venules to open (1), which gives the bacteria a greater chance to invade the circulatory system.

Of interest also is the previous description that exudative fluid accumulation in inflamed tissues may serve as a nutrient for bacteria and as a vehicle for their dissemination (32). In this regard, enhanced vascular permeability induced by bradykinin might contribute to increased tissue hydrostatic pressure. The hypertensive hydrodynamic pressure in the tissue would push not only bacterial cells but also their toxins into vascular lumens. Bacterial cells and toxins might enter the circulatory system via the lymphatic ducts, inasmuch as lymph flow is increased by bradykinin.

Data indicate that bacterial proteases have tissue-destructive activity in bacterial infections, such as pulmonary and corneal infections (2, 17) and dermal infection of burned sites (8). It was reported previously that some bacterial proteases effectively degrade the extracellular matrix proteins, e.g., elastin, collagen, and fibronectin (13, 23). This result is further substantiated by our recent observation that bacterial exoproteases produced by Vibrio spp. and P. aeruginosa strongly activated matrix metalloproteases, such as progelatinase and procollagenase (26). Furthermore, various bacterial proteases are known to destroy a wide variety of the host’s defense-oriented substances, such as immunoglobulin A (6, 13, 14), complement (25, 29), and a plasma serine protease inhibitor (14, 22, 24). Therefore, the effect of bradykinin on intravascular bacterial dissemination might synergistically contribute to facilitation of septicemia through the action of bacterial proteases.

In conclusion, bradykinin generated in infectious foci is crucial in the pathogenesis of V. vulnificus septicemia in that it facilitates bacterial dissemination and exacerbates septicemia. Septicemia and sepsis syndrome are often encountered as serious complications of opportunistic infections caused by gram-negative bacilli with multiple drug resistance. Therefore, modulation of the host’s derived factors, such as kinins and the endogenous protease cascade, appears to be beneficial in suppressing or preventing such a fatal complication of bacterial infection.