INTRODUCTION

Top Abstract Introduction References

NEONATES with early-onset sepsis, particularly that caused by group B streptococci (GBS), frequently develop persistent pulmonary hypertension of the newborn with accompanying severe hypoxemia (11). Bradykinin (BK) is known to be present in increased amounts during transitional circulation and is believed to play an important role in the early decrease in pulmonary vascular resistance after birth (1, 12, 15). In addition, BK binds to at least two receptors, designated B₁ and B₂, which may be found on endothelium and smooth muscle cells and which, when stimulated in vitro, result in vasoconstriction and dilation, respectively (9). BK binds predominantly to B₂-receptors on the endothelial cells, initiating the release of nitric oxide (NO) and prostacyclin (7), potent vasodilators, which have been shown to modulate pulmonary vascular tone at the time of delivery (3, 13). Under normal conditions, the number of B₁-receptors present on vascular smooth muscle is low, and the response to a specific B₁-receptor agonist is negligible (14). However, investigators have demonstrated an upregulation of B₁-receptor expression in vascular smooth muscle after lipopolysaccharide (LPS) exposure (23), leading to increased in vitro contraction of vascular smooth muscle tissue preparations (5, 18) in response to specific B₁-receptor agonists. Specifically, an increase in in vitro vascular smooth muscle contractile response to des-Arg⁹-BK, an active metabolite of BK (8) and a specific B₁-receptor agonist (14), has demonstrated this increase in receptor expression. During the inflammatory process, there is increased carboxypeptidase N activity that results in increased metabolism of BK to des-Arg⁹-BK (2). The latter has increased affinity for B₁ receptors and is ~10 times more potent than BK (22) at B₁-receptor sites. On the basis of this result, we hypothesized that pulmonary artery (PA) rings exposed to GBS in vitro would have a decreased endothelium-mediated relaxation response to BK secondary to upregulation of BK B₁ receptors on vascular smooth muscle.

The aims of this study were 1) to determine the effect of BK on the in vitro resting tension of piglet PA rings after exposure to GBS, 2) to ascertain whether GBS decreases the in vitro vasodilating effect of BK in preconstricted piglet PAs, and 3) to evaluate the effect of a specific B₁-receptor agonist, des-Arg⁹-BK, on the vasoreactivity of GBS-exposed PA rings.

	METHODS

Yorkshire piglets <7 days old were euthanized with CO₂ inhalation and exsanguinated. The heart-lung block was rapidly removed and placed in ice-cold Krebs solution, which had the following composition (in mM): 118 NaCl, 5 KCl, 2.5 CaCl₂, 1.2 MgSO₄, 1.2 NaH₂PO₄, 2.5 NaHCO₃, 0.03 EDTA, and 11.0 glucose. The intralobar PAs, taken just distal to the fifth branch point in the lower lobes, were dissected free from the surrounding lung parenchyma; cleaned of loose connective tissue, fat, and blood; and cut into rings 4-5 mm in length. Special care was taken not to touch the luminal surface. In some rings, the endothelium was deliberately removed by gently rubbing the luminal surface with a wooden stick. Endothelial cell removal was confirmed by the absence of a relaxation response to acetylcholine (10⁶ M). The rings were then suspended between two stainless steel stirrups, one fixed within the organ chamber and the other attached to a force transducer (Kent Scientific, Litchfield, CT) connected via an analog-to-digital converter to a computer for monitoring of isometric tension (Workbench Software, Strawberry Tree, Sunnyvale, CA; Macintosh Centris 650 Computer, Apple Computers, Cupertino, CA). The organ chambers were filled with 10 ml of Krebs solution, bubbled with 95% O₂-5% CO₂ and maintained at 37°C. Immediately after they were mounted, the PAs were stretched to 1.2 g of tension, the optimal resting tension as determined by a tension-response curve to Krebs solution containing KCl (20 mM). Indomethacin (10⁵ M) was added to all rings before each experiment to inhibit prostacyclin production. The rings were allowed to equilibrate for 60 min before further experimentation. Rings with and without endothelium were studied in parallel except for protocol 1.

GBS was prepared as follows. A culture of group B -hemolytic streptococci (type Ic) in the logarithmic phase of growth were inoculated into 1,200 ml of Todd-Hewitt broth. The culture was incubated for 18 h in a shaking incubator set at 30°C and 100 rpm. The bacteria were collected by centrifugation at 2,600 g for 20 min, and the supernatant was discarded. The bacterial pellet was resuspended in phosphate saline (pH 7.3) and washed twice with this solution. The bacteria were then resuspended in 40 ml of lactated Ringer solution, divided into 10-ml aliquots, and frozen at 70°C until the day of the experiment, at which time the bacterial suspension was thawed at 37°C, washed twice in Krebs solution, and resuspended in 1 ml of Krebs solution, yielding a bacterial concentration of 8.55 × 10¹¹ colony-forming units (CFU)/ml. The following four study protocols were then performed.

Protocol 1: PA dose-response curve to BK. Paired PA rings with endothelium were mounted in organ baths under resting tension as described. One ring of each pair was incubated for 4 h with Krebs solution containing GBS (8.55 × 10⁹ CFU) or Krebs solution alone (Control). After the incubation period, rings were rinsed with Krebs, and incremental doses of BK were added (after the rings reached steady-state contraction or relaxation, usually ~3-6 min) at concentrations of 10⁹ M to 1.2 × 10⁵ M, and the change from resting tension was recorded.

Protocol 2: response of preconstricted PA rings to BK with and without nitro-L-arginine-methyl ester (L-NAME). Paired PA rings (with and without endothelium) were incubated for 4 h at 37°C in Krebs containing GBS (8.55 × 10⁹ CFU) or Krebs solution alone (Control). After a 4-h incubation period, prostaglandin F₂ (PGF₂) (10⁷ M) was added to contract the PA rings. Steady-state contraction was achieved, BK (10⁶ M) was added, and maximum relaxation was recorded. The rings were washed with Krebs solution and allowed to relax to baseline. Rings were again contracted with PGF₂ (10⁷ M), and, at steady-state contraction, L-NAME (10⁵ M) was added to the organ bath and allowed to incubate for 5 min. There was no further contraction of the rings after the addition of L-NAME. Maximal relaxation to the subsequent addition of BK (10⁶ M) was then recorded.

Protocol 3: PA response to des-Arg⁹-BK. Rings were mounted in organ baths as described above. Rings (with and without endothelium) were incubated for 4 h at 37°C in Krebs containing GBS (8.55 × 10⁹ CFU) or Krebs solution alone (Control). Rings were rinsed with Krebs solution and PGF₂ (10⁷ M) was then added to each bath. When the contractile response had reached steady state, des-Arg⁹-BK was added in incremental concentrations from 10⁹ to 10⁶ M. The rings were allowed to reach a steady state (usually ~3-6 min) before each incremental dose of des-Arg⁹-BK was added. The changes in tension from the contractile steady-state tension were recorded at each dose.

Protocol 4: PA response to des-Arg⁹-BK after incubation with des-Arg⁹-[Leu⁸]-BK. Rings were prepared as in protocol 3. After a 4-h incubation period, rings were rinsed with Krebs solution, and PGF₂ (10⁷ M) was then added to each bath. At steady-state contraction, the B₁-receptor antagonist des-Arg⁹-[Leu⁸]-BK (10⁶ M) was added to each organ bath and allowed to incubate for 5 min before the experiment was continued. des-Arg⁹-[Leu⁸]-BK is a competitive B₁ antagonist. Because the affinity of des-Arg⁹-[Leu⁸]-BK for the B₁ receptor equals that of des-Arg⁹-BK, we chose a concentration equal to the maximal response observed for des-Arg⁹-BK. The des-Arg⁹-BK was added at incremental concentrations of 10⁹ M to 10⁶ M, and the change in tension from contractile steady state was recorded at each dose.

Reagents. PGF₂, BK, des-Arg⁹-BK, des-Arg⁹-[Leu⁸]-BK, indomethacin, NaCl, KCl, CaCl₂, MgCl₂, NaH₂PO₄, NaHCO₂, EDTA, and glucose were obtained from Sigma Chemical (St. Louis, MO). L-NAME was obtained from Bachem (Torrance, CA).

Statistics. The data are presented as means ± SE. Statistical analysis was performed by using the SYSTAT (Evanston, IL) software package version 5.0. The four data sets were analyzed by both factorial and repeated-measures ANOVA models. The repeated-measures factor was level of dose, and the groups were treatment/Control and endothelium present/absent.

	RESULTS

Protocol 1: PA dose-response curve to BK. BK caused a dose-dependent response in both the Control and GBS-exposed PA rings with endothelium (n = 7 sets of rings from 7 animals) (Fig. 1). Control rings displayed a decrease in resting tension (1.2 g) with a maximal decrease of 0.60 ± 0.19 g at 10⁶ M BK, whereas the GBS ring had an increase in tension of 0.10 ± 0.13 g at 10⁶ M BK. The difference between Control and GBS-treated rings was statistically significant (P < 0.001). There was a significant difference in response to increasing doses of BK in the GBS-treated rings compared with the Control rings (P < 0.001).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Protocol 1. Dose-response curve to bradykinin (BK). Pulmonary artery (PA) rings were adjusted to a resting tension of 1.2 g, and then the recorder was tared to 0 g. Values are means () ± SE; n, no. of animals. In rings with endothelium, there was a dose-dependent response in both Control and group B streptococci (GBS)-treated PA rings to BK. GBS overall significantly attenuated the endothelial-mediated relaxing effect of BK (P < 0.001) compared with Control group and caused a contractile response to increasing doses.

Protocol 2: relaxation to BK. Addition of BK resulted in relaxation in both the GBS-treated and Control PA rings with intact endothelium (n = 4 sets of rings from 4 animals; Fig. 2). The vasodilator response to BK (10⁶ M) was significantly reduced in PA rings exposed to GBS compared with Controls. Relaxation in the Control group was 115 ± 11% compared with 62 ± 12% in the GBS-treated group (P < 0.05). Addition of L-NAME (10⁵ M) reduced relaxation in the Control group to 60 ± 22% and in the GBS-exposed group to 18 ± 26% (P < 0.05) at the same BK concentration. There was no difference in the degree of inhibition of relaxation by L-NAME in both the Control and GBS-treated groups.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Protocol 2. Relaxation to BK. BK produced relaxation in both GBS-treated and Control PA rings. Values are means ± SE; n, no. of animals. Vasodilator response to BK was significantly reduced in PA rings with endothelium treated with GBS compared with Control rings (P < 0.05). NG-nitro-L-arginine methyl ester (L-NAME) significantly reduced relaxation in GBS-treated and Control groups to the same degree (P < 0.05). Control and GBS-treated rings without endothelium were not significantly changed from steady state by BK administration.

6

Protocol 3: PA response to des-Arg⁹-BK. The mean contractions to PGF₂ at 10⁷ M before the addition of des-Arg⁹-BK were not significantly different between GBS and Control groups.

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Protocol 3. Dose-response curve to des-Arg9-BK. There was a dose-dependent increase in contraction to des-Arg9-BK in GBS-treated PA rings with (E+) and without (E) endothelium (P < 0.0001). Values are means ± SE (bars); n, no. of animals. GBS-treated rings, both E+ and E, had significantly more tension than did Controls (P < 0.0001).

Protocol 4: PA response to des-Arg⁹-BK after incubation with des-Arg⁹-[Leu⁸]-BK. When des-Arg⁹-[Leu⁸]-BK (10⁶ M) was added to the Control and GBS-exposed rings before addition of des-Arg⁹-BK, there was significant inhibition of the contractile response of des-Arg⁹-BK (P < 0.05; n = 6 sets of rings from 6 animals; Fig. 4). Relaxation was noted in all rings at lower doses of des-Arg⁹-BK and only in Control rings with endothelium at higher doses.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Protocol 4. Dose-response curve to des-Arg9-BK after addition of des-Arg9 [Leu8]-BK. Values are means ± SE; n, no. of animals. des-Arg9 [Leu8]-BK (106 M) when added to Control and GBS-treated rings before addition of des-Arg9-BK, significantly attenuated contractile response of des-Arg9-BK (P < 0.001).

	DISCUSSION

The results of this study indicate that GBS increases the in vitro contraction of PA rings to BK and the BK metabolite des-Arg⁹-BK in a neonatal animal model. This effect may play an important role in the development of pulmonary hypertension in response to GBS sepsis or pneumonia. Because BK is increased during transitional circulation (1, 12, 15) and is believed to contribute to the fall in pulmonary vascular resistance occurring after birth, any change in its metabolism or receptor expression may alter the natural course of transitional circulation. Our data support the hypothesis that GBS increases the pulmonary vascular response to BK in part via upregulation of B₁-receptors. Furthermore, these data show that L-NAME, a specific inhibitor of endothelial NO synthesis, reduced relaxation secondary to BK both in the GBS and Control rings with endothelium to the same degree; this suggests that GBS does not affect BK-induced release of NO.

In addition, there was a BK-induced contractile response in both the GBS-exposed and Control rings without endothelium in the presence of L-NAME. The contractile response in the GBS-exposed rings without endothelium was not statistically different from the response of Controls. We postulate that L-NAME inhibits the basal release of NO by smooth muscle cells, thereby altering the balance of constricting and relaxing mediators to effect a net contraction. The small amount of relaxation to BK in the GBS-exposed and Control rings without endothelium before the addition of L-NAME may be due to basal release of NO from smooth muscle cells or to incomplete removal of the endothelium. However, the latter is unlikely in view of the significant decrease in relaxation compared with rings in which the endothelium was left intact and, additionally, the presence of a contractile response to acetylcholine (10⁶ M).

The results of the BK dose-response portion of this study demonstrate that BK increased resting tension in GBS-exposed vascular rings while decreasing that of Control rings. This suggests that BK has a direct effect on smooth muscle cells to cause contraction of the pulmonary vascular rings, because protocol 2 showed that L-NAME inhibited relaxation in both the GBS and Control rings to the same degree. This contraction is most likely due to stimulation of B₁ receptors on vascular smooth muscle cells. Furthermore, there was no difference in contraction between GBS-treated and Control rings in response to PGF₂, indicating that there was no change in spontaneous tone after GBS exposure.

PA rings at resting tension and exposed to GBS showed a contractile response to increasing doses of BK, whereas those rings that were preconstricted with PGF₂ relaxed in response to BK, although significantly less than Control. The reason for the difference in response between resting conditions and contractile conditions is not known, but one possible explanation is that there could be an interaction between PGF₂ and BK-receptors, causing a decrease in PGF₂ response in addition to a minimal NO response. Rings at higher tensions may also have a greater response to low levels of NO than do rings at lower tensions. These findings warrant further investigation.

Under normal conditions, B₁ stimulation does not result in contraction, due to the overriding B₂ response and the relatively small numbers of B₁ receptors. However, Regoli et al. (23) have shown that LPS induces the formation of B₁ receptors in rabbit aorta strips. In addition, deBlois et al. (5) demonstrated an increased response to des-Arg⁹-BK in rabbit aortic strips exposed to LPS. In our study, there was a dose-dependent contractile response in the GBS-treated PA rings to des-Arg⁹-BK similar to that noted by deBlois et al., suggesting that GBS induces an upregulation of B₁ receptors in the piglet PA.

In our study, addition of increasing doses of des-Arg⁹-BK to PAs demonstrated a dose-dependent increase in contraction above the steady-state PGF₂ contraction in both the GBS-exposed and Control rings. However, GBS-exposed rings were significantly more constricted than were the Control rings, supporting the hypothesis that GBS induces an upregulation of B₁ receptors. Further support implicating B₁-receptor involvement in the increased contractile response to des-Arg⁹-BK was the significant reduction in contraction to des-Arg⁹-BK at all doses after addition of the specific B₁-receptor antagonist des-Arg⁹-[Leu⁸]-BK. The relaxation at low doses of des-Arg⁹-BK in both the GBS-exposed and Control rings may be secondary to basal release of NO from the endothelium and smooth muscle cells. At higher doses of des-Arg⁹-BK, there is contraction in the rings exposed to GBS and in Control rings without endothelium. This may be due to a concentration-dependent incomplete block of the competitive B₁-receptor antagonist des-Arg⁹-[Leu⁸]-BK at the concentration used in the presence of an increased dose of des-Arg⁹-BK.

We postulate that during sepsis there is an increased release of des-Arg⁹-BK in the circulation, particularly the pulmonary circulation, because carboxypeptidase N, a plasma enzyme that has been shown to be elevated in inflammatory exudates (2, 20), and carboxypeptidase M, an enzyme found in various pulmonary cell types, including endothelial cells (17), metabolize BK to des-Arg⁹-BK. The increased generation of BK from circulating blood kininogen (21) during inflammatory reactions, in combination with its conversion to des-Arg⁹-BK and upregulation of B₁ receptors, may lead to increased pulmonary vascular resistance secondary to pulmonary arterial contraction. In addition, des-Arg⁹-BK has increased affinity for B₁ receptors, which is ~10 times greater than that of BK (14). This further supports the possible involvement of des-Arg⁹-BK in the production and maintenance of pulmonary hypertension.

Expression of B₁ receptors has been linked to the cytokine network, especially interleukin-1 (IL-1) (4, 6), which is released from human umbilical vein endothelial cells when exposed to LPS (19). This suggests that pathological conditions associated with IL-1 production (such as sepsis) may result in increased numbers of B₁ receptors. Although we did not expose our rings to IL-1, it is known that this cytokine is elevated in patients with gram-negative bacteremia (10, 27), neonatal piglets exposed to GBS (26), and neonates with perinatal complications (16), suggesting another mechanism for GBS-induced upregulation of B₁ receptors.

In summary, exposure of PA rings of the neonatal piglet to GBS decreases the vasodilator response to BK secondary to upregulation of B₁ receptors. We speculate, therefore, that the changes in BK metabolism in concert with upregulation of B₁-receptor expression during GBS exposure may play an important role in the etiology of persistent pulmonary hypertension of the newborn during GBS sepsis or pneumonia.