Neurokinin B induces oedema formation in mouse lung via tachykinin receptor-independent mechanisms

Abstract

The tachykinin neurokinin B (NKB) has been implicated in the hypertension that characterises pre-eclampsia, a condition where tissue oedema is also observed. The ability of NKB, administered intradermally or intravenously, to induce oedema formation (assessed as plasma extravasation) was examined by extravascular accumulation of intravenously injected ¹²⁵I-albumin in wild-type and tachykinin NK₁ receptor knockout mice. Intradermal NKB (30-300 pmol) caused dose-dependent plasma extravasation in wild-type (P < 0.05) but not NK₁ knockout mice, indicating an essential role for the NK₁ receptor in mediating NKB-induced skin oedema. Intravenous administration of NKB to wild-type mice produced plasma extravasation in skin, uterus, liver (P < 0.05) and particularly in the lung (P < 0.01). Surprisingly, the same doses of NKB led to plasma extravasation in the lung and liver of NK₁ knockout mice. By comparison, the tachykinin substance P induced only minimal plasma extravasation in the lungs of wild-type mice. The plasma extravasation produced by NKB in the lungs of NK₁ receptor knockout mice was unaffected by treatment with the NK₂ receptor antagonist SR48968 (3 mg kg⁻¹), by the NK₃ receptor antagonists SR142801 (3 mg kg⁻¹) and SB-222200 (5 mg kg⁻¹) or by the cyclo-oxygenase (COX) inhibitor indomethacin (20 mg kg⁻¹). L-Nitro-arginine methyl ester (15 mg kg⁻¹), an inhibitor of endothelial nitric oxide synthase (eNOS), produced only a partial inhibition. We conclude that NKB is a potent stimulator of plasma extravasation through two distinct pathways: via activation of NK₁ receptors, and via a novel neurokinin receptor-independent pathway specific to NKB that operates in the mouse lung. These findings are in keeping with a role for NKB in mediating plasma extravasation in diseases such as pre-eclampsia.

NKB is a decapeptide of the tachykinin family, a group of neuropeptides including substance P and neurokinin A which share a common carboxy terminal pentapeptide sequence (Kangawa et al. 1983). They are primarily synthesised within neurons, and so the mammalian tachykinins are commonly known as neurokinins. Substance P was the first to be discovered and is now a well-established pro-inflammatory neuropeptide, localised to sensory nerves throughout the body. It is a potent mediator of increased microvascular permeability, leading to plasma extravasation and tissue oedema formation, through activation of NK₁ receptors located on post-capillary venule endothelial cells in a variety of tissues, including lung and skin (Lembeck et al. 1992; Emonds-Alt et al. 1993). There is also strong evidence that substance P is involved in nociceptive nerve pathways (Cao et al. 1998; De Felipe et al. 1998), as well as in mediating nausea and anxiety, again through NK₁ receptor activation (Saria, 1999). Until recently, NKB was considered to be restricted to the CNS, and an extensive search in a variety of peripheral tissues in the rat (e.g. heart, lung, stomach, skin, colon, eye, liver, although not uterus) failed to find NKB (Moussaoui et al. 1992). The role of NKB in the CNS remains unclear, although it has been suggested to play a role in anxiety (Ribeiro et al. 1999) and sensory transmission (Zerari et al. 1997).

Three neurokinin receptors, the NK₁ receptor, the NK₂ receptor and the NK₃ receptor, have been identified by molecular cloning and sequence analysis. They are members of the seven transmembrane-domain rhodopsin-like super family (see Maggi, 1995b for review). All three neurokinins have high affinity for, and full agonist activity on, the three receptor types. The receptors can be distinguished by the rank order of potency of the neurokinins (e.g. NKB is more potent than NKA or SP at the NK₃ receptor). Both peptide and non-peptide antagonists at all three receptors have been produced. Interestingly, although NKB is expressed almost exclusively in the CNS, NK₃ receptors have also been identified in the vasculature of several mammalian species, and their activation leads to increased heart rate in the dog (Thompson et al. 1998), contraction of the rat hepatic portal vein (Mastrangelo et al. 1987) and constriction of the mesenteric venous beds in the rat (D'Orleans-Juste et al. 1991).

In a recent paper, Page and co-workers (2000) identified the presence of neurokinin B (NKB) mRNA in the human placenta, the first time NKB had ever been identified outside the brain and spinal cord. NKB has also been observed in the rat uterus, where its levels increase with age (Cintado et al. 2001). The presence of NKB in the human placenta is particularly interesting as it was found to be elevated in women suffering from pre-eclampsia, with plasma NKB levels correlating well with blood pressure. Pre-eclampsia is the primary cause of maternal mortality and morbidity during pregnancy. Its defining sign is hypertension coupled to proteinuria, but a variety of additional, potentially life threatening, symptoms can appear suddenly and without warning over days or weeks. These include cerebral and peripheral oedema (Cunningham & Twickler, 2000), renal failure, pulmonary oedema (Davison, 1997), liver capsule distension and nausea (Martin et al. 1999). The sheer variety of symptoms associated with pre-eclampsia have made efforts to determine the underlying cause particularly difficult. The one factor which links all cases of pre-eclampsia is the presence of a placenta, and the evidence for placental involvement is strong.

Page and colleagues found that plasma concentrations of NKB were elevated in a group of eight women affected by pre-eclampsia, compared to levels in normotensive pregnant women (Page et al. 2000). The authors suggest that the hypertension observed in pre-eclampsia is caused by elevated secretion of NKB from the syncytiotrophoblasts of the placenta in response to defective trophoblastic invasion. Page et al. (2000) suggest that one mechanism by which NKB causes the hypertension is through activation of the peripheral NK₃ receptors. As described above, this would cause increased heart rate (Thompson et al. 1998), contraction of the hepatic portal vein (Mastrangelo et al. 1987) and constriction of the mesenteric venous beds (D'Orleans-Juste et al. 1991). Together, this combination of increased heart rate and vasoconstriction could lead to hypertension. Another major pathology associated with pre-eclampsia is the formation of oedema, particularly in the lungs, brain and extremities. We have recently confirmed, using NK₁ receptor knockout mice, that substance P acts purely via NK₁ receptors to mediate increased microvascular permeability and inflammatory oedema formation in skin (Cao et al. 1999; Grant et al. 2002). Earlier work by both this and other groups indicates that neurokinin B is also a potent mediator of oedema formation, although the nature of the receptors involved has not been fully investigated (Brain & Williams, 1989; Campos & Calixto, 2000). Thus we hypothesised that neurokinin B may play a role in the oedema formation, in addition to the hypertension, observed in pre-eclampsia. We have established techniques to assess plasma extravasation in a murine model, and have adapted these to measure NKB-induced oedema. The mechanisms mediating the effects of NKB were then investigated using both selective antagonists to the neurokinin receptors, and the NK₁ receptor knockout mouse.

METHODS

Animals

Wild-type (+/+) and NK₁ receptor knockout (−/−) Sv129+C57BL/6 mice were obtained from Perlmutter Laboratory (Children's Hospital, Boston, MA, USA) then bred in-house. Female mice (22-27 g) were used in this study. All were maintained on normal diet, with free access to food and water, in a climatically controlled environment. Both strains displayed normal growth and behavioural characteristics. All experiments were carried out in accordance with the Animals (Scientific Procedures) Act 1986. The mice were anaesthetised by intraperitoneal (i.p.) injection of urethane (7 μg g⁻¹).

Measurement of oedema formation in the skin after intradermal NKB injection

The dorsal skin was shaved, and the injection pattern marked out. Oedema formation was measured by the extravascular accumulation of intravenously injected ¹²⁵I-labelled bovine serum albumin (BSA). ¹²⁵I-BSA (45 kBq) was injected (i.v.) with saline (0.1 ml) into the tail vein and then flushed through with 0.1 ml saline. After 5 min, NKB (10-300 pmol) or Tyrode solution (vehicle control) was injected intradermally (i.d.), with one uninjected site to allow a comparison of injected sites with naïve skin. Oedema development was allowed for a 30 min period. A blood sample (0.2-0.6 ml) was then taken by cardiac puncture, and the animal killed by cervical dislocation. The blood samples were centrifuged at 6000 g for 4 min, after which plasma was taken for measurement of plasma radioactivity in a gamma counter. The dorsal skin was removed and the injection sites punched out using a 14 mm punch. The sites were weighed, and their radioactivity measured. Skin oedema was expressed as microlitres of plasma per gram of tissue, after subtraction of the value for uninjected skin to control for the basal level of radioactivity.

Measurement of oedema formation after systemic NKB administration

¹²⁵I-BSA (45 kBq), and either NKB (1-30 nmol), substance P (3-30 nmol) or vehicle were injected with saline (i.v.; 0.1 ml final volume) into the tail vein and then flushed through with 0.1 ml saline. Oedema development was allowed for a 30 min period. A blood sample (0.2-0.6 ml) was then taken by cardiac puncture, and the animal killed by cervical dislocation. The blood samples were centrifuged at 6000 g for 4 min, after which plasma was taken for measurement of plasma radioactivity in a gamma counter. An area of dorsal skin was removed and a 14 mm diameter site punched out. Pieces of lung, liver and uterus were also taken. The samples were weighed, and their radioactivity measured, as above. Oedema was expressed as microlitres of plasma per gram of tissue.

The NK₂ receptor antagonist SR48968 (3 mg kg⁻¹), the NK₃ receptor antagonist SR142801 (3 mg kg⁻¹) or the NK₃ receptor antagonist SB-222200 (5 mg kg⁻¹) were administered in solution (100 μl total volume i.v.) into the tail vein. The doses chosen were of the same order of magnitude as those used in previous studies (e.g. Inoue et al. 1996; Sarau et al. 2000). After 15 min ¹²⁵I-BSA (45 kBq) and NKB (10 nmol) were administered (0.1 ml total volume) into the tail vein, and oedema measurements carried out as above.

The nitric oxide synthase (NOS) inhibitor l-NAME (15 mg kg⁻¹; 10 min prior to NKB; Hu et al. 1997) or the cyclo-oxygenase (COX) inhibitor indomethacin (20 mg kg⁻¹; 30 min prior to NKB; Fujii et al. 1995) were administered in solution (i.v.; 100 μl total volume) into the tail vein. After an incubation period, ¹²⁵I-BSA (45 kBq) and NKB (10 nmol) were administered (0.1 ml total volume) into the tail vein, and oedema measurements carried out as above.

Contribution of intravascular volume to apparent plasma extravasation

Wild-type mice were used in this protocol. NKB (10 nmol) was injected (0.1 ml final volume) into the tail vein, and oedema allowed to develop for 30 min. ¹²⁵I-BSA (45 kBq) was then injected i.v. into the tail vein. A blood sample was taken immediately by cardiac puncture, and the animal killed by cervical dislocation. The organs were collected and the volume of plasma remaining within the vasculature assessed in the same way as for the measurement of plasma extravasation.

Materials

¹²⁵I-bovine serum albumin was purchased from ICN (Basingstoke, UK), BSA, urethane, l-NAME, indomethacin and NKB were obtained from Sigma Chemicals. BSA, urethane and l-NAME were dissolved in saline. Indomethacin was dissolved in 5 % NaHCO₃ solution. NKB was dissolved in a minimum volume of ultrapure water, then diluted with 0.01 % BSA solution. SR142801 and SR48968 were gifts from Dr X. Emonds-Alt, Sanofi, Toulouse, France. SB-222200 was a gift from Dr A. Medhurst, GlaxoSmithKline, Harlow, UK. They were dissolved in a minimum amount of ethanol, then made up to the final volume with saline.

Analysis of data and statistical analysis

All oedema results were expressed as microlitres of plasma per gram of tissue. Statistical analysis of dose-response data was by ANOVA followed by Dunnett's multiple comparisons test. Comparisons between treated and untreated mice were carried out by Student's unpaired t tests. Results were all expressed as means ± s.e.m.

RESULTS

Oedema formation induced by neurokinin B, assessed as plasma extravasation, was studied under two different conditions. Initially, in keeping with our previous studies, the dose-dependent ability of NKB to mediate oedema formation in skin was investigated after its intradermal injection. We then went on to administer neurokinin B, via an intravenous route, in order to more closely mimic the scenario of raised plasma levels circulating in pre-eclampsia.

The ability of NKB to produce peripheral oedema after i.d. administration was assessed in both wild-type and NK₁ receptor knockout mice, in order to determine whether NKB could induce plasma extravasation in this strain of mice, and whether the extravasation is dependent on NK₁ receptor activation. Intradermal injection of NKB (10-300 pmol) caused significant oedema at doses of 30 pmol and above in wild-type mice. The same doses failed to produce any response when injected intradermally into the knockout mice (Fig. 1) in keeping with results obtained in the same strain of mouse using substance P (Cao et al. 1999).

Figure 1. Plasma extravasation to intradermal neurokinin B.

Effect of neurokinin B (NKB; 10–300 pmol i.d.) on plasma extravasation in the skin of wild-type and NK₁ receptor knockout mice, measured after 30 min. Results are expressed as plasma extravasation, mean ± s.e.m., n = 8 for wild-type, n = 5 for knockout. *P < 0.05, **P < 0.01 compared to Tyrode-treated control values.

The ability of NKB to mediate oedema formation when present in plasma after i.v. administration was then studied. It is believed that the oedema formation, both peripheral and systemic, produced by tachykinin administration is primarily mediated by the NK₁ receptor, located on the endothelial cells of post-capillary venules. It was hypothesised that intravenous injection of NKB into NK₁ receptor knockout mice would produce much less oedema than in the wild-type mice, or that the oedema would be entirely abolished. Surprisingly, injection of NKB (10-30 nmol, i.v.) produced significant plasma extravasation in the lungs of both wild-type and NK₁ receptor knockout mice, as shown in Fig. 2. NKB at these doses was also able to induce plasma extravasation in the livers of both wild-type and knockout mice (see Table 1). In contrast to the lung and liver, NKB was only able to induce plasma extravasation in the skin and uterus of wild-type animals and had no effect on these tissues in NK₁ receptor knockout animals (Table 1). However, uterine plasma extravasation was only apparent in the wild-type mice at 3 nmol NKB, and the response did not appear to be dose-dependent.

Figure 2. Tachykinin-induced plasma extravasation in the lung.

Effect of neurokinin B (1-30 nmol, i.v.) or substance P (SP; 3–30 nmol, i.v.) on plasma extravasation in the lungs of wild-type and NK₁ receptor knockout mice, measured after 30 min. Results are expressed as plasma extravasation, mean ± s.e.m., n = 10. *P < 0.05, **P < 0.01 compared to respective Tyrode-treated control values.

Table 1.

Neurokinin B-induced plasma extravasation

	Skin		Liver		Uterus
Dose (nmol)	WT	KO	WT	KO	WT	KO
0	23.2 ± 1.2	30.5 ± 3.7	176.6 ± 22.0	228.4 ± 26.0	126.2 ± 23.0	154.3 ± 31.7
NKB (1)	32.3 ± 6.7	28.6 ± 4.6	241.5 ± 44.8	286.6 ± 40.1	99.9 ± 19.5	90.5 ± 12.5
NKB (3)	122.5 ± 11.0**	36.4 ± 4.4†††	544.6 ± 71.1*	242.2 ± 23.6	373.4 ± 47.9**	132.2 ± 24.1†
NKB (10)	158.6 ± 18.8**	43.24 ± 6.6††	434.8 ± 36.9	366.1 ± 36.1*	290.1 ± 66.5	142.4 ± 20.7
NKB (30)	106.4 ± 21.6*	43.7 ± 3.0††	691.2 ± 192*	492.9 ± 54.3**	276.0 ± 82.2	144.0 ± 18.0

Effect of neurokinin B (1−30 nmol, I.V.)on accumulation of ¹²⁵I-BSA radioactivity in the skin, liver and uterus of wild-type (WT)and NK₁ receptor knockout (KO) mice, measured after 30 min. Results are expressed as plasma extravasation, μlg⁻¹, mean ± S.E.M., n =10.

^*P < 0.05

^**P < 0.01 compared to respective Tyrode-treated control values

^†P < 0.05

^††P < 0.01

^†††P < 0.001 compared to NKB-treated wild-type mice.

The effect of substance P at equimolar doses to NKB was then examined to determine whether it could also induce plasma extravasation. It was hypothesised that as they both interact with the same set of receptors, they would both possess the ability to stimulate oedema formation. However, it was apparent that substance P (3-30 nmol i.v.) was much less potent than NKB at stimulating plasma extravasation in the lung, with barely a 2-fold increase (P < 0.01) over control at a dose of 30 nmol, compared to a 16-fold increase with NKB (see Fig. 2). This lack of potency was also seen in the liver, where substance P (3-30 nmol i.v.) failed to induce significant plasma extravasation (data not shown). Again, a difference was seen between the effects on the skin and uterus, and those on the lung and liver. Substance P (SP; 10 nmol) was able to induce plasma extravasation in the skin (SP: 81.1 ± 6.0 vs. control: 36.4 ± 2.8 μl g⁻¹, P < 0.01) and uterus (SP: 321.7 ± 48.3 vs. control: 163.1 ± 18.3 μl g⁻¹, P < 0.01) of wild-type mice.

It has been shown that activation of systemic NK₁ receptors has a potent vasodilator activity (Maggi, 1995a). It is possible that, in this model, injection of NKB produces a significant degree of vasodilatation through NK₁ receptor activation. This would lead to a greater amount of blood being retained within the microvasculature, with an associated increase in radioactivity. In turn, this would lead to an artificially high assessment of the plasma extravasation that had occurred. In order to assess the contribution of intravascular plasma to the apparent plasma extravasation, a separate series of experiments were carried out in which ¹²⁵I-albumin was injected 30 min after the NKB. A sample of plasma was taken immediately by cardiac puncture, and the animal killed, so the labelled BSA was distributed throughout the vasculature, but it had not had time to enter the tissues. The difference between the apparent blood volume measured by this protocol and that measured when ¹²⁵I-albumin was injected along with NKB at 0 min was assessed as the true value from plasma extravasation. Figure 3 shows the apparent plasma extravasation to NKB (10 nmol i.v.) after ¹²⁵I-albumin injection at 0 and 30 min. In all four tissues examined there was a significant difference between the two values, indicating that NKB does produce pronounced plasma extravasation, in addition to any vasodilatation.

Figure 3. Contribution of intravascular plasma to the apparent plasma extravasation.

Comparison of the radioactive plasma volume within the tissue after ¹²⁵I-BSA injection at 0 (¹²⁵I-BSA distributed throughout exudate and organ vasculature) and 30 min (¹²⁵I-BSA distributed throughout organ vasculature alone) after NKB (10 nmol i.v.) injection in wild-type mice. Results are expressed as plasma volume, mean ± s.e.m., n = 7. **P < 0.01, ***P < 0.001 compared to BSA injection at 30 min.

The unexpected observation that NKB induces plasma extravasation in the lungs and livers of both wild-type and NK₁ receptor knockout mice suggests that NK₁ receptors are not the primary receptors which mediate oedema formation to NKB in these tissues. It is known that NKB is also a full agonist at the NK₂ and NK₃ receptors (Nakanishi, 1991; Maggi, 1995b), both of which may be able to produce the observed oedema. To determine whether an NK₂ or NK₃ component to the oedema was present, the NK₂ antagonist SR48968 (3 mg kg⁻¹) or the NK₃ antagonist SR142801 (3 mg kg⁻¹) were administered to NK₁ knockout mice (removing any possibility of an NK₁-component), along with NKB (10 nmol), and the effect on plasma extravasation in the lung was measured. These data are shown in Fig. 4. The doses were chosen as they have been previously shown to inhibit oedema formation in the mouse (Inoue et al. 1996). Neither antagonist had any effect on the plasma extravasation, suggesting that it is mediated through an unknown neurokinin receptor or an alternative system. There is little evidence for non-tachykinin receptor-mediated effects of the tachykinins, so this result was surprising. The effect of a second selective NK₃ receptor antagonist (SB-222200; 5 mg kg⁻¹; as described by Sarau et al. 2000) was also examined, in order to provide further evidence that the pulmonary plasma extravasation was not mediated through the NK₃ receptor. NKB (10 nmol) produced a 5-fold plasma extravasation increase over the control value in both the presence and absence of SB-222200, demonstrating that this antagonist was also unable to reduce plasma extravasation in the lung induced by NKB (10 nmol NKB: 1670.7 ± 207.1 vs. NKB + SB-222200: 1461.1 ± 134.9 μl g⁻¹, n = 7, n.s.).

Figure 4. Lack of effect of tachykinin antagonists.

Effects of NK₂ (SR48968, 3 mg kg⁻¹) and NK₃ (SR142801, 3 mg kg⁻¹) antagonists on plasma extravasation to neurokinin B (10 nmol, i.v.) in the lung of NK₁ receptor knockout mice, measured after 30 min. Results are expressed as plasma extravasation, mean ± s.e.m., n = 5–10. **P < 0.01 compared to respective Tyrode-treated control values.

Two mediators which are known to have potent vasoactive properties and may be involved in the plasma extravasation to NKB are NO and the prostaglandins. To test whether these compounds contribute to the observed oedema formation, NK₁ receptor knockout mice were pre-treated with the NO synthase inhibitor l-NAME (15 mg kg⁻¹) or the cyclo-oxygenase inhibitor indomethacin (20 mg kg⁻¹) prior to injection of NKB (10 nmol). The results are shown in Fig. 5. l-NAME significantly, but only partially, inhibited the plasma extravasation, but indomethacin had no effect.

Figure 5. Effect of l-NAME and indomethacin.

Effects of l-NAME (15 mg kg⁻¹) and indomethacin (INDO; 20 mg kg⁻¹) on plasma extravasation to neurokinin B (10 nmol, i.v.) in the lung of NK₁ receptor knockout mice, measured after 30 min. Results are expressed as plasma extravasation, mean ± s.e.m., n = 8–10. **P < 0.01 compared to respective Tyrode-treated control values; † P < 0.05 compared to NKB treatment alone.

Discussion

The results demonstrate that NKB is a potent mediator of inflammatory oedema formation but that two different mechanisms appear to be present, in that the plasma extravasation observed in response to intradermal administration in skin is mediated via NK₁ receptors, whilst plasma extravasation observed in the lung after intravenous administration of NKB is mediated via a tachykinin receptor-independent mechanism, as observed using selective antagonists in NK₁ receptor knockout mice. In addition, evidence is provided to show that neither vasoactive prostaglandins nor nitric oxide play a major role in mediating this response.

Responses to intradermal NKB injection

Our previous results have shown that exogenous and endogenous substance P act solely via the NK₁ receptor to mediate plasma extravasation in experiments where the selective NK₁ receptor antagonist SR140333 or NK₁ knockout mice have been used (Cao et al. 1999; Grant et al. 2002). Plasma extravasation mediated by other agents (e.g. histamine) was not affected in the knockout mice, or by the presence of the antagonist. The present study demonstrates for the first time that neurokinin B, in addition to substance P, acts solely via the NK₁ receptor to mediate plasma extravasation after its injection into mouse skin.

Responses to intravenous administration

The intravenous administration of NKB results in plasma accumulation in the lung, liver, uterus and skin as shown in Fig. 2 and Table 1. This mode of administration was chosen in order to mimic the elevated plasma levels seen in pre-eclampsia more closely than extravascular administration. The results in Fig. 3 confirmed that the majority of the plasma extravasation is real, and not an artefact due to vasodilatation. The great potency of NKB as a stimulator of plasma extravasation in the lung is particularly apparent, with the 30 nmol dose producing a greater than 15-fold increase in plasma extravasation, compared to the control value. We hypothesised that the plasma extravasation produced by systemic administration of NKB would, as with intradermal NKB injection, be mediated via NK₁ receptor activation. In this case, plasma extravasation produced by i.v. injection of NKB into NK₁ knockout mice would either be greatly reduced or abolished. In keeping with this hypothesis, plasma extravasation was not observed in the skin or uterus of NK₁ knockout mice (Table 1). Surprisingly though, NKB was able to cause significant oedema both in the lung (Fig. 2) and liver (Table 1) of the knockout mice. Again, the potent ability of NKB to induce plasma extravasation in the lung is particularly striking.

These results suggest that, as predicted, the plasma extravasation in the skin and uterus is dependent on activation of the NK₁ receptor by NKB. Previous studies (Lembeck et al. 1992; Emonds-Alt et al. 1993) have shown that these receptors are localised to the endothelial cells of post-capillary venules and act to increase vessel permeability. However, plasma extravasation to NKB was apparent in the lung and liver of both wild-type and knockout mice, suggesting that in these organs it is not NK₁ receptor-mediated. This result may indicate that plasma extravasation in the lung and liver is not mediated by the NK₁ receptor in either the wild-type or knockout mice, implying a role for other neurokinin receptors (e.g. the NK₃ receptor, which is known to be expressed peripherally) as NKB is not known to interact with non-neurokinin receptors.

In addition to NKB, we also examined the ability of substance P to induce plasma extravasation. Substance P and NKB are both members of the mammalian tachykinin family, and have closely related structures, sharing a common carboxy terminal sequence. They are both full agonists at all three neurokinin receptors. It was expected that systemic substance P would activate the same receptors as NKB, producing comparable, if not identical, effects in wild-type mice. Substance P was able to induce plasma extravasation in the skin and uterus of wild-type mice, tissues in which the effects of NKB appear to be mediated by the NK₁ receptor, the preferred receptor for substance P. However it had little, if any, effect on the lung or liver, tissues where plasma extravasation to NKB is NK₁ receptor-independent. This suggests that there are two separate, tissue-specific mechanisms by which NKB is stimulating plasma extravasation. In the skin and uterus it is NK₁ receptor-dependent, and can also be activated by another tachykinin, substance P. A separate mechanism exists in the lung and liver, which is independent of NK₁ receptor activation, and seems to be specific to NKB, rather than being a general tachykinin-activated pathway.

Mechanisms of NKB activity

The most obvious mechanism by which NKB may be inducing NK₁ receptor-independent plasma extravasation in the lung and liver is through activation of either an NK₂ or NK₃ receptor. To examine the role of other neurokinin receptors in modulating the response to NKB in the knockout mice, selective antagonists were used to try and block the plasma extravasation to NKB in the lung. Neither the NK₂ antagonist SR48968 nor the NK₃ receptor antagonists SR142801 or SB-222200 had any effect on the plasma extravasation. This suggests that the mechanism by which NKB produces plasma extravasation in the lung is specific to NKB, rather than being a general tachykinin pathway, and also independent of any currently identified neurokinin receptors.

The prostaglandins and NO are also known to play important roles in several inflammatory conditions, and so may be contributing to the plasma extravasation produced by NKB either directly, or by promoting blood flow to the tissue. To investigate this hypothesis, l-NAME was used to inhibit NO synthase, and indomethacin was used to inhibit the production of prostaglandins by COX. Treatment with l-NAME was able to attenuate the plasma extravasation in the knockout mouse lung to NKB by around 30 %. However, this is most likely to be a side effect of its vasoconstrictor activity reducing the blood supply to the lung rather than direct inhibition of the plasma extravasation, although a direct effect cannot be ruled out at this stage. Indomethacin had no effect on the plasma extravasation.

Taken together, these results provide the first evidence for a novel pathway by which NKB can stimulate plasma extravasation, in addition to the widely described mechanism involving activation of NK₁ receptors which can be triggered by any of the tachykinins. This new pathway seems to be seems to be specific to NKB, and does not involve neurokinin receptor activation. It does not seem to involve prostaglandins and may be independent of NO production. A further possibility may be that NKB is stimulating degranulation of mast cells, leading to plasma extravasation. A tachykinin receptor-independent mechanism for mast cell activation has been proposed, in addition to that via the NK₁ receptor, but previous work from this group (Cao et al. 1999; Grant et al. 2002) has demonstrated the total abolition of skin oedema in the NK₁ receptor knockout mice. This implies that tachykinin activation of mast cells, at least in this mouse strain, is purely NK₁ receptor-dependent. It is therefore very unlikely that activation of mast cells by NKB is contributing to the plasma extravasation observed in the lungs in this study. There have been very few studies of the peripheral effects of NKB in vivo as it was previously thought to be confined to the brain and nervous system, so further studies are necessary to fully elucidate this novel NKB mechanism.

In conclusion, this manuscript provides evidence that NKB is a potent mediator of oedema formation, acting via NK₁ receptors in mouse skin. However, the results suggest that NKB also has potent activity in the mouse lung via an, as yet, unidentified mechanism. It is intriguing that this mechanism appears to be independent of neurokinin receptors, and also not shared by substance P. This led us to conclude that NKB acts via a novel mechanism to induce plasma extravasation in the mouse lung. The oedema inducing activity of NKB may be of direct relevance to the pathophysiology of pre-eclampsia if a similar phenomenon is observed in human lung. Multi-organ oedema is one of the defining signs of pre-eclampsia, and acute pulmonary oedema is a major cause of death in those affected by pre-eclampsia (Sibai et al. 1987). Further work is necessary to identify the mechanisms leading to the elevated secretion of NKB during pre-eclampsia, and whether it truly is the primary causative agent of the observed pathologies.