Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The pathophysiology of bronchial asthma is regulated by the release of cytokines from inflammatory cells. There is increasing evidence that one of these cytokines, tumor necrosis factor- (TNF), is directly linked to airway inflammation and the hyper-responsiveness observed in asthma (Broide et al., 1992; Amrani et al., 2000a). In vivo pretreatment of airways in several species, including human, to aerosolized TNF produced significant increases in bronchial airway resistance when challenged with endogenous agonists (Kips et al., 1992; Thomas et al., 1995). These observations, together with pharmacological evidence (Renzetti et al., 1996), suggest that TNF may be responsible for the bronchial smooth muscle hyper-responsiveness observed in asthma. However, the cellular mechanisms by which this TNF-induced hyper-responsiveness occurs is not clear. Longer incubations (18 h) with TNF in airway smooth muscle cells can produce increases in agonist-stimulated intracellular Ca²⁺ release (Amrani et al., 1996), possibly through increased expression of signaling proteins (Amrani et al., 1997; Hotta et al., 1999). However, in vivo effects are observed with TNF incubations of 1 h (Kips et al., 1992; Thomas et al., 1995). These effects could be caused by release of further inflammatory agents (Moore et al., 1991) and/or may be direct effects on bronchial smooth muscle cells.

It is generally accepted that the initiation of smooth muscle contractility is predominantly controlled by a Ca²⁺-dependent increase in myosin light chain₂₀ (MLC₂₀) phosphorylation (Sellers, 1991). However, other pathways have now been described that may regulate the contractility of smooth muscle by regulating the phosphorylation level of MLC₂₀ independently of a rise in intracellular Ca²⁺ (Himpens et al., 1990). These pathways are generally stimulated by contractile agonists that activate heterotrimeric G protein-coupled receptors, probably via G_12/13 stimulation of Rho guanine exchange factors (GEFs) (Hart et al., 1998; Kozasa et al., 1998). Activation of the monomeric G protein RhoA by seven-transmembrane G protein-coupled receptors may be of particular importance in many smooth muscle types (Somlyo and Somlyo, 2000). Activation of RhoA leads to subsequent activation of a recently isolated downstream target of Rho, p160 Rho-associated coiled-coil-containing protein kinase (Rho-kinase) (Leung et al., 1995; Matsui et al., 1996). Rho-kinase directly phosphorylates the regulatory subunit (MYPT-1) of the smooth muscle myosin light chain phosphatase (Feng et al., 1999), either directly or via an additional myosin phosphatase-associated kinase (Borman et al., 2002). This phosphorylation results in an inhibition of phosphatase activity leading to increased accumulation of phosphorylated MLC₂₀ (Kimura et al., 1996) and subsequently an increased Ca²⁺ sensitivity to contraction. We have recently uncovered the first evidence of the cellular mechanism whereby TNF pretreatment of bronchial smooth muscle directly increases maintained agonist-induced responses (Parris et al., 1999). It is now clear that TNF, although not a Ca²⁺-releasing agonist, potentiates the Ca²⁺-activated contractile response via an increase in MLC₂₀ phosphorylation. This TNF-induced increase in the Ca²⁺ sensitivity of MLC₂₀ phosphorylation is the result of TNF receptor activation and intracellular signal transduction pathways that ultimately regulate smooth muscle contractility. TNF is known to activate a variety of signaling cascades in airway smooth muscle (Emala et al., 1997; Amrani et al., 2000b, 2001; McFarlane et al., 2001). Because there is no experimental evidence to suggest that TNF receptor subtypes (TNF-R1 and TNF-R2) can couple to heterotrimeric G proteins, the activation of these Ca²⁺-sensitizing pathways is likely to occur via as yet undescribed signaling cascades, possibly involving either a potentiation of myosin light chain kinase activity or inhibition of the myosin light chain phosphatase activity.

In this study, we show that the TNF-induced Ca²⁺ sensitization of MLC₂₀ phosphorylation in ASM cells from both guinea pig and human is regulated by activation of RhoA in a time course compatible with the TNF-induced airway hyper-responsiveness. This is predominantly via the TNF-R1 receptor. This Ca²⁺ sensitization is induced by an inhibition of the smooth muscle myosin light chain phosphatase activity, probably via Rho-kinase activation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. Recombinant human TNF- was purchased from R & D Systems (Abingdon, UK). Mutated forms ("muteins") of human TNF-, which allow selective activation of either TNF-R1 (R32W, S86T/R1-TNF) or TNF-R2 (D143N, A145R/R2-TNF), have been described previously (Van Ostade et al., 1994). Monoclonal antibodies against RhoA (26C4) and TNF-R1 (H5) and a polyclonal antibody against TNF-R2 were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Polyclonal antibodies against MYPT-1 and phospho-MYPT-1 (Thr696) were from Upstate Biotechnology (Lake Placid, NY). The Rho-kinase inhibitor Y27632 was purchased from BIOMOL Research Laboratories (Plymouth Meeting, PA).

Cell Isolation and Culture. Male Duncan Hartley guinea pigs were killed by cervical dislocation, and the trachea was removed and placed in Hanks' balance salt solution. After removal of fat and connective tissue, the smooth muscle was cut into small pieces and incubated in serum-free DMEM containing 1 mg/ml collagenase (type II), 0.2 mg/ml elastase (type IV), and 50 µg/ml soybean trypsin inhibitor at 37°C. The tissue was triturated every 30 min until complete dispersal had occurred (3-4 h); transferred into an 80-cm² tissue culture flask containing DMEM, 20% FBS, 2 mM L-glutamine, penicillin (10,000 units/ml), and streptomycin (10 mg/ml); and incubated at 37°C in a humidified 5% CO₂ atmosphere. After 24 h, the medium was removed and cells were transferred to fresh DMEM, 10% FBS, 2 mM L-glutamine, penicillin (10,000 units/ml), and streptomycin (10 mg/ml). Cells were used for experimentation between passages 4 and 8.

Cell Permeabilization and MLC₂₀ Phosphorylation. Smooth muscle cells grown to 80% confluence in six-well tissue culture plates were transferred to 1 mM EGTA relaxing solution (G1) with pCa < 8.0 (1 mM EGTA, 30 mM PIPES, 10 mM creatine phosphate, 7.3 mM Na₂ATP, and 85.8 mM potassium methane sulfonate) and permeabilized with 50 µg/ml Staphylococcus aureus -toxin (Sigma-Aldrich, St. Louis, MO), for 2 h at RT. A23187 (5 µM) was added to release calcium from internal stores. After removal of -toxin, cells were incubated in fresh G1 without or with TNF (200 ng/ml) for 1 h at RT, and then transferred to pCa 6.8 for 2 min. Details of the solutions have been described previously (Horiuti et al., 1986). The Ca²⁺ concentration of pCa buffers was regulated by the ratio of K₂EGTA and CaEGTA. The incubation in pCa 6.8 was chosen to activate myosin light chain kinase. Two minutes is the peak time course of MLC₂₀ phosphorylation (data not shown). In some experiments, cells were preincubated with the rho-kinase inhibitor Y27632 (10 µM) for 30 min before the addition of TNF. Treatments were terminated by the addition of HClO₄ to a final concentration of 0.3 M, and cells were scraped from the dishes and centrifuged at 10,000g for 2 min. Cell pellets were washed with acetone to remove acid, air-dried, and extracted in 8 M urea, 20 mM Tris, 22 mM glycine, and 10 mM dithiothreitol for 1 h at RT. Cellular debris was removed by centrifugation at 10,000g for 2 min at room temperature and supernatants were analyzed by two-dimensional electrophoresis as described previously (Parris et al., 1999).

RhoA Activation Assay. RhoA activity was measured using a pull-down assay, based on the binding of active, GTP-bound RhoA to the Rho-binding domain of rhotekin (Ren et al., 1999). Smooth muscle cells, grown to 80% confluence in 10-cm² dishes, were transferred to serum-free medium for 48 h before treatment with TNF, TNF-R1 mutein, or TNF-R2 mutein (200 ng/ml) for the indicated times. Cells were washed twice with ice-cold Tris-buffered saline (10 mM Tris-HCl, pH 7.4, and 150 mM NaCl) and extracted with 50 mM Tris, pH 7.2, 500 mM NaCl, 10 mM MgCl₂, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 10 µg/ml leupeptin, 10 µg/ml aprotinin, and 1 mM PMSF. Cell extracts were incubated with 20 µg of rhotekin Rho-binding domain (RBD) coupled to glutathione-agarose (Upstate Biotechnology) for 45 min at 4°C. The beads were washed three times with 50 mM Tris-HCl, pH 7.2, 1% Triton X-100, 150 mM NaCl, 10 mM MgCl₂, 0.1 mM PMSF, 10 µg/ml leupeptin, and 10 µg/ml aprotinin. Samples were analyzed by immunoblotting using a RhoA-specific monoclonal antibody. RhoA activity is determined as the amount of rhotekin-bound RhoA (GTP-RhoA) compared with the total amount of RhoA in cell lysates. Increases in RhoA activation for each experiment are expressed as fold increases from control (zero time point), normalized to "1".

Analysis of MYPT-1 Phosphorylation. Human airway smooth muscle cells, grown to 80% confluence in 10-cm² dishes, were deprived of serum for 48 h, before treatment with TNF (200 ng/ml) for the indicated times. Cells were lysed in 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.25% Na-deoxycholate, 0.1% SDS, 1 mM PMSF, 10 µg/ml leupeptin, 10 µg/ml aprotinin, and 10 µg/ml pepstatin; and cellular debris was removed by centrifugation at 15,000g for 15 min at 4°C, before analysis by immunoblotting.

Immunoprecipitation. Human airway smooth muscle cells, treated without or with TNF, as described above, were lysed with 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% Triton X-100, 1 mM EDTA, 1 mM PMSF, 10 µg/ml leupeptin, 10 µg/ml aprotinin, and 10 µg/ml pepstatin and clarified by centrifugation at 15,000g for 15 min at 4°C. Supernatants were incubated with either control IgG, anti-RhoA, or anti-TNF-R1 antibody. Immune complexes were collected by incubation with protein G-Sepharose beads, washed three times with lysis buffer, boiled in SDS sample buffer, and analyzed by immunoblotting. Control experiments demonstrated that, under these conditions, the anti-TNF-R1 antibody can coprecipitate proteins normally associated with TNF-R1.

SDS-Polyacrylamide Gel Electrophoresis and Immunoblotting. Cell lysates were mixed with an equal volume of 2× SDS sample buffer and incubated at 100°C for 5 min. Lysates were fractionated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. After blocking with 5% nonfat milk powder in Tris-buffered saline, pH 7.4, containing 0.1% Tween 20 for 1 h at RT, blots were incubated with primary antibody for 1 h at RT or overnight at 4°C, washed, and incubated with horseradish peroxidase-conjugated secondary antibody for 1 h at RT. Immunoreactive species were visualized using enhanced chemiluminescence and quantitated by scanning densitometry using a GS-690 imaging densitometer (Bio-Rad, Hercules, CA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

TNF-Induced Increase in Ca²⁺-Activated MLC₂₀ Phosphorylation in Cultured Guinea Pig ASM Cells. We have previously demonstrated that TNF treatment of permeabilized guinea pig bronchial smooth muscle strips results in Ca²⁺ sensitization of the myofilaments and a corresponding increase in MLC₂₀ phosphorylation (Parris et al., 1999). In attempting to dissect the underlying mechanism of this novel signaling pathway, we have examined the ability of TNF to stimulate MLC₂₀ in permeabilized cultured guinea pig ASM. In initial studies, the effect of GTPS, a nonhydrolysable analog of GTP, known to produce an increase in the Ca²⁺ sensitivity of MLC₂₀ phosphorylation in permeabilized smooth muscle (Fu et al., 1998), was determined. After a 2-min incubation in pCa 6.8 with 100 µM GTPS, the MLC₂₀ phosphorylation was significantly increased compared with pCa 6.8 alone (Fig. 1). Because GTPS is membrane-impermeant, these results demonstrate that the cells are effectively permeabilized and that the intracellular signaling pathways are intact.

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Effect of TNF and GTPS on Ca2+-activated MLC20 phosphorylation in permeabilized guinea pig ASM cells. A, typical colloidal gold-stained membranes showing nonphosphorylated (NP), monophosphorylated (P1), and diphosphorylated (P2) MLC20 after a 2-min stimulation with pCa 6.8. Permeabilized cells were preincubated with 200 ng/ml TNF in G1 for 45 min and phosphorylation compared with controls (incubated in G1 alone for 45 min). Permeabilized cells were also incubated with 100 µM GTPS for 2 min simultaneously with pCa 6.8. Controls were only incubated in pCa 6.8. As assessed by densitometric analyses, TNF preincubation increased the Ca2+-activated MLC20 phosphorylation compared with control. GTPS also increased the Ca2+-activated MLC20 phosphorylation. B, graph showing mean MLC20 phosphorylation expressed as a percentage of the total MLC20. Data are mean ± S.E.M. n = 8 for TNF-treated samples, and n = 5 for GTPS-treated cells.

Involvement of the RhoA/Rho-Kinase Pathway in the TNF-Induced Ca²⁺ Sensitization of Cultured Guinea Pig ASM Cells. In smooth muscle, Ca²⁺ sensitization by a number of agonists has been shown to be dependent on the small G protein RhoA (Gong et al., 1997) and its downstream effector Rho-kinase. To examine the role of the RhoA/Rho-kinase pathway in TNF-induced MLC₂₀ phosphorylation, permeabilized guinea pig ASM cells were treated with the Rho-kinase inhibitor Y27632, before incubation with 200 ng/ml TNF and subsequent stimulation with pCa 6.8 buffer as described above. The TNF-induced increase in the Ca²⁺-activated MLC₂₀ phosphorylation was significantly inhibited by preincubation with Y27632 (Fig. 2A). Y27632 alone had no significant effect on Ca²⁺-activated MLC₂₀ phosphorylation in the absence of TNF.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Effect of Rho-kinase inhibition on Ca2+-activated MLC20 phosphorylation and TNF-induced RhoA activation in guinea pig ASM cells. A, graph showing the inhibition of the TNF-induced increase in Ca2+-activated MLC20 phosphorylation by Y27632. Permeabilized ASM cells were incubated with 10 µM Y27632 for 30 min followed by the addition of 200 ng/ml TNF for a further 45 min. Cells were then challenged with pCa 6.8 for 2 min. Mean MLC20 phosphorylation data are expressed as a percentage of the total MLC20. n = 5 for each point. Asterisk denotes statistical significance. B, TNF-induced increase in GTP-bound RhoA from guinea pig ASM cells. Typical blots and graph of mean data showing GTP-bound RhoA (n = 5). The respective blot of total RhoA in each sample is also shown. Cells were stimulated at different time points with 200 ng/ml TNF and the rhotekin RBD binding measured. TNF caused an increase in GTP-bound (active) RhoA within 5 min, reaching a peak after 15 min of stimulation. This declined to baseline levels by approximately 30 min.

Involvement of the RhoA/Rho-Kinase Pathway in TNF-Induced Ca²⁺ Sensitization of Human Cultured ASM Cells. Because human TNF may have an altered affinity and/or efficacy when bound to guinea pig TNF receptors, the effects of human TNF on primary cultured human ASM cells were also investigated. In human cells, as in guinea pig ASM cells, TNF induced an increase in GTP-bound RhoA as assessed by the GST-rhotekin RBD pull-down assay (Fig. 3A). RhoA was activated more rapidly in human than guinea pig ASM cells, with a 2-fold increase in the level of GTP-RhoA detected within 1 min. Maximal RhoA activation (4-fold) was reached within 5 min, and thereafter activation declined, returning to basal levels by 60 min.

View larger version (34K): [in this window] [in a new window]   Fig. 3.   TNF-induced RhoA activation and phosphorylation of MYPT-1 in human ASM cells. A, TNF-induced increase in GTP-bound RhoA from human ASM cells. Typical blots and graph of mean data showing GTP-bound RhoA (n = 5). The respective blot of total RhoA in each sample is also shown. Cells were stimulated at different time points with 200 ng/ml TNF and the rhotekin RBD binding measured. TNF caused an increase in GTP-bound (active) RhoA within 1 min, reaching a peak after 5 min of stimulation. This declined to baseline levels by approximately 60 min. B., typical blot and graph of mean data of TNF-induced phosphorylation of MYPT-1 in human ASM cells. Cells were stimulated at different time points with 200 ng/ml TNF and blotted with anti-phospho MYPT-1 antibody. Total MYPT-1 protein in each respective sample is also shown (n = 3). MYPT-1 phosphorylation peaked at 5 min and decreased slowly thereafter. Levels were still above baseline after 60 min of TNF incubation C, typical blot and graph of mean data showing inhibition of MYPT-1 phosphorylation by Y27632 in human ASM cells stimulated with 200 ng/ml TNF for 5 min. The rho-kinase inhibitor had no effect under control conditions (no TNF). However, a 30-min preincubation with 10 µM Y27632 almost completely abolished the TNF-induced phosphorylation of MYPT-1.

TNF Receptor Subtypes Involved in the TNF-Induced Activation of RhoA. The effects of TNF are mediated by two receptors, TNF-R1 and TNF-R2 (MacEwan, 2002). To examine the expression of these receptors in cultured ASM cells, cell extracts were analyzed by immunoblotting with antibodies specific for the receptor subtypes. Hela cells, which have a previously defined expression of TNF-R1 and TNF-R2 (Grell et al., 1998), were used as a comparison. Both TNF-R1 and TNF-R2 were detected in human ASM cells and had apparent molecular weights identical to those in Hela cells (Fig. 4A). Human ASM cells contained approximately 3-fold less TNF-R1 and 1.5-fold more TNF-R2, compared with Hela cells. As previously published, Hela cells have a TNF receptor subtype ratio of approximately 95% TNF-R1:5% TNF-R2 (Grell et al., 1998).

View larger version (27K): [in this window] [in a new window]   Fig. 4.   TNF receptor subtypes involved in RhoA activation in ASM cells. A, expression of TNF receptor subtypes in human ASM cells. Typical blot of TNF-R1 and TNF-R2 expression in whole cell preparations with anti-human TNF receptor subtype-specific antibodies. Human ASM cells are compared with Hela cells. Human ASM cells express both TNF-R1 and TNF-R2. The immunoreactive bands corresponding to the TNR receptor subtypes have the same molecular weight as those observed in Hela cells. B, Rho activation in human ASM cells stimulated with TNF-R1 mutein. Typical blots and graph of mean data showing GTP-bound RhoA (n = 4). The respective blot of total RhoA in each sample is also shown. Cells were stimulated at different time points with 200 ng/ml TNF-R1 mutein and the rhotekin RBD binding measured. Selective activation of TNF-R1 produced an increase in GTP-bound RhoA in a time course and magnitude similar to recombinant human TNF: RhoA activation was increased at 1 min, reaching a maximum at 5 min. C, Rho activation in human ASM cells stimulated with 200 ng/ml TNF-R2 mutein. Typical blots and graph of mean data showing GTP-bound RhoA (n = 6). The respective blot of total RhoA in each sample is also shown. Cells were stimulated at different time points with TNF-R2 mutein and the rhotekin RBD binding measured. Selective activation of TNF-R2 produced only small increases in GTP-bound RhoA.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   TNF-R1 does not immunoprecipitate with RhoA in human ASM cells. Typical blot showing immunoprecipitation in ASM cell homogenates using anti-TNF-R1 antibody, anti-RhoA antibody, or control IgG. Total extract shows total RhoA present in the homogenate. As a positive control, anti-RhoA antibody was capable of immunoprecipitating RhoA. In nonstimulated cells, immunoprecipitation with the anti-TNF-R1 antibody did not pull down detactable levels of RhoA. This was also the case in ASM cells incubated with 200 ng/ml recombinant TNF for 5 min.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

There is considerable in vivo evidence of an important role for TNF in airway hyper-responsiveness (Kips et al., 1992; Thomas et al., 1995; Renzetti et al., 1996). The present study describes a molecular mechanism that may be, at least in part, responsible for producing acute airway hyper-responsiveness (Thomas et al., 1995). TNF can increase the Ca²⁺ sensitivity of MLC₂₀ phosphorylation in ASM cells, from both guinea pigs and humans, via a Rho-kinase-dependent pathway. In addition, TNF can activate RhoA in a time course compatible with a role in airway hyper-responsiveness and the TNF-induced potentiation of contractility previously observed in guinea pig bronchial smooth muscle (Parris et al., 1999). This activation is predominantly via TNF-R1.

Previous studies on the intracellular effects of TNF in ASM cells have concentrated predominantly on elucidating the mechanisms involved in longer term effects resulting in alterations of protein expression (e.g., heterotrimeric GTP-binding proteins) (Hotta et al., 2000) and intracellular adhesion molecule-1 (Amrani et al., 2000b, 2001). These TNF-induced effects are likely to be of importance in chronic inflammatory responses and possibly ASM cell proliferation; however, they do not address the potential mechanisms of acute airway hyper-responsiveness induced by TNF in a much shorter time course (Kips et al., 1992; Thomas et al., 1995). Our previous study suggested such a mechanism may occur via a TNF-induced increase in the Ca²⁺ sensitivity of ASM contraction (Parris et al., 1999), although TNF itself is not a contractile agonist and does not increase intracellular Ca²⁺ in ASM cells; i.e., TNF is acting only as a sensitizing agent. In permeabilized guinea pig bronchial smooth muscle strips, TNF incubation for 45 min significantly increased the Ca²⁺-activated contractility via an increase in the Ca²⁺-activated MLC₂₀ phosphorylation. The increase in the MLC₂₀ phosphorylation in guinea pig ASM cells in the present study was inhibited by preincubation with Y27632, a Rho-kinase inhibitor. Rho-kinase is known to play an important role in regulating contractility of airway (Iizuki et al., 1999, 2000; Ito et al., 2001) and other types of smooth muscle (Kureishi et al., 1997; Fu et al., 1998). To date, this regulation has been associated with contractile agonists activating seven-transmembrane receptors and heterotrimeric G proteins (Somlyo and Somlyo, 2000). This study now demonstrates a role for TNF, a noncontractile agonist, in regulating ASM contractility via activation of Rho-kinase. Further evidence for this mechanism is provided by TNF-induced activation of RhoA in both guinea pig and human ASM cells. The slower time of peak Rho activation in guinea pig ASM cells may represent a decreased relative efficacy and/or affinity of recombinant human TNF on guinea pig TNF receptors. However, this may also represent physiological differences in the TNF-induced Ca²⁺ sensitization in guinea pig versus humans. The fold levels of increase in GTP-bound RhoA observed here are similar in magnitude to those observed in other smooth muscles stimulated with contractile agonists such as U46619, norepinephrine, and endothelin-1 (Sakurada et al., 2001). Although our initial study first revealed the TNF-induced increase in Ca²⁺ sensitivity of MLC₂₀ phosphorylation in guinea pig ASM, it is important to determine whether this effect occurs in human. TNF does produce an in vivo airway hyper-responsiveness in humans (Thomas et al., 1995). In cultured human ASM cells, it was not possible to determine TNF-induced MLC₂₀ phosphorylation using two-dimensional gel electrophoresis until the apparently multiple isoforms of MLC₂₀ have been further characterized. Phosphorylation of the MYPT-1 from the smooth muscle myosin light chain phosphatase, however, was assessed as an indicator of Ca²⁺ sensitization. TNF incubation increased phosphorylation of MYPT-1 in a time course compatible with TNF-induced Ca²⁺ sensitization; however, the phosphorylation levels remained elevated for a longer time period and were above control levels even after 1 h. This is in agreement with a recent study showing that dephosphorylation of MYPT-1 is relatively slow (Takizawa et al., 2002) and consistent with the prolonged Ca²⁺ sensitization observed after TNF incubation (Parris et al., 1999). Phosphorylation of MYPT-1 was blocked by the Rho-kinase inhibitor Y27632, suggesting that it is the result of Rho-kinase activation. This study therefore suggests that TNF induces activation of the RhoA/Rho-kinase pathway, leading to phosphorylation and inhibition of the smooth muscle myosin light chain phosphatase and a subsequent increase in the Ca²⁺ sensitivity of MLC₂₀ phosphorylation.

TNF effects are mediated by two receptor subtypes, TNF-R1 and TNF-R2 (MacEwan, 2002). Both TNF-R1 and TNF-R2 receptors are single transmembrane glycoproteins with 28% sequence homology and a marked delineation of function (Grell et al., 1994). As shown in this study by immunoblotting, both these receptor subtypes are present in cultured human ASM cells (Amrani et al., 2000b) and have been previously identified by us in guinea pig ASM cells (McFarlane et al., 2001). TNF receptor engagement activates a variety of signaling molecules. This study is the first to report activation of RhoA by TNF in ASM cells, predominantly via activation of the TNF-R1, with a smaller proportion via the TNF-R2. The intracellular mechanism by which TNF activates RhoA in these cells is not clear. TNF receptors activate downstream signaling pathways via interaction with a family of adaptor proteins (MacEwan, 2002). These proteins, known as TRAFs, bind to the cytoplasmic domain of TNF receptors (Wajant et al., 1999). To date, no direct interaction of TRAFs with either Rho or one of the several GEFs, which mediate Rho activation, has been identified. TNF-induced RhoA activation has previously been implicated in cytoskeletal reorganization in serum-starved endothelial cells and fibroblasts over a relatively slow time course (approximately 30 min) (Wojciak-Stothard et al., 1998; Puls et al., 1999). This is possibly via a hierarchical cascade of structurally related monomeric G proteins (Puls et al., 1999). This study has directly measured the time course of TNF-induced RhoA activation and this reveals a rapid increase (within 1 min). This fast time course indicates that a relatively direct signaling pathway is probably involved. A direct interaction of RhoA has been previously observed with the neurotrophin receptor p75 (Yamashita et al., 1999), a related member of the TNF receptor family. We investigated whether this may also be the case for the TNF-R1 receptor. Under the conditions used in this study no interaction was observed between RhoA and TNF-R1, or the principal TNF-R1 adaptor proteins TRAF-2, receptor interacting protein and TNF receptor-associated death domain binding protein. A recent study in guinea pig ASM cells has suggested that the TNF-induced increase in MLC₂₀ phosphorylation may be via formation of reactive oxygen species (Thabut et al., 2002), although the mechanism linking reactive oxygen species generation to increased MLC₂₀ phosphorylation is not known. It is interesting to note that some effects of reactive oxygen species in the lung occur via Rho-mediated effects (Chiba et al., 2001). It is likely that the TNF receptor, predominantly TNF-R1 (or associated TRAFs), may activate one or more GEFs involved in Rho activation, although this remains to be established.

In conclusion, this study clearly demonstrates that TNF can activate RhoA in ASM cells. In addition, we have also shown that activation of the RhoA/Rho-kinase pathway contributes significantly to the TNF-induced Ca²⁺ sensitization of MLC₂₀ phosphorylation. These effects are not species-specific and were observed in both guinea pig and human ASM cells. This TNF-activated intracellular mechanism may therefore contribute to the airway hyper-responsiveness observed in inflammatory diseases such as asthma.