Differential roles of CXCL2 and CXCL3 and their receptors in regulating normal and asthmatic airway smooth muscle cell migration

Abstract

Structural cell migration plays a central role in the pathophysiology of several diseases, including asthma. Previously, we established that IL-17-induced (CXCL1, CXCL2, and CXCL3) production promoted airway smooth muscle cell (ASMC) migration, and consequently we sought to investigate the molecular mechanism of CXC-induced ASMC migration. Recombinant human CXCL1, CXCL2, and CXCL3 were used to assess migration of human primary ASMCs from normal and asthmatic subjects using a modified Boyden chamber. Neutralizing Abs or small interfering RNA (siRNA) knockdown and pharmacological inhibitors of PI3K, ERK1/2, and p38 MAPK pathways were used to investigate the receptors and the signaling pathways involved in CXC-induced ASMC migration, respectively. We established the ability of CXCL2 and CXCL3, but not CXCL1, to induce ASMC migration at the tested concentrations using normal ASMCs. We found CXCL2-induced ASMC migration to be dependent on p38 MAPK and CXCR2, whereas CXCL3-induced migration was dependent on p38 and ERK1/2 MAPK pathways via CXCR1 and CXCR2. While investigating the effect of CXCL2 and CXCL3 on asthmatic ASMC migration, we found that they induced greater migration of asthmatic ASMCs compared with normal ones. Interestingly, unlike normal ASMCs, CXCL2- and CXCL3-induced asthmatic ASMC migration was mainly mediated by the PI3K pathway through CXCR1. In conclusion, our results establish a new role of CXCR1 in ASMC migration and demonstrate the diverse mechanisms by which CXCL2 and CXCL3 mediate normal and asthmatic ASMC migration, suggesting that they may play a role in the pathogenesis of airway remodeling in asthma.

FIGURE 1.

CXCL2 and CXCL3, but not CXCL1, induce ASMC migration. Migration of normal ASMCs was investigated by Boyden chamber with either 1) 10-fold doses (0.01–100 ng/ml) of recombinant human (A) CXCL1, (B) CXCL2, or (C) CXCL3, or 2) 2-fold doses (0.125–8 ng/ml) of (D) CXCL2 and (E) CXCL3. CXCL2 and CXCL3, but not CXCL1, were able to induce significant migration of ASMCs. (F) The magnitude of ASMC migration induced by CXCL2 and CXCL3 is comparable to that induced by PDGF-BB, CXCL8, and CCL11. Studies included four to six independent experiments using four to six subjects. *p < 0.05, **p < 0.01, ***p < 0.001 compared with nonstimulated control; ^$$p < 0.01 compared with PDGF-BB. Results are expressed as means ± SEM.

FIGURE 2.

CXCR1 and CXCR2 differentially regulate CXCL2- and CXCL3-induced ASMC migration. To inhibit the functionality of CXCR1 and CXCR2, serum-starved ASMCs were either 1) incubated with Abs against CXCR1 or CXCR2 for 1 h prior to migration experiment, or 2) transfected with siRNA to knock down CXCR1 or CXCR2 expression. ASMC migration toward (A) CXCL2 was completely abolished following treatment with CXCR2, but not CXCR1, neutralizing Ab. Similar results were shown after using (C) CXCR2, but not CXCR1, siRNA knockdown. In contrast, (B) CXCL3-induced ASMC migration was inhibited after neutralization of either CXCR1 or CXCR2. The results were similar with (D) CXCR1 and CXCR2 siRNA knockdown. Studies included three to four independent experiments using three to four subjects. *p < 0.05, **p < 0.01, ***p < 0.001 compared with nonstimulated controls; ^$$p < 0.01, ^$$$p < 0.001 compared with isotype or scrambled siRNA controls. Results are expressed as means ± SEM.

FIGURE 3.

Differential regulation of CXCL2- and CXCL3-induced ASMC migration by p38 and ERK1/2 MAPK pathways. Serum-starved ASMCs were incubated with the following pharmacological inhibitors; BIRB0796 (BIRB, for p38 MAPK), PD184352 (PD, for ERK1/2), or PI103 (PI, for PI3K) for 1 h prior to migration experiment with 4 ng/ml CXCL2 or CXCL3. Both (A) CXCL2- and (B) CXCL3-induced ASMC migration were significantly inhibited after inhibition of p38 MAPK pathway using BIRB0796 (BIRB). However, only (B) CXCL3-induced ASMC migration was significantly reduced after inhibition of ERK1/2 MAPK pathway using PD184352 (PD). Studies included independent experiments using five to seven subjects. **p < 0.01, ***p < 0.001 compared with nonstimulated control; ^$$p < 0.01, ^$$$p < 0.001 compared with vehicle control (DMSO). Results are expressed as means ± SEM.

FIGURE 4.

CXCL2 and CXCL3 treatment activates ERK1/2 and p38 MAPK, but not PI3K, pathways in ASMCs. Whole-cell lysates from ASMCs stimulated with 4 ng/ml CXCL2 or CXCL3 for 0, 5, 15, 30, 45, and 60 min were used in Western blot. The activation of (A) p38 MAPK and (B) ERK1/2 MAPK pathways in ASMCs was strongly detected after 5 min of CXCL2 and CXCL3 stimulation. However, the significant activation of MAPKs was longer after CXCL3 stimulation. In contrast, (C) the PI3K pathway was not significantly activated by either chemokine. *p < 0.05, **p < 0.01, ***p < 0.001 compared with nonstimulated control. Studies included five to six independent experiments using five to six subjects. Results are expressed as means ± SEM.

FIGURE 5.

CXCR1 and CXCR2 play different roles in CXCL2- and CXCL3-induced p38 and ERK1/2 MAPK activation. ASMCs were transfected with siRNA to knock down CXCR1 or CXCR2 expression, after which they were serum-starved and stimulated with 4 ng/ml (0.5 nM) CXCL2 or CXCL3 for 0, 5, 15, 30, 45, and 60 min. Whole-cell lysates were then collected and used in Western blot for detection of p38 and ERK1/2 MAPK pathways. CXCR2, but not CXCR1, siRNA knockdown efficiently inhibited CXCL2- induced (A) p38 and (B) ERK1/2 MAPK pathway activation. In contrast, CXCL3-induced (C) p38 and (D) ERK1/2 MAPK pathway activation significantly decreased after siRNA knockdown of either CXCR1 or CXCR2. Studies included four independent experiments using four subjects. *p < 0.05, **p < 0.01, ***p < 0.001 compared with nonstimulated controls; ^$p < 0.05, ^$$p < 0.01, ^$$$p < 0.001 compared with scrambled siRNA. Results are expressed as means + SEM.

FIGURE 6.

CXCL2 and CXCL3 induce higher migratory response in asthmatic ASMCs. Serum-starved asthmatic ASMC migration was investigated by Boyden chamber. Higher concentrations (8 ng/ml) of (A) CXCL2 were able to induce significant migration of asthmatic ASMCs compared with normal ASMCs. In contrast, lower concentrations (0.125 ng/ml) of (B) CXCL3 were able to induce significant migration of asthmatic ASMCs compared with normal ASMCs. CXCL2- versus CXCL3-induced migration exhibits (C) dissociated curves of asthmatic ASMCs compared with (D) semi-superimposed curves of normal ASMCs. Studies included three to six independent experiments using three to six asthmatic subjects. *p < 0.05, **p < 0.01, ***p < 0.001 compared with nonstimulated control; ^$p < 0.05, ^$$p < 0.01, ^$$$p < 0.001 between asthmatics and nonasthmatics under similar concentrations; ^#p < 0.05, ^##p < 0.01, ^##p < 0.001 between CXCL2 and CXCL3 under similar concentrations. Results are expressed as means ± SEM.

FIGURE 7.

CXCL2- and CXCL3-induced asthmatic ASMC migration is mainly CXCR1- and PI3K pathway–dependent. To assess the involvement of receptors and signaling pathways in asthmatic ASMC migration, serum-starved ASMCs were incubated with Abs against CXCR1 or CXCR2 for 1 h prior to migration experiments. ASMC migration toward (A) 4 ng/ml CXCL2 was completely abolished following treatment with CXCR2, but not CXCR1, neutralizing Ab. In contrast, neutralizing CXCR1, but not CXCR2, was effective in reducing ASMC migration toward higher concentrations (B) 8 ng/ml CXCL2. ASMC migration toward either (C) 4 ng/ml or (D) 0.125 ng/ml CXCL3 was significantly inhibited following blockade of CXCR1, but not CXCR2, in asthmatic ASMCs. ASMC migration induced by 4 and 8 ng/ml of CXCL2 [(E) and (F), respectively] and 0.125 and 4 ng/ml of CXCL3 [(G) and (H), respectively] was significantly reduced after inhibition of the PI3K pathway using PI103 (PI), but not after inhibition of p38 or ERK1/2 MAPK pathways (BIRB0796 [BIRB] and PD184352 [PD], respectively). Studies included three independent experiments using three subjects. *p < 0.05, **p < 0.01, ***p < 0.001 compared with nonstimulated controls; ^$p < 0.05, ^$$p < 0.01, ^$$$p < 0.001 compared with isotype or DMSO control. Results are expressed as means ± SEM.

FIGURE 8.

CXCL2- and CXCL3-induced PI3K signaling pathway activation in asthmatic ASMCs is mainly regulated by CXCR1. Whole-cell lysates from asthmatic ASMCs stimulated with CXCL2 (4 and 8 ng/ml) or CXCL3 (0.125 and 4 ng/ml) for 0, 5, 15, 30, 45, and 60 min were used in Western blot. Significant activation of the PI3K pathway was detected 5 min after stimulation with CXCL2 at (A) 4 ng/ml and (B) 8 ng/ml or with CXCL3 at (C) 0.125 ng/ml and (D) 4 ng/ml. To evaluate the involvement of receptors in the PI3K activation in asthmatic ASMCs, serum-starved ASMCs were incubated with Abs against CXCR1 or CXCR2 for 1 h prior to collecting cell lysates after stimulation with CXCL2 and CXCL3. Blockade of CXCR2 completely abolished PI3K activation after stimulation with (E) 4 ng/ml CXCL2, whereas blockade of CXCR1 was effective in inhibiting PI3K activation following treatment (F) with 8 ng/ml CXCL2. Similar to the latter effect, neutralizing CXCR1, but not CXCR2, was effective in reducing PI3K activation after stimulating asthmatic ASMCs with (G) 0.125 ng/ml and (H) 4 ng/ml CXCL3. Studies included three independent experiments using three subjects. *p < 0.05, **p < 0.01 compared with nonstimulated controls; ^$p < 0.05, ^$$p < 0.01 compared with isotype control. Results are expressed as means ± SEM.