Heat Shock Protein 70 Prevents Hyperoxia-Induced Disruption of Lung Endothelial Barrier via Caspase-Dependent and AIF-Dependent Pathways

Abstract

Exposure of pulmonary artery endothelial cells (PAECs) to hyperoxia results in a compromise in endothelial monolayer integrity, an increase in caspase-3 activity, and nuclear translocation of apoptosis-inducing factor (AIF), a marker of caspase-independent apoptosis. In an endeavor to identify proteins involved in hyperoxic endothelial injury, we found that the protein expression of heat-shock protein 70 (Hsp70) was increased in hyperoxic PAECs. The hyperoxia-induced Hsp70 protein expression is from hspA1B gene. Neither inhibition nor overexpression of Hsp70 affected the first phase barrier disruption of endothelial monolayer. Nevertheless, inhibition of Hsp70 by using the Hsp70 inhibitor KNK437 or knock down Hsp70 using siRNA exaggerated and overexpression of Hsp70 prevented the second phase disruption of lung endothelial integrity. Moreover, inhibition of Hsp70 exacerbated and overexpression of Hsp70 prevented hyperoxia-induced apoptosis, caspase-3 activation, and increase in nuclear AIF protein level in PAECs. Furthermore, we found that Hsp70 interacted with AIF in the cytosol in hyperoxic PAECs. Inhibition of Hsp70/AIF association by KNK437 correlated with increased nuclear AIF level and apoptosis in KNK437-treated PAECs. Finally, the ROS scavenger NAC prevented the hyperoxia-induced increase in Hsp70 expression and reduced the interaction of Hsp70 with AIF in hyperoxic PAECs. Together, these data indicate that increased expression of Hsp70 plays a protective role against hyperoxia-induced lung endothelial barrier disruption through caspase-dependent and AIF-dependent apoptotic pathways. Association of Hsp70 with AIF prevents AIF nuclear translocation, contributing to the protective effect of Hsp70 on hyperoxia-induced endothelial apoptosis. The hyperoxia-induced increase in Hsp70 expression and Hsp70/AIF interaction is contributed to ROS formation.

Fig 1. Hyperoxia increases Hsp70 expression.

PAECs were exposed to normoxia (room air, 5% CO₂) or to hyperoxia (95% oxygen, 5% CO₂) for 1 to 24 h or 48 h after which the protein and mRNA levels of Hsp70 were measured. (A) Representative immunoblots of Hsp70. (B) Bar graph show the changes in Hsp70 protein levels quantified by scanning densitometry. (C) Bar graph showing the changes in the mRNA levels of hspA1A, hspA1B and hspA2 quantified by quantitative real-time PCR. Results are expressed as mean ± SE; n = 4. *P<0.05 vs. normoxia.

Fig 2. Hsp70 inhibitor KNK437exaggerates hyperoxia-induced lung endothelial barrier disruption.

PAECs were treated with and without KNK437 (50 μM) and exposed to normoxia and hyperoxia for 48 h. TEER was continuously monitored as described in Materials and Methods. Results are expressed as mean ± SE; n = 4. * P<0.05 vs normoxia+vehicle.

Fig 3. Overexpression of Hsp70 protects against hyperoxia-induced lung endothelial barrier disruption.

PAECs were transduced with sham adenovirus or adenovirus containing Hsp70 gene. After incubation for 48 h, cells were then exposed to hyperoxia or normoxia for 60 h. TEER was continuously monitored as described in Materials and Methods. Results are expressed as mean ± SE; n = 3. * P<0.05 vs normoxia+sham virus.

Fig 4. Apoptosis contributes to hyperoxia-induced disruption of lung endothelial barrier.

(A and B) PAECs were exposed to normoxia or to hyperoxia for 24 to 48 h after which caspase-3 activity (A) and nuclear AIF protein level (B) was assayed. (C) PAECs were treated with and without the caspase-3 inhibitor Z-VAD-FMK (50 μM) and exposed to normoxia and hyperoxia for 48 h. TEER was continuously monitored as described in Materials and Methods. Results are expressed as mean ± SE; n = 4. * P<0.05 vs normoxia; **P<0.05 vs. hyperoxia+vehicle; #P<0.05 vs. normoxia+Z-VAD.

Fig 5. KNK437 exaggerates hyperoxia-induced endothelial apoptosis and the increases in caspase-3 activity and nuclear AIF protein level.

PAECs were exposed to normoxia and hyperoxia in the absence and presence of KNK437 (50–100 μM) for 48 h after which TUNEL staining, caspase-3 activity and nuclear AIF protein level were determined. (A) Representative images of TUNEL staining of PAECs exposed to hyperoxia and/or KNK437 for 48 h. (B) Bar graph depicting the changes in the numbers of TUNEL-positive cells. (C) Changes in caspase-3 activity. (D) Representative immunoblots of AIF. (C) Bar graph depicting the changes in nuclear AIF protein levels. Results are expressed as mean ± SE; n = 4. *P<0.05 vs. normoxia; **P<0.05 vs. normoxia+vehicle; #P<0.05 vs. hyperoxia+vehicle.

Fig 6. Knocking down Hsp70 exaggerates hyperoxia-induced endothelial apoptosis and the increase in nuclear AIF protein level.

PAECs were transfected with control siRNA or siRNA against Hsp70 mRNA (hspA1A and hspA1B) and then exposed to normoxia (nor) and hyperoxia (hyper) for 48 h after which TUNEL staining and nuclear AIF protein level were assayed. (A) Changes in the numbers of TUNEL-positive cells. (B) Representative immunoblots of AIF and Hsp70. (C) Bar graph depicting the changes in Hsp70 and nuclear AIF protein levels. Results are expressed as mean ± SE; n = 4. *P<0.05 vs. normoxia. **P< 0.05 vs. normoxia+control siRNA; #P<0.05 vs. hyperoxia+control siRNA.

Fig 7. Overexpression of Hsp70 prevents hyperoxia-induced endothelial apoptosis and the increases in caspase-3 activity and nuclear AIF protein level.

PAECs were transduced with sham (RFP) adenovirus or adenovirus containing Hsp70 gene. After incubation for 48 h, cells were then exposed to normoxia (nor) and hyperoxia (hyper) for 48 h after which TUNEL staining, caspase-3 activity and nuclear AIF protein level were determined. (A) Changes in the numbers of TUNEL-positive cells. (B) Changes in caspase-3 activity. (C) Representative immunoblots of Hsp70 and AIF. (D) Bar graph depicting the changes in Hsp70 and nuclear AIF protein levels. Results are expressed as mean ± SE; n = 3. *P<0.05 vs. normoxia; #P<0.05 vs. hyperoxia+RFP virus.

Fig 8. Hsp70/AIF interaction and the effects of NAC and KNK437 on the Hsp70/AIF interaction in hyperoxic PAECs.

PAECs were exposed to normoxia (nor) and hyperoxia (hyper) in the absence and presence of NAC (5 mM) or KNK437 (50–100 μM) for 48 h after which co-immunoprecipitations of Hsp70 and AIF in the cytosolic fraction of cell lysates were performed. (A and C) Representative immunoblots of Hsp70 and AIF. (B and D) Bar graph depicting the changes in Hsp70/AIF ratio. Results are expressed as mean ± SE; n = 4. *P<0.05 vs. normoxia; #P<0.05 vs. hyperoxia+vehicle.

Fig 9. Hyperoxia-induced increase in Hsp70 expression is ROS-dependent.

PAECs were treated with and without NAC (5 mM) and exposed to normoxia (nor) and hyperoxia (hyper) for 48 h after which Hsp70 protein level was measured. (A) Representative immunoblots of Hsp70. (B) Bar graph show the changes in Hsp70 protein levels quantified by scanning densitometry. Results are expressed as mean ± SE; n = 4. *P<0.05 vs. normoxia; #P<0.05 vs. hyperoxia+vehicle.

Fig 10. A schematic pathway illustrating the role of Hsp70 in hyperoxic lung endothelial barrier disruption.

Hyperoxia induces an increase in Hsp70 expression which plays a protective role against endothelial barrier disruption and lung endothelial apoptosis via caspase- and AIF-dependent mechanism in hyperoxic lung endothelial injury.