A nonapoptotic endothelial barrier-protective role for caspase-3

Abstract

Noncanonical roles for caspase-3 are emerging in the fields of cancer and developmental biology. However, little is known of nonapoptotic functions of caspase-3 in most cell types. We have recently demonstrated a disassociation between caspase-3 activation and execution of apoptosis with accompanying cytoplasmic caspase-3 sequestration and preserved endothelial barrier function. Therefore, we tested the hypothesis that nonapoptotic caspase-3 activation promotes endothelial barrier integrity. Human lung microvascular endothelial cells were exposed to thrombin, a nonapoptotic stimulus, and endothelial barrier function was assessed using electric cell-substrate impedance sensing. Actin cytoskeletal rearrangement and paracellular gap formation were assessed using phalloidin staining. Cell stiffness was evaluated using magnetic twisting cytometry. In addition, cell lysates were harvested for protein analyses. Caspase-3 was inhibited pharmacologically with pan-caspase and a caspase-3-specific inhibitor. Molecular inhibition of caspase-3 was achieved using RNA interference. Cells exposed to thrombin exhibited a cytoplasmic activation of caspase-3 with transient and nonapoptotic decrease in endothelial barrier function as measured by a drop in electrical resistance followed by a rapid recovery. Inhibition of caspases led to a more pronounced and rapid drop in thrombin-induced endothelial barrier function, accompanied by increased endothelial cell stiffness and paracellular gaps. Caspase-3-specific inhibition and caspase-3 knockdown both resulted in more pronounced thrombin-induced endothelial barrier disruption. Taken together, our results suggest cytoplasmic caspase-3 has nonapoptotic functions in human endothelium and can promote endothelial barrier integrity.

Keywords: barrier function; caspase-3; cytoskeleton; endothelium; human lung microvascular endothelial cells; nonapoptotic; thrombin.

Fig. 1.

Thrombin induces cytoplasmic caspase-3 (Casp 3) activation without cell death. A: human lung microvascular endothelial cells (HLMVECs) were exposed to 1.25 U/ml of thrombin for 15 min, and cells were harvested for cytoplasmic and nuclear preparations for assessment of caspase-3 activity. Cytoplasmic fractions demonstrate higher caspase-3 activity than nuclear fractions. Thrombin stimulation leads to an increase in caspase-3 activity only within the cytoplasmic fraction. Purity of cytosolic (C) and nuclear (N) fractions was assessed by exclusion of GAPDH and histone deacetylase (HDAC)-2 immunoblotting, respectively (inset); n = 3. B: nuclear and cytosolic fractions were immunoprobed for caspase-3 expression. Caspase-3 is present within the cytosolic fraction under control conditions and remains in the cytosol even after thrombin stimulation. C and D: HLMVECs were exposed to 1.25 U/ml of thrombin for 2 h, and Hoechst 33342 staining was used to measure apoptosis (by nuclear condensation and fragmentation) and cell dropout (by cell counts). Caspase inhibition in HLMVECs did not result in increased apoptosis with thrombin exposure (C) or cell dropout (D); n = 4; 4,925–5,502 cells counted/group. E: HLMVECs were exposed to 1.25 U/ml of thrombin for 24 h, and Hoechst 33342 staining was used to measure apoptosis. Thrombin exposure did not result in increased apoptosis; n = 4. RLU, relative light units; qVD, q-VD-OPH. *P < 0.05 vs. control.

Fig. 2.

Caspase inhibition worsens thrombin-induced endothelial barrier disruption. Human lung microvascular endothelial cells (HLMVECs) were exposed to 1.25 U/ml of thrombin, and endothelial barrier integrity was measured. A and B: HLMVECs grown in full media or basal media (10 and 2.5% fetal bovine serum, respectively) show a drop in transendothelial resistance (TER) in response to thrombin. Caspase inhibition with q-VD-OPH (qVD) leads to a more significant drop in TER compared with thrombin alone. Representative tracing for each condition is plotted. C: quantification of maximal drop in TER; n = 5 separate experiments; 8–11 individual wells/condition. R/R₀, resistance/resistance at baseline.

Fig. 3.

Caspase inhibition exacerbates thrombin-induced endothelial cellular contraction and paracellular gap formation. A: human lung microvascular endothelial cells (HLMVECs) were grown in basal media, and magnetic twisting cytometry was used to assess cellular stiffness at baseline and after thrombin stimulation (1.25 U/ml). Caspase inhibition with q-VD-OPH (qVD) leads to stiffer cells compared with thrombin alone. Summation of individual cell stiffness for each condition is plotted; n = 5 separate experiments; 148–235 cells analyzed/condition. B and C: HLMVECs grown in basal media and after 30 min of thrombin stimulation (time point of maximum drop in TER and increased cell stiffness); cells were fixed and stained using phalloidin, and paracellular gaps were quantified. B: fluorescent microscopy demonstrates at baseline there is strong cortical actin staining, and in response to thrombin there is marked actin stress fiber formation with resultant paracellular gaps. In caspase-inhibited cells, there is increased paracellular gap formation following thrombin. C: quantification of paracellular gaps demonstrates a significant increase in paracellular gaps in caspase-inhibited cells compared with thrombin alone. Representative images of two experiments taken at ×10 magnification. More than 105 images were analyzed. R/R₀, resistance/resistance at baseline; TER_max, maximum drop in transendothelial electrical resistance; Casp7, caspase-7; Casp3, caspase-3. *P < 0.05 vs. thrombin alone.

Fig. 4.

Caspase-3 promotes endothelial barrier integrity during thrombin-induced disruption. Human lung microvascular endothelial cells (HLMVECs) were exposed to 1.25 U/ml of thrombin, and endothelial barrier integrity was measured. A and B: HLMVECs grown in basal media or serum-free media (2.5 and 0% fetal bovine serum, respectively) show a drop in transendothelial resistance (TER) in response to thrombin. Caspase-3 inhibition with z-DEVD-FMK (DEVD) leads to a more significant drop in TER compared with thrombin alone. Summation of all individual wells for each condition is plotted. C: quantification of maximal drop in TER. D: HLMVECs were grown in 6-well plates and treated with nontargeting small-interfering RNA (siRNA) or siRNA against caspase-3. Following siRNA exposure (24 h), HLMVECs were plated on Electrical Cell-substrate Impedance Sensing System (ECIS) plates; an aliquot of cells was replated for immunoblotting. Cells were harvested at 48 h after siRNA treatment, at the time of ECIS experiments. Left, immunoblotting for caspase-7 shows no nonspecific knockdown of caspase-7 expression. Right, immunoblotting for caspase-3 shows significant knockdown of caspase-3 expression. E: quantification of maximal drop in TER shows knockdown of caspase-3 leads to significant thrombin-induced endothelial barrier disruption; n = 3–4 separate experiments; 6–14 individual wells/condition.

Fig. 5.

Schematic of dual and divergent location-specific functions of caspase-3 (Casp 3). Cytoplasmic activation of caspase-3 promotes endothelial barrier integrity although the molecular targets are unknown and warrant further investigation. Nuclear translocation of caspase-3 promotes apoptosis and endothelial barrier disruption; the molecular mechanism by which caspase-3 translocates to the nucleus is still under investigation.