An ATF6-tPA pathway in hepatocytes contributes to systemic fibrinolysis and is repressed by DACH1

Abstract

Tissue-type plasminogen activator (tPA) is a major mediator of fibrinolysis and, thereby, prevents excessive coagulation without compromising hemostasis. Studies on tPA regulation have focused on its acute local release by vascular cells in response to injury or other stimuli. However, very little is known about sources, regulation, and fibrinolytic function of noninjury-induced systemic plasma tPA. We explore the role and regulation of hepatocyte-derived tPA as a source of basal plasma tPA activity and as a contributor to fibrinolysis after vascular injury. We show that hepatocyte tPA is downregulated by a pathway in which the corepressor DACH1 represses ATF6, which is an inducer of the tPA gene Plat Hepatocyte-DACH1-knockout mice show increases in liver Plat, circulating tPA, fibrinolytic activity, bleeding time, and time to thrombosis, which are reversed by silencing hepatocyte Plat Conversely, hepatocyte-ATF6-knockout mice show decreases in these parameters. The inverse correlation between DACH1 and ATF6/PLAT is conserved in human liver. These findings reveal a regulated pathway in hepatocytes that contributes to basal circulating levels of tPA and to fibrinolysis after vascular injury.

Graphical abstract

Figure 1.

DACH1 deletion in hepatocytes increases liver tPA, plasma tPA and systemic fibrinolytic activity in mice. (A) Experimental scheme for depleting hepatocyte DACH1 in adult mice and immunoblot of liver DACH1 in these mice, with β-actin as loading control. Dach1^fl/fl mice were injected IV with AAV8-TBG-Cre (Cre) to create HC-DACH1–KO mice, with AAV8-TBG-LacZ–injected Dach1^fl/fl mice (LacZ) serving as controls. (B) Plat mRNA in liver normalized to Rplp0 and expressed as relative to the value in LacZ mice. (C) Plasma tPA concentration by ELISA. (D) Plasma tPA activity by enzymatic assay. (E) Plasma PAI-1–free tPA concentration by ELISA. (F) Plasma fibrinogen-fibrin antigen concentration by ELISA. (G) Plasma clottable fibrinogen concentration by Clauss fibrinogen assay. The data are expressed relative to the value in plasma of LacZ mice. The absolute values using human fibrinogen for the standard curve were 2.27 ± 0.14 (LacZ) and 1.43 ± 0.14 (Cre) mg/mL. (H) Concentration of fibrin degradation products in plasma by ELISA. (I) Fibrinolytic activity measured by euglobulin clot lysis assay from plasma. Horizontal lines in dot-density plots indicate mean values. n = 8 to 10 mice per group (A-H); 4 mice per group (I). *P < .05, **P < .01, ***P < .001, Mann-Whitney U test (B), Student t test (C-F,I).

Figure 2.

DACH1 deletion in hepatocytes increases bleeding time and time to thrombotic carotid occlusion in mice. (A) Tail bleeding time and representative images of bleeding patterns on filter paper. Black arrows indicate beginning bleeding time course, and red arrows depict episodes of rebleeding. Time to occlusive carotid arterial thrombosis induced by 10% FeCl₃ injury (B) or Rose bengal/laser photochemical injury (C). (B) Representative blood flow pattern. The red arrows depict rapid increases in blood flow after transient occlusions, suggestive of transient recanalization of the clotted vessel. **P < .01, 2-tailed Student t test (n = 8-10 mice per group).

Figure 3.

Silencing of hepatocyte Plat mRNA in HC-DACH1–KO mice decreases liver tPA, plasma tPA, and systemic fibrinolytic activity. Liver Plat mRNA and plasma tPA protein concentration and activity (A) and tail bleeding time and time to occlusive carotid arterial thrombosis induced by photochemical injury (B) in Dach1^fl/fl mice administrated LacZ or Cre as in Figure 1, together with control AAV8-H1 virus (Ctrl) or AAV8-H1-shPlat. Horizontal lines indicate mean values. *P < .05, **P < .01, 1-way analysis of variance, followed by the Tukey test (n = 6 mice per each group).

Figure 4.

Silencing of hepatocyte Plat mRNA in WT mice decreases liver tPA, plasma tPA, and systemic fibrinolytic activity. Plat mRNA (A), tPA concentration by ELISA (B), and tPA activity by enzymatic assay (C) in the liver and carotid arterial lysates of WT mice injected with control AAV8-H1 virus or AAV8-H1-shPlat. Results are shown as mean ± SEM (n = 4 mice per each group). Plasma tPA protein concentration (D), plasma tPA activity (E), tail bleeding time (F), and time to occlusive carotid arterial thrombosis (G) induced by photochemical injury in WT mice injected with control AAV8-H1 virus or AAV8-H1-shPlat (n = 8-10 mice per each group). (H) Plasma tPA concentration 1 week before (basal) and 20 minutes after FeCl₃-induced carotid artery thrombosis; tPA release was calculated by subtracting the basal value from the after-FeCl₃ value for each mouse. Horizontal lines in dot-density plots indicate mean values. *P < .05, **P < .01, ***P < .001, 2-tailed Student t test (A-C,E-H), Mann-Whitney U test (D). n.s., not significant (P ≥ .05).

Figure 5.

DACH1 decreases tPA expression in hepatocytes by repressing the Plat inducer ATF6. (A) Human primary hepatocytes were transduced with adeno-LacZ or adenovirus expressing dominant-negative ΔDS-DACH1 and then assayed for PLAT and ATF6 mRNA and for tPA and ATF6 by immunoblot. (B) tPA concentration by ELISA and tPA activity by enzymatic assay in the culture medium of the cells in (A). In (A-B), results are shown as mean ± SEM (n = 3 sets of cells per each group). (C) A conserved ATF6 binding consensus sequence in exon 9 of the PLAT gene. (D) Mouse liver nuclear extracts were subjected to a chromatin immunoprecipitation assay using anti-ATF6 or control immunoglobulin G. The exon 9 region containing the ATF6 binding sequence and a nonconsensus sequence in the Plat gene as control were amplified by quantitative polymerase chain reaction and normalized to the values obtained from input DNA. Results are shown as mean ± SEM (n = 3 mice per group). (E) Immunoblots of ATF6 and tPA and quantification of relative PLAT mRNA in primary human hepatocytes transduced with adeno-LacZ or adeno-shATF6. (F) Immunoblots of ATF6 and tPA and PLAT mRNA in primary human hepatocytes treated with control adeno-LacZ or adeno–ATF6-N. In (E-F), results are shown as mean ± SEM (n = 3 sets of cells per each group). Shorter film exposure time was used for developing immunoblots in panel D vs panel E to obtain signals within linear range. *P < .05, **P < .01, 2-tailed Student t test. n.s., not significant (P ≥ .05).

Figure 6.

ATF6 deletion in hepatocytes decreases systemic fibrinolytic activity, bleeding time, and time to thrombotic carotid occlusion. Liver Plat mRNA (A), plasma tPA protein concentration (B), plasma tPA activity (C), tail bleeding time (D), and time to occlusive carotid arterial thrombosis (E) induced by photochemical injury in Atf6^fl/fl mice treated with AAV8-TBG-LacZ or AAV8-TBG-Cre. Horizontal lines in dot-density plots indicate mean values. *P < .05, **P < .01, ***P < .01, 2-tailed Student t test (n = 7 mice per each group).

Figure 7.

Relationships among DACH1, ATF6, and PLAT in human liver. (A) Immunoblot of DACH1, with β-actin as a loading control, in lysates of liver specimens from 25 human subjects (supplemental Table 1). (B) ATF6 or PLAT mRNA was normalized to RPLP0. Liver DACH1:β-actin was quantified by densitometric analysis of the immunoblots in (A). Data were analyzed by the Fisher exact 2-tail tests for DACH1 <4 or >4 vs ATF6 <4 or >4 and vs PLAT <15 or >15. Statistically significant inverse associations were found for the gray areas in each graph, with the P values indicated. (C) Correlation of liver ATF6 and PLAT mRNA. The data were analyzed by linear regression, with the r² and P values indicated.