Adenosine-dependent activation of hypoxia-inducible factor-1 induces late preconditioning in liver cells

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

The cellular mechanisms by which ischemic preconditioning increases liver tolerance to ischemia/reperfusion injury are still poorly understood. This study investigated the role of the hypoxia-inducible factor-1 (HIF-1) in the protection associated with the late phase of liver preconditioning. Late preconditioning was induced in primary cultured rat hepatocytes by a transient (10 minute) hypoxic stress or by 15 minutes incubation with the adenosine A2A receptors agonist CGS21680 24 hours before exposure to 90 minutes of hypoxia in a serum-free medium. Late preconditioning induced the nuclear translocation of HIF-1 and the expression of carbonic anhydrase IX (CAIX), a HIF-1–regulated transmembrane enzyme that catalyzes bicarbonate production. Such effects were associated with prevention of hepatocyte killing by hypoxia and the amelioration of intracellular acidosis and Na+ accumulation. The inhibition of PKC-mediated and PI3-kinase–mediated signals with, respectively, chelerythrine and wortmannin abolished HIF-1 activation and blocked both CAIX expression and the protective action of late preconditioning. CAIX expression was also prevented by interfering with the transcriptional activity of HIF-1 using a dominant negative HIF-1β subunit. The inhibition of CAIX with acetazolamide or the block of bicarbonate influx with disodium-4-acetamido-4′-isothiocyanato-stilben-2,2′-disulfonate also reverted the protective effects of late preconditioning on intracellular acidosis and Na+ accumulation. Conclusion: The stimulation of adenosine A2A receptors induced late preconditioning in liver cells through the activation of HIF-1. HIF-1–induced expression of CAIX increases hepatocyte tolerance to ischemia by maintaining intracellular Na+ homeostasis. These observations along with the importance of HIF-1 in regulating cell survival indicates HIF-1 activation as a possible key event in liver protection by late preconditioning. (HEPATOLOGY 2008.)

A transient interruption of blood flow confers tissue protection against ischemia-reperfusion injury, a process known as ischemic preconditioning.1, 2 Liver ischemic preconditioning is receiving increasing attention for its capacity to improve reperfusion damage during hepatic surgery and transplantation.3, 4 The protective effects of ischemic preconditioning occur in two phases: an early phase that immediately follows the preconditioning stimulus (early preconditioning) and involves the modulation of different cellular functions and a delayed phase (late preconditioning) that is evident several hours after the preconditioning stimulus and is associated with gene transcription and “de novo” protein synthesis.1, 2, 5

In recent years, several studies have characterized the mechanisms responsible for the development of early preconditioning. In the liver, adenosine released by hepatocytes is an important trigger of early preconditioning.6 By interacting with adenosine A2A-receptors (A2AR), adenosine activates different intracellular signal pathways involving, among others, phosphatidylinositol 3-kinase (PI3K), protein kinase B (PKB/Akt), the isofoms δ and ϵ of protein kinase C (PKCδ/ϵ) and p38 mitogen activated protein kinase (p38MAPK).6, 7 However, the factors that modulate the gene expression during the delayed phase of hepatic preconditioning are less well characterized.6

The hypoxia-inducible factor-1 (HIF-1) is the main transcription factor responsible for the tissue adaptation to ischemia.8, 9 HIF-1 binds to hypoxia responsive elements (HRE) located in the regulatory regions of a vast array of genes that regulate energy metabolism, neovascularization, hematopoiesis and cell migration.9, 10 Moreover, recent evidence points out HIF-1 influence on cell survival responses.8, 10 In particular, the expression in rat cardiomyocytes of stable hybrid forms of HIF-1α prevent cell death induced by simulated ischemia/reperfusion,11 indicating the involvement of HIF-1–regulated genes in enhancing cell tolerance against reoxygenation injury. HIF-1 is a helix-loop-helix transcription factor that requires for its activity the dimerization of HIF-1α and HIF-1ß subunits.9 HIF-1ß is a constitutively expressed nuclear protein, while HIF-1α is a cytoplasmatic protein that is continuously degraded by oxygen-dependent HIF-prolyl-4-hydroxylase and arginyl-hydroxylase factor inhibiting- HIF-1α.9 Thus, hypoxia prevents HIF-1α hydroxylation allowing its nuclear translocation and the transcription of HIF-1 target genes. However, increasing evidence indicates that HIF-1α can be also activated under nonhypoxic conditions by oxidative stress, growth factors, and cytokines.12 In particular, we have recently demonstrated that the stimulation of adenosine A2A-receptors triggers an oxygen-independent induction of HIF-1 in macrophages.13 These observations and the notion that adenosine is a key inducer of liver preconditioning, prompted us to investigate the possible involvement of HIF-1 in hepatic late preconditioning.

Abbreviations

A2AR, adenosineA2A receptor; CAIX, carbonic anhydrase IX; EMSA, electrophoretic mobility-shift assay; HIF-1, hypoxia-inducible factor-1; PI3K, phosphatidylinositol 3-kinase; PKC, protein kinase C.

Materials and Methods

Materials.

CGS21680, acetazolamide, disodium 4-acetamido-4′-isothiocyanato-stilben-2,2′-disulfonate, gentamycin, penicillin, streptomycin, 8-cyclopently-1,3-dipropylxanthine (DPCPX), VUF5574, PD 98059, H89, chelerythine, wortmannin and dexferroxamine were purchased from Sigma-Aldrich, Milan, Italy. ZM241325 was obtained from Tocris's Cookson Ltd. (Bristol, UK).

Preparation and Treatment of Cultured Hepatocytes.

Liver cells were isolated from the livers of fed male Wistar rats (180–250 g weight) (Harlan-Nossan, Italy) by collagenase perfusion7 and hepatocytes were purified by centrifugation at 50g for 5 minutes followed by a further 2 minutes of centrifugation at 350g through a layer of Percoll (1.06 final density). Cell purity was assessed according to Benten et al.14 The use and care of the animals were approved by the Italian Ministry of Health. Hepatocyte suspension (purity >95%) were plated on collagen-coated culture dishes and cultured for 48 hours in Dulbecco's modified Eagle medium (DMEM-HAM F12) containing 10% fetal bovine serum and 1% penicillin/streptomycin and 1% glutamine. Hepatocytes were preconditioned in Krebs-Henseleit-HEPES (KHH) medium by exposure to 10 minutes of hypoxia (95% N2, 5% CO2) followed by 10 minutes of reoxygenation (95% air, 5% CO2) or by 15 minutes of treatment with 1 μmol/L CGS21680. At the end of the preconditioning treatments, the cells were transferred into fresh DMEM-HAM F12 and cultured for 1–24 hours.

The development of late preconditioning was evaluated 24 hours after cell stimulation by assessing hepatocyte tolerance to the killing induced by 90 minutes of incubation in Krebs-Henseleit-HEPES (KHH) buffer (final cell density of 106/mL) under a hypoxic atmosphere.7 Control cells were maintained in normoxic DMEM-HAM F12. The different inhibitors were added 15 minutes before preconditioning and removed 2 hours afterward. Acetazolamide and disodium 4-acetamido-4′-isothiocyanato-stilben-2,2′-disulfonate, were added 30 minutes before the hypoxic incubation. The HTC rat hepatoma cell line was obtained from the European Collection of Cell Cultures and cultured in DMEM-HAM F12 medium containing 10% fetal bovine serum and 1% penicillin/streptomycin and 1% glutamine. The preconditioning procedures and treatments were the same as for primary hepatocytes.

Electrophoretic Mobility Shift Assay (EMSA).

Nuclear extracts were prepared as described15 and aliquots were incubated with γ-[32P]ATP-labeled oligonucleotides (Primm, Milan, Italy) encompassing the binding sites for HIF-1 (5′-AGCGTACGTGCCTCAGGA-3′) and oct-1 (5′-TGCGAATGCAAATCACTAGAA-3′), and then electrophoresed and autoradiographed.15 The specificity of the assay was demonstrated by the disappearance of the signals after the addition of a 50-fold excess of specific and not nonspecific unlabeled oligonucleotides. The quantitative determinations were performed by direct nuclear counting using an InstantImager (Packard Instruments, Milan, Italy), and the values were normalized to the activity of oct-1.

Transient Transfection Assay.

Subconfluent HTC cells in 24-well tissue culture dishes were transiently transfected using the TransIT- (Mirus, Tema Ricerca, Italy) with a 5:1 mixture of pGL3PGK6TKp (100 ng) vector containing a HRE multimer (gift from P.J. Ratcliffe, Wellcome Trust Center for Human Genetics, Oxford, UK) and the pRL-TK reporter vector containing renilla luciferase, which was used to normalize transfection efficiency. After 5 hours, the medium was replaced with fresh medium, and the cells were exposed to 5μM CGS21680. When appropriate, the cells were cotransfected with 1μg of the expression vector pcDNA3ARNTdelta_b (ΔARNT) coding for the dominant negative mutant form of ARNT subunit (obtained from M. Schwarz, University of Tübingen, Germany). After 24 hours, the cells were collected and luciferase activities were measured in a Promega luminometer using the Dual-Luciferase Reporter Assay System (Promega, Milan, Italy) according to the manufacturer's instructions. The empty vectors showed practically undetectable luciferase activity. All of the transfection experiments were carried out on duplicate plates and repeated at least three times.

Evaluation of HIF-1 Expression.

HIF-1α was determined in nuclear extracts from hepatocytes and HTC cells. Equal amounts of proteins were electrophoresed in acrylamide-SDS gels and electroblotted to Hybond-ECL membranes (Amersham, Milan, Italy). After assessing transfer by means of Ponceau S staining, the membranes were incubated with monoclonal anti-human HIF-1α antibody (H1α67, dilution 1:1000; Novus Biologicals, Littleton, CO). The anti-TFIID antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was used to assess equal protein loading. The antigens were detected using an immunodetection kit (ECL Plus; Amersham) and quantified by videodensitometry.

Evaluation of Carbonic Anhydrase IX Expression.

Cell lysates were prepared as described previously,7 were centrifuged 10 minutes at 13,000g. Aliquots (30μg) of the protein extract were electrophoresed on a 10% SDS-polyacrylamide gel and blotted onto nitrocellulose membranes. CAIX was revealed using a policlonal anti-CAIX antiserum (dilution 1:500; Santa Cruz Biotechnology, CA). The anti-actin monoclonal antibody (Sigma-Aldrich, Milan, Italy) was used to assess for equal protein loading. The antigens were detected using an immunodetection kit (ECL Plus; Amersham). The relative intensity of CAIX and actin bands was measured by videodensitometry and the results were expressed as ratios.

Determination of Cell Viability.

Cell viability was estimated by microscope-counting the hepatocyte excluding Trypan blue and by the determination of nuclear fluorescence staining with propidium iodide.7

Measurement of Cytosolic pH and Intracellular Na+ Content.

Cytosolic pH and intracellular Na+ content were evaluated using the respective fluorescent dyes 2′,7′-bis(carboxyethyl)-5,6-carboxyfluorescein-acetoxymethyl ester (BCECF-AM) and benzofuran isophthalate acetoxymethyl ester (SBFI-AM) (Molecular Probes, Eugene, OR) as previously described.16

Data Analysis and Statistical Calculations.

The data were expressed as means ± SD. Statistical analysis was performed by Instat-3 statistical software (GraphPad Software Inc., San Diego, CA) using one-way analysis of variance test with Bonferroni's correction for multiple comparisons when more than two groups were analyzed. Distribution normality of all groups was preliminary verified by Kolmogorov and Smirnov test. Significance was taken at 5% level.

Results

The Stimulation of Adenosine A2A Receptors Induces Late Preconditioning in Rat Hepatocytes.

Primary cultured rat hepatocytes were exposed to a transient (10 minute) hypoxic stress followed by the return to normoxic conditions and the development of the late phase of preconditioning was evaluated 24 hours later. Figure 1A shows that late preconditioning significantly increased hepatocyte tolerance against cell killing induced by 90 minutes of hypoxia in a serum-free medium. Previous studies demonstrated the specific role of adenosine A2A receptors (A2AR) in triggering the early phase of hypoxic preconditioning.6, 17, 18 In agreement with these results, we also observed that the protection given by late preconditioning was reverted by the specific A2AR antagonist ZM241385 (1 μmol/L). Moreover, late preconditioning was evident 24 hours after a short (15 minute) hepatocyte incubation with the A2AR agonist CGS21680 (1 μmol/L) (Fig. 1A). The tolerance toward hypoxic death in hepatocytes undergoing late preconditioning was associated with the prevention of both intracellular acidosis and Na+ accumulation (Fig. 2).

Figure 1.

Late preconditioning protects hepatocytes from killing induced prolonged hypoxia: role of adenosine A2A receptors (A2AR) and PI3K- and PKC-mediated signals. Late preconditioning was induced by exposing cultured rat hepatocytes to transient (10 minutes) hypoxic stress (PC) or by 15 minutes incubation with the A2AR agonist CGS21680 (CGS, 1 μmol/L). Preconditioned and non preconditioned hepatocytes were further cultured in normoxic conditions for 24 hours before the exposure to hypoxia in serum-free medium. ZM241385(1 μmol/L), wortmannin (WM, 250 nmol/L), or chelerythrine (Chel, 50 μmol/L) were added to the culture medium 15 minutes before the preconditioning procedures and removed 2 hours after. The results are means of at least five different experiments ± SD. Statistical significance: *P < 0.001 versus hypoxia or ZM241385+PC+hypoxia; #P < 0.001 versus PC+hypoxia+WM, or PC+ hypoxia+Chel.

Figure 2.

Late preconditioning prevents intracellular acidosis and Na+ accumulation in ischemic hepatocytes. Late preconditioning was induced by hypoxic stress (PC) or by 15 minute incubation with CGS21680 (1 μmol/L; CGS). Preconditioned and non preconditioned hepatocytes were further cultured in normoxic conditions for 24 hours before the exposure to hypoxia. Wortmannin (WM, 250 nmol/L) or chelerythrine (Chel, 50 μmol/L) were was added to the culture medium 15 minutes before the preconditioning procedures and removed 2 hours after. (A) Intracellular pH was measured following 30 minutes of cell incubation in hypoxic or normoxic conditions. The changes in intracellular Na+ were measured following the induction of late preconditioning by (B) hypoxic stress or (C) CGS21680. The results are means of at least four different experiments ± SD. Statistical significance: *P < 0.05 versus PC+hypoxia+WM or PC+hypoxia+Chel; # P < 0.002 versus hypoxia, PC+hypoxia+WM or PC+hypoxia+Chel.

The Stimulation of Adenosine A2A Receptor Activates HIF-1 in Preconditioned Hepatocytes.

EMSA assays using nuclear extracts of hepatocytes preconditioned by a transient (10 minutes) hypoxic stress revealed an increase in the DNA binding of HIF-1 2-3 hours after the preconditioning stimulus (Fig. 3). The effect of preconditioning on HIF-1 activity was comparable to that induced by 3 hours of incubation under hypoxia or by the iron chelation with desferrioxamine (Fig. 3), a well-known HIF-1 inducer.19 In the same assays the complex that migrated faster than HIF-1α represented constitutive factors closely related or identical to the transcription factors ATF-1 and CREB-120 that have been previously shown to be induced by hypoxia, iron chelation, or adenosine.21, 22 HIF-1 activation in preconditioned hepatocytes was not the response to the transient oxygen deprivation because no HIF-1 DNA binding was appreciable in cells exposed to 30 minutes of hypoxia (Fig. 3). HIF-1 induction by hypoxic preconditioning was abolished by the A2AR antagonist ZM241385 (Fig. 3) that did not affected HIF-expression induced by 3 hours of hypoxia or desferrioxamine (not shown). Conversely, A2AR stimulation by CGS21680 promoted HIF-1 DNA binding in oxygenated cells mimicking the effects of preconditioning (Fig. 3). The specificity of HIF-1 DNA binding was confirmed by competition experiments using unlabeled oligonucleotides (Fig. 3). Immunoblotting of nuclear proteins from cells exposed to hypoxic preconditioning or CGS21680 confirmed an increase in the nuclear content of HIF-1α protein which paralleled with HIF-1 DNA binding (Fig. 4A,B). Also in these experiments, the A2AR antagonist ZM241385 prevented the nuclear translocation of HIF-1α (Fig. 4C), whereas the block of adenosine A1 and A3 receptors with, respectively, DPCPX (100 μmol/L) or VUF5574 (100 μmol/L) was ineffective (Fig. 4E). The capacity of A2AR to trigger nonhypoxic HIF-1 response in hepatocytes was confirmed by in vivo experiments where CGS21680 was administered (0.5 mg/kg body weight, intraperitoneally)23 to rats 3 hours before liver cell preparation. Figure 4F shows that HIF-1 nuclear content in hepatocyte isolated from CGS21680-treated rat increased by about five-fold as compared to hepatocytes from untreated rats.

Figure 3.

Late preconditioning promotes HIF-1 DNA binding through the stimulation of adenosine A2A receptor. EMSA analysis of the HIF-1 binding activity in nuclear extracts from untreated hepatocytes (C), hepatocytes treated with CGS21680 (CGS, 1 μmol/L), or exposed to preconditioning by a 10-minute hypoxic stress (PC). As positive controls, HIF-1 DNA binding was assessed in hepatocytes incubated for 3 hours under hypoxia or for 20 hours with desferrioxamina (DFO, 100 μmol/L). ZM241385 (1 μmol/L), was added to the cell culture medium before PC. The arrow indicates the inducible HIF-1 complex, whereas the arrowhead indicates the constitutive complex (const). s.c. indicates specific competition of the 3 hour CGS sample with 50-fold excess of unlabeled specific oligonucleotides. The binding activity of the constitutively expressed transcription factor oct-1 was used to assess equal loading. The bars indicate the fold difference ± SD in relation to untreated control. The results are representative of three independent experiments. Statistical significance: *P < 0.001 versus C, CGS 1h, PC2h+ZM241385, CGS 3h+s.c.

Figure 4.

Late preconditioning increases the nuclear translocation of HIF-1. Hepatocytes were preconditioned by transient hypoxic stress (PC) or by treatment with CGS21680 (1μmol/L; CGS). Preconditioned and untreated cells (C) were then cultured in normoxic conditions for 1-4 hours. (A, B) Time course of the nuclear accumulation of HIF-1α protein following the different preconditioning treatments. (A,B) Effect of blocking different adenosine receptors or protein kinases on the nuclear localization of HIF-1α in preconditioned hepatocytes. ZM241385(1 μmol/L), DPCPX (100 μmol/L), VUF5574 (100 μmol/L), wortmannin (WM, 250 nmol/L), chelerythrine (Chel, 50μmol/L), PD98059 (20 μmol/L), or H89 (0.1 μmol/L) were added 15 minutes before hepatocyte preconditioning. Preconditioned and control cells (C) were then cultured in normoxic conditions for 3 hours. (F) Effect of “in vivo” administration of CGS21680 on the nuclear translocation of HIF-1α in isolated hepatocytes. Rats received CGS21680 (0.5 mg/kg body weight) intraperitoneally 3 hours before hepatocyte preparation. HIF-1α protein levels were evaluated by immunoblot analysis of nuclear extracts from primary hepatocytes. TFDII was used as a loading control. The values are expressed as arbitrary units after normalization at 1 to the control samples. The results are means of at least four different experiments ± SD. Statistical significance: *P < 0.002 versus C, CGS 1h or PC 1h; §P < 0.05 versus CGS 2h or PC 2h; #P < 0.002 versus C, CGS+WM, CGS+Chel, PC+ ZM241385, PC+WM or PC+Chel.

Studies in vascular smooth muscle cells have shown that the oxygen-independent activation of HIF-1 by angiotensin II involves transcriptional/translational mechanisms depending on the PI3K/AKT-dependent and PKC-dependent signals.24 Because both these kinases are involved in the transduction of A2AR-mediated signals in hepatocytes,7, 16, 17 we have explored their role in HIF-1 activation by preconditioning. Figure 4C,D show that the inhibition of PI3K with wortmannin (250 nmol/L) or the block of PKC with chelerythrine (50 μmol/L) abolished the nuclear translocation of HIF-1 induced by either hypoxic preconditioning or CGS21680. No effect was observed by blocking ERK1/2 with PD98059 (20 μmol/L), whereas the inhibition of PKA with H89 (0.1 μmol/L), previously shown to affect PI3K activation by A2AR,7 prevented HIF-1 nuclear translocation (Fig. 4E). Wortmannin and chelerythrine also reverted the protective effects of late preconditioning against hypoxic injury (Figs. 1B and 2).

HIF-1–Induced Carbonic Anhydrase IX Expression in Preconditioned Hepatocytes.

Carbonic anhydrase (CAIX) is a transmembrane enzyme that catalyses the conversion of carbon dioxide and water to carbonic acid and is a major target of HIF-1 activity.25, 26 This prompted us to investigate CAIX as a marker of HIF-1-mediated action in hepatocytes undergoing late preconditioning. Immunoblotting revealed that control rat hepatocytes expressed negligible levels of CAIX (Fig. 5A). Conversely, 2-4 hours after hypoxic preconditioning or transient incubation with CGS21680 CAIX expression increased in a time-dependent manner up to 24 hours (Fig. 5A,B). The block of A2AR, PI3K and PKC with, respectively, ZM241385, wortmannin and chelerythrine, abolished CAIX expression in preconditioned hepatocytes (Fig. 5C,D). Parallel experiments showed that the stimulation of the HTC rat hepatoma cell line with CGS21680 (5 μmol/L) induced the nuclear accumulation of HIF-1 and the expression of CAIX with a kinetic comparable to that observed in preconditioned primary hepatocytes (Fig. 6A,B). To obtain further insight into the role of HIF-1 in CAIX induction by hepatic preconditioning, HTC cells were transfected with a luciferase reporter gene controlled by a DNA fragment containing multiple consensus HREs, previously shown to drive HIF-1–dependent transcription.27 the expression of the reporter gene increased more than 2.5-fold in response to CGS21680 addition (Fig. 6C). Such an effect was almost abolished upon cotransfection with a plasmid expressing a dominant negative form of HIF-1β subunit (HRE/ΔARNT) (Fig. 6C), that forms a heterodimer with HIF-1α, but cannot bind to DNA.15 The transfection with the dominant negative HIF-1β also prevented the expression of CAIX in HTC cells exposed to CGS21680 (Fig. 6D), confirming that HIF-1 was involved in CAIX induction by preconditioning.

Figure 5.

Late preconditioning induces CAIX expression in hepatocytes. Hepatocytes were preconditioned by transient hypoxic stress (PC) (A, C) or with CGS21680 (CGS, 1 μmol/L) (B, D). Preconditioned and control cells (C) were then cultured in normoxic conditions for (C, D) 8 hours or (A, B) up to 24 hours. ZM241385(1 μmol/L), wortmannin (WM, 250 nmol/L), chelerythrine (Chel, 50 μmol/L) were added before preconditioning. CAIX expression was evaluated by western blotting. The relative intensity of CAIX and actin related bands was measured by videodensitometry and the results were expressed as ratios after normalization at 1 of the control. The results are means of at least four different experiments ± SD. Statistical significance: *P < 0.05 versus C (A), PC 2h, PC 4h; or versus CGS 6h or CGS 8h (B); **P < 0.001 versus C, PC 2h, PC 4h, PC 6h (A), or C, CGS 2h, CGS 4h (B) or versus PC+ZM241385, PC+WM, or PC+chel (C); or versus C, CGS+WM or CGS+chel (D).

Figure 6.

HIF-1-mediated CAIX induction in HTC cell lines. (A) Immunoblotting of the nuclear content of HIF-1α protein in control (C) or CGS21680-treated (5 μmol/L; CGS) HTC cells. Equal amounts of proteins, as assessed by using TFIID as loading control. The values are expressed as arbitrary units after normalization at 1 for the control sample and are the means of at least three different experiments ± SD. Statistical significance: *P < 0.004 versus controls. (B) CAIX expression was evaluated by immunoblotting in control (C) or CGS21680-treated (5 μmol/L; CGS) HTC cells. The relative intensity of CAIX and actin related bands was measured by videodensitometry and the results were expressed as ratios after normalization at 1 of the control. The results are means of at least four different experiments ± SD. Statistical significance: *P < 0.002 versus controls. (C) The HTC cells were transiently transfected with the empty pGL2 basic vector (empty vector) or a construct in which luciferase is controlled by an HRE multimer (HRE), and were left untreated (C), or exposed to CGS 21680 (5 μmol/L; CGS) for 20 hours. When appropriate, they were also cotransfected with an expression vector coding for a dominant negative mutant of the constitutive HIF-1β subunit (ΔARNT). The cells were cotransfected with a control vector containing the Renilla luciferase gene. The chemioluminescence, corrected for transfection efficiency on the basis of Renilla luciferase activity, was normalized to 1 in comparison to the activity in untreated cells. Mean values ± SD of three independent experiments. **P < 0.001 versus empty vector/C or empty vector/CGS; *P < 0.004 versus HRE/CGS. (D) CAIX expression in HTC cells cotransfected with an expression vector coding for a dominant negative mutant of the constitutive HIF-1β subunit (ΔARNT). The results are means of at least three different experiments ± SD. Statistical significance: *P < 0.002 versus C, C/ΔRNT or ΔRNT/CGS.

Carbonic Anhydrase IX Expression Mediates the Protective Effects of Late Preconditioning Against Hypoxic Injury.

Several studies have proposed a role of CAIX in regulating extracellular and intracellular pH.10, 25, 26, 28 Thus, we investigated the possible contribution of CAIX in mediating the protective effects of late preconditioning against hypoxic injury. Figure 7 shows that blocking CAIX with acetazolamide (100 μmol/L) reverted the protection induced by late preconditioning toward intracellular acidosis and cell death. Intracellular Na+ accumulation was also evident in preconditioned hepatocytes receiving acetazolamide (Fig. 7C). Acetazolamide did not affect intracellular pH, Na+ homeostasis and cell viability in control hepatocytes exposed to normoxic or hypoxic conditions. The protective action of late preconditioning was also abolished by inhibiting hepatocyte bicarbonate uptake through the Cl/HCOmath image exchanger using disodium 4-acetamido-4′-isothiocyanato-stilben-2,2′-disulfonate (SITS) (50 μmol/L) (Fig. 7). This indicated that in hepatocytes undergoing late preconditioning the combined activities of CAIX and Cl/HCOmath image exchanger were responsible for maintaining intracellular H+ and Na+ homeostasis during hypoxia.

Figure 7.

CAIX mediates the protective effects of late preconditioning against hypoxic injury. Late preconditioning was induced by hypoxic stress (PC) or by 15 minutes of incubation with CGS21680 (1 μmol/L; CGS). Preconditioned and control cells (C) were cultured for 24 hours in normoxic conditions. CAIX inhibitor acetazolamide (Acet, 100 μmol/L) or the Cl/HCO3 exchanger blocker SITS (50 μmol/L) were added 30 minutes before the hypoxic incubation. (A) Hepatocyte viability was evaluated after 60 minutes of incubation in normoxic or hypoxic conditions. Results are mean of at least four independent experiments. Statistical significance: *P < 0.001 versus preconditioned cells+inhibitors. (B) Intracellular pH was evaluated after 30 minutes of incubation in normoxic or hypoxic. Results are the mean of at least three independent experiments. Statistical significance: *P < 0.04 versus preconditioned cells+inhibitors. (C) Intracellular Na+ concentration was evaluated after 30 minutes of incubation in normoxic or hypoxic. Results are the mean of at least three independent experiments. Statistical significance: *P < 0.002 versus preconditioned cells+inhibitors.

Discussion

The protective action associated with the late phase of preconditioning is well-documented in the heart and the brain.2, 5 In the liver, late preconditioning can be induced by transient ischemia/reperfusion, short-term hyperthermia, oxidative stress, and atrial natriuretic peptide6 and is associated with the induction of a number of cytoprotective genes, including inducible nitric oxide synthase (iNOS), Bcl-2, Bcl-XL, heat shock proteins, and heme-oxygenase-1.6, 29, 30 Hepatic late preconditioning ameliorates postischemic sinusoidal perfusion, leukocyte infiltration, bile production and aminotransferase release up to 48 hours after the application of the preconditioning stimulus.6 Because of its sustained duration, the pharmacological induction of late preconditioning might be particularly relevant to improve liver surgery.3, 4 However, the present knowledge of the mechanisms involved is quite preliminary.

A role for HIF-1 in preconditioning has emerged from studies in rodent brain showing that 1-6 hours of nonlethal hypoxia followed by 24 hours of reoxygenation induce the expression of elevated levels of HIF-1 protein and of HIF-1–regulated genes that are associated with the prevention of the injurious effects of a subsequent ischemia.31, 32 Moreover, Cai and coworkers have recently reported the complete loss of the cardioprotection by ischemic preconditioning in mice with partial deficiency of HIF-1α.33 In hepatocytes, HIF-1 DNA binding is evident 2 hours after the preconditioning stimulus in parallel with the increase in the nuclear content of HIF-1α. HIF-1 activation is abolished by blocking adenosine A2A receptors, while the stimulation of oxygenated hepatocytes with the A2A receptors agonist CGS21680 mimics the action of the transient hypoxia. This indicates that HIF-1 stimulation during hepatic preconditioning involves oxygen-independent mechanisms triggered by adenosine-mediated signals. Adenosine is a well recognized mediator of hepatic preconditioning17, 18 and enhances liver tolerance to ischemia/reperfusion.34 The capacity of adenosine to promote the oxygen-independent activation of HIF-1 in hepatocytes is consistent with recent observations concerning the role of adenosine in triggering HIF-1 activity in human macrophages and glioblastoma cells.13, 35, 36 Growing evidence indicates that growth factors, cytokines and vascular hormones can lead to an oxygen-independent induction of HIF-1 in many cell types.12 Differently from the stabilization of HIF-1α occurring in response to oxygen deprivation, the mechanisms implicated in such an oxygen-independent induction of HIF-1 involve an increase in the translation of HIF-1α mRNA that shifts the balance between the synthesis and the degradation towards an accumulation of HIF-1α.12 Studies from several laboratories, have identified the role of PI3K, PKB/AKT, mTOR, and p70S6 in regulating HIF-1α mRNA translation through the phosphorylation of the S6 protein in the 40S ribosomal units12. Moreover, Page and coworkers24 have shown that PKC-dependent signals are also responsible for an increased transcription of HIF-1α mRNA in vascular smooth muscle cells stimulated with angiotensin II. It is noteworthy, that both adenosine-induced and angiotensin II–induced stimulation of HIF-1 are evident after 2 hours from the addition of the agonist and require the transduction of PI3K/AKT-dependent and PKC-dependent signals. These analogies, along with the notion that hepatocyte adenosine A2A receptors are coupled with both PI3K and PKCϵ/δ signaling,7, 16, 17 suggest the possibility that the adenosine-mediated activation of HIF-1 might involve an increased transcription/translation of HIF-1α (Fig. 8).

Figure 8.

Molecular mechanisms involved in the prevention of hypoxic injury in rat hepatocytes exposed to late preconditioning. By inducing the autocrine stimulation of the adenosine A2A receptors (A2aR) hypoxic preconditioning promotes a PI3K-dependent and PKC-dependent activation of the nuclear transcription factor HIF-1. In turn, HIF-1 induces the expression of carbonic anhydrase IX (CAIX) that converts CO2 in bicarbonate in the extracellular milieu. Bicarbonate is transported into the hepatocytes through the Cl/COmath image exchanger and neutralizes the intracellular acids, thus maintaining cytosolic pH and preventing Na+ accumulation. The inhibitors used to elucidate the various steps are indicated.

The role of HIF-1 in protecting liver cells against ischemic damage is also consistent with recent observations concerning the capacity of HIF-prolyl-hydroxylase inhibitors in preventing stroke2 and myocardial infarction.37 In this study, we observed that the activation of HIF-1 triggered by the stimulation of adenosine A2A receptors is associated with an increased expression of carbonic anhydrase IX (CAIX). CAIX is normally expressed in the gastrointestinal tract where facilitates H+/COmath image transport coupled to gastric secretion.38 Under normoxic conditions CAIX expression is negligible or absent in the tissues outside the gastrointestinal tract, but it is strongly induced by hypoxia.26, 28 CAIX activity has been proposed to play a role in the acidification of extracellular milieu as well as in maintaining intracellular pH via bicarbonate supply.25, 26, 28 CAIX transcription is tightly regulated by HIF-1 and “in vitro” studies have shown that CAIX can be used as a tool for monitoring HIF-1 activity.26 Accordingly, we report that in liver cells CAIX expression is prevented by overexpressing a defective HIF-1β subunit that interfere with HIF-1 binding to DNA. In agreement with the role of CAIX in regulating cellular pH, CAIX induction in preconditioned hepatocytes ameliorates intracellular acidosis induced by ischemia by promoting bicarbonate uptake through the Cl/HCOmath image exchanger (Fig. 8). We have previously shown that the activation of acid buffering systems (Na+/H+ exchanger and Na+/HCO3 cotransporter) in response to the intracellular acidosis causes an irreversible accumulation of Na+ that precipitates osmotic hepatocyte lysis by altering the cell volume regulation.39 In this contest, the agents that prevent intracellular acidosis or Na+ overload protect against hepatocyte death.40, 41 We now demonstrate that CAIX induction during the late phase of hepatic preconditioning ameliorates intracellular Na+ overload and prevents hepatocyte killing by hypoxia. This mechanism is consistent with the role of CAIX in the development of cancer cell resistance to hypoxia.42 Nonetheless, the importance of HIF-1 activation by preconditioning can have a broader significance in ameliorating hepatic reperfusion injury; Plock and co-workers have recently reported that the activation of HIF-1 in mouse livers exposed to moderate hypoxia up-regulates cytoprotective genes and prevents hepatocyte apoptosis induced by Fas ligand.43

In conclusion, our results show that: (1) the activation of HIF-1 by adenosine-mediated signals is involved in the development of the late phase of hepatic preconditioning; (2) the expression of HIF-1–regulated CAIX contributes to increased hepatocyte tolerance to hypoxia by preventing the alteration of intracellular pH and Na+ homeostasis that lead ischemic cell death. These observations, along with the capacity of HIF-1 to down-modulate cell responses to proapoptotic signals2, 9, 10, 43 suggest HIF-1 as potential target for the pharmacological induction of late preconditioning in the liver.

Ancillary