Alpha-1-antitrypsin inhibits acute liver failure in mice


  • Nils Jedicke,

    1. Department of Gastroenterology, Hepatology and Endocrinology, Hannover Medical School, Hannover, Germany
    2. Chronic Infection and Cancer Group, Helmholtz Center for Infection Research, Braunschweig, Germany
    3. Division of Translational Gastrointestinal Oncology, Department of Internal Medicine I, University Hospital Tuebingen, Tübingen, Germany
    Search for more papers by this author
  • Nina Struever,

    1. Department of Gastroenterology, Hepatology and Endocrinology, Hannover Medical School, Hannover, Germany
    Search for more papers by this author
  • Nupur Aggrawal,

    1. Department of Respiratory Medicine, Hannover Medical School, Hanover, Germany
    Search for more papers by this author
  • Tobias Welte,

    1. Department of Respiratory Medicine, Hannover Medical School, Hanover, Germany
    Search for more papers by this author
  • Michael P. Manns,

    1. Department of Gastroenterology, Hepatology and Endocrinology, Hannover Medical School, Hannover, Germany
    Search for more papers by this author
  • Nisar P. Malek,

    1. Department of Internal Medicine I, University Hospital Tuebingen, Tübingen, Germany
    Search for more papers by this author
  • Lars Zender,

    1. Department of Gastroenterology, Hepatology and Endocrinology, Hannover Medical School, Hannover, Germany
    2. Chronic Infection and Cancer Group, Helmholtz Center for Infection Research, Braunschweig, Germany
    3. Division of Translational Gastrointestinal Oncology, Department of Internal Medicine I, University Hospital Tuebingen, Tübingen, Germany
    Search for more papers by this author
  • Sabina Janciauskiene,

    Corresponding author
    1. Department of Respiratory Medicine, Hannover Medical School, Hanover, Germany
    • Address reprint requests to: T. Wuestefeld, Division of Translational Gastrointestinal Oncology, Department of Internal Medicine I, University Hospital Tuebingen, Ottfried-Mueller-Strasse 10, 72076 Tuebingen, Germany. E-mail: or S. Janciauskiene, Department of Respiratory Medicine, Hannover Medical School, Carl-Neuberg-Strasse 1, 30625 Hanover, Germany. E-mail:

    Search for more papers by this author
    • These authors contributed equally to this work.

  • Torsten Wuestefeld

    Corresponding author
    1. Department of Gastroenterology, Hepatology and Endocrinology, Hannover Medical School, Hannover, Germany
    2. Chronic Infection and Cancer Group, Helmholtz Center for Infection Research, Braunschweig, Germany
    3. Division of Translational Gastrointestinal Oncology, Department of Internal Medicine I, University Hospital Tuebingen, Tübingen, Germany
    • Address reprint requests to: T. Wuestefeld, Division of Translational Gastrointestinal Oncology, Department of Internal Medicine I, University Hospital Tuebingen, Ottfried-Mueller-Strasse 10, 72076 Tuebingen, Germany. E-mail: or S. Janciauskiene, Department of Respiratory Medicine, Hannover Medical School, Carl-Neuberg-Strasse 1, 30625 Hanover, Germany. E-mail:

    Search for more papers by this author
    • These authors contributed equally to this work.

  • Potential conflict of interest: Dr. Janciauskiene received grants from Grifols and Baxter. She has intellectual property rights with Grifols.

  • Supported by the German Research Foundation, DFG [(Emmy Noether Programme ZE 545/2-1 to L.Z.), “REBIRTH” cluster of Excellence, project “Liver regeneration”, SFB/TRR77, project B4) to L.Z.], the European Commission (project “HEPTROMIC”), the Wilhelm Sander Stiftung (2009.005.2 to L.Z.), the Bear Necessities Pediatric Cancer Foundation and the Federal German Ministry for Education and Research (BMBF) (ARCHES AWARD to L.Z.). N. Jedicke was supported by Hannover Biomedical Research School StrucMed Programme and SFB TRR77 German Liver Cancer Stipend; S. Janciauskiene was supported by Deutsche Forschungsgemeinschaft (SFB 587, A18) and Deutsches Zentrum für Lungenforschung (DZL).


Acute liver failure remains a critical clinical condition, with high mortality rates, and increased apoptosis of hepatocytes represents a key event in the cause of liver failure. Alpha-1-antitrypsin (AAT) is synthesized and secreted mainly by hepatocytes, and plasma purified AAT is used for augmentation therapy in patients with AAT deficiency. Because AAT therapy exerts antiinflammatory and immune modulatory activities in various experimental models, and it was recently suggested that AAT exerts antiapoptotic activities, we aimed to explore whether administration of AAT may represent a therapeutic strategy to treat acute liver failure in mice. Well-established preclinical models of acute liver failure such as the Jo2 FAS/CD95 activating model and models of acetaminophen and α-amanitin poisoning were used. Therapeutic effects of AAT were evaluated by monitoring animal survival, histopathological changes, measurement of caspase activity, and serum cytokine levels. Systemic treatment with AAT significantly decreased Jo2-induced liver cell apoptosis and prolonged survival of mice. Native and oxidized (lacking elastase inhibitory activity) forms of AAT were equally effective in preventing acute liver injury and showed direct inhibition of active caspase-3 and −8 in liver homogenates and in a cell-free system in vitro. Concomitantly, mice treated with AAT showed significantly lower serum levels of tumor necrosis factor alpha (TNF-α), which also paralleled the reduced activity of ADAM17 (TACE). Noticeably, the increased survival and a reduction of apoptotic hepatocytes were also observed in the α-amanitin and acetaminophen-induced liver injury mouse models. Conclusion: Our data suggest that systemic administration of AAT can be a promising therapy to treat acute liver failure and clinical studies to explore this treatment in humans should be initiated. (Hepatology 2014;59:2299–2308)




acute liver failure


oxidized form of alpha-1-antitrypsin


TNF-alpha converting enzyme


tumor necrosis factor-alpha


TdT-mediated dUTP nick end labeling

Acute liver failure is characterized by a sudden and massive death of liver cells and remains a disease with high mortality and limited therapeutic options, often demanding liver transplantation.[1] The injured hepatocyte may itself aggravate and exacerbate liver injury by using different mechanisms, such as oxidative stress, mitochondrial dysfunction, apoptotic or necrotic cell death, and immune activation, often leading to a systemic inflammatory response syndrome as the most common cause of death.[2, 3]

Several mouse models of acute liver failure have been established and extensively characterized, such as the concanavalin-A model,[4] the Jo2 Fas (CD95)-antibody model[5] or poisoning with hepatotoxins such as acetaminophen (Paracetamol) or α-amanitin produced by the fungus Amanita phalloides. Different cell populations in the liver like hepatocytes, cholangiocytes, activated stellate cells, and Kupffer cells express high levels of cell-death-inducing receptors like Fas (CD95).[6] Hepatocytes are very sensitive to Fas-induced apoptosis, and the administration of anti-Fas agonistic antibody Jo2 to mice results in rapid death of the animals due to fulminant hemorrhagic hepatitis, mimicking forms of acute liver failure (ALF) in humans.[7]

Apoptosis is mediated through a highly regulated sequence of steps involving the activation of caspases. Caspase-8 acts as the most upstream caspase in the apoptosis signaling cascade. It can be directly activated by the death domain of the Fas receptor complex (called DISC),[8] one of the key inducers of FAS-mediated ALF.[9] Caspase-8 directly activates caspase-3, which is the central effector caspase. Alternatively, caspase-3 can be activated by the mitochondrial pathway involving caspase-9, in which caspase-8 cleaves the BH3-interacting domain death agonist.[8] Consistent with this, Bid-deficient mice are shown to be resistant to Fas-induced hepatocellular apoptosis.[10] However, Bak/Bax-deficient mice are reported to show a delayed onset in caspase-dependent Fas-mediated hepatocyte apoptosis by reactivating the extrinsic pathway.[11] Therefore, caspase-3 is considered a key downstream effector caspase orchestrating intrinsic and extrinsic apoptosis cascades and is an attractive target to reduce hepatic damage.

Alpha1-antitrypsin (AAT) is an acute phase protein mainly synthesized by hepatocytes.[12] For a while it has been the prevailing view that inhibition of neutrophil elastase and proteinase 3 are the key functions of AAT.[13] However, recent studies demonstrate that AAT inhibits other intracellular and cell-surface proteases.[14] In addition, it becomes clear that AAT plays an important role in regulating inflammation, proteostasis, and possibly cellular senescence program independently of elastase inhibition.[15, 16] AAT is one of the few serum proteins that can protect against serum depletion-induced cell apoptosis.[17] Novel studies also suggest that AAT expresses intracellular antiproteolytic activity by binding and inactivating active caspase-3, protecting primary lung microvascular endothelial cells from apoptosis.[18] Furthermore, studies by Zhang et al.[19] showed that AAT protects pancreatic β-cells against apoptosis through inhibition of caspase-3.

Given the importance of apoptosis in the development of liver failure, we hypothesized that AAT treatment may represent a strategy to treat ALF in mice. The therapeutic efficiency of AAT treatment was probed in mouse models of Jo2-, acetaminophen-, and α-amanitin-induced ALF. Notably, AAT treatment effectively decreased liver damage and prolonged survival in all model systems. Our data suggest that AAT treatment may represent an attractive strategy to treat patients with ALF of different etiologies and clinical trials should be initiated.

Materials and Methods


C57BL/6J female mice, age 4-8 weeks, purchased from Janvier (Le Genest-St-Isle, France), were housed in accordance with the institutional guidelines of Hannover Medical School, Helmholtz Centre for Infection Research, and the University of Tübingen, Germany. All animal experiments were approved by the German legal authorities.

Preparation of AAT/oxAAT

AAT, commercially available purified human plasma-pooled AAT (Prolastin, Grifols, Barcelona, Spain), was further purified, reconstituted in phosphate-buffered saline (PBS, pH 7.2), and stored at −80°C. For the preparation of the oxidized form of AAT (oxAAT), AAT was coincubated with N-chlorosuccinimide (Sigma-Aldrich, Taufkirchen, Germany) at a 1:25 molar ratio for 30 minutes and purified by centrifugal microconcentrator centricon-30 (Millipore, Billerica, MA). Native and oxAAT were tested for their ability to form complexes with elastase. AATs and porcine pancreatic elastase (Sigma-Aldrich) were coincubated at a molar ratio 1.2:1 (AAT:elastase) for 30 minutes at room temperature and analyzed on 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). As expected, oxAAT did not form a complex with elastase, but it is cleaved as a substrate.

Animal Experiments and Survival Studies

Mice were intraperitoneally (i.p.) injected with 0.4 μg/g mouse weight of Jo2 antibody (BD Pharmingen, San Diego, CA) diluted in 0.9% NaCl. Controls were injected with 0.9% NaCl. Supplemental experiments were performed using a different Jo2 batch. For acetaminophen and alpha-amanitin-induced ALF experiments, mice were i.p.-injected with 0.6 mg/g mouse weight of acetaminophen or 0.6 μg/g alpha-amanitin (Sigma Chemical, St. Louis, MO), diluted in 0.9% NaCl.

In a model of Jo2-induced liver injury, 1 hour after Jo2 injection mice received i.p. injection of 0.5 mg/mouse AAT or oxAAT. Controls received saline. In the supplementary studies, AAT was administered either simultaneously with 0.2 μg/g Jo2 or 2 hours before Jo2 injection and mice were sacrificed after 7.5 hours. Additional experiments were performed with multiple AAT interventions. In these experiments mice were treated with AAT or saline 1 and 6 hours after acetaminophen or alpha-amanitin injection, or 1, 2, and 3.5 hours after Jo2 injection. For phenocopy experiments, 1 hour after Jo2 injection 3 μg/g caspase-3 inhibitor Ac-DEVD-CHO (Biomol, Hamburg, Germany) dissolved in dimethyl sulfoxide (DMSO) and diluted in 0.9% NaCl was i.p. administered. For controls, carrier alone was injected.

For survival studies, mice were continuously monitored. For other analysis mice were sacrificed 105 minutes after Jo2 injection. This timepoint was empirically determined from our preliminary survival studies as the earliest timepoint of hepatic encephalopathy onset in mice treated with the present batch of Jo2. Key phenotype changes were defined as significant alterations in motor behavior and reflex, especially lethargy. For acetaminophen or α-amanitin experiments, mice were sacrificed after 24 hours.

Caspase Activity Measurements

Lysates of frozen liver tissue were prepared using AFC-lysis buffer (detailed protocols available on request). Protein concentrations were determined using the Bradford method. Caspase activity was determined adopting protocols from commercially available ApoAlert fluorimetric caspase-3 /-8 assay kits (Clontech, Mountain View, CA).[9] The cleavage of the specific substrates Ac-DEVD-AFC for caspase-3, Ac-IETD-AFC for caspase-8, and Ac-YVAD-AFC for caspase-1 (all from Biomol) was monitored within 2 hours at 37°C. AFC release was measured with a fluorescence plate reader (SpectraFluor Plus Fluorimeter; Tecan, Crailsheim, Germany) using an excitation filter of 390 nm and an emission filter of 510 nm wavelength. Values were normalized for the protein concentration.

Caspase-3 and -8 Cell-Free Activity Assay

Per assay, 0.5U active recombinant mouse caspase-3 or -8 (Biovision, Mountain View, CA) was coincubated with 60 μg (0.5 mg/mL) of either AAT, oxAAT, albumin, oxidized albumin (oxAlbumin), 0.625 μg Ac-DEVD-CHO, Ac-IETD-CHO, or PBS in a total volume of 120 μL reaction buffer for 30 minutes. Caspase-3 and −8 activity was determined by adding fluorogenic caspase-3 or −8 substrate Ac-DEVD-AFC or Ac-IETD-AFC and measured as described above.

TACE/ADAM-17 Activity Assay

Activity of TACE/ADAM-17 in vivo was determined using the AnaSpec SensoLyte 520 TACE Activity Assay, following the manufacturers instructions (AnaSpec, San Jose, CA).

Statistical Data Analysis

All statistical analyses were performed using the two-tailed Student t test, log rank-test or log rank test for trend using the software GraphPad Prism 4 (GraphPad Software, La Jolla, CA). The results were considered significant at P < 0.05.

Additional materials and methods are provided in the Supporting Materials.


AAT Attenuates Fas (CD95)-Mediated Acute Liver Failure

As death-receptor mediated apoptosis represents a key mechanism in acute liver failure, we took advantage of the well-characterized Jo2 model. Jo2 antibody was delivered i.p. at a dose of 0.4 μg/g mouse body weight, a dose leading to extensive hemorrhagic liver damage (Fig. 1). Jo2-mediated tissue damage was predominantly found in the liver, as no overt histopathological changes were observed in other organs (data not shown).

Figure 1.

AAT attenuates Fas (CD95-) induced ALF. (A) Experimental layout: Mice were i.p. challenged with 0.4 μg/g Jo2 antibody, therapeutic intervention was performed with 0.5 mg AAT or oxAAT 60 minutes after liver failure induction, and mice were sacrificed after 105 minutes. Control mice received AAT only. (B) Liver sections (stained with TUNEL/DAPI or H&E magnification 200×) show massive apoptosis and hemorrhagic liver damage 105 minutes after liver failure induction. AAT and oxAAT markedly attenuates liver damage. (C) TUNEL quantification shows significantly reduced apoptotic cells in AAT and oxAAT-treated mice (P = 0.0092 compared to untreated mice). Data are represented as mean ± SEM (n = 6 for control, n = 7 for Jo2, n = 8 for Jo2 + AAT and n = 8 for Jo2 + oxAAT). (D) Kaplan-Meier survival curve: multiple interventions with AAT (indicated on x-axis) significantly increase survival time and overall survival in Jo2-induced liver damage, single intervention delays liver failure onset (P = 0.0138, log rank-test). (E) Representative western blots stained for serum AAT, liver AAT, and liver actin control showing enhanced accumulation of AAT in the liver after dosing.

To determine whether AAT treatment is sufficient to counteract the lethal effect of the Jo2 antibody, mice were treated with the native or oxidized (lacking elastase inhibitory activity) form of AAT (oxAAT) (Fig. 1 A) 1 hour after Jo2 administration. Hematoxylin and eosin (H&E) stainings and histological analyses revealed that a single dose of native or oxAAT was able to significantly reduce Jo2-mediated liver damage in all injected mice compared to controls (Fig. 1B). Quantification of apoptotic cells by TdT-mediated dUTP nick end labeling (TUNEL) revealed a reduction in the number of apoptotic cells in AAT-treated mice (78% and 90%, respectively, compared to mice injected with Jo2 alone, P < 0.01) (Fig. 1B,C).

AAT Dose- and Application Time-Dependent Effect on Mice Survival

Suppression of hepatocyte apoptosis was also reflected by an increase of overall survival time to a median of 210 minutes (mean [SD] = 229 [37] minutes, n = 5, 0/5 mice surviving) compared to nontreated mice (median 162.5 minutes, mean [SD] = 176 [47.5] minutes, n = 6, 0/6 mice surviving) (Fig. 1D). To follow up on the observed phenotype, we next used three repeated AAT injections at 60, 120, and 210 minutes after Jo2 antibody-induced liver damage. As illustrated in Fig. 1D, this led to an even further significant increase in survival time (median 286.5 minutes, mean [SD] = 253 [67] minutes, n = 6, one in six mice survived) (log rank-test P < 0.014, log rank-trend P < 0.006).

To determine whether the hepatoprotective effect of AAT depends on the application time, we either pretreated mice with AAT 2 hours prior to Jo2 injection or simultaneously injected AAT and Jo2 with a Jo2 dosage of 0,2 μg/g mouse weight. Interestingly, mice survival time and survival rates did not differ between mice pretreated with AAT and untreated groups (303 min mean survival time, median [SD] = 293 [137.5] minutes, n = 9, 22% overall survival compared to 273 minutes mean survival, median [SD] = 310 [105] minutes, n = 19, 21% overall survival). However, when mice were treated with Jo2 and AAT simultaneously, a marked increase in mean survival time (up to 402 min) and nearly twice overall survival with 36% mice surviving was observed (median [SD] = 479 [122.5] minutes, n = 11) (Supporting Fig. 1A). Based on histopathological analyses, simultaneous injection of Jo2 and AAT more efficiently reduced hemorrhages and hepatocyte apoptosis compared to mice pretreated with AAT prior to Jo2 injection (Supporting Fig. 1B,C). Similarly, TUNEL/DAPI staining confirmed more profound reduction in apoptosis (by 50%, n = 5) in mice treated with Jo2 plus AAT than in mice pretreated with AAT (by 17%, n = 5-6) (Supporting Fig. 1B-E). Our results suggest a very short half-life of the AAT protein in vivo. Consistent with this we detected a fast increase in serum AAT directly after its application and reduction at 4 hours postinjection (Supporting Fig. 2). Similarly, AAT levels strongly increased in the liver in the intervention setting (Fig. 1E).

Figure 2.

Caspase-3 plays a central role in AAT-mediated liver protection in Jo2 model: caspase-3 activity is lowered in vivo and in vitro by direct inhibition. (A) In vivo activity of caspase-3 was fluorometrically determined from whole-cell extracts. Caspase-3 activity is widely reduced in AAT and oxAAT-treated mice compared to nontreated (P = 0.0039), oxAAT showing more potent effects than AAT (P = 0.0471). (B) AAT directly inhibits caspase-3 activity in vitro: cell-free coincubation of 0.5U recombinant mouse caspase-3 and 0.5 mg AAT/oxAAT or appropriate controls for 30 minutes. Caspase activity was calculated relative to control as 100%. AAT, oxAAT, and Ac-DEVD-CHO reduce caspase-3 activity significantly (P = 0.0005, P = 0.0113, P<0.0001). (C) Liver sections (stained with TUNEL/DAPI or H&E, magnification 200×) show hepatoprotective effects for specific caspase-3 inhibitor Ac-DEVD-CHO (injected 60 minutes after liver failure induction) 105 minutes after liver failure induction (n = 8). (D) TUNEL quantification reveals significant reduction of apoptotic cell counts in Ac-DEVD-CHO treated mice compared to untreated (P = 0.0071). The level of TUNEL-positive cells is comparable to AAT-treated mice, showing no significant difference (P = 0.7681). (E) Caspase-8 activity is significantly lowered by all treatments to a comparable level (P = 0.0213). Inflammatory caspase-1 is not lowered significantly by any treatments (P = 0.5567). All percentages were calculated using untreated mice as 100% caspase activity. All data are represented as means ± SEM. (F) AAT directly inhibits caspase-8 activity in vitro: comparable experimental layout as in (B). AAT, oxAAT, and Ac-IETD-CHO directly reduce caspase-8 activity in a cell-free system (P = 0.0469; P = 0.0713; P = 0.0040).

AAT Directly Inhibits Caspase-3 Activity

Analysis of whole liver lysates obtained from AAT-treated mice revealed that caspase-3 activity was more than 80% reduced compared to nontreated mice (P = 0.004). Remarkably, mice treated with oxAAT showed a 6.63% higher reduction than mice treated with native AAT (P = 0.0471) (Fig. 2A).

Therefore, our data suggest that even after triggering the FAS pathway, AAT could inhibit apoptosis, and we therefore investigated a possible direct inhibition of caspase-3. We performed a cell-free assay wherein active recombinant mouse caspase-3 was coincubated 30 minutes with native or oxAAT and appropriate controls such as albumin, oxidized albumin, or the direct inhibitor of caspase-3 Ac-DEVD-CHO. The activity of caspase-3 was then assessed over a 2-hour time course at 37°C.

AAT reduced caspase-3 activity by 37.4% (100% [±3.33 SEM] for caspase-3 + albumin versus 62.57% [±1.50] for caspase-3 + AAT) (P = 0.0005). Remarkably, oxAAT also reduced caspase-3 activity by 36.8% (P = 0.0113) (100% [±8.64] for caspase-3 + oxidized albumin versus 63.19% [±1.05] caspase-3 + oxAAT). Expectedly, Ac-DEVD-CHO reduced caspase-3 activity by more than 97% (to 2.30% [±0.14]). (Fig. 2B). Hence, AAT seems to directly inhibit caspase-3 activity, although we did not detect a stable complex formation between AAT and caspase-3 (Supporting Fig. 3).

Figure 3.

AAT treatment ameliorates serum TNF-α levels and TACE/ADAM-17 activity. (A) Serum TNF-α levels are significantly reduced in all treated mice 105 minutes after liver failure onset (P = 0.0194). Ac-DEVD-CHO lowers serum TNF-α levels less pronounced than AAT (P = 0.1337). Data are presented as means ± SEM. (B) AAT, oxAAT, and Ac-DEVD-CHO significantly inhibit TACE/ADAM-17 activity in vivo (P < 0.0001; P = 0.0167; P = 0.0076).

AAT and Ac-DEVD-CHO, a Direct Inhibitor of Caspase-3, Show Similar Protective Effects Against Fas Receptor-Mediated Acute Liver Injury

As we demonstrated that AAT directly inhibits caspase-3, we wanted to compare the hepatoprotective effects of AAT with the direct inhibitor of caspase-3, Ac-DEVD-CHO in vivo. Instead of AAT, 1 hour after Fas-induced liver damage we conducted a therapeutic intervention with 3 μg/g Ac-DEVD-CHO. The key phenotypic changes in mice treated with the caspase-3 inhibitor closely resembled those treated with AAT or oxAAT (Fig. 2C-E). The recapitulation of our AAT phenotype strongly supports a major contribution of AAT-mediated inhibition of caspase-3 to the observed attenuation of liver damage and failure.

Activity of Caspases-8 and -1 in Liver Homogenates and In Vitro

Caspase-8 represents an upstream key enzyme in the death receptor-mediated apoptosis pathways. As AAT exerts fast effects on apoptosis regulation, we further investigated potential upstream effects. Indeed, the activity of caspase-8 in whole liver cell lysates was reduced by more than 60% (from 100% [±17.49] for nontreated mice to 39.25% [±13.39] and 33.42% [±7.09]) in AAT- and oxAAT-treated mice, respectively. Treatment with caspase-3 specific Ac-DEVD-CHO lowered caspase-8 activity in liver homogenates to a similar level of 34.17% (±10.38) (P = 0.0053). Similarly, as for the cell-free caspase-3 assay, we investigated a potential inhibitory effect of AAT on caspase-8. Indeed, AAT reduced caspase-8 activity by 30% (Fig. 2F). Interestingly, our experiments revealed that the anti-caspase-8 effect of AAT in vivo is far stronger than in vitro. These results indicate that the inhibition of caspase-3 probably also inhibits caspase-8 by way of the known feedback loop in hepatocytes.[20] Caspase-1 activity was not affected by AAT treatment (Fig. 2E).

Treatment With AAT Decreases Serum Levels of TNF-α

TNF-α is a secreted proinflammatory cytokine which plays a central role in Jo2-induced hepatotoxicity.[21, 22] Therefore, we analyzed serum TNF-α levels. As shown in Fig. 3A, in both AAT- and oxAAT-treated mice, serum levels of TNF-α were reduced by 52% and 38%, respectively (mean [SEM] 19.4 ± 3.5 pg/mL, n = 7 and 25.0 ± 4.6 pg/mL, n = 7) compared to controls (40.5 ± 6.4 pg/mL, n = 9, P = 0.0194). Under these experimental conditions the caspase-3 inhibitor Ac-DEVD-CHO reduced serum TNF-α levels by only 30% (28.1 ± 4.0 pg/mL), suggesting that AAT treatment exerts therapeutic effects beyond capsase-3 inhibition. We further investigated the effect of AAT on TNF-α by analyzing ADAM17 activity in whole liver cell lysates. ADAM17 functions as a TNF-α converting enzyme (TACE) to release autocrine TNF-α. Consistent with the reduction in TNF-α, AAT inhibits TACE activity. Interestingly, the caspase-3 inhibitor Ac-DEVD-CHO inhibited TACE activity to some extent (Fig. 3B.) This finding might be consistent with reported caspase dependency of TACE activity.[23]

AAT Attenuates Liver Damage in Models of Acetaminophen- and alpha-Amanitin-Induced ALF and Increases Survival

Finally, we wanted to test the antiapoptotic effects of AAT in additional preclinical models of acute liver failure, acetaminophen (Paracetamol) and alpha-amanitin intoxication. Again we used an intervention set up, applying AAT after triggering liver damage. In both settings, treatment with AAT showed significant hepatoprotective effects (Fig. 4A-D).

Figure 4.

AAT treatment is effective in acetaminophen and α-amanitin-induced hepatic failure. (A) H&E-stained liver slides show attenuated liver damage in AAT-treated mice 24 hours after acetaminophen (APAP) or α-amanitin intoxication. (B) TUNEL/DAPI staining and quantification show significantly lower apoptotic cell count in AAT-treated mice in APAP hepatoxicity (P = 0.049) and α-amanitin-induced liver damage (P = 0.038), comparable to control mice, receiving AAT only. All data represented as mean ± SEM. (C) Kaplan-Meier survival curves indicating markedly increased survival time and overall survival for AAT and oxAAT-treated mice in acetaminophen (APAP)-induced acute liver failure (P = 0.049, log rank-trend). (E) Kaplan-Meier survival curves showing significantly increased survival time and overall survival for AAT and oxAAT-treated mice in α-amanitin-induced acute liver failure (P = 0.0136).

To test the effect of AAT on mice survival, we used our therapeutic regimen of multiple interventions in both models. Especially in the USA and UK, acetaminophen toxicity represents the most common cause of ALF,[24] and therefore a major health problem. Acetaminophen-induced acute liver failure was markedly attenuated, as mice repeatedly treated with AAT or oxAAT showed increased survival time and overall survival up to 100% for oxAAT (log rank-trend: P < 0.05) compared to only 40% survival of mice treated with carrier alone (Fig. 4C). In the α-amanitin model, AAT treatment resulted in a later onset of liver failure and enhanced survival to 20% compared to 0% in the controls. This clearly indicates the potential power of AAT treatment against liver failure of different etiologies.


ALF involves the abrupt loss of hepatocellular function in patients with previously normal liver function.[1] Possible triggers for ALF are intoxication by acetaminophen or mushroom toxins, drug-induced hepatitis, or genetic diseases such as Wilson's disease.[25] In the U.S. and many west European countries, drug-induced liver injury predominates.[26] As ALF is associated with high morbidity and mortality without liver transplantation,[27] and as the availability of donor organs is limited, new strategies to handle this disease are urgently needed.

While different studies have suggested antiapoptotic effects of AAT, our study shows for the first time that 1) AAT can be used in vivo to treat ALF; 2) AAT treatment prolongs mice survival in different ALF models; and 3) AAT directly inhibits caspase-8.

Our results imply that systemic treatment with AAT increases resistance to acute liver injury and therefore could prolong the time until transplantation is needed, or even allow the endogenous regenerative capacity of the liver to rescue the organ.

Regarding the mechanism, how AAT treatment protects against ALF, our study reveals a direct interaction between AAT and active caspase-3 and -8. It has been reported that caspase inhibitors can protect mice against Jo2-induced apoptosis of hepatocytes and lethality.[28] Therefore, to further verify that the antiapoptotic effect of AAT is, at least in part, mediated by way of inhibition of caspase-3, we compared the effects of AAT with a specific caspase-3 inhibitor, Ac-DEVD-CHO. When Ac-DEVD-CHO was administered 1 hour after the Jo2 injection, liver cell damage was reduced in a similar manner to AAT-treated mice. Taken together, our data provide evidence that the protective effects of AAT during Jo2-induced acute liver injury are related to caspase inhibition and highlighted the therapeutic possibilities for inhibiting an already activated apoptosis cascade. Importantly, in contrast to synthetic, preclinical caspase-3 inhibitors, ALF treatment using AAT may rapidly be translated into clinical use, as AAT therapy has been FDA-approved for 25 years and is widely used for patients with AAT deficiency-related emphysema. Preparations of AAT show no apparent toxicity and provide excellent long-term safety.[29]

Notably, native and oxidized forms of AAT mediated similar protective effects in the Jo2-model. Oxidation of two methionines of AAT, Met358 and Met351, results in loss of affinity of AAT for elastase.[30] Thus, our findings suggest that caspase-3 and -8 inhibition by AAT does not involve Met at the active site, which is necessary for the inhibition of neutrophil elastase and other serine proteases.

While inhibition of caspase-3 and -8 likely represents a key mechanism by which AAT attenuates experimentally induced acute liver injury, other protective mechanisms clearly contribute to our phenotype, as the in vivo effects are more potent than suggested by the in vitro data. For acetaminophen-induced ALF c-Jun-kinases,[31] apoptosis-signal-regulating-kinase132 and mitochondrial oxidant stress are important mediators of toxicity. APAP liver damage activates neutrophils,[33] and recent publications identified the ability of AAT to regulate key neutrophil inflammatory responses including cell chemotaxis,[34] partly by inhibiting ADAM17 activity. In addition, AAT was shown to abrogate reactive oxygen species (ROS) production elicited by a variety of stimulants,[35] and to protect against cytokine-induced apoptosis.[36] AAT affects monocytes to reduce the release of TNF-α and increase the release of the antiinflammatory cytokine IL-10,[37] an antiinflammatory action of AAT independent of its antiprotease activity.[38] In APAP-induced liver failure, AAT should provide its protective role through a combination of direct caspase-3, -8 inhibition, TNF-α (activates JNKs) reduction and its immune-modulatory functions.

Alpha-amanitin, on the other hand, inhibits hepatic transcription and thereby sensitizes hepatocytes strongly to the cytotoxic action of TNF-α[39] and TNF-α-induced murine hepatocyte apoptosis requires transcriptional arrest.[40] Furthermore, α-amanitin-induced apoptosis of cultured human hepatocytes is p53 and caspase-3-dependent.[41] Therefore, in this system the inhibitory effects of AAT on caspase-3, -8, and TNF-α should work synergistically to reduce liver damage.

As the effects on caspase-3 and -8 are not so profound in these models, perhaps AAT also influences RIP kinase-mediated necrosis, termed necroptosis. For necroptosis, it was shown that the deletion of the apoptotic executioner caspases-3 or -7 has no impact on TNF-induced systemic inflammatory response syndrome (SIRS)[42] and SIRS represents the most common cause of death in ALF.

TNF-receptor and Fas ligation are well-characterized processes leading to massive apoptosis in various cellular models.[21, 43] The Fas-mediated secretion of TNF-α aggravates apoptosis.[2] Second, Fas activation might increase hepatic chemokines that recruit TNF-α-secreting neutrophils to the liver.[44] TNF-α plays an important role in ALF, as supported by the observation that mice lacking TNF-α receptors 1 and 2 are resistant to death and fulminant liver injury induced by agonistic anti-Fas antibody.[22] As our experiments focused on the short-term effects, it is unlikely that the described inhibitory effect of AAT on TNF-α induced self-expression contributes significantly to the protection against ALF. We suggest that the inhibitory effect of AAT on ADAM17 (shown by us in the Jo-2 model and reported previously[34]) which cleaves various transmembrane proteins including TNF-α,[45] contributes to the reduction of TNF-α levels.

In summary, we show a novel protective effect of AAT therapy during acute liver injury and provide a framework for future studies addressing how exogenous AAT is internalized by liver cells and delays acute liver failure. Although the liver itself is a major source of AAT synthesis, local production of AAT may not be sufficient during liver damage, and exogenously applied AAT may exert an important role in the maintenance of hepatocyte survival, in part by modulating inflammatory processes. Our data also suggest investigating the therapeutic potential of AAT in conditions like acute hepatitis, as we envisage a protective impact by its immune modulatory functions.

Against the background of the demonstrated immune-regulatory and antiinflammatory properties of AAT (reviewed[46]), commercial preparations of human plasma AAT (such as Prolastin) are being increasingly studied in various models of acute inflammation. For example, novel studies demonstrate that therapy with AAT prevents murine islet cell allografts from rejection, suppresses alloreactivity in allogeneic marrow transplantation models, and protects the heart from ischemia-reperfusion injury in a mouse model of acute myocardial infarction. In other models, AAT therapy reduced TNF-α- or endotoxin-induced lethality.[47] Therefore, our data are consistent with the expanding knowledge of AAT application as a therapeutic option beyond augmentation therapy for emphysema. Commercially used AAT is already proven to be well tolerated, with minor negative effects.[32] Hence, AAT might represent a new therapeutic option for ALF of a variety of etiologies with the prospect of fast translation into clinical use.


The authors thank Helena Lickei for technical assistance with western blots and ELISA, Tetyana Yevsa for technical contribution and critical discussion and the Lars Zender Lab, the Sabina Janciauskiene Lab, and the Florian Kuehnel Lab for supporting N. Jedicke performing the experiments. We further would like to thank Grifols, Barcelona, Spain for providing us with Prolastin.