Human hepatocarcinoma-intestine-pancreas/pancreatitis-associated protein cures fas-induced acute liver failure in mice by attenuating free-radical damage in injured livers


  • Nicolas Moniaux,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM), U785, Centre Hépatobiliaire, Villejuif, France
    2. Université Paris-Sud, Faculté de Médecine, Villejuif, France
    Search for more papers by this author
  • Haiyan Song,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM), U785, Centre Hépatobiliaire, Villejuif, France
    2. Université Paris-Sud, Faculté de Médecine, Villejuif, France
    Search for more papers by this author
    • These authors contributed equally to this manuscript.

  • Marion Darnaud,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM), U785, Centre Hépatobiliaire, Villejuif, France
    2. Université Paris-Sud, Faculté de Médecine, Villejuif, France
    Search for more papers by this author
    • These authors contributed equally to this manuscript.

  • Kévin Garbin,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM), U785, Centre Hépatobiliaire, Villejuif, France
    2. Université Paris-Sud, Faculté de Médecine, Villejuif, France
    Search for more papers by this author
  • Michelle Gigou,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM), U785, Centre Hépatobiliaire, Villejuif, France
    2. Université Paris-Sud, Faculté de Médecine, Villejuif, France
    Search for more papers by this author
  • Claudia Mitchell,

    1. INSERM, U567, Institut Cochin, Université Paris Descartes, Paris, France
    Search for more papers by this author
  • Didier Samuel,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM), U785, Centre Hépatobiliaire, Villejuif, France
    2. Université Paris-Sud, Faculté de Médecine, Villejuif, France
    Search for more papers by this author
  • Laure Jamot,

    1. Alfact Innovation, Paris, France
    Search for more papers by this author
  • Paul Amouyal,

    1. Alfact Innovation, Paris, France
    Search for more papers by this author
  • Gilles Amouyal,

    1. Alfact Innovation, Paris, France
    Search for more papers by this author
  • Christian Bréchot,

    1. Institut National de la Santé et de la Recherche Médicale (INSERM), U785, Centre Hépatobiliaire, Villejuif, France
    2. Université Paris-Sud, Faculté de Médecine, Villejuif, France
    Search for more papers by this author
  • Jamila Faivre

    Corresponding author
    1. Institut National de la Santé et de la Recherche Médicale (INSERM), U785, Centre Hépatobiliaire, Villejuif, France
    2. Université Paris-Sud, Faculté de Médecine, Villejuif, France
    • INSERM U785, Hepatobiliary Center, Paul-Brousse Hospital, 94800 Villejuif, France
    Search for more papers by this author
    • fax: (33) 1 45 59 60 80

  • Potential conflict of interest: Nicolas Moniaux, Haiyan Song, Marion Darnaud, Kévin Garbin, Michelle Gigou, Claudia Mitchell, and Didier Samuel have no potential conflict of interest. Laure Jamot is an employee of Alfact Innovation. Paul Amouyal is Chairman of Alfact Innovation. Gilles Amouyal is CEO of Alfact Innovation. Christian Brechot is Vice-President at Mérieux Alliance. Jamila Faivre is a scientific consultant to Alfact Innovation.

  • J.F. received grant support from the Agence Nationale pour la Recherche (ANR-05-EMPB-005-02 and RPV07137LS), the Fondation de l'Avenir pour la Recherche Médicale Appliquée (ET7-473), the Association Française pour l'Etude du Foie (RAK08120LLA), and the Programme National de Recherche en Hépato-Gastroentérologie Inserm/AFEF (ASE07127LSA). N.M. was supported by grants from the Institut National de la Santé et la Recherche Médicale (INSERM) and the Institut National contre le Cancer (INCA).


Acute liver failure (ALF) is a rare syndrome with a difficult clinical management and a high mortality rate. During ALF, several molecular pathways governing oxidative stress and apoptosis are activated to induce massive tissue injury and suppress cell proliferation. There are few anti-ALF drug candidates, among which is the C-type lectin Reg3α, or human hepatocarcinoma-intestine-pancreas/pancreatitis-associated protein (HIP/PAP), which displayed promising properties for tissue regeneration and protection against cellular stress in transgenic mice. We report on substantial preclinical and clinical advances in the development of a recombinant (rc) full-length human HIP/PAP protein as an anti-ALF drug. The curative effects and mechanisms of action of rcHIP/PAP were investigated in murine Fas-induced ALF. Primary hepatocytes were cultured with cytotoxic doses of tumor necrosis factor α/actinomycin-D, transforming growth factor β, agonistic Fas antibody or hydrogen peroxide, and various concentrations of rcHIP/PAP. Cell viability, proliferation index, apoptosis, and oxidation were monitored. We found that rcHIP/PAP significantly improved survival in Fas-intoxicated mice in a dose-dependent and time-dependent manner, with optimum effects when it was injected at advanced stages of ALF. Primary hepatocytes were efficiently protected against multiple cell death signals by rcHIP/PAP. This survival benefit was linked to a depletion of oxidized biomolecules in injured liver cells due to a strong reactive oxygen species scavenging activity of rcHIP/PAP. Clinically, an escalating dose phase 1 trial demonstrated a good tolerability and pharmacokinetic profile of rcHIP/PAP in healthy subjects. Conclusion: The rcHIP/PAP protein exhibited significant curative properties against ALF in mice. It is a free-radical scavenger that targets a broad spectrum of death effectors and favors liver regeneration. The good safety profile of rcHIP/PAP during a phase 1 trial encourages evaluation of its efficacy in patients with ALF. (HEPATOLOGY 2011:53:618-627)

Acute liver failure (ALF) is a rare syndrome characterized by a sudden liver dysfunction of diverse etiologies and varying in course and outcome. This variability and the scarcity of biomarkers of disease progression make the management of ALF difficult.1 The global mortality rate is approximately 30%. During ALF, a number of molecular pathways governing oxidative stress and apoptosis are activated in concert to induce massive tissue injury and suppress liver cell proliferation. Liver oxidative stress follows from an increase in the production of reactive oxygen and nitrogen species (ROS/RNS), generated by the causal agents of ALF and exceeding the antioxidative capacity of the hepatocytes. The involvement of ROS in death signaling was established for the TNFα (tumor necrosis factor), TRAIL (TNF-related apoptosis-inducing ligand), and CD95 (APO-1/Fas) death receptors.2 Pharmaceutical care of patients with ALF remains problematic. Several molecules were found to counteract ALF induced by death ligands or toxics in animals, but their clinical evaluation has not yet given conclusive results.3-7 Liver transplantation remains the only recognized therapy of proven benefit against ALF, except for acetaminophen-induced ALF, which is treatable by a specific antidote, N-acetylcysteine (NAC). Other possible indications for NAC were recently suggested for non–acetaminophen-induced ALF,8 emphasizing the importance of oxidative stress in the pathogenesis of ALF.

Among anti-ALF drug candidates, the human secreted C-type lectin Reg3α (also called hepatocarcinoma-intestine-pancreas/pancreatitis-associated protein [HIP/PAP]) has a promising potential for tissue regeneration and protection against apoptosis and cellular stress. In animal models, HIP/PAP was shown to improve survival against pancreatic9 and hepatic6, 7, 10 inflammatory diseases, and promote axonal guidance in axotomized motor neurons.11In vitro, HIP/PAP conferred strong resistance against TNFα-induced apoptosis to pancreas and liver cells.10 Several survival pathways, such as the JAK/STAT3 and PI3/Akt pathways, were shown to be activated by HIP/PAP,11, 12 to exert these functions. Regarding ALF, it was shown that HIP/PAP transgenic mice were effectively protected against acetaminophen- and Fas-induced ALFs that were highly lethal in knockout mice.6, 7, 10 Parenteral administration of a recombinant HIP/PAP protein to wild-type mice prior to acetaminophen intoxication had a substantial effect against ALF and death.6 This finding represented a first step toward a pharmacological use of HIP/PAP. This report deals with the drug development of a cGMP recombinant full-length HIP/PAP protein called rcHIP/PAP. We first relate a preclinical assessment of the therapeutic value and mechanism of action of rcHIP/PAP. A single dose of rcHIP/PAP administered to wild-type mice with Fas-induced ALF resulted in a significant survival gain, which was correlated with a lower cell death rate, and a higher proliferation index. This therapeutic effect was obtained at advanced stages of ALF, and was associated with a substantial reduction of the concentration of oxidized biomolecules in damaged livers. In vitro studies established that rcHIP/PAP acts against several cell death effectors by scavenging the ROS generated during cell death. We also give account of a phase 1 (safety and pharmacokinetics) clinical trial of rcHIP/PAP establishing its tolerability up to a maximum recommended starting dose of 20 mg daily in human healthy volunteers, and thus warranting a phase 2 evaluation of rcHIP/PAP in patients with ALF.


ActD, actinomycin D; ALF, acute liver failure; ATP, adenosine triphosphate; BrdU, 5-bromo-2′-deoxyuridine; DNP, 2,4-dinitrophenylhydrazone; rcHIP/PAP, recombinant human hepatocarcinoma-intestine-pancreas/pancreatitis-associated protein; MDA, malondialdehyde; NAC, N-acetylcysteine; ROS, reactive oxygen species; SEM, standard error of the mean; TBA, thiobarbituric acid; TGFβ, transforming growth factor β; TNFα, tumor necrosis factor α.

Materials and Methods


Human recombinant TGF-β1 (P100-21) was purchased from Abcys, agonistic Fas antibody (Jo2), hereafter called Fas, from BD Biosciences, William's medium from Invitrogen, mouse recombinant TNFα (T7539), actinomycin D (A1410), H2O2 30% solution (H1009) and collagenase type IV (C5138) from Sigma Aldricht.

Recombinant Human HIP/PAP Protein (rcHIP/PAP).

The rcHIP/PAP protein is a new biological drug that corresponds to the addition of one amino-terminal methionine to the sequence of the secreted (i.e., lacking the 26–amino acid signal sequence) form of the human endogenous HIP/PAP (NP_620355). It was produced in Escherichia coli, purified to ≥99% and released in batches in compliance with the clinical grade manufacturing process by PX'Therapeutics.

Primary Hepatocyte Cultures.

Hepatocytes from 6-week-old C57Bl/6J mice (Charles River) were isolated by in situ collagenase perfusion of livers and plated in William's medium containing 10% fetal bovine serum. Cellular death was induced using TNFα/ActD (20 ng/mL / 50 ng/mL), Fas/ActD (40 ng/mL / 50 ng/mL), TGFβ (10 ng/mL), or with a 1 hour H2O2 1 mM pretreatment. Viability studies with rcHIP/PAP were performed when the cell death rates in stressed control cultures were of 55% ± 10%. Hepatocytes were stimulated with death inducers and rcHIP/PAP or an equivalent volume of buffer (vehicle). Cells were fixed and stained with Hoechst 33342 (5 μg/mL). Living and apoptotic cells were quantified using a Zeiss Axio Imager M1. Cell quantification was performed on 20 fields per slide. The normalized cell viability is defined as (nHIP-nCont)/(nBas-nCont), where nHIP, nCont and nBas are the numbers of living cells in rcHIP/PAP-treated, vehicle and nonintoxicated (basal) cultures at 24 hours after stress, respectively. Bromodeoxyuridine (BrdU) tests were performed using a BrdU labeling and detection kit (Roche). Cell counting was performed in 10 random microscopic fields. The cell proliferation index is defined as the ratio of the numbers of BrdU-positive nuclei in rcHIP/PAP-treated and vehicle-treated cultures.

Immunoblot Analysis.

Whole-cell lysates were prepared in ice cold buffer containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1% Nonidet P-40, 0.25% Na-deoxycholate, 1 mM Na3VO4, 20 mM NaF, 1 μg/mL aprotinin, 10 μg/mL pepstatin, 10 μg/mL leupeptin and 1 μM PMSF. Membranes were immunoblotted with antibodies against Bid, Flip, cleaved caspase-3, phospho-Akt, total Akt, Bcl-xL, phospho-Bad, total Bad (Cell Signaling Technology) and rcHIP/PAP (laboratory-made rabbit polyclonal antibody purified by affinity chromatography).

Mouse Model of ALF.

Six-week-old C57BL/6J mice (Charles River) were maintained for 1 week at 22°C before experiments. Animals received humane care in compliance with Institutional Guidelines. ALF was induced by intravenous injection of Fas at 170 μg/kg, which is a lethal dose for approximately 60% of the mice within 24 hours. Human rcHIP/PAP was intravenously injected following a dose response or a time course protocol. The half-life of rcHIP/PAP in the mice, as measured during regulatory preclinical studies, was of less than 2 hours. The control groups were injected with equivalent volumes of vehicle. Mice were monitored for survival at 24 hours, and the normalized survival increase, i.e., the ratio between the numbers of live animals in the rcHIP/PAP-treated and vehicle groups, was assessed. A semiquantitative scoring of hepatic lesions was performed in five livers per group (untreated, vehicle, and rcHIP/PAP) and five slices per liver. BrdU was administered to all surviving animals 2 hours before sacrifice at 48 hours after Fas intoxication. The cell proliferation index was assessed in 10 random microscope fields of tissue sections.

Measurement of Lipid Peroxidation and Protein Carbonylation.

Malondialdehyde (MDA), an end product of lipid peroxidation, was dosed using the thiobarbituric acid (TBA) procedure with the OxiSelect TBARS Assay kit (Cell Biolabs). Carbonyl groups of oxidized proteins were derivatized to 2,4-dinitrophenylhydrazone (DNP) by reaction with 2,4-dinitrophenylhydrazine using the OxyBlot Protein Oxidation Detection kit (Millipore).

Measurement of ROS Scavenging Activities.

Hydroxyl radicals were generated by the Fe3+-ascorbate-EDTA-H2O2 (Fenton) reaction. Hydroxyl radical scavenging activity was quantified through the degradation product of 2-deoxyribose condensed with TBA. Superoxide radicals were generated by the phenazine methosulfate-nicotinamide adenine dinucleotide (PMS/NADH) system. Superoxide radical scavenging activity was monitored by the reduction of nitro blue tetrazolium to purple formazan. Hydrogen peroxide scavenging activity was monitored using the FOX2 reagent after 30 minutes incubation of rcHIP/PAP protein with 750 μM H2O2.

Phase 1 Clinical Trial.

The study was designed as a first-in-man phase 1 clinical trial, and approved by the Persons Protection Committee of Ile-de-France in July 2009. The clinical trial authorization was granted by the French health products safety Agency (AFSSAPS) in August 2009. The trial was a monocentric, double-blind, placebo-controlled, randomized, ascending dose investigation. Thirty-two male volunteers between 18 and 45 years of age were included. All gave written informed consent prior to inclusion and were healthy according to medical history, medical examination and laboratory tests. The subjects received a 10-minute infusion of a single dose of rcHIP/PAP or placebo. The doses were of 0.105 (n = 4, group 1), 1.05 (n = 4, group 2), 3.99 (n = 8, group 3), 10.05 (n = 8, group 4) or 19.95 mg (n = 8, group 5). The number of patients receiving placebo was of one in groups 1 and 2, and two in groups 3 to 5. Safety variables and adverse events were recorded. Pharmacokinetic parameters were determined from plasma samples collected from the 24 rcHIP/PAP-treated subjects prior to dosing and at 0, 20, 30, 45, 60, and 90 minutes after dose, and at 2, 4, 6, 12, and 24 hours after dose. The data were subjected to noncompartmental pharmacokinetic analysis using WinNonLin and SAS softwares.

Statistical Analysis.

The statistical analyses were carried out using Graph-Pad software. The data from in vitro studies were analyzed using Kruskal-Wallis'/Dunn's, two-way ANOVA/Bonferroni's and Mann-Whitney tests. The survival data, which include their own control group, were analyzed using Fischer's exact test.


rcHIP/PAP Cures Murine Fas-Induced ALF.

We investigated the therapeutic potential of rcHIP/PAP against ALF in wild-type C57BL6/J mice injected with lethal doses of agonistic anti-CD95/Fas antibody (Fas). Single intravenous injections of rcHIP/PAP at doses ranging from 50 to 2500 μg/kg were made at the fixed time of 6 hours after intoxication. The normalized survival increase was of 9%, 39% (P < 0.02), 66% (P < 0.003), and 41% (P < 0.01) at the doses of 50, 250, 750, and 2500 μg/kg of rcHIP/PAP, respectively (Table 1 and Fig. 1A). It is seen that rcHIP/PAP improved 24-hour survival in a dose dependent manner, the most efficient dose being of 750 μg/kg. To determine the therapeutic time window, rcHIP/PAP was then administered at the fixed dose of 750 μg/kg at 2, 4, 6, 9, or 11 hours after intoxication. The effect of rcHIP/PAP on the 24-hour survival was insignificant after injection at 2 hours and 4 hours post intoxication, but substantial at later injection times. The normalized survival increases were of 66% (P < 0.008), 75% (P < 0.018) and 122% at 6, 9, and 11 hours after intoxication, respectively (Table 2 and Fig. 1B). The fact that the survival increase at 11 hours after intoxication does not appear statistically significant is probably only due to the very small number of surviving animals prior to rcHIP/PAP injection at this time point. To sum up, these observations demonstrated a strong curative effect of a single injection of rcHIP/PAP performed at an advanced stage of Fas-induced acute liver failure.

Figure 1.

rcHIP/PAP increases survival of wild-type mice bearing a Fas-induced ALF. (A) Normalized survival increase at 24 hours in Fas-intoxicated mice injected with rcHIP/PAP 6 hours after intoxication at the indicated doses. For cohort sizes, see Table 1. *P < 0.05 or **P < 0.01. (B) Normalized survival increase at 24 hours in mice injected with 750 μg/kg of rcHIP/PAP at the indicated injection times after intoxication. For cohort sizes, see Table 2. *P < 0.05 or **P < 0.01. (C) BrdU incorporation in liver tissues from Fas-intoxicated mice injected with 750 μg/kg of rcHIP/PAP or vehicle at 6 hours post intoxication. Left: BrdU immunolabeled sections. BrdU-positive nuclei are green, total nuclei are blue (Hoechst dye). Scale bar: 50 μm. Right: Normalized proliferation index in liver tissue sections as a function of the rcHIP/PAP injection time after Fas intoxication. *P < 0.05 or **P < 0.01.

Table 1. Dose-Response Study of the 24-Hour Survival of Fas-Intoxicated Wild-Type Mice Treated 6 Hours After Intoxication With a Single Dose of rcHIP/PAP or an Equivalent Volume of Vehicle
Injected Dose (μg/kg)TreatmentAnimalsAlive at 24 HoursSurvival RatioP Values
  • *

    P < 0.05,

  • **

    P < 0.01.

Table 2. Time-Response Study of the 24-Hour Survival and Liver Proliferation Index of Fas-Intoxicated Wild-Type Mice Treated 6 Hours After Intoxication With a Single Dose of rcHIP/PAP (750 μg/kg) or an Equivalent Volume of Vehicle
Injection Time (Hours)TreatmentAnimalsAlive at 24 HoursSurvival RatioP ValueBrdU-Positive NucleiProliferation IndexP Value
  • Proliferation index is the ratio of the numbers of BrdU-positive nuclei in rcHIP/PAP-treated, and vehicle-treated mice at 48 hours after intoxication.

  • *

    P < 0.05,

  • **

    P < 0.01. NA, insufficient sample size.


The gain in 24-hour survival due to rcHIP/PAP was preserved at the 48-hour post intoxication time point, which corresponds to the proliferation peak of liver cells in the model under study. We measured the proliferation index at 48 hours in livers from mice injected with 750 μg/kg of rcHIP/PAP or vehicle at different times post intoxication. In all the studied groups, rcHIP/PAP-treated mice harbored a statistically significant increase in 48-hour proliferation index compared to control mice (Fig. 1C and Table 2). BrdU-labeled hepatocytes were scattered throughout the hepatic lobes, predominantly in the midzonal region. The normalized ratio of BrdU-positive nuclei in rcHIP/PAP treated mice was of 1.65 (P < 0.01), 1.96 (P < 0.004), 2.24 (P < 0.04), 2.12 (P < 0.009), and 2.14 for times of rcHIP/PAP injection of 2, 4, 6, 9, and 11 hours after intoxication, respectively. Thus, rcHIP/PAP effectively promoted regeneration in injured livers regardless of its injection time. It is worth noting that rcHIP/PAP injected at an early stage of Fas-induced ALF did not significantly improve survival, despite a substantial mitogenic activity.

rcHIP/PAP Decreases Lipid Peroxidation in Fas-Injured Mouse Livers.

To cast light on the molecular mechanisms underlying the curative action of rcHIP/PAP, we studied the modulation, by rcHIP/PAP, of the levels of cell survival, proliferation or apoptosis markers and oxidized biomolecules in Fas-intoxicated livers and primary hepatocytes. Fas-intoxicated mice were treated with 750 μg/kg of rcHIP/PAP at 6 hours post intoxication. Control mice received an equivalent volume of buffer. Mice were sacrificed 2 hours after injection of rcHIP/PAP or buffer, and livers analyzed for Flip, BclxL, Bid, caspase-3, Bad, Akt, Erk. Figure 2A illustrates the macroscopic time evolution of livers from Fas-intoxicated mice without any treatment. The first signs of damage appeared at 4 hours after intoxication, and then the livers became globally dark red, indicative of massive necrosis and hemorrhage. At 6 hours, the calculated area of living hepatocytes was of 36%. At 8 hours, it was of 18% in vehicle-injected mice, whereas the livers that were treated with rcHIP/PAP for 2 hours had kept their macroscopic and histological aspect unchanged with 38% of living hepatocytes (P = 0.022; Fig. 2B,C). Immunoblots revealed a lesser activation of cleaved caspase-3 in rcHIP/PAP-treated than in control livers, and no significant difference in activation or expression for any of the other studied markers (Fig. 2D and data not shown). These molecules were not modulated by rcHIP/PAP in livers collected 24 and 48 hours after intoxication (data not shown). We then measured the concentration of MDA in livers, which were intoxicated with Fas for different duration times, and treated for 2 hours with rcHIP/PAP or buffer prior to the sacrifice of mice. In Fas-induced ALF, the relative increase in hepatic MDA concentration with respect to the baseline value remained essentially constant at approximately 70% until 6 hours, and rose rapidly to approximately 150% in the following 2 hours. A 2-hour treatment with rcHIP/PAP significantly reduced the increase in the MDA level only when it was applied at 6 hours. Then the MDA level at 8 hours was 45% smaller in rcHIP/PAP-treated than in buffer-injected mice (P < 0.0029) (Fig. 2E). Overall, these observations indicate that the survival gain provided by a late injection of rcHIP/PAP in Fas-induced ALF mostly resulted from an antioxidant activity of the protein counteracting the dynamics of ROS production in this model. The lipid peroxidation kinetic measurements showed a late rise of oxidative damage correlated with a progression of necrotic inflammatory lesions during Fas-induced ALF suggesting the existence of an oxidative-damage threshold, above which the liver destruction process is dramatically accelerated. The lack of protective effect of early rcHIP/PAP injections shows that rcHIP/PAP is only effective at the moment of the threshold crossing.

Figure 2.

rcHIP/PAP acts as a major antioxidant against induced ALF. Six hours after Fas intoxication, mice were treated with 2500 μg/kg of rcHIP/PAP or vehicle for 2 hours. (A) Time series of macroscopic views of livers during untreated Fas-induced ALF. (B) Macroscopic views of livers and hematoxylin and eosin–stained tissue sections. Boundaries between necrotic inflammatory areas (pink) and living areas (purple) are delineated with broken lines. PT: portal tract. Scale bar: 50 μm. (C) Percentage of living hepatocytes in the indicated groups of animals at different time points. (D) Immunoblots using the indicated markers of apoptosis in liver extracts. rcHIP/PAP and actin immunoblots are shown as controls. (E) Concentration of MDA, a lipid peroxidation product, in liver extracts at five time points of untreated Fas-induced ALF and 2 hours after injection of rcHIP/PAP or vehicle at the 6-hour time point. Means of two independent experiments ± standard error of the mean (SEM). **P < 0.01.

rcHIP/PAP Protects Primary Hepatocytes Against Multiple Cell Death Signals.

To substantiate the conjecture of an antioxidant function of rcHIP/PAP, we investigated the rcHIP/PAP-induced survival of primary hepatocytes subjected to lethal doses of TNFα/ActD, TGFβ, or Fas/ActD, and escalating doses of rcHIP/PAP. H2O2 was provided during 1 hour before rcHIP/PAP addition. A significant overall beneficial effect of rcHIP/PAP treatment was found for hepatocytes stressed with H2O2, TGFβ, and TNFα/ActD (Fig. 3A). The P values of viable cell counts cumulated over rcHIP/PAP doses were the following: H2O2 (P = 0.015), TGFβ (P = 0.011), and TNFα/ActD (P = 0.014) in mice; H2O2 (P = 0.011), TGFβ (P = 0.019), and TNFα/ActD (P = 0.019) in rat. The normalized cell viability was rcHIP/PAP dose dependent culminating at 55% (P < 0.01) at 1 μg/mL for H2O2, 26% (P < 0.01) at 4 μg/mL for TGFβ, and 23% (P < 0.01) at 4 μg/mL for TNFα/ActD in mice, and of 91% (P < 0.01) at 4 μg/mL for H2O2, 63% (P < 0.05) at 4 μg/mL for TGFβ, and 60% (P < 0.05) at 4 μg/mL for TNFα/ActD in rats (Fig. 3B). No viability gain due to rcHIP/PAP was observed in FasL/ActD-stressed cells, indicating that rcHIP/PAP does not interfere with the Type I Fas/CD95 death signaling pathway activated by Fas in cultured primary hepatocytes.13 To sum up, these results show that rcHIP/PAP protects liver cells against a number of different death pathways. This fact suggests that rcHIP/PAP displays its cytoprotective properties through a signaling cascade that involves mitochondria, the central hubs for apoptotic and nonapoptotic death signals.

Figure 3.

rcHIP/PAP protects murine primary hepatocytes from apoptotic and nonapoptotic death. (A) Hoechst-stained nuclei of nonstressed (basal) hepatocytes and hepatocytes stressed with the indicated inducers and cultured with vehicle or rcHIP/PAP. Scale bar: 50 μm. (B) Normalized cell viability at 16 hours after stress with the indicated inducers for different doses of rcHIP/PAP. Means of three independent experiments ± SEM (C) Normalized cell viability of TNFα/ActD-stressed or H2O2-stressed hepatocytes treated with increasing doses of rcHIP/PAP alone or in combination with 30 mM of NAC. Means of three independent experiments ± SEM. *P < 0.05 or **P < 0.01 or ***P < 0.005 versus control.

Next, we compared the cellular-protection efficiency of rcHIP/PAP to that of NAC, which is the current gold standard treatment of acetaminophen-induced ALF. Murine hepatocytes stressed with H2O2 or TNFα/ActD were treated with rcHIP/PAP and NAC alone or a combination. rcHIP/PAP alone rescued cells from both the TNFα/ActD and H2O2 lethal challenges, as mentioned above, whereas NAC alone restored viability of H2O2-stressed cells, but not of TNFα/ActD-challenged cells. The two drugs had similar quantitative efficiencies against H2O2, given separately or in combination (Fig. 3C). The efficiency of rcHIP/PAP against TNFα/ActD was not significantly altered by NAC. This indicates that rcHIP/PAP may be used in conjunction with NAC in the treatment of ALF.

rcHIP/PAP Is a Free-Radical Scavenger.

Primary hepatocytes challenged with H2O2, TGFβ, or TNFα/ActD were treated with 4 μg/mL of rcHIP/PAP. No significant difference in the levels of expression and activation between rcHIP/PAP-treated and control cells was observed for any of the studied apoptotic molecules at any time point, suggesting that the survival benefit provided by rcHIP/PAP implies no direct modulation of the apoptotic paths (Fig. 4A). An activation of Akt and Erk rapidly developed in response to lethal challenges but no significant difference in their expression/activation levels between rcHIP/PAP-treated and control cells was observed (Fig. 4A). Next we investigated the influence of rcHIP/PAP on cell proliferation by performing time course and dose response experiments in primary hepatocytes using BrdU incorporation assays. Figure 2B shows that rcHIP/PAP had no significant mitogenic activity at 24, 48, and 72 hours after plating, indicating that the strong enhancement of cell viability observed at 24 hours after plating was not attributable to an increase in cell proliferation. The mitogenic activity of rcHIP/PAP appeared only at the final time of 96 hours, with a significant increase in proliferation index at the doses of 0.5 (P < 0.01), 1 (P < 0.001) and 4 μg/mL (P < 0.001) of rcHIP/PAP (Fig. 4B). We measured the concentrations of oxidized lipids and carbonylated proteins in primary hepatocytes stressed with H2O2, TNFα/ActD or Fas/ActD and incubated with rcHIP/PAP or buffer for different duration times. The level of lipid peroxidation in buffer-incubated cells displayed a rapid increase during the first hour during exposure to TNFα/ActD and H2O2 (Fig. 4C). The rcHIP/PAP-stimulated cells displayed no increase in lipid peroxidation level on exposure to TNFα/ActD and H2O2 stresses. The Fas/ActD induced death was not associated with any change in lipid peroxidation in vitro, suggesting that it was independent of mitochondrial change. The amount of total carbonylated proteins was clearly smaller in rcHIP/PAP-stimulated than in control cells after a 8 hours of exposure to TNFα/ActD or H2O2 (Fig. 4D). We investigated the influence of rcHIP/PAP on c-Jun N-terminal kinase (JNK) activation, which is known to be sustained by ROS via MAP kinase phosphatases inhibition,14 in TNFα/ActD-stressed primary hepatocytes. Stimulation by rcHIP/PAP decreased TNFα/ActD-induced JNK activation by 25% at all incubation times (Fig. 4E). Finally, we measured the concentrations of O2 and OH free radicals and of a nonradical ROS (H2O2) in oxidation reagents as a function of rcHIP/PAP concentration. The positive controls used were: a 20 mM solution of ascorbic acid that exhibited 20% of scavenging activity against O2, a 30 mM solution of glycerol that exhibited 90% of activity against OH, and a 20 mM solution of sodium pyruvate that displayed 87% of activity against H2O2. We found that rcHIP/PAP acted as a scavenger for O2 and OH in a dose-dependent manner (Fig. 4F). At the highest dose of 18 μM, rcHIP/PAP displayed quenching activities of 16% against O2 and 42% against OH, and no detectable scavenging activity against H2O2. These results demonstrate a sufficiently high free-radical scavenging capability of rcHIP/PAP to account for most of its curative properties against ALF.

Figure 4.

rcHIP/PAP is a free-radical scavenger that inhibits oxidative damage and favors liver cell proliferation. Murine primary hepatocytes were stressed with different death inducers and treated with 4 μg/mL of rcHIP/PAP or a corresponding volume of vehicle for different incubation durations. (A) Immunoblots performed with the indicated markers of apoptosis and cell survival in TNFα/ActD-stressed, H2O2-stressed, and TGFβ-stressed hepatocytes. Immunoblots of rcHIP/PAP and actin are shown as controls. (B) Left: cell proliferation index of non-stressed hepatocytes incubated with BrdU and rcHIP/PAP at the indicated doses for different incubation durations. Means of three independent experiments ± SEM. *P < 0.05 versus control. Right: Hoechst dye-stained rHIP/PAP-treated and vehicle-treated hepatocytes at 72 hours of incubation. Arrows: mitotic cells. Scale bars: 20 μm. (C) Normalized quantity of MDA, a lipid peroxidation marker, in TNFα/ActD-, H2O2-, and Fas/ActD-stressed hepatocytes for different incubation durations of rcHIP/PAP. Time 0 corresponds to nonstressed hepatocytes. The measurements were normalized to nonstressed vehicle-treated cultures. Means of three independent experiments ± SEM. *P < 0.05 versus control. (D) Anti-DNP immunoblots displaying total carbonylated proteins from TNFα/ActD-stressed and H2O2-stressed hepatocytes for different incubation durations of rcHIP/PAP. (E) Left: immunoblots of total and phosphorylated JNK from TNFα/ActD-stressed hepatocytes. Right: densitometry quantification of the immunoblots. (F) ROS scavenging activities of rcHIP/PAP. The data represent the relative decrease in concentration of the indicated ROS following additions of rcHIP/PAP at different concentrations. Means of two independent experiments ± SEM.

rcHIP/PAP Is a Safe and Well-Tolerated Drug Candidate.

A phase 1 clinical trial was conducted to evaluate the safety, tolerability and pharmacokinetic parameters of the full-length rcHIP/PAP protein given in a single intravenous dose. Healthy volunteers received a 10-minute infusion of doses of rcHIP/PAP ranging from 0.105-19.95 mg. Safety variables and adverse events were recorded. Pharmacokinetic parameters were determined from plasma samples collected at various postdose times (Table 3). No serious adverse event, no adverse event requiring the withdrawal of a subject and no out-of-range values of laboratory parameters were recorded. The maximum plasma concentrations were observed at the end of infusion. The exposures, measured by the area under curve (AUC), and maximal plasma concentrations increased in a dose-dependent manner, and were not significantly different among dose groups. The inter-individual variability was low. The terminal plasma half-life was of approximately 4 hours at all doses. Clearance and volume distribution were close to 4 L/hour and 25 L, respectively. Overall, these data indicate a good safety profile of the full-length rcHIP/PAP protein.

Table 3. Summary of Mean (±SEM) Plasma Pharmacokinetic Parameters Following Single Intravenous Administration of rcHIP/PAP to Healthy Human Subjects
ParameterDose (mg)
  1. AUC0-∞, integral of plasma concentration; Cl, clearance; Cmax, peak plasma concentration; ND, not determined; tmax, time required for the plasma concentration to reach Cmax; t1/2, terminal plasma half-life; Vd, volume of distribution.

Subjects (n)33666
Cmax (ng/mL)ND246 ± 35758 ± 482389 ± 1253943 ± 207
tmax (h)ND0.
t1/2 (h)ND3.684.734.774.21
AUC0-∞ (ng.h/mL)ND237 ± 31819 ± 292228 ± 754310 ± 261
Cl (L/h)ND1.6 ± 1.13.8 ± 0.83.6 ± 0.63.8 ± 0.4
Vd (L)ND12 ± 1124 ± 525 ± 423 ± 3


We have shown that the recombinant rcHIP/PAP protects liver cells from death mediated by the TNFα and Fas/CD95 pathways, TGFβ or H2O2 oxidative stress. The rcHIP/PAP-induced viability gain of primary hepatocytes subjected to these different types of cellular stress was substantial, and increased as the added dose of rcHIP/PAP increased. A single injection of rcHIP/PAP to mice bearing a Fas-induced ALF resulted in a significant survival gain that correlated with less necrotic inflammatory lesions, a lower cell death rate and a higher proliferation index. These curative effects were observed when rcHIP/PAP was administered at an advanced stage of ALF when massive oxidative stress was rapidly leading to liver necrosis. In our model of Fas-induced ALF, ROS accumulate until a lesion threshold occurring at approximately 6 hours. Our recombinant protein has a half-life of less than 2 hours. It appears that rcHIP/PAP influences the outcome of the disease only if it is administered less than 2 hours before crossing the threshold. This points to the dynamic nature of the balance between the accumulating oxidative damage and the antioxidative activities (including the one induced by rcHIP/PAP) of the liver. Reportedly, spontaneous recovery from ALF is associated with apoptotic cell death, whereas the most severe forms of ALF are characterized by nonapoptotic cell death.15 The capability of rcHIP/PAP to fight both apoptotic and nonapoptotic death signals suggests that it should be an efficient drug against the multiple forms of ALF, including the most severe ones.

Moreover, we have evidence that the broad cytoprotective activity of rcHIP/PAP in the liver was associated with a reduction of the concentration of oxidized proteins and lipids, but not with significant modulations of molecules belonging to key apoptosis and survival pathways. The attenuation of the oxidative damage by rcHIP/PAP resulted from its scavenging activity against two deleterious ROS, namely, the superoxide and hydroxyl free radicals, which are abundantly generated during the numerous cell death processes activated in ALF. A similar ROS scavenging activity involved in the protection of murine hematopoietic stem cells has recently been reported for dietary lectins.16 The chemical reactions that are involved in the ROS scavenging activity of rcHIP/PAP remain to be studied. The endogenous HIP/PAP binds divalent cations, such as zinc and copper,17 which may allow rcHIP/PAP to act as a superoxide dismutase (SOD)-like enzyme or exchange these cations with SOD, favoring the activity of the latter. Catalase-like or glutathione reductase-like activities are also possible.

It should be noted that the observed mitogenic and antiapoptotic properties of rcHIP/PAP in injured livers are likely to be due to its antioxidant activity preserving the functionality of proteins involved in cell survival, apoptosis and proliferation pathways. The main therapeutic goal to be assigned to an anti-ALF drug is to slow down the loss of hepatocellular functions long enough to allow liver regeneration to occur. In experimental models of ALF, rcHIP/PAP and other antioxidant molecules1, 7, 8 act to maintain liver cells in a functional state by drastically reducing oxidative damage. Therefore, they constitute a promising class of anti-ALF drugs. In this category, specific features of rcHIP/PAP are a broad spectrum of targeted death effectors, and a mitogenic activity supporting liver regeneration, which, theorically, should allow rcHIP/PAP to prevent the progression of acetaminophen-induced as well as non–acetaminophen-induced ALF. This, combined with the good clinical safety profile of rcHIP/PAP found during this study, led us to design a multicentric phase 2 efficacy and safety trial in ALF, which is currently in progress.