Introduction

Acute liver failure (ALF) is a clinical syndrome characterized by peripheral vasodilation, encephalopathy and coagulopathy culminating in multi-organ dysfunction and death [1]. There are two main causes for ALF in Europe: hepatitis and medical intoxication [2, 3]. In the USA and other western countries the most frequent cause for ALF with about 40% is the acetaminophen (acetyl-para-aminophenol, APAP)—intoxication [2, 4,5,6,7]. Especially, the combination of drugs with APAP leads to an unintentional and chronic APAP overdose [8].

The pathomechanisms of APAP are still not completely understood [9,10,11,12,13,14,15]. It is widely accepted that APAP is metabolized in the liver by cytochrome P450 family to N-acetyl-p-benzoquinone (NAPQI), which consumes glutathione [12]. When this pathway is saturated, NAPQI generates covalent protein adducts especially on mitochondrial proteins and leads to mitochondrial dysfunction [12, 16,17,18]. These protein adducts seem to be the most relevant cause for APAP toxicity [19,20,21]. Further, a sterile inflammation and also a perfusion deficit are described [22, 23], which may lead to APAP-induced liver damage. Until now, it is not clear which form of cell death prevails in APAP toxicity. Accordingly, there are studies, which show that APAP-induced liver cell death is commonly brought upon by necrosis [15, 24,25,26], however some studies postulate that apoptosis may occur as well [27, 28].

Accordingly, APAP-induced apoptosis is separated in two major pathways, the death receptor-dependent or extrinsic pathway and the mitochondrial-dependent or intrinsic pathway [29,30,31]. It is characterized by ATP-dependent biochemical mechanisms and apparent morphological changes such as cell shrinking, DNA fragmentation and membrane budding [31]. The apoptotic pathway is linked to the BH3-interacting domain death agonist (Bid), a BH3 only protein, which belongs to the Bcl-2 family and has a pro-apoptotic effect. Bid is commonly activated through death receptor dependent caspase activation, especially caspase-8 [32] or, as a recent study suggested [33], also by other pathomechanisms including mitochondria-related oxidative stress. Truncated Bid, tBid, translocates to the mitochondria. However, at present the molecular events leading to APAP-induced liver injury and failure are not fully understood [15, 26, 34]. It is known that Bid transfers the peripheral apoptotic signals through direct or indirect mechanisms [35,36,37,38]. The most noted way to pass on apoptotic signals is the activation of the pro-apoptotic Bcl-2 family members Bak and Bax [39,40,41], followed by cytochrome c release, which is located in the inner mitochondria membrane. This then deteriorates the respiratory chain and consequently induces oxidative stress. Therefore, mitochondria seem to play an important role in APAP-induced liver damage.

Several studies already showed strategies to inhibit APAP-induced ALF, which however mainly focused on the prevention of APAP-induced liver apoptotic cell death by gene manipulating approaches [28, 42, 43]. For instance, Badmann et al. [43] demonstrated that gene ablation of the pro-apoptotic Bim protein (another Bcl2 family member) substantially protected mice from APAP-induced liver injury [43].

Furthermore, it was shown that liver sinusoidal endothelial cells of Bid knockout mice are significantly protected against APAP toxicity [42]. Also silencing of Bid expression in vitro with short interfering RNA (siRNA) protected liver sinusoidal endothelial cells from cytotoxicity and support the therapeutic hypothesis, that targeting Bid could prevent apoptosis-dependent liver injury [42]. SiRNA therapeutics hold great therapeutic potential, as it is now possible to design and chemically synthesize siRNAs for the safe and efficient targeting of specific genes by RNA interference [44,45,46]. In addition, targeting gene expression by RNA interference is transient, and can be applied ad hoc in adult mice, thereby reducing the effects of compensatory gene regulation or variable genetic background associated with studies using gene-modified organisms. Therefore, we investigated whether silencing Bid in the liver by short term siRNA application would have hepatoprotective effects in vivo. To inhibit Bid expression in the liver, Bid targeting siRNAs were formulated with a liver-specific siRNA delivery vehicle, DBTC [47] and tested in a murine model for APAP-induced liver injury.

Materials and methods

Animals and in vivo experiments

Male C57BL/6 J mice (Charles River Laboratories, Sulzfeld, Germany) were used at 6–8 weeks of age with a body weight of approximately 20–30 g. Animals were provided water and standard laboratory chow ad libitum. During the night before APAP application, mice were fasted to reduce hepatic glutathione levels. The experimental protocol was approved by the local committee (LALLF 7221.3-1.1-016/14) and all animals received human care according to the German legislation on protection of animals and the Guide for the Care and Use of Laboratory Animals (NIH publication 86–23 revised 1985).

Mice (n = 60) were injected either with in phosphate-buffered saline (PBS) dissolved APAP purchased from SIGMA (99% pure; 300 mg/kg body weight intraperitoneally (bw i.p.) for induction of ALF (n = 30) or with PBS (n = 30) as control and were studied 6 h thereafter (+ 6 h) as elucidated in Fig. 1a. The animals received 48 h prior to injection of either APAP or PBS a liver-specific small interfering RNA delivery system (DBTC lipoplex), prepared either with Bid siRNA (DBTC/siRNABid, n = 20), with non-targeting control siRNA (DBTC/siRNALuci, n = 20) or with vehicle (vehicleHEPES, n = 20), which served as controls regarding the silencing regimen. For induction of ALF, concentrations of APAP and PBS were used in accordance with work published previously by other groups [34, 48]. The choice to pre-treatment 48 h prior APAP induction is based on the previous study of our group [49]. Since APAP-induced liver cell death was maximally pronounced at + 6–12 h [34], we used this time point for the readout of liver damage.

Oligonucleotides

The siRNA molecules used in this study are blunt-ended, double-stranded RNA oligonucleotides (Table 1), stabilized by alternating 2′-O-methyl or 2′-F-fluoro modifications, respectively, on both strands as previously described [50, 51] and were synthesized by Biospring (Frankfurt a.M., Germany). For in vivo application, siRNAs were formulated with DBTC, a liver specific siRNA delivery system developed by Silence Therapeutics GmbH for targeting specifically the liver [47].

Table 1 List of siRNA molecules: mA, mU, mC, mG: 2´O-Methyl RNA and fA, fU, fC, fG: 2′deoxy-2′-fluoro RNA
Full size table

Intravital fluorescence microscopy

For in vivo analysis of hepatocellular apoptosis, intrahepatic leukocyte accumulation and sinusoidal perfusion failure, fluorescence microscopy was performed 6 h after APAP exposure in ketamine/xylazine-anesthetized animals (75/25 mg/kg [bw, ip]) in accordance with work previously published by our group [52, 53]. Further details are provided in supplemental materials and methods.

Sampling and assays

After in vivo microscopy, animals were exsanguinated by puncture of the vena cava inferior for immediate separation of EDTA plasma. The degree of hepatic disintegration was assessed by spectrophotometric determination of plasma alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activities using commercially available reaction kits (Roche Diagnostics, Mannheim, Germany). Cyclophilin A (indicator for necrotic cell death) and HMGB1 (indicator for necrotic cell death and sterile inflammation) were measured using ELISA Kits according to the manufacturer´s instructions (HMGB1: IBL International GmbH, Hamburg, Germany and cyclophilin A: LSBio, LifeSpab BioScience, Inc., Seattle, WA). Liver tissue was sampled for Western blot, real time-PCR as well as hepatic GSH analysis and histology.

Real time-PCR (RT-PCR) analysis

Approximately 20 mg of tissue was homogenized in a Mixer Mill MM 301 (Retsch GmbH, Haan, Germany) using tungsten carbide beads (Qiagen). Total RNA was isolated via the Invisorb Spin Tissue RNA Mini Kit (Invitek, Berlin, Germany). Depending on the tissue, 25–100 ng total RNA was used for quantitative TaqMan Real Time (RT)-PCR with the amplicon set]s (for Bid, cytochrome P450 2E1 (cyp2E1), ApoB, see Table 2) obtained from BioTez GmbH, Berlin, Germany: The TaqMan RT-PCR reactions were carried out with an ABI PRISM 7700 Sequence Detector (Software: Sequence Detection System v1.6.3 (ABI)) or StepOnePlus™ RT-PCR Sytem (ABI) using a standard protocol for RT-PCR as described previously [50] and probes at a concentration of 300 and 100 nM respectively. TaqMan data were calculated by using the comparative CT method.

Table 2 List of amplicon sets
Full size table

Western blot analysis

Target protein expression was assessed by Western blotting of whole tissue lysates. Snap frozen tissues were homogenized in lysis buffer (10 mM Tris pH 7.5, 10 mM NaCl, 0.1 mM EDTA, 0.5% Triton-X 100, 0.02% NaN3), and 0.2 mM PMSF (a protease inhibitor cocktail), incubated for 30 min on ice and centrifuged for 10 min at 4 °C and 10,000×g. Protein contents were assayed by the bicinchoninic acid method (Pierce Biotechnology) with 2.5% BSA (Pierce Biotechnology) as standard. On 14% SDS gels, 40 µg protein from liver tissue was separated and transferred to a polyvinyldifluoride membrane (Immobilon-P; Millipore). After blockade with 2.5% BSA (Pierce Biotechnology), membranes were incubated overnight at 4 °C with following antibodies: a mouse monoclonal anti-Bax (1:250, BD Pharmingen, Heidelberg, Germany), a mouse monoclonal anti-bcl2 (1:500, BD Pharmingen), a mouse monoclonal anti-Bid (1:1000, Santa Cruz Biotechnology, Texas, USA) and a mouse monoclonal anti-ß-actin (1:20,000; Sigma-Aldrich, Taufkirchen, Germany) or rabbit alpha-actinin antibody (1:500, Cell Signaling, Frankfurt) for loading control. Afterwards, secondary peroxidase-linked anti-mouse antibodies (Bax and Bcl2; 1:20,000; ß-actin; 1:60,000) or anti-rabbit antibodies (1:40,000) were applied. Protein expression was visualized by means of luminol-enhanced chemiluminescence (ECL plus; Amersham Pharmacia Biotech) and digitalized with ChemiDoc™ XRS System (Bio-Rad Laboratories). Signals were densitometrically assessed (Quantity One; Bio-Rad Laboratories) and normalized to ß-actin. Bid and alpha-actinin blots were analysed using Stella camera system and AIDA image analyser software 4.25 from Raytest (Mannheim, Germany).

Hepatic glutathione (GSH) analysis

For measurement of hepatic GSH content, livers were homogenized with 50 mM phosphate puffer containing 1 mM EDTA (pH 6-7) and deproteinized with metaphosphoric acid and triethanolamine. The GSH content was analysed by using the glutathione assay kit method according to the manufacturer’s instructions (Cayman Chemical Company, MI, USA) and is given as µmol/g liver tissue.

Histology

Liver tissue was fixed in 4% phosphate-buffered formalin for 2–3 days and then embedded in paraffin. 4 µm sections were fixed on glass slides and stained with haematoxylin and eosin (H & E). For histomorphometric analysis of necrotic tissue images of twenty random low-power fields (× 10 magnification, Olympus BX 51, Hamburg, Germany) were acquired with a Colour View II FW camera (Colour View, Munich, Germany). The quotient of the focal necrosis surface to the total liver section area was assessed and given in percent.

Statistical analysis

All data are expressed as means + SEM. For statistics, one-way analysis of variance, including all groups, was used to assess significant differences between groups. Subsequently, post hoc pairwise comparison test including Bonferroni correction for multiple comparisons was applied to identify which group differs to each other. Data were considered significant if p < 0.05. Statistical analysis was performed using the SigmaStat software package (Jandel Corporation, San Rafael, CA, USA). The results were presented with the program SigmaPlot 11.0 (Jandel Corporation, San Rafael, CA, USA).

Results

Effects of Bid silencing in mice treated with APAP

In order to evaluate whether hepatic silencing of Bid was effective, Bid mRNA and protein levels were assessed in liver samples collected 54 h after mice were treated with DBTC/siRNABid formulation by i.v. administration (Fig. 1b, c). At this time point, Bid mRNA levels were significantly decreased compared to the control groups treated with the vehicleHEPES or with non-targeting control lipoplex, DBTC/siRNALuci (Fig. 1b). Similarly, hepatic Bid protein levels were markedly decreased in siRNABid-versus vehicleHEPES-treated mice as shown by a representative Western blot (Fig. 1c). The quantitative densitometric analysis revealed an about 70% reduction of Bid protein levels (siRNABid vs. vehicleHepes n = 8 each, p = 0.013, data not shown). Furthermore, siRNABid treatment attenuated the APAP-induced upregulation of Bax protein expression, as shown by an only twofold rise of Bax protein expression versus a three to fourfold rise in siRNALuci and vehicleHEPES pre-treated mice (Fig. 2a, p < 0.05 vs. siRNALuci). Bcl2 protein expression was only slightly increased in livers of siRNABid pre-treated mice while APAP caused a 1.5-to twofold rise in siRNALuci and vehicleHEPES pre-treated mice (Fig. 2b). To evaluate whether Bid silencing before APAP exposure has the potential for off-target effects, we investigated hepatic cyp2E1 mRNA expression and hepatic GSH content. APAP exposure led to a reduction of relative cyp2E1 mRNA expression in all three groups with comparable values in vehicleHEPES/APAP-(0.80 ± 0.07), siRNALuci/APAP-(0.85 ± 0.07; p = 0.007) and siRNABid/APAP-(0.78 ± 0.07) exposed mice when compared to PBS-exposed mice (relative cyp2E1 mRNA expression 1.15 ± 0.09 at average). Similarly, the content of hepatic GSH with 6.6 ± 0.1 µmol/g liver tissue in the siRNABid/APAP group was almost unchanged when compared to values of the vehicleHEPES/APAP (6.3 ± 0.05 µmol/g) or siRNALuci/APAP (6.5 ± 0.06 µmol/g) groups.

Fig. 1
figure 1

a Schematic illustration of the experimental design and b quantitative RT-PCR analysis of Bid mRNA expression in livers of treated C57BL/6 J mice (n = 60). All animals received 42 h prior to injection of either APAP (n = 30) or PBS (n = 30, 300 mg/kg body weight per mice intraperitoneally (bw i.p.)) a liver-specific small interfering RNA delivery system (DBTC/siRNABid (n = 20) or DBTC/siRNALuci (n = 20) or DBTC/vehicleHEPES (n = 20)). For the control group we did the same pre-treatment but with equivalent volumes of phosphate-buffered saline (PBS) instead of APAP. Mice were studied 6 h thereafter (+6 h). c Representative western blot of the hepatic Bid protein expression with loading control of actinin in livers of DBTC/vehicleHEPES + APAP-(n = 4) and DBTC/siRNALuci + APAP-(n = 4) treated mice. Signals were corrected to that of ApoB. Values are given as mean + SEM; ANOVA, post hoc pairwise comparison tests, Bonferroni correction: *p < 0.05 versus DBTC/siRNALuci; §p < 0.05 versus DBTC/vehicleHEPES

Full size image
Fig. 2
figure 2

Representative western blot images (upper panels) and quantitative analysis of hepatic protein expression of a Bax and b Bcl2 in livers of treated C57BL/6 J mice (n = 60). Signals were corrected to that of ß-actin. All animals received 42 h prior to injection of either APAP (n = 30) or PBS (n = 30, 300 mg/kg body weight per mice intraperitoneally (bw i.p.)) a liver-specific small interfering RNA delivery system (DBTC/siRNABid (n = 20) or DBTC/siRNALuci (n = 20) or DBTC/vehicleHEPES (n = 20)). For the control group we did the same pre-treatment but with equivalent volumes of phosphate-buffered saline (PBS) instead of APAP. Mice were studied 6 h thereafter (+ 6 h). Values are given as mean + SEM; ANOVA, post hoc pairwise comparison tests, Bonferroni correction: *p < 0.05 versus DBTC/siRNALuci

Full size image

Bid silencing decreased APAP-induced necrotic cell death and sterile inflammation

Hepatocellular DNA fragmentation was almost absent in livers of PBS-treated mice, while APAP caused a dramatic rise in cell death, evident by the high number (200–300 hepatocytes/mm2) with DNA fragmentation in siRNALuci and vehicleHEPES pre-treated control mice (Fig. 3a, right panel, p < 0.05 vs. PBS). Administration of siRNABid significantly reduced the APAP-associated cell death to only approx. 70 cells/mm2 (Fig. 3a, p < 0.05 vs. PBS, siRNALuci and vehicleHEPES). The reduction of cell death by Bid silencing let to a general reduction of typical signs of liver damage. This is illustrated by a significant reduction of necrotic tissue in liver sections (Fig. 3b, p < 0.05 vs. PBS, siRNALuci and vehicleHEPES) and a decrease in the plasma levels of the transaminases ALT and AST when compared to mice treated with control siRNAs (Fig. 4a, b). The plasma concentration of HMGB1, a strong indicator of necrotic cell death and sterile inflammation, was also significantly reduced in siRNABid-treated APAP-exposed mice (Fig. 4c, p < 0.05 vs. siRNALuci). Furthermore, the concentration of cyclophilin A tended to be lowest in siRNABid-treated mice (Fig. 4d).

Fig. 3
figure 3

Quantitative analysis of a DNA fragmented hepatocytes as well as representative images of intravital fluorescence microscopy (right panel; original magnification × 200) and of b necrotic tissue area (in %) as well as representative H&E stained liver specimen (right panel; original magnification × 100). C57BL/6 J mice (n = 60) received 42 h prior to injection of either APAP (n = 30) or PBS (n = 30, 300 mg/kg body weight per mice intraperitoneally (bw i.p.)) a liver-specific small interfering RNA delivery system (DBTC/siRNABid (n = 20) or DBTC/siRNALuci (n = 20) or DBTC/vehicleHEPES (n = 20)). For the control group we did the same pre-treatment but with equivalent volumes of phosphate-buffered saline (PBS) instead of APAP. Mice were studied 6 h thereafter (+ 6 h). Values are given as mean + SEM; ANOVA, post hoc pairwise comparison tests, Bonferroni correction: #p < 0.05 vs. PBS; *p < 0.05 versus DBTC/siRNALuci; §p < 0.05 versus DBTC/vehicleHEPES

Full size image
Fig. 4
figure 4

Plasma activities of a alanine aminotransferase (ALT) and b aspartate aminotransferase (AST) as well as plasma concentrations of c HMGB1 and of d cyclophilin A. C57BL/6 J mice (n = 60) received 42 h prior to injection of either APAP (n = 30) or PBS (n = 30, 300 mg/kg body weight per mice intraperitoneally (bw i.p.)) a liver-specific small interfering RNA delivery system (DBTC/siRNABid (n = 20) or DBTC/siRNALuci (n = 20) or DBTC/vehicleHEPES (n = 20)). For the control group we did the same pre-treatment but with equivalent volumes of phosphate-buffered saline (PBS) instead of APAP. Mice were studied 6 h thereafter (+ 6 h). Values are given as mean + SEM; ANOVA, post hoc pairwise comparison tests, Bonferroni correction: #p < 0.05 versus PBS; *p < 0.05 versus DBTC/siRNALuci

Full size image

Bid silencing did not affect APAP-induced non-sterile inflammation and sinusoidal perfusion failure

In vivo microscopy of livers of APAP-exposed mice revealed characteristic features of acute non-sterile inflammation (Fig. 5a–c) with markedly increased numbers of both rolling and adherent leukocytes in postsinusoidal venules (Fig. 5a, b, p < 0.05 vs. PBS) as well as of tissue-infiltrating leukocytes (Fig. 5c, p < 0.05 vs. PBS) when compared to PBS-treated mice. APAP-induced non-sterile inflammation as assessed by intrahepatic leukocyte flow behaviour did not significantly differ between the different siRNA treated groups (Fig. 5a–c).

Fig. 5
figure 5

Quantitative analysis of a rolling vs. floating leukocytes, b adherent leukocytes in postsinusoidal venules as well as c of tissue infiltrating leukocytes with representative images of intravital fluorescence microscopy (+ 6 h) (right middle panel, original magnification × 200). Quantitative analysis of d sinusoidal perfusion with representative images of intravital fluorescence microscopy (right lower panel, original magnification × 200). C57BL/6 J mice (n = 60) received 42 h prior to injection of either APAP (n = 30) or PBS (n = 30, 300 mg/kg body weight per mice intraperitoneally (bw i.p.)) a liver-specific small interfering RNA delivery system (DBTC/siRNABid (n = 20) or DBTC/siRNALuci (n = 20) or DBTC/vehicleHEPES (n = 20)). For the control group we did the same pre-treatment but with equivalent volumes of phosphate-buffered saline (PBS) instead of APAP. Mice were studied 6 h thereafter (+ 6 h). Values are given as mean + SEM; ANOVA, post hoc pairwise comparison tests, Bonferroni correction: #p < 0.05 versus PBS

Full size image

Next to non-sterile inflammation, livers of APAP-exposed mice revealed a marked sinusoidal perfusion failure. Of interest, perfusion deficit was highest in vehicleHEPES and siRNALuci pre-treated mice with approx. 32% non-perfused (Fig. 5d, right panel, p < 0.05 vs. PBS) sinusoids, while siRNABid pre-treated mice showed ~ 20% less perfusion failure (Fig. 5d, right panel).

Discussion

The worldwide leading cause of acute liver failure (ALF) is intoxication by APAP. However, the pathogenesis of hepatic injury under APAP treatment is still not completely clarified [15, 22]. On the one hand side, recent studies [34, 54, 55] discovered that non-sterile inflammation represented by infiltration of inflammatory cells occurs in APAP-induced liver damage, but is not the main contributor of ALF [22, 34, 54, 55]. On the other hand, many studies described that APAP-induced liver pathology is characterized by sterile inflammation wherein the innate immune response is activated by pathogenic-derived molecules such as HMGB1 [56, 57]. The present study confirms now that both non-sterile and sterile inflammation upon APAP exposure occurs as indicated by intrahepatic leukocyte accumulation as well as raised HMGB1 plasma concentrations.

Beside inflammation, the microcirculation is deteriorated in APAP-induced liver failure [58]. Likewise, hepatic sinusoids of APAP-exposed livers showed failure of perfusion. This may directly trigger cytotoxic effector mechanisms [42, 58] contributing to apoptotic and necrotic cell death. Until today, there is ongoing discussion on the predominant form of cell death in APAP-induced ALF with reports on apoptosis, necrosis, necroptosis and regulated necrosis [22, 59,60,61]. Latest research discovered DNA-fragmented cells during cell death in regulated necrosis [25, 31, 62], a feature that was originally assigned to apoptosis [27, 28, 33, 63]. In the present study we used bisbenzimide for in vivo staining of hepatic tissue, which showed DNA fragmentation and condensation, most probably depicting cells undergoing apoptosis [27, 28]. At the same time, significant increases of plasma AST and ALT activities and the presence of large and confluent areas of necrotic tissue in APAP-treated mice indicate that also necrosis occurs as an additional pathway upon APAP intoxication. In this context, it has to be mentioned that male mice -as used in the present study- are particular susceptible to APAP-induced ALF and showing significantly higher increases of plasma AST and ALT activities than female mice [64]. Besides transaminase, plasma concentrations of HMGB1 and cyclophilin A were significantly increased upon APAP exposure. Necrosis as the predominant part of APAP-induced ALF was further affirmed by failure of caspase-3 activation (data not shown) which was already shown by many other studies [28, 60, 62, 65]. In case of ATP deficiency cells shift from apoptosis to necrosis [33, 66]. Upon APAP-treatment mitochondrial proteins are especially affected [12, 16,17,18], leading to impaired respiratory chain, oxidative stress and in consequence to lack of ATP. Thus, cell showing DNA fragmentation will further undergo necrotic cell death.

In regulated necrosis newest studies described pathway upon APAP treatment, such as translocation of the BH3-interacting domain death agonist (Bid) and Bcl-2-associated X protein (Bax), which were first attributed to apoptosis but are now becoming even more important in necrosis [33, 67,68,69]. Thus, the marked increase of Bax protein suggests that APAP-induced ALF is mainly characterized by necrotic signals.

In previous in vivo studies using Bid knockout mice [42] and in vitro studies employing Bid silencing [42] it was proposed that inhibition of the Bid pathway has therapeutic potential to reduce ALF. Although the present study does not represent a therapeutic approach, we confirmed and extent the current literature by showing that transient liver-specific targeting of Bid expression using liposomal siRNA delivery has hepatoprotective effects. This formulation was already used in a previous study silencing Fas expression 48 h prior to induction of ALF [49] and showing a protection against apoptotic and necrotic cell death as well as microcirculatory dysfunction. Several lipid nanoparticles for oligonucleotide delivery are currently in preclinical and clinical development [70, 71] and the first siRNA lipid nanoparticle has been approved by the US Food and Drug administration in August 2018 [72]. In addition, N-acetylgalactosamine (GalNAc) siRNA conjugates for liver targeting hold great promise for safe and efficacious therapeutic compounds with the ease of subcutaneous administration and with many GalNAc siRNA conjugates in clinical development [72,73,74]. In the present study, an about 70% reduction of Bid protein expression was sufficient to ameliorate liver injury, indicating that Bid could be a sensible target for siRNA-based therapeutics or other targeting approaches. The failure of siRNABid to influence hepatic cyp2E1 expression and GSH content implies that this strategy lacks off-target effects and seems to be specific.

We could demonstrate that the number of bisbenzimide-stained DNA-fragmented hepatocytes were markedly reduced upon Bid silencing, supporting Bid as a major player in APAP-induced necrotic liver tissue damage. Since Bax-silenced mice were protected from APAP-induced ALF [48], reduced Bax protein expression upon Bid silencing might also contribute to the attenuated liver damage. The fact that mitochondrial translocations of tBid and Bax are closely related and have similar effects on the molecular pathomechanism of necrosis [33, 67, 68], it is reasonable to assume that necrosis rather than apoptosis triggered APAP-induced liver tissue damage. This view is further supported by the reduction of necrotic tissue area, HMGB1 concentrations, as well as plasma activities of AST and ALT. Thus, necrosis might be the predominant cell death pathway as reported by many other studies observing a form of regulated necrosis upon APAP intoxication [22, 59,60,61, 75].

In summary, Bid silencing mediated tissue protection against APAP most probably via reduced (i) expression of Bid and Bax, (ii) execution of necrotic cell death and (iii) release of HMGB1 as a marker of sterile inflammation.