Since reactive oxygen species (ROS) derived from Kupffer cells (KC), especially CD68+ KC, play a key role in the induction of hepatic oxidative stress and injuries, we developed a polythiolated- and mannosylated human serum albumin (SH-Man-HSA), which functions as a novel nanoantioxidant for delivering thiol to CD68+ KC. In vitro electron paramagnetic resonance coupled with pharmacokinetics and immunohistochemical studies showed that SH-Man-HSA possessed powerful radical-scavenging activity and rapidly and selectively delivered thiols to the liver via mannose receptor (CD206) on CD68+ cells. SH-Man-HSA significantly improved the survival rate of concanavalin-A (Con-A)–treated mice. Moreover, SH-Man-HSA exhibited excellent hepatoprotective functions, not by decreasing tumor necrosis factor or interferon-γ production that is closely associated with Con-A–induced hepatitis, but by suppressing ROS production. Interestingly, the protective effect of SH-Man-HSA was superior to N-acetyl cysteine (NAC). This could be attributed to the difference in the inhibition of hepatic oxidative stress between the two antioxidants depending on their potential for thiol delivery to the liver. Similar results were also observed for acetaminophen (APAP)-induced hepatopathy models. Flow cytometric data further confirmed that an increase in F4/80+/ROS+ cells was dramatically decreased by SH-Man-HSA. The administration of SH-Man-HSA at 4 hours following a Con-A or APAP injection also exhibited a profound hepatoprotective action against these hepatitis models, whereas this was not observed for NAC. It can be concluded therefore that SH-Man-HSA has great potential for use in a rescue therapy for hepatopathy as a nanoantioxidant because of its ability to efficiently and rapidly deliver thiols to CD68+/CD206+ KC.
Oxidative stress plays a key role in the onset and progression of various liver diseases, including viral hepatitis (Korenaga et al., 2005), autoimmune hepatitis (Moreno-Otero, 2013), alcoholic hepatopathy (Tsukamoto and Lu, 2001), and drug-induced hepatic impairment (Laskin and Laskin, 2001). Reactive oxygen species (ROS) produced upon activation of Kupffer cells (KC), and subsequent production of inflammatory cytokines, such as tumor necrosis factor (TNF)–cytokines triggered by the presence of ROS, have been shown to be involved in the progression of a number of liver pathologic conditions (Roberts et al., 2007; Jaeschke, 2011). Kinoshita et al. (2010) reported on the subclassification of KC on the basis of their surface markers and functions, namely, CD68+, residential, and CD11b+ KC, monocyte-derived. CD68+ KC have a potent capacity to produce ROS, whereas CD11b+ cells did not. Conversely, CD11b+ KC have a strong capacity for the production of TNF, which was much less prominent in CD68+ cells. Moreover, ROS produced by CD68+ KC have been identified as an aggravating factor for the onset and progression of hepatitis in a concanavalin-A (Con-A)–induced hepatitis mouse model (Nakashima et al., 2008). In addition, mice that were pretreated with gadolinium chloride (GdCl3), a KC inhibitor, especially CD68+ cells, showed a decreased acetaminophen (APAP)-induced hepatotoxicity, possibly by suppressing the induction of oxidative stress (Laskin et al., 1995; Michael et al., 1999). This was particularly evident in the case of CD68+ KC. It therefore appears that KC, especially CD68+, play a major role in the progression of liver pathologic conditions. In fact, it had been shown that the intensity of KC activation, on the basis of CD68 expression, is directly correlated with the degree of disease severity in patients with alcoholic liver disease (Chedid et al., 2004). Interestingly, mannose receptors (CD206) are present on the KC surface, including CD68+ (Hu et al., 2012). An effective ROS-scavenging agent that can be delivered to CD68+ KC in the liver through CD206 would therefore be predicted to result in promising therapeutic outcomes. However, as far as we know, no carrier that targets CD68+/CD206+ KC is currently available.
Human serum albumin (HSA) is a simple protein with no oligosaccharide chain structures. However, the insertion of a consensus sequence for an oligosaccharide chain into the albumin gene results in the production of a protein that contains an oligosaccharide chain, as in some reported genetic variants (Minchiotti et al., 2001). Using three HSA variants—Albumin Dalakarlia, Casebrook, and Redhill (Kragh-Hansen et al., 2001)—as a template for designing a recombinant glycosylated-HSA, we recently succeeded, employing a yeast expression system, in producing a recombinant mannosylated-HSA (Man-HSA) containing a biosynthetic oligosaccharide chain with a high content of 12 mannose residues per HSA molecule (Supplemental Fig. 1) (Hirata et al., 2010). Therefore, Man-HSA would be anticipated to be efficiently distributed to CD68+ KC via CD206.
Antioxidants containing a thiol (SH) group, such as glutathione (GSH) and N-acetyl cysteine (NAC), possess excellent ROS-scavenging activities (Saito et al., 2010). In fact, NAC is currently the only remedy available for the treatment of APAP-induced hepatotoxicity (Lee et al., 2008). Hence, polythiolated-Man-HSA (SH-Man-HSA), produced by introducing SH groups into Man-HSA, has the potential to function as a novel nanoantioxidant that targets ROS derived from KC, especially CD68+/CD206+ and, thereby, as a candidate rescue therapy for both acute and chronic hepatopathy.
Materials and Methods
Male C57BL/6 mice (8 weeks old) were purchased from Japan SLC (Shizuoka, Japan). All animal experiments were conducted in accordance with the guidelines of Kumamoto University for the care and use of laboratory animals.
DTPA (diethylenetriamine-pentaacetic acid) and DTNB [5,5′-dithiobis (2-nitrobenzoic acid)] were obtained from Dojindo Laboratories (Kumamoto, Japan). DMPO (5,5-dimethyl-1-pyrroline N-oxide) and POBN [α-(4-pyridyl-1-oxide)-N-tert-butylnitrone] were purchased from Alexis Biochemicals (Lausen, Switzerland). All other chemicals were of the highest analytical grade available.
Production and Purification of Recombinant HSAs.
The protocol used to express the recombinant HSAs (rHSAs), Man-HSA and HSA, by Pichia pastoris was a modification of a previously published procedure (Hirata et al., 2010). The eluted rHSAs were deionized and defatted by charcoal treatment, freeze-dried, and then stored at −80°C until used. Sample purity was estimated by a density analysis of the Coomassie brilliant blue–stained protein bands on 10% SDS-PAGE gels. The rHSAs were estimated to be more than 97% pure. Matrix-assisted laser desorption/ionization–time of flight mass spectroscopy analysis revealed that approximately 12 mannose (N-acetyl glucosamine = 2) units are present in the glycan chain structure of the Man-HSA (Supplemental Fig. 1).
Synthesis of SH-rHSAs and Determination of Thiolation Efficiency.
Terminal SH groups were added to the rHSAs molecule by incubating 0.15 mM rHSAs with 8 mM 2-iminothiolane in 100 mM potassium phosphate buffer (KPB) containing 0.5 mM DTPA, pH 7.8, for 1 hour at room temperature (Katayama et al., 2008). The amount of SH groups in the SH-rHSAs was quantified using the DTNB method. First, 20-μl aliquots of SH-rHSAs solution and reduced glutathione (standard) were incubated in a 96-well plate with 0.2 ml of 100 mM KPB, pH 7.0, containing 1 mM DTPA and 0.5 mM DTNB, for 15 minutes at room temperature. The absorbance was then measured at 405 nm. The number of SH groups attached to Man-HSA or HSA was estimated to be 7.52 ± 0.01 mol SH/mol Man-HSA or 7.82 ± 0.02 mol SH/mol HSA, respectively (Supplemental Fig. 2B).
Scavenging Activity of SH-Man-HSA against ⋅OH Generated by H2O2/UV System.
The reaction solution contained 100 μM DTPA, 9 mM DMPO, and 500 μM H2O2 in the absence or presence of 75 μM HSA, Man-HSA or SH-Man-HSA in phosphate-buffered saline (PBS) (pH 7.4). After the sample was irradiated with UV (254 nm) for 1 minute in the electron paramagnetic resonance (EPR) flat cell, EPR spectra were measured immediately.
Pharmacokinetic Analysis of SH-Man-HSA.
SH-Man-HSA was radiolabeled with 111In using the bifunctional chelating reagent DTPA anhydride according to the method of Hnatowich et al. (1982). Mice received tail vein injections of 111In-labeled SH-Man-HSA in saline, at a dose of 10.0 nmol SH/kg. In the early period after injection, the efflux of 111In radioactivity from organs is assumed to be negligible, because the degradation products of 111In-labeled ligands using DTPA anhydride cannot easily pass through biologic membranes. This assumption was confirmed by the finding that 111In was not detectable in the urine throughout the 120-minute period. At appropriate intervals after the injection, blood was collected from the vena cava under ether anesthesia and plasma was obtained by centrifugation (3000g, 10 minutes). The liver, kidney, spleen and lung were excised, rinsed with saline, and weighed. The radioactivity of each sample was measured in a well-type NaI scintillation counter (ARC-500; Hitachi Aloka Medical, Ltd., Tokyo, Japan).
Preparation of Bromobimane-Labeled SH-Man-HSA.
A solution of 75 μM bromobimane in 0.1 M KPB was added to a solution of 100 μM SH-Man-HSA in 0.1 M KPB, followed by incubation for 2 hours at 4°C. The resulting solution was purified using a Pharmacia Bio-Gel PD-10 column (GE Healthcare, Little Chalfont Bucks, UK) and concentrated by Vivaspin (Sartorius Stedim Biotech S.A., Aubagne, France).
Evaluation of CD68+/CD206+ KC Distribution of SH-Man-HSA.
The prepared bromobimane-labeled SH-Man-HSA (20.0 μmol SH/kg) was injected into the tail vein of mice. Mannan (6.0 mg/mouse) was injected intravenously 30 minutes before the bromobimane-labeled SH-Man-HSA administration. At 1 and 12 hours after the administration of bromobimane-labeled SH-Man-HSA, the liver was removed, covered with optimal cutting temperature compound, and frozen at −80°C. Fresh-frozen sections (4-μm thickness) of the liver were cut on a cryostat (CM3000II; Leica, Wetzlar, Germany), collected on slides, and immediately dried. The sections were fixed with phosphate-buffered 4% paraformaldehyde and washed. After incubation with 1% Block Ace (DS Pharma Biomedical, Osaka, Japan) for 10 minutes, the slides were incubated overnight with anti-HSA (Bethyl Laboratories, Inc., Montgomery, TX), CD68 (FA-11; BioLegend, San Diego, CA), and CD206 (C068C2; BioLegend) antibodies diluted 100, 100, or 200 times, respectively. After the reaction, the slide was observed using a microscope (BZ-8000; Keyence, Osaka, Japan).
Effect of SH-Man-HSA on Survival Rate of Lethal Con-A–Treated Mice.
A lethal Con-A–treated mouse model was produced using C57BL/6 mice that were intravenously injected with Con-A (1.0 mg/mouse). The same SH content (20.0 μmol SH/kg) of SH-Man-HSA and NAC was administered intravenously just prior to the Con-A injection.
Hepatoprotective Effect of SH-Man-HSA on Con-A– or APAP-Treated Mice.
Con-A– and APAP-induced liver injury mouse model was produced as previously reported (Ayoub et al., 2004; Nakashima et al., 2008; Saito et al., 2010). The animal experiment protocol is shown in Supplemental Figs. 4 and 5. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activity levels were determined using a transaminase C2 test kit from Wako Chemicals (Saitama, Japan). In case of Con-A, the mice received 0.25 mg of Con-A intravenously (Sigma-Aldrich, St. Louis, MO). Just prior to Con-A injection, 5.0, 10.0, and 20.0 μmol SH/kg of SH-Man-HSA and 20.0 μmol SH/kg of NAC were administered intravenously to the mice. Likewise, SH-HSA (20.0 μmol SH/kg) or Man-HSA (2.66 μmol SH/kg) was administered intravenously. In case of APAP, the mice received 300 mg/kg of APAP intraperitoneally (Sigma-Aldrich). The same SH content (20.0 μmol SH/kg) of SH-Man-HSA and NAC was administered intravenously to the mice just prior to APAP injection. Furthermore, we administered SH-Man-HSA or NAC (intravenously) at 2 and 4 hours after the injection of Con-A or APAP. To inhibit the KC, especially CD68+ cells, 200 μg of GdCl3 was administered (Sigma-Aldrich) intravenously via the caudal vein 24 hours before Con-A or APAP injection. All mice were sacrificed 12 hours after Con-A or APAP injection.
Histologic Examination of Liver Tissues.
Twelve hours after Con-A or APAP injection, the whole liver was removed. For the histologic analyses, the liver tissues were fixed in 10% neutral-buffered formalin (Wako Pure Chemical Industries, Osaka, Japan), embedded in paraffin (Sakura Finetek Japan, Tokyo, Japan), and sectioned at a 4-μm thickness. The liver sections were subjected to hematoxylin and eosin staining for morphologic analysis, terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining for cell apoptosis, and immunohistochemistry for 8-OHdG (8-hydroxy-2′-deoxyguanosine) [15A3] (Santa Cruz Biotechnology, Dallas, TX) and nitrotyrosine (NO2-Tyr) (Millipore, Billerica, MA). All assays were performed in triplicate. For TUNEL staining, sections were stained using the In Situ Cell Death Detection Kit, Fluorescein (Roche, Basel, Switzerland), according to the manufacturer’s protocol for paraffin-embedded sections. For the immunohistochemistry of 8-OHdG and NO2-Tyr, first, antigen retrieval was performed by means of an immunosaver (Nisshin EM Corporation, Tokyo, Japan). T-TB solution containing 50 mM Tris-HCl (pH 7.4) and 0.1% Tween-20 was then used to solubilize the liver slices, followed by blocking with Block Ace (Dainippon Pharmaceutics, Osaka, Japan) at room temperature for 10 minutes. Next, reaction with the primary antibody was carried out overnight at a temperature below 4°C. In addition, the primary antibody against 8-OHdG or NO2-Tyr was diluted with 0.5% bovine serum albumin in PBS 50 times before use. The liver slices were then washed with T-TB solution, followed by reaction with the secondary antibody at room temperature for 1.5 hours. For 8-OHdG immunostaining, Alexa Fluor 488 goat anti-mouse IgG (H+L) (Invitrogen, Tokyo, Japan) and for NO2-Tyr immunostaining, Alexa Fluor 546 goat anti-mouse IgG (H+L) (Invitrogen/Life Technologies) were diluted with 0.5% bovine serum albumin in PBS 200 times before use. After the reaction, the slide was observed under a microscope (BZ-8000; Keyence).
Quantification of Liver SH and GSH Levels.
Tissue homogenates were prepared by homogenizing the tissues with 5% (w/v) sulfosalicylic acid. The homogenates were then centrifuged at 10,000g for 10 minutes at 4°C. SH levels were determined according to the DTNB method. Reduced and oxidized glutathione levels in the livers were determined using a reduced/oxidized GSH ratio (GSH/GSSG) quantification kit (Dojinkagaku, Kumamoto, Japan).
Determination of ROS in Liver Using EPR.
The in vivo EPR analysis was performed as described previously with some modifications (Sato et al., 2002). The EPR spectra of POBN spin adducts of lipid peroxide radical LOO˙ were determined at 4 hours after Con-A or APAP injection. POBN (1.0 g/kg) was injected intraperitoneally at 30 minutes before the mice were sacrificed. The livers were homogenized in 1 ml of saline on an ice bath. The homogenates were added in 2 ml of 2:1 chloroform/methanol, shaken, then centrifuged at 2800 rpm for 10 minutes. The chloroform layer was isolated. After evaporating the sample by bubbling with N2 in a vial, the sample was resuspended in organic solvent (200 μl of chloroform/methanol). EPR spectra were immediately recorded at room temperature in a JES-TE200 spectrometer (JEOL, Tokyo, Japan) under the following conditions: modulation frequency, 100 kHz; microwave frequency, 9.43 GHz; microwave power, 40 mW; scanning field, 333.5 mT; sweep time, 2 minutes; field modulation width, 0.25 mT; receiver gain, 630; and time count, 0.3 seconds. Every buffer and solution of the reaction mixture used for EPR measurement were treated with Chelex 100 resin (Bio-Rad Laboratories, Hercules, CA) before use to remove metals.
Isolation of Hepatic Mononuclear Cells and Flow Cytometry.
Mononuclear cells were prepared as described previously (Dobashi et al., 1999). Liver specimens were minced with scissors. After adding a 0.25% collagenase solution, the specimens were shaken for 20 minutes in a 37°C constant-temperature bath. The liver specimen was then filtered through a nylon mesh (108 μm). After mixing in 33% Percoll solution containing heparin (1%), the sample was centrifuged for 20 minutes at 500g at room temperature. After removing the supernatant, the pellet was resuspended in a red blood cell lysis solution and then was washed twice in RPMI 1640 with 10% fetal bovine serum. For identification of KC, mononuclear cells were stained by phycoerythrin Cy5-conjugated anti-F4/80 antibody (eBioscience, San Diego, CA). For the detection of intracellular ROS, CM-H2DCFDA (Invitrogen, Carlsbad, CA), an ROS-sensitive fluorescent dye used as probe. Also, for the detection of macrophage polarization status, anti–inducible NO synthase (Santa Cruz Biotechnology, Santa Cruz, CA) and CD206 (C068C2; BioLegend) antibody were used. F4/80+ KC were isolated by positive selection according to the manufacturer’s protocol (MACS isolation system; Miltenyi Biotec, Auburn, CA). Flow cytometry was performed using a FACSCalibur (BD Biosciences, San Jose, CA).
Quantification of TNF and Interferon-γ Levels.
TNF and interferon (IFN)-γ levels in plasma at 1 and 12 hours, respectively, after Con-A (0.25 mg/mouse) injection were measured by enzyme-linked immunosorbent assay, following the manufacturer’s recommended protocol.
All data are expressed as mean ± S.E. The means for groups were compared by analysis of variance followed by Tukey’s multiple comparison and Student’s t test. The survival rates were compared using Kaplan-Meier survival curves and the log-rank test. A probability value of P < 0.05 was considered significant.
Radical-Scavenging Activity of SH-Man-HSA.
We developed SH-Man-HSA (7.5 mol SH/mol of Man-HSA) using 2-iminothiolane (Supplemental Fig. 2A) and its radical-scavenging activity against ·OH radicals was evaluated using an EPR method. Figure 1A shows the EPR signal derived from ·OH radicals in the absence (control) and presence of HSA, Man-HSA, and SH-Man-HSA. In the presence of HSA and Man-HSA, the signal intensity decreased to 40 and 50% of the control, respectively, whereas SH-Man-HSA inhibited the intensity by nearly 80%, indicating that SH-Man-HSA has more potent ROS-scavenging activity than HSA and Man-HSA (Fig. 1B).
Pharmacokinetic Analysis of SH-Man-HSA.
A pharmacokinetic analysis of the SH-Man-HSA showed that 50% of the administered dose was distributed to the liver within about 20 minutes, similar to previously reported findings for Man-HSA (Fig. 2A) (Hirata et al., 2010). Furthermore, the hepatic distribution of SH-Man-HSA was investigated by means of a fluorescence-imaging technique. In this procedure, the SHs of SH-Man-HSA were completely modified by the treatment with bromobimane. Liver sections were observed at 1 hour after the administration of bromobimane-labeled SH-Man-HSA (20.0 μmol SH/kg), and modified SH groups were detected with a fluorescence probe in the form of a blue emission. As shown in Fig. 2B, substantial fluorescence was observed in the liver sections, especially near the sinusoidal regions. Such a fluorescence was largely inhibited by the preadministration of a two times higher dose of mannan, a typical substrate for CD206, suggesting the importance of this receptor for distributing SH-Man-HSA to the liver.
At the region near the sinusoid, there are several cells that express CD206, such as KC, endothelial cells, and dendritic cells (Kogelberg et al., 2007). Our previous study revealed that Man-HSA was preferentially distributed to KC, but not to endothelial cells (Hirata et al., 2010). It has been demonstrated that the C-type lectin–like domain binds complex carbohydrates in which the chains are terminated with mannose, such as Man(n)GlcNAc2 (n, numbers of mannose residues). For instance, the carbohydrate-recognition domain of Dectin-2 is a C-type lectin with a very high affinity for high-mannose structures, especially structures that contain more than seven mannose residues, and its specificity is enhanced with an increase in the number of mannose residues (McGreal et al., 2006). Interestingly, Dectin-2 expression was predominantly restricted to CD206-positive macrophages (Sun et al., 2013). Because SH-Man-HSA possesses an oligosaccharide chain with Man(12)GlcNAc2, it would be expected to be preferentially recognized by Dectin-2 on KC that express CD206, but not endothelial cells.
Evaluation of CD68+/CD206+ KC Distribution of SH-Man-HSA.
To further characterize the distribution of SH-Man-HSA to KC subtype, liver sections were subjected to immunostaining with anti-HSA, CD68, and CD206 antibodies. The anti-HSA antibody used in this study did not crossreact with the mouse species. At 1 hour after the administration of bromobimane-labeled SH-Man-HSA, the fluorescence derived from bromobimane (blue) was merged with both the fluorescence of the anti-HSA antibody (red) and the anti-CD68 antibody (green) as shown by white spots in the inset image (Fig. 3A). Similar white spots that had merged with the fluorescence of bromobimane, the anti-HSA antibody, and the anti-CD206 antibody (green) were also observed (Fig. 3B). These data indicate that SH-Man-HSA was efficiently distributed to CD68+/CD206+ KC within 1 hour after the administration of bromobimane-labeled SH-Man-HSA depending on its unique oligosaccharide moiety. However, the fluorescence of bromobimane did not become merged with both the fluorescences of the anti-HSA antibody and the anti-CD68 antibody at 12 hours after the administration of bromobimane-labeled SH-Man-HSA (Fig. 3C, right panel), suggesting that SH-Man-HSA is susceptible to lysosomal degradation in CD68+/CD206+ KC, and that the majority of the SH moieties were liberated from SH-Man-HSA and moved to the outside of CD68+/CD206+ KC.
Hepatoprotective Effect of SH-Man-HSA on Con-A–Treated Mice.
To examine the effect of SH-Man-HSA on liver injury, we used a Con-A–induced hepatitis model, because it has been widely used as a model for liver injury, such as autoimmune and viral types. As shown in Fig. 4A, all of the saline-treated mice died within 28 hours after a lethal dose (1.0 mg/mouse i.v.) of Con-A, whereas 20% of the NAC (20.0 μmol SH/kg)-treated mouse group survived. Interestingly, the administration of SH-Man-HSA (20.0 μmol SH/kg) resulted in a significant increase in survival rate, with 70% of the mice being alive at 60 hours after the Con-A injection.
The hepatoprotective effects of SH-Man-HSA were evaluated to monitor ALT and AST levels at 12 hours after the injection of a low dose (0.25 mg/mouse i.v.) of Con-A because the plasma ALT and AST levels were at a maximum at 12 hours post Con-A injection (Supplemental Fig. 6). SH-Man-HSA suppressed the elevation of plasma ALT and AST levels that had been induced by Con-A in a dose-dependent manner. In particular, 20.0 μmol SH/kg of SH-Man-HSA inhibited the elevation of ALT and AST plasma levels by 98 and 99% for Con-A–treated mice (Fig. 4B). Therefore, 20.0 μmol SH/kg of SH-Man-HSA was adopted in the subsequent experiments (Supplemental Fig. 3).
The elevation in levels of plasma ALT and AST in Con-A–treated mice was also suppressed by 20.0 μmol SH/kg of NAC, but the effect was much less than by SH-Man-HSA. On the other hand, pretreatment with GdCl3 at 24 hours before the Con-A injection resulted in a hepatoprotective effect that was equivalent to 20.0 μmol SH/kg of SH-Man-HSA (Fig. 4B).
Histologic examinations of liver tissue in Con-A–induced hepatitis were also performed with H&E and TUNEL staining (Fig. 4C). H&E staining of liver sections showed evidence of massive necrosis within the liver lobules in the saline treatment group of the Con-A–challenged mice. Similar results were obtained for TUNEL staining. In the saline treatment groups, a number of TUNEL-positive hepatocytes were observed for the Con-A–induced model, whereas no evidence for such positive cells was found in the case of the SH-Man-HSA and GdCl3 treatment groups. NAC also reduced TUNEL-positive cells, but to a lesser extent compared with SH-Man-HSA.
The contribution of each functional component was examined in the Con-A–induced model using two HSA derivatives, namely, Man-HSA, which contains mannose residues but no additional SH groups, and SH-HSA, which contains additional SH moieties, but no mannose residues. As shown in Fig. 4D, Man-HSA (2.66 μmol SH/kg) and SH-HSA (20.0 μmol SH/kg) significantly attenuated the Con-A–induced elevation in plasma ALT and AST levels. The extent of attenuation was in the following order: Man-HSA < SH-HSA < SH-Man-HSA.
Effect of SH-Man-HSA on Hepatic SH Level and Oxidative Stress Induced by Con-A.
SH levels in liver homogenates were determined 12 hours after the injection of Con-A (Fig. 5A) to evaluate the efficacy of SH delivery to the liver by SH-Man-HSA and NAC, both of which had the same SH content (20.0 μmol SH/kg). The SH level of the saline-treated groups was decreased by Con-A injection, and it recovered in the following treatment order: NAC < GdCl3 < SH-Man-HSA. Furthermore, as shown in Fig. 5B, SH-Man-HSA suppressed the decrease in GSH/GSSG, indicating that SH-Man-HSA is superior to NAC in terms of delivering SH to the liver.
The effects of SH-Man-HSA or NAC on Con-A–induced free radical formation in the liver were directly monitored in vivo using an EPR technique (Fig. 5C), in which LOO˙ levels were determined in a liver homogenate. Radical production increased in a time-dependent manner, reaching a maximum at 4 or 5 hours after the Con-A injection (Supplemental Fig. 7). As shown in Fig. 5D, more than 80% of the EPR signal induced by the Con-A challenges was inhibited by both SH-Man-HSA and GdCl3, whereas less than 40% of the signal intensity was decreased by NAC at 4 hours after the Con-A injection. Similar antioxidative effects of SH-Man-HSA were also confirmed by immunostaining of 8-OHdG and NO2-Tyr, markers for oxidative and nitrative stress, respectively, in the liver (Fig. 5E). In addition, Con-A markedly increased the fluorescence intensity of ROS derived from F4/80+/ROS+–detecting probe (168 ± 27), whereas it was significantly inhibited by both SH-Man-HSA (57 ± 6) and GdCl3 (70 ± 5) (Fig. 5F). Such similar inhibitions of ROS production in F4/80+/ROS+ cells between SH-Man-HSA and GdCl3 could reflect comparable action against Con-A–induced hepatic damage.
On the other hand, the possibility that the ROS-scavenging activity of SH-Man-HSA might be mediated via the inhibition of TNF or IFN-γ that is closely associated with Con-A–induced hepatic damage cannot be completely excluded. However, SH-Man-HSA did not decrease to the levels of TNF or IFN-γ (Fig. 5G).
Hepatoprotective Effect of SH-Man-HSA on APAP-Treated Mice.
The hepatoprotective effect of SH-Man-HSA was also examined against an APAP-induced hepatic injury model that was prepared by an intraperitoneal injection of 300 mg/kg of APAP (Fig. 6). As in the case of Con-A–induced hepatitis, equal SH contents (20.0 μmol SH/kg) of SH-Man-HSA or NAC were administered (intravenously) just before APAP injection, and the plasma ALT and AST levels were measured at 12 hours after APAP injection (Supplemental Fig. 8). As a result, SH-Man-HSA largely inhibited APAP-induced elevation of plasma ALT and AST levels by 96 and 94%, respectively. NAC also significantly suppressed the plasma ALT and AST levels by 61 and 57%, respectively, but this effect was much less than that of SH-Man-HSA. Pretreatment with GdCl3 showed hepatoprotective effects similar to SH-Man-HSA administration (Fig. 6A). This is consistent with the previous finding that KC play an important role in the development of APAP-induced hepatic injury.
H&E staining of liver sections indicated that the APAP-induced massive necrosis within the centralobular region of the livers, as evidenced by the loss of basophilic staining, vacuolization, cell swelling, and karyolysis around the centrilobular veins, as previously reported (Fig. 6B, upper panel). However, liver sections from SH-Man-HSA administration showed little damage, and a nearly normal morphology was preserved. Similar results were also obtained for a GdCl3 pretreatment. H&E staining of tissue from the NAC administration group showed some improvement in the integrity of liver sections against APAP-induced hepatotoxicity, but its efficacy was less than that of SH-Man-HSA and GdCl3. Similar results were obtained for TUNEL staining (Fig. 6B, lower panel). In the saline treatment group, a number of TUNEL-positive hepatocytes were observed, whereas such positive cells could not be found after SH-Man-HSA and GdCl3 administration groups. NAC also reduced TUNEL-positive cells, but to a lesser extent compared with SH-Man-HSA.
Effect of SH-Man-HSA on Hepatic SH Level and Oxidative Stress Induced by APAP.
Figure 7A shows SH levels in liver homogenates 12 hours after the injection of mice with APAP. The SH levels in liver were significantly decreased as a result of the APAP injection, but the level was completely recovered by the subsequent SH-Man-HSA administration or a GaCl3 pretreatment. In contrast, the administration of NAC was ineffective in recovering APAP-induced SH depletion in the liver. Furthermore, as shown in Fig. 7B, SH-Man-HSA suppressed the decrease in GSH/GSSG, indicating that SH-Man-HSA is superior to NAC in delivering SH to the liver.
Figure 7, C and D, shows the effect of SH-Man-HSA on the lipid peroxide radical in liver homogenate after APAP injection to mice. As observed in the Con-A–induced hepatopathy model, a remarkable EPR signal was detected at 4 hours after APAP injection (Supplemental Fig. 9). This EPR signal was significantly suppressed by the SH-Man-HSA or GdCl3 pretreatment by approximately 70%, but NAC treatment only decreased the signal intensity by approximately 40%.
Previous findings revealed that oxidative and nitrative toxicity in liver induced by APAP included lipid peroxidation and protein nitration (Michael et al., 1999). As shown in Fig. 7E, our immunostaining data against 8-OHdG and NO2-Tyr demonstrated that SH-Man-HSA administration effectively inhibited the accumulation of those oxidized products in the liver induced by APAP, whereas NAC was less effective in inhibiting the accumulation of those oxidative stress markers. Furthermore, to clarify the antioxidative activity of SH-Man-HSA in KC of APAP-induced hepatic injury mice, the F4/80+/ROS+ KC population was estimated at 4 hours after APAP injection (300 mg/kg i.p.) by flow cytometry (Fig. 7F). F4/80+/ROS+ KC increased following the APAP injection, whereas it was decreased by SH-Man-HSA administration. Pretreatment with GdCl3 also decreased the content of F4/80+/ROS+ KC, to the lesser extent.
Effect of Postadministration of SH-Man-HSA on Con-A– or APAP-Treated Mice.
To investigate the potential of SH-Man-HSA as a rescue therapy after Con-A or APAP challenges, equal SH contents (20.0 μmol SH/kg) of SH-Man-HSA or NAC were administered 2 and 4 hours following an injection of Con-A (intravenously) or APAP (intraperitoneally).
As shown in Fig. 8, the administration of SH-Man-HSA 2 hours after Con-A or APAP injection largely inhibited the elevation in ALT and AST levels. In addition, even 4 hours after the Con-A or APAP injection, SH-Man-HSA significantly reduced the elevation of ALT and AST. In contrast, the administration of NAC 2 or 4 hours later did not exhibit sufficient hepatoprotective action in either of the models.
A previous study reported that the administration of a high dose (650 μmol SH/kg) of NAC (intravenously) at 1.5 hours after an APAP injection (300 mg/kg i.p.) exhibited a significant hepatoprotective action (Saito et al., 2010). Thus, the effect of 650 μmol SH/kg of NAC on APAP-induced hepatic injury at 2 or 4 hours after an APAP injection (300 mg/kg i.p.) was also examined. At 2 hours following the APAP injection, a high dose of NAC showed a greater suppressive effect on ALT (84%) and AST (80%) elevation than a low dose (20.0 μmol SH/kg), as shown in Fig. 8B. However, a high dose of NAC showed only a small inhibitory effect against ALT and AST elevations at 4 hours after the APAP injection. These data indicate that SH-Man-HSA was a more effective agent for treating these experimental acute liver injuries compared with NAC.
To confirm whether the hepatoprotective action derived from the postadministration of SH-Man-HSA resulted from the inhibition of hepatic oxidative stress as well as the pretreatment, we carried out histopathological and oxidative stress analyses after a post-treatment with SH-Man-HSA to Con-A– and APAP-induced hepatopathy models. The liver sections of H&E and the 8-OHdG staining of the SH-Man-HSA treatment at 2 hours after the Con-A and APAP injections clearly showed that SH-Man-HSA effectively inhibited liver damage and hepatic oxidative stress (Supplemental Fig. 10), confirming that, even though a post-treatment, SH-Man-HSA showed hepatoprotective action via the inhibition of hepatic oxidative stress similar to that observed for the pretreatment.
The main goal of the present study was to design a system for delivering a nanoantioxidant containing SH groups to KC, especially CD68+/CD206+, using mannose as a recognition element for use as a rescue therapeutic agent in treating various acute liver injuries. HSA is a simple protein and contains no oligosaccharide chain structures. However, the insertion of a consensus sequence for an oligosaccharide chain into the albumin gene (D63N, A320T, D494N) results in a protein that contains an oligosaccharide chain, as in some reported genetic variants (Minchiotti et al., 2001). The pharmacokinetic properties and biologic activities of these glycosylated mutants of HSA are similar to the wild-type molecule. Yeast expression systems were used to produce the Man-HSA because generally glycoproteins that are expressed in yeast possess high mannose-type glycan chains, which are different from those expressed by mammalian cells. We previously showed that Man-HSA is suitable for use as a drug delivery system carrier in targeting KC owing to its pharmacokinetic properties, which include its rapid and efficient distribution to the liver and the fact that it is selectively distributed to KC by CD206, unlike other chemically modified mannosylated carriers (Hirata et al., 2010). The subsequent addition of an average of 7.5 mol SH groups to each Man-HSA molecule did not alter its unique hepatic distribution properties, and furthermore, the immunohistochemical evaluations reported here clearly showed that, among the KC subtypes, SH-Man-HSA is efficiently distributed to CD68+/CD206+ KC (Fig. 3). As far as we know, this is the first study to report on the development of a carrier for targeting CD68+/CD206+ KC. Thus, SH-Man-HSA meets the requirement for use as a CD68+/CD206+ KC directed nanoantioxidant because it confers more potent ROS-scavenging activity than HSA (Fig. 1).
Con-A– (Nakashima et al., 2008) and APAP-induced hepatitis (Ayoub et al., 2004; Saito et al., 2010) mouse models were used to evaluate the efficacy of SH-Man-HSA as a CD68+ KC-directed antioxidant in the present study. The most frequent single cause of acute liver failure is an APAP overdose, accounting for 46% of the reported cases in the United States (Lee et al., 2008). In addition, the characteristics of an APAP-induced hepatitis mouse model of liver injury are similar to those for human patients on the basis of clinical observations and biochemical and histopathological data, and has been used in the past to identify potential therapeutic interventions. In fact, the present study demonstrated that the efficient delivery of SH to CD68+/CD206+ KC exhibited excellent hepatoprotective action against two models of acute hepatopathy. As mentioned in the introduction section, it is generally thought that ROS generated by CD68+ KC play an important role in the initiation or progression of liver pathologic conditions. Our results provide more insights into involvement of CD68+/CD206+ KC in the development of acute liver injury. In addition, the present study also demonstrated that at an equal dose of SH (20.0 μmol SH/kg), SH-Man-HSA was superior to NAC in terms of improving the survival rate of Con-A–treated mice (Fig. 4A), and of alleviating liver damage in Con-A– and APAP-induced hepatitis models as evidenced by the suppression of ALT and AST (Figs. 4B and 6A) and H&E staining (Figs. 4C and 6B). Because the hepatoprotective effect of Man-HSA and SH-HSA were not comparable to that of SH-Man-HSA (Fig. 4D), this suggests that both mannose residues and SH are essential for the therapeutic effect of SH-Man-HSA against a hepatic injury. The hepatoprotective effect observed for SH-Man-HSA was similar to that for a GdCl3 pretreatment (Figs. 4B and 6A), indicating that SH-Man-HSA, when delivered to CD68+ KC, could efficiently scavenge ROS generated in response to Con-A or APAP. In addition, excellent ROS-scavenging activity was evident for SH-Man-HSA, as evidenced by in vivo EPR, flow cytometry, and oxidative or nitrative stress markers (8-OHdG, NO2-Tyr) (Figs. 5, C–F, and 7, C–F).
On the other hand, it is well known that Con-A activates T cells to produce IFN-γ via binding to T-cell receptors. IFN-γ, in turn, activates CD68+ KC to produce TNF that interacts with the TNF receptor on the CD68+ KC to produce ROS (Nakashima et al., 2008; Kinoshita et al., 2010). The possibility that the ROS-scavenging activity of SH-Man-HSA might be mediated via the inhibition of TNF or IFN-γ cannot be completely excluded. Interestingly, SH-Man-HSA levels did not decrease to the levels of these inflammatory factors. This result indicates that SH-Man-HSA attenuates Con-A–induced hepatopathy regardless of the levels of these inflammatory factors (Fig. 5G). Therefore, SH-Man-HSA exhibited its hepatoprotective effect, not by influencing TNF or IFN-γ production, but by suppressing ROS production produced by CD68+/CD206+ KC, which occurs downstream from the production of these cytokines. Similar results were reported for other antioxidants, such as edarabone and lecithinized superoxide dismutase, in Con-A–induced hepatitis (Nakashima et al., 2008).
There is a possibility that mannose residues or HSA molecules of SH-Man-HSA directly interact with the injected Con-A molecules, thus leading to suppression of the toxicity of Con-A. If this is correct, TNF and IFN-γ, which are involved in the development of Con-A–induced hepatopathy, should not be increased after the coadministration of Con-A and SH-Man-HSA. However, we found similar elevations of these inflammatory factors between Con-A injection and coadministration of SH-Man-HSA and Con-A, indicating that Con-A still preserved the potency that causes the induction of hepatic damage when SH-Man-HSA and Con-A were coadministrated. This clearly excludes the possibility that mannose residues or HSA of SH-Man-HSA directly inactivate the hepatic toxicity of Con-A. In fact, the postadministration of SH-Man-HSA also ameliorated the Con-A–induced liver injury via the mechanism similar to that of the pretreatment (Supplemental Fig. 10).
The role of ROS in APAP toxicity has been a subject of debate for decades and is not without its controversies (James et al., 2003). However, the pathologic role of ROS has not been fully understood because its effects can vary, depending on experimental conditions. For example, in their comprehensive review of that drug’s toxicity, Jaeschke et al. (2012) mentioned that ROS can induce an inflammatory response in immune cells, including KC, and these extracellularly generated oxidants can diffuse into hepatocytes and trigger various types of mitochondrial dysfunction and oxidant stress, which subsequently induce cell death. In addition, Botta et al. (2006) commented on the importance of glutathione synthesis by glutamate-cysteine ligase against APAP-induced liver injury using glutamate-cysteine ligase transgenic mice. On the other hand, Lewerenz et al. (2003) reported that antioxidants protect primary rat hepatocyte cultures against APAP-induced DNA strand breaks but not against APAP-induced cytotoxicity. Moreover, the knockout of Cu, Zn-superoxide dismutase, a major intracellular antioxidant enzyme, was reported to be resistant to APAP-induced hepatotoxicity (Lei et al., 2006). In this study, we demonstrated that both the pre- and postadministration of SH-Man-HSA that targets ROS derived from CD68+/CD206+ KC significantly improve the Con-A– and APAP-induced hepatopathy mouse model via the inhibition of hepatic oxidative stress. This implies that ROS derived from CD68+/CD206+ KC could be responsible for the development of liver damage in both disease models. In addition, we found that the SH-Man-HSA treatment did not significantly influence the early metabolic pathway of APAP-induced liver injury (data not shown). A similar result was also reported by Michael et al. (1999), that is, a pretreatment of GdCl3 prevented APAP-induced hepatopathy without a decrease in the APAP-protein adducts. These findings indicate that ROS derived from CD68+/CD206+ KC are probably not involved in the early event of APAP toxicity and that SH-Man-HSA functions at the stage after the formation of N-acetyl-p-benzoquinone imine (NAPQI). Moreover, hypoxia-inducible factor-1α (HIF-1α) is a critical transcription factor in response to oxidative stress (James et al., 2006). It has been revealed that ROS are important contributors to the early HIF-1α stabilization that enhances APAP toxicity (Sparkenbaugh et al., 2011). Since NAC prevented HIF-1α accumulation via the inactivation of the NAPQI formation, it would be interesting to know whether SH-Man-HSA also inhibits the stabilization of HIF-1α or whether ROS derived from CD68+/CD206+ KC are involved in the HIF-1α accumulation. As mentioned above, SH-Man-HSA does not significantly influence the early metabolic pathway of APAP toxicity, SH-Man-HSA or ROS derived from KC may not be directly involved in the HIF-1α accumulation. These findings led us to conclude that, even though the early events of the onset of acute liver injury are different depending on the disease, ROS derived from CD68+/CD206+ KC might be a common contributor for the development of acute liver injury. Future investigations will be necessary to clarify these interesting issues in terms of understanding the pathologic role of ROS derived from CD68+/CD206+ KC. In addition, CD206 has been frequently used as a marker of M2 macrophages that are a type of anti-inflammatory immune cell (Choi et al., 2010). Thus, it is possible that SH-Man-HSA inhibits hepatic injury via an increase in the population of M2 macrophages. However, SH-Man-HSA did not significantly influence the balance of M1/M2 macrophages in both disease models (Supplemental Fig. 11). This indicates that the hepatoprotective effect of SH-Man-HSA is not attributable to the changes in the polarization of macrophages in liver.
Clinically, NAC is only effective if given to patients the first few hours after ingestion of an overdose of APAP. In the case of the APAP hepatotoxicity mouse model, NAC was reported to be the least beneficial if administered 3 or 4 hours after APAP injection (Saito et al., 2010). Consistent with previous reports, we also found that, even at a high dose (650 μmol SH/kg), NAC was no longer effective at 4 hours after APAP injection (Fig. 8B). Surprisingly, 20.0 μmol SH/kg of SH-Man-HSA, a dose that was approximately 33 times less than the NAC dose (650 μmol SH/kg), exhibited a significant hepatoprotective effect, even though it was administered 4 hours later in both pathologic models. In vivo EPR data clearly showed that ROS were mainly generated in the liver at 4 to 5 hours after the injection of Con-A or APAP and was completely inhibited by SH-Man-HSA or GdCl3. Taking into account the rapid and efficient delivery of SH into CD68+/CD206+ KC by SH-Man-HSA (Fig. 3) and the elevation in the hepatic SH content (Figs. 5A and 7A), these findings suggest that the difference in the therapeutic effect between SH-Man-HSA and NAC given after an injection of Con-A or APAP could be attributable to the difference in the rate and site of distribution to the liver, especially CD68+/CD206+ KC, between these two drugs.
The issue of whether the recovery of the hepatic SH content to control levels by SH-Man-HSA can be attributed to the SH moieties in SH-Man-HSA or suppression of the consumption of GSH, an intracellular substance with SH, or both, remains unknown. In contrast to the hepatic SH content, the hepatic GSH/GSSG ratio was not recovered to the control level as the result of the SH-Man-HSA treatment (Figs. 5B and 7B). Such an inconsistency might explain why, initially, GSH is consumed as the result of a Con-A or APAP challenge, and then SH derived from SH-Man-HSA makes up for the deficit in hepatic SH levels. In fact, immunohistochemical analyses (Figs. 2 and 3) and the SH levels in liver homogenates (Figs. 5A and 7A) are supportive of the hypothesis that SH liberated from SH-Man-HSA may still be retained within the liver, but not CD68+ KC, at 12 hours after SH-Man-HSA administration (Fig. 3C). SH-Man-HSA would be hydrolyzed in lysosomes of CD68+/CD206+ KC after endocytosis by CD206, and the SH groups would then be liberated from SH-Man-HSA because 2-iminothiolane, the SH labeling agent for HSA used in this study, is also hydrolyzed at a low pH, such as pH 5 (Singh et al., 1996). These SH groups then move to the outside of the CD68+/CD206+ KC and would contribute to the elevation of the hepatic SH levels in disease models. Therefore, SH-Man-HSA is not only able to scavenge ROS in the liver more efficiently than NAC, but also ROS-scavenging activity is retained in the liver longer than NAC. It might be possible to treat a liver impairment with SH-Man-HSA at a lower dose and a lesser dosing frequency than is needed for NAC.
In conclusion, SH-Man-HSA is a unique and powerful nanoantioxidant that targets ROS derived from CD68+/CD206+ KC because of its efficient and rapid delivery of SH to CD68+/CD206+ KC. SH-Man-HSA exhibited an excellent hepatoprotective action against two experimental acute hepatitis models that was superior to NAC because of its sufficient suppression of oxidative stress. Therefore, it has great potential for use as a rescue therapy for acute hepatopathy. Moreover, the therapeutic impact of SH-Man-HSA in chronic hepatic diseases in which CD68+ KC contribute to the onset and progression of diseases, such as in nonalcoholic steatohepatitis (Chedid et al., 2004) and alcoholic hepatitis (Gadd et al., 2014), deserves further examination.
Participated in research design: Maeda, Hirata, Watanabe, Ishima, Inatsu, Kinoshita, Otagiri, Maruyama.
Conducted experiments: Maeda, Hirata, Tanaka, Sasaki.
Contributed new reagents or analytic tools: Maeda, Watanabe, Ishima, Inatsu, Kinoshita, Tanaka, Sasaki, Maruyama.
Performed data analysis: Maeda, Watanabe, Ishima, Inatsu, Kinoshita, Tanaka, Sasaki, Maruyama.
Wrote or contributed to the writing of the manuscript: Maeda, Watanabe, Ishima, Chuang, Taguchi, Kinoshita, Tanaka, Sasaki, Otagiri, Maruyama.
- Received August 24, 2014.
- Accepted November 3, 2014.
This research was supported, in part, by Grant-in-Aid for Scientific Research from Japan Society for the Promotion of Science (JSPS) (KAKENHI 21390177).
- alanine aminotransferase
- aspartate aminotransferase
- mannose receptor
- 5,5-dimethyl-1-pyrroline I-oxide
- 5,5′-dithiobis-(2-nitrobenzoic acid)
- diethylenetriamine-pentaacetic acid
- electron paramagnetic resonance
- reduced/oxidized glutathione ratio
- hypoxia-inducible factor-1α
- human serum albumin
- Kupffer cells
- potassium phosphate buffer
- N-acetyl cysteine
- N-acetyl-p-benzoquinone imine
- phosphate-buffered saline
- recombinant HSAs
- reactive oxygen species
- tumor necrosis factor
- terminal deoxynucleotidyl transferase dUTP nick-end labeling
- Copyright © 2014 by The American Society for Pharmacology and Experimental Therapeutics