Gene transfer of high-mobility group box 1 box-A domain in a rat acute liver failure model

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.JournalofSurgicalResearch.com

Gene transfer of high-mobility group box 1 box-A domain in a rat acute liver failure model Masayuki Tanaka, MD,a Masahiro Shinoda, MD, PhD,a,* Atsushi Takayanagi, PhD,b Go Oshima, MD, PhD,a Ryo Nishiyama, MD, PhD,a Kazumasa Fukuda, PhD,a Hiroshi Yagi, MD, PhD,a Tetsu Hayashida, MD, PhD,a Yohei Masugi, MD, PhD,c Koichi Suda, MD, PhD,a Shingo Yamada, PhD,d Taku Miyasho, DVM, PhD,e Taizo Hibi, MD, PhD,a Yuta Abe, MD, PhD,a Minoru Kitago, MD, PhD,a Hideaki Obara, MD, PhD,a Osamu Itano, MD, PhD,a Hiroya Takeuchi, MD, PhD,a Michiie Sakamoto, MD, PhD,c Minoru Tanabe, MD, PhD,f Ikuro Maruyama, MD, PhD,g and Yuko Kitagawa, MD, PhDa a

Department of Surgery, Keio University School of Medicine, Shinjuku, Tokyo, Japan Department of Molecular Biology, Keio University School of Medicine, Shinjuku, Tokyo, Japan c Department of Pathology, Keio University School of Medicine, Shinjuku, Tokyo, Japan d Central Institute, Shino-Test Corporation, Sagamihara, Kanagawa, Japan e Department of Veterinary Science, School of Veterinary Medicine, Rakuno Gakuen University, Ebetsu, Hokkaido, Japan f Department of Surgery, Tokyo Medical and Dental University, Bunkyo, Tokyo, Japan g Department of Laboratory and Vascular Medicine, Kagoshima University Graduate School of Medicine and Dental Sciences, Kagoshima, Kagoshima, Japan b

article info

abstract

Article history:

Background: High-mobility group box 1 (HMGB1) has recently been identified as an important

Received 25 May 2014

mediator of various kinds of acute and chronic inflammation. The protein encoded by the

Received in revised form

box-A domain of the HMGB1 gene is known to act as a competitive inhibitor of HMGB1. In this

13 October 2014

study, we investigated whether box-A gene transfer results in box-A protein production in

Accepted 13 November 2014

rats and assessed therapeutic efficacy in vivo using an acute liver failure (ALF) model.

Available online 22 November 2014

Materials and methods: Three types of adenovirus vectors were constructedda wild type and two mutantsdand a mutant vector was then selected based on the secretion from HeLa

Keywords:

cells. The secreted protein was subjected to a tumor necrosis factor (TNF) production in-

Acute liver failure

hibition test in vitro. The vector was injected via the portal vein in healthy Wistar rats to

High-mobility group box 1

confirm box-A protein production in the liver. The vector was then injected via the portal

Adenovirus vectors

vein in rats with ALF.

Box-A

Results: Western blot analysis showed enhanced expression of box-A protein in HeLa cells

Gene delivery

transfected with one of the mutant vectors. The culture supernatant from HeLa cells transfected with the vector inhibited TNF-a production from macrophages. Expression of

* Corresponding author. Department of Surgery, Keio University School of Medicine, 35 Shinanomachi, Shinjuku, Tokyo 160 8582, Japan. Tel.: þ81 3 3353 1211/ x. 62334; fax: þ81 3 3355 4707. E-mail address: [email protected] (M. Shinoda). 0022-4804/$ e see front matter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2014.11.022

572

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

box-A protein was confirmed in the transfected liver at 72 h after transfection. Transfected rats showed decreased hepatic enzymes, plasma HMGB1, and hepatic TNF-a messenger RNA levels, and histologic findings and survival were significantly improved. Conclusions: HMGB1 box-A gene transfer results in box-A protein production in the liver and appears to have a beneficial effect on ALF in rats. ª 2015 Elsevier Inc. All rights reserved.

1.

Introduction

High-mobility group box 1 (HMGB1) has recently been identified as an important mediator of various kinds of acute and chronic inflammation [1e3]. In animal models of shock, tissue injury and endotoxin-induced lethality, HMGB1 is released from necrotic tissues and causes systemic inflammation [1,4e7]. HMGB1 protein is a single-chain polypeptide of 215 amino acids consisting of a box-A domain, a box-B domain, and a C terminus. In 2003, Yang et al. [8] reported that injection of Box-A protein antagonized the inflammatory activity of HMGB1 in a murine sepsis model. The domain of box-A binds with receptors such as advanced glycation end products and toll-like receptor (TLR) without causing receptor activation or production of inflammatory mediators. HMGB1 and box-A have equal receptor binding affinities, which indicate that box-A acts as a competitive inhibitor of HMGB1. Subsequent to the report mentioned previously by Yang et al., other investigators have shown protective effects of recombinant box-A protein in various inflammatory conditions [9e12]. Gene transfer using an adenovirus vector is a widely used method for promoting expression of a protein encoded by the transferred gene in the liver [13e16]. Adenoviral gene transfer enables continuous production of protein from the liver, although the enhancement of protein production is not permanent, for example, lasting for several days only [15]. Therefore, adenoviral gene transfer is best suited for acute

inflammatory diseases. Recombinant box-A protein is commercially available, but is currently expensive at the doses required for large animals and humans. Box-A gene transfer represents a potentially cost-effective alternative to administration of the protein. In this study, we hypothesized that gene transfer of HMGB1 box-A results in production of box-A protein, thereby attenuating HMGB1-induced inflammation in vivo, and conducted an experiment using a rat model of acute liver failure (ALF), in which the role of HMGB1 has been thoroughly investigated [15e19].

2.

Materials and methods

2.1. Construction of secretory human HMGB1 box-A complimentary DNA and construction of adenoviral vector Human HMGB1 box-A complementary DNA (cDNA) was amplified from the Human Liver Marathon-Ready cDNA library (Clontech, CA) by polymerase chain reaction (PCR) with the following primers: 50 - GAGAGAAAGCTTATGGGCAAA GGAGATCCTAAGAAGCCGA - 30 and 50 - CTCTCGAATTCATT ATGTCTCCCCTTTGGGAGGGATA -3’ (underlining indicates the restriction-enzyme sites of HindIII and EcoRI, respectively). A plasmid vector, p3xFLAG-sHMGB1 box-A, which expresses secretory human HMGB1 box-A with the preprotrypsin signal peptide and consequent N-terminal 3xFLAG tag (Fig. 1), was

Fig. 1 e Recombinant HMGB1 box-A genes in the adenovirus vectors. The wild-type vector encodes 86 amino acids of box-A protein. Mutant vectors were also constructed replacing aspartic acid at position 37 or serine at position 39 with alanine.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

constructed by inserting the amplified fragment digested with HindIII and EcoRI into the HindIII-EcoRI site of the plasmid p3xFLAG-CMV-8 (SigmaeAldrich, MO). The sequence of the insert was then confirmed. In addition to wild-type box-A cDNA, two mutant cDNAs (N37 A and S39 A shown in Fig. 1), which disrupted a putative glycosylation site, were constructed by PCR and cloned similarly. In the first mutant (N37 A), amino acid 37 was switched from aspartic acid to alanine. In the second mutant (S39 A), amino acid 39 was switched from serine to alanine. The replication-defective adenoviral vector containing the CAG promoter, Escherichia coli lacZ gene, and poly(A) sequence (the adenovirus vector encoding [Adex] LacZ) was kindly provided by Dr I Saito, Institute of Medical Science, University of Tokyo [20]. Another replication-defective adenoviral vector (Adex box-A [wild type, W]), which expresses the secretory human HMGB1 box-A gene derived from p3xFLAG-sHMGB1 box-A, was constructed using an adenovirus expression vector kit (Takara Bio, Tokyo, Japan). The amplified fragment containing the coding region with the primers 50 - TTGATTTATCGATTTGCCACCATGTCTGC ACTTATGATCCTAGCTC - 30 and 50 -ACTAGTTTATAATTTGG GGAGGGGTCACAGGGATGCCA-30 from p3xFLAG-sHMGB1 box-A and two plasmids with the mutants were inserted into the unique SwaI site of the adenovirus genome in the cassette cosmid pAxCAwtit2. After sequencing of the coding region, the cosmid bearing an expression unit was cotransfected into the human embryonic kidney 293 cells, together with the adenovirus DNA-terminal protein complex. The cloned recombinant adenoviruses (designated as Adex box-A [W], Adex box-A [N37 A], and Adex box-A [S39 A], respectively) were purified by anion-exchange chromatography using Sartobind Q100X (Sartorius Stedim Biotech, Gottingen, Germany) as follows: human embryonic kidney 293 cells in 5  225 cm2 culture flasks with medium were harvested 4 d after full infection (multiplicity of infection > 50). The cell culture medium and the cell pellets were separated by centrifugation at 1000g. The supernatants were mixed of 2.5 M NaCl, 20% (w/v) polyethylene glycol (average molecular weight 6000 Da, SigmaeAldrich) at a 4:1 ratio. Adenoviruses precipitated by centrifugation at 10,000g after overnight cooling at 4 C were suspended in Hanks Balanced Salt Solution (Sigmae Aldrich). The cell pellets were resuspended in three volumes of Hanks Balanced Salt Solution with 1:1000 Benzonase Nuclease (purity >99%, Novagen, WI) and 1 mM MgCl2, and then fractured by three cycles of freeze-thawing. After centrifugation at 10,000g, the resulting supernatants were mixed with the corresponding adenovirus solution from cell culture supernatants and filtered through a 0.45-mm membrane filter. The clarified adenovirus solutions were adjusted to contain 380-mM NaCl by adding 0.25-M Tris-HCl (pH 8), 2.5M NaCl, and 150-mg/L phenol red at a 1:10 ratio. The adenoviruses were bound to Sartobind Q100X columns equilibrated with 20-mM Tris-HCl (pH 8) and 380-mM NaCl. After washing with 20 column volumes (60 mL) of 20 mM Tris-HCl (pH 8) and 450 mM NaCl, the viruses were eluted with 15 mL of a solution consisting of 20-mM Tris-HCl (pH 8), 620-mM NaCl, and 15mg/L phenol red. The purified virus was concentrated and resuspended in Dulbecco phosphate-buffered saline (Gibco, New York, MA) containing 10% glycerol by Amicon Ultra-15 Ultracel-100 (Millipore, Billerica). The adenoviruses were

573

aliquoted and stored at 80 C. Virus titers were assessed with an Adeno-X Rapid Titer Kit (Clontech) and expressed as plaque-forming units (pfu).

2.2. In vitro gene delivery and biologic activity assay of secreted box-A protein HeLa cells were infected with Adex box-A (W), Adex box-A (N37 A), and Adex box-A (S39 A) (multiplicity of infection ¼ 10) in Dulbecco modified Eagle’s medium (Gibco) with 10% fetal bovine serum (Gibco) and incubated for 48 h at 37 C. Cells were lysed with Laemmli sodium dodecyl sulfate sample buffer (Invitrogen, CA). The supernatant and cell extracts were collected from HeLa cells transfected with each vector type and subjected to Western blot analysis. The samples were boiled at 100 C for 5 min and separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis on 5%e20% gradient gels (Oriental Instruments, Kanagawa, Japan). Blotting onto Immobilon-P polyvinylidene difluoride membranes (Millipore) and blocking with 5% Perfect-Block (MoBiTec GmbH, Gottingen, Germany) were performed, and the membranes were then incubated with anti-FLAG monoclonal antibody M2 (cat# F3165; SigmaeAldrich) as a primary antibody and horseradish peroxidase-conjugated goat antiemouse immunoglobulin (cat# P0447; Dako, Glostrup, Denmark) as a secondary antibody. Detection was performed using ECL Prime Western blotting detection reagent (GE Healthcare, WI) and digitized using a LAS1000plus imaging apparatus (Fujifilm, Tokyo, Japan). The biologic activity of HMGB1 box-A protein secreted from transfected HeLa cells was assessed using a previously reported method [8]. Murine macrophage-like RAW 264.7 cells (RAW 264.7, kindly provided by Dr Horiuchi K, Department of Orthopedic Surgery, Keio University School of Medicine, Tokyo, Japan) were cultured in Roswell Park Memorial Institute 1640 medium (Gibco) supplemented with 10% fetal bovine serum, penicillin, and streptomycin (Gibco). Subconfluent cultures of RAW 264.7 cells were incubated in serum-free Opti-MEM I medium (Gibco) with recombinant HMGB1 protein (1 mg/mL) for 8 h with the following supplementation: i) no culture supernatant supplementation; ii), iii) culture supernatant from HeLa cells transfected with Adex box-A (S39 A) (10 mL/1000 mL and 100 mL/1000 mL, respectively); and iv), v) culture supernatant from HeLa cells transfected with adenovirus vector encoding E coli lacZ (Adex LacZ) (10 mL/1000 mL and 100 mL/1000 mL, respectively). The conditioned medium was assayed for tumor necrosis factor (TNF)-a using a commercial enzyme-linked immunosorbent assay kit (R&D Systems, MN).

2.3.

In vivo gene delivery into the rat liver

Male Wistar rats, weighing 200e250 g, were used for this study. All animals were acclimated to the animal research laboratory for 5 days before experiments and were maintained in a light-controlled room (12-h lightedark cycle) at an ambient temperature of 25 C with chow diet and water ad libitum. The animals were cared for in accordance with the guidelines set forth by Laboratory Animal Resources of Keio University School of Medicine.

574

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

Healthy rats were anesthetized by inhalation of isoflurane. Adex box-A (S39 A) was injected via the portal vein (2.6  109 pfu per injection, in 2-mL saline). Livers were harvested at 6, 24, and 72 h after injection of vector via the portal vein for immunofluorescence staining and Western blot analysis. For the immunofluorescence staining of FLAG tag (the integrated marker of box-A protein), paraformaldehyde-fixed liver specimens were embedded in paraffin. Sections 3-mm thick were autoclaved in 10-mM sodium citric acid (pH 6) and then incubated at room temperature with a primary antibody (mouse anti-FLAG M2 monoclonal antibody cat# F1804, SigmaeAldrich) at a 1:800 dilution for 120 min and then a secondary antibody (Alexa Fluor 488-labeled goat antiemouse IgG cat# A11001, Invitrogen) at a 1:800 dilution for 60 min. Sections were observed under a Nikon E1000 fluorescence microscope (Nikon, New York, NY) using a B-2A filter, 450 nme490 nm wavelength for excitation and 505 nm for emission. For the observation of cell nuclei (40 ,6-diamidino-2-phenylindole cat# D1306; Molecular Probes, OR), staining was used (a UV-1A filter). For Western blot analysis, the liver was minced in 3-mm particles and then homogenized in radio-immunoprecipitaion assay lysis buffer. The homogenate was centrifuged (1000g) for 10 min at 4 C to remove cell debris (MX300; TOMY, Tokyo, Japan). The supernatant was subjected to Western blot analysis according to the method described previously.

2.4.

Effects of box-A gene delivery on ALF in rats

2.4.1.

Experimental design

Male Wistar rats, weighing 200e250 g, were randomly assigned to two groups as follows: Adex box-A and Adex LacZ. In the Adex box-A group, Adex box-A (S39 A) (2.6  109 pfu per injection) was injected in the same way described previously 72 h before ALF induction. In the Adex LacZ group, Adex LacZ was injected instead. To induce ALF, D-galactosamine (SigmaeAldrich) (freshly dissolved in physiological saline and adjusted to pH 6.8 with 1-N NaOH) was injected intraperitoneally at doses of 2.0 g/kg and 2.8 g/ kg for the blood and liver collection study and the survival observation study, respectively. After the D-galactosamine injection, the rats had free access to food and water until sample collection. Blood samples were collected immediately before and at 12, 24, and 48 h after ALF induction by punctuating the abdominal aorta. Liver samples were collected immediately before and at 24 and 48 h after ALF induction for real-time quantitative PCR analysis and histologic examination. A total of 16 rats per group were used for blood collection and 9 per group were used for liver collection. Survival was assessed every 12 h for 10 d after ALF induction. Rats were sacrificed for sample collection at each time point. Six rats per group were used in the survival study.

2.4.2.

Determination of blood biochemistry and HMGB1

Plasma was separated by centrifugation and stored immediately at 70 C before the determination of the levels of total bilirubin (TB), aspartate aminotransferase (AST), alanine aminotransferase (ALT) (SRL laboratory, Tokyo, Japan), and HMGB1 (Shino-Test, Kanagawa, Japan).

2.4.3. Real-time PCR analysis for interleukin-6 and TNF-a in hepatic tissue Real-time quantitative PCR analysis was performed according to the following methods. Total RNA was extracted using a commercial kit (RNeasy; Qiagen, CA). The concentration of total RNA was determined using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, CA). First-strand cDNA was synthesized from total RNA using ReverTra Ace reverse transcriptase (Toyobo, Osaka, Japan) according to the manufacturer’s manual. PCR reaction mixes were prepared using template cDNA samples and Eagle Taq Master Mix with ROX (Roche Applied Science, IN), and the expression of human interleukin (IL)-6 and TNF-a was analyzed using an Applied Biosystems 7500 Fast real-time PCR system (Applied Biosystems, CA). TaqMan gene expression assay primer and probe mixes were used to detect glyceraldehyde 3-phosphate dehydrogenase, IL-6, and TNF-a (assay IDs: Hs99999905_m1, Rn01410330_m1, and Rn01525859_g1, respectively, produced by Applied Biosystems). The thermal cycling reaction included incubation at 95 C for 20 s followed by 40 cycles at 95 C for 3 s and 60 C for 30 s. Data were collected using analytical software (Applied Biosystems). The expression level of each gene was determined using the ⊿⊿CT method.

2.4.4.

Histology and immunohistochemistry

For the histologic examination, the liver samples were processed as described in the following. Paraformaldehyde-fixed specimens were embedded in paraffin. Sections 5-mm thick were cut and stained with hematoxylin and eosin. For the immunohistochemical examination of HMGB1, sections were incubated with primary antibody (1:1000 dilution, 60 min at room temperature; mouse antieHMGB1 monoclonal antibody; Shino-Test) and a secondary antibody (1:1000 dilution, 60 min at room temperature; peroxidaseconjugated donkey antiemouse IgG cat# 715-035-151, Jackson ImmunoResearch Laboratories, PA). The sections were visualized with 3,3-diaminobenzidine tetrahydrochloride solution (Dako). Using the same specimens, control sections, which were processed in parallel except for the incubation with primary antibody, were prepared for each sample. All histologic findings were carefully assessed by two pathologists.

2.4.5.

Image analysis of immunohistochemistry for HMGB1

Images of immunohistochemistry for HMGB1 were analyzed using a semiquantitative method reported by Takano et al. [17]. In brief, immunohistochemical staining images were digitally converted to monochrome images using ImageJ software (rsb.info.nih.gov/ij/). In a 400 view field, the gray levels were measured at both cytoplasmic and nuclear domains in individual hepatocytes. The intensity of each HMGB1-positive nucleus was defined by the following equation: intensity of HMGB1-positive nuclei ¼ (gray level of nuclear domain)/(gray level of cytoplasmic domain). The average intensity of each HMGB1-positive nucleus was calculated, and the number of HMGB1-positive nuclei was counted. Finally, the total intensity of HMGB1-positive nuclei was assessed multiplying average intensity of HMGB1-positive nuclei and the number of HMGB1-positive nuclei. Five different sites for each rat’s liver were randomly chosen to calculate a representative value per rat.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

575

Fig. 2 e In vitro gene delivery and biologic activity assay of HMGB1 box-A. (A) Western blot analysis of HMGB1 box-A proteins. Recombinant HMGB1 box-A proteins in the culture supernatant and cell extracts of HeLa cells infected with the indicated adenovirus vectors were detected with anti-FLAG antibody. (B) Biologic activity assay of HMGB1 box-A. HMGB1 indicates TNF-a amounts produced by RAW 264.7 cells per 1 mL after 8-h incubation with recombinant full-length HMGB1 protein (1 mg/mL). HMGB1DAdex box-A10 and HMGB1DAdex box-A100 indicate TNF-a amounts when culture supernatant of HeLa cells transfected with Adex box-A (S39 A; 10 mL/1,000 mL and 100 mL/1,000 mL, respectively) is supplemented. HMGB1DAdex LacZ10 and HMGB1DAdex LacZ100 indicate TNF-a amounts when culture supernatant of HeLa cells transfected with Adex LacZ (10 mL/1,000 mL and 100 mL/1,000 mL, respectively) is supplemented. Results are expressed as mean D SD (n [ 4). #P < 0.05 versus HMGB1.

2.5.

Statistical analysis

Results are expressed as mean  standard deviation. For parametric data, differences between groups were evaluated using Student t-test for unpaired data, based on the assumption that the data were derived from populations with equal SDs. If the difference between the two SDs was significant, Welch t-test was used. Animal survival data for 10 d were evaluated using the KaplaneMeier test and the generalized Wilcoxon test. Differences were considered significant at P values <0.05.

3.

Results

3.1. In vitro box-A protein secretion from transfected HeLa cells To select a vector from the three types of vectors for this study, we assessed box-A protein secretion from transfected HeLa cells. Western blot analysis showed strong expressions of box-A protein in the cell extracts of transfected HeLa cells

Fig. 3 e Immunofluorescence staining of liver transfected with vector. (A) 6 h, (B) 24 h, and (C) 72 h after injection of the vector (Adex box-A [S39 A], 1.0 3 109 pfu) via the portal vein. Green immunofluorescence represents staining of FLAG tag, the integrated marker of box-A protein. Blue immunofluorescence represents 40 ,6-diamidino-2-phenylindole-staining of cell nuclei. Pictures of green immunofluorescence and blue immunofluorescence are digitally combined in a single picture. Original magnification is 3200. A representative sample is presented for each time point. (D) Western blot analysis of liver transfected with vector at 72 h after injection of the vector (Adex box-A [S39 A], 2.6 3 109 pfu) via the portal vein.

576

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

Fig. 4 e Effects of box-A gene transfer on blood parameters and cytokine mRNA. Plasma levels of (A) TB, (B) AST, (C) ALT, and (D) HMGB1 immediately before and at 12, 24, 48 h after ALF induction. (E) TNF-a mRNA and (F) IL-6 mRNA expression in liver tissue at 24 h after ALF induction. Adex box-A (S39 A) (solid line) or Adex LacZ (broken line) (2.6 3 109 pfu in each group) vector was injected via the portal vein 72 h before ALF induction. Data represent mean D SD or mean L SD in (AeD) (n [ 4 per group at each time point); and mean D SD in (E) and (F) (n [ 3 per group at each time point). The mRNA levels are normalized relative to the levels immediately before ALF induction. #P < 0.05 versus Adex LacZ at the same time point.

with all the vector types. Expression in the culture supernatant was most enhanced by transfection of Adex box-A (S39 A) (Fig. 2A). Therefore, Adex box-A (S39 A) was used in the following in vitro and in vivo experiments.

3.2. TNF production-inhibiting activity of supernatant from transfected HeLa cells

significantly suppressed when culture supernatant of HeLa cells transfected with Adex box-A (S39 A) was added to the culture (Fig. 2B). On the other hand, the concentration did not show marked changes when culture supernatant of HeLa cells transfected with Adex LacZ, which serves as a control for Adex box-A (S39 A), was added to the culture.

3.3. To assess the biologic activity of HMGB1 box-A protein secreted from transfected HeLa cells, we performed a procedure previously reported by Yang et al. [8]. In this assay, the biologic activity of HMGB1 box-A was considered to be represented by the ability of box-A protein to inhibit production of TNF-a from macrophages that had been stimulated (24 h at 37 C) with recombinant HMGB1 protein. The TNF-a concentration in the culture supernatant of stimulated macrophage was 402  22 pg/mL and was dose-dependently and

Box-A protein production in rat liver

Immunofluorescence staining for FLAG tag (the integrated marker of box-A protein) of the liver showed a few green fluorescence cells at 6 h after injection of Adex box-A via the portal vein (Fig. 3A). The number of green fluorescence cells increased in a time-dependent manner at 24 and 72 h after transfection (Fig. 3B and C). Western blot analysis also showed a clear expression of box-A protein in the liver at 72 h after transfection (Fig. 3D).

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

577

Fig. 5 e Effects of box-A gene transfer on liver histology. Findings of hematoxylin and eosin staining of the liver at 48 h after ALF induction in the Adex LacZ group (A and B) and in the Adex box-A group (C and D). Immunohistochemical staining for HMGB1 in liver immediately before ALF induction (E); at 48 h after ALF induction in the Adex box-A group (F); and at 48 h

578 3.4.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

Effects of gene transfer of box-A on ALF in rats

3.4.1. Blood biochemistry, plasma HMGB1, and TNF-a and IL-6 messenger RNA levels The levels of TB, AST, and ALT were increased from baseline in a time-dependent manner at 12, 24, 48 h after ALF induction in the Adex LacZ group. The levels were significantly lower at 48 h after ALF induction in the Adex box-A group than those in the Adex LacZ group (Fig. 4AeC). A similar trend with the plasma levels of HMGB1 was observed in both groups. HMGB1 levels differed significantly between groups at 48 h after ALF induction (Fig. 4D). The TNF-a and IL-6 messenger RNA (mRNA) levels were markedly suppressed in the Adex box-A group at 24 h after ALF induction compared with those in the Adex LacZ group (Fig. 4E and F). There was a significant difference in TNF-a mRNA levels between groups.

3.4.2.

Histology and immunohistochemistry

Hematoxylin and eosin staining of the liver showed hepatocyte ballooning, single cell necrosis, and the presence of inflammatory cells such as plasma cells and lymphocytes in the Adex LacZ group at 48 h after ALF induction (Fig. 5A and B). In contrast, these findings were markedly improved in the Adex box-A group at 48 h after ALF induction (Fig. 5C and D). To assess HMGB1 preservation in the nuclei, we performed immunohistochemical staining for HMGB1of the liver. The staining showed strong and clear staining in the nuclei of hepatocytes before ALF induction (Fig. 5E). Staining was preserved to some extent in the nuclei in the Adex Box-A group (Fig. 5F), but was almost absent in the nuclei and the cytoplasm of hepatocytes in the Adex LacZ group at 48 h after ALF induction (Fig. 5G). To determine changes in the HMGB1 preservation in the nuclei, we used a previously reported semiquantitative method [17]. The analysis showed that the average intensity of HMGB1-positive nuclei, number of HMGB1-positive nuclei, and total intensity of HMGB1-positive nuclei were significantly better preserved in the Adex box-A group compared with those in the Adex LacZ group at 48 h after ALF induction (Fig. 5HeJ).

3.4.3.

Survival

Since the data mentioned previously suggested that adenovirus (S39 A) treatment would palliate liver inflammation, we investigated the effect of the adenovirus treatment on survival after ALF induction in rats (Fig. 6). The survival curves showed that the adenovirus treatment extended the life of ALF rats significantly (33% and 67% at 48 h and 0% and 50% at 96 h after ALF induction in the Adex LacZ group and the Adex box-A group, respectively).

=

4.

Discussion

HMGB1 induces inflammation by binding to multiple separate receptors: for example, glycation end products, TLR2, TLR4, TLR9, Mac-1, syndecan-1, phosphacan protein-tyrosine phosphatase-z/b, and CD24 [21e33]. Two different cascades for intracellular signaling through the receptors have been proposed; one involves the small GTPases Rac and Cdc42 and the other involves the Ras-mitogen-activated protein kinase pathway and subsequent nuclear factor-kB nuclear translocation-mediated inflammation [2,34]. Because box-A protein inhibits HMGB1 from binding to receptors, it is possible that box-A protein produced by gene transfer also inhibits HMGB1 to bind with some or all the receptors and suppresses the downstream portion of the signaling pathway described previously. Box-A protein is a component of HMGB1, but it is not produced physiologically under either normal or inflammatory conditions. It was unknown whether it would be feasible to use the gene delivery method to promote the production and secretion of this “unphysiological” protein. We were concerned that unexpected glycosylation could inhibit secretion and function of the protein because amino acid sequence 37e39 of box-A protein, Asn-Phe-Ser, matched the consensus sequence (Asn-X-Ser), which sometimes leads to N-linked glycosylation [35]. It was noteworthy that replacement of a constituent amino acid in cDNA increased protein secretion from host cells. We also performed an in vitro experiment to confirm whether secreted box-A protein possesses biological activity against HMGB1 using the protocol reported by Yang et al. [8]. This study successfully showed that the secreted mutant box-A protein potentially functions to inhibit HMGB1 and prevent it from stimulating macrophages. To investigate whether transfection of Adex box-A (S39 A) results in box-A protein production in vivo, we injected the vector via the portal vein in healthy rats. The findings from immunofluorescence staining and Western blot analysis strongly suggested that box-A protein was produced in the liver in the acute phase after transfection. Investigators have reported successful production of box-A protein in vitro using plasmid vector [36], but, to our knowledge, this is the first report of successful production of box-A protein in vivo by means of adenoviral gene transfer. To investigate potential beneficial effects of gene transfer using the vector in vivo, we used a D-galactosamine-induced ALF model in rats [15e19]. As we reported previously [17], administration of D-galactosamine at a dose of 2.0 g/kg resulted in severe intoxication in rats. However, pretreatment with the vector 72 h before ALF induction led to remarkable outcomes; for example, suppression of TB, AST, and ALT; improvement of histology; and

after ALF induction in the Adex LacZ group (G). The images of immunohistochemical staining for HMGB1 are digitally converted as monochrome images and analyzed as in the following using the previously reported method [18]. Original magnification; 3100 in (A) and (C); and 3400 in (B), (DeG). A representative sample is presented for each examination. Image analysis parameters of average intensity of HMGB1-positive nuclei (H), number of HMGB1-positive nuclei (I), and total intensity of HMGB1-positive nuclei (J) before and at 48 h after ALF induction. The results indicated that the HMGB1 content, which was well contained in the nuclei before ALF induction, was better preserved in the Adex box-A group compared with that in the Adex LacZ group at 48 h after ALF induction. Results are expressed as mean D SD. #P < 0.05 versus before ALF induction. *P < 0.05 versus Adex box-A at 48 h after ALF induction. (Color version of figure is available online.)

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

579

encoded box-A protein are needed, the strategy of box-A protein gene delivery may have great therapeutic significance for not only the treatment of ALF but also other inflammatory conditions such as sepsis and ischemiae reperfusion injury.

Acknowledgment

Fig. 6 e Effect of box-A gene transfer on survival in the rat ALF model. Survival curves of the Adex LacZ group (broken line) and Adex box-A group (solid line) (n [ 6).

improvement of survival rate. The significant suppression of these parameters suggests that this novel therapeutic approach, gene transfer of box-A, would be effective for suppressing inflammation, as would box-A protein administration. As for the mechanism underlying the beneficial effects observed in the ALF model, we speculate that box-A protein was successfully produced in the liver and that it bound to receptors on hepatocytes, interrupted downstream intracellular signaling in hepatocytes, and decreased inflammatory cytokine RNA transcription. In fact, suppression of IL-6 and TNF-a mRNAs was confirmed in this study. We have previously shown that blockade of inflammatory cytokines improves hepatic tissue injury and survival rate in a rat ALF model [15e17]. Less injury to hepatic tissue might result in better preservation of HMGB1 in nuclei, decreased release of HMGB1 from injured hepatocytes, and decreased plasma HMGB1 level. We are interested in investigating various receptors and signaling mediators in this ALF model using box-A gene transfer in a future study. We are also encouraged to conduct further studies of box-A gene transfer using other inflammatory models such as those for sepsis and ischemiaereperfusion. Another area of interest is the biological safety of this therapeutic approach. Although we have observed no major systemic adverse effects in healthy rats in which mutant box-A protein is produced by the gene transfer method, investigating the effects of box-A gene transfer on normally functioning cells and confirming the biological safety of the produced mutant protein should also be included in a future study.

5.

Conclusions

In conclusion, it was possible to produce HMGB1 box-A protein in the liver of rats using our adenovirus vector. Production of HMGB1 box-A protein in the liver appeared to have a therapeutic effect in a rat model of ALF. Although optimization of vector construction and the amino acid sequence of the

This work was partially supported by grants-in-aid from the Ministry of Education, Science, and Culture of Japan (#23591875). The authors thank the late Dr Akitoshi Ishizaka for his invaluable advice during the investigation and Ms Katsuko Sano for her technical assistance. Author’s contributions: M.T. performed the experiment and wrote the article. M.S. and A.T. designed the experiment. K.S., T.H., Y.A., M.K., H.O., O.I., H.T., M.T., M.S., A.T., Y.K., and I.M. supervised the writing of the article. G.O., R.N., K.F., H.Y., T.H., S.Y., and T.M. performed the experiment. Y.M. and M.S. were concerned with pathology. K.S., T.H., Y.A., M.K., H.O., O.I., H.T., M.T., and I.M. provided the study concept. Y.K. represents our surgical department. All authors read and approved the final article.

Disclosure The authors reported no proprietary or commercial interest in any product mentioned or concept discussed in the article.

references

[1] Edward A, John A, Aaron C, Haichao W, Tracey KJ. HMG-1 as a mediator of acute lung inflammation. J Immunol 2000;165: 2950. [2] Andersson U, Tracey KJ. HMGB1 is a therapeutic target for sterile inflammation and infection. Annu Rev Immunol 2011; 29:139. [3] Harris HE, Andersson U, Pisetsky DS. HMGB1: a multifunctional alarmin driving autoimmune and inflammatory disease. Nat Rev Rheumatol 2012;8:195. [4] Suda K, Kitagawa Y, Ozawa S, et al. Anti-high-mobility group box chromosomal protein 1 antibodies improve survival of rats with sepsis. World J Surg 2006;30:1755. [5] Tsung A, Sahai R, Tanaka H, et al. The nuclear factor HMGB1 mediates hepatic injury after murine liver ischemiareperfusion. J Exp Med 2005;201:1135. [6] Wang H, Bloom O, Zhang M, et al. HMG-1 as a late mediator of endotoxin lethality in mice. Science 1999;285:248. [7] Watanabe T, Kubota S, Nagaya M, et al. The role of HMGB-1 on the development of necrosis during hepatic ischemia and hepatic ischemia/reperfusion injury in mice. J Surg Res 2005; 124:59. [8] Yang H, Ochani M, Li J, et al. Reversing established sepsis with antagonists of endogenous high-mobility group box 1. Proc Natl Acad Sci U S A 2004;101:296. [9] Andrassy M, Volz HC, Igwe JC, et al. High-mobility group box1 in ischemia-reperfusion injury of the heart. Circulation 2008;117:3216. [10] Muhammad S, Barakat W, Stoyanov S, et al. The HMGB1 receptor RAGE mediates ischemic brain damage. J Neurosci 2008;28:12023.

580

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 4 ( 2 0 1 5 ) 5 7 1 e5 8 0

[11] Sitia G, Iannacone M, Muller S, Bianchi ME, Guidotti LG. Treatment with HMGB1 inhibitors diminishes CTL-induced liver disease in HBV transgenic mice. J Leukoc Biol 2007;81:100. [12] Urbonaviciute V, Furnrohr BG, Meister S, et al. Induction of inflammatory and immune responses by HMGB1nucleosome complexes: implications for the pathogenesis of SLE. J Exp Med 2008;205:3007. [13] Harada H, Wakabayashi G, Takayanagi A, et al. Transfer of the interleukin-1 receptor antagonist gene into rat liver abrogates hepatic ischemia-reperfusion injury. Transplantation 2002;74:1434. [14] Ono S, Obara H, Takayanagi A, et al. Suppressive effects of interleukin-18 on liver function in rat liver allografts. J Surg Res 2012;176:293. [15] Shinoda M, Tilles AW, Kobayashi N, et al. A bioartificial liver device secreting interleukin-1 receptor antagonist for the treatment of hepatic failure in rats. J Surg Res 2007;137:130. [16] Shinoda M, Tilles AW, Wakabayashi G, et al. Treatment of fulminant hepatic failure in rats using a bioartificial liver device containing porcine hepatocytes producing interleukin-1 receptor antagonist. Tissue Eng 2006;12:1313. [17] Takano K, Shinoda M, Tanabe M, et al. Protective effect of high-mobility group box 1 blockade on acute liver failure in rats. Shock 2010;34:573. [18] Oshima G, Shinoda M, Tanabe M, et al. Increased plasma levels of high mobility group box 1 in patients with acute liver failure. Eur Surg Res 2012;48:154. [19] Nishiyama R, Shinoda M, Tanabe M, et al. Hemadsorption of high-mobility group box 1 using a column for large animals. Eur Surg Res 2013;51:181. [20] Miyake S, Makimura M, Kanegai Y, et al. Efficient generation of recombinant adenoviruses using adenovirus DNAterminal protein complex and a cosmid bearing the fulllength virus genome. Proc Natl Acad Sci U S A 1996;93:1320. [21] Yang H, Tracey KJ. Targeting HMGB1 in inflammation. Biochim Biophys Acta 2010;1799:149. [22] Yu M, Wang H, Ding A, et al. HMGB1 signals through toll-like receptor (TLR) 4 and TLR2. Shock 2006;26:174. [23] Park JS, Svetkauskaite D, He Q, et al. Involvement of toll-like receptors 2 and 4 in cellular activation by high mobility group box 1 protein. J Biol Chem 2004;279:7370. [24] Park JS, Gamboni-Robertson F, He Q, et al. High mobility group box 1 protein interacts with multiple toll-like receptors. Am J Physiol, Cell Physiol 2006;290:C917.

[25] Apetoh L, Ghiringhelli F, Tesniere A, et al. Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Nat Med 2007; 13:1050. [26] Tian J, Avalos AM, Mao SY, et al. Toll-like receptor 9dependent activation by DNA-containing immune complexes is mediated by HMGB1 and RAGE. Nat Immunol 2007;8:487. [27] Chen GY, Tang J, Zheng P, Liu Y. CD24 and Siglec-10 selectively repress tissue damage-induced immune responses. Science 2009;323:1722. [28] Kuhla A, Norden J, Abshagen K, et al. RAGE blockade and hepatic microcirculation in experimental endotoxaemic liver failure. Br J Surg 2013;100:1229. [29] Tsung A, Klune JR, Zhang X, et al. HMGB1 release induced by liver ischemia involves toll-like receptor 4 dependent reactive oxygen species production and calcium-mediated signaling. J Exp Med 2007;204:2913. [30] Tsung A, Zheng N, Jeyabalan G, et al. Increasing numbers of hepatic dendritic cells promote HMGB1-mediated ischemiareperfusion injury. J Leukoc Biol 2007;81:119. [31] Fan J, Li Y, Levy RM, et al. Hemorrhagic shock induces NAD(P) H oxidase activation in neutrophils: role of HMGB1-TLR4 signaling. J Immunol 2007;178:6573. [32] Yang H, Hreggvidsdottir HS, Palmblad K, et al. A critical cysteine is required for HMGB1 binding to toll-like receptor 4 and activation of macrophage cytokine release. Proc Natl Acad Sci U S A 2010;107:11942. [33] Apetoh L, Ghiringhelli F, Tesniere A, et al. The interaction between HMGB1 and TLR4 dictates the outcome of anticancer chemotherapy and radiotherapy. Rev Immunol 2007;220:47. [34] Huttunen HJ, Fages C, Rauvala H. Receptor for advanced glycation end products (RAGE)-mediated neurite outgrowth and activation of NF-kappa B require the cytoplasmic domain of the receptor but different downstream signaling pathways. J Biol Chem 1999;274:19919. [35] Moremen KW, Tiemeyer M, Nairn AV. Vertebrate protein glycosylation: diversity, synthesis and function. Nat Rev Mol Cell Biol 2012;13:448. [36] Kikuchi H, Yagi H, Hasegawa H, et al. Therapeutic potential of transgenic mesenchymal stem cells engineered to mediate anti-high mobility group box 1 activity: targeting of colon cancer. J Surg Res 2014;190:134.