Using a Liver-Directed Vector Manipulating Multiple

Efficient Attenuation of NK Cell−Mediated Liver Injury through Genetically Manipulating Multiple Immunogenes by Using a Liver-Directed Vector This information is current as of September 30, 2021. Jianlin Geng, Xuefu Wang, Haiming Wei, Rui Sun and Zhigang Tian J Immunol published online 3 April 2013 http://www.jimmunol.org/content/early/2013/04/03/jimmun ol.1203129 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2013/04/03/jimmunol.120312

Material 9.DC1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 30, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published April 3, 2013, doi:10.4049/jimmunol.1203129 The Journal of Immunology

Efﬁcient Attenuation of NK Cell–Mediated Liver Injury through Genetically Manipulating Multiple Immunogenes by Using a Liver-Directed Vector

Jianlin Geng, Xuefu Wang, Haiming Wei, Rui Sun, and Zhigang Tian

Adenovirus or adenoviral vectors were reported to induce serious liver inflammation in an NK cell–dependent manner, which limits its clinical applicability for liver gene therapy. We tried to develop an efficient liver-directed therapeutic approach to control hepatic NK cell function via simultaneously manipulating multiple immune genes. Based on our previous study, we found that CCL5 knockdown synergistically enhanced the attenuating effect of silencing CX3CL1 (fractalkine [FKN]) in adenovirus-induced acute liver injury. In addition, the combined treatment of human IL-10 expression with FKN knockdown would further strengthen the protective effect of silencing FKN. We used a hepatocyte-specific promoter to construct a hepatocyte-specific Downloaded from multiple function vector, which could simultaneously overexpress human IL-10 and knock down CCL5 and FKN expression. This vector could attenuate adenovirus-induced acute hepatitis highly efficiently by reducing liver NK cell recruitment and serum IFN-g and TNF-a. The multiple function vectors could be delivered by nonviral (hydrodynamic injection) and viral (adenovirus) approaches, and maintained long-term function (more than 1 month in mice). Our results suggest a possible strategy to ameliorate the acute liver injury induced by adenovirus by modulating multiple immune genes. The novel multifunction vector has an extensive and practical use for polygenic and complex liver diseases such as malignancies and hepatitis, which correlate with http://www.jimmunol.org/ multiple gene disorders. The Journal of Immunology, 2013, 190: 000–000.

he liver is the largest solid organ in the body, and its ef- been developed and demonstrated to be an attractive gene therapy ficient protein synthetic capacity makes the liver an at- vector with highly reduced toxicity and long-term transduction ef- T tractive target organ for gene therapy (1, 2). Although ex- ficiency (5, 6). Second, chemically engineered vectors such as poly- citing results were obtained in animal models and in some clinical ethylene glycol modified adenoviruses exhibit receptor-specific trials, many obstacles remain in clinical practice. Among these cell transduction and reduced vector toxicity (7). Third, because obstacles, liver-specificity and immune injury are the most chal- adenovirus-induced strong innate responses are mainly mediated by lenging (3). Due to its liver tropism feature, ability to infect qui- innate immune cells such as NK cells, cell activities and functions by guest on September 30, 2021 escent hepatocytes, and high capacity for transgenes, the adenoviral can be regulated to ameliorate the liver injury. Blocking of NKG2D vector is now the most widely used vector in clinical trials (2). receptor inhibited NK cell activation upon adenovirus infection, Systemic infection of high-dose adenoviruses will initiate innate leading to a delay in adenovirus clearance and NK cell–mediated and adaptive immune responses against adenovirus that causes se- cytolysis (8). Knockdown of chemokines such as fractalkine (FKN) vere allergic reaction and inactivation of adenoviral vectors, which to inhibit NK cell recruitment and activity could protect the acute is the major drawback of adenovirus-mediated gene therapy (4). liver injury induced by adenoviral vector (9). Inflammatory cyto- A variety of strategies have been used to reduce the immunoge- kines such as IFN and TNF had important roles in adenovirus- nicity of adenovirus and inhibit the liver injury induced by adeno- induced acute liver injury; therefore, overexpression of inhibitory virus. First, gutless adenovirus lacking all viral coding regions has cytokines IL-10 has also been used to modulate the injury (10, 11). Finally, to achieve liver specificity in gene therapy, liver-specific promoters cannot only achieve prolonged transgene expression but Department of Immunology, School of Life Sciences, University of Science and Technology of China, Hefei, Anhui 230027, China; and Hefei National Laboratory also reduce side effects caused by transgene expression in non–liver for Physical Sciences at Microscale, Hefei, Anhui 230027, China cells (12, 13). Various liver-specific promoter elements and chi- Received for publication November 12, 2012. Accepted for publication March 4, meric promoters have been evaluated for high liver specificity and 2013. activity (14). This work was supported by the Ministry of Science and Technology of China (973 Because liver diseases, such as malignancies and hepatitis, Basic Science Project 2010CB911901), the Natural Science Foundation of China correlate with several immune gene disorders, a vector containing (91029303, 30911120480, and 31021061), and National Science and Technology Major Projects (2012ZX10002014, 2012ZX10002006). combinational gene intervention (e.g., simultaneous overexpression Address correspondence and reprint requests to Dr. Zhigang Tian and Dr. Rui Sun, and knockdown of multiple genes) is urgently needed to modulate School of Life Sciences, University of Science and Technology of China, 443 the inflammatory liver microenvironment during disease (15). Huangshan Road, Hefei 230027, China. E-mail addresses: [email protected] (Z.T.) Adenovirus-mediated coexpression of TRAIL and a short hairpin and [email protected] (R.S.). RNA (shRNA) containing a sequence to silence COX-2 exhibited The online version of this article contains supplemental material. great synergistic antitumor activity in experimental hepatocellular Abbreviations used in this article: ALT, alanine transaminase; ALP, alkaline phosphatase; AST, aspartate aminotransferase; BILI, bilirubin; CRE, creatinine; DC, carcinoma (16). Thus, development of vectors that can handle mul- dendritic cell; FCM, flow cytometry; FKN, fractalkine; GGT, g-glutamyltranspepti- tiple genes is practically needed. dase; huIL-10, human IL-10; miRNA, microRNA; RNAi, RNA interference; shRNA, In the current study, we found that synergistically interfering short hairpin RNA; VP, viral particle. CCL5 and FKN or the combined overexpression of human IL-10 Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 (huIL-10) with FKN knockdown could enhance the protective ef-

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1203129 2 ATTENUATING LIVER INJURY THROUGH MANIPULATING MULTIPLE GENES fect of silencing FKN alone in adenovirus-induced acute liver injury. Cryostat sections (6 mm thick) were prepared using a Leica CM1950 Next, we constructed a novel multifunction plasmid vector using an cryostat. EGFP and DsRed2 fluorescence protein expression was analyzed efficient hepatocyte-specific promoter, which could simultaneously with a fluorescence microscope (Carl Zeiss AXIOSCOP2, Oberkochen, Germany). overexpress and knock down multiple immune genes in a liver- directed manner. Our data show that simultaneous overexpression Mouse hepatic cell isolation of huIL-10 and knockdown of endogenous CCL5 and FKN highly Liver leukocytes were isolated similar to what has been described previ- attenuated adenovirus-mediated acute liver injury by inhibiting NK ously (18). Livers were passed through a 200-gauge stainless steel mesh. cell recruitment and activation and inflammatory cytokines pro- Cells were resuspended in 40% Percoll (Life Technologies BRL) and 3 duction. This vector could be delivered by viral and nonviral ap- centrifuged at 1260 g for 15 min at room temperature. Pelleted leukocytes were washed in PBS and used to evaluate cellularity and for flow proaches and exert function in vivo for a long time. cytometry (FCM) staining. Materials and Methods NK cell depletion Mice and i.v. injection Mice were injected i.v. with 50 mg anti-asialo-GM1 Ab (Wako, Tokyo, Japan) or 50 mg rabbit IgG control (Calbiochem, La Jolla, CA) per mouse C57BL/6 mice were purchased from the Experimental Animal Center 24 h before adenovirus infection. The elimination of NK cells was con- (Chinese Academy of Science, Shanghai, China). All mice were maintained firmed with FCM. under specific pathogen-free and controlled conditions (22˚C, 55% hu- midity, and 12-h day/night rhythm), and received care in compliance with Flow cytometric analysis the guidelines outlined in the Guide for the Care and Use of Laboratory

Animals. Mice were used between 6 and 8 wk of age. In high-dose The mAbs and isotypes used in this study included (all from BD Phar- Downloaded from adenovirus-induced liver acute injury model, mice were injected with mingen): FITC-labeled anti-CD11b, RatIgG2bl, anti-CD19, RatIgG2a k, 1 3 1011 viral particles (VP) Ad5-EGFP or mixed adenoviruses into the tail anti–TCR-gd, ArHIgG, anti-NK1.1, MsIgG2a k, and anti-CD3e; PE- vein in a 300-ml volume. Plasmid vectors were hydrodynamically injected labeled anti-CD69, ArHIgG l3, anti-Ly6G, and RatIgG2a k;PerCP- into the tail vein in a 2-ml volume, as described previously (17). Generally, Cy5.5–labeled anti-CD45.2, MsIgG2a k, anti-F4/80, Rat IgG2a k,and a total of 50 mg plasmid vector was injected into the tail veins of 6–8-wk- anti-NK1.1; and APC-labeled anti-CD3, ArHIgG1 k, anti-CD11c, ArHIgG old C57BL/6 mice in a volume of saline equivalent to 10% of the body l2, anti-Gr-1, and RatIgG2b k. Cells were stained with the indicated mass of the mouse (e.g., 2 ml for mouse of 20 g). The total volume was fluorescence-labeled mAbs according to standard protocol and analyzed delivered within 5–8 s. using a FACSCalibur (Becton Dickinson, Franklin Lakes, NJ). Data were http://www.jimmunol.org/ analyzed with FlowJo software version 7.6. Vector construction and adenoviral vector preparation Analysis of liver injury pLIVE vector (Mirus Bio, Madison, WI) was modified to delete original restricted enzyme sites Spe I and Xba I and introduce new sites into the Liver injury was assessed at the indicated time points after adenovirus intron 2 region of the vector. Template 97mer oligonucleotides containing infection by measuring serum alanine transaminase (ALT), aspartate ami- microRNA (miRNA)-30–based shRNA structure (Supplemental Table I) notransferase (AST), alkaline phosphatase (ALP), g-glutamyltranspeptidase was ligated into the shuttle vector MSCV-LMP (Open Biosystems, Hunts- (GGT), creatinine (CRE) and bilirubin (BILI). The normal serum levels ville, AL) and then constructed into pLIVE vector. More cassettes could of each parameter are as follows: ALT, ,40 IU/L; AST, ,40 IU/L; ALP, be assembled in a multiple-target manner, as Xba I and Spe I were iso- 53–128 IU/L; GGT, 7–32 IU/L; CRE, 53–108 mmol/L; and BILI, 5.1–19 caudomers. The IRES element and DsRed2 were amplified from pIRES2- mmol/L). All these liver injury parameters were measured using a test kit by guest on September 30, 2021 DsRed2 (Clontech, Mountain View, CA), whereas huIL-10 was obtained (Rong Sheng, Shanghai, China) following the manufacturer’s instruc- from pMD18T–huIL-10 (Sino Biological, Beijing, China). All constructs tions. were confirmed by DNA sequencing (Invitrogen, Shanghai, China). The recombinant replication-deficient E1, E3-deleted type 5 Adeno-X H&E staining expression system was purchased from Clontech. shNeg-Ad and Multi- Ad were constructed by homologous recombination. Recombinant ade- For histologic analysis, liver sections were fixed in 10% buffered formalin noviruses were packaged and propagated in HEK293 cells and purified and embedded in paraffin. Tissue sections were affixed to slides, depar- by CsCl discontinuous density gradient centrifugation. Typical vector titers affinized, stained with H&E, and examined using light microscopy. ranged between 0.4–1.9 3 1011 viral particles per milliliter as determined by Adeno-X qPCR Titration Kit (Clontech). The particle-to-PFU ratios, ELISA for cytokine detection determined in HEK293 cells using Adeno-X Rapid Titer Kit (Clontech) Mouse serum was harvested at indicated time points after hydrodynamic ranged from 8–20. Replication-deficient adenoviruses containing the EGFP injection of plasmids or adenovirus infection. huIL-10 was analyzed using gene (Ad5-EGFP) were purchased from 5+ MMI (5+MMI, Beijing, China). an huIL-10 sandwich ELISA kit (Cusabio, Wuhan, China). Mouse IFN-g For quantification of adenovirus burden in the liver, genomic DNA was and TNF-a were analyzed using mouse IFN-g and TNF-a sandwich ELISA isolated from liver tissue using EasyPure Genomic DNA Kit (TransGen, kits (Dawei, Shenzhen, China). Beijing, China) following the manufacturer’s instructions. The adenovirus DNA copies were determined using an Adeno-X qPCR Titration Kit. RNA preparation and real-time quantitative PCR Whole-body imaging of adenovirus infection Total RNA was isolated from Hepa1–6 cells or mouse hepatocytes using Mice infected with multifunction adenoviral vector or control vector were Trizol Reagent (Invitrogen) according to the manufacturer’s instructions. placed into an IVIS imaging chamber (Caliper Life Sciences) when fully The same quantity of total RNA was reverse-transcribed to cDNA using an anesthetized by isoflurane. DsRed2 fluorescence protein expression was M-MLV transcriptase (Invitrogen). Quantitative PCR was performed using imaged and calculated with Living Image software. SYBR Green I (Takara) for 45 cycles at 95˚C for 15 s and 60˚C for 60 s with ABI-Prism 7000 (Applied Biosystems) according to the manufacturer’s in- Cell cultures and DNA transfection structions. Primer sequences were as follows: b-actin: sense, 59-CACAGC- TTCTTTGCAGCTCCTT-39;antisense,59-ATGCCGGAGCCGTTGTC-39. Hepa1-6 (mouse hepatoma) cell lines were purchased from Shanghai Cell FKN: sense, 59-ACACCCCAACTCCAGTGAAC-39; antisense, 59-CGGA- Bank (Shanghai, China). Cells were grown in DMEM (Life Technologies, GAGCTCCAGAAAACAC-39. CCL5: sense, 59-ATATGGCTCGGACACC- Grand Island, NY) supplemented with 10% heat-inactivated FBS, 100 U/ml ACTC-39;antisense,59- GATGCCCATTTTCCCAGGAC-39. Results were penicillin, and 100 mg/ml streptomycin, and were incubated at 37˚C in 5% analyzed using the DDCt method as described previously (19). Values were CO2. Transient transfection was performed with Lipofectamine 2000 expressed as fold change compared with control. (Invitrogen, Carlsbad, CA) following the manufacturer’s instructions. Statistical analysis Frozen sections Results were analyzed by Student t test when appropriate. All data are Liver tissue from mice treated with plasmids and adenoviral vectors was expressed as mean 6 SEM. A p value ,0.05 was considered statistically embedded in Tissue-Tek OCT compound and snap frozen in liquid nitrogen. significant. The Journal of Immunology 3

Results high level of ALT was detected after system infection with high- Adenovirus induces acute liver injury by recruiting NK cells dose adenovirus, which was accompanied by an increase in liver Systemic delivery of high-dose, replication-deficient recombinant leukocytes and liver NK cell proportion (Fig. 1A–C). Depletion adenovirus causes acute liver injury, as demonstrated in mice and of NK cells before adenovirus administration could significantly in clinical trials (20, 21). Toxicity associated with adenovirus in- prevent acute liver injury, with a much lower serum ALT level and fection has been linked to the activation of both innate and fewer liver leukocytes (Fig. 1B, 1C), and produce less lymphocyte adaptive immune responses (22–24). In the present study a robust, infiltration and tissue necrosis (Fig 1D). NK depletion also led to Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 1. Acute liver injury induced by systemic infection of high dose adenovirus is mainly mediated by NK cells. (A)C57BL/6micewere injected i.v. with saline or a high dose of replication-deficient adenovirus (Ad5-EGFP, 1 3 1011 VP/mouse). Serum ALT levels were measured at indicated time points. Values are expressed as the mean 6 SEM (n =3).(B) C57BL/6 mice were administered 50 mg anti-asialo-GM1 or Rabbit IgG control i.v. 1 d before adenovirus infection. Mice were subsequently infected with 13 1011 VP Ad5-EGFP i.v. Serum ALT levels were measured 24 h later. Absolute leukocyte (CD45.2+) numbers per liver were counted (B) and liver NK cell (NK1.1+CD32) proportion in total mononuclear cells (MNCs) was analyzed with FCM (C). Values are expressed as the mean 6 SEM (n =4).*p , 0.05, ***p , 0.001 versus mice treated with rabbit IgG control. (D) Liver paraffin sections were prepared and stained with H&E to detect lymphocyte infiltration and hepatocyte necrosis (original magnification 3200; scale bar, 50 mm). (E) Mice were sacrificed, and frozen liver sections were made 1 d after infection. EGFP expression was examined by fluorescence microscopy (original magnification 3100; scale bar, 100 mm). A representative experiment is shown. (F) Total genomic DNA was isolated from the liver and analyzed for adenovirus DNA copies by quantitative real-time PCR. Values are expressed as the mean 6 SEM (n =4).*p , 0.05 versus mice treated with rabbit IgG control. 4 ATTENUATING LIVER INJURY THROUGH MANIPULATING MULTIPLE GENES higher levels of EGFP expression and larger amounts of adeno- shRNA structure, we constructed RNA interference (RNAi) vec- viral DNA in the liver (Fig. 1E, 1F), which indicated that NK cells tors targeting CCL5 or FKN (each containing three different RNAi might have a critical role in innate immune elimination of ade- sequences to enhance the interference efficiency) and test their noviral vectors and acute hepatitis. silence effects in vitro (Supplemental Fig. 1A). Consistent with liver injury, a robust, high level of ALT was detected upon high- Simultaneously silencing CCL5 and FKN enhances the dose adenovirus infection, and hydrodynamic injection of FKN protective effect of silencing FKN alone in adenovirus-induced silence vector pLIVE-(shFKN*3)mir decreased ALT level and acute liver injury absolute number of liver leukocytes. Synergistically, silence CCL5 Chemokines are important for NK cell recruitment into the liver; and FKN exerted a stronger protective effect, with lower ALT indeed, we previously found that knocking-down FKN prevented level, liver leukocytes numbers, and NK cell proportion (Fig. 2A– NK cell-mediated acute liver injury from adenovirus infection in C). High-dose adenovirus infection also caused severe pathologic mice. Expression of multiple chemokines such as FKN and CCL5 damage, whereas simultaneous knockdown of CCL5 and FKN was enhanced after adenovirus infection (9, 24) both in vitro and exhibited the least lymphocyte infiltration and tissue necrosis (Fig. in vivo, which was also demonstrated in this study (Supplemental 2D). Mice treated with both CCL5 and FKN RNAi vectors had a Fig. 1A, 1D). Therefore, we speculated whether the protective lower level of serum IFN-g than did mice with only FKN knock- effect would be more significant when more chemokines were si- down, although the reduction of serum TNF-a was not significant lenced. Using a hepatocyte-specific promoter and miRNA-based (Fig. 2E). Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 2. CCL5 silence synergistically enhances the attenuating effect of silencing FKN in adenovirus-induced acute liver injury. Total vectors (50 mg; 25 mg per vector when two vectors were mixed) was delivered to the liver of C57BL/6 mice by hydrodynamic tail-vein injection 3 d before adenovirus infection. Mice were injected i.v. with a high dose of replication-deficient adenovirus (Ad5-EGFP, 13 1011 VP/mouse). Liver injury was analyzed 1 d after infection. (A) Serum ALT levels of C57BL/6 mice, treated with the indicated vectors and infected by adenovirus for 24 h, were measured. Values are expressed as the mean 6 SEM (n = 5). ***p , 0.001 versus mice treated with pLIVE vector and adenovirus, *p , 0.05 versus mice treated with pLIVE- (shFKN*3)mir, pLIVE, and adenovirus. (B) Absolute leukocyte numbers per liver were counted (**p , 0.01 versus mice treated with pLIVE vector and adenovirus), and liver NK cell proportion in total mononuclear cells (MNCs) was analyzed by FCM (C). *p , 0.05 versus mice treated with pLIVE vector and adenovirus. (D) Liver paraffin sections were prepared and stained with H&E to detect lymphocytes infiltration and hepatocytes necrosis. Arrows indicate areas of hepatocyte necrosis (original magnification 3200; scale bar, 100 mm). (E) Serum IFN-g and TNF-a were measured with ELISA. Values are expressed as the mean 6 SEM (n = 5). **p , 0.01 versus mice treated with pLIVE vector and adenovirus. The Journal of Immunology 5

Simultaneous huIL-10 overexpression and FKN knockdown acute liver injury induced by high-dose adenovirus, the liver in- enhance the attenuating effect of knocking down FKN alone in flammation was greatly inhibited with the lowest ALT lever (Fig. adenovirus-induced acute liver injury 4A). As a result, we constructed a liver-directed multifunction IL-10 functions as a critical regulator of NK cell–related inflam- vector, named Multi-Vector, by simultaneously overexpressing mation (11, 25, 26). We tried to determine whether overexpression huIL-10 and knocking down CCL5 and FKN (Supplemental Fig. of huIL-10 would further enhance the attenuating effect of si- 1B). To test the ability of the Multi-Vector to control gene ex- lencing FKN in adenovirus-induced acute liver injury. The com- pression under acute hepatic injury conditions in vivo, it was bination of huIL-10 overexpression with FKN knockdown showed hydrodynamically injected into C57BL/6 mice 3 d before ad- a stronger protective effect than FKN knockdown alone, as evi- enovirus infection. The function of Multi-Vector was verified denced by a much lower ALT level and significantly reduced liver 1 d later; it could overexpress both huIL-10 and DsRed2 (as leukocytes and NK proportion (Fig. 3A–C). Lymphocyte infiltra- a report gene) while simultaneously silencing CCL5 and FKN tion and tissue necrosis were consistent with the serum ALT level expression in vivo (Supplemental Fig. 1C–E). Consistent with (Fig. 3D). huIL-10 was detected only in mice treated with pLIVE– liver injury, hydrodynamic injection of control shNeg-Vector huIL-10, which also had the lowest serum IFN-g and TNF-a in resulted in a robust high level of ALT, AST, GGT, and ALP these mice (Fig. 3E, 3F). upon high-dose adenovirus infection (Fig. 4B). Using the therapeutic vectors, huIL-10 overexpression, CCL5 knockdown, or Simultaneous manipulation of FKN, CCL5, and huIL-10 using FKN knockdown alone decreased the liver injury level to some a hepatocyte-directed vector highly efficiently prevents against extent, but the Multi-Vector treatment exhibited the most signifi- adenovirus-induced acute hepatitis via downregulating NK cant protective effect (Fig. 4B). Although adenovirus infection Downloaded from cells markedly increased the absolute leukocyte number in control mice As expected, when the three vectors silencing CCL5 or FKN or livers, the number only mildly increased in the therapeutic Multi- over-expressing huIL-10 were mixed together and used for treating Vector–treated mice. For NK cells in particular, Multi-Vector http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 3. HuIL-10 overexpression enhances the attenuating effect of silencing FKN in adenovirus-induced acute liver injury. (A) Serum ALT levels of C57BL/6 mice, treated with the indicated vectors and infected by adenovirus for 24 h, were measured. Values are expressed as the mean 6 SEM (n = 5). *p , 0.05 versus mice treated with pLIVE vector and adenovirus. (B) Absolute leukocyte numbers per liver were counted, and liver NK cell proportion in total mononuclear cells (MNCs) was analyzed by FCM (C). ***p , 0.001, *p , 0.05 versus mice treated with pLIVE vector and adenovirus. (D) Liver paraffin sections were prepared and stained with H&E to detect lymphocytes infiltration and hepatocytes necrosis. Arrows indicate areas of hepatocyte necrosis (original magnification 3 200; scale bar, 100 mm). (E) Serum huIL-10 and (F) IFN-g and TNF-a were measured with ELISA. Values are expressed as the mean 6 SEM (n = 5). The symbol # means serum huIL-10 was undetectable. **p , 0.01, *p , 0.05 versus mice treated with pLIVE vector and adenovirus. ***p , 0.001 versus mice treated with pLIVE-(shFKN*3)mir, pLIVE, and adenovirus. 6 ATTENUATING LIVER INJURY THROUGH MANIPULATING MULTIPLE GENES Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 4. Hepatocyte-specific hydrodynamic injection of multifunction vector protects against adenovirus-induced acute liver injury. (A) Total vectors (50 mgor75mg; 25 mg per vector when different vectors were mixed) were delivered to the liver of C57BL/6 mice by hydrodynamic tail-vein injection 3 d before adenovirus infection. Mice were injected i.v. with a high dose of replication-deficient adenovirus (Ad5-EGFP, 13 1011 VP/mouse). Serum ALT levels of C57BL/6 mice, treated with the indicated vectors and infected by adenovirus for 24 h, were measured. Values are expressed as the mean 6 SEM (n = 5). ***p , 0.001 versus mice treated with pLIVE vector and adenovirus. *p , 0.05 versus mice treated with pLIVE-(shFKN*3)mir, pLIVE-huIL-10, and adenovirus. (B) Serum ALT, AST, GGT, and ALP levels of C57BL/6 mice, treated with the indicated vectors and infected by adenovirus for 24 h, were measured. Values are expressed as the mean 6 SEM (n = 6). *p , 0.05 versus mice treated with huIL-10–Vector and adenovirus, ***p , 0.001 versus mice treated with shNeg-Vector and adenovirus. (C) Absolute leukocyte numbers per liver were counted, and liver NK cell proportion in total mononuclear cells (MNCs) was analyzed with FCM. *p , 0.05 versus mice treated with huIL-10–Vector and adenovirus, ***p , 0.001, **p , 0.01 versus mice treated with shNeg-Vector and adenovirus. (D) Liver paraffin sections were prepared and stained with H&E (original magnification 3200; scale bar, 100 mm). Arrows indicate areas of heavily infiltrated lymphocytes. (E) Serum levels of IFN-g and TNF-a were measured with ELISA. Values are expressed as mean 6 SEM (n = 4). *p , 0.05, **p , 0.01, and ***p , 0.001 versus mice treated with shNeg-Vector and adenovirus.

treatment exhibited the strongest inhibition of NK recruitment Multifunction vector has a long-lasting effect on NK cell (Fig. 4C); no significant differences in recruitment were observed recruitment and protects against liver injury induced by among other lymphocytes, including CD3+ T cells, NKT cells, gd high-dose adenovirus infection T cells, neutrophils, and Kupffer cells (Supplemental Fig. 2). In- We tested whether the multifunction vector had a long-lasting effect terestingly, mice treated with shNeg-Vector had the lowest den- on NK cell recruitment after adenovirus infection. As shown in Fig. dritic cell (DC) populations, whereas mice treated with huIL-10– 5, NK cells were recruited to the liver shortly after adenovirus in- Vector and Multi-Vector had similar DC populations with saline- fection. Multi-Vector could inhibit NK cell recruitment to the liver, treated control mice (Supplemental Fig. 2). It has been suggested and the inhibition effect lasted for at least 1 wk. Liver functionality that DCs might have a protective role in acute liver injury induced by high-dose adenovirus infection. High-dose adenovirus infec- was assessed at longer time points following multifunction vector tion also caused severe immune cell infiltration and liver tissue administration and adenovirus infection. Different liver injury necrosis. Upon single-vector treatment, there were fewer necrotic parameters (e.g., serum ALT, AST, ALP, GGT, CRE, BILI) were areas, although Multi-Vector treatment exhibited the least damage measured at weekly intervals. Multi-Vector was able to suppress the with the lowest proinflammatory IFN-g and TNF-a cytokine adverse consequences measured at 1–35 d after infection (Fig. 6A, levels in the serum (Fig. 4D, 4E). Supplemental Fig. 3A). There were no significant adverse effects. The Journal of Immunology 7

that occurs upon adenovirus administration when given at the same time.

Discussion Normally, the administration of replication-deﬁcient recombinant adenovirus elicits NK cell–mediated innate immunity and causes acute inﬂammation, which is a major drawback when used for gene therapy. In this study, we successfully prevented and treated adenovirus-induced acute hepatitis using a liver-directed multifunction vector that simultaneously knocked down and overex- pressed multiple immune genes. By using this multifunctional vector, recruitment of NK cells to the liver was successfully inhibited by synergistically silencing CCL5 and FKN while NK cy- totoxicity was elaborately ameliorated by overexpressing inhibitory cytokine IL-10. Adenoviruses are among the most commonly used vectors for liver gene therapy because of their liver tropism, high capacity for

therapeutic genes, high operability, and productivity; however, ad- Downloaded from enovirus activate host innate immunity and result in acute lethal inﬂammation (4). Strategies to eliminate the side effect are therefore important. Besides vector modiﬁcation, such as deletion of viral coding sequences and PEGylation (covalent attaching polyethylene glycol polymers to adenoviral vector), some strategies have been

used to modify the local immune microenvironment. Restricting http://www.jimmunol.org/ transgene expression to hepatocytes decreases the risk of Ab for- mation against transgene products because of the immunotolerant environment of the liver (12). Although hepatotropic adenovirus preferentially infects hepatocytes, it also might infect other tissues like lung and lymph nodes. Therefore, hepatocyte-specific pro- FIGURE 5. Multifunction vector has a long-lasting effect on NK cell moters for cell-specific expression might overcome these obstacles. recruitment in adenovirus-induced acute liver injury. shNeg-Vector or In our previous work, FKN knockdown prevented mice from Multi-Vector (50 mg) was delivered to the liver of C57BL/6 mice by hydrodynamic tail-vein injection 3 d before adenovirus infection. Mice were acute liver injury induced by hydrodynamic adenovirus vector by guest on September 30, 2021 injected i.v. with a high dose of replication-deficient adenovirus (Ad5- injection (9). However, the protection mediated by FKN knock- EGFP, 13 1011 VP/mouse). Absolute leukocyte numbers per liver were down was still not strong enough, likely because of silencing only counted (A), and liver NK cell proportion in total mononuclear cells a single gene and using a non–hepatocyte-specific RNA polymer- (MNCs) was analyzed by FCM (B) at the indicated time after adenovirus ase III promoter. Indeed, multiple chemokines play critical roles in infection. Values are expressed as mean 6 SEM (n = 3). *p , 0.05 versus adenovirus-induced liver injury, including MCP-1, CCL5, and FKN mice treated with shNeg-Vector and adenovirus. (24). Compared with traditional approaches generating shRNA, miRNA-based shRNA expression from a Pol II promoter was more Moreover, in mice pretreated with shNeg-Vector or Multi-Vector safe and effective (27, 28). This miRNA-based shRNA expressed for 30 d and then infected with high-dose adenovirus, Multi-Vector from a RNA Pol II promoter-AFP enhancer/albumin promoter can still protected against high-dose adenovirus-induced acute hepatitis silence genes in a hepatocyte-specific manner, helping to diminish with a lower ALT levels and NK cell proportion in the liver (Fig. the off-target and other side effects of RNAi. In addition, many 6B, 6C), indicating that the protective effects mediated by the more miRNA-based shRNA cassettes can be incorporated into this Multi-Vector could last for at least 1 mo. vector before reaching its capacity. A single transcript containing as To explore whether the use of our multifunction vector can be many as six miRNA-based shRNAs was tested in our work (Sup- delivered by virus but not only by hydrodynamic injection, we plemental Fig. 1B). packaged the vector into an adenoviral vector by homologous re- Inflammation caused by systemic infection of high-dose adeno- combination (named shNeg-Ad or Multi-Ad) and infected C57BL/6 virus comes from the complex immunopathologic factors, such as mice with a mixture of vector-free adenovirus (Ad5-EGFP) and a the enhanced expression of multiple chemokines, recruitment of high dose of 1 31011 VP/mouse of adenovirus at different ratios. innate immune cells, and secretion of several inflammatory cyto- Adenovirus infection and function was monitored in real time with kines. Therefore, we can target multiple function genes to modulate DsRed2 expression by whole body imaging (Fig. 7A). There is the liver inflammation. In our study, hepatic CCL5 and FKN ex- no significant difference of adenoviral burden between different pression greatly increased 24 h after adenovirus infection (Sup- groups (Fig. 7B). Liver inflammation was examined 24 h later. As plemental Fig. 1D). This could explain why simultaneous CCL5 and shown in Fig. 7C–E, Multi-Ad significantly attenuated adenovirus- FKN inhibition more significantly protected against liver injury than induced liver injury, even at the low dose of vector-containing silencing either single gene alone (Figs. 2 and 4). In addition, be- adenovirus, by decreasing leukocyte and NK cell recruitment into cause IL-10 exerts strong inhibitory effects on NK cells, the huIL- the liver as well as serum IFN-g and TNF-a (Fig. 7D, 7E). Sim- 10 overexpression vector more strongly attenuated adenovirus- ilarly, Multi-Ad was able to suppress the adverse consequences induced liver injury and synergized with chemokine inhibition measured at 1–28 d after infection (Supplemental Fig. 3B). These (Figs. 3 and 4). Moreover, our liver-directed multifunction vector data suggested that a multifunction vector can be delivered by viral can be packaged into adenovirus and other viral vector systems, vectors and that this multifunction vector can control inflammation and the DsRed2 reporter gene can help us to monitor the infection, 8 ATTENUATING LIVER INJURY THROUGH MANIPULATING MULTIPLE GENES

FIGURE 6. The protective effect of multifunction vector against adenovirus-induced acute liver injury lasts for at least 1 mo. (A) A different vector (50 mg) was delivered to the liver of C57BL/6 mice by hydrodynamic tail-vein injection 3 d before adenovirus infection. Mice were injected i.v. with a high dose of replication- deﬁcient adenovirus (Ad5-EGFP, 13 1011 VP/mouse). Serum ALT levels were measured at weekly intervals 1–35 d after infection. Values are expressed as the mean 6 SEM (n = 4). Mice were pretreated with 50 mg shNeg-Vector or Multi-Vector by hydrodynamic injection. Thirty days later, a dose of 13 1011 VP/mouse adenovirus was used to infect mice by tail vein injection. (B)SerumALT Downloaded from level was examined 24 h after infection. Values are expressed as the mean 6 SEM (n =3).**p , 0.01 versus mice treated with shNeg- Vector and adenovirus. (C)Theabso- lute number of liver mononuclear cells

(MNCs) was counted, and the liver http://www.jimmunol.org/ NK cell proportion was determined by FCM. *p , 0.05 versus mice treated with shNeg-Vector and adenovirus. distribution, and function of the vectors in a real-time manner. To tiple immune genes, which could modulate the liver immune mi- our knowledge, this report is the ﬁrst description of a hepatocyte- croenvironment and inhibit acute liver injury induced by high-dose speciﬁc multifunction vector for simultaneously manipulating mul- adenovirus. by guest on September 30, 2021

FIGURE 7. Multifunction vector delivered by adenoviral vector protects against adenovirus-induced acute liver injury. The adenoviral vector was mixed with vector-free adenovirus in different proportions, and mice were infected by tail vein injection at a dose of 13 1011 VP/mouse. (A) DsRed2 expression was monitored in real time, and ﬂuorescence signal emission from mice was quantiﬁed and analyzed using the IVIS imaging system 24 h after infection. (B)Liver samples were harvested 24 h after infection, and total genomic DNA was isolated from the liver. Adenovirus DNA copies were analyzed by quantitative real- time PCR. Values are expressed as the mean 6 SEM (n = 3). (C) Serum ALT levels from C57BL/6 mice treated for 24 h with the adenoviral vector and vector- free adenovirus mixture were measured. Values are expressed as the mean 6 SEM (n = 3). ***p , 0.001 versus mice treated with the shNeg-Ad and vector-free adenovirus mixture. (D) Absolute leukocyte numbers per liver were counted, and the proportion of liver NK cells was analyzed with FCM. **p , 0.01, *p , 0.05 versus mice treated with the shNeg-Ad and vector-free adenovirus mixture. (E) Serum levels of IFN-g and TNF-a were measured by ELISA. Values are expressed as mean 6 SEM (n = 3). ***p , 0.001, **p , 0.01, *p , 0.05 versus mice treated with the shNeg-Ad and vector-free adenovirus mixture. The Journal of Immunology 9

Acknowledgments 13. Nathwani, A. C., J. T. Gray, C. Y. Ng, J. Zhou, Y. Spence, S. N. Waddington, E. G. Tuddenham, G. Kemball-Cook, J. McIntosh, M. Boon-Spijker, et al. 2006. We thank our colleagues for the constructive suggestion on the current Self-complementary adeno-associated virus vectors containing a novel liver- study. specific human factor IX expression cassette enable highly efficient transduction of murine and nonhuman primate liver. Blood 107: 2653–2661. 14. Kramer, M. G., M. Barajas, N. Razquin, P. Berraondo, M. Rodrigo, C. Wu, Disclosures C. Qian, P. Fortes, and J. Prieto. 2003. In vitro and in vivo comparative study of The authors have no financial conflicts of interest. chimeric liver-specific promoters. Mol. Ther. 7: 375–385. 15. Blum, H. E. 2007. Molecular therapy and prevention of liver diseases. Adv. Med. Sci. 52: 29–36. 16. Chen, Q., W. Lou, J. Shen, L. Ma, Z. Yang, L. Liu, J. Luo, and C. Qian. 2010. References Potent antitumor activity in experimental hepatocellular carcinoma by adenovirus- 1. Gao, B., W. I. Jeong, and Z. Tian. 2008. Liver: An organ with predominant mediated coexpression of TRAIL and shRNA against COX-2. Clin. Cancer Res. innate immunity. Hepatology 47: 729–736. 16: 3696–3705. 2. Suda, T., K. Kamimura, T. Kubota, Y. Tamura, M. Igarashi, H. Kawai, Y. Aoyagi, 17. Liu, F., Y. Song, and D. Liu. 1999. Hydrodynamics-based transfection in animals and D. Liu. 2009. Progress toward liver-based gene therapy. Hepatol. Res. 39: by systemic administration of plasmid DNA. Gene Ther. 6: 1258–1266. 325–340. 18. Sun, R., Z. Tian, S. Kulkarni, and B. Gao. 2004. IL-6 prevents T cell-mediated 3. Nguyen, T. H., and N. Ferry. 2004. Liver gene therapy: advances and hurdles. hepatitis via inhibition of NKT cells in CD4+ T cell- and STAT3-dependent Gene Ther. 11(Suppl 1): S76–S84. manners. J. Immunol. 172: 5648–5655. 4. Marshall, E. 1999. Gene therapy death prompts review of adenovirus vector. 19. Witting, S. R., M. Brown, R. Saxena, S. Nabinger, and N. Morral. 2008. Helper- Science 286: 2244–2245. dependent adenovirus-mediated short hairpin RNA expression in the liver acti- 5. Morral, N., W. O’Neal, K. Rice, M. Leland, J. Kaplan, P. A. Piedra, H. Zhou, vates the interferon response. J. Biol. Chem. 283: 2120–2128. R. J. Parks, R. Velji, E. Aguilar-Co´rdova, et al. 1999. Administration of helper- 20. Liu, Q., A. K. Zaiss, P. Colarusso, K. Patel, G. Haljan, T. J. Wickham, and dependent adenoviral vectors and sequential delivery of different vector serotype D. A. Muruve. 2003. The role of capsid-endothelial interactions in the innate for long-term liver-directed gene transfer in baboons. Proc. Natl. Acad. Sci. USA immune response to adenovirus vectors. Hum. Gene Ther. 14: 627–643. 96: 12816–12821. 21. Raper, S. E., N. Chirmule, F. S. Lee, N. A. Wivel, A. Bagg, G. P. Gao, Downloaded from 6. Wang, L., R. Hernańdez-Alcoceba, V. Shankar, M. Zabala, S. Kochanek, J. M. Wilson, and M. L. Batshaw. 2003. Fatal systemic inflammatory response B. Sangro, M. G. Kramer, J. Prieto, and C. Qian. 2004. Prolonged and inducible syndrome in a ornithine transcarbamylase deficient patient following adenoviral transgene expression in the liver using gutless adenovirus: a potential therapy for gene transfer. Mol. Genet. Metab. 80: 148–158. liver cancer. Gastroenterology 126: 278–289. 22. Worgall, S., G. Wolff, E. Falck-Pedersen, and R. G. Crystal. 1997. Innate im- 7. Mok, H., D. J. Palmer, P. Ng, and M. A. Barry. 2005. Evaluation of polyethylene mune mechanisms dominate elimination of adenoviral vectors following in vivo glycol modification of first-generation and helper-dependent adenoviral vectors administration. Hum. Gene Ther. 8: 37–44. to reduce innate immune responses. Mol. Ther. 11: 66–79. 23. Jooss, K., H. C. Ertl, and J. M. Wilson. 1998. Cytotoxic T-lymphocyte target

8. Zhu, J., X. Huang, and Y. Yang. 2010. NKG2D is required for NK cell acti- proteins and their major histocompatibility complex class I restriction in re- http://www.jimmunol.org/ vation and function in response to E1-deleted adenovirus. J. Immunol. 185: sponse to adenovirus vectors delivered to mouse liver. J. Virol. 72: 2945–2954. 7480–7486. 24. Muruve, D. A., M. J. Barnes, I. E. Stillman, and T. A. Libermann. 1999. Ade- 9. Chen, Q., H. Wei, R. Sun, J. Zhang, and Z. Tian. 2008. Therapeutic RNA si- noviral gene therapy leads to rapid induction of multiple chemokines and acute lencing of Cys-X3-Cys chemokine ligand 1 gene prevents mice from adenovirus neutrophil-dependent hepatic injury in vivo. Hum. Gene Ther. 10: 965–976. vector-induced acute liver injury. Hepatology 47: 648–658. 25. Ouyang, W., S. Rutz, N. K. Crellin, P. A. Valdez, and S. G. Hymowitz. 2011. 10. Zhu, J., X. Huang, and Y. Yang. 2008. A critical role for type I IFN-dependent Regulation and functions of the IL-10 family of cytokines in inflammation and NK cell activation in innate immune elimination of adenoviral vectors in vivo. disease. Annu. Rev. Immunol. 29: 71–109. Mol. Ther. 16: 1300–1307. 26. Gaddi, P. J., M. J. Crane, M. Kamanaka, R. A. Flavell, G. S. Yap, and 11. Oberholzer, C., A. Oberholzer, S. K. Tschoeke, R. M. Minter, F. R. Bahjat, T. P. Salazar-Mather. 2012. IL-10 mediated regulation of liver inflammation D. LaFace, B. Hutchins, and L. L. Moldawer. 2004. Influence of recombinant during acute murine cytomegalovirus infection. PLoS ONE 7: e42850. adenovirus on liver injury in endotoxicosis and its modulation by IL-10 ex- 27. Silva, J. M., M. Z. Li, K. Chang, W. Ge, M. C. Golding, R. J. Rickles, D. Siolas,

pression. J. Endotoxin Res. 10: 393–401. G. Hu, P. J. Paddison, M. R. Schlabach, et al. 2005. Second-generation shRNA by guest on September 30, 2021 12. Pastore, L., N. Morral, H. Zhou, R. Garcia, R. J. Parks, S. Kochanek, libraries covering the mouse and human genomes. Nat. Genet. 37: 1281–1288. F. L. Graham, B. Lee, and A. L. Beaudet. 1999. Use of a liver-speciﬁc promoter 28. Giering, J. C., D. Grimm, T. A. Storm, and M. A. Kay. 2008. Expression of reduces immune response to the transgene in adenoviral vectors. Hum. Gene shRNA from a tissue-speciﬁc pol II promoter is an effective and safe RNAi Ther. 10: 1773–1781. therapeutic. Mol. Ther. 16: 1630–1636.