Cytoglobin Overexpression Protects against Damage-Induced Fibrosis Ruian Xu, Phillip M. Harrison, Miao Chen, Linyan Li, Tung-yu Tsui, Peter C.W. Fung, Pik-to Cheung, Guangji Wang, Hua Li, Yong Diao, Geoffrey W. Krissansen, Sue Xu, Farzin Farzaneh  Molecular Therapy  Volume 13, Issue 6, Pages 1093-1100 (June 2006) DOI: 10.1016/j.ymthe.2005.11.027 Copyright © 2006 The American Society of Gene Therapy Terms and Conditions

FIG. 1 In vitro transduction of primary rat HSC with rAAV. (A) Primary rat HSC, 24 h postisolation, were stained with an anti-desmin antibody to determine purity. (B) The 570-bp cDNA encoding the open reading frame of rat cytoglobin was cloned into the EcoRI and NotI sites of the replication-deficient rAAV-2 vector. The vector also contained the AAV-2 ITRs, a CAG promoter, and the woodchuck hepatitis virus posttranslational regulatory element to facilitate expression. The in vitro infection of isolated rat (C) HSC and (D) hepatocytes with the rAAV/eGFP (at an m.o.i. of 5 × 104 (C0 and 2.5 × 105 particles/ml (D)) demonstrates the efficient transduction of HSC, but not hepatocytes, at 60 h. Molecular Therapy 2006 13, 1093-1100DOI: (10.1016/j.ymthe.2005.11.027) Copyright © 2006 The American Society of Gene Therapy Terms and Conditions

FIG. 2 Cytoglobin protects primary rat hepatic stellate cells against oxidative stress-induced activation. (A) HSC, 60 h posttransduction with rAAV (m.o.i. of 5 × 104 particles/cell) encoding Cygb (left) or eGFP (right), were stained with a Cygb-specific antibody. (B) Oxidative stress was induced in HSC by exposure to 50 μM ferric nitrilotriacetate and 20 μM arachidonic acid (Fe/AA), 48 h after transduction with rAAV encoding Cygb or eGFP. Levels of lipid peroxidation were assessed by production of MDA and 4-HNE. (C) The total oxyradical scavenging capacity of HSC was measured following exposure to H2O2, after transduction with rAAV encoding Cygb or eGFP. (D) HSC were treated as in (B) and expression of TGF-β1 and TIMP-1 mRNA transcripts was determined by RT-PCR. (E) HSC were treated as in (B) and electrophoretic gel mobility shift analysis was used to measure DNA binding activities for AP-1 (left) and NF-κB (right). (F) Collagen production by HSC was determined by incorporation of [3H]proline (10 μCi/ml) into collagenase-sensitive protein [26]. Molecular Therapy 2006 13, 1093-1100DOI: (10.1016/j.ymthe.2005.11.027) Copyright © 2006 The American Society of Gene Therapy Terms and Conditions

FIG. 3 Expression of cytoglobin in HSC transduced with rAAV/Cygb in vivo. Four weeks after 5 × 1011particles of rAAV/Cygb were administered directly into the portal vein of male Sprague–Dawley rats, DIG nonradioactive in situ hybridization with a Cygb-specific cRNA probe was used to detect Cygb transcripts in liver sections from (A) normal and (B) rAAV/Cygb-treated rats. Sections were developed using alkaline phosphatase NBT–BCIP detection. Arrows point to the Cygb transcript-positive cells (B). Liver sections were also stained with anti-Cygb (green) and anti-desmin (red) antibodies. Representative liver sections from (C) normal and (D) rAAV/Cygb rats are shown. Double-stained cells are indicated by arrows. (E) Animals were sacrificed 1, 7, and 14 days after pv injection of 5 × 1011 particles of rAAV/Cygb. The desmin-positive HSC were separated from parenchymal liver cells by standard methodology. Intracellular staining for Cygb was performed using rabbit anti-Cygb antibody visualized by goat anti-rabbit antibody labeled with PE and levels were detected and analyzed using FACSCalibur. Less than 10% of the desmin-negative cell population expressed Cygb at any time point. Molecular Therapy 2006 13, 1093-1100DOI: (10.1016/j.ymthe.2005.11.027) Copyright © 2006 The American Society of Gene Therapy Terms and Conditions

FIG. 4 Prophylactic overexpression of Cygb inhibits injury-induced activation of HSC in vivo. (A) 5 × 1011 particles of rAAV encoding either Cygb or eGFP, or an equivalent volume of PBS carrier, were delivered to male SD (n = 10 for each group) rats via pv injection 2 weeks before induction of hepatic necrosis by twice weekly intraperitoneal injection of CCl4 0.5 ml/kg, mixed in an equal volume of olive oil. Following 8 weeks of CCl4 administration, liver sections were stained with Masson's trichrome and hematoxylin–eosin. Representative liver sections are shown from normal controls (top, left) or CCL4-treated rats receiving PBS (top, right), rAAV/eGFP (bottom, left), or rAAV/Cygb (bottom, right). (B) As in (A) or (D) but RT-PCR was used to detect PC-1, TGF-β1, and TIMP-1 transcripts in mRNA isolated from whole liver. (C) As in (A) but Western blot analysis was used to detect α-SMA and TGF-β1 protein levels in whole liver tissue lysates. (D) rAAV vectors were administered as in (A), 3 days before induction of cholestatic liver injury by BDL. Twenty-eight days after BDL, liver sections were stained with Masson's trichrome and hematoxylin–eosin. Representative liver sections are shown from normal controls (top, left) or BDL rats receiving PBS (top, right), rAAV/eGFP (bottom, left), or rAAV/Cygb (bottom, right). The number of rats in each group was at least 7. Molecular Therapy 2006 13, 1093-1100DOI: (10.1016/j.ymthe.2005.11.027) Copyright © 2006 The American Society of Gene Therapy Terms and Conditions

FIG. 5 Overexpression of Cygb after the onset of liver injury reduces liver fibrosis and induces a quiescent HSC phenotype. Rats were exposed to CCl4 for 8 weeks or BDL for 12 days and then injected via pv with 5 × 1011 particles of rAAV encoding either Cygb or eGFP or an equivalent volume of PBS carrier. Animals were sacrificed 4 weeks (CCl4) or 12 days (BDL) after transduction with the indicated AAV vectors. Representative liver sections, stained with Masson's trichrome and hematoxylin–eosin, are shown for (A) CCl4-treated and (B) BDL rats: controls (left), rAAV/eGFP (middle), or rAAV/Cygb (right). (C) Real-time RT-PCR was used to measure levels of TGFβ-1 and PC-1 mRNA in HSC from BDL rats, isolated at the time of sacrifice (1, sham operated; 2, BDL-rAAV/eGFP; 3, BDL-rAAV/Cygb; 4, no-template control). For each group n = 5 for the BDL- and n = 10 for the CCl4-treated rats. Molecular Therapy 2006 13, 1093-1100DOI: (10.1016/j.ymthe.2005.11.027) Copyright © 2006 The American Society of Gene Therapy Terms and Conditions