Portal de Periódicos da CAPES

Small Interfering RNA Targeting Heme Oxygenase-1 Enhances Ischemia-Reperfusion-induced Lung Apoptosis

2004; Elsevier BV; Volume: 279; Issue: 11 Linguagem: Inglês

10.1074/jbc.m312941200

1083-351X

Xuchen Zhang, Peiying Shan, Dianhua Jiang, Paul W. Noble, Nader G. Abraham, Attallah Kappas, Patty J. Lee,

Mechanical Circulatory Support Devices

Heme oxygenase-1 (HO-1) is emerging as an important cytoprotective enzyme system in a variety of injury models. To optimize future therapeutic applications of HO-1, it is necessary to delineate the precise functions and mechanisms as well as modes of externally regulating HO-1 expression. Investigations have been limited by difficulties with the generation of HO-1 null mice and the lack of specific HO-1 inhibitors. Lung ischemia-reperfusion (I-R) injury is the inciting event in acute lung failure following transplantation, surgery, and shock. To study the function of HO-1 in I-R-induced lung injury, we designed small interfering RNA (siRNA) sequences that effectively suppress HO-1 expression both in vitro and in vivo in an organ-specific manner. In this study we show that there is enhanced apoptosis, via increased Fas expression and caspase 3 activity, in the presence of HO-1 siRNA in endothelial cells and mouse lung during I-R injury, whereas HO-1 overexpression attenuates apoptosis. To the best of our knowledge, we are the first to demonstrate that lung-specific siRNA delivery can be achieved by intranasal administration without the need for viral vectors or transfection agents in vivo, thereby obviating potential concerns for toxicity if siRNA technology is to have clinical application in the future. Heme oxygenase-1 (HO-1) is emerging as an important cytoprotective enzyme system in a variety of injury models. To optimize future therapeutic applications of HO-1, it is necessary to delineate the precise functions and mechanisms as well as modes of externally regulating HO-1 expression. Investigations have been limited by difficulties with the generation of HO-1 null mice and the lack of specific HO-1 inhibitors. Lung ischemia-reperfusion (I-R) injury is the inciting event in acute lung failure following transplantation, surgery, and shock. To study the function of HO-1 in I-R-induced lung injury, we designed small interfering RNA (siRNA) sequences that effectively suppress HO-1 expression both in vitro and in vivo in an organ-specific manner. In this study we show that there is enhanced apoptosis, via increased Fas expression and caspase 3 activity, in the presence of HO-1 siRNA in endothelial cells and mouse lung during I-R injury, whereas HO-1 overexpression attenuates apoptosis. To the best of our knowledge, we are the first to demonstrate that lung-specific siRNA delivery can be achieved by intranasal administration without the need for viral vectors or transfection agents in vivo, thereby obviating potential concerns for toxicity if siRNA technology is to have clinical application in the future. HO-1 1The abbreviations used are: HO, heme oxygenase; siRNA, small interfering RNA; PAEC, pulmonary artery endothelial cells; A-R, anoxia-reoxygenation; I-R, ischemia-reperfusion; FACS, fluorescence-activated cell sorter; TUNEL, terminal deoxynucleotidyltransferase dUTP nick end-labeling; Ad, adenovirus.1The abbreviations used are: HO, heme oxygenase; siRNA, small interfering RNA; PAEC, pulmonary artery endothelial cells; A-R, anoxia-reoxygenation; I-R, ischemia-reperfusion; FACS, fluorescence-activated cell sorter; TUNEL, terminal deoxynucleotidyltransferase dUTP nick end-labeling; Ad, adenovirus. is one of three isoforms of HO, the rate-limiting enzyme in the degradation of heme to biliverdin and eventually to bilirubin. HO-1 expression is induced in multiple cell types and organs in response to injury. This induction is postulated to have protective properties; however, the mechanisms remain elusive (1Otterbein L.E. Choi A.M. Am. J. Physiol. 2000; 279: L1029-L1037Crossref PubMed Google Scholar, 2Otterbein L.E. Kolls J.K. Mantell L.L. Cook J.L. Alam J. Choi A.M. J. Clin. Investig. 1999; 103: 1047-1054Crossref PubMed Scopus (468) Google Scholar, 3Abraham N.G. Lavrovsky Y. Schwartzman M.L. Stoltz R.A. Levere R.D. Gerritsen M.E. Shibahara S. Kappas A. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6798-6802Crossref PubMed Scopus (316) Google Scholar, 4Lee P.J. Otterbein L.E. Sethi J. Sasidhar M. Choi A.M. Marczin N. Kharitonov S. Yacoub M. Barnes P. Disease Markers in Exhaled Breath. Vol. 170. Marcel Dekker, Inc., New York2002: 117-127Google Scholar). HO-2 is primarily constitutive and has been found to be important in the central nervous system (5Dore S. Takahashi M. Ferris C.D. Hester L.D. Guastella D. Snyder S.H. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 2445-2450Crossref PubMed Scopus (612) Google Scholar). The function of HO-3 is yet unknown. Synthetic heme analogues such as protoporphyrins that competitively inhibit the activity of all HO isoforms are commonly used to study HO-1 function but are limited by the lack of specificity and can have the paradoxical effect of up-regulating HO-1 protein expression (6Jozkowicz A. Dulak J. Acta Biochim. Pol. 2003; 50: 69-79Crossref PubMed Scopus (49) Google Scholar, 7Sardana M.K. Kappas A. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 2464-2468Crossref PubMed Scopus (153) Google Scholar). The HO inhibitor tin mesoporphyrin is approved by the Federal Drug Administration for the treatment of hyperbilirubinemia but, again, lacks specificity (8Kappas A. Abraham N.G. Alam J. Nath K.A. Heme Oxygenase in Biology and Medicine. Kluwer Academic/Plenum Publishing Corp., New York2002: 3-17Crossref Google Scholar). The recent emergence of siRNA technology to silence mammalian genes allows for highly specific analysis of gene function and has potential clinical application. Among the limited number of reports on siRNA administration in vivo, all use systemic delivery, transfection chemicals, or viral vectors, and none have been organ-specific to the best of our knowledge (9Hasuwa H. Kaseda K. Einarsdottir T. Okabe M. FEBS Lett. 2002; 532: 227-230Crossref PubMed Scopus (224) Google Scholar, 10McCaffrey A.P. Meuse L. Pham T.T. Conklin D.S. Hannon G.J. Kay M.A. Nature. 2002; 418: 38-39Crossref PubMed Scopus (955) Google Scholar, 11Reich S.J. Fosnot J. Kuroki A. Tang W. Yang X. Maguire A.M. Bennett J. Tolentino M.J. Mol. Vis. 2003; 9: 210-216PubMed Google Scholar, 12Sorensen D.R. Leirdal M. Sioud M. J. Mol. Biol. 2003; 327: 761-766Crossref PubMed Scopus (427) Google Scholar, 13Zender L. Hutker S. Liedtke C. Tillmann H.L. Zender S. Mundt B. Waltemathe M. Gosling T. Flemming P. Malek N.P. Trautwein C. Manns M.P. Kuhnel F. Kubicka S. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 7797-7802Crossref PubMed Scopus (357) Google Scholar). We demonstrate that intranasal siRNA delivery, without a vector or transfection agent, has lung specificity and that HO-1 potently regulates lung apoptosis. Lung-specific siRNA will not only be a useful tool to study gene function but may have therapeutic applications for a wide range of lung diseases. Furthermore, HO siRNA may become the basis of modulating severe hyperbilirubinemia of newborns and the severe jaundice of Crigler-Najjar type I patients where there is excessive bilirubin formation and for whom specific therapy does not currently exist. PAEC Anoxia-Reoxygenation (A-R) and Mouse Lung I-R Models— Primary rat pulmonary artery endothelial cells (PAEC) were maintained in Dulbecco’s modified Eagle’s medium (Invitrogen) with 10% fetal bovine serum (Hyclone, Logan, UT) and 0.1% gentamycin (Invitrogen). Dr. Troy Stevens, University of Alabama, generously provided the PAEC. All data using PAEC were collected before passage 20. PAEC were exposed to anoxia (95% N2, 5% CO2) in a sealed modular chamber (Billup-Rothberg, Del Mar, CA) with continuous monitoring and automated adjustments to maintain <0.5% O2 (BioSpherix, Redfield, NY) for 24 h, as described previously (14Zhang X. Bedard E.L. Potter R. Zhong R. Alam J. Choi A.M. Lee P.J. Am. J. Physiol. 2002; 283: L815-L829Crossref PubMed Scopus (38) Google Scholar). For anoxia-reoxygenation, the media were changed after 24 h of anoxia, and cells were exposed to normoxia at 37 °C in a humidified atmosphere of 5% CO2 for 1 h. For mouse lung I-R, after an intraperitoneal injection of urethane (180 mg/kg of body weight), the mice were intubated via tracheostomy and ventilated with a Harvard ventilator (respiratory rate, 120/min; tidal volume, 0.5 cm). A left hilar clamp was placed for 30 min of unilateral ischemia to the left lung, and the clamp was then released for 2 h of reperfusion as described previously (14Zhang X. Bedard E.L. Potter R. Zhong R. Alam J. Choi A.M. Lee P.J. Am. J. Physiol. 2002; 283: L815-L829Crossref PubMed Scopus (38) Google Scholar). All of the mouse experiments used male C57Bl/6J mice at 6-8 weeks of age weighing 20-25 g as purchased from Jackson Laboratory (Bar Harbor, ME). The Animal Care and Use Committee at Yale University approved this protocol in accordance with their guidelines. siRNA Design and Preparation—The design of siRNAs was based on the characterization of siRNA by Elbashir et al. (15Elbashir S.M. Lendeckel W. Tuschl T. Genes Dev. 2001; 15: 188-200Crossref PubMed Scopus (2658) Google Scholar, 16Elbashir S.M. Harborth J. Weber K. Tuschl T. Methods. 2002; 26: 199-213Crossref PubMed Scopus (1022) Google Scholar). siRNAs were synthesized in 2′-deprotected, duplexed, desalted, and purified form by Dharmacon Research, Inc. (Lafayette, CO). The sense and antisense strands of rat and mouse HO-1 siRNA were: sequence 1, 5′-AAGGACAUGGCCUUCUGGUAUdTdT-3′ (sense) and 5′-AUACCAGAAGGCCAUGUCCUUdTdT-3′ (antisense); sequence 2, 5′-AAUGAACACUCUGGAGAUGACdTdT-3′ (sense) and 5′-GUCAUCUCCAGAGUGUUCAUUdTdT-3′ (antisense); sequence 3, 5′-AAGACCAGAGUCCCUCACAGAdTdT-3′ (sense) and 5′-UCUGUGAGGGACUCUGGUCUUdTdT-3′ (antisense); sequence 4, 5′-AAGCCACACAGCACUAUGUAAdTdT-3′ (sense) and 5′-UUACAUAGUGCUGUGUGGCUUdTdT-3′ (antisense); sequence 5, 5′-AAGCCGAGAAUGCUGAGUUCAdTdT-3′ and 5′-UGAACUCAGCAUUCUCGGCUUdTdT-3′ (antisense). The sense and antisense strands of human heme oxygenase-1 siRNA were: sequence 1, 5′-GGAGAUUGAGCGCAACAAGdTdT-3′ (sense) and 5′-CUUGUUGCGCUAAAUCUCCdTdT-3′ (antisense); sequence 2, 5′-UGAUAGAAGAGGCCAAGACdTdT-3′ (sense) and 5′-GUCUUGGCCUCUUCUAUCAdTdT-3′ (antisense); sequence 3, 5′-CUGCGUUCCUGCUCAACAUdTdT-3′ (sense) and 5′-AUGUUGAGCAGGAACGCAGdTdT-3′ (antisense). 5′-Biotin-labeled rodent HO-1 siRNA sequence 4 and nonspecific siRNA scrambled duplex (sense, 5′-GCGCGCUUUGUAGGAUUCGdTdT-3′; antisense, 5′-CGAAUCCUACAAAGCGCGCdTdT-3′) were also synthesized by Dharmacon Research, Inc. Nonspecific 5′-fluorescein-labeled siRNA (sense, 5′-UUCUCCGAACGUGUCACGUdTdT-3′; antisense, 5′-ACGUGACACGUUCGGAGAAdTdT-3′) was synthesized by Qiagen (Germantown, MD) and used to determine transfection efficiency. Transfection of siRNA Duplexes in Vitro and in Vivo—PAEC were seeded into 6- or 12-well plates 1 day prior to transfection using Dulbecco’s modified Eagle’s tissue culture medium supplemented with 10% fetal bovine serum without antibiotics. At the time of transfection with siRNA, the cells were 50-60% confluent. Oligofectamine reagent (Invitrogen) was used as the transfection agent, and cells were then incubated for 6 h. Next, 30% fetal bovine serum/Dulbecco’s modified Eagle’ medium was added to reach a final concentration of 10% fetal bovine serum in the wells. Cells were exposed to A-R 24 h after transfection. For in vivo studies, each mouse was anesthetized with methoxyflurane and then given intranasal HO-1 siRNA (2 mg/kg of body weight) or equivalent doses of nonspecific control siRNA duplex or recombinant adenovirus containing rat HO-1 cDNA (Ad-HO-1, a generous gift from Dr. Leo Otterbein, University of Pittsburgh) or the recombinant adenovirus containing the β-galactosidase gene (Ad-X-LacZ, purchased from BD Biosciences Clontech) in a volume of 50 μl. Generation of Stable PAEC Cell Line Overexpressing Human HO-1— The human HO-1-expressing replication-deficient retrovirus vector LSN-HHO-1 has been described previously (17Quan S. Yang L. Abraham N.G. Kappas A. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 12203-12208Crossref PubMed Scopus (51) Google Scholar, 18Yang L. Quan S. Abraham N.G. Am. J. Physiol. 1999; 277: L127-L133Crossref PubMed Google Scholar). Exponentially growing PA317 packaging cells were used for transfection and preparation of viral particles. Individual G418-resistant clones were selected, and initial viral titer assays were measured by infecting NIH-3T3 cells as described previously (18Yang L. Quan S. Abraham N.G. Am. J. Physiol. 1999; 277: L127-L133Crossref PubMed Google Scholar). PA317/LSN-HHO-1 and the empty viral control cells, PA317/LXSN, were grown until subconfluence was reached. The supernatants were harvested and used to infect PAEC. After selection with G418 a stably transfected cell line of PAEC overexpressing human HO-1, designated LSN/HO-1, and a retroviral vector control cell line, LXSN, were obtained. Western Blot Analysis—As described previously (14Zhang X. Bedard E.L. Potter R. Zhong R. Alam J. Choi A.M. Lee P.J. Am. J. Physiol. 2002; 283: L815-L829Crossref PubMed Scopus (38) Google Scholar) protein was extracted from cell or lung tissue lysates, electrotransferred, and immunoblotted with monoclonal HO-1 antibody (Stressgen Biotechnologies, Victoria, Canada). Detection was performed with a Phototope-horseradish peroxidase Western detection system (Cell Signaling Technology, Beverly, MA). Equivalent sample loading was confirmed by stripping membranes with blot restore membrane rejuvenation solution (Chemicon International, Inc., Temecula, CA) and reprobed with anti-β-tubulin antibody. Apoptosis Assays—We used a fluorescence-activated cell sorter (FACS) to detect annexin V-fluorescein isothiocyanate labeling (BD Biosciences) according to the manufacturer’s instruction. Briefly, after PAEC were washed with cold phosphate-buffered saline and resuspended with binding buffer (10 mm HEPES/NaOH (pH 7.4), 140 mm NaCl, 2.5 mm CaCl2), a solution containing 1 × 105 cells was transferred to a 5-ml tube, and 5 μl each of annexin V and propidium iodide were added. Binding buffer was then added to each tube and analyzed by FACS (BD Biosciences). The annexin V-fluorescein isothiocyanate signal was detected by FL1 (fluorescein isothiocyanate detector) at 518 nm, and the propidium iodide signal was detected by FL2 (phycoerythrin fluorescence detector) at 620 nm. Mouse lung sections were subjected to terminal deoxynucleotidyltransferase dUTP nick end-labeling (TUNEL) assay using the in situ cell death detection kit (Roche Diagnostics) as described previously (19Zhang X. Shan P. Otterbein L.E. Alam J. Flavell R.A. Davis R.J. Choi A.M. Lee P.J. J. Biol. Chem. 2003; 278: 1248-1258Abstract Full Text Full Text PDF PubMed Scopus (249) Google Scholar). Fas Expression and Caspase 3 Activity—For Fas expression PAEC were washed twice in cold phosphate-buffered saline, pelleted, suspended in phosphate-buffered saline containing Fas (1:100 dilution) or control rat IgG (1:100 dilution) antibodies (Santa Cruz Biotechnology, Santa Cruz, CA), and incubated on ice for 45 min. The cells were resuspended in phosphate-buffered saline containing anti-rat fluorescein isothiocyanate (1:50 dilution) antibody (Santa Cruz Biotechnology) and fixed in 1% paraformaldehyde prior to FACS analysis. Caspase 3 activity was measured with colorimetric assays using the CaspACE assay system (Promega, Madison, WI). PAEC lysates were centrifuged, and the supernatants were incubated with colorimetric substrate, Ac-Asp-Glu-Val-Asp-p-nitroanilide (Ac-DEVD-pNA). The release of p-nitroanilide from Ac-DEVD-pNA was measured at 405 nm using a spectrophotometer. Statistics—Data are expressed as mean ± S.E. and analyzed by Student’s t test. The significant difference was accepted at the p < 0.05 level. HO-1 siRNA Inhibits A-R-induced HO-1 Protein Expression in PAEC—We have shown previously (14Zhang X. Bedard E.L. Potter R. Zhong R. Alam J. Choi A.M. Lee P.J. Am. J. Physiol. 2002; 283: L815-L829Crossref PubMed Scopus (38) Google Scholar) that mouse lungs subjected to I-R injury or endothelial cells exposed to A-R induce HO-1 expression. However, the function of HO-1 during lung I-R injury is unknown. To delineate the role of HO-1 in lung I-R injury, we sought to knock down HO-1 induction in vitro and in vivo using HO-1 siRNA. We designed five siRNA sequences (1Otterbein L.E. Choi A.M. Am. J. Physiol. 2000; 279: L1029-L1037Crossref PubMed Google Scholar, 2Otterbein L.E. Kolls J.K. Mantell L.L. Cook J.L. Alam J. Choi A.M. J. Clin. Investig. 1999; 103: 1047-1054Crossref PubMed Scopus (468) Google Scholar, 3Abraham N.G. Lavrovsky Y. Schwartzman M.L. Stoltz R.A. Levere R.D. Gerritsen M.E. Shibahara S. Kappas A. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6798-6802Crossref PubMed Scopus (316) Google Scholar, 4Lee P.J. Otterbein L.E. Sethi J. Sasidhar M. Choi A.M. Marczin N. Kharitonov S. Yacoub M. Barnes P. Disease Markers in Exhaled Breath. Vol. 170. Marcel Dekker, Inc., New York2002: 117-127Google Scholar, 5Dore S. Takahashi M. Ferris C.D. Hester L.D. Guastella D. Snyder S.H. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 2445-2450Crossref PubMed Scopus (612) Google Scholar) directed against rodent HO-1, according to the methods of Elbashir et al. (20Elbashir S.M. Harborth J. Lendeckel W. Yalcin A. Weber K. Tuschl T. Nature. 2001; 411: 494-498Crossref PubMed Scopus (8015) Google Scholar), and transfected them into PAEC. We determined that sequence 4 was the most effective in inhibiting A-R-induced HO-1 protein expression (Fig. 1A). The silencing effect of sequence 4 on HO-1 protein induction during A-R was dose-dependent, whereas incremental doses of nonspecific siRNA had no effect on HO-1 expression (Fig. 1B). Using 5′-biotin-labeled HO-1 siRNA, we demonstrated the efficient entry of the siRNA into PAEC as denoted by the diffuse brown staining in the cytoplasm (Fig. 1C). We also quantitated transfection efficiency to be greater than 88%, using 5′-fluorescein-labeled siRNA detection by FACS (data not shown). All subsequent in vitro and in vivo studies utilized sequence 4 HO-1 siRNA. HO-1 siRNA Inhibits I-R-induced HO-1 Protein Expression in Mouse Lung—We tested the efficacy of the HO-1 siRNA in the mouse lung by intranasal administration. A time course of HO-1 siRNA delivery prior to subjecting the mouse to lung I-R injury shows that HO-1 siRNA has maximal effect in attenuating HO-1 protein induction if given 16 h prior to I-R injury (Fig. 2A, lane 5) and persists for at least 72 h (data not shown). Therefore, intranasal HO-1 siRNA was given for 16 h prior to I-R lung injury in the subsequent mouse experiments. Lung I-R injury causes HO-1 protein induction in other organs such as the kidney, liver, and heart, but intranasal siRNA administration is lung-specific (Fig. 2B, lane 3). As expected, intranasal administration of nonspecific siRNA, at the same dose, had no effect on I-R-induced HO-1 expression in the lung or other organs (Fig. 2B, lane 4). We confirmed lung delivery of the siRNA by intranasally administering a 5′-biotin-labeled HO-1 siRNA that confers a brown stain to cells that have incorporated the siRNA. We detected brown-stained cells diffusely in the airway and lung parenchyma as early as 4 h and persisting to 16 h (Fig. 2C). HO-1 Overexpression in PAEC and Mouse Lung—To contrast the effects of HO-1 siRNA with HO-1 overexpression, we used a human HO-1 gene in a replication-defective retroviral vector, which has been described previously (18Yang L. Quan S. Abraham N.G. Am. J. Physiol. 1999; 277: L127-L133Crossref PubMed Google Scholar), to stably transfect PAEC. We achieved significantly increased HO-1 protein expression in the HO-1 overexpressors (LSN/HO-1) compared with empty vector transfection (LXSN) in room air (Fig. 3A). We used intranasal administration of rat HO-1 adenovirus (Ad-HO-1) to achieve HO-1 overexpression in mouse lungs. We determined HO-1 protein levels at 48 h, 72 h, 5 days, and 1 week after administration of Ad-HO-1 and observed increased HO-1 protein after 48-72 h (Fig. 3B, upper panel, lanes 2 and 3). As expected, the empty vector (Ad-X-LacZ) had no effect on HO-1 induction in the mouse lung (Fig. 3B, lower panel, lane 3). To effectively silence the human HO-1 gene in the stably transfected PAEC, we designed three human HO-1 siRNA sequences (1Otterbein L.E. Choi A.M. Am. J. Physiol. 2000; 279: L1029-L1037Crossref PubMed Google Scholar, 2Otterbein L.E. Kolls J.K. Mantell L.L. Cook J.L. Alam J. Choi A.M. J. Clin. Investig. 1999; 103: 1047-1054Crossref PubMed Scopus (468) Google Scholar, 3Abraham N.G. Lavrovsky Y. Schwartzman M.L. Stoltz R.A. Levere R.D. Gerritsen M.E. Shibahara S. Kappas A. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6798-6802Crossref PubMed Scopus (316) Google Scholar). Transfection experiments revealed that all three siRNA sequences, especially at higher doses, inhibited HO-1 expression in HO-1 overexpressing PAEC if given for 24-72 h (Fig. 3C). At 72 h of siRNA transfection, lower doses (300 nm) of siRNA appeared to be less effective in suppressing HO-1 protein, which is likely because of siRNA degradation. Sequence 1 at 600 nm for 24 h was used in subsequent experiments with human HO-1-overexpressing PAEC (LSN/HO-1). HO-1 Modulates Apoptosis in PAEC and Mouse Lung during A-R and I-R Injury, Respectively—Apoptosis is a pivotal mechanism of I-R-induced organ injury (21Stammberger U. Gaspert A. Hillinger S. Vogt P. Odermatt B. Weder W. Schmid R.A. Ann. Thorac. Surg. 2000; 69: 1532-1536Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar, 22Cao G. Pei W. Ge H. Liang Q. Luo Y. Sharp F.R. Lu A. Ran R. Graham S.H. Chen J. J. Neurosci. 2002; 22: 5423-5431Crossref PubMed Google Scholar, 23Calvillo L. Latini R. Kajstura J. Leri A. Anversa P. Ghezzi P. Salio M. Cerami A. Brines M. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 4802-4806Crossref PubMed Scopus (543) Google Scholar). We demonstrated the biological effect of HO-1 on apoptosis in PAEC during A-R and in mouse lung during I-R. When we knocked down HO-1 expression in PAEC with siRNA directed against rodent HO-1, there was a dramatic increase in A-R-induced apoptosis in PAEC (quantitated by FACS analysis) to 33.8 ± 2.2% during anoxia alone and 41.3 ± 1.4% during A-R compared with wild type controls and cells transfected with nonspecific siRNA (Fig. 4A, lanes 1-3). Given that the absence of HO-1 induction during A-R was deleterious, we expected that HO-1 overexpression would attenuate apoptosis. HO-1-overexpressing cells (LSN/HO-1) subjected to A-R exhibited significantly less apoptosis (8.6 ± 0.2%) during A-R compared with wild type controls and cells that express empty vector (LXSN) (Fig. 4A, lanes 1, 4, and 5). This effect on apoptosis appeared to be specific to HO-1, because we were able to reverse the anti-apoptotic effect of HO-1 overexpression by introducing HO-1 siRNA directed against the human HO-1-overexpressing vector (Fig. 4A, lane 6). However, given that the HO-1 siRNA was directed against the exogenously transfected human HO-1 gene and considering our previous demonstration that there was endogenous induction of rat HO-1 in PAEC (a rat cell line) during A-R injury (14Zhang X. Bedard E.L. Potter R. Zhong R. Alam J. Choi A.M. Lee P.J. Am. J. Physiol. 2002; 283: L815-L829Crossref PubMed Scopus (38) Google Scholar), we proceeded to simultaneously inhibit both rodent and human HO-1. Notably, the presence of both rodent and human siRNA led to more apoptosis than human siRNA alone (41.4 ± 2.4 versus 26.8 ± 0.7%) (Fig. 4A, lanes 6 and 7, respectively). This was likely because of the fact that even in the presence of human siRNA, which only silenced the transfected human HO-1 gene expression, endogenous HO-1 induction in the PAEC overexpressors continued to have an anti-apoptotic effect. Only when both the exogenous human HO-1 and the endogenous HO-1 induction were silenced did apoptosis reach the levels seen in wild type PAEC treated with rodent HO-1 siRNA (Fig. 4A, lane 3). Nonspecific siRNA transfection at equivalent doses had no effect on A-R-induced apoptosis. These findings demonstrated the species specificity of the siRNA sequences and highlighted the anti-apoptotic effect of HO-1. We proceeded to correlate our in vitro findings with in vivo studies by subjecting mice to lung I-R injury and assessing TUNEL staining. HO-1 siRNA significantly increased the number of TUNEL-positive cells compared with naïve mice, mice subjected to I-R alone, and mice with nonspecific siRNA subjected to I-R (Fig. 4B, A-D). On the other hand, mice that overexpressed HO-1 in the lung (Ad-HO-1, Fig. 4B, F) exhibited significantly less TUNEL staining during lung I-R compared with mice subjected to I-R alone and mice given empty vector (Ad-X-LacZ) (Fig. 4B, B and E, respectively). The quantitation of TUNEL positive cells paralleled the PAEC results in that HO-1 siRNA in vivo caused significantly increased lung apoptosis, whereas HO-1 overexpression with intranasal administration of HO-1 adenoviral vector decreased apoptosis during lung I-R (Fig. 4C). Taken together, our data indicate that in both endothelial cells and mouse lung, HO-1 expression has profound anti-apoptotic properties during A-R and I-R injury, respectively. HO-1 Modulates Apoptosis via Fas and Caspase 3-dependent Mechanisms in PAEC during A-R Injury—We have demonstrated previously (24Zhang X. Shan P. Alam J. Davis R.J. Flavell R.A. Lee P.J. J. Biol. Chem. 2003; 278: 22061-22070Abstract Full Text Full Text PDF PubMed Scopus (154) Google Scholar) that Fas/FasL expression and caspase 3 activity are increased in PAEC during A-R. To delineate potential mechanisms of HO-1-modulated apoptosis, we investigated the effects of HO-1 siRNA and overexpression on Fas and caspase 3 expression in PAEC during A-R. Fig. 5A is a representative FACS analysis of Fas expression in PAEC during A-R, which demonstrated that HO-1 siRNA increased A-R-induced Fas expression, whereas stable HO-1 overexpression decreased Fas expression. An antibody to rat IgG was used as a negative control. Fig. 5B is a graphical representation of three independent FACS analyses of Fas expression. The y axis depicts the percentage of cells compared with the total that express Fas during the various conditions. In the presence of HO-1 siRNA (Fig. 5B, lane 3), there was significantly increased Fas expression (50.2 ± 2.3%) compared with the wild type control cells (31.6 ± 0.9%) and cells transfected with nonspecific siRNA (30.4 ± 1.6%) during A-R. However, in the presence of stable HO-1 overexpression using the human HO-1 gene (LSN/HO-1), there was a dramatic attenuation of A-R-induced Fas expression to 16.9 ± 1.3% compared with cells stably transfected with empty vector (LXSN) (Fig. 5B, lanes 4 and 5). We were able to reverse the effects of HO-1 overexpression on Fas expression by transfecting the HO-1 overexpressors with siRNA directed against the human HO-1 vector (Fig. 5B, lane 6). In parallel with our apoptosis studies in PAEC (Fig. 4A), the silencing of both the human HO-1 overexpression vector and the endogenous rodent HO-1 induction in PAEC resulted in even greater Fas expression in PAEC during A-R (Fig. 5B, lane 7). In addition to Fas, HO-1 also modulated caspase 3 activity during A-R. HO-1 overexpression significantly attenuated A-R-induced caspase 3 activity in PAEC (1.6 ± 0.1%) (Fig. 5C, lane 5). On the other hand, HO-1 siRNA, especially if directed against both the exogenous human HO-1 vector and endogenous rodent HO-1, dramatically increased caspase 3 activity in PAEC during A-R (4.6 ± 0.2%) (Fig. 5C, lane 7).Fig. 5HO-1 siRNA increased whereas HO-1 overexpression decreased Fas expression and caspase 3 activity in PAEC during A-R. PAEC were untransfected (control), transfected with nonspecific siRNA (nonspecific siRNA), transfected with HO-1 siRNA directed against rodent siRNA (siRNA), stably transfected with empty retroviral vector (LXSN), stably transfected with human HO-1-overexpressing retroviral vector (LSN/HO-1), HO-1 overexpressors transfected with HO-1 siRNA directed against human HO-1 (LSN/HO-1/Human siRNA), or HO-1 overexpressors transfected with HO-1 siRNAs directed against both human and rodent HO-1 (LSN/HO-1/Double siRNA), and then the cells were exposed to room air (RA), 24-h anoxia (24A) or 24-h anoxia followed by 1-h reoxygenation (A-R). A, PAEC were stained with anti-Fas or anti-rat IgG (Negative Control) antibody during A-R, and Fas expression was detected by FACS analysis. The data are representative of three independent experiments. B, graphical quantitation of the percentage of total cells that express Fas during the various conditions. The data represent the mean of three independent experiments ± S.E. *, p < 0.05, compared with corresponding siRNA 24-h anoxia and A-R; #, p < 0.05, compared with corresponding LSN/HO-1/Human siRNA 24-h anoxia and A-R. C, caspase 3 activity in PAEC was detected during A-R and is represented graphically. Data represent the mean of three independent experiments ± S.E. *, p < 0.05, compared with corresponding siRNA 24-h anoxia and A-R; #, p < 0.05, compared with corresponding LSN/HO-1/Human siRNA 24-h anoxia and A-R.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Our studies are the first to utilize the highly specific technology of siRNA to directly demonstrate that HO-1 has a dramatic effect on apoptosis during lung I-R injury via Fas and caspase 3-dependent pathways. HO-1 overexpression by gene transfer successfully attenuated I-R-induced apoptosis and may potentially have therapeutic value in I-R-induced lung injury for which specific therapies do not currently exist. Lung I-R injury is critical to the pathogenesis of acute lung injury during lung transplantation/surgery, pulmonary embolism, and re-expansion pulmonary edema. The inhibition of apoptosis during I-R injury in other organ systems has been shown to be cytoprotective and to promote organ survival (25Chanalaris A. Sun Y. Latchman D.S. Stephanou A. J. Mol. Cell. Cardiol. 2003; 35: 257-264Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar, 26Yaoita H. Ogawa K. Maehara K. Maruyama Y. Circulation. 1998; 97: 276-281Crossref PubMed Scopus (564) Google Scholar). Therefore, identifying genes that modulate I-R-induced apoptosis would have potential therapeutic relevance to a variety of organ systems. In addition, we demonstrate that lung-specific siRNA delivery, without the presence of viral vectors or transfection agents, is biologically effective and may have significant implications for future therapeutic interventions. Our data show that the transfer of exogenous human HO-1 cDNA to endothelial cells or adenovirally-mediated HO-1 overexpression in mouse lung dramatically attenuates A-R- and I-R-induced apoptosis. Conversely, specific knockdown of HO-1 expression using siRNA in vitro and in vivo significantly increased A-R- and I-R-induced apoptosis, respectively. Furthermore, we use HO-1 overexpression vectors and HO-1 siRNA to demonstrate that HO-1 specifically modulates endothelial cell apoptosis by attenuating A-R-induced Fas expression and caspase 3 activity. Thus far, most investigations into the consequences of HO-1 deficiency have utilized synthetic heme analogues such as the protoporphyrins, which are not specific for HO-1 and have the paradoxical effect of inducing HO-1 expression (6Jozkowicz A. Dulak J. Acta Biochim. Pol. 2003; 50: 69-79Crossref PubMed Scopus (49) Google Scholar, 7Sardana M.K. Kappas A. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 2464-2468Crossref PubMed Scopus (153) Google Scholar). Reports using HO-1-deficient mice have been limited, likely because of breeding difficulties, and dominant-negative constructs have not been effective. The use of HO-1 siRNA will greatly facilitate the precise identification of the HO-1 gene function. We show that HO-1 siRNA has biologic functions both in vitro as well as in vivo and can be effective in an organ-specific manner. RNA interference mediated by siRNA is a powerful technology that allows the silencing of genes with great specificity and potency. Although it is used extensively in plants, nematodes, and Drosophila to determine gene function, until recently RNA interference was not applicable in mammalian cells. The discovery that transfection of short 21-23-nucleotide siRNA into mammalian cells specifically interferes with gene expression and does not induce nonspecific responses opened this technique to the study of mammalian gene function and may have potential therapeutic applications (20Elbashir S.M. Harborth J. Lendeckel W. Yalcin A. Weber K. Tuschl T. Nature. 2001; 411: 494-498Crossref PubMed Scopus (8015) Google Scholar). There are increasing reports on the use of siRNA in vitro (27Shen C. Buck A.K. Liu X. Winkler M. Reske S.N. FEBS Lett. 2003; 539: 111-114Crossref PubMed Scopus (195) Google Scholar, 28Barton G.M. Medzhitov R. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 14943-14945Crossref PubMed Scopus (260) Google Scholar, 29Devroe E. Silver P.A. BMC Biotechnol. 2002; 2: 15-19Crossref PubMed Scopus (192) Google Scholar, 30Abbas-Terki T. Blanco-Bose W. Deglon N. Pralong W. Aebischer P. Hum. Gene Ther. 2002; 13: 2197-2201Crossref PubMed Scopus (281) Google Scholar). However the in vivo use of siRNA has been limited. Researchers have found that systemic delivery of caspase 8 siRNA or Fas siRNA into mice protects against liver injury (13Zender L. Hutker S. Liedtke C. Tillmann H.L. Zender S. Mundt B. Waltemathe M. Gosling T. Flemming P. Malek N.P. Trautwein C. Manns M.P. Kuhnel F. Kubicka S. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 7797-7802Crossref PubMed Scopus (357) Google Scholar, 31Song E. Lee S.K. Wang J. Ince N. Ouyang N. Min J. Chen J. Shankar P. Lieberman J. Nat. Med. 2003; 9: 347-351Crossref PubMed Scopus (1014) Google Scholar). Similarly, McCaffrey et al. (32McCaffrey A.P. Nakai H. Pandey K. Huang Z. Salazar F.H. Xu H. Wieland S.F. Marion P.L. Kay M.A. Nat. Biotechnol. 2003; 21: 639-644Crossref PubMed Scopus (583) Google Scholar) showed that the administration of hepatitis B virus siRNA effectively inhibited hepatitis B virus replication in cultured cells and mouse liver, indicating that siRNA could be useful in the treatment of viral liver diseases. Subretinal siRNA can also be achieved using a chemical transfection agent (11Reich S.J. Fosnot J. Kuroki A. Tang W. Yang X. Maguire A.M. Bennett J. Tolentino M.J. Mol. Vis. 2003; 9: 210-216PubMed Google Scholar), and transgenic approaches as well as viral constructs have been used to deliver siRNA in vivo (9Hasuwa H. Kaseda K. Einarsdottir T. Okabe M. FEBS Lett. 2002; 532: 227-230Crossref PubMed Scopus (224) Google Scholar, 33Rubinson D.A. Dillon C.P. Kwiatkowski A.V. Sievers C. Yang L. Kopinja J. Rooney D.L. Ihrig M.M. McManus M.T. Gertler F.B. Scott M.L. Van Parijs L. Nat. Genet. 2003; 33: 401-406Crossref PubMed Scopus (1336) Google Scholar). The reports thus far of siRNA in vivo utilize systemic delivery of the siRNA and/or require the use of a transfection chemical or viral vector, which potentially raises concerns for toxicity if used clinically. Furthermore, although systemically delivered siRNA can be detected in multiple organs, including the lung, it is clear that the biologic activity of systemically administered siRNA is not equally effective in all organs. For example, systemic injection of caspase 8 siRNA, although detected in most organs, inhibited Fas-induced liver apoptosis but not lung or kidney apoptosis (13Zender L. Hutker S. Liedtke C. Tillmann H.L. Zender S. Mundt B. Waltemathe M. Gosling T. Flemming P. Malek N.P. Trautwein C. Manns M.P. Kuhnel F. Kubicka S. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 7797-7802Crossref PubMed Scopus (357) Google Scholar). In our present study, intranasal administration of HO-1 siRNA, in the absence of transfection agents, had significant lung-specific effects in modulating apoptosis. We observed the presence of HO-1 siRNA diffusely in the lung airway and alveoli. In the near future the use of tissue-specific siRNA may become feasible (with polymerase II rather than polymerase III promoters), opening up the possibility of targeting specific lung cell types. The ability to apply siRNA in an organ or cell-specific manner will be of paramount interest not only for lung diseases but also for a broad range of clinical processes.