Effective treatment of vascular endothelial growth factor ... - Nature

Gene Therapy (2006) 13, 1342–1350 & 2006 Nature Publishing Group All rights reserved 0969-7128/06 $30.00 www.nature.com/gt

ORIGINAL ARTICLE

Effective treatment of vascular endothelial growth factor refractory hindlimb ischemia by a mutant endothelial nitric oxide synthase gene HS Qian1, P Liu1, L-Y Huw1, A Orme2, M Halks-Miller1, SM Hill1, F Jin2, P Kretschmer2, E Blasko3, L Cashion2, P Szymanski2, R Vergona1, R Harkins2, J Yu4, WC Sessa4, WP Dole5, GM Rubanyi2 and K Kauser2 1 Department of Pharmacology, Berlex Biosciences, Richmond, CA, USA; 2Department of Gene Therapy, Berlex Biosciences, Richmond, CA, USA; 3Department of Immunology, Berlex Biosciences, Richmond, CA, USA; 4Yale University, New Haven, CT, USA and 5 Cardiovascular Research, Berlex Biosciences, Richmond, CA, USA

Gene delivery of angiogenic growth factors is a promising approach for the treatment of ischemic cardiovascular diseases. However, success of this new therapeutic principle is hindered by the lack of critical understanding as to how disease pathology affects the efficiency of gene delivery and/or the downstream signaling pathways of angiogenesis. Critical limb ischemia occurs in patients with advanced atherosclerosis often exhibiting deficiency in endothelial nitric oxide production. Similar to these patients, segmental femoral artery resection progresses into severe ischemic necrosis in mice deficient in endothelial nitric oxide synthase (ecNOS-KO) as well as in balb/c mice. We used these models to evaluate the influence of severe ischemia on

transfection efficiency and duration of transgene expression in the skeletal muscle following plasmid injection in combination with electroporation. Subsequently, we also explored the potential therapeutic effect of the phosphomimetic mutant of ecNOS gene (NOS1177D) using optimized delivery parameters, and found significant benefit both in ecNOS-KO and balb/c mice. Our results indicate that NOS1177D gene delivery to the ischemic skeletal muscle can be efficient to reverse critical limb ischemia in pathological settings, which are refractory to treatments with a single growth factor, such as vascular endothelial growth factor. Gene Therapy (2006) 13, 1342–1350. doi:10.1038/ sj.gt.3302781; published online 27 April 2006

Keywords: angiogenesis/revascularization; critical limb ischemia; skeletal muscle gene delivery; gene expression imaging; endothelial nitric oxide synthase

Introduction Peripheral arterial occlusive disease is a highly prevalent disease lacking adequate treatment especially for its most advanced stage, critical limb ischemia (CLI).1 Local delivery of genes encoding angiogenic factors such as vascular endothelial growth factor (VEGF), fibroblast growth factor or hepatocyte growth factor emerged as a promising new therapeutic approach for ischemic cardiovascular disease.2 However, clinical results so far are lacking unequivocal proof for efficacy by therapeutic angiogenesis. It may be due to inefficient gene delivery or because of some patients are refractory to exogenously administered growth factors.2 Gene expression following naked plasmid DNA delivery is generally very low compared to viral vectors,

Correspondence: Dr K Kauser, Boehringer Ingelheim Pharmaceuticals, Inc., 175 Briar Ridge Road, Ridgefield, CT 06877, USA. E-mail: [email protected] Received 29 April 2005; revised 31 January 2006; accepted 11 March 2006; published online 27 April 2006

therefore its suitability for clinical therapeutic application is questionable.3 Transfection efficiency achieved by plasmid DNA delivery has been significantly improved by the application of electroporation.4 However, the efficiency and the longevity of transgene expression by this method have not been evaluated under conditions of peripheral limb ischemia. In healthy young individuals, myocardial ischemia induces collateral vessel development, which provides protection from subsequent coronary events.5 In the presence of cardiovascular risk factors, including advanced age, angiogenesis is impaired, which may contribute to the increased severity of cardiovascular diseases in this patient population.6–10 Studies in experimental mouse models of hindlimb ischemia demonstrated impaired revascularization and CLI like symptoms (i.e. tissue necrosis) in old animals8,11 and in animals with compromised immune system.9,12 Diminished angiogenesis in these animal models corresponded with decreased expression of VEGF.8,9,13 Interestingly, substituting VEGF by exogenously delivered protein or gene has not shown therapeutic benefit in either aged mice11 or in balb/c mice,12 unlike in other mouse models.9 Similarly, VEGF refractory, severe hindlimb

NOS1177D gene transfer in CLI HS Qian et al

ischemia was also reported in mice deficient of endothelial nitric oxide synthase (ecNOS-KO).14 Cardiovascular risk factors impair (e.g. hypercholesterolemia, diabetes) the functional integrity of the endothelium and production of endothelial nitric oxide (EDNO) by ecNOS.15 In addition to the well-described vasculoprotective effects,16 EDNO also plays a critical role in angiogenesis.17,18 EcNOS-KO mice develop severe ischemic necrosis following femoral artery ligation indicating the importance of EDNO in vascular homeostasis, angiogenesis, collateral formation and maintenance of tissue integrity.14,19 Improvement in blood flow recovery and angiogenesis in response to ecNOS gene transfer has been demonstrated in normal rats following surgically induced hindlimb ischemia.20 However, the effects of ecNOS gene therapy has not been investigated in a model of CLI leading to limb loss owing to severe ischemic injury. It is uncertain whether restoring EDNO by ecNOS gene delivery can prevent the consequences of hindlimb ischemia in severe disease models. In the ecNOS-KO mice, mechanisms, such as developmental defects21 may also play a role in the progression of the severe ischemic phenotype. The present study has focused on two major goals (1) to investigate the efficiency and duration of transgene expression in murine hindlimb skeletal muscle following transfection of plasmid DNA by electroporation under ischemic conditions using a non-invasive bioluminescence imaging system and (2) to evaluate the potential therapeutic benefit of ecNOS gene transfer using a phosphomimetic mutant of ecNOS (NOS1177D)22,23 in mouse models of CLI, where the therapeutic effect of VEGF has been compromised.

1343

Results Optimization of gene delivery to ischemic hindlimb skeletal muscle Correlation between ex vivo and non-invasive in vivo detection of gene expression. The evaluation of the efficiency of a gene delivery system needs reliable monitoring of the transgene expression in target tissues. Most methods utilize enzymatic assays performed on tissue samples ex vivo, an end point measurement that does not allow for easy investigation of the duration kinetics of gene expression. To examine the persistence of gene expression in vivo, we have used a bioluminescence imaging system equipped with a highly sensitive ‘cooled charged-coupled detector’ (CCCD) camera. To validate the accuracy and correlation of this non-invasive method with the ex vivo assay, we have injected different doses (10–80 mg) of the marker gene luciferase into the adductor muscle of both hindlimbs of four intact C57Bl/6J mice (n ¼ 8 limbs) using a plasmid vector (pLuc). Animals were also injected with a control plasmid (pNull) (80 mg DNA). Four days following gene delivery, muscle samples were collected to measure luciferase activity from muscle homogenate immediately after the detection of the bioluminescence signal in vivo in anesthetized animals. We found a strong correlation between the in vivo bioluminescence signal and the ex vivo luciferase enzyme activity (R2 ¼ 0.92) (Figure 1a–d). Comparison of vectors for skeletal muscle gene delivery Success of angiogenic gene therapy is dependent on the efficiency of gene transfer. To evaluate different gene

Figure 1 Quantitation of skeletal muscle gene expression using firefly luciferase reporter gene in vivo and in vitro. Transfection efficiency was determined by cooled charged-coupled detector camera (a and b) or measuring luciferase activity in the adductor muscle homogenates (c) 4 days after intramuscular injection of different doses (10–80 mg) of plasmid DNA into the intact skeletal muscle of C57Bl/J6 mice (n ¼ 8). pcDNA3.1/hygro plasmid was used as control (pNull) at 80 mg concentration. The two measurements showed strong correlation (d). Gene Therapy


1344

delivery methods, we have injected the marker gene luciferase into the adductor muscle of intact and operated C57Bl/6J and balb/c mice using plasmid or viral vectors. Four days following gene delivery, muscle samples were collected to measure luciferase activity from muscle homogenate. We found that 80 mg plasmid delivery in combination with electroporation results in a 300-fold greater gene expression than gene delivery to the mouse hindlimb by intramuscular injection of 5 1010 PFU replication-incompetent type 5 adenovirus (Ad5), which was used in most mouse surgical hindlimb ischemia experiments previously (Figure 2a). Compared to naked plasmid injection, electroporation resulted in a 1000-fold enhancement in transfection efficiency (Figure 2b).

Figure 2 Comparison of different skeletal muscle gene delivery methods. Plasmid DNA, pLuc (80 mg), alone resulted in almost undetectable luciferase activity compared to Ad5Luc (5 1010 PFU). Significant 1000-fold improvement was observed combining the pLuc injection with electroporation (pLuc+EP). Panel a illustrates results with the in vivo bioluminescence measurement. Panel b depicts results obtained using in vitro luciferase activity measurements. In each group, four animals were used and injections were made in both hindlimbs, n ¼ 8. Asterisks (*) indicate statistically significant (Po0.05) differences between naked plasmid (pLuc) and Ad5Luc and pLuc+EP. The plus sign (+) illustrates statistically significant differences between Ad5Luc and pLuc+EP.

Timing of gene delivery in relation to acute ischemia Intramuscular gene delivery immediately following the surgery to ligate the femoral artery, which was performed to induce ischemia of the hindlimb (n ¼ 6), resulted in significantly attenuated gene expression at the operated side. Using the bioluminescence imaging system with the CCCD camera, we were able to follow the duration of expression in the same animal simultaneously in the operated and unoperated hindlimb (Figure 3a–b). Whereas the level of expression was significantly decreased at the operated site, the duration of expression followed a similar pattern in both hindlimbs (Figure 3c). By delaying gene delivery for 3 or 7 days following surgery, the level of gene expression increased significantly, as determined by measuring reporter gene expression 4 days following plasmid

Figure 3 Effect of hindlimb ischemia on skeletal muscle gene expression following surgical femoral artery ligation using firefly luciferase reporter gene. Gene expression was monitored non-invasively in balb/c mice following plasmid DNA (pLuc) delivery with electroporation into the adductor muscle of both hindlimbs on the same day of surgical unilateral femoral artery ligation, which results in severe ischemia in this model. Duration of gene expression was followed from day 1 to day 28 with the cooled charged-coupled detector camera (panel a) in n ¼ 6 animals. Bioluminescent signal was detected as early as 24 h after gene transfer. The signal peaked at day 7 and reached approximately 15% of the maximal level by day 28 (panel b). The level of gene expression was significantly (*Po0.05) lower at the ischemic site (panel b), but followed the same duration pattern (panel c). Gene Therapy


injection, that is, 7 and 11 days after the onset of ischemia (Figure 4).

Endothelial nitric oxide synthase expression in skeletal muscle following gene delivery Skeletal muscle expression of pNOS1177D was tested in ecNOS-KO mice using intramuscular plasmid injection in combination with electroporation. There was no detectable ecNOS enzyme-linked immunosorbent assay (ELISA) signal in ecNOS-KO mice before NOS1177D gene delivery. In C57Bl/6J mice, the average level of ecNOS expression was approximately 30 pg/mg skeletal muscle wet weight. Quantitation of ecNOS protein expression from skeletal muscle homogenates by ecNOS-specific ELISA showed that the transgene expression after gene delivery could reach or exceed levels seen in control (C57Bl/J6) mice (Figure 5). NOS1177D gene transfer results in blood flow recovery and limb salvage in mice deficient in endothelial nitric oxide synthase Blood flow recovery and gross pathology of hindlimbs. To evaluate the therapeutic potential of local EDNO delivery by the mutant human ecNOS gene, two groups of 6-month-old ecNOS-KO mice (n ¼ 8 in each group) were treated either with NOS1177D plasmid (pNOS1177D) or empty vector (pNull). Treatment with the mutant ecNOS gene resulted in a significantly improved blood flow recovery compared to the pNulltreated animals (Figure 6a). The onset of hindlimb ischemia was manifested in all mice as discoloration within hours following femoral artery resection. Without improvement in hindlimb perfusion, four out of the eight pNull-treated mice lost the operated limb

Figure 4 Optimization of timing of skeletal muscle gene delivery following surgical hindlimb ischemia using firefly luciferase reporter gene. Skeletal muscle gene delivery using plasmid DNA with electroporation was performed in intact and operated mouse hindlimbs either on the same day or 3 and 7 days following surgery in C57Bl/J6 mice (n ¼ 6). Performing gene delivery on the same day of surgery resulted in significant loss of gene expression compared to intact animals, similar to that shown in balb/c mice. Delaying gene expression following acute ischemia restored and exceeded the efficiency of gene expression to those seen in intact mice. Gene delivery, administered 3 and 7 days post-ischemia resulted in significantly higher expression levels compared to intact mice. Gene expression was measured 4 days following gene delivery, that is, 7 and 11 days following the onset of ischemia. Asterisks (*) indicate statistically significant (Po0.05) differences between gene expression in the intact and operated animals.

by day 28 (Figure 6b). In contrast, treatment with pNOS1177D prevented limb loss in all treated animals (Figure 6b).

1345

Histological findings and morphometry At day 28 after femoral artery ligation, mice were killed and both intact (non-operated) and operated hindlimbs were collected from pNull- and pNOS1177D-treated animals for histomorphometric evaluations (Figure 7). At the microscopic level, muscle groups from all operated hindlimbs (regardless of treatment group) showed some evidence of chronic response to ischemia. The most common finding was replacement/repair of damaged muscle fibers by adipocytes. Acute responses with active degeneration of muscle tissue were not common in these tissues (harvested on day 28 after surgery). However, of the three most severe cases, where areas of inflammation comprised over 10% of the hindlimb volume, all were in the pNull-treated group. These lesions were characterized by inflammatory infiltrates surrounding degenerating muscle fibers as well as the fatty replacement tissues within muscle (Figure 7b). Immunohistochemical stains for CD34, a marker of vascular endothelium, showed few capillaries in areas of muscle degeneration (Figure 7c, inset). There was a statistically significant decrease in total limb volume in the pNull-treated group as compared to the pNOS1177D-treated group, where total volume was similar to that seen in untreated limbs (Figure 8). The total volume of healthy muscle fibers per limb was also significantly decreased in pNull- versus pNOS1177Dtreated limbs. Although the volume of inflammatory infiltrates was increased in the pNull group as compared to the pNOS1177D group (Figure 8), this change did not reach statistical significance (P ¼ 0.0525). The amount of adipose tissue replacing muscle mass was similar in both treatment groups as was the amount of ‘other tissues’ (i.e., bone, connective tissue, peripheral nerves).

Figure 5 Quantitation of endothelial nitric oxide synthase (ecNOS) protein expression following intramuscular NOS1177D delivery to ecNOS-KO mouse hindlimb. ecNOS protein levels were determined by an ecNOS-specific enzyme-linked immunosorbent assay. There was no detectable signal in ecNOS-KO muscles without gene delivery, represented by the sign: in the figure. Results indicate that ecNOS expression following delivery of pNOS1177D with electroporation, represented by the sign: +, reaches ecNOS levels seen in control (C57Bl/J6) mice (B30 pg/mg, skeletal muscle ¼ SKM). Gene Therapy


1346

therapy can be effective in severe ischemic conditions, even in the absence of genetic deficiency in ecNOS.

Discussion Efficiency of the different gene delivery methods and the capacity of the host environment to execute the complex process of angiogenesis or arteriogenesis are key determinants of the clinical success with therapeutic revascularization using gene transfer. The effect of transfection efficiency and the consequences of the underlying disease on transgene expression and activity may be underestimated owing to lack of relevant experimental results. The present study had two main goals: (1) to investigate the potential impact of acute ischemia on skeletal muscle gene delivery/transgene expression efficiency and (2) to evaluate the therapeutic effect of gene transfer using a phosphomimetic ecNOS mutant (NOS1177D) in two different, exogenous VEGF resistant, experimental models of severe limb ischemia.

Figure 6 Therapeutic effect of NOS1177D gene delivery in endothelial nitric oxide synthase (ecNOS)-KO mice following unilateral hindlimb ischemia. Blood flow was measured by laser Doppler perfusion imaging (LDPI) following unilateral femoral artery occlusion and gene delivery (panel a). Bar graphs represent ischemic muscle perfusion normalized by the perfusion of the contralateral limb. The data are expressed as means7s.e.m. of n ¼ 8 animals. NOS1177D gene transfer resulted in significantly improved LDPI flow signal on days 7 and 28. Asterisks (*) indicate statistically significant (Po0.05) differences between pNOS1177D-treated mice (filled columns) compared to the pNull-treated groups (open columns). Ischemic tissue damage of the hindlimb was evaluated by taking photographs of the legs on the same days of the LDPI measurement (panel b). By day 28, there was significantly greater incidence of leg loss in the pNull (4/8) compared to the pNOS1177D-treated animals (0/8).

Effect of NOS1177D gene transfer in a mouse CLI model without genetic deficiency of ecNOS Similar to ecNOS-KO mice, balb/c mice also respond to unilateral surgical femoral artery occlusion with similarly severe, VEGF refractory hindlimb ischemia and autoamputation.12 Therefore, we chose this mouse strain to evaluate the effect of NOS1177D gene transfer in a model without genetic deficiency of ecNOS expression. Delivery of NOS1177D resulted in significantly improved blood flow recovery in balb/c mice in n ¼ 10 animals (Figure 9a). Elevated blood flow persisted even in the absence of maintained transgene expression measured by an ecNOS-specific Taqman polymerase chain reaction (PCR), ruling out the possibility that the blood flow increase was due to the vasodilating activity of EDNO (Figure 9b). These results confirm that NOS1177D gene Gene Therapy

Optimization of gene delivery to ischemic hindlimb musculature Comparing different methods of gene delivery to intact skeletal muscle we demonstrated that electroporation of plasmid DNA resulted in transfection efficiency 300-fold higher than after adenoviral (Ad) delivery. The dose of Ad was chosen based on earlier reports.9,24 Gene delivery to the skeletal muscle has been the preferred approach for therapeutic angiogenesis to treat peripheral vascular disease. Investigating the transfection efficiency and duration of gene expression in mouse models of hindlimb ischemia, we found that acute, severe ischemia significantly attenuated transgene expression following naked plasmid delivery with electroporation. We also showed, that this reduction in gene expression can be avoided by delivering the gene 3 or, even more efficiently, 7 days after surgery. Gene delivery, which was administered 7 days post-ischemia resulted in significantly higher expression levels compared to intact mice. Day 7 following surgical hindlimb ischemia is the peak of the post-ischemic angiogenic response involving infiltration of leukocytes.25 It is likely that non-muscular cells have also contributed to the high transgene expression at that time resulting in significant transgene expression increase compared to gene delivery to intact skeletal muscle. Similar finding has been reported in a rat model of hindlimb ischemia.26 Interestingly, the duration of gene expression, determined by measuring luciferase enzyme activity, was not different between the ischemic and contralateral limb. This may reflect, among other possibilities, that survival of the skeletal muscle following the surgical intervention is not affected adversely by gene delivery with electroporation. Therapeutic efficacy of NOS1177D gene transfer in VEGF refractory severe hindlimb ischemia Cardiovascular risk factors and age coincide with impaired revascularization.5 These clinical conditions are also accompanied by EDNO deficiency.15 Endothelial nitric oxide is a key contributor to angiogenesis, as a downstream mediator of growth factor(s).14 Endothelial dysfunction resulting in decreased EDNO availability,


1347

Figure 7 Representative histological sections of endothelial nitric oxide synthase (ecNOS)-KO mouse hindlimb. Sections were stained with hematoxylin and eosin for histomorphometric evaluation. Panels a and b illustrate a comparison of non-operated (right limb) to operated (left limb) sections in animals treated with null-vector. Note large areas of adipose tissue (*) within the quadriceps muscle of left leg. (These sections of hind limb are mirror images of each other.) Panel c shows an area of acute muscle degeneration from a null-treated mouse with a very severe lesion, which is characterized by a dense infiltration of leukocytes surrounding necrotic muscle fibers. Inset shows a sister section from the boxed region that was immunostained with antibodies against CD34, a marker of vascular endothelium. A few capillaries can be seen extending from the rim of the lesion into the avascular necrotic center. Bar ¼ 50 m.

Figure 8 Histomorphometric evaluation of the therapeutic effect of NOS1177D gene delivery in endothelial nitric oxide synthase (ecNOS)-KO mice. Histomorphometric quantitation of the skeletal muscle isolated from intact (n ¼ 3), ischemic pNull- and pNOS1177D-treated mice (n ¼ 8 for each vector treatment) at the end of the 28 days treatment period revealed a significantly (*Po0.05) larger overall volume of the pNOS1177D-treated limbs compared to the pNull samples, indicating a restoration of the tissue volume to that seen in intact (control) limbs (far right column). The volume % of healthy skeletal muscle was also significantly (*Po0.05) higher in the pNOS1177D group compared to the pNull group. There was a trend towards a decrease in inflammation in the pNOS1177D-treated group without statistical significance (P ¼ 0.053).

likely limits the efficiency of angiogenesis triggered either by endogenously expressed or exogenously delivered growth factors. Delivering a constitutively active EDNO-producing enzyme, like NOS1177D, may become a more effective therapeutic approach than delivery of a single angiogenic growth factor. During the angiogenesis process, VEGF induces an Akt-dependent phosphorylation of ecNOS, which results in simultaneous EDNO release from the VEGF-stimulated endothelium.22,23 The importance of EDNO in VEGF-induced angiogenesis/arteriogenesis has been elegantly demonstrated using gene transfer of VEGF in ecNOS-KO mice.14 The lack of VEGF-induced revascularization in ecNOS-KO mice suggests that EDNO is

Figure 9 Therapeutic effect of NOS1177D gene delivery in mice without genetic deficiency of endothelial nitric oxide synthase (ecNOS). Blood flow was measured by laser Doppler perfusion imaging (LDPI) following unilateral femoral artery occlusion and gene delivery in balb/c mice. Graph represents ischemic muscle perfusion normalized to the perfusion of the contralateral limb. The data are expressed as means7s.e.m. of n ¼ 10 animals. NOS1177D gene transfer resulted in significantly improved LDPI flow signal (panel a). Asterisks (*) indicate statistically significant (Po0.05) differences between NOS1177D-treated mice (filled columns) compared to pNull-treated animals (open columns). Elevated blood flow persisted even in the absence of maintained transgene expression in the pNOS1177D-treated limb compared to pNulltreated control (panel b).

indispensable for the full angiogenic/arteriogenic activity of VEGF for post-developmental vascular remodeling.14 The lack of effect in hindlimb blood flow recovery following VEGF gene delivery in ecNOS-KO mice is Gene Therapy


1348

similar to data reported using balb/c mice.12 Although balb/c mice are not deficient in ecNOS, they are also refractory to VEGF treatment, which may be the result of an impaired VEGF-receptor-2 (KDR) induction in this strain of mice following hindlimb ischemia.13 The ecNOS gene used in our studies, NOS1177D, is a phosphomimetic mutant of the full-length human ecNOS, which produces high local levels of EDNO without the need for a post-translational activation by phosphorylation.22,23 Indeed, overexpression of NOS1177D protected aged cells from apoptosis without additional stimulation of EDNO release, such as shear stress, unlike transfection with the wild-type ecNOS enzyme.27 NOS1177D substitutes for Akt-phosphorylated ecNOS, which results from KDR-mediated VEGF activation. VEGF-receptor-2 is a key receptor contributing to the angiogenic effect of VEGF.28 Besides mediating VEGF-induced ecNOS phosphorylation leading to subsequent EDNO release, KDR signaling is responsible for other Akt-mediated EDNO-independent downstream pathways including survival and migration of endothelial cells.29 As balb/c mice are not genetically deficient in ecNOS, the severe ischemic injury in this model likely reflects the importance of these additional KDR-mediated components of angiogenesis. Our results indicate that, unlike the overexpression of VEGF,12 delivery of the NOS1177D gene can overcome the ischemic revascularization deficiency in balb/c mice. We document here for the first time, that NOS1177D gene therapy is effective in animal models of VEGF refractory CLI with or without genetic deficiency of endogenous ecNOS. Efficient gene transfer with NOS1177D in the ischemic hindlimb significantly augmented blood flow, reduced skeletal muscle necrosis and improved limb salvage in two different murine models of CLI. Histomorphometric evaluation indicated augmented protection and/or regeneration of the pNOS1177D-treated skeletal muscle, which may be the result of a local anti-inflammatory effect of EDNO. The present study was not designed to assess the precise molecular mechanisms how EDNO exerts its therapeutic benefit. Owing to the multiple roles of EDNO in the maintenance of vascular homeostasis and endothelial cell survival,16 it is possible that NOS1177D gene therapy simultaneously contributes to the stimulation of angiogenesis, amelioration of microvascular dysfunction, vasodilation of existing vessels and remodeling of existing collaterals (arteriogenesis). The morphometric evaluation of angiogenesis at the end of the study at day 28 was not considered feasible for several reasons: (1) variablity in the extent of muscle degeneration and adipose cell replacement would not allow for relevant comparisons to uninvolved muscle as the metabolic needs of fat cells are likely different from those of myocytes; (2) regions of active inflammation have a blood supply architecture that is very different from that seen in intact skeletal muscle, again making comparisons difficult to interpret. The fact that overexpression of ecNOS was not detected at the end of the observation period in balb/c mice argues that the permanently elevated blood flow in these animals cannot be only due to vasodilation by EDNO.

Gene Therapy

In summary, we demonstrated that intramuscular delivery of NOS1177D using electroporation and optimized timing of gene delivery produces therapeutic benefit in experimental models of CLI, refractory to treatment with exogenous VEGF. NOS1177D-derived NO appears to be an effective component of ischemic revascularization, which could be applied to patients with severe CLI.

Materials and methods Animals and surgical procedure Male C57Bl/J6, balb/c and ecNOS-KO mice (Jackson Laboratory, Bar Harbor, ME, USA) between 3 and 6 months of age were used for the experiments. The ecNOS-KO mice have been backcrossed to C57Bl/J6 mice for more than 10 generations, therefore we used agematched C57Bl/J6 mice as control, when comparison needed to be made with wild-type animals. All mice were housed under controlled temperature (241C) and lighting (1400:1000 h light–dark cycle) conditions with free access to food (normal chow) and water. The experiments were conducted according to protocols approved by the Animal Care Committee at Berlex Biosciences, in agreement with the recommendation of the American Association for the Accreditation of Laboratory Animal Care. For surgery, animals were anesthetized with 1.5% isoflurane inhalation. A skin incision (2 mm) was performed at the upper portion of the left hindlimb overlying the femoral artery. Unilateral hindlimb ischemia was established by resection of a proximal segment of the femoral artery leaving the femoral vein intact. Gene expression analysis following skeletal muscle gene delivery To evaluate the efficiency of skeletal muscle gene delivery by different methods, the firefly luciferase gene driven by a cytomegalovirus (CMV) promoter was used in the pcDNA3.1 vector (pLuc) or in recombinationdeficient type 5 adenovirus (Ad5Luc). Luciferase expression was quantitated as an end point measurement by ELISA (Promega Corp., Madison, WI, USA) determining luciferase activity from skeletal muscle homogenates. To examine the duration of gene expression non-invasively, a bioluminescence imaging system was used called CCCD camera. It consisted of a camera (model LN/ CCD-1300EB) equipped with a 50-mm Nikon lens (Roper Scientific, Princeton Instruments, Trenton, NJ, USA), a light-tight specimen chamber, a camera controller (ST133) and a computer system for the data analysis. This system detects bioluminescent photons that are emitted from tissues of living animals expressing the luciferase gene and allows continuous, non-invasive monitoring of gene expression. Mice were anesthetized for the measurements and light emission was monitored placing the animals in a dark box. Two minutes before the detection, animals were injected with beetle luciferin (Promega Corp.,), 125 mg/kg body weight, into the peritoneal cavity. Light measurements were taken under the same conditions including time (2 min) and distance of lenses from the mice. pcDNA3.1/hygro plasmid was used as an empty vector control (pNull) at maximal DNA concentration (80 mg).


Therapeutic gene transfer agents and protocols for skeletal muscle gene delivery Plasmids encoding the NOS1177D mutant (pNOS1177D) of the human ecNOS gene was generated using the pcDNA3.1 vector with a CMV promoter. NOS1177D was constructed by point mutation of the serine at 1177 position in the full-length human ecNOS cDNA to an aspartic acid (S1177D) mimicking the phosphorylated protein.22 pcDNA3.1/hygro plasmid was used as control (pNull). Gene delivery was performed by intramuscular injection of the plasmid vectors (80 mg) in 50 ml of phosphate-buffered saline during anesthesia. Injections were made into the adductor magnus and quadriceps muscles of the ischemic hindlimb, 50 ml each, followed by electroporation using a caliper electrode (200 V/cm, 20 ms, 1 Hz, 8 pulses) (BTX, ECM 830, San Diego, CA, USA).

period, and stored at 801C until processed. Tissues were diced and homogenized using a Duall ground glass homogenizer in lysis buffer containing 25 mM Tris pH 7.8, 10% glycerol, 0.2% NP40, Roche Protease Inhibitor mini-tablet (containing: 1 mM ethylene diaminetetraacetic acid and inhibitors of serine and cysteine proteases). Cofactors for NOS enzyme activity were also added: 4 mM each of FAD, FMN, BH4, 3 mM dithiothreitol, 3 mM CaCl2 and 0.125 mM calmodulin. After 5 min of centrifugation using an Eppendorf centrifuge at maximal speed, the supernatant was saved and the pellet homogenized again. After a second centrifugation, the two supernatants were combined resulting in a final concentration of 80 mg muscle per ml with final volume of 300–950 ml. Concentration of ecNOS protein was determined using a specific ELISA kit from R&D Systems (cat#DEN00) according to the manufacturer’s instruction.

Non-invasive blood flow measurement Laser Doppler perfusion imaging (LDPI) was used to evaluate blood perfusion (Moore Instruments, Wilmington, DE, USA) before and immediately after surgery; and on days 1, 3, 7, 10, 14, 21 and 28 following the operation. Consecutive measurements, for each time point described, were performed by scanning over the same region of the hindlimb. The procedure was performed under anesthesia and placing the animals on a heating plate (371C) to control for variables introduced by body movement and temperature. Blood flow was displayed by colored pixels from minimal (dark blue) to maximal (red). The average perfusion of the ischemic and nonischemic leg was calculated on the basis of colored histogram pixels and final blood flow values were expressed as the ratio of ischemic to non-ischemic hindlimb perfusion. Ischemic tissue damage of the hindlimb was evaluated by taking digital images of the legs at days of the LDPI measurement.

Gene expression analysis in skeletal muscle samples Samples from the adductor muscles of the operated pNull- and pNOS1177D-treated limbs of balb/c mice were collected at the end of the experiment. The tissue pieces were homogenized in 600 ml RLT buffer and total RNA was then isolated by RNeasy kit with DNase I digestion (Qiagen, Valecia, CA, USA). Relative abundance of ecNOS as well as internal control 18S rRNA were measured by real-time quantitative PCR performed on ABI PRISM 7700 Sequence Detector (Applied Biosystems, Foster City, CA, USA). Primers and probe for ecNOS (GenBank accession number, NM_008713) were: upper primer, 50 -CGTCATCGGCGTGCT-30 (nt 3436– 3450), lower primer, 50 -ACCTCCTGGGTGCGC-30 (nt 3510–3496) and probe, 50 -6FAM-CGGGATCAGCAACG CTACCA-TAMRA-30 (nt 3452–3471). Endothelial nitric oxide synthase expression is calculated against a standard curve with serial dilutions of total RNAs from murine hemangioendothelioma.30 Measurements were made in quadruplicate for each sample. Endothelial nitric oxide synthase expression in the operated and treated limbs is shown in comparison to levels, which was detected in skeletal muscle homogenates of the contralateral, non-ischemic leg.

Histomorphometric evaluation Mice were killed 28 days after gene delivery. The hind quarters diarticulated from the body, stripped of skin, fixed in 10% phosphate-buffered formalin, placed in decalcifying solution for 2 days, and then blocked in a systematic random manner. In brief, the feet (if remaining) were removed and sectioned between the second and third digits, with both sections embedded in a single cassette. The hindlimbs were then sectioned perpendicular to the femur bone at 2 mm intervals to enable histomorphometric evaluation of treatment effects. All hindlimb sections were placed in a single cassette, dehydrated and embedded with the lateral face of each section at the cutting face of the block. Five-micron-thick sections were cut and stained with hematoxylin and eosin for histomorphometric evaluation using the C.A.S.T. stereology system (Olympus, Denmark). The area occupied by each different histologic feature under study was estimated for each limb using systematic random tissue sections. This allowed for estimation of relative volumes as the systematic random sections are separated by a known distance, in this case, 2 mm. Protein assays in skeletal muscle homogenates In some experiments, samples from the adductor muscle were collected for ELISA at the end of the observation

1349

Calculations and statistical analysis All results are expressed as mean7standard error of the mean. Statistical significance was evaluated using unpaired Student’s t-test or analysis of variance (ANOVA) for comparisons between two means with the aid of a computer software (Statistica 6.0) or by ANOVA followed by the Fisher protected least significance difference post hoc test using StatView software (SAS, Institute, Cary, NC, USA). Linear regression analysis between two variables was performed by using Microsoft Excel. A value of Po0.05 was considered as statistically significant.

References 1 Rissanen TT, Vajanto I, Yla-Herttuala S. Gene therapy for therapeutic angiogenesis in critically ischaemic lower limb – on the way to the clinic. Eur J Clin Invest 2001; 31: 651–666. 2 Khan TA, Sellke FW, Laham RJ. Gene therapy progress and prospects: therapeutic angiogenesis for limb and myocardial ischemia. Gene Therapy 2003; 10: 285–291. Gene Therapy


1350 3 Li S, Huang L. Nonviral gene therapy: promises and challenges. Gene Therapy 2000; 7: 31–34. 4 Mir LM, Bureau MF, Gehl J, Rangara R, Rouy D, Caillaud JM et al. High efficiency gene transfer into skeletal muscle mediated by electric pulses. Proc Natl Acad Sci USA 1999; 96: 4262–4267. 5 Kodama K, Kusuoka H, Sakai A, Adachi T, Hasegawa S, Ueda Y et al. Collateral channels that develop after an acute myocardial infarction prevent subsequent left ventricular dilation. J Am Coll Cardiol 1996; 27: 1133–1139. 6 Anversa P, Li P, Sonnenblick EH, Olivetti G. Effects of aging on quantitative structural properties of coronary vasculature and microvasculature in rats. Am J Physiol 1994; 267: H1062–H1073. 7 Gerhard M, Roddy MA, Creager SJ, Creager MA. Aging progressively impairs endothelium-dependent vasodilation in forearm resistance vessels of humans. Hypertension 1996; 27: 849–853. 8 Rivard A, Fabre JE, Silver M, Chen D, Murohara T, Kearney M et al. Age-dependent impairment of angiogenesis. Circulation 1999; 99: 111–120. 9 Couffinhal T, Silver M, Kearney M, Sullivan A, Witzenbichler B, Magner M et al. Impaired collateral vessel development associated with reduced expression of vascular endothelial growth factor in ApoE/ mice. Circulation 1999; 99: 3188–3198. 10 Panchal VR, Rehman J, Nguyen AT, Brown JW, Turrentine MW, Mahomed Y et al. Reduced pericardial levels of endostatin correlate with collateral development in patients with ischemic heart disease. J Am Coll Cardiol 2004; 43: 1383–1387. 11 Pola R, Ling LE, Aprahamian TR, Barban E, Bosch-Marce M, Curry C et al. Postnatal recapitulation of embryonic hedgehog pathway in response to skeletal muscle ischemia. Circulation 2003; 108: 479–485. 12 Masaki I, Yonemitsu Y, Yamashita A, Sata S, Tanii M, Komori K et al. Angiogenic gene therapy for experimental critical limb ischemia: acceleration of limb loss by overexpression of vascular endothelial growth factor 165 but not of fibroblast growth factor2. Circ Res 2002; 90: 966–973. 13 Fukino K, Sata M, Seko Y, Hirata Y, Nagai R. Genetic background influences therapeutic effectiveness of VEGF. Biochem Biophys Res Commun 2003; 310: 143–147. 14 Murohara T, Asahara T, Silver M, Bauters C, Masuda H, Kalka C et al. Nitric oxide synthase modulates angiogenesis in response to tissue ischemia. J Clin Invest 1998; 101: 2567–2578. 15 Zeiher A M, Drexler H, Wollschla¨ger H, Just H. Endothelial dysfunction of coronary microvasculature is associated with impaired coronary blood flow regulation in patients with early atherosclerosis. Circulation 1991; 84: 1984–1992. 16 Kauser K, Rubanyi GM. ‘Nitric oxide deficiency’ in cardiovascular diseases. Cardiovascular protection by restoration of endothelial nitric oxide production. In: Rubanyi GM (ed).

Gene Therapy

17 18 19

20

21

22

23

24

25

26

27

28

29

30

Mechanisms of Vasculoprotection. Springer-Verlag: NY, 2002, pp 1–31. Ziche M, Morbidelli L. Nitric oxide and angiogenesis. J Neurooncol 2000; 50: 139–148. Katusic ZS. Therapeutic angiogenesis. Arterioscler Thromb Vasc Biol 2002; 22: 1254–1255. Qian HS, Liu P, Kauser K, Rubanyi GM. Nitric oxide deficiency leads to impaired angiogenesis and severe dysfunction of microcirculation in a mouse hind limb ischemia model. In: Proceedings of the 7th World Congress of Microcirculation. Monduzzi Editore: Sydney, Australia, 2001, pp 525–529. Smith RS, Lin K-F, Agata J, Chao L, Chao J. Human endothelial nitric oxide synthase gene delivery promotes angiogenesis in a rat model of hindlimb ischemia. Arterioscler Thromb Vasc Biol 2002; 22: 1279–1285. Gregg AR, Schauer A, Shi O, Liu Z, Lee CG, O’Brien WE. Limb reduction defects in endothelial nitric oxide synthase-deficient mice. Am J Physiol 1998; 275: H2319–H2324. Fulton D, Gratton JP, McCabe TJ, Fontana J, Fujio Y, Walsh K et al. Regulation of endothelium derived nitric oxide production by the protein kinase Akt. Nature 1999; 399: 597–601. Dimmeler S, Dernbach E, Zeiher AM. Phosphorylation of the endothelial nitric oxide synthase at ser-1177 is required for VEGF-induced endothelial cell migration. FEBS Lett 2000; 477: 258–262. Rivard A, Silver M, Chen D, Kearney M, Magner M, Annex B et al. Rescue of diabetes-related impairment of angiogenesis by intramuscular gene therapy with adeno-VEGF. Am J Pathol 1999; 154: 355–363. Couffinhal T, Silver M, Zheng LP, Kearney M, Witzenbichler B, Isner JM. Mouse model of angiogenesis. Am J Pathol 1998; 152: 1667–1679. Takeshita S, Isshiki T, Sato T. Increased expression of direct gene transfer into skeletal muscles observed after acute ischemic injury in rats. Lab Invest 1996; 74: 1061–1065. Hoffmann J, Haendeler J, Aicher A, Rossig L, Vasa M, Zeiher AM et al. Aging enhances the sensitivity of endothelial cells toward apoptotic stimuli: Important role of nitric oxide. Circ Res 2001; 89: 709–715. Endo A, Fukuhara S, Masuda M, Ohmori T, Mochizuki N. Selective inhibition of vascular endothelial growth factor receptor-2 (VEGFR-2) identifies a central role for VEGFR-2 in human aortic endothelial cell responses to VEGF. J Recept Signal Transduct Res 2003; 23: 239–254. Sato Y, Kanno S, Oda N, Abe M, Ito M, Shitara K et al. Properties of two VEGF receptors, Flt-1 and KDR, in signal transduction. Ann NY Acad Sci 2000; 902: 201–205. Obeso J, Weber J, Auerbach R. A hemangioendotheliomaderived cell line: its use as a model for the study of endothelial cell biology. Lab Invest 1990; 63: 259–269.