In vitro expression of calcitonin gene-related peptide in human endothelial cells transfected with plasmid and retroviral vectors

False positive results in chimeraplasty for von Willebrand Disease

Chimeraplasty or the use of chimeric RNA/DNA oligonucleotides (RDOs) to correct single-base mutations emerged in the field of gene therapy with reported base pair conversions of up to 40%. We investigated the applicability of chimeraplasty to correct a point mutation in the von Willebrand Factor (VWF) gene resulting in a von Willebrand Disease (VWD) type 3 phenotype. Although we have access to VWD type 3 dogs, we used wild type endothelial cells for in vitro studies, as isolation of endothelial cells from VWD type 3 dogs is not straightforward due to the bleeding diathesis. RDOs to convert the wild type VWF gene into VWD type 3 gDNA were constructed and used in various transfection conditions. However, no gene conversion could be detected either in the RNA or in the DNA isolated from transfected cells, not even with the sensitive colony hybridisation technique, despite the presence of RDOs in the cell nucleus. On the other hand, sequence analysis of isolated DNA of transfected cells did reveal the presence of VWF type 3 DNA. However, this apparent conversion is very likely not the result of RDO-mediated nucleotide conversion as the same VWF type 3 DNA sequence was also detected in negative control experiments where no RDO was used. Our negative results are in line with the emerging reports of chimeraplasty failure and can contribute to the call for an international "chimeraplasty consortium" with free exchange of results to clarify the controversy about the applicability of the RDO-mediated base conversion.

Neuronal expression and regulation of CGRP promoter activity following viral gene transfer into cultured trigeminal ganglia neurons

We have examined the regulation of calcitonin gene-related peptide (CGRP) promoter activity in primary cultures of rat trigeminal ganglia neurons. A viral vector was used to circumvent the potential complication of examining only a small subpopulation of cells in the heterogeneous cultures. Infection with high titers of recombinant adenovirus containing 1.25 kb of the rat CGRP promoter linked to the beta-galactosidase reporter gene (AdCGRP-lacZ) yielded expression in about 50% of the CGRP-expressing neurons. The CGRP-lacZ reporter gene was preferentially expressed in neurons, with 91% co-expression with endogenous CGRP. In contrast, an adenoviral vector containing a CMV-lacZ reporter was predominantly expressed in non-neuronal cells, with only 29% co-expression with CGRP. We then asked whether the CGRP promoter in the viral vector could be regulated by serotonin receptor type 1 (5-HT(1)) agonists. Promoter activity was decreased two- to threefold by treatment with five 5-HT(1B/D) agonists, including the triptan drugs sumatriptan, eletriptan, and rizatriptan that are used for migraine treatment. As controls, CMV promoter activity was not affected, and 5-HT(1B/D) receptor antagonists blocked the repression caused by sumatriptan and eletriptan. Thus, adenoviral gene transfer can be used in trigeminal ganglia neurons for studying the mechanisms of triptan drug action on CGRP synthesis.

AP-4F, antennapedia peptide linked to an amphipathic alpha helical peptide, increases the efficiency of Lipofectamine-mediated gene transfection in endothelial cells

Typically, endothelial cells are difficult to transfect. In this study, we report that antennapedia peptide (AP) linked to L-4F, a water-soluble, amphipathic alpha helical peptide that avidly binds lipids (AP-4F) increases Lipofectamine 2000-mediated transfection of bovine coronary endothelial cell cultures. Transfection efficiency was monitored by flow cytometry and fluorescent microscopy. Lipofectamine 2000 transfection of endothelial cell cultures with green fluorescence protein (GFP)-DNA typically yields transfection efficiencies of 35.4+/-3.3% with low levels of cell death (8.1+/-1.0%). Pre-treatment of the Lipofectamine 2000-GFP-DNA complexes with AP-4F for 5 min increased transfection to 58.2+/-2.8% without increasing cell death. AP-4F increases Lipofectamine 2000-mediated transfection in a time-dependent fashion (within 10-20 min). Systematic studies reveal that the individual components of AP-4F, i.e., AP and L-4F alone, are ineffective in increasing Lipofectamine 2000-mediated transfection and that AP-4F must be directly associated with DNA liposomes prior to transfection for optimal uptake by endothelial cells. These observations demonstrate that AP-4F may be useful for increasing the transfection efficiency of endothelial cell cultures with standard commercially available reagents.

Gene-transfer systems for human endothelial cells. stewart.martin@nottingham.ac.uk

By virtue of its location and importance in a number of pathophysiological processes the endothelium represents an attractive target tissue for gene-transfer and gene-therapy strategies. Although it is important to maximise gene-transfer to endothelial cells in such strategies primary human endothelial cells have proven to be rather intransigent to a variety of transfection techniques both in vitro and in vivo. We report on the variety of techniques in current use, revealing their strengths and weaknesses, indicate the steps that should ideally be taken to optimise expression and discuss the usefulness and future directions for viral mediated transduction.

Vascular endothelial cells synthesize and secrete brain-derived neurotrophic factor

Brain-derived neurotrophic factor (BDNF) is an abundant neurotrophin in brain and peripheral nerves, where it affects neural development, survival and repair after injury. BDNF has been detected in rat and human blood, but the source of circulating BDNF is not established. BDNF messenger and peptide were detected in cultured cells and in the culture medium of human umbilical vein endothelial cells. The expression of BDNF was up-regulated by elevation of intracellular cAMP and down-regulated by Ca(2+) ionophore, bovine brain extract and laminar fluid shear stress. These results suggest that vascular endothelial cells may contribute to circulating BDNF.

Adrenomedullin receptor antagonism by calcitonin gene-related peptide(8-37) inhibits carotid artery neointimal hyperplasia after balloon injury

Intimal injury by angioplasty results in a series of changes, including smooth muscle cell hyperplasia, that lead to vascular restenosis. Adrenomedullin, a potent vasodilator peptide, has natriuretic effects, and its plasma concentration is elevated in cardiovascular diseases. Adrenomedullin is secreted by endothelial and vascular smooth muscle cells, but its role in neointimal hyperplasia after balloon injury has not been previously described. We investigated the role of endogenous adrenomedullin in neointimal hyperplasia using an in vivo rat model of postinjury vascular restenosis. In the injured rats, bromodeoxyuridine-labeled nuclei in the media of untreated common carotid arteries were increased 2 days after injury, which were suppressed by in vivo treatment with the adrenomedullin receptor antagonist calcitonin gene-related peptide (CGRP)(8-37). Inhibition of neointimal hyperplasia by CGRP(8-37) was distinct at 7 and 14 days, whereas CGRP(1-37) had no effect. The expression of adrenomedullin in the media of both untreated and treated common carotid arteries was elevated at 2 days and further enhanced in hyperplastic intima of untreated common carotid arteries at 7 days. Our findings suggest a novel role for endogenous adrenomedullin in balloon injury-induced restenosis and indicate that CGRP(8-37) may be useful for the prevention of vascular restenosis.

Targeting of a Heterologous Protein to a Regulated Secretion Pathway in Cultured Endothelial Cells

The stimulation of regulated exocytosis in vascular endothelial cells (EC) by a variety of naturally occurring agonists contributes to the interrelated processes of inflammation, thrombosis, and fibrinolysis. The Weibel-Palade body (WPB) is a well-described secretory granule in EC that contains both von Willebrand factor (vWF) and P-selectin, but the mechanisms responsible for the targeting of these proteins into this organelle remain poorly understood. Through adenoviral transduction, we have expressed human growth hormone (GH) as a model of regulated secretory protein sorting in EC. Immunofluorescence microscopy of EC infected with GH-containing recombinant adenovirus (GHrAd) demonstrated a granular distribution of GH that colocalized with vWF. In contrast, EC infected with an rAd expressing the IgG1 heavy chain (IG), a constitutively secreted protein, did not demonstrate colocalization of IG and vWF. In response to phorbol ester, GH as well as endogenously synthesized vWF were rapidly released from GHrAd-infected EC. By immunofluorescence microscopy, granular colocalization of GH with endogenous tissue-type plasminogen activator (tPA) was also demonstrated, and most of the tPA colocalized with vWF. These data indicate that EC are capable of selectively targeting heterologous proteins, such as GH, to the regulated secretory pathway, which suggests that EC and neuroendocrine cells share common protein targeting recognition signals or receptors.

Two distinct HCO(-)(3)-dependent H(+) efflux pathways in human vascular endothelial cells

Intracellular pH (pH(i)) regulation in human umbilical vein endothelial cells (HUVEC) was investigated. The pH(i) was recorded using seminaphthorhodafluor-1 (SNARF-1). Cells were intracellularly acid loaded with NH(4)Cl prepulse. In HEPES-buffered Tyrode (nominally HCO(-)(3) free), pH(i) recovery from acid load was inhibited by 1.5 mM amiloride or Na(+)-free solution. Additionally, in HCO(-)(3)-buffered Tyrode, a HCO(-)(3)-dependent pH(i) recovery from acidosis was evident in the presence of 1.5 mM amiloride, which mediated complete recovery of pH(i) (7.26). In Na(+)-free solution, the HCO(-)(3)-dependent acid extruder mediated pH(i) recovery after an acid load but only back to 7.09. These results suggest that there are two HCO(-)(3)-dependent acid extruders in the HUVEC. One is Na(+) dependent, and the other is Na(+) independent. The former was further shown to be completely inhibited by 0.5 mM DIDS, whereas the latter was only inhibited by 24.6%. In Cl(-)-free solution, both of the HCO(-)(3)-dependent pathways were inhibited. In conclusion, one HCO(-)(3)-dependent acid extruder in the HUVEC resembles the Na(+)-dependent Cl(-)/HCO(-)(3) exchange found in other tissues, and the other is Cl(-) dependent but Na(+) independent.

Optimization of transfection of human endothelial cells

Recently developed transfection methods for mammalian cells provide a powerful means for the study of gene function. Unfortunately, human endothelial cells were relative refractory to the classic transfection techniques. In this study we compared the usability of calcium phosphate, DEAE-dextran transfection, transferrinfection, lipofection, and electroporation for the transfection of early passage HUVECs and for the human endothelial cell lines ECV 304 and EA.hy 926. The classic transfection methods resulted in no or only marginal expression of the reporter gene E. coli beta-galactosidase. For lipofection experiments we compared the commercially available liposome formulations DOTAP and Transfectam with liposomes prepared of dimethyldioctadecylammoniumbromide (DDAB) or 1,2-dimyristyloxypropyl-3-dimethylhydroxyethyl ammonium bromide (DMRIE) as the cationic lipid compound and dioleylphosphatidylethanolamine (DOPE) or Azolectin (a crude fraction of soybean lipids, commercially available as phosphatidylcholine II) as neutral co-lipid. Because the protocol for the chemical synthesis of DMRIE has not been published yet, we developed a protocol for the chemical synthesis of this cationic lipid. With transfection protocols optimized for each cell line, we could achieve transfection efficiencies up to 2%. Compared to the other methods used, the lipofection proved to be a reliable technique for the efficient transfection of the human endothelial cell lines ECV 304 and EA.hy 926. Although we achieved a maximum transfection efficiency of 0.45% for the lipofection of HUVEC, the electroporation seemed to be the better choice for these cells.

Cardiovascular gene transfer

The cardiovascular system is a promising site for gene delivery because of its large surface area and direct contact with the blood stream. Gene transfer into vascular cells by adenoviral and retroviral vectors or liposomes has been demonstrated in several cell types and animal models. These methods have been somewhat successful; however, many improvements must be made before use in humans is possible. Some of these vectors exhibit toxicity and mediate variable expression of the transferred gene. New generations of viral vectors, new lipid formulations, other gene transfer methods, and more effective delivery methods must be developed to combat these problems.