Eur J Cardiothorac Surg 1998;13:460-466
© 1998 Elsevier Science NL
Gene transfer into rat heart-derived endothelial cells1
Marc Hein,
Markus Ernst,
Frank Möller,
Dieter Regensburger
Department of Cardiovascular Surgery, University of Kiel, Arnold Heller Str. 7, 24105 Kiel, Germany
Received 20 October 1997;
received in revised form 3 December 1997;
accepted 27 January 1998.
Corresponding author. Tel.: +49 431 5974401; fax: + 49 431 5974402; e-mail: MarcHein@compuserve.com
 |
Abstract
|
|---|
Objective: Progressive graft arteriosclerosis is responsible for the majority of late deaths in cardiac transplant recipients. Despite many investigations, the pathogenesis of this disease remains undetermined and its control inadequate. A somatic gene transfer during the cold ischemic time and thus before transplantation might be a new therapeutic tool. This approach allows a long incubation time of the DNA and a safe transfer with liposomes and transferrin with less adverse effects for the organ recipient. Methods: The target cells (microvascular endothelial cells (MVECs)) for this gene transfer were isolated from rat hearts by perfusion with collagenase via an aortic cannulae. The cells were purified by changing the medium 30 min after subcultivation in order to remove fibroblasts and smooth muscle cells. The endothelial cells (ECs) were identified by typical morphology and the uptake of Dil-Ac-LDL. The gene transfer was carried out with a ß-galactosidase reporter plasmid (pCMVß), cationic liposomes (Lipofectin®), and transferrin. Different transfection solutions were prepared with or without serum, and with different plasmid–liposome ratios and transferrin concentrations. The transfer rate was monitored with a semiquantitative orthonitrophenyl-ß-D-galactoside (ONPG) assay and histologically by X-Gal staining. The cytotoxicity of this procedure was determined with a colorimetric ELISA with Alamar blue®. The cardioplegic property of the transfection solution was tested in a Langendorff perfusion system monitoring the coronary blood flow over time after a cold ischemic time of 4 h. Results: The maximal gene expression could be detected after transfection with 4 µl Lipofectin, 2 µg pCMVß, and 16 µg transferrin/200 µl transfection solution. Under these conditions 60% of the cells showed a blue staining with X-Gal. Only 20% of the cells died during transfection. The lowest cytotoxicity during cold ischemic time for ECs was assessed with normal cell culture medium and the Buckberg solution. The best coronary flow rates after 4 h cold ischemia of the heart were measured for cardioplegia with St. Thomas and Buckberg solutions. In summary, the best transfection solution with a good cardioplegic property was the Buckberg solution. Conclusions: Finally, the results of this study show that an effective DNA delivery with a low toxicity into ECs is possible with a combination of liposomes and transferrin. This method might be useful for a safe and effective gene transfer into solid organs during the cold ischemic time and thus a therapeutic tool for chronic rejection.
Key Words: Gene transfer • Heart transplantation • Liposomes
|
Introduction
|
|---|
Gene transfer might be a useful strategy for the therapy of untreatable human diseases. Important requirements are a high transfection efficiency with a long expression time of the introduced genes with less adverse effects for the recipient. These goals have not been achieved yet. Transplant is widely applied and has revealed significant potential for gene therapy. Gene transfer might be feasible during the cold ischemic time at organ transportation. By adding genes to the organ perfusate the incubation time would allow for a high local gene concentration. Adverse effects for the recipient might be minimized by removal of the transfection solution prior to reperfusion.
At the moment, few investigations describe a gene transfer via the vasculature during the cold ischemia of solid organs. In a heterotopic rat heart transplantation model an effective, but poorly quantified, gene transfer with liposomal vectors for a reporter plasmid was described
[1]. In the rabbit an intracoronary gene transfer of interleukin-10 and transforming growth factor-ß1 was carried out with replication defective adenoviral vectors. Gene product expression could be detected in the allograft and in the recipient serum
[2]. A gene transfer technique for rat hearts with cationic liposomes and the hemagglutinating virus of Japan liposome (HVJ) during cold ischemia was described previously
[3]. Both vectors only allowed for a low level of gene expression in endothelial cells (ECs). Inhibition of neointimal formation after cardiac transplantation by administration of antisense ODN against CDK2 kinase using this gene transfer technique was observed
[4]. At this time, the benefit of viral vectors was not clearly proven, and adverse effects in the recipient could not be excluded
[5]. Successful retrovirus-mediated gene transfer to vascular cells requires the induction of proliferating target cells. Thus these viruses are not applicable to transfect quiescent ECs. Immune response to adenoviral proteins is a major limitation to the use of these vectors
[6]. However, adenovirus vectors in current use evoke non-specific inflammatory response leading to cell proliferation
[5].
Liposomes offer many advantages as gene transfer vectors: they can be used to transfer expression cassettes of unlimited size, they do not replicate or recombine to form an infectious agent, and they may evoke fewer inflammatory or immune responses. The only disadvantage at the moment is the lower efficiency compared with viral vectors. The present study was designed to optimize a liposome-mediated gene transfer into cultured microvascular endothelial cells (MVECs). These are the target cells for an in vivo gene transfer into solid organs during cold ischemia. Higher transfection rates might be achieved by varied liposome–plasmid ratios or concentrations, and changes in transfection solutions. The transfection solution has to guarantee for three properties: high transfection efficiency, low cytotoxicity, and a good organ conservation during cold ischemia. Therefore MVECs were isolated from rat hearts and transfected in vitro with a ß-galactosidase reporter plasmid using a cationic liposome and transferrin. The ß-galactosidase reporter plasmid was chosen in most investigations involving gene transfer into vascular cells. Transferrin is known to improve liposomal gene transfer into HeLa cells in vitro
[7]. Different liposome and plasmid concentrations, and transfection solutions were tested. The transfection efficiency was monitored histologically and semiquantitatively with an ELISA. Cytotoxicity of the transfection procedure in cell culture was analyzed with an ELISA technique measuring the reduction of an indicator dye by cell proliferation and metabolism. The cardioplegic properties of the different transfection solutions were tested in a Langendorff perfusion system.
|
Materials and methods
|
|---|
Isolation of microvascular endothelial cells
Hearts from WKY rats (n=10) were explanted and perfused via an aortic cannulae with 10 ml calcium-free PBS, followed by perfusion with collagenase CLS II (Biochrome, USA; 1 mg/ml in 30 ml calcium-free PBS containing 1% BSA) for 30 min at 37°C. The perfusate was centrifugated and resuspended in M199 supplemented with ECGF (Boehringer, Germany), 10% fetal calf serum (FCS), and penicillin/streptomycin, seeded on 6-well plates and cultivated at 37°C in a 5% CO2 incubator. The medium was changed the first time after 30 min to remove non-endothelial cells. The cells were subcultivated at the time of confluence followed by early medium exchange after 30 min. MVECs were identified by their typical morphology and uptake of Dil-Ac-LDL (Molecular Probes, USA). MVEC were seeded into Lab-Tek chamber slides and grown to confluence. Dil-Ac-LDL was added to a final concentration of 10 µg/ml and cells were cultivated for another 12–24 h at 37°C. After washing the cells three times with PBS they were fixated with 0.5% glutaraldehyde in PBS for 10 min at 4°C and thereafter examined with fluorescence microscopy at 571 nm (Zeiss Axiovert, Germany). The cell cultures were investigated for contaminating fibroblasts or smooth muscle cells by staining them with a monoclonal 
-smooth muscle actin antibody (Sigma, USA), and the specific fibroblast antibody ‘mAK AS02’ (Dianova, Germany). For all gene transfer experiments MVECs isolated from one heart were used.
Gene transfer protocol
For transfection experiments the cells were grown to confluence in 96-well plates and Lab-Tek chamber slides (Costar). The transfection solution was prepared according to the following protocol: 8 µg of the ß-galactosidase reporter plasmid pCMVß (Qiagen, Germany) was diluted into 100 µl serum-free M199; 4–24 µl Lipofectin (Gibco, USA) was diluted into 100 µl serum-free M199 with or without 8–64 µg transferrin (Becton Dickinson, USA). Both solutions were mixed after 30 min at room temperature. After 15 min, 800 µl of M199 with or without 10% FCS, was added. For some experiments M199 (n=4) was diluted with Buckberg cardioplegic solution 4:1 (Köhler Chemie, Germany). A serial 1:1 dilution of the transfection agent was prepared. The cell layers were washed once with serum-free M199 and 200 µl of the transfection solution was added per well. After an incubation time of 4 h at 37 and 4°C, the DNA-containing medium was replaced with 250 µl normal growth medium and incubated for a total of 48 h at 37°C.
To optimize the gene transfer, transfection solutions were prepared with different liposome–plasmid ratios (1:2, 1:1, 2:1, 3:1) and with or without 16 µg transferrin. The liposome–plasmid complexes (LPCs) were added to different transfection solutions: cell culture medium (M199) with or without FCS, and a blood–cardioplegic solution (Buckberg), which was diluted with M199 at a ratio of 1:4.
Analysis of transfection efficiency
For semiquantitative analysis of the transfection efficiency, an ELISA assay with orthonitrophenyl-ß-D-galactoside (ONPG) (Sigma, USA) in 96-well plates was carried out. ONPG is a chromogenic substrate for the ß-galactosidase. The cleavage of ONPG results in a yellow color. The cell layers, were 48 h after transfection, fixed with 0.5% glutaraldehyde for 10 min at 4°C, washed twice with PBS, and incubated with 200 µl ONPG solution (8.8 mg/10 ml PBS with 0.1 M MgCl2) at 37°C for 12 h. The optic density correlates with the total enzymatic activity representing the transfection rate. It was measured at 420 nm.
For microscopy the cell layers were stained with X-Gal (ß-Gal Staining Kit, Invitrogen, USA). The gene product in transfected cells, the ß-galactosidase, catalyzes the hydrolysis of X-Gal, producing a blue color that can be visualized under a microscope and the percentage of transfected cells could be calculated.
Analysis of cytotoxicity of transfection solutions in cell culture
The cytotoxicity of the LPCs to MVEC layers was monitored with the Alamar Blue Assay (BioSource, USA). This assay incorporates an oxidation–reduction indicator that changes color in response to chemical reduction of growth medium resulting from cell metabolism and proliferation. Reduction related to growth causes the indicator to change from oxidized (blue color) to reduced (red color) form. After removal of the transfection solution, normal cell culture medium supplemented with 10% Alamar blue was added to the 96-well plates. The change in color could be monitored with an ELISA reader at 620 nm at different time points. Cell proliferation correlates with the difference in change of the optic density measured after 0 and 12 h (
OD). Thus, a high cytotoxicity is associated with a low OD.
Different solutions, which might be useful as carriers for the DNA–liposome complexes and cardioplegia, were tested for cytotoxicity to MVEC cultures. Cells were grown to confluence in 96-well plates and after removal of cell culture medium, 250 µl of cold PBS, M199, M199 with 10%FCS, St. Thomas solution and Buckberg solution (Köhler Chemie, Germany) were added. After 4 h at 4°C the cells were fed with normal cell growth medium supplemented with 10% Alamar blue. Cytotoxicity was monitored by changes in optic density over time as described above.
Analysis of cytotoxicity of transfection solutions during cardioplegia
Different transfection solutions and cardioplegic solutions were tested for cardiac conservation in a Langendorff perfusion system. Rat hearts were explanted and perfused with Krebs–Henseleit buffer for 30 min in a Langendorff system to stabilize. Cardiac arrest was induced by perfusion with 10 ml 4°C cold M199 (+8 mM KCl) with or without serum, St. Thomas or Buckberg solution for blood cardioplegia (M199 and FCS:Buckberg; 4:1). After 4 h of cold ischemia the hearts were reperfused with 37°C Krebs–Henseleit buffer. Coronary blood flow (ml/min) was monitored by collecting the perfusate for 1 min and compared with control hearts, which had no cold ischemia. A decline in coronary blood flow was marked for poor cardiac conservation during cold ischemic time.
All values are expressed as mean±S.E. Comparisons between different groups were carried out using the Student’s t-test.
|
Results
|
|---|
Microvascular endothelial cells
Pure MVEC cultures could be obtained by repeated subcultivation with early medium exchange after 30 min. MVEC formed a homogenous contact-inhibited monolayer with a typical cobblestone cell shape (
Fig. 1
A). Specific endothelial staining with Dil-Ac-LDL could be demonstrated for more than 95% of the cells (
Fig. 1B). Only a few cells showed a positive staining for actin, demonstrating a contamination with fibroblasts or smooth muscle cells (
Fig. 1C).
View larger version (148K):
[in this window]
[in a new window]
|
Fig. 1. ECs from the microvasculature of rat hearts. The cells were cultured on Lab-Tek chamber slides with M199, 10% FCS, and ECGF. Typical morphology was demonstrated with phase contrast (A) and specific uptake of Dil-Ac-LDL with fluorescence microscopy (B). Only a few cells showed a positive staining for actin (C).
|
|
Transfection efficiency
The transfection efficiency as monitored with a semiquantitative ELISA assay depends on the concentration of the LPC and the ratio of both substances. The best results were obtained for a 2:1 liposome–plasmid mixture (
Fig. 2
Fig. 3
A). The transfection could be enhanced twice with transferrin (
Fig. 2). Increasing the LPC concentration enhances the transfection rate (
Fig. 3A) and the cytotoxicity leading to cell death as monitored with the Alamar blue assay (
Fig. 3B). The optimal transfection efficiency was found at the highest possible transfection rate and the lowest possible cytotoxicity. At this concentration nearly 60% of the cells showed positive staining with X-Gal (
Fig. 4
).
View larger version (35K):
[in this window]
[in a new window]
|
Fig. 2. Results of one representative ONPG ELISA (n=5) for transfection efficiency. Relative expression levels depended on the Lipofectin–plasmid ratios and LPCs. Transferrin is able to enhance transfection.
|
|
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 3. Comparison of transfection efficiency (ONPG ELISA, A) and cytotoxicity (Alamar blue ELISA, B) of gene transfer in vitro (one representative experiment of five). (A) Highest expression levels were obtained for a 2:1 Lipofectin–pCMVß ratio and reached a maximum at a 1:1 dilution. A further increase could not be detected because of an increase in cytotoxicity (B). Cell viability was significantly suppressed above a 1:2 dilution of the complex and depends upon the Lipofectin–pCMVß ratio.
|
|
View larger version (153K):
[in this window]
[in a new window]
|
Fig. 4. Transfection of confluent MVEC layers. After incubation of the cells with the optimal liposome–pCMVß concentration and transferrin, 60% of the cells are showing a positive staining with X-Gal. Different color intensity reflects different expression levels.
|
|
Cytotoxicity of transfection solutions in cell culture
After 4 h of cold incubation the lowest cytotoxicity and highest proliferation rate was observed for normal cell culture medium and for the Buckberg cardioplegic solution. The highest cytotoxicity was observed after incubation with St. Thomas solution (
Fig. 5 A).
View larger version (39K):
[in this window]
[in a new window]
|
Fig. 5. Cytotoxic effects of the transfection procedure. (A) ECs were incubated at 4°C with different media and cardioplegic solutions. The proliferation correlates with the change in optic density (OD) in the Alamar blue ELISA. The lowest cytotoxicity provides the normal cell culture medium M199 with 10% FCS and the blood cardioplegic Buckberg solution containing 10% FCS (* P<0.05 vs. PBS). (B) Relevant transfection solutions were tested for their cardioplegic property in a Langendorff perfusion system. Coronary blood flow was measured over time after a cold ischemic time at 4°C of 4 h. Compared with the control, best flow rates were detected after cardioplegia with St. Thomas and Buckberg solution.
|
|
Cytotoxicity of transfection solutions during cardioplegia
After cold ischemia of 4 h, all hearts were beating at the end of the experiment (130 min). Within the first 30 min of reperfusion the coronary blood flow rapidly declined from 12 to 5–10 ml/min. The highest and most enduring flow rates were found after cardioplegia with St. Thomas or Buckberg solution (7 ml/min after 120 min). Normal cell culture medium was not useful for cardioplegia of the rat heart (3–5 ml/min flow after 120 min) (
Fig. 5B).
Gene transfer in vitro with cardioplegic solutions
A gene transfer with liposomes at 4°C is possible, but it only succeeds in 60% of the better levels obtained at 37°C. Transferrin is able to enhance gene transfer at both temperatures. Comparable gene expression levels were measured for normal cell culture medium and the Buckberg cardioplegic solution (
Fig. 6 ).
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 6. Transfection efficiency at 4/37°C monitored with the ONPG ELISA. Lipofectin and pCMVß (2:1) and transferrin were added to ECs in different solutions at different temperatures. An enhancement of the transfection with transferrin could be detected. The cardioplegic Buckberg solution produces comparable gene expression levels to normal cell culture medium.
|
|
|
Discussion
|
|---|
Despite many investigations, ECs are difficult to transfect culturally. After transfection of human umbilical vein endothelial cells (HUVEC) with liposomes, 70% of the cells are viable with 0% of the cells transfected. Of the HUVECs, 50% could be transfected with a 19% viability by electroporation
[8]. Adenoviral vectors allowed for a much higher transfer rate (90–100%)
[9]. Compared with current studies our data demonstrate a higher transfer rate into ECs using a non-viral viral vector system (about 60%). This could be achieved by testing different concentrations of the plasmid, liposome and transferrin in different transfection solutions. To introduce a cytokine or hormone into a solid organ, a transfection of 100% of the target cells does not seem to be necessary to achieve a detectable effect. In vivo transfection into uninjured ECs has its limitation in continuous blood flow and is therefore lacking in incubation. In vivo, adenoviral vectors are able to transfer the entire endothelium of sheep carotid arteries
[10]. The transfection efficiency of adenoviral vectors and liposomes was compared in porcine coronary arteries. The gene levels obtained with the Ad5 vector was more than two orders of magnitude higher than obtained with Lipofectin. In a heterotopic heart transplantation model in the mouse, an effective gene transfer during cold ischemia could be achieved with liposomes. Gene expression could be detected in ECs, some arterial smooth muscle cells, and some myocytes, but it was not quantified
[1]. In comparable models higher expression levels were obtained with adenoviral vectors and therapeutic effects were detectable
[4].
Our results for a liposome-mediated gene transfer demonstrate a much higher transfer rate in ECs. This could be obtained by optimizing the transfection efficiency with different liposome and plasmid concentrations and the use of transferrin. An optimal dosage was found by looking for the maximal transfer rate with an acceptable cytotoxicity. One disadvantage of the used preparation technique for the ECs was that the cell cultures contain ECs from epicardial coronary arteries and MVECs. The contamination rate with non-endothelial cells was very low.
The transfection solution for solid organ transplantation plays another important role. For clinical use a good cardiac conservation for 4 h at 4°C is mandatory. Normal transfection solutions (cell culture media like M199) are not useful for cardioplegia and in turn cardioplegic solutions (St. Thomas) are not useful for gene transfer. In contrast the Buckberg solution provides for both properties.
It remains to be proven whether transfection of solid organs is as feasible as it is in cell cultures with this gene transfer protocol. If comparable levels of gene expression can be found this might result in a powerful tool in transplant medicine. Patients might be less prone to chronic rejection or the rejection problem as a whole. Furthermore, with our method, the disadvantages of adenoviral expression systems would be excluded. One therapy for the treatment of chronic rejection might be the gene transfer of the constitutive nitric oxide synthase, using the antiproliferative effect of NO to vascular smooth muscle cells.
|
Footnotes
|
|---|
Presented at the 11th Annual Meeting of the European Association for Cardio-thoracic Surgery, Copenhagen, Denmark, 28 September–1 October, 1997. 
|
Appendix A. Conference discussion
|
|---|
Dr J. Vaage (Stockholm, Sweden): I think this, and also the previous paper, clearly shows that molecular biology will, within the next few years, be a part of cardiac surgery. And maybe we’ll end up doing molecular surgery. Before declamping we’ll just fill up the heart with a solution for gene transfer and change some of the genes of the heart. I just wanted to ask you, then, what you know about contamination by other cells in your EC culture. Did you have a pure EC culture?
Dr M. Hein: More than 95% of the culture are ECs because of negative staining for fibroblasts with -actin.
Dr J. Vaage: One of the problems with a gene transfection is that it’s temporary. How long did your cells have the ability?
Dr M. Hein: I didn’t look at the duration of gene expression in my experiments. What is known from other studies is that you would expect a transfection time of 3–4 weeks at maximum using liposomes.
|
References
|
|---|
- Ardehali A, Fyfe A, Laks H, Drinkwater DC, Qiao JH, Lusis AJ Direct gene transfer into donor hearts at the time of harvest. J Thorac Cardiovasc Surg 1995;109:716-720.[Abstract/Free Full Text]
- Brauner R, Wu L, Laks H, Nonoyama M, Scholl F, Shvarts O, Berk A, Drinkwater DC, Wang JL Intracoronary gene transfer of immunosuppressive cytokines to cardiac allografts: method and efficacy of adenovirus-mediated transduction. J Thorac Cardiovasc Surg 1997;113(6):1059-1066.[Abstract/Free Full Text]
- Sawa Y., Kadoba K, Bai HZ, Kaneda Y, Shirakura R, Matsuda H Efficient gene transfer method into the whole heart through the coronary artery with hemagglutinating virus of Japan liposome. J Thorac Cardiovasc Surg 1997;113(3):512-518.[Abstract/Free Full Text]
- Suzuki J, Isobe M, Morishita R, Aoki M, Horie S, Okubo Y, Kaneda Y, Sawa Y, Matsuda H, Ogihara T, Sekiguchi M Prevention of graft coronary arteriosclerosis by antisense cdk2 kinase oligonucleotide. Nat Med 1997;3(8):900-903.[Medline]
- Crystal RG Transfer of genes to humans: early lessons and obstacles to success. Science 1995;270:404-410.[Abstract/Free Full Text]
- Nabel EG Gene therapy for cardiovascular disease. Circulation 1995;91(2):541-548.[Free Full Text]
- Cheng PW Receptor ligand-facilitated gene transfer: enhancement of liposome-mediated gene transfer and expression by transferrin. Hum Gene Ther 1996;7:275-282.[Medline]
- Nathwani AC, Gale KM, Pemberton KD, Crossmann DC, Tuddenham EGD, McVey JH Efficient gene transfer into human umbilical vein endothelial cells allows functional analysis of the human tissue factor gene promoter. Br J Haematol 1994;88:122-128.[Medline]
- Chen SJ, Wilson JM, Muller DWM Adenovirus-mediated gene transfer of soluble vascular cell adhesion molecule to porcine interposition vein grafts. Circulation 1994;89(5):1922-1928.[Abstract/Free Full Text]
- Lemarchand P, Jones M, Yamada I, Crystal RG In vivo gene transfer and expression in normal uninjured blood vessels using replication-deficient recombinant adenovirus vectors. Circ Res 1993;72(5):1132-1138.[Abstract/Free Full Text]
Top
Abstract
Introduction
