Functional Analysis of Genomic DNA, cDNA, and Nucleotide Sequence of the Mature C-Type Natriuretic Peptide Gene in Vascular Cells
Objective— The aim of this study was to investigate the effect of various C-type natriuretic peptide (CNP) sequences (genomic DNA [CNPDNA], cDNA derived from mRNA [CNPcDNA], and sequence coding for 22 amino acids of the mature CNP [CNP22aa]) on the growth of primary porcine vascular cells.
Methods and Results— Gene transfer was performed with cationic lipid DOCSPER or linear polyethylenimine. All 3 CNP sequences led to significant inhibition of smooth muscle cell (SMC) proliferation. In contrast, significant stimulation of cell growth was observed in endothelial cells (ECs) using CNPDNA or CNPcDNA but not CNP22aa. In a porcine restenosis model, a significant reduction in neointima hyperplasia was found 3 months after application of the CNPcDNA vector compared with the control transfection.
Conclusions— The results demonstrate that the first intron in the CNP sequence does not contain any additional enhancer-binding sites. However, the signal sequence is indispensable for secretion of CNP and its appropriate physiological function. Furthermore, the results show for the first time the therapeutic effect of CNP using liposome-mediated gene transfer and local adventitial delivery. Advantages of the CNP gene are its dual effects with inhibition of SMC proliferation and simultaneous promotion of EC growth.
C-type natriuretic peptide (CNP) is a promising therapeutic agent for prevention of restenosis after angioplasty.1–3 CNP inhibits smooth muscle cell (SMC) proliferation, decreases extracellular matrix synthesis, promotes revascularization, and positively influences vascular remodeling.2,3
The CNP gene has been localized to chromosome 2 in the human genome, and its sequence is the most highly conserved of the natriuretic peptides across many species.4,5 Sequence analysis reveals that there are a total of 3 exons and 2 introns. In the coding region for prepro CNP, the first 2 exons are separated by an intron.6 Exon 3, coding for poly A, is removed during the processing pathway within the nucleus. The gene is flanked in the 5′-region by cis-acting regulatory elements that are not part of the other natriuretic peptide genes, suggesting that the regulation mechanisms for CNP gene transcription are different compared with the other natriuretic peptides.6 The function of the first intron regarding transcription regulation of this gene is so far unknown. However, a number of genes are known to be regulated by enhancer elements outside the coding region, and many of them contain regulatory sequences within the first intron.7,8 The CNP gene encodes a 126-residue precursor peptide that is processed to generate 22- and 53-aa peptides (human CNP22 and human CNP53, respectively).9 CNP22 is more widely and abundantly expressed and is more potent than CNP53.5,9 The amino terminal of prepro CNP contains a signal peptide sequence that allows the synthesized peptide to pass from the endoplasmic reticulum to the Golgi complex, where it is packed into granules from which CNP is then released at the cell surface by exocytosis. The signal peptide sequence is generally necessary for protein secretion. However, some proteins, such as fibroblast growth factor-2 (FGF-2), are secreted into the extracellular medium despite lacking any signal sequence by a still unknown mechanism.10 For CNP, the necessity of the signal peptide sequence was not yet investigated.
Therefore, we constructed 3 DNA vectors containing different CNP sequences (genomic DNA including 2 exons and the first intron, cDNA derived from mRNA, and a construct containing only the nucleotide sequence of the mature CNP with 22 amino acids) and performed functional analyses of these sequences regarding their gene expression, growth effect, and toxicity in primary vascular cells in vitro. The most suitable vector for the CNP gene was then used for prevention of restenosis formation in a porcine restenosis model.
A detailed Methods section is available online at http://atvb.ahajournals.org.
Plasmid Vectors and Cloning Strategies
The vectors pTRACER-CMV2 (pTr; Clontech), pCR3.1 (Invitrogen), and pSEAP2-Control (Clontech) were used. Three different CNP vectors (pTrCNPDNA, pTrCNPcDNA, and pTrCNP22aa) were constructed (Figure 1). The whole CNPDNA was gained either from human umbilical cord vein or from porcine endothelial cells (ECs) and amplified using primers CNP-L1 and CNP-R1 (Table). CNPcDNA and CNP22aa were synthesized by Medigenomix and amplified using primers CNP-L2/CNP-R2 and CNP-L3/CNP-R3, respectively (Table). All 3 CNP sequences were first subcloned into the pCR3.1 vector, digested with KpnI/XbaI, and subcloned into pTr, building the desired vectors pTrCNPDNA, pTrCNPcDNA, and pTrCNP22aa. The control vector pTrSEAP was digested with HindIII/SalI, and the SEAP fragment was subcloned into HindIII/XhoI of pCR3.1, cut with HindIII/XbaI, and integrated into EcoRI/XbaI of the pTr vector, resulting in pTrSEAP.
For in vitro experiments, human urinary bladder carcinoma cell line ECV-304 (DSMZ Nr. ACC 310) from the umbilical cord vein, primary porcine SMCs, and ECs were used and cultivated as described previously.12–14
In Vitro Gene Transfer
mRNA Analysis of the Transfected Genes and NP Receptors In Vitro and In Vivo
The following primers were applied (Table): GAPDH-L1/GAPDH-R for the GAPDH gene (450 bp); GFP-L/GFP-R for the GFP gene (450 bp), CNP-L4/CNP-R3 for the CNP gene (69 bp); NPRB-L/NPRB-R for the NPR-B gene (293 bp); NPRC-L/NPRC-R for the NPR-C gene (492 bp); and βGal-L/βGal-R for the β-galactosidase gene (786 bp).
RT-PCR Product Quantification
mRNA expression quantification was determined by normalizing CNP and NPR amplification levels against GAPDH. Expression levels for CNP, NPR-B, and NPR-C were calculated as a percentage of nontreated cells using Brilliant SYBR Green QPCR master mix (Stratagene).
CNP Protein Expression
CNP protein concentration in supernatant and cell lysate was determined using a special CNP immunoassay (Bachem).
Cell Proliferation and Cell Toxicity Inhibition
Proliferation rate was determined using the 5-bromodeoxyuridine method (Roche) as described previously.13 For cytotoxicity evaluation, the CytoTox-One method was used (Promega). cANF (Bachem) was used in concentration of 10−7−10−6 M.16
Balloon-Injured Porcine Peripheral Artery Gene Transfer In Vivo
Gene transfer to a pig restenosis model was performed in Pietrain pigs (n=6) using CNPcDNA vector and the needle injection catheter as described previously.14,17,18 Treated vessels were explanted 3 months after treatment and analyzed as described.2,12–15,17,18
Transferred Plasmid DNA Analysis
Plasmid DNA was detected by PCR amplifying a partial sequence of the plasmids used: pCMVβ-L/pCMVβ-R (647 bp), pTrCMV2-L/CNP-R3 (460 bp), and GAPDH-L2/GAPDH-R (550 bp; Table). All PCR reactions were performed using the following conditions: 45 s 94°C, 1 minute 60°C, and 45 s at 72°C for 35 cycles.
All values are expressed as means±SE. One-way ANOVA was used for comparison of differences among the groups (SPSS). Statistical significance was defined as P<0.05.
Various CNP Sequence Gene Expression
The vectors containing various subcloned CNP sequences (CNPDNA, CNPcDNA, and CNP22aa) were first transfected into SMCs and ECs. For CNPDNA, human and porcine DNA sequence was used for transfection. No differences were seen between these 2 sequences regarding biological effect. Therefore, in view of possible future applications in patients, the human CNP sequence was tested in subsequent experiments. For quantitative analysis of expressed GFP and CNP genes, both comprised in the CNP vectors, semiquantitative RT-PCR was performed (Figure I, available online at http://atvb.ahajournals.org). Results demonstrated no significant difference in GFP expression or the various CNP sequences subcloned. Consequently, gene transfer was successful and independent of the vector size and CNP vector used.
CNP Expression Optimization in Vascular Cells
To achieve maximal gene transfer efficiency in vascular cells, first, optimization experiments were performed with cationic lipid DOCSPER or polycationic molecule PEI22. For SMCs, the most efficient gene transfer was observed using 1.2 μg cationic lipid DOCSPER and 0.2 μg plasmid-DNA per well (weight ratio 1/6), whereas for ECs, maximal transfer conditions were found using 0.078 μg PEI22 and 0.1 μg plasmid DNA per well (N/P ratio 1/6). Therefore, for all further gene transfer experiments in vitro, SMCs were transfected with DOCSPER, whereas ECs were transfected with PEI22 with the described adjuvant concentrations.
To determine maximal expression of the transfected plasmid-DNA over time, accumulation and protein turnover of SEAP and CNP in supernatant of the transfected vascular cells were analyzed. Protein accumulation was determined at days 1, 2, 3, and 4, and protein turnover for every 24 hours up to 4 days (Figure II, available online at http://atvb.ahajournals.org). For SEAP, gene product accumulation increased continuously, whereas maximal protein turnover was achieved 3 days after gene transfer (Figure IIa and IIb). For CNPcDNA, CNP peptide accumulation in supernatant was maximal at day 3 after transfection, and maximal protein turnover was observed between days 2 and 3 (Figure IIc and IId). Over time, these expression patterns were observed for SMCs and ECs.
CNP Concentration Analysis
Maximal protein expression of transfected DNA was detected 3 days after transfection (Figure II). Therefore, concentration of CNP peptide using the different CNP nucleic acid sequences was determined 3 days after gene transfer in supernatant and in cell lysate of the transfected vascular cells (Figure 2). CNPDNA and CNPcDNA demonstrated significantly higher concentrations of CNP in supernatant of the transfected SMCs and ECs compared with nontreated cells or cells treated with the control gene for SEAP. In contrast, transfection with CNP22aa did not significantly enhance CNP concentration in the supernatant of the transfected cells compared with control transfection (SEAP). In cell lysate of the transfected SMCs, CNP concentration enhancement was only found using the CNP-22aa vector (Figure 2). For ECs, an increase of the CNP concentration in cell lysate was observed using all 3 CNP vector constructs compared with the control transfection with SEAP.
CNP Receptor Quantitative Analysis
Quantitative RT-PCR analysis of NPR-B and NPR-C receptors was performed before and 1, 2, and 3 days after gene transfer in SMCs and ECs (Figure 3). Expression of mRNA of both receptors and corresponding GAPDH was detected in all samples analyzed (Figure 3a). The expression level of these receptors before and 1 day after transfection is demonstrated in Figure 3b. No significant differences in NPR-B and NPR-C mRNA expression were found before and after gene transfer in both cell types independent of the vector construct used (Figure 3b). Furthermore, no significant changes in expression were seen between 1 and 3 days after transfection (data not shown). Expression levels of NPR-B were observed to be similar for both cell types. The same expression level of NPR-C was also found in ECs. In contrast, NPR-C mRNA expression in SMCs was ≈10-fold lower compared with expression of this receptor in ECs and with NPR-B expression in both cell types.
Cell Proliferation and Cytotoxicity
Optimal gene transfer conditions for maximal protein expression were applied as described in the previous paragraphs (Figure II). The effect of various CNP sequences on growth of vascular cells in vitro under these conditions was analyzed 2 and 3 days after gene transfer (Figure 4). In SMCs, all 3 CNP sequences demonstrated significant proliferation rate reduction (Figure 4a). Inhibition of cell proliferation 2 days after transfection was 0.76±0.11 for CNPDNA, 0.79±0.13 for CNPcDNA, and 0.85±0.10 for CNP22aa. Three days after transfection, no additional significant changes were seen. For ECs, increase in proliferation rate was observed after gene transfer with the vectors for CNPDNA or CNPcDNA (Figure 4a). The greatest increase of proliferation was found 2 days after transfection (1.11±0.03 for CNPDNA and 1.12±0.02 for CNPcDNA). Gene transfer using the CNP22aa vector did not lead to cell growth enhancement. In contrast, a slight but nonsignificant reduction of EC proliferation was observed (Figure 4a).
Experiments were performed as described above as well as in the presence of cANF, a specific ligand for NPR-C. No changes were observed in the proliferation rate 2 and 3 days after gene transfer compared with vascular cells without cANF treatment (Figure 4b). These results were found for SMCs and ECs.
For determination of toxicity of the various CNP sequences after gene transfer, lactate dehydrogenase release was assayed (Figure 4c). In SMCs, CNPDNA and CNPcDNA demonstrated no significant differences in cellular toxicity compared with the control vector for SEAP. The overall toxic rate was 0.07±0.01 except for the CNP22aa construct (0.11±0.01; Figure 4c). In ECs transfected with CNPDNA or CNPcDNA, toxicity increased continuously over time, with 0.09±0.01 2 days after transfection and 0.14±0.2 after 3 days. Using CNP-22aa, significant higher cellular toxicity was observed compared with the other CNP vector constructs used with 0.14±0.2 2 days after transfection and 0.17±0.3 after 3 days.
In Vivo Gene Transfer in a Porcine Restenotic Model
For analysis of the therapeutic effect of CNP in vivo, a porcine restenotic model involving local delivery into adventitial tissue of peripheral arteries and nonviral gene transfer was applied. Because there was no significant difference in gene transfer efficiency in vitro between CNPDNA and CNPcDNA vectors, the smaller gene construct containing CNPcDNA complexed with cationic lipid DOCSPER was used. For detection of the transgene DNA and expression of the transfected gene, PCR and RT-PCR were performed 3 months after gene transfer (Figure III, available online at http://atvb.ahajournals.org). The transgene DNA and transgene-specific expression were detected in tissue surrounding the treated arteries (Figure III). In accordance with maximum exogenous DNA found in analyzed segments, the highest expression was detected in the same segments for both transfected genes, the control gene for β-galactosidase (Figure III, left) and the therapeutic gene for CNP (Figure III, right). Furthermore, as demonstrated in Figure 5, 3 months after application, neointimal formation was suppressed significantly in arteries transfected with the CNP gene compared with the control side transfected with the β-galactosidase vector (Figure 5a and 5b). Stenosis reduction after application of the CNP gene determining the intima/media ratio was 7.69±1.78 compared with the control artery transfected with the gene for β-galactosidase (Figure 5d).
In the present study, functional analysis of different CNP nucleic acid sequences (Figure 1) was performed in vascular cells using nonviral gene transfer. Somatic gene delivery with either cationic lipid or polycationic molecules have enjoyed increasing interest in recent years.14,15,17–23 Therefore, efficient agents of both groups, the cationic lipid DOCSPER and the cationic polymer PEI, were tested in porcine primary vascular cells. DOCSPER was shown to have high transfer efficiencies in SMCs under optimized conditions in vitro and in vivo.14 However, in ECs, significantly lower gene transfer efficiency was achieved compared with SMCs.13 The cationic molecule PEI is known for great potential to transfect different cell types efficiently.20,21 Results demonstrated that in SMCs, DOCSPER allowed for higher gene transfer efficiencies than PEI22. In contrast, in ECs, PEI22 was more efficient than DOCSPER. Discrepancies in transfer efficiency for DOCSPER versus PEI can be explained by different behavior of cationic lipids or polycationic molecules in DNA complexes after cellular uptake and releasing mechanisms within the cell as reviewed previously and was not the aim of our study.21,24
Results demonstrated no significant differences in expression of the various CNP sequences subcloned, independent of the CNP vector construct used. These results confirmed our previous data demonstrating that vector size does not significantly influence transfection using liposome-mediated gene transfer.14 In addition, all 3 CNP sequences were expressed successfully in vascular cells. After gene transfer of vectors for CNPDNA and CNPcDNA in SMCs and ECs, a significant increase of the CNP concentration was found in supernatant and cell lysate. However, there was no significant difference in CNP concentration using these 2 CNP sequences. These results support the notion that there are no additional transcription regulatory elements in the first intron of the CNP gene. In contrast, using the CNP22aa sequence, an enhanced CNP concentration was found in cell lysate but not in the supernatant. Therefore, there is no active secretion of this mature peptide by the cell. These results demonstrate the necessity of a signal sequence in the CNP gene for peptide secretion and proper biological function. The active removal of CNP from growth medium was further confirmed by comparing the absolute concentration of CNP peptide in the supernatant with the amount of SEAP protein in the control transfection, where the CNP concentration was significantly lower (43-fold for SMCs and 21-fold for ECs). SEAP in contrast to CNP is a stable protein with no active degradation. This is an important finding, indicating that CNP gene expression for longer periods of time is necessary to achieve therapeutic effects, thus, in the case of CNP, privileging the use of gene transfer over the use of CNP peptide administration.
According to the results found for maximal expression of the transgene DNA, maximal inhibition of SMC proliferation after gene transfer with different CNP vectors was found 2 and 3 days after transfection. Interestingly, all 3 various CNP sequences demonstrated cell proliferation inhibition (0.76±0.11 for CNPDNA, 0.79±0.13 for CNPcDNA, and 0.85±0.10 for CNP22aa). SMC growth inhibition after transfection with the CNP22aa construct contradicted the results demonstrating that no enhanced concentration of CNP was found in the supernatant of transfected cells. The reason for the inhibition of cell proliferation is likely because of a higher cytotoxicity of the short mature CNP vector construct. Toxicity is probably caused by the high concentration of CNP in the cytoplasm or cell vesicles lacking the signal peptide sequence, thus without possibility to be secreted. CNPDNA and CNPcDNA demonstrated comparable results regarding SMC growth inhibition. These data further support the hypothesis that the intron sequence is not necessary for optimal expression of the CNP gene. In contrast to SMCs, in ECs, both CNPDNA and CNPcDNA increased cell proliferation after gene transfer (1.11±0.03 for CNPDNA and 1.12±0.02 for CNPcDNA). However, transfection with the CNP22aa vector in ECs resulted in a slight reduction in cell proliferation. These data are in accordance with the results achieved in SMCs, confirming that the enhanced cytotoxicity after gene transfer with the CNP-22aa construct is responsible for cell growth inhibition as demonstrated by lactate dehydrogenase release. Similar results as described above were observed also in the presence of cANF, the specific ligand for NPR-C that inhibits this clearance receptor. This can be explained by the high concentration of CNP in the medium, which abounded to maximally activate the cGMP signaling pathway through NPR-B. However, because of removal of CNP by NPR-C, constant supply of CNP peptide is necessary for long-lasting biological effects.
The reason for the different responses of CNP on proliferation of SMCs and ECs is not yet understood, but several mechanisms may be involved. First, there is differential expression of NP receptors on SMCs and ECs.25,26 This was confirmed by our analysis of NPR-B and NPR-C expression in these cells. The expression of NPR-B was found to be the same in SMCs and ECs. However, NPR-C expression was markedly lower in SMCs compared with ECs. These variations in NP receptor expression may lead to discrepancies in the measured growth inhibition or stimulation of vascular cells after CNP application. Although the NPR-C receptor has apparently no intrinsic cyclase activity, there are some studies indicating the role of this receptor via signal transduction by cAMP.4 Thus, the enhanced expression of NPR-C on ECs compared with SMCs may contribute to the differences in response to CNP, resulting in differential growth of these cells. Second, the molecular mechanism involved in the antiproliferative effects of NPs in SMCs include inhibition of different growth-stimulating factors such as extracellular signal-regulated kinase, c-Jun N-terminal kinase, and activator protein-1 binding sites.27 These effects are reproduced by cGMP-signaling pathway. In ECs, there may also be other mechanisms or signaling pathways via cGMP, thus promoting proliferation of these cells. Third, in ECs, there may also be an indirect effect via CNP, where other growth factors such as FGF-2 or PDGF are stimulated, thus enhancing mitogenic effect of CNP on ECs.28 In addition, CNP enhances NO production in SMCs and augments expression of CNP receptors in the injured artery, which also contributes to the highly site-specific CNP effects.
The in vivo experiments in a porcine restenosis model demonstrated for the first time the therapeutic effect of CNP using nonviral gene transfer system and local adventitial delivery. Despite relatively low efficiency in vivo using liposome-mediated gene transfer, significant reduction of neointima hyperplasia was observed using the CNPcDNA gene compared with the control transfection with the gene for β-galactosidase. Reasons for this successful result are the bystander effect of the secreted CNP and its dual mechanism of inhibiting SMC proliferation and, in parallel, accelerating endothelial regeneration.1,3 Accelerated re-endothelialization may also help to prevent subacute thrombosis, which is often seen after neointima suppression using drug-eluting stents.29,30
In summary, these experiments demonstrate for the first time that the CNP gene can significantly inhibit restenosis formation after angioplasty using local adventitial delivery and nonviral gene transfer. Furthermore, the functional analysis of 3 different CNP nucleic acid sequences revealed that the signal peptide sequence in the CNP gene is necessary to secrete the CNP peptide to achieve a therapeutic effect. Moreover, data demonstrate that the first intron in the CNP sequence does not contain any additional enhancer-binding sites. These findings may be helpful for developing more efficient and safe gene therapeutic strategies for restenosis prevention.
This work was funded by Ernst und Berta Grimmke-Stiftung, and we thank this foundation for the possibility to perform our scientific work. We also acknowledge the help of BERNINA Co. for providing the cationic lipid DOCSPER.
- Received May 4, 2004.
- Accepted May 28, 2004.
Doi K, Ikeda T, Itoh H, Ueyama K, Hosoda K, Ogawa Y, Yamashita J, Chun TH, Inoue M, Masatsugu K, Sawada N, Yamahara K, Kook H, Komeda M, Ueda M, Nakao K. C-type natriuretic peptide induces redifferentiation of vascular smooth muscle cells with accelerated re-endothelialization. Arterioscler Thromb Vasc Biol. 2001; 21: 930–936.
Morishige K, Shimokawa H, Yamawaki T, Miyata K, Eto Y, Kandabashi T, Yogo K, Higo T, Egashira K, Ueno H, Takeshita A. Local adenovirus-mediated transfer of C-type natriuretic peptide suppresses vascular remodeling in porcine coronary arteries in vivo. J Am Coll Cardiol. 2000; 35: 1040–1047.
Ohno N, Itoh H, Ikeda T, Ueyama K, Yamahara K, Doi K, Yamashita J, Inoue M, Masatsugu K, Sawada N, Fukunaga Y, Sakaguchi S, Sone M, Yurugi T, Kook H, Komeda M, Nakao K. Accelerated re-endothelialization with suppressed thrombogenic property and neointimal hyperplasia of rabbit jugular vein grafts by adenovirus-mediated gene transfer of C-type natriuretic peptide. Circulation. 2002; 105: 1623–1626.
Suzuki T, Yamazaki T, Yazaki Y. The role of the natriuretic peptides in the cardiovascular system. Cardiovasc Res. 2001; 51: 489–494.
Warnecke C, Willich T, Holzmeister J, Bottari SP, Fleck E, Regitz-Zagrosek V. Efficient transcription of the human angiotensin II type 2 receptor gene requires intronic sequence elements. Biochem J. 1999; 340: 17–24.
Kone BC. Molecular biology of natriuretic peptides and nitric oxide synthases. Cardiovasc Res. 2001; 51: 429–441.
Pelisek J, Engelmann MG, Golda A, Fuchs A, Armeanu S, Shimizu M, Mekkaoui C, Rolland PH, Nikol S. Optimization of nonviral transfection: variables influencing liposome-mediated gene transfer in proliferating vs. quiescent cells in culture and in vivo using a porcine restenosis model. J Mol Med. 2002; 80: 724–736.
Komatsu Y, Itoh H, Suga S, Ogawa Y, Hama N, Kishimoto I, Nakagawa O, Igaki T, Doi K, Yoshimasa T, Nakao K. Regulation of endothelial production of C-type natriuretic peptide in coculture with vascular smooth muscle cells. Role of the vascular natriuretic peptide system in vascular growth inhibition. Circ Res. 1996; 78: 606–614.
Farb A, Heller PF, Shroff S, Cheng L, Kolodgie FD, Carter AJ, Scott DS, Froehlich J, Virmani R. Pathological analysis of local delivery of paclitaxel via a polymer-coated stent. Circulation. 2001; 104: 473–479.
Liistro F, Colombo A. Late acute thrombosis after paclitaxel eluting stent implantation. Heart. 2001; 86: 262–264.