- Email: [email protected]
Rational promoter selection for gene transfer into cardiac cells
Journal of Molecular and Cellular Cardiology 35 (2003) 823–831 www.elsevier.com/locate/yjmcc
Original Article
Rational promoter selection for gene transfer into cardiac cells Alexander Maass a,b, Stephen J. Langer a, Silke Oberdorf-Maass a,b, Sebastian Bauer b,1, Ludwig Neyses b,2, Leslie A. Leinwand a,* a
Department of Molecular, Cellular, and Developmental Biology, University of Colorado, Campus Box 347, Boulder, CO 80309 0347, USA b Department of Medicine, University of Wuerzburg, Germany Received 26 November 2002; received in revised form 14 March 2003; accepted 3 April 2003
Abstract Cardiomyocytes (CMCs) are extremely difficult to transfect with non-viral techniques, but they are efficiently infected by adenoviruses. The most commonly used promoters to drive protein expression in cardiac myocytes are of viral origin, since they are believed to be constitutively active and minimally regulated by physiological or pharmacological challenge of cells. In recombinant adenoviruses, we systematically compared three different promoters: the cytomegalovirus (CMV), the Rous sarcoma virus (RSV), and a synthetic promoter with three MEF2 transcription factor-binding sites upstream of the heat-shock protein 68 minimal promoter. We determined their basal activity in primary cardiac cells as well as their possible stimulation by commonly used agonists. The CMV promoter was activated up to 60-fold by the phorbol ester phorbol myristate acetate (PMA) and/or forskolin in neonatal rat CMCs and cardiac fibroblasts. Primary adult rat CMCs had higher basal expression from the CMV promoter that was not activated by PMA or forskolin. The RSV promoter was less affected by agonists and was more active in cardiac myocytes compared to cardiac fibroblasts. The MEF2-responsive promoter showed high basal expression in both myocytes and fibroblasts, and minimal induction by phorbol esters and forskolin. The relevance of reporter gene induction was confirmed with a contractile protein, troponin T (TnT). The CMV promoter driving TnT could be induced more than 15-fold with phenylephrine or forskolin to replace the endogenous protein almost to completion at a multiplicity of infection of 10. These results suggest the following use of the tested promoters: an inducible system (CMV), a myocyte-enriched system (RSV), or a stable control system (MEF2). © 2003 Elsevier Science Ltd. All rights reserved. Keywords: Cytomegalovirus; Rous sarcoma virus; MEF2; Transcription factor; Promoter; Phorbol ester; Forskolin; Gene transfer; Cardiomyocytes; Fibroblasts
1. Introduction Cardiomyocytes (CMCs) are generally considered to be one of the most difficult cell types to transfect with DNA constructs. Adenovirus vectors have facilitated gene transfer into both adult and neonatal rat CMCs in vitro as well as neonatal and adult myocardium in vivo [1–5]. To drive highlevel expression of proteins of interest in cardiovascular cells, most investigators use viral promoters in recombinant adenovirus vectors. Previous experiments have shown that the cytomegalovirus (CMV) promoter can be activated up to 23-fold by forskolin and phorbol ester in human vascular * Corresponding author. Tel.: +1-303-492-7606; fax: +1-303-492-8907. E-mail address: [email protected] (L.A. Leinwand). 1
Present address: Department of Medicine, West German Cancer Center, University of Essen Medical School, Germany. 2 Present address: Department of Medicine, University of Manchester, Manchester, UK.
© 2003 Elsevier Science Ltd. All rights reserved. DOI: 10.1016/S0022-2828(03)00140-8
smooth muscle cells [6]. Regulation of other viral promoters has been less extensively studied in cardiac myocytes. Different viral promoters have never been compared directly in isolated cardiac cells with regard to (1) their specificity for myocyte expression vs. cardiac fibroblasts, (2) the amount of protein produced, and (3) their regulation by physiological or pharmacological intervention. We constructed adenovirus vectors with either the luciferase or b-galactosidase cDNA, driven by three different promoters, while maintaining the termination and polyadenylation signals constant (Fig. 1). The first is the 800 bp CMV immediate-early promoter, the second is the Rous sarcoma virus (RSV) long-terminal repeat (LTR), both extensively used for gene transfer experiments [7]. The third vector contains an artificial promoter consisting of a heatshock protein 68 (hsp68) minimal promoter and three binding sites from the desmin gene for the transcription factor MEF2. MEF2 is a transcription factor that is expressed primarily in developing muscle and brain tissue. A transgenic
824
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
2.3. Recombinant adenoviruses
Fig. 1. DNA constructs used to generate recombinant adenoviruses for this study. Boxes at the start and the end represent the inverted terminal repeats; W: packaging signal; CMV: cytomegalovirus immediate-early promoter; RSV: Rous sarcoma virus long-terminal repeat; MEF2 × 3: binding sites for the transcription factor MEF2 from the murine desmin promoter in triplicate; hsp68: minimal promoter from the heat-shock protein 68 gene; SV40 pA: polyadenylation and termination signals from the simian virus 40 genome. Figure is not to scale.
mouse expressing this construct showed expression in embryonic muscle and brain tissues [8]. In the adult, MEF2 activity was very low in the heart and increased with pathologic conditions such as hypertrophy associated with overexpression of a constitutive calmodulin-dependent protein kinase [9]. Our study demonstrates differential basal activity and stimulation of the tested promoters that should lead to differential use of these promoters depending on the experimental design.
2. Materials and methods 2.1. Isolation and culture of neonatal rat CMCs and cardiac fibroblasts Hearts from several litters of 1–3-d-old Wistar rats were excised and cells were isolated by trypsin digestion (0.1%) as previously described [10]. To enrich for CMCs, the cell suspension was preplated on regular polystyrene dishes for 30–45 min. CMCs were cultured in serum-free MEM supplemented with transferrin (1 mg/ml), insulin (1 µg/ml), and bovine serum albumin (BSA) (1 mg/ml) for 24–48 h before experiments were carried out. Dishes from preplating were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), passaged to enrich for the faster dividing fibroblasts from other non-myocytes, and used for experiments from passages 3 to 5.
Ad-CMV-LacZ and Ad-RSV-LacZ were generous gifts from Dr. Jerome Schaack (UCHSC-Denver) and are firstgeneration adenovirus vectors constructed using the Addl327 background. Ad-CMV-LacZ contains the 1.1 kb immediate-early promoter from human CMV driving the expression of the Escherichia coli b-galactosidase gene (LacZ). Ad-RSV-LacZ differs from Ad-CMV-LacZ by substitution of the RSV LTR for the CMV promoter. Ad-CMVLuc was derived by subcloning the BglII-BamHI fragment containing the luciferase gene from pJD205 into the polylinker of pACCMV.LpA [12,13]. A SalI fragment containing three head-to-tail copies of the desmin MEF2binding site fused to the hsp68 promoter and driving the expression of the b-galactosidase gene was subcloned into a left-end adenovirus plasmid construct, containing the adenovirus serotype 5 sequence from 1358 bp (SacII) and from 3330 (BglII) to 5700 bp (XhoI), within a polylinker inserted between the SacII and BglII sites to generate pAd-MEF2 × 3hsp68LacZ [8]. Methods for generating recombinant adenovirus LacZ expression vectors have been described previously [14]. 2.4. Amplification and purification of recombinant adenoviruses All the viral stocks were amplified from plaque-purified viral stocks. Thirty-six 100-mm dishes were seeded at 5 × 106 cells/dish and infected at a multiplicity of infection (MOI) of 25 plaque-forming units (pfu)/cell. After 48 h, virus was concentrated by low-speed centrifugation to pellet infected cells, resuspended in 20 ml of growth media, and then freeze– thawed for four cycles to release progeny virus. The freeze–thaw cell lysate was clarified by low-speed centrifugation and virus particles purified by banding on CsCl buoyant density gradients. The purified virus was diluted 5-fold in virus storage buffer (250 mM NaCl, 100 mM Tris (pH 7.5), 50% glycerol, 1 mg/ml BSA, 1 mM MgCl2) and stored at –20 °C. Virus concentration was determined by plaquetitering and b-galactosidase expressing plaques detected by addition of X-Gal as previously described [14]. The predicted genomic structure was verified by restriction analysis of isolated viral genomic DNA. Particle/pfu ratios were routinely around 50. No attempt was made to recover replication-competent adenovirus, although experiments were not conducted in human cells where this might have presented a problem. 2.5. Luciferase and b-galactosidase assays
2.2. Isolation and culture of adult rat CMCs The hearts of 3-month-old male Sprague–Dawley rats (250–300 g) were excised and CMCs were isolated by perfusion with collagenase as previously described [11]. Cells were cultured on laminin-coated plates in serum-free DMEM supplemented with BSA (1 mg/ml) and insulin (1 µg/ml).
After the indicated incubation period, cells were washed twice with phosphate-buffered saline (PBS) and lysed in 1× passive lysis buffer (Promega, Madison, WI). Protein concentrations were measured by the modified Bradford method (Bio-Rad, Hercules, CA) and luciferase activity determined after mixing with 100 µl assay reagent (Promega, Madison,
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
825
WI) in a standard luminometer (Turner Designs, Sunnyvale, CA). Relative light units (RLU) were normalized to microgram protein content in the sample. Cells infected with b-galactosidase containing constructs were washed twice with PBS, fixed for 10 min in 2% formaldehyde and stained in potassium ferricyanide/potassium ferrocyanide and BCIG overnight. b-Galactosidase-positive cells were counted by phase-contrast microscopy. Quantitative determination of b-galactosidase amounts was carried out with a commercially available kit (Galacto-Star, Tropix, Bedford, MA) according to the manufacturer’s recommendation. Each independent experiment was done in triplicate dishes. 2.6. Western blotting Protein was isolated from cells as previously described [15]. Briefly, cells were homogenized in ice-cold buffer A (50 mM KCl, 10 mM KPO4, 2 mM MgCl2, 0.5 mM EDTA, 2 mM DTT, and 0.1 mM PMSF) centrifuged at 14 K for 10 min at 4 °C and the pellet resuspended in an equal volume of buffer A. Protein concentrations were determined by a Bradford Assay (Bio-Rad, Hercules, CA) and 20 µg were run on a 15% polyacrylamide-SDS gel. The gel was blotted overnight on a PVDF membrane (Hybond, Amersham, Piscataway, NJ). Protein levels were determined by incubation with monoclonal antibody against troponin T (TnT) JLT-12 (Sigma, St. Louis, MO) or c-MYC 9E10.2 (American Type Culture Collection, Rockville, MD) and alkaline-phosphate anti-mouse IgG secondary antibody (Jackson Laboratories, West Grove, PA), treated with ECF (Bio-Rad, Hercules, CA) and quantified after exposure on a phosphorimager screen (Molecular Dynamics, Amersham, Piscataway, NJ). 3. Results In order to develop the optimal promoter system for use in different applications in primary cardiac myocytes, we tested three promoters with respect to their basal activity and response to agonist stimulation. Using a replication-deficient adenovirus vector containing the luciferase cDNA under control of the CMV promoter (Ad-CMV-Luc), we tested inducers of two different signal transduction pathways for their effects on the CMV promoter in neonatal and adult rat CMCs. Phorbol myristate acetate (PMA) was used to stimulate protein kinase C and its downstream effectors. Forskolin was used to activate cAMP and its downstream effectors, protein kinase A and cAMP response element-binding protein (CREBP). Neonatal rat CMCs were infected with Ad-CMV-Luc at an MOI of 10 infectious particles/cell and the cells were stimulated with either 100 nM PMA, 1 µM forskolin, or both. These concentrations showed maximal activation in earlier experiments with adenovirus-assisted plasmid transfection in CMCs (unpublished observations). PMA increased luciferase activity 10-fold, forskolin led to a 20-fold increase, and both substances together led to a 40-fold induction
Fig. 2. Induction of luciferase activity in neonatal and adult rat CMCs after infection with recombinant adenovirus at increasing multiplicities of infection. Luciferase activity after infection with an adenovirus vector encoding a luciferase cDNA under control of the CMV promoter at a multiplicity of infection of 10 (MOI10), 100 (MOI100), or 1000 (MOI1000) pfu/cell. Cells were infected for 18 h and then stimulated with PMA (100 nM) and forskolin (Forsk, 1 µM); CO: control, no stimulation. Bars represent mean ± S.E.M. n = 5.
(Fig. 2). These dramatic inductions led us to test 5% FBS as a maximal growth inducer for neonatal CMCs to see if this increase was mainly due to an increase in general protein synthesis. The highest possible induction with serum was about 5-fold (luciferase data not shown, see Figs. 5–7). These results suggest the induction of the CMV promoter by PMA and forskolin occurred above the stimulation of general transcription and translation. When we used higher MOIs of 100 or 1000, the stimulation by forskolin and PMA was preserved (Fig. 2), even though it decreased in magnitude at an MOI of 1000, possibly because of a toxic effect of massive transgene production. In fact, we observed toxicity at an MOI of 1000 not in control but in the forskolin and PMA-treated cells (data not shown, see Fig. 3). Interestingly, cells infected with an MOI of 10 and stimulated with PMA and forskolin showed similar luciferase activity as cells infected at an MOI of 1000 without treatment (Fig. 2) suggesting saturation already at low MOIs. We repeated these experiments in adult rat CMCs. Surprisingly, these cells showed no significant induction of luciferase activity after stimulation with PMA and forskolin (Fig. 2). Even though the quantitative response of PKC or cAMP might be different in adult vs. neonatal cells, these substances have been shown to act on both cell types and, therefore, this does not explain the total lack of response of the CMV promoter in adult cells. Serum did not lead to an increase in luciferase activity and did not act synergistically with either of the two substances (data not shown, for b-galactosidase see Fig. 4). At higher MOIs, there was a linear increase in luciferase activity, but no significant stimulation by PMA and forskolin (Fig. 2). Due to the difference in promoter activation in neonatal and adult CMCs, we tested the hypothesis that the CMV promoter was present but inactive in unstimulated neonatal CMCs. We analyzed the number of cells expressing
826
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
Fig. 3. b-Galactosidase expression in neonatal rat CMCs. Representative photomicrographs of neonatal rat CMCs after infection with an adenovirus encoding the b-galactosidase cDNA under control of the CMV promoter. MOI10, 100, 1000: 10, 100, or 1000 pfu/cell; FBS: fetal bovine serum, 5%; CO: control, no stimulation; PMA: phorbol myristate acetate, 100 nM; Forsk: forskolin, 1 µM.
b-galactosidase after infection with Ad-CMV-b-gal. Starting with an MOI of 10, 14% of neonatal rat CMCs were b-galactosidase positive, which increased after stimulation with PMA and/or forskolin to over 90% (Fig. 3). Using an MOI of 100, the stimulation caused an increase from 75% to 92% positive cells, whereas at an MOI of 1000, 94% of control cells stained blue in comparison to 98% after stimulation. Adult rat CMCs showed infection efficiencies of close
to 80% at an MOI of 10 with no significant stimulation by either substance (Fig. 4). The surprising magnitude of induction of the CMV promoter in neonatal CMCs led us to test a different viral promoter, the RSV LTR, as well as a non-viral promoter, an hsp68 minimal promoter fused with three binding sites for the transcription factor MEF2 taken from the murine desmin promoter (MEF2), each driving the b-galactosidase cDNA.
Fig. 4. b-Galactosidase expression in adult rat CMCs. Representative photomicrographs of adult rat CMCs after infection with an adenovirus encoding the b-galactosidase cDNA under control of the CMV promoter. MOI10, 100, 1000: 10, 100, or 1000 pfu/cell; FBS: fetal bovine serum, 5%; CO: control, no stimulation; PMA: phorbol myristate acetate, 100 nM; Forsk: forskolin, 1 µM.
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
827
Fig. 5. Comparison of different promoters in neonatal rat CMCs. RSV, CMV, MEF2: adenovirus-encoding b-galactosidase under control of the Rous sarcoma virus promoter, cytomegalovirus promoter, or the synthetic promoter with MEF2-binding sites; FBS: fetal bovine serum; PMA: phorbol myristate acetate; Forsk: forskolin. (a) b-Galactosidase activity. Bars represent mean ± S.E.M. Numbers correspond to fold increase vs. control. n = 5. (b) Representative photomicrographs.
To be able to directly compare them to the CMV promoter, they were cloned into the same construct keeping the other regulatory elements identical (Fig. 1). In neonatal rat CMCs, the RSV and MEF2 promoters were stronger than the CMV promoter under basal conditions, whereas the CMV promoter was stronger in cells treated with FBS (Fig. 5a). The CMV promoter was the only one that responded to PMA and forskolin in these cells with a 60-fold (vs. 40-fold that was observed with the luciferase reporter) stimulation of b-galactosidase activity over basal conditions when treated with PMA and forskolin. Both RSV and MEF2 promoters stayed at the level of serum-treated
cells when stimulated with PMA/forskolin (Fig. 5a). Fig. 5b shows representative photomicrographs of treated and untreated cells infected with the different constructs. Adult rat CMCs showed different basal activities of the three promoters: CMV was about 10-fold higher than in neonatal cells, whereas RSV and MEF2 showed 5- and 2-fold lower activities in adult vs. neonatal cells, respectively (Fig. 6a). None of the agonists used led to a significant increase of any of the tested promoters in these cells. In fact, forskolin suppressed the RSV promoter. Fig. 6b shows representative photomicrographs of infected adult CMCs. Whereas the CMV promoter led to expression in virtually all
828
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
Fig. 6. Comparison of different promoters in adult rat CMCs. RSV, CMV, MEF2: adenovirus-encoding b-galactosidase under control of the Rous sarcoma virus promoter, cytomegalovirus promoter, or the synthetic promoter with MEF2-binding sites; FBS: fetal bovine serum; PMA: phorbol myristate acetate; Forsk: forskolin. (a, top) b-Galactosidase activity. Bars represent mean ± S.E.M. n = 5. (b, bottom) Representative photomicrographs.
cells, the RSV promoter showed expression in only about 20% of the cells, suggesting an imbalance of negative vs. positive transcriptional regulators for this promoter in adult but not neonatal cells. We extended the cell type of investigation to include primary cardiac fibroblasts, a cell type that commonly contaminates CMC cell cultures and might interfere with the interpretation of experimental data. This cell type is less susceptible than CMCs to adenoviral infection with lower activity of all three tested promoters. This was most striking for the RSV promoter, which showed about 100-fold reduced activity vs. neonatal rat CMCs, whereas CMV and MEF2 were only about 5–10-fold lower in fibroblasts. Again, only the CMV promoter showed significant inducibility with PMA and/or forskolin showing a maximal inducibility of about 5-fold with both PMA and forskolin (Fig. 7a). There was a trend towards increase of the RSV promoter with PMA and forskolin but this was not statistically significant. Fig. 7b shows the lower number of infected cells and the response to treatment. Cells infected with the CMV promoter showed an increase in the amount of b-galactosidase per cell as well as an increase in the number of expressing cells. The last set of experiments was aimed at confirming the results with an endogenous cardiac protein rather than a reporter protein. We constructed a recombinant adenovirus expressing the murine TnT cDNA with an N-terminal Myc tag under control of the CMV promoter. We infected neonatal
Fig. 7. Comparison of different promoters in rat cardiac fibroblasts. RSV, CMV, MEF2: adenovirus encoding b-galactosidase under control of the Rous sarcoma virus promoter, cytomegalovirus promoter, or the synthetic promoter with MEF2-binding sites; FBS: fetal bovine serum; PMA: phorbol myristate acetate; Forsk: forskolin. (a, top) b-Galactosidase activity. Bars represent mean ± S.E.M. Numbers correspond to fold increase vs. control. n = 5. (b, bottom) Representative photomicrographs.
rat CMCs with increasing MOIs of this virus and determined protein levels of endogenous and transgenic protein by western blotting with TnT and Myc-specific antibodies. Since there can only be a certain amount of TnT incorporated into the sarcomere, there is no alteration of the stoichiometry of cTnT but free TnT is degraded rapidly. Endogenous TnT is, therefore, replaced by adenovirally-encoded protein. Since the Myc-tagged TnT migrates a little slower on polyacrylamide gels because of its slightly higher molecular weight, there are two bands when probed with a TnT-specific antibody (Fig. 8a), allowing quantification of the relative amounts of endogenous vs. adenovirally expressed protein. With MOIs of 2, 10, or 50 the fraction of Myc-tagged TnT increased from 2% to 40% to 99%. Forskolin or phenylephrine increased this percentage to 33% or 35% (MOI2), and 98% or 99% (MOI10), respectively (Fig. 8b). Forskolin leads to a suppression of the endogenous protein, which might change the ratio of mutant/wild-type protein but it was in-
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
Fig. 8. Dose-dependent replacement of endogenous TnT by adenoviral gene transfer of Myc-tagged TnT. CO: control, no treatment; PE: phenylephrine, 20 µM; Fo: forskolin, 1 µM; MOI 2, 10, 50: infection with 2, 10, or 50 pfu of recombinant adenovirus expressing Myc-tagged murine TnT under control of the CMV promoter. (a, top) Representative western blot from neonatal CMC lysates. TnT: western blot with monoclonal anti-troponin T JLT-12 antibody; upper band: Myc-tagged TnT protein; lower band: endogenous TnT. Myc: western blot with monoclonal anti-MYC 9E10 antibody. Bars represent mean ± S.E.M. n = 4. (b, bottom) Percentage of Myc-tagged vs. endogenous TnT quantified from western blots. n = 3.
ferred from the reporter gene experiments that it also induces the mutant TnT. These results demonstrate that very low amounts of adenovirus are sufficient to lead to a replacement of a large proportion of an endogenous sarcomeric protein and that the amount can be controlled by treatment with various agonists. We introduced phenylephrine as a more relevant stimulus since it also induces hypertrophy of neonatal CMCs with the accumulation of additional sarcomeres and, therefore, stimulates transcription and translation of the endogenous TnT. 4. Discussion This study was designed to compare two commonly used viral promoters and a synthetic promoter in an adenoviral context with regard to their expression in primary cardiac cells and their regulation by pharmacological intervention that are commonly used in cardiac myocytes. The CMV promoter, thought to be maximally active at basal conditions, has been chosen for many gene transfer experiments. However, it has more recently been shown to be regulated by many stimuli, such as a interferon (downregulation), PMA/forskolin (induction), and methyl methanesulfonate (induction) in different cells types and has also
829
been shown to be inactivated after in vivo gene transfer into the mouse liver [6,16–18]. In this study, we show its dramatic induction by PMA and forskolin in neonatal CMCs. The mechanism of this induction remains unknown, but a number of hypotheses exist. An increase in the actual uptake of the complexes or adenoviruses and, thereby, increase in gene transfer efficiency is unlikely because the substances are added 18 h after the transfection/infection and the cells are washed with fresh media removing any external virus. One might argue about complexes or adenovirus still attached to the cell membrane that could then be taken up. It has been shown, however, that adenovirus uptake into cells is very rapid and should be completed after a few hours [19]. Furthermore, this induction does not occur for the RSV and MEF2/hsp68 promoters. The induction of the CMV promoter by PMA and forskolin is also not likely to be merely due to generalized increase in transcription because the effect is also observed when cellular growth is maximally induced by FBS. An increase of the half-life of luciferase in mammalian cells of about 3 h is also not likely since the effect is also seen with b-galactosidase, a protein with a half-life of more than 20 h [20,21]. A more likely explanation is the induction of specific transcription factors stimulating transcription from the CMV early promoter. Binding sites for multiple transcription factors have been identified in this promoter, some being implicated in stimulation (NFjB, CREBP) and another (YY1) is thought to be a negative regulator [22]. Phorbol esters are inducers of the signal transduction molecule protein kinase C that induces multiple cellular pathways including the transcription factor NFjB in CMCs [23]. Forskolin leads to increased cellular cAMP levels that induce protein kinase A and the transcription factor CREBP. Adult cells, however, seem to have a much higher basal transcription from the CMV promoter, producing about 10 times more luciferase per cell than neonatal cells at the same MOI. A possible explanation lies in the higher expression of a negative transcription factor in neonatal vs. adult cells with the transcription factor YY1 being a candidate. Neonatal CMCs express high levels of YY1, whereas expression levels in adult cells is low unless there is a pathological state (C.S. Long, personal communication; Ref. [24]). An interesting and surprising finding of this study is that the induction of the CMV promoter is not restricted to cells that already express a gene from this promoter to a detectable extent, but that a high percentage of neonatal rat CMCs are “silently infected”. This transcriptional inactivity can be overcome with treatment with PMA and/or forskolin. Adult rat CMCs, on the other hand, express CMV-driven b-galactosidase in the majority of cells already at a low MOI of 10 and have a higher basal activity. Inactivation of the CMV promoter has been described in vivo after adenoviral infection of the mouse liver. It could be reactivated after reinfection of the liver with an “empty” (i.e. transgene- and promoter-less) replication-deficient adenovirus. A transient induction of the transcription factor NFjB after adenoviral infection was shown to be the likely molecular mechanism
830
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
[18]. This transcription factor is expressed at low levels in CMCs that increase with (patho)physiology, such as ischemia, aging, or hypertrophy [25,26]. The other transcription factors with demonstrated binding to the CMV promoter are also of low abundance/activity in untreated CMCs, possibly leaving an imbalance between positive and negative transcriptional regulators towards the latter [27]. The RSV promoter has also been extensively used for gene transfer to the myocardium. This promoter, however, is less well characterized. In transgenic animals it is highly expressed in cardiac and skeletal muscle [28]. In contrast to the CMV promoter, it has been shown not to be responsive to increased cAMP levels [29]. Transcription from this promoter has been shown to be dependent on YY1, which, under certain circumstances, can function as a transcriptional activator [30]. This is in striking contrast to the CMV promoter and might be an explanation for the observed differences in basal promoter activity between neonatal and adult CMCs. The synthetic MEF2-responsive promoter did not behave as expected from the transgenic animal results [8]. We expected activity to be low in CMCs under basal conditions and in fibroblasts but to be induced by hypertrophic stimuli. We did, however, observe high-basal activity, even in fibroblasts and little responsiveness. A possible explanation is transactivation by a cryptic promoter in the adenovirus genome rather than the hsp68–MEF2 promoter [31]. Further experiments are aimed at insulating this promoter and using a plasmid construct to test this hypothesis. The 60-fold induction of the CMV promoter in neonatal CMCs by agonists suggests using caution when employing this promoter as a normalization for studies involving promoter analysis of other cellular promoters. This phenomenon is not restricted to cardiac cells but is also found in other cells. As stated earlier, an inactivation of the CMV promoter has also been demonstrated in vivo [18]. Therefore, when using this promoter in an in vivo setting, it is important to be aware of this phenomenon. We and others have seen a significant induction of the CMV promoter after i.m. injection in a model of muscle degeneration/regeneration (Allen at al., unpublished observations; Ref. [32]). In addition to activation of promoters by physiological or pathological stimuli, promoter inactivation is an increasingly recognized phenomenon and is a vector-independent problem that might inhibit therapeutic use of several promoters [33]. The relevance of our reporter gene experiments for gene transfer was confirmed with one the contractile protein, TnT, that was Myc-tagged to differentiate it from the endogenous protein. Infection of neonatal CMCs with this adenovirus led to replacement of the endogenous protein that was dose dependent. As stoichiometry of contractile proteins cannot be altered in cells, overexpression cannot be achieved at protein levels but endogenous protein is replaced by adenovirally expressed proteins [15]. At very low MOIs, the adenovirally expressed protein was only 2% of total, whereas stimulation increased this to 33–35%. This result emphasizes the use of the CMV promoter as an inducible system for relevant
proteins and might be important for potentially toxic gene products where expression can be timed by induction. For contractile proteins, exact titration of protein expression is important for the study of mutated proteins that are thought to act through dominant-negative interaction with wild-type proteins [15]. Despite the cautionary statements made above, we believe the properties of these promoters can be exploited in cardiac cells in the following fashion: The CMV promoter is inducible in both primary cardiac fibroblasts and neonatal CMCs. It shows low basal activity in these cells and could be used when high-level inducible expression of a gene is desired. It might be very useful for adult cardiac myocytes to induce dose-dependent but noninducible expression. The RSV promoter shows higher basal activity than CMV and less inducibility in neonatal CMCs and very low activity in fibroblasts, making it a promoter of choice if it is used as a normalizing standard for activity of cardiac promoters. It shows both cardiac selectivity and it is particularly useful if (pharmacological) treatment of cells is required. It is less active in adult CMCs than CMV. The synthetic MEF2-responsive promoter shows the least inducibility and high-basal activity in all cell types (higher than RSV in non-myocytes) and may be a good promoter to use as a control in co-infection experiments when put into an adenoviral context as in our construct. In addition, we have confirmed that CMCs are very susceptible to adenoviral infection in that virtually all neonatal cells are transgene positive at very low MOIs but if the CMV promoter is used, it needs specific transcriptional activation. These results emphasize that the current focus on finding optimal vector systems needs to be extended by characterization of the regulatory elements in gene transfer constructs to optimize gene expression in the target tissues.
Acknowledgements This work was supported by grants from the Deutsche Forschungsgemeinschaft (Ma 2185/1-1 to A. Maass and TPB6, SFB355 to L. Neyses) and by NIH grant HL50560 to. L.A. Leinwand. The authors would like to thank R.C. Thompson and M. Buvoli for continuous discussion and support.
References [1]
Kirshenbaum LA, MacLellan WR, Mazur W, French BA, Schneider MD. Highly efficient gene transfer into adult ventricular myocytes by recombinant adenovirus. J Clin Invest 1993;92:381–7.
[2]
Stratford-Perricaudet LD, Makeh I, Perricaudet M, Briand P. Widespread long-term gene transfer to mouse skeletal muscles and heart. J Clin Invest 1992;90:626–30.
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831 [3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
Kass-Eisler A, Falck-Pedersen E, Alvira M, Rivera J, Buttrick P, Wittenberg B, et al. Quantitative determination of adenovirusmediated gene delivery to rat cardiac myocytes in vitro and in vivo. Proc Natl Acad Sci USA 1993;90:11498–502. Zhang JCL, Woo YJ, Chen J-A, Swain JL, Sweeney HL. Efficient transmural cardiac gene transfer by intrapericardial injection in neonatal mice. J Mol Cell Cardiol 1999;31:721–32. Barr E, Carroll J, Kalynych AM, Tripathy SK, Kozarsky K, Wilson JM, et al. Efficient catheter-mediated gene transfer into the heart using replication-defective adenovirus. Gene Ther 1994;1:51–8. Clesham GJ, Browne H, Efstathiou S, Weissberg PL. Enhancer stimulation unmasks latent gene transfer after adenovirus-mediated gene delivery into human vascular smooth muscle cells. Circ Res 1996;79: 1188–95. Pasleau F, Tocci MJ, Leung F, Kopchick JJ. Growth hormone gene expression in eukaryotic cells directed by Rous sarcoma virus long terminal repeat or cytomegalovirus immediate-early promoter. Gene 1985;38:227–32. Naya FJ, Wu C, Richardson JA, Overbeek P, Olson EN. Transcriptional activity of MEF2 during mouse embryogenesis with a MEF2dependent transgene. Development 1999;126:2045–52. Passier R, Zeng H, Frey N, Naya FJ, Nicol RL, McKinsey TA, et al. CaM kinase signaling induces cardiac hypertrophy and activates the MEF2 transcription factor in vivo. J Clin Invest 2000;105:1395–406. Simpson P, McGrath A, Savion S. Myocyte hypertrophy in neonatal rat heart cultures and its regulation by serum and catecholamines. Circ Res 1982;51:787–801. Wollnik B, Kubisch C, Maass A, Vetter H, Neyses L. Hyperosmotic stress induces immediate-early gene expression in ventricular adult cardiomyocytes. Biochem Biophys Res Commun 1993;194:642–6. DeWet JR, Wood KV, DeLuca M, Helinski DR, Subramani S. Cloning of the firefly luciferase cDNA and its expression in mammalian cells. Mol Cell Biol 1987;7:725–37. Becker TC, Noel RJ, Coats WS, Gomez-Foix AM, Alam T, Gerard RD, et al. Use of recombinant adenovirus for metabolic engineering of mammalian cells. Meth Cell Biol 1994;43:161–89. Schaack J, Langer S, Guo X. Efficient selection of recombinant adenoviruses by vectors that express beta-galactosidase. J Virol 1995; 69:3920–3. Tardiff JC, Factor SM, Tompkins BD, Hewett TE, Palmer BM, Moore RL, et al. A truncated troponin T molecule in transgenic mice suggests multiple cellular mechanisms for familial hypertrophic cardiomyopathy. J Clin Invest 1998;101:2800–11. Acsadi G, O’Hagan D, Lochmüller H, Prescott S, Larochelle N, Nalbantoglu J, et al. Interferons impair early transgene expression by adenovirus-mediated gene transfer in muscle cells. J Mol Med 1998; 76:442–50. Nabirochkina EN, Ivanov AV. Treatment by methyl methanesulfonate induces up-regulation of cytomegalovirus immediate/early promoter. Biotechniques 2000;29:732–6.
831
[18] Loser P, Jennings GS, Strauss M, Sandig V. Reactivation of the previously silenced cytomegalovirus major immediate-early promoter in the mouse liver: involvement of NFkappaB. J Virol 1997;72: 180–90. [19] Greber UF, Willets M, Webster P, Helenius A. Stepwise dismantlingof adenovirus 2 during entry into cells. Cell 1993;75:477–86. [20] Thompson JF, Hayes LS, Lloyd DB. Modulation of firefly luciferase stability and impact on studies of gene regulation. Gene 1991;103: 171–7. [21] Bachmair A, Finley D, Varshavsky A. In vivo half-life of a protein is a function of its amino-terminal residue. Science 1986;234:179–86. [22] Liu R, Baillie J, Sissons JGP, Sinclair JH. The transcription factor YY1 binds to negative regulatory elements in the human cytomegalovirus major immediate early enhancer/promoter and mediates repression in non-permissive cells. Nucl Acids Res 1994;22:2453–9. [23] Rouet-Benzineb P, Gontero B, Dreyfus P, Lafuma C. Angiotensin II induces nuclear factor-jB activation in cultured neonatal rat cardiomyocytes through protein kinase C signaling pathway. J Mol Cell Cardiol 2000;32:1767–78. [24] Patten M, Wang W, Aminololama-Shakeri S, Burson M, Long CS. IL-1b increases abundance and activity of the negative transcriptional regulator yin yang-1 (YY1) in neonatal rat cardiac myocytes. J Mol Cell Cardiol 2000;32:1341–52. [25] Helenius M, Hänninen M, Lehtinen SK, Salminen A. Aging-induced up-regulation of nuclear binding activities of oxidative stress responsive NF-jB transcription factor in mouse cardiac muscle. J Mol Cell Cardiol 1996;28:487–98. [26] Purcell NH, Tang G, Yu C, Mecurio F, DiDonato JA, Lin A. Activation of NF-jB is required for hypertrophic growth of primary rat neonatal ventricular cardiomyocytes. Proc Natl Acad Sci USA 98:6668–73. [27] Muller FU, Neumann F, Schmitz W. Transcriptional regulation by cAMP in the heart. Mol Cell Biochem 2000;212:11–7. [28] Conti FG, Powell R, Pozzi L, Zezze G, Faraggiana T, Gannon F, et al. A novel line of transgenic mice (RSV/LTR-bGH) expressing growth hormone in cardiac and striated muscle. Growth Regul 1995;5:101–8. [29] Mellon PL, Clegg CH, Correll LA, McKnight GS. Regulation of transcription by cyclic AMP-dependent protein kinase. Proc Natl Acad Sci USA 1989;86:4887–91. [30] Bhalla SS, Robitaille L, Nemer M. Cooperative activation by GATA-4 and YY1 of the cardiac B-type natriuretic peptide promoter. J Biol Chem 2001;276:11439–45. [31] Buvoli M, Langer SJ, Bialik S, Leinwand LA. Potential limitations of transcription terminators used as transgene insulators in adenoviral vectors. Gene Ther 2002;9:227–31. [32] Danko I, Fritz JD, Jiao S, Hogan K, Latendresse JS, Wolff JA. Pharmacological enhancement of in vivo foreign gene expression in muscle. Gene Ther 1994;1:114–21. [33] Bestor TH. Gene silencing as a threat to the success of gene therapy. J Clin Invest 2000;105:409–11.
Original Article
Rational promoter selection for gene transfer into cardiac cells Alexander Maass a,b, Stephen J. Langer a, Silke Oberdorf-Maass a,b, Sebastian Bauer b,1, Ludwig Neyses b,2, Leslie A. Leinwand a,* a
Department of Molecular, Cellular, and Developmental Biology, University of Colorado, Campus Box 347, Boulder, CO 80309 0347, USA b Department of Medicine, University of Wuerzburg, Germany Received 26 November 2002; received in revised form 14 March 2003; accepted 3 April 2003
Abstract Cardiomyocytes (CMCs) are extremely difficult to transfect with non-viral techniques, but they are efficiently infected by adenoviruses. The most commonly used promoters to drive protein expression in cardiac myocytes are of viral origin, since they are believed to be constitutively active and minimally regulated by physiological or pharmacological challenge of cells. In recombinant adenoviruses, we systematically compared three different promoters: the cytomegalovirus (CMV), the Rous sarcoma virus (RSV), and a synthetic promoter with three MEF2 transcription factor-binding sites upstream of the heat-shock protein 68 minimal promoter. We determined their basal activity in primary cardiac cells as well as their possible stimulation by commonly used agonists. The CMV promoter was activated up to 60-fold by the phorbol ester phorbol myristate acetate (PMA) and/or forskolin in neonatal rat CMCs and cardiac fibroblasts. Primary adult rat CMCs had higher basal expression from the CMV promoter that was not activated by PMA or forskolin. The RSV promoter was less affected by agonists and was more active in cardiac myocytes compared to cardiac fibroblasts. The MEF2-responsive promoter showed high basal expression in both myocytes and fibroblasts, and minimal induction by phorbol esters and forskolin. The relevance of reporter gene induction was confirmed with a contractile protein, troponin T (TnT). The CMV promoter driving TnT could be induced more than 15-fold with phenylephrine or forskolin to replace the endogenous protein almost to completion at a multiplicity of infection of 10. These results suggest the following use of the tested promoters: an inducible system (CMV), a myocyte-enriched system (RSV), or a stable control system (MEF2). © 2003 Elsevier Science Ltd. All rights reserved. Keywords: Cytomegalovirus; Rous sarcoma virus; MEF2; Transcription factor; Promoter; Phorbol ester; Forskolin; Gene transfer; Cardiomyocytes; Fibroblasts
1. Introduction Cardiomyocytes (CMCs) are generally considered to be one of the most difficult cell types to transfect with DNA constructs. Adenovirus vectors have facilitated gene transfer into both adult and neonatal rat CMCs in vitro as well as neonatal and adult myocardium in vivo [1–5]. To drive highlevel expression of proteins of interest in cardiovascular cells, most investigators use viral promoters in recombinant adenovirus vectors. Previous experiments have shown that the cytomegalovirus (CMV) promoter can be activated up to 23-fold by forskolin and phorbol ester in human vascular * Corresponding author. Tel.: +1-303-492-7606; fax: +1-303-492-8907. E-mail address: [email protected] (L.A. Leinwand). 1
Present address: Department of Medicine, West German Cancer Center, University of Essen Medical School, Germany. 2 Present address: Department of Medicine, University of Manchester, Manchester, UK.
© 2003 Elsevier Science Ltd. All rights reserved. DOI: 10.1016/S0022-2828(03)00140-8
smooth muscle cells [6]. Regulation of other viral promoters has been less extensively studied in cardiac myocytes. Different viral promoters have never been compared directly in isolated cardiac cells with regard to (1) their specificity for myocyte expression vs. cardiac fibroblasts, (2) the amount of protein produced, and (3) their regulation by physiological or pharmacological intervention. We constructed adenovirus vectors with either the luciferase or b-galactosidase cDNA, driven by three different promoters, while maintaining the termination and polyadenylation signals constant (Fig. 1). The first is the 800 bp CMV immediate-early promoter, the second is the Rous sarcoma virus (RSV) long-terminal repeat (LTR), both extensively used for gene transfer experiments [7]. The third vector contains an artificial promoter consisting of a heatshock protein 68 (hsp68) minimal promoter and three binding sites from the desmin gene for the transcription factor MEF2. MEF2 is a transcription factor that is expressed primarily in developing muscle and brain tissue. A transgenic
824
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
2.3. Recombinant adenoviruses
Fig. 1. DNA constructs used to generate recombinant adenoviruses for this study. Boxes at the start and the end represent the inverted terminal repeats; W: packaging signal; CMV: cytomegalovirus immediate-early promoter; RSV: Rous sarcoma virus long-terminal repeat; MEF2 × 3: binding sites for the transcription factor MEF2 from the murine desmin promoter in triplicate; hsp68: minimal promoter from the heat-shock protein 68 gene; SV40 pA: polyadenylation and termination signals from the simian virus 40 genome. Figure is not to scale.
mouse expressing this construct showed expression in embryonic muscle and brain tissues [8]. In the adult, MEF2 activity was very low in the heart and increased with pathologic conditions such as hypertrophy associated with overexpression of a constitutive calmodulin-dependent protein kinase [9]. Our study demonstrates differential basal activity and stimulation of the tested promoters that should lead to differential use of these promoters depending on the experimental design.
2. Materials and methods 2.1. Isolation and culture of neonatal rat CMCs and cardiac fibroblasts Hearts from several litters of 1–3-d-old Wistar rats were excised and cells were isolated by trypsin digestion (0.1%) as previously described [10]. To enrich for CMCs, the cell suspension was preplated on regular polystyrene dishes for 30–45 min. CMCs were cultured in serum-free MEM supplemented with transferrin (1 mg/ml), insulin (1 µg/ml), and bovine serum albumin (BSA) (1 mg/ml) for 24–48 h before experiments were carried out. Dishes from preplating were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), passaged to enrich for the faster dividing fibroblasts from other non-myocytes, and used for experiments from passages 3 to 5.
Ad-CMV-LacZ and Ad-RSV-LacZ were generous gifts from Dr. Jerome Schaack (UCHSC-Denver) and are firstgeneration adenovirus vectors constructed using the Addl327 background. Ad-CMV-LacZ contains the 1.1 kb immediate-early promoter from human CMV driving the expression of the Escherichia coli b-galactosidase gene (LacZ). Ad-RSV-LacZ differs from Ad-CMV-LacZ by substitution of the RSV LTR for the CMV promoter. Ad-CMVLuc was derived by subcloning the BglII-BamHI fragment containing the luciferase gene from pJD205 into the polylinker of pACCMV.LpA [12,13]. A SalI fragment containing three head-to-tail copies of the desmin MEF2binding site fused to the hsp68 promoter and driving the expression of the b-galactosidase gene was subcloned into a left-end adenovirus plasmid construct, containing the adenovirus serotype 5 sequence from 1358 bp (SacII) and from 3330 (BglII) to 5700 bp (XhoI), within a polylinker inserted between the SacII and BglII sites to generate pAd-MEF2 × 3hsp68LacZ [8]. Methods for generating recombinant adenovirus LacZ expression vectors have been described previously [14]. 2.4. Amplification and purification of recombinant adenoviruses All the viral stocks were amplified from plaque-purified viral stocks. Thirty-six 100-mm dishes were seeded at 5 × 106 cells/dish and infected at a multiplicity of infection (MOI) of 25 plaque-forming units (pfu)/cell. After 48 h, virus was concentrated by low-speed centrifugation to pellet infected cells, resuspended in 20 ml of growth media, and then freeze– thawed for four cycles to release progeny virus. The freeze–thaw cell lysate was clarified by low-speed centrifugation and virus particles purified by banding on CsCl buoyant density gradients. The purified virus was diluted 5-fold in virus storage buffer (250 mM NaCl, 100 mM Tris (pH 7.5), 50% glycerol, 1 mg/ml BSA, 1 mM MgCl2) and stored at –20 °C. Virus concentration was determined by plaquetitering and b-galactosidase expressing plaques detected by addition of X-Gal as previously described [14]. The predicted genomic structure was verified by restriction analysis of isolated viral genomic DNA. Particle/pfu ratios were routinely around 50. No attempt was made to recover replication-competent adenovirus, although experiments were not conducted in human cells where this might have presented a problem. 2.5. Luciferase and b-galactosidase assays
2.2. Isolation and culture of adult rat CMCs The hearts of 3-month-old male Sprague–Dawley rats (250–300 g) were excised and CMCs were isolated by perfusion with collagenase as previously described [11]. Cells were cultured on laminin-coated plates in serum-free DMEM supplemented with BSA (1 mg/ml) and insulin (1 µg/ml).
After the indicated incubation period, cells were washed twice with phosphate-buffered saline (PBS) and lysed in 1× passive lysis buffer (Promega, Madison, WI). Protein concentrations were measured by the modified Bradford method (Bio-Rad, Hercules, CA) and luciferase activity determined after mixing with 100 µl assay reagent (Promega, Madison,
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
825
WI) in a standard luminometer (Turner Designs, Sunnyvale, CA). Relative light units (RLU) were normalized to microgram protein content in the sample. Cells infected with b-galactosidase containing constructs were washed twice with PBS, fixed for 10 min in 2% formaldehyde and stained in potassium ferricyanide/potassium ferrocyanide and BCIG overnight. b-Galactosidase-positive cells were counted by phase-contrast microscopy. Quantitative determination of b-galactosidase amounts was carried out with a commercially available kit (Galacto-Star, Tropix, Bedford, MA) according to the manufacturer’s recommendation. Each independent experiment was done in triplicate dishes. 2.6. Western blotting Protein was isolated from cells as previously described [15]. Briefly, cells were homogenized in ice-cold buffer A (50 mM KCl, 10 mM KPO4, 2 mM MgCl2, 0.5 mM EDTA, 2 mM DTT, and 0.1 mM PMSF) centrifuged at 14 K for 10 min at 4 °C and the pellet resuspended in an equal volume of buffer A. Protein concentrations were determined by a Bradford Assay (Bio-Rad, Hercules, CA) and 20 µg were run on a 15% polyacrylamide-SDS gel. The gel was blotted overnight on a PVDF membrane (Hybond, Amersham, Piscataway, NJ). Protein levels were determined by incubation with monoclonal antibody against troponin T (TnT) JLT-12 (Sigma, St. Louis, MO) or c-MYC 9E10.2 (American Type Culture Collection, Rockville, MD) and alkaline-phosphate anti-mouse IgG secondary antibody (Jackson Laboratories, West Grove, PA), treated with ECF (Bio-Rad, Hercules, CA) and quantified after exposure on a phosphorimager screen (Molecular Dynamics, Amersham, Piscataway, NJ). 3. Results In order to develop the optimal promoter system for use in different applications in primary cardiac myocytes, we tested three promoters with respect to their basal activity and response to agonist stimulation. Using a replication-deficient adenovirus vector containing the luciferase cDNA under control of the CMV promoter (Ad-CMV-Luc), we tested inducers of two different signal transduction pathways for their effects on the CMV promoter in neonatal and adult rat CMCs. Phorbol myristate acetate (PMA) was used to stimulate protein kinase C and its downstream effectors. Forskolin was used to activate cAMP and its downstream effectors, protein kinase A and cAMP response element-binding protein (CREBP). Neonatal rat CMCs were infected with Ad-CMV-Luc at an MOI of 10 infectious particles/cell and the cells were stimulated with either 100 nM PMA, 1 µM forskolin, or both. These concentrations showed maximal activation in earlier experiments with adenovirus-assisted plasmid transfection in CMCs (unpublished observations). PMA increased luciferase activity 10-fold, forskolin led to a 20-fold increase, and both substances together led to a 40-fold induction
Fig. 2. Induction of luciferase activity in neonatal and adult rat CMCs after infection with recombinant adenovirus at increasing multiplicities of infection. Luciferase activity after infection with an adenovirus vector encoding a luciferase cDNA under control of the CMV promoter at a multiplicity of infection of 10 (MOI10), 100 (MOI100), or 1000 (MOI1000) pfu/cell. Cells were infected for 18 h and then stimulated with PMA (100 nM) and forskolin (Forsk, 1 µM); CO: control, no stimulation. Bars represent mean ± S.E.M. n = 5.
(Fig. 2). These dramatic inductions led us to test 5% FBS as a maximal growth inducer for neonatal CMCs to see if this increase was mainly due to an increase in general protein synthesis. The highest possible induction with serum was about 5-fold (luciferase data not shown, see Figs. 5–7). These results suggest the induction of the CMV promoter by PMA and forskolin occurred above the stimulation of general transcription and translation. When we used higher MOIs of 100 or 1000, the stimulation by forskolin and PMA was preserved (Fig. 2), even though it decreased in magnitude at an MOI of 1000, possibly because of a toxic effect of massive transgene production. In fact, we observed toxicity at an MOI of 1000 not in control but in the forskolin and PMA-treated cells (data not shown, see Fig. 3). Interestingly, cells infected with an MOI of 10 and stimulated with PMA and forskolin showed similar luciferase activity as cells infected at an MOI of 1000 without treatment (Fig. 2) suggesting saturation already at low MOIs. We repeated these experiments in adult rat CMCs. Surprisingly, these cells showed no significant induction of luciferase activity after stimulation with PMA and forskolin (Fig. 2). Even though the quantitative response of PKC or cAMP might be different in adult vs. neonatal cells, these substances have been shown to act on both cell types and, therefore, this does not explain the total lack of response of the CMV promoter in adult cells. Serum did not lead to an increase in luciferase activity and did not act synergistically with either of the two substances (data not shown, for b-galactosidase see Fig. 4). At higher MOIs, there was a linear increase in luciferase activity, but no significant stimulation by PMA and forskolin (Fig. 2). Due to the difference in promoter activation in neonatal and adult CMCs, we tested the hypothesis that the CMV promoter was present but inactive in unstimulated neonatal CMCs. We analyzed the number of cells expressing
826
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
Fig. 3. b-Galactosidase expression in neonatal rat CMCs. Representative photomicrographs of neonatal rat CMCs after infection with an adenovirus encoding the b-galactosidase cDNA under control of the CMV promoter. MOI10, 100, 1000: 10, 100, or 1000 pfu/cell; FBS: fetal bovine serum, 5%; CO: control, no stimulation; PMA: phorbol myristate acetate, 100 nM; Forsk: forskolin, 1 µM.
b-galactosidase after infection with Ad-CMV-b-gal. Starting with an MOI of 10, 14% of neonatal rat CMCs were b-galactosidase positive, which increased after stimulation with PMA and/or forskolin to over 90% (Fig. 3). Using an MOI of 100, the stimulation caused an increase from 75% to 92% positive cells, whereas at an MOI of 1000, 94% of control cells stained blue in comparison to 98% after stimulation. Adult rat CMCs showed infection efficiencies of close
to 80% at an MOI of 10 with no significant stimulation by either substance (Fig. 4). The surprising magnitude of induction of the CMV promoter in neonatal CMCs led us to test a different viral promoter, the RSV LTR, as well as a non-viral promoter, an hsp68 minimal promoter fused with three binding sites for the transcription factor MEF2 taken from the murine desmin promoter (MEF2), each driving the b-galactosidase cDNA.
Fig. 4. b-Galactosidase expression in adult rat CMCs. Representative photomicrographs of adult rat CMCs after infection with an adenovirus encoding the b-galactosidase cDNA under control of the CMV promoter. MOI10, 100, 1000: 10, 100, or 1000 pfu/cell; FBS: fetal bovine serum, 5%; CO: control, no stimulation; PMA: phorbol myristate acetate, 100 nM; Forsk: forskolin, 1 µM.
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
827
Fig. 5. Comparison of different promoters in neonatal rat CMCs. RSV, CMV, MEF2: adenovirus-encoding b-galactosidase under control of the Rous sarcoma virus promoter, cytomegalovirus promoter, or the synthetic promoter with MEF2-binding sites; FBS: fetal bovine serum; PMA: phorbol myristate acetate; Forsk: forskolin. (a) b-Galactosidase activity. Bars represent mean ± S.E.M. Numbers correspond to fold increase vs. control. n = 5. (b) Representative photomicrographs.
To be able to directly compare them to the CMV promoter, they were cloned into the same construct keeping the other regulatory elements identical (Fig. 1). In neonatal rat CMCs, the RSV and MEF2 promoters were stronger than the CMV promoter under basal conditions, whereas the CMV promoter was stronger in cells treated with FBS (Fig. 5a). The CMV promoter was the only one that responded to PMA and forskolin in these cells with a 60-fold (vs. 40-fold that was observed with the luciferase reporter) stimulation of b-galactosidase activity over basal conditions when treated with PMA and forskolin. Both RSV and MEF2 promoters stayed at the level of serum-treated
cells when stimulated with PMA/forskolin (Fig. 5a). Fig. 5b shows representative photomicrographs of treated and untreated cells infected with the different constructs. Adult rat CMCs showed different basal activities of the three promoters: CMV was about 10-fold higher than in neonatal cells, whereas RSV and MEF2 showed 5- and 2-fold lower activities in adult vs. neonatal cells, respectively (Fig. 6a). None of the agonists used led to a significant increase of any of the tested promoters in these cells. In fact, forskolin suppressed the RSV promoter. Fig. 6b shows representative photomicrographs of infected adult CMCs. Whereas the CMV promoter led to expression in virtually all
828
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
Fig. 6. Comparison of different promoters in adult rat CMCs. RSV, CMV, MEF2: adenovirus-encoding b-galactosidase under control of the Rous sarcoma virus promoter, cytomegalovirus promoter, or the synthetic promoter with MEF2-binding sites; FBS: fetal bovine serum; PMA: phorbol myristate acetate; Forsk: forskolin. (a, top) b-Galactosidase activity. Bars represent mean ± S.E.M. n = 5. (b, bottom) Representative photomicrographs.
cells, the RSV promoter showed expression in only about 20% of the cells, suggesting an imbalance of negative vs. positive transcriptional regulators for this promoter in adult but not neonatal cells. We extended the cell type of investigation to include primary cardiac fibroblasts, a cell type that commonly contaminates CMC cell cultures and might interfere with the interpretation of experimental data. This cell type is less susceptible than CMCs to adenoviral infection with lower activity of all three tested promoters. This was most striking for the RSV promoter, which showed about 100-fold reduced activity vs. neonatal rat CMCs, whereas CMV and MEF2 were only about 5–10-fold lower in fibroblasts. Again, only the CMV promoter showed significant inducibility with PMA and/or forskolin showing a maximal inducibility of about 5-fold with both PMA and forskolin (Fig. 7a). There was a trend towards increase of the RSV promoter with PMA and forskolin but this was not statistically significant. Fig. 7b shows the lower number of infected cells and the response to treatment. Cells infected with the CMV promoter showed an increase in the amount of b-galactosidase per cell as well as an increase in the number of expressing cells. The last set of experiments was aimed at confirming the results with an endogenous cardiac protein rather than a reporter protein. We constructed a recombinant adenovirus expressing the murine TnT cDNA with an N-terminal Myc tag under control of the CMV promoter. We infected neonatal
Fig. 7. Comparison of different promoters in rat cardiac fibroblasts. RSV, CMV, MEF2: adenovirus encoding b-galactosidase under control of the Rous sarcoma virus promoter, cytomegalovirus promoter, or the synthetic promoter with MEF2-binding sites; FBS: fetal bovine serum; PMA: phorbol myristate acetate; Forsk: forskolin. (a, top) b-Galactosidase activity. Bars represent mean ± S.E.M. Numbers correspond to fold increase vs. control. n = 5. (b, bottom) Representative photomicrographs.
rat CMCs with increasing MOIs of this virus and determined protein levels of endogenous and transgenic protein by western blotting with TnT and Myc-specific antibodies. Since there can only be a certain amount of TnT incorporated into the sarcomere, there is no alteration of the stoichiometry of cTnT but free TnT is degraded rapidly. Endogenous TnT is, therefore, replaced by adenovirally-encoded protein. Since the Myc-tagged TnT migrates a little slower on polyacrylamide gels because of its slightly higher molecular weight, there are two bands when probed with a TnT-specific antibody (Fig. 8a), allowing quantification of the relative amounts of endogenous vs. adenovirally expressed protein. With MOIs of 2, 10, or 50 the fraction of Myc-tagged TnT increased from 2% to 40% to 99%. Forskolin or phenylephrine increased this percentage to 33% or 35% (MOI2), and 98% or 99% (MOI10), respectively (Fig. 8b). Forskolin leads to a suppression of the endogenous protein, which might change the ratio of mutant/wild-type protein but it was in-
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
Fig. 8. Dose-dependent replacement of endogenous TnT by adenoviral gene transfer of Myc-tagged TnT. CO: control, no treatment; PE: phenylephrine, 20 µM; Fo: forskolin, 1 µM; MOI 2, 10, 50: infection with 2, 10, or 50 pfu of recombinant adenovirus expressing Myc-tagged murine TnT under control of the CMV promoter. (a, top) Representative western blot from neonatal CMC lysates. TnT: western blot with monoclonal anti-troponin T JLT-12 antibody; upper band: Myc-tagged TnT protein; lower band: endogenous TnT. Myc: western blot with monoclonal anti-MYC 9E10 antibody. Bars represent mean ± S.E.M. n = 4. (b, bottom) Percentage of Myc-tagged vs. endogenous TnT quantified from western blots. n = 3.
ferred from the reporter gene experiments that it also induces the mutant TnT. These results demonstrate that very low amounts of adenovirus are sufficient to lead to a replacement of a large proportion of an endogenous sarcomeric protein and that the amount can be controlled by treatment with various agonists. We introduced phenylephrine as a more relevant stimulus since it also induces hypertrophy of neonatal CMCs with the accumulation of additional sarcomeres and, therefore, stimulates transcription and translation of the endogenous TnT. 4. Discussion This study was designed to compare two commonly used viral promoters and a synthetic promoter in an adenoviral context with regard to their expression in primary cardiac cells and their regulation by pharmacological intervention that are commonly used in cardiac myocytes. The CMV promoter, thought to be maximally active at basal conditions, has been chosen for many gene transfer experiments. However, it has more recently been shown to be regulated by many stimuli, such as a interferon (downregulation), PMA/forskolin (induction), and methyl methanesulfonate (induction) in different cells types and has also
829
been shown to be inactivated after in vivo gene transfer into the mouse liver [6,16–18]. In this study, we show its dramatic induction by PMA and forskolin in neonatal CMCs. The mechanism of this induction remains unknown, but a number of hypotheses exist. An increase in the actual uptake of the complexes or adenoviruses and, thereby, increase in gene transfer efficiency is unlikely because the substances are added 18 h after the transfection/infection and the cells are washed with fresh media removing any external virus. One might argue about complexes or adenovirus still attached to the cell membrane that could then be taken up. It has been shown, however, that adenovirus uptake into cells is very rapid and should be completed after a few hours [19]. Furthermore, this induction does not occur for the RSV and MEF2/hsp68 promoters. The induction of the CMV promoter by PMA and forskolin is also not likely to be merely due to generalized increase in transcription because the effect is also observed when cellular growth is maximally induced by FBS. An increase of the half-life of luciferase in mammalian cells of about 3 h is also not likely since the effect is also seen with b-galactosidase, a protein with a half-life of more than 20 h [20,21]. A more likely explanation is the induction of specific transcription factors stimulating transcription from the CMV early promoter. Binding sites for multiple transcription factors have been identified in this promoter, some being implicated in stimulation (NFjB, CREBP) and another (YY1) is thought to be a negative regulator [22]. Phorbol esters are inducers of the signal transduction molecule protein kinase C that induces multiple cellular pathways including the transcription factor NFjB in CMCs [23]. Forskolin leads to increased cellular cAMP levels that induce protein kinase A and the transcription factor CREBP. Adult cells, however, seem to have a much higher basal transcription from the CMV promoter, producing about 10 times more luciferase per cell than neonatal cells at the same MOI. A possible explanation lies in the higher expression of a negative transcription factor in neonatal vs. adult cells with the transcription factor YY1 being a candidate. Neonatal CMCs express high levels of YY1, whereas expression levels in adult cells is low unless there is a pathological state (C.S. Long, personal communication; Ref. [24]). An interesting and surprising finding of this study is that the induction of the CMV promoter is not restricted to cells that already express a gene from this promoter to a detectable extent, but that a high percentage of neonatal rat CMCs are “silently infected”. This transcriptional inactivity can be overcome with treatment with PMA and/or forskolin. Adult rat CMCs, on the other hand, express CMV-driven b-galactosidase in the majority of cells already at a low MOI of 10 and have a higher basal activity. Inactivation of the CMV promoter has been described in vivo after adenoviral infection of the mouse liver. It could be reactivated after reinfection of the liver with an “empty” (i.e. transgene- and promoter-less) replication-deficient adenovirus. A transient induction of the transcription factor NFjB after adenoviral infection was shown to be the likely molecular mechanism
830
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831
[18]. This transcription factor is expressed at low levels in CMCs that increase with (patho)physiology, such as ischemia, aging, or hypertrophy [25,26]. The other transcription factors with demonstrated binding to the CMV promoter are also of low abundance/activity in untreated CMCs, possibly leaving an imbalance between positive and negative transcriptional regulators towards the latter [27]. The RSV promoter has also been extensively used for gene transfer to the myocardium. This promoter, however, is less well characterized. In transgenic animals it is highly expressed in cardiac and skeletal muscle [28]. In contrast to the CMV promoter, it has been shown not to be responsive to increased cAMP levels [29]. Transcription from this promoter has been shown to be dependent on YY1, which, under certain circumstances, can function as a transcriptional activator [30]. This is in striking contrast to the CMV promoter and might be an explanation for the observed differences in basal promoter activity between neonatal and adult CMCs. The synthetic MEF2-responsive promoter did not behave as expected from the transgenic animal results [8]. We expected activity to be low in CMCs under basal conditions and in fibroblasts but to be induced by hypertrophic stimuli. We did, however, observe high-basal activity, even in fibroblasts and little responsiveness. A possible explanation is transactivation by a cryptic promoter in the adenovirus genome rather than the hsp68–MEF2 promoter [31]. Further experiments are aimed at insulating this promoter and using a plasmid construct to test this hypothesis. The 60-fold induction of the CMV promoter in neonatal CMCs by agonists suggests using caution when employing this promoter as a normalization for studies involving promoter analysis of other cellular promoters. This phenomenon is not restricted to cardiac cells but is also found in other cells. As stated earlier, an inactivation of the CMV promoter has also been demonstrated in vivo [18]. Therefore, when using this promoter in an in vivo setting, it is important to be aware of this phenomenon. We and others have seen a significant induction of the CMV promoter after i.m. injection in a model of muscle degeneration/regeneration (Allen at al., unpublished observations; Ref. [32]). In addition to activation of promoters by physiological or pathological stimuli, promoter inactivation is an increasingly recognized phenomenon and is a vector-independent problem that might inhibit therapeutic use of several promoters [33]. The relevance of our reporter gene experiments for gene transfer was confirmed with one the contractile protein, TnT, that was Myc-tagged to differentiate it from the endogenous protein. Infection of neonatal CMCs with this adenovirus led to replacement of the endogenous protein that was dose dependent. As stoichiometry of contractile proteins cannot be altered in cells, overexpression cannot be achieved at protein levels but endogenous protein is replaced by adenovirally expressed proteins [15]. At very low MOIs, the adenovirally expressed protein was only 2% of total, whereas stimulation increased this to 33–35%. This result emphasizes the use of the CMV promoter as an inducible system for relevant
proteins and might be important for potentially toxic gene products where expression can be timed by induction. For contractile proteins, exact titration of protein expression is important for the study of mutated proteins that are thought to act through dominant-negative interaction with wild-type proteins [15]. Despite the cautionary statements made above, we believe the properties of these promoters can be exploited in cardiac cells in the following fashion: The CMV promoter is inducible in both primary cardiac fibroblasts and neonatal CMCs. It shows low basal activity in these cells and could be used when high-level inducible expression of a gene is desired. It might be very useful for adult cardiac myocytes to induce dose-dependent but noninducible expression. The RSV promoter shows higher basal activity than CMV and less inducibility in neonatal CMCs and very low activity in fibroblasts, making it a promoter of choice if it is used as a normalizing standard for activity of cardiac promoters. It shows both cardiac selectivity and it is particularly useful if (pharmacological) treatment of cells is required. It is less active in adult CMCs than CMV. The synthetic MEF2-responsive promoter shows the least inducibility and high-basal activity in all cell types (higher than RSV in non-myocytes) and may be a good promoter to use as a control in co-infection experiments when put into an adenoviral context as in our construct. In addition, we have confirmed that CMCs are very susceptible to adenoviral infection in that virtually all neonatal cells are transgene positive at very low MOIs but if the CMV promoter is used, it needs specific transcriptional activation. These results emphasize that the current focus on finding optimal vector systems needs to be extended by characterization of the regulatory elements in gene transfer constructs to optimize gene expression in the target tissues.
Acknowledgements This work was supported by grants from the Deutsche Forschungsgemeinschaft (Ma 2185/1-1 to A. Maass and TPB6, SFB355 to L. Neyses) and by NIH grant HL50560 to. L.A. Leinwand. The authors would like to thank R.C. Thompson and M. Buvoli for continuous discussion and support.
References [1]
Kirshenbaum LA, MacLellan WR, Mazur W, French BA, Schneider MD. Highly efficient gene transfer into adult ventricular myocytes by recombinant adenovirus. J Clin Invest 1993;92:381–7.
[2]
Stratford-Perricaudet LD, Makeh I, Perricaudet M, Briand P. Widespread long-term gene transfer to mouse skeletal muscles and heart. J Clin Invest 1992;90:626–30.
A. Maass et al. / Journal of Molecular and Cellular Cardiology 35 (2003) 823–831 [3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
Kass-Eisler A, Falck-Pedersen E, Alvira M, Rivera J, Buttrick P, Wittenberg B, et al. Quantitative determination of adenovirusmediated gene delivery to rat cardiac myocytes in vitro and in vivo. Proc Natl Acad Sci USA 1993;90:11498–502. Zhang JCL, Woo YJ, Chen J-A, Swain JL, Sweeney HL. Efficient transmural cardiac gene transfer by intrapericardial injection in neonatal mice. J Mol Cell Cardiol 1999;31:721–32. Barr E, Carroll J, Kalynych AM, Tripathy SK, Kozarsky K, Wilson JM, et al. Efficient catheter-mediated gene transfer into the heart using replication-defective adenovirus. Gene Ther 1994;1:51–8. Clesham GJ, Browne H, Efstathiou S, Weissberg PL. Enhancer stimulation unmasks latent gene transfer after adenovirus-mediated gene delivery into human vascular smooth muscle cells. Circ Res 1996;79: 1188–95. Pasleau F, Tocci MJ, Leung F, Kopchick JJ. Growth hormone gene expression in eukaryotic cells directed by Rous sarcoma virus long terminal repeat or cytomegalovirus immediate-early promoter. Gene 1985;38:227–32. Naya FJ, Wu C, Richardson JA, Overbeek P, Olson EN. Transcriptional activity of MEF2 during mouse embryogenesis with a MEF2dependent transgene. Development 1999;126:2045–52. Passier R, Zeng H, Frey N, Naya FJ, Nicol RL, McKinsey TA, et al. CaM kinase signaling induces cardiac hypertrophy and activates the MEF2 transcription factor in vivo. J Clin Invest 2000;105:1395–406. Simpson P, McGrath A, Savion S. Myocyte hypertrophy in neonatal rat heart cultures and its regulation by serum and catecholamines. Circ Res 1982;51:787–801. Wollnik B, Kubisch C, Maass A, Vetter H, Neyses L. Hyperosmotic stress induces immediate-early gene expression in ventricular adult cardiomyocytes. Biochem Biophys Res Commun 1993;194:642–6. DeWet JR, Wood KV, DeLuca M, Helinski DR, Subramani S. Cloning of the firefly luciferase cDNA and its expression in mammalian cells. Mol Cell Biol 1987;7:725–37. Becker TC, Noel RJ, Coats WS, Gomez-Foix AM, Alam T, Gerard RD, et al. Use of recombinant adenovirus for metabolic engineering of mammalian cells. Meth Cell Biol 1994;43:161–89. Schaack J, Langer S, Guo X. Efficient selection of recombinant adenoviruses by vectors that express beta-galactosidase. J Virol 1995; 69:3920–3. Tardiff JC, Factor SM, Tompkins BD, Hewett TE, Palmer BM, Moore RL, et al. A truncated troponin T molecule in transgenic mice suggests multiple cellular mechanisms for familial hypertrophic cardiomyopathy. J Clin Invest 1998;101:2800–11. Acsadi G, O’Hagan D, Lochmüller H, Prescott S, Larochelle N, Nalbantoglu J, et al. Interferons impair early transgene expression by adenovirus-mediated gene transfer in muscle cells. J Mol Med 1998; 76:442–50. Nabirochkina EN, Ivanov AV. Treatment by methyl methanesulfonate induces up-regulation of cytomegalovirus immediate/early promoter. Biotechniques 2000;29:732–6.
831
[18] Loser P, Jennings GS, Strauss M, Sandig V. Reactivation of the previously silenced cytomegalovirus major immediate-early promoter in the mouse liver: involvement of NFkappaB. J Virol 1997;72: 180–90. [19] Greber UF, Willets M, Webster P, Helenius A. Stepwise dismantlingof adenovirus 2 during entry into cells. Cell 1993;75:477–86. [20] Thompson JF, Hayes LS, Lloyd DB. Modulation of firefly luciferase stability and impact on studies of gene regulation. Gene 1991;103: 171–7. [21] Bachmair A, Finley D, Varshavsky A. In vivo half-life of a protein is a function of its amino-terminal residue. Science 1986;234:179–86. [22] Liu R, Baillie J, Sissons JGP, Sinclair JH. The transcription factor YY1 binds to negative regulatory elements in the human cytomegalovirus major immediate early enhancer/promoter and mediates repression in non-permissive cells. Nucl Acids Res 1994;22:2453–9. [23] Rouet-Benzineb P, Gontero B, Dreyfus P, Lafuma C. Angiotensin II induces nuclear factor-jB activation in cultured neonatal rat cardiomyocytes through protein kinase C signaling pathway. J Mol Cell Cardiol 2000;32:1767–78. [24] Patten M, Wang W, Aminololama-Shakeri S, Burson M, Long CS. IL-1b increases abundance and activity of the negative transcriptional regulator yin yang-1 (YY1) in neonatal rat cardiac myocytes. J Mol Cell Cardiol 2000;32:1341–52. [25] Helenius M, Hänninen M, Lehtinen SK, Salminen A. Aging-induced up-regulation of nuclear binding activities of oxidative stress responsive NF-jB transcription factor in mouse cardiac muscle. J Mol Cell Cardiol 1996;28:487–98. [26] Purcell NH, Tang G, Yu C, Mecurio F, DiDonato JA, Lin A. Activation of NF-jB is required for hypertrophic growth of primary rat neonatal ventricular cardiomyocytes. Proc Natl Acad Sci USA 98:6668–73. [27] Muller FU, Neumann F, Schmitz W. Transcriptional regulation by cAMP in the heart. Mol Cell Biochem 2000;212:11–7. [28] Conti FG, Powell R, Pozzi L, Zezze G, Faraggiana T, Gannon F, et al. A novel line of transgenic mice (RSV/LTR-bGH) expressing growth hormone in cardiac and striated muscle. Growth Regul 1995;5:101–8. [29] Mellon PL, Clegg CH, Correll LA, McKnight GS. Regulation of transcription by cyclic AMP-dependent protein kinase. Proc Natl Acad Sci USA 1989;86:4887–91. [30] Bhalla SS, Robitaille L, Nemer M. Cooperative activation by GATA-4 and YY1 of the cardiac B-type natriuretic peptide promoter. J Biol Chem 2001;276:11439–45. [31] Buvoli M, Langer SJ, Bialik S, Leinwand LA. Potential limitations of transcription terminators used as transgene insulators in adenoviral vectors. Gene Ther 2002;9:227–31. [32] Danko I, Fritz JD, Jiao S, Hogan K, Latendresse JS, Wolff JA. Pharmacological enhancement of in vivo foreign gene expression in muscle. Gene Ther 1994;1:114–21. [33] Bestor TH. Gene silencing as a threat to the success of gene therapy. J Clin Invest 2000;105:409–11.