Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Aug 15.
Published in final edited form as: Gene Ther. 2008 May 22;15(15):1107–1115. doi: 10.1038/gt.2008.83

Cell-specific nuclear import of plasmid DNA in smooth muscle requires tissue-specific transcription factors and DNA sequences

AM Miller 1, DA Dean 1,2
PMCID: PMC3744158  NIHMSID: NIHMS99581  PMID: 18496575

Abstract

Two shortcomings of nonviral gene therapy are a lack of tissue-specific targeting of vectors and low levels of gene transfer. Our laboratory has begun to address these limitations by designing plasmids that enter the nucleus of specific cell types in the absence of cell division, thereby enhancing expression in a controlled manner. We have shown that a 176 bp portion of the smooth muscle γ-actin (SMGA) promoter can mediate plasmid nuclear import specifically in smooth muscle cells (SMCs). Here, we demonstrate that the binding sites for serum response factor (SRF) and NKX3-1/3-2 within this DNA nuclear targeting sequence (DTS) are required for plasmid nuclear import. Knockdown of these factors with siRNA abrogates plasmid nuclear import, indicating that they are necessary cofactors. In addition, coinjection of recombinant SRF and Nkx3.2 with the vector in TC7 epithelial cells rescues import. Finally, we show that the SRF nuclear localization sequence (NLS) is required for vector nuclear import. We propose that SRF and NKX3-1/3-2 bind the SMGA DTS in the cytoplasm, thus coating the plasmid with NLSs that mediate translocation across the nuclear pore complex. This discovery could aid in the development of more efficient nonviral vectors for gene transfer to SMCs.

Keywords: nuclear import, plasmid, nuclear localization sequence, transfection, gene delivery

Introduction

Nonviral gene therapy has achieved only modest therapeutic success due to low levels of gene transfer. This is not surprising, given the numerous cellular barriers that hinder the delivery of nonviral vectors to the nucleus. In nondividing cells, perhaps the greatest barrier is the nuclear envelope. Using microinjection, several groups have demonstrated that in the absence of cell division, pDNA injected into the cytoplasm achieves very little expression compared to equivalent amounts of pDNA injected directly into the nucleus.1–3 Unlike a dividing cell where the breakdown of the nuclear envelope permits DNA to enter the nucleus and undergo transcription, in nondividing cells this barrier remains intact and precludes nuclear delivery. As most target tissues are nondividing in vivo, the nuclear membrane is a major obstacle to gene transfer and must be circumvented to improve expression from nonviral vectors.3–8

Our laboratory has begun to design plasmid vectors that achieve efficient gene expression by targeting constructs to the nuclear compartment of specific cell types.9–13 Initially, we demonstrated that inclusion of as little as 72 bp of the simian virus 40 (SV40) promoter/enhancer in plasmid vectors facilitates nuclear import in all cell types tested.12 Due to the ability of this sequence to target DNA to the nucleus, we refer to it as a DNA nuclear targeting sequence (DTS). We, and others, have subsequently shown that nuclear import of plasmid DNA is sequence dependent, requires karyopherins (importins) and the small GTPase RAN, and occurs through the nuclear pore complex (NPC).14–18 In our model of plasmid nuclear import, we propose that transcription factors present in the cytoplasm bind to the DTS and coat the plasmid with nuclear localization sequences (NLSs) that utilize importins to cross the NPC, which regulates the nucleocytoplasmic shuttling of macromolecules during interphase.19–21 We have also developed a tissue-specific DTS, utilizing the smooth muscle γ-actin (SMGA) promoter, that mediates nuclear import of pDNA specifically in smooth muscle cells (SMCs).10 The SMGA DTS was shown to drive SMC-specific plasmid nuclear import in rat mesenteric vessels following electroporation into living animals, proving its viability for in vivo gene transfer. In addition, inclusion of the SMGA DTS in a luciferase construct increased the expression by 40-fold in arrested cells.10 On the basis of the presence of binding sites for several SMC-specific transcription factors in the SMGA DTS, we hypothesize that SMC-specific factors bind the SMGA DTS to mediate cell-specific nuclear import. Indeed, overexpression of SRF in non-SMCs increased import, supporting this model.10 Despite the progress we have made using the SMGA DTS to enhance SMC-specific gene transfer, we have yet to confirm the mechanism of import.

In this study, we have determined that plasmid nuclear import in SMCs is sequence-specific, requiring the binding sites for serum response factor (SRF) and the NK3 family of transcription factors within the SMGA DTS. In addition, we demonstrate that these smooth muscle-specific transcription factors are necessary and sufficient to drive import of pDNA containing the SMGA DTS. Finally, we show that a functional NLS on SRF is required to mediate import. Further investigation of this route of gene transfer could facilitate the development of SMC-specific gene therapy vectors for therapeutic use.

Results

Nuclear import of pDNA in smooth muscle is sequence specific

We employed a microinjection strategy to ascertain the sequence requirements for plasmid nuclear import mediated by the SMGA DTS in human pulmonary artery SMCs (hSMCs). Previously, we demonstrated that a 404 bp portion of the SMGA promoter cloned into the pGL3-basic plasmid (Promega, Madison, WI, USA) is sufficient to drive plasmid nuclear import.10 This sequence has been shown to contain binding sites (SRE2, NKE1 and SRE1) for the SMC-specific transcription factors SRF and Nkx3.1 (a Drosophila bagpipe homologue) by gel shift analysis.22 The other known mammalian NK3 family member, Nkx3.2, is also known to bind to the NKE1 consensus element.23 These transcription factors are coexpressed only in vascular SMCs and bear NLSs that allow them to shuttle into the nucleus. This lead us to hypothesize that these factors bind to the SMGA DTS in SMCs and mediate import specifically in this tissue.24–26 By extension, mutation of these binding sites should preclude binding of SRF and NK3 family members to the DTS and prevent plasmid nuclear import. To test this hypothesis, we first demonstrated that a 176 bp proximal fragment of the SMGA promoter that contains these sites was also able to mediate nuclear import of a 5 kbp plasmid in hSMCs. hSMCs were cytoplasmically microinjected with pSMGA176 DTS and assessed for pDNA nuclear import 8 h later by in-situ hybridization (Figure 1). Next, we cytoplasmically microinjected hSMCs with pGL3 vectors containing an SRE2 mutant, an NKE1 mutant or an SRE1 mutant and 8 h later the localization of the plasmid DNA was determined similarly. As a negative control for import, pBR322 DNA (which lacks a DTS) was microinjected to determine baseline levels of plasmid nuclear import that results from cell division during the experiment or inadvertent injection of pDNA directly into the nuclei of cells (data not shown). A mutation of the SRE2, NKE1 or SRE1 binding sites in the 176 bp SMGA DTS decreased the level of import to baseline, demonstrating that plasmid nuclear import in SMCs is sequence specific, requiring the binding sites for SRF and NK3 transcription factors (Figure 1). These results strongly suggest that the SMC-specific transcription factors SRF and NKX3-1/3-2 are involved in the nuclear import of pDNA containing the SMGA promoter.

Figure 1.

Figure 1

Open in a new tab

Sequence requirements for plasmid nuclear import in smooth muscle cells (SMCs). (a) For microinjection experiments, plasmid DNA (0.5 mg ml−1) was microinjected into the cytoplasm of adherent cells and the localization of the DNA was determined 8 h later using fluorescence microscopy (Nuclear Import vs No Import). (b) Plasmid constructs containing the: (1) SMGA176 DTS, (2) SRE2 mutant, (3) NKE1 mutant or (4) SRE1 mutant were cytoplasmically microinjected into human pulmonary artery SMCs (hSMCs) and 8 h later the cellular localization of the constructs was determined using in-situ hybridization. At least 100 cells were scored for each condition and the experiment was repeated three times (mean % nuclear import±s.d.). *P<0.001 vs pBR322 control.

The SMC-specific transcription factors SRF and NKX3-1/3-2 are required for import

Although mutation of the binding sites for SRF and NKX3-1/3-2 in the SMGA DTS eliminated plasmid nuclear import in SMCs, this finding does not substantiate that they are necessary for import. Therefore, we decided to use siRNA to knockdown the expression of these proteins in hSMCs and determine the effect on the nuclear import of microinjected pDNA. First, immunoblots of total cell lysates from hSMCs transfected with siRNAs against SRF and the two mammalian NK3 homologues (NKX3-1 and NKX3-2), revealed successful knockdown of each transcription factor at 48 h (Figure 2a). In all cases, knockdown of the transcription factors was between 70 and 90% (Figure 2a and not shown). Following optimization of the siRNA protocol, cells were treated with siRNA for 48 h and then cytoplasmically microinjected with: (1) pBR322 (negative control for plasmid import), (2) pCMV-GFP-SV40 DTS (positive control for plasmid import) or (3) pCMV-GFP-SMGA404 DTS that had been labeled with a Cy3-PNA clamp. After 8 h, the cellular localization of the pDNA was determined using fluorescence microscopy in living cells. Control cytoplasmic injections with pBR322 DNA established baseline levels of plasmid import that resulted from cell division or inadvertent microinjection of pDNA directly into nuclei (Figure 2b, empty bar). Cells treated with nontargeting siRNAs (NT) that were microinjected with pCMV-GFP-SMGA404 DTS had significantly elevated levels of nuclear import (45.3±6.4%) compared to pBR322 control (7.7±1.5%), demonstrating that the SMGA DTS successfully mediates import in treated cells. In contrast, cells treated with SRF siRNA, exhibited significantly reduced levels of import (27.4±1.4%) compared to the NT control. For cells treated individually with either NKX3-1 or NKX3-2 siRNA, import was not affected (44.7±2.5 and 41.8±3.7%, respectively). However, when both NK3 factors were knocked down, levels of import of cytoplasmically microinjected pCMV-GFP-SMGA404 DTS (23.5±3.1%) were significantly reduced compared to the NT control, suggesting that the NK3 homologues compensate for each other. When all three factors were knocked down, import of cytoplasmically microinjected pCMV-GFP-SMGA404 DTS (11.7±4.8%) was reduced to baseline levels. These results indicate that SRF and NK3 transcription factors are necessary to drive plasmid nuclear import in SMCs and likely operate in a cooperative manner. Finally, to rule out the possibility that knockdown of SRF and/or the NK3 factors could inhibit plasmid nuclear import by an indirect mechanism (for example, by somehow altering nuclear pore structure), cells treated with siRNAs for all three factors (SRF, NKX3-1 and NKX3-2) were microinjected with pDNA containing the universal SV40 DTS (pCMV-GFP-SV40 DTS). The SV40 DTS is bound by a separate set of ubiquitously expressed transcription factors that then facilitate nuclear import through NPC and therefore should not be affected by the knockdown of SRF and NKX3-1/3-2. This plasmid successfully translocated into the nucleus of hSMCs treated with all three siRNAs (50.5±4.6%), indicating that plasmid nuclear import is not grossly compromised by SRF and NK3 factor knockdown (Figure 2b, gray bars).

Figure 2.

Figure 2

Open in a new tab

Serum response factor (SRF) and NK3 transcription factors are required for nuclear import in smooth muscle cells (SMCs). (a) Immunoblots of human pulmonary artery SMCs (hSMCs) transfected with nontargeting (NT) siRNA or siRNA targeting SRF, NKX3-1 and NKX3-2 reveal successful knockdown of each factor at 48 h. (b) pBR322, pCMV-EGFP-SV40 DTS and pCMV-EGFP-SMGA404 DTS (0.5 mg ml−1) were labeled with Cy3-PNA and cytoplasmically microinjected into hSMCs that were treated with siRNA for 48 h. After 8 h, the cellular localization of the constructs was determined using fluorescence microscopy. At least 100 cells were injected for each condition and the experiment was repeated three times (mean % nuclear import±s.d.). *P<0.001 vs NT control.

Recombinant His-SRF and GST-Nkx3.2-HA are sufficient to drive import

We have previously shown that pDNA constructs containing the SMGA DTS microinjected into the cytoplasm of TC7 epithelial cells do not undergo nuclear import.10 Immunoblots of total cell lysates of TC7 cells reveal that SRF, NKX3-1 and NKX3-2 are not expressed in this cell line, providing further evidence that these factors are necessary for import (Figure 3a). To test if SRF and NK3 transcription factors are sufficient to drive nuclear import of pDNA containing the SMGA DTS in TC7 cells, we expressed recombinant His-SRF and GST-Nkx3.2-HA in bacteria, purified these factors and complexed them (and/or bovine serum albumin (BSA)) with: (1) pBR322 (negative control for plasmid import), (2) pCMV-GFP-SV40 DTS (positive control for plasmid import) or (3) pCMV-GFP-SMGA404 DTS that had been labeled with a Cy3-PNA clamp, and injected the complexes into the cytoplasm of TC7 cells. After 8 h, the cellular localization of the pDNA was determined using fluorescence microscopy in living cells (Figure 3b). The levels of nuclear import of pCMV-GFP-SV40 DTS were significantly higher than the pBR322 control across all conditions and were not affected by the coinjection of recombinant protein (gray bars vs empty bars). Coinjection of control protein (BSA), GST-Nkx3.2-HA or His-SRF alone with pCMV-GFP-SMGA404 DTS yielded levels of nuclear import that were not statistically different from baseline. However, coinjection of both His-SRF and GST-Nkx3.2-HA with pCMV-GFP-SMGA404 DTS resulted in drastically increased levels of nuclear import in these non-SMCs, approximately twofold greater than that seen with pDNA containing the SV40 DTS. These results suggest that SRF and NK3 factors are sufficient to drive nuclear import of pDNA containing the SMGA DTS and that these factors act in a synergistic manner to rescue import in non-SMCs.

Figure 3.

Figure 3

Open in a new tab

Recombinant serum response factor (SRF) and Nkx3.2 are sufficient to rescue import in TC7 epithelial cells. (a) Immunoblots of cellular lysates reveal that SRF, NKX3-1 and NKX3-2 are expressed in human pulmonary artery smooth muscle cells (hSMCs) but not TC7 epithelial cells. (b) pBR322, pCMV-EGFP-SV40 DTS and pCMV-EGFP-SMGA404 DTS (0.5 mg ml−1) were labeled with Cy3-PNA and cytoplasmically microinjected with combinations of (1) bovine serum albumin (BSA), (2) His-SRF and/or (3) GST-Nkx3.2-HA (totaling 0.3 mg ml−1) into the cytoplasm of TC7 cells. After 8 h, the cellular localization of the constructs was determined using fluorescence microscopy. At least 100 cells were injected for each condition and the experiment was repeated three times (mean % nuclear import±s.d.). *P<0.001 vs pBR322 control. **P<0.001 vs SV40 DTS.

The SRF nuclear localization sequence is required for import

Although the siRNA experiments provide compelling evidence that SRF and NKX3-1/3-2 are required for SMC plasmid nuclear import, it is possible that knockdown of these transcription factors over a 48-h period inhibits nuclear import via an indirect mechanism (for example, altering the expression pattern of proteins regulated by these transcription factors). To antagonize import directly, we electroporated adherent hSMC with: (1) pCMV-EGFP-mNLS-SRF, which expresses a dominant-negative EGFP–SRF fusion protein that lacks a functional NLS or (2) pCMV-EGFP-SRF, which expresses the wild-type fusion protein. This allowed us to express pools of exogenous mutant or wild-type SRF in SMCs that could be tracked by fluorescence microscopy without altering the expression of the endogenous SRF. After 48 h, transfected cells were identified using live-cell fluorescent microscopy and were cytoplasmically microinjected with pCMV-GFP-SMGA404 DTS that had been labeled with a fluorescent Cy3-PNA clamp. After 8 h, the cellular localization of the fusion proteins and pDNA was visualized (Figure 4). As can be seen in the figure, EGFP-SRF colocalized with the pDNA in the nucleus, supporting our model in which EGFP-SRF binds to the vector and that the resulting pDNA–protein complex successfully undergoes nuclear import. In contrast, EGFP-mNLS-SRF colocalized with the microinjected pDNA in the cytoplasm in all cells observed, demonstrating that mutant SRF binds to the vector and sequesters it in the cytoplasm. These results imply that a functional SRF NLS is required to translocate the vector into the nucleus.

Figure 4.

Figure 4

Open in a new tab

Mutant serum response factor (SRF) lacking a functional nuclear localization sequence (NLS) sequesters pDNA in the cytoplasm of human pulmonary artery smooth muscle cells (hSMCs). Adherent hSMCs were electroporated with constructs expressing either EGFP-SRF or EGFP-mNLS-SRF. Forty-eight hours post transfection, cells expressing either the EGFP-SRF or EGFP-mNLS-SRF fusion protein were cytoplasmically microinjected with pCMV-EGFP-SMGA404 DTS (0.5 mg ml−1) labeled with Cy3-PNA. After 8 h, the cellular localization of the pDNA and the fusion protein was determined using fluorescence microscopy.

To provide a further line of evidence that the SRF NLS is required for the plasmid nuclear import mediated by the SMGA DTS, we injected Cy3-labeled pCMV-GFP-SMGA404 DTS precomplexed with: (1) BSA, (2) BSA, GST-Nkx3.2-HA and GST-SRF or (3) BSA, GST-Nkx3.2-HA and GST-mNLS-SRF into the cytoplasm of TC7 epithelial cells and quantified the levels of nuclear import (Figure 5). As we noted in the previous rescue experiment, plasmid DNA coinjected with BSA alone was not imported into the nucleus of TC7 cells (9.3±5.8%). In contrast, vector DNA coinjected with BSA, GST-Nkx3.2-HA and GST-SRF was imported in almost all of the injected cells (78.5±12.5%). However, when vector DNA was coinjected with BSA, GST-Nkx3.2-HA and GST-mNLS-SRF, levels of import returned to baseline levels (5.2±5.4%). Taken together, this data clearly demonstrates that the SRF NLS is necessary for import in SMCs and to rescue import in TC7 cells.

Figure 5.

Figure 5

Open in a new tab

The serum response factor (SRF) nuclear localization sequence (NLS) is required to rescue import in TC7 epithelial cells. pCMV-EGFP-SMGA404 DTS (0.5 mg ml−1) was labeled with Cy3-PNA and cytoplasmically microinjected with combinations of (1) bovine serum albumin (BSA), (2) GST-Nkx3.2-HA, (3) GST-SRF or (4) GST-mNLS-SRF (totaling 0.3 mg ml−1) into the cytoplasm of TC7 cells. After 8 h, the cellular localization of the constructs was determined using fluorescence microscopy. At least 100 cells were injected for each condition and the experiment was repeated three times (mean % nuclear import±s.d.). *P<0.001 vs BSA-only control.

Discussion

In this study, we demonstrate the sequence and protein cofactor requirements for the nuclear import of pDNA containing the SMGA DTS in SMCs. Using microinjection of mutant vectors into the cytoplasm of human vascular SMCs, we have shown that the binding sites for SRF and NKX3-1/3-2 within the SMGA DTS (SRE2, NKE1 and SRE1) are all required for nuclear import. All three of these transcription factors are coexpressed specifically in vascular SMCs, which provides a plausible explanation for why import of pDNA containing the SMGA DTS occurs specifically in SMCs.24–26 Consistent with this finding, when we knocked down SRF in hSMCs using siRNA, nuclear import of microinjected pDNA containing the SMGA DTS was inhibited. It should be noted that residual levels of SRF were still present despite siRNA treatment, which likely explains why import was not completely reduced to baseline (that is, nuclear import resulting from inadvertent injection into nuclei or cell division). Next, when either of the two mammalian bagpipe homologues (NKX3-1 and NKX3-2) was knocked down with siRNA, import was not effected. However, when both were jointly knocked down, import was inhibited. These results indicate that NKX3-1 and NKX3-2 likely play overlapping roles in mediating nuclear import. As residual levels of NKX3-1 and NKX3-2 were still present despite siRNA treatment, we again noted that nuclear import was not reduced to baseline. Interestingly, when all three factors were knocked down simultaneously, nuclear import was completely abrogated. These results imply that SRF and NK3 transcription factors act synergistically to drive plasmid nuclear import in SMCs. Indeed, although injection of recombinant His-SRF, GST-Nkx3.2-HA or control BSA individually complexed with pDNA containing the SMGA DTS into the cytoplasm of TC7 cells did not rescue import, injection of recombinant His-SRF and GST-Nkx3.2-HA complexed together with the vector into the cytoplasm of TC7 cells resulted in nuclear import in virtually all of the cells. This provides another line of evidence that SRF and NK3 factors act synergistically to mediate import.

Our finding that SRF and NK3 act synergistically to mediate plasmid nuclear import in SMCs is consistent with transcriptional studies of the regulation of the SMGA promoter. Several groups have shown that SRF physically interacts with the NK class of homeodomain proteins (NXX3-1, NKX3-2 and NKX2-5) via its MADS box motif to regulate transcription in smooth and cardiac muscle.22,26 Zimmer and coworkers first reported that the NKE1 site adjacent to the SRE1 element in the SMGA promoter binds Nkx3.1 and promotes binding of SRF to the promoter via a homeodomain–MADS box interaction.22 This combination of Nkx3.1 and SRF led to synergistic coactivation of the promoter in CV1 epithelial cells. In addition, they found that SRF exhibits only weak binding to the SMGA DTS unless Nkx3.1 was present.22 The homeodomain of Nkx3.2 has also been shown to physically interact with the SRF-MADS box and to bind to the NKE1 consensus element.23 In fact, Nishida et al.26 demonstrated that a triad of SRF, Nkx3.2 and GATA-6 formed a complex with their corresponding cis-elements and cooperatively transactivated smooth muscle genes, including α(1) integrin, SM22α and caldesmon. They concluded that NK families confer muscle specificity to the binding of SRF to SRE sites within these promoters. From these studies, it appears that NK3 transcription factors bind to the NKE1 site in the SMGA DTS and facilitate SRF recruitment.

Based on our findings, we have devised a working model of SMC-specific plasmid translocation (Figure 6). We propose that SRF and NK3 transcription factors bind to the DTS in the cytoplasm, thus attaching NLSs to the pDNA for recognition and transport across the NPC by the NLS-dependent import machinery.10,27,28 Similar NLS-mediated nuclear import models have been proposed for viral genomes, including HIV and adenovirus.29,30 In addition, several groups have shown that by complexing plasmids with proteins, such as histone proteins and nuclear factor κB, or with peptide nucleic acids (PNAs) that contain nuclear localization signals, nuclear import and expression are enhanced.31–34 As SRF and NK3 factors are coexpressed specifically in SMCs, other cell types are unable to form an import complex on the pDNA, making the SMGA promoter a tissue-specific DTS.

Figure 6.

Figure 6

Open in a new tab

Model for plasmid DNA nuclear import in smooth muscle cells (SMCs). SMC-specific plasmid nuclear import is accomplished by incorporating smooth muscle γ-actin (SMGA) promoter fragments that contain binding sites for transcription factors that are uniquely coexpressed in this tissue (that is, serum response factor (SRF) and NK3 factors). Once plasmid DNA has entered the cytoplasm, these factors bind to the vector and coat it with nuclear localization sequences (NLSs) that interact with importins to drive import through the nuclear pore complex (NPC). In other cell types, these factors are not expressed and therefore cannot form a protein–pDNA complex that is capable of nuclear import.

Implicit to our hypothesis is that either SRF or NK3 factors contain an NLS that remains exposed upon binding to the SMGA DTS in the cytoplasm, thus permitting interaction with the nuclear import machinery. Fortunately, the SRF NLS has been characterized and occupies a spatially distinct region on the protein than the DNA-binding region.35,36 This permitted the use of dominant-negative SRF constructs to investigate the importance of the NLS in mediating nuclear import of pDNA containing the SMGA DTS. When we microinjected SMCs expressing EGFP-mNLS-SRF (a dominant-negative fusion protein where the NLS has been mutated) with our plasmid vector, it was sequestered in the cytoplasm and failed to undergo import. In addition, although coinjection of our vector with GST-SRF and GST-Nkx3.2-HA in TC7 epithelial cells rescued import, coinjection of GST-mNLS-SRF and GST-Nkx3.2-HA with our vector failed to rescue import. This data indicates that the SRF NLS is required for nuclear import of our SMGA DTS vector.

Unfortunately, the NLS for NK3 transcription factors has not been mapped, preventing us from conducting a similar set of experiments for this class of transcription factor. However, a putative NLS does exist in the third helix of the NK3 homeodomain and is identical to the NLS previously characterized in another homeobox transcription factor, PDX-1.37 As this NLS is in the third helix of the homeodomain (that is, the region responsible for the protein’s DNA-binding activity), it is likely that the NLS is masked in the cytoplasm upon binding of the pDNA vector, thus precluding interaction with importins. This begs the question of why we found the NKE binding site and the NKX3-1/3-2 transcription factors to be necessary for nuclear import in our present study. We propose that NK3 is necessary to facilitate SRF binding to the vector for nuclear import, as has been shown in transcriptional studies of promoter transactivation.22,26 In the absence of NK3, there is insufficient affinity of SRF for the SMGA DTS and the vector does not get coated with the NLSs required for interaction with the endogenous import machinery.

Our results may also shed light on why some DNA elements, including the strong viral promoters for cytomegalovirus (CMV) and respiratory syncytial virus, despite the fact that they bind to multiple transcription factors, do not display plasmid nuclear import activity in cytoplasmically microinjected cells.10 It is possible that the transcription factors that bind to these sequences contain NLSs that are either hidden when bound to vector DNA in the cytoplasm or are spatially inaccessible to the importin family members. In future studies, we plan to investigate if certain classes of transcription factors (that is, those with distinct NLS and DNA-binding regions) are required to mediate plasmid nuclear import and if there is an ideal spatial arrangement of transcription factor binding sites within DTSs. Such knowledge should facilitate the development of future cell-specific DTSs for efficient gene transfer.

Materials and methods

Cell culture and electroporations

TC7 cells, a line of African Green Monkey kidney epithelial cells, and hSMCs were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 2 mml-glutamine. For electroporations, cells were grown to confluency in six-well dishes and rinsed twice in serum- and antibiotic-free media. Plasmid DNA (10 μg) in 500 μl of serum- and antibiotic-free DMEM were added to each well and one 165 mV square wave electric pulse was applied using a PetriPulser electrode (BTX, San Diego, CA, USA).

Plasmids

The pGL-3 basic plasmid (Promega) containing the 176 bp SMGA DTS sequence (pSMGA176 DTS) and the constructs bearing binding site mutations in the SRF and NK3 binding sites (SRE2, NKE1 and SRE1) were a gift from Warren Zimmer (Texas A&M University, College Station, TX, USA). The SMGA SRE1 was disrupted by changing the wild-type element from 5′-CCTATTTAGG-3′ to 5′-CCCTATCCCGG-3′ and the SRE2 sequence (5′-CCTATATGG-3′) mutated to 5′-CCTTATGTTT-3′. The SMGA NKE1 was disrupted by changing the wild-type element from 5′-CACTTAGCCT-3′ to 5′-CACCCCCCCT-3′, as previously described.22 Plasmids pCMV-GFP- SV40 DTS and pCMV-GFP- SMGA404 DTS express green fluorescent protein from the CMV immediate early promoter and contain the 72 bp SV40 DNA nuclear targeting sequence or the 404 bp SMGA DNA nuclear targeting sequence (respectively) downstream of the reporter gene.9,38 Derivatives of pBR322, pCMV-GFP-SV40 DTS and pCMV-GFP-SMGA404 DTS that contain tandem binding sites for the GeneGrip Cy3-labeled PNA (Gene Therapy Systems, San Diego, CA, USA) were used for microinjection studies and were labeled with Cy3-PNA as previously described.39 The pCMV-GFP-SMGA404 DTS was constructed by PCR-amplification of the 404 bp SMGA DTS from pCAT-404, along with flanking NotI and XbaI sites, and insertion into the corresponding sites of pEGFP-N1 (Clontech, Mountain View, CA, USA), where the SV40 enhancer had been replaced by the 101 bp PNA linker (Gene Therapy Systems).10 The EGFP–SRF fusion protein construct (pCMV-EGFP-SRF) and the dominant-negative fusion construct (pCMV-EGFP-mNLS-SRF), where amino acids 95R and 96R within the nuclear localization signal (95-RRGLKR-100) were mutated to 95E and 96E, were the kind gift of Julian Solway (University of Chicago, Chicago, IL, USA).40 The bacterial expression construct for His-SRF (pET-SRF) and GST-SRF (pGEX-SRF) were a gift from Warren Zimmer (Texas A&M University) and the GST-Nkx3.2-HA bacterial expression construct (pGEX-Nkx3.2-HA) was a gift from Andrew Lassar (Harvard University, Boston, MA, USA).22,23 The bacterial expression construct for GST-mNLS-SRF was created by excising the mutant NLS sequence from pCMV-EGFP-mNLS-SRF using flanking XmaI sites and ligating the fragment into the corresponding region of pGEX-SRF where the wild-type NLS had been excised. All plasmids were purified from Escherichia coli using Promega Maxiprep kits as described by the manufacturer (Promega).

Expression and purification of bacterially expressed recombinant protein

Constructs for His-SRF (pET-SRF), GST-SRF (pGEX-SRF), GST-mNLS-SRF (pGEX-mNLS-SRF) and GST-Nkx3.2-HA (pGEX-Nkx3.2-HA) were transformed into BL21 (DE3) cells (Novagen, San Diego, CA, USA). Freshly transformed cells were grown in 500 ml of Luria broth containing 100 μg ml−1 of ampicillin at 37 °C for 2 h. Isopropyl-β-d-thiogalactopyranoside was added at a 1 mm concentration and growth was allowed to continue for 4 h. Cells were harvested, resuspended in column buffer (20 mm 4-(2-hydroxyethyl)-1-piperazineethanesul-fonic acid, 135 mm potassium acetate, 10 mm sodium acetate, 2.5 mm magnesium acetate) and sonicated, and cellular debris were removed by centrifugation. The fusion proteins were bound to Ni-NTA agarose beads (Qiagen, Valencia, CA, USA) or glutathione-sepharose beads (Amersham Biosciences, Piscataway, NJ, USA), packed into a column, washed extensively and eluted with fractions of column buffer containing 500 mm imidazole or 10 mm glutathione, respectively. Eluates were concentrated using YM-3 Centricon devices (Millipore, Billerica, MA, USA) and their final concentration was determined by Bradford protein assay and SDS–polyacrylamide gel electrophoresis (PAGE) followed by Coomassie staining.

siRNA methods

Pools of ON-TARGETplus siRNAs for human SRF (no. L-009800-00), NKX3-1 (no. L-015422-00), NKX3-2 (no. L-011216-00) and nontargeting siRNAs (no. D-001810-10-05) were purchased from Dharmacon (Lafayette, CO, USA). According to the manufacturer’s instructions, hSMCs were plated at 30% confluency 24 h prior to transfection and DharmaFECT reagent 2 (no. T-2002-03) was used to transfect cells with 100 nm of each siRNA. Transfected knockdown cells were allowed to grow for 48 h prior to microinjection.

Immunoblotting

Cells were lysed in radioimmuno precipitation assay buffer (10 mm Tris-HCL pH 7.5, 1% sodium deoxycho-late, 1% NP40, 150 mm NaCl) and concentrated using YM-3 centricon devices (Millipore). Concentrated samples were then boiled 1:1 in Laemmli sample buffer. Extracts were separated by SDS–PAGE, followed by electrotransfer to nitrocellulose membranes. Mouse monoclonal anti-SRF antibody (Millipore), rabbit polyclonal anti-Nkx3.1 antibody (Santa Cruz Biotechnologies, Santa Cruz, CA, USA), goat polyclonal anti-Nkx3.2 antibody (Santa Cruz Biotechnologies) and horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnologies) were used for western blots followed by chemiluminescence detection, according to the manufacturer’s instructions (Pierce Chemical Co., Rockford, IL, USA).

Microinjection

TC7 cells and hSMCs were grown on cover slips and microinjected with plasmid constructs (0.5 mg ml−1 in phosphate buffered saline) along with recombinant protein or BSA (0.1 mg ml−1 per transcription factor in addition to BSA to bring total protein concentration to 0.3 mg l−1) using an Eppendorf Femtojet system as previously described.39

In-situ hybridizations

Cells were rinsed with 1× phosphate buffered saline (PBS; 10 mm Phosphate, 137 mm NaCl, 2.7 mm KCl, pH 7.4), permeabilized in 1× PBS containing 0.5% Triton X-100 for 45 s, fixed at −20 °C in a 1:1 methanol:acetone solution for 5 min, and placed in 70% ethanol at 4 °C overnight. Plasmid in-situ hybridizations were carried out using nick-translated DNA probe from appropriate plasmid backbones with an Alexa 488 fluorophore (Invitrogen, Carlsbad, CA, USA) as previously described.12 Coverslips were mounted on slides with 4′,6-diamidino-2-phenylindole dihydrochloride and the anti-fade reagent, DABCO (Invitrogen).

Microscopy and image analysis

Images were captured in live cells on a Leica DMIRB inverted microscope with a ×40 objective. Nuclei were stained with Hoescht 33258 and detected using a 340–380 band-pass excitation filter and a 425 nm long pass filter, and Cy3-labeled plasmid DNA was detected using a 545 nm excitation filter and a 610/75 nm band-pass filter and OpenLab software (Improvision, Lexington, MA, USA). Fixed cells and in-situ hybridizations were observed under a Leica DMRXA2 epifluorescence microscope with a ×100 objective (NA 1.35; Leica). Images were acquired using a Hamamatsu Orca-ER 12-bit, cooled CCD camera (Hamamatsu, Japan) using OpenLab software.

Statistical analysis

Results are expressed as mean±s.d. of the mean. Differences between multiple groups were analyzed using one-way analysis of variance with the Student–Newman–Keuls post hoc test for all pairwise comparisons using SigmaStat software (SPSS, Chicago, IL, USA).

Acknowledgements

We thank James DeGiulio, Erin Vaughan and Warren Zimmer for insightful discussions and technical advice. We also thank Andrew Lassar, Julian Solway and Warren Zimmer for the generous gifts of constructs. This work was supported in part by grants HL59956 and by a predoctoral fellowship from the American Heart Association, Greater Midwest Affiliate (AMM).

References

  • 1.Capecchi MR. High efficiency transformation by direct micro-injection of DNA into cultured mammalian cells. Cell. 1980;22:479–488. doi: 10.1016/0092-8674(80)90358-x. [DOI] [PubMed] [Google Scholar]
  • 2.Graessman M, Menne J, Liebler M, Graeber I, Graessman A. Helper activity for gene expression, a novel function of the SV40 enhancer. Nucleic Acids Res. 1989;17:6603–6612. doi: 10.1093/nar/17.16.6603. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Zabner J, Fasbender AJ, Moninger T, Poellinger KA, Welsh MJ. Cellular and molecular barriers to gene transfer by a cationic lipid. J Biol Chem. 1995;270:18997–19007. doi: 10.1074/jbc.270.32.18997. [DOI] [PubMed] [Google Scholar]
  • 4.Brunner S, Sauer T, Carotta S, Cotten M, Saltik M, Wagner E. Cell cycle dependence of gene transfer by lipoplex, polyplex and recombinant adenovirus. Gene Therapy. 2000;7:401–407. doi: 10.1038/sj.gt.3301102. [DOI] [PubMed] [Google Scholar]
  • 5.Liu F, Song Y, Liu D. Hydrodynamics-based transfection in animals by systemic administration of plasmid DNA. Gene Therapy. 1999;6:1258–1266. doi: 10.1038/sj.gt.3300947. [DOI] [PubMed] [Google Scholar]
  • 6.Barron LG, Szoka FC. The perplexing delivery mechanism of lipoplexes. In: Huang L, Hung M-C, Wagner E, editors. Nonviral Vectors for Gene Therapy. Academic Press; San Diego: 1999. pp. 230–266. [Google Scholar]
  • 7.Godbey WT, Barry MA, Saggau P, Wu KK, Mikos AG. Poly(ethylenimine)-mediated transfection: a new paradigm for gene delivery. J Biomed Mater Res. 2000;51:321–328. doi: 10.1002/1097-4636(20000905)51:3<321::aid-jbm5>3.0.co;2-r. [DOI] [PubMed] [Google Scholar]
  • 8.Grosse S, Thevenot G, Monsigny M, Fajac I. Which mechanism for nuclear import of plasmid DNA complexed with polyethy-lenimine derivatives? J Gene Med. 2006;8:845–851. doi: 10.1002/jgm.915. [DOI] [PubMed] [Google Scholar]
  • 9.Dean DA, Dean BS, Muller S, Smith LC. Sequence requirements for plasmid nuclear entry. Exp Cell Res. 1999;253:713–722. doi: 10.1006/excr.1999.4716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Vacik J, Dean BS, Zimmer WE, Dean DA. Cell-specific nuclear import of plasmid DNA. Gene Therapy. 1999;6:1006–1014. doi: 10.1038/sj.gt.3300924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Dean BS, Byrd JN, Jr, Dean DA. Nuclear targeting of plasmid DNA in human corneal cells. Curr Eye Res. 1999;19:66–75. doi: 10.1076/ceyr.19.1.66.5344. [DOI] [PubMed] [Google Scholar]
  • 12.Dean DA. Import of plasmid DNA into the nucleus is sequence specific. Exp Cell Res. 1997;230:293–302. doi: 10.1006/excr.1996.3427. [DOI] [PubMed] [Google Scholar]
  • 13.Young JL, Byrd JN, Wyatt CR, Dean DA. Endothelial cell-specific plasmid nuclear import. Mol Biol Cell. 1999;10S:443a. [Google Scholar]
  • 14.Dowty ME, Williams P, Zhang G, Hagstrom JE, Wolff JA. Plasmid DNA entry into postmitotic nuclei of primary rat myotubes. Proc Natl Acad Sci USA. 1995;92:4572–4576. doi: 10.1073/pnas.92.10.4572. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Wilson GL, Dean BS, Wang G, Dean DA. Nuclear import of plasmid DNA in digitonin-permeabilized cells requires both cytoplasmic factors and specific DNA sequences. J Biol Chem. 1999;274:22025–22032. doi: 10.1074/jbc.274.31.22025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Mesika A, Grigoreva I, Zohar M, Reich Z. A regulated, NFkappaB-assisted import of plasmid DNA into mammalian cell nuclei. Mol Ther. 2001;3:653–657. doi: 10.1006/mthe.2001.0312. [DOI] [PubMed] [Google Scholar]
  • 17.Vaysse L, Harbottle R, Bigger B, Bergau A, Tolmachov O, Coutelle C. Development of a self-assembling nuclear targeting vector system based on the tetracycline repressor protein. J Biol Chem. 2004;279:5555–5564. doi: 10.1074/jbc.M311894200. [DOI] [PubMed] [Google Scholar]
  • 18.Dean DA, Strong DD, Zimmer WE. Nuclear entry of nonviral vectors. Gene Therapy. 2005;12:881–890. doi: 10.1038/sj.gt.3302534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Gorlich D. Nuclear protein import. Curr Opin Cell Biol. 1997;9:412–419. doi: 10.1016/s0955-0674(97)80015-4. [DOI] [PubMed] [Google Scholar]
  • 20.Moroianu J. Distinct nuclear import and export pathways mediated by members of the karyopherin beta family. J Cell Biochem. 1998;70:231–239. [PubMed] [Google Scholar]
  • 21.Stoffler D, Fahrenkrog B, Aebi U. The nuclear pore complex: from molecular architecture to functional dynamics. Curr Opin Cell Biol. 1999;11:391–401. doi: 10.1016/S0955-0674(99)80055-6. [DOI] [PubMed] [Google Scholar]
  • 22.Carson JA, Fillmore RA, Schwartz RJ, Zimmer WE. The smooth muscle gamma-actin gene promoter is a molecular target for the mouse bagpipe homologue, mNkx3-1, and serum response factor. J Biol Chem. 2000;275:39061–39072. doi: 10.1074/jbc.M006532200. [DOI] [PubMed] [Google Scholar]
  • 23.Kim D-W, Kempf H, Chen R, Lassar A. Characterization of Nkx3.2 DNA binding specificity and its requirement for somitic chondrogenesis. J Biol Chem. 2003;278:27532–27539. doi: 10.1074/jbc.M301461200. [DOI] [PubMed] [Google Scholar]
  • 24.Tanaka M, Kasahara H, Bartunkova S, Schinke M, Komuro I, Inagaki H, et al. Vertebrate homologs of tinman and bagpipe: roles of the homeobox genes in cardiovascular development. Dev Genet. 1998;22:239–249. doi: 10.1002/(SICI)1520-6408(1998)22:3<239::AID-DVG6>3.0.CO;2-7. [DOI] [PubMed] [Google Scholar]
  • 25.Tanaka M, Lyons GE, Izumo S. Expression of the Nkx3.1 homobox gene during pre and postnatal development. Mech Dev. 1999;85:179–182. doi: 10.1016/s0925-4773(99)00084-2. [DOI] [PubMed] [Google Scholar]
  • 26.Nishida W, Nakamura M, Mori S, Takahashi M, Ohkawa Y, Tadokoro S, et al. A triad of serum response factor and the GATA and NK families governs the transcription of smooth and cardiac muscle genes. J Biol Chem. 2002;277:7308–7317. doi: 10.1074/jbc.M111824200. [DOI] [PubMed] [Google Scholar]
  • 27.Dean DA, Vacik J, Dean BS, Zimmer WE. Cell-specific nuclear import of plasmids containing the smooth muscle gamma-actin promoter. Mol Biol Cell. 1998;9S:188a. [Google Scholar]
  • 28.Colin M, Moritz S, Fontanges P, Kornprobst M, Delouis C, Keller M, et al. The nuclear pore complex is involved in nuclear transfer of plasmid DNA condensed with an oligolysine-RGD peptide containing nuclear localisation properties. Gene Therapy. 2001;8:1643–1653. doi: 10.1038/sj.gt.3301572. [DOI] [PubMed] [Google Scholar]
  • 29.Bukrinsky MI, Sharova N, Dempsey MP, Stanwick TL, Bukrinsky AG, Haggerty S, et al. Active nuclear import of human immunodeficiency virus type 1 preintegration complexes. Proc Natl Acad Sci USA. 1992;89:6580–6584. doi: 10.1073/pnas.89.14.6580. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Greber UF, Kasamatsu H. Nuclear targeting of SV40 and adenovirus. Trends Cell Biol. 1996;6:189–195. doi: 10.1016/0962-8924(96)10016-7. [DOI] [PubMed] [Google Scholar]
  • 31.Fritz JD, Herweijer H, Zhang G, Wolff JA. Gene transfer into mammalian cells using histone-condensed plasmid DNA. Hum Gene Ther. 1996;7:1395–1404. doi: 10.1089/hum.1996.7.12-1395. [DOI] [PubMed] [Google Scholar]
  • 32.Mesika A, Kiss V, Brumfeld V, Ghosh G, Reich Z. Enhanced intracellular mobility and nuclear accumulation of DNA plasmids associated with a karyophilic protein. Hum Gene Ther. 2005;16:200–208. doi: 10.1089/hum.2005.16.200. [DOI] [PubMed] [Google Scholar]
  • 33.Branden LJ, Mohamed AJ, Smith CI. A peptide nucleic acid-nuclear localization signal fusion that mediates nuclear transport of DNA. Nat Biotechnol. 1999;17:784–787. doi: 10.1038/11726. [DOI] [PubMed] [Google Scholar]
  • 34.Cartier R, Reszka R. Utilization of synthetic peptides containing nuclear localization signals for nonviral gene transfer systems. Gene Therapy. 2002;9:157–167. doi: 10.1038/sj.gt.3301635. [DOI] [PubMed] [Google Scholar]
  • 35.Rech J, Barlat I, Veyrune J, Vie A, Blanchard J. Nuclear import of serum response factor (SRF) requires a short amino-terminal nuclear localization sequence and is independent of the casein kinase II phosphorylation site. J Cell Sci. 1994;107:3029–3036. doi: 10.1242/jcs.107.11.3029. [DOI] [PubMed] [Google Scholar]
  • 36.Gauthier-Rouvière C, Vandromme M, Lautredou N, Cai QQ, Girard F, Fernandez A, et al. The serum response factor nuclear localization signal: general implications for cyclic AMP-dependent protein kinase activity in control of nuclear localization. Mol Cell Biol. 1995;15:433–444. doi: 10.1128/mcb.15.1.433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Hessabi B, Ziegler P, Schmidt I, Hessabi C, Walther R. The nuclear localization signal (NLS) of PDX-1 is part of the homeodomain and represents a novel type of NLS. Eur J Biochem. 1999;263:170–177. doi: 10.1046/j.1432-1327.1999.00481.x. [DOI] [PubMed] [Google Scholar]
  • 38.Young JL, Benoit JN, Dean DA. Effect of a DNA nuclear targeting sequence on gene transfer and expression of plasmids in the intact vasculature. Gene Therapy. 2003;10:1465–1470. doi: 10.1038/sj.gt.3302021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Gasiorowski JZ, Dean DA. Postmitotic nuclear retention of episomal plasmids is altered by DNA labeling and detection methods. Mol Ther. 2005;12:460–467. doi: 10.1016/j.ymthe.2005.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Camoretti-Mercado B, Fernandes D, Dewundara S, Churchill J, Ma L, Kogut P, et al. Inhibition of TGF beta-enhanced SRF-dependent transcription by SMAD7. J Biol Chem. 2006;281:20383–20392. doi: 10.1074/jbc.M602748200. [DOI] [PubMed] [Google Scholar]

RESOURCES