The influence of mediators of intracellular trafficking on transgene expression efficacy of polymereplasmid DNA complexes

A B S T R A C T

Polymer-mediated gene delivery is an attractive alternative to viral vectors, but is limited by low effi- cacies of transgene expression. We report that polymers possess differential efficacies for transfecting closely related human prostate cancer cells, which correlates with dramatically different intracellular fate of nanoscale cargo in these cells. Sequestration of nanoscale cargo (27 nm quantum dots and 150e250 nm polyplexes) at a single location near the microtubule organizing compartment (MTOC) in PC3ePSMA human prostate cancer cells correlated with lower polymer-mediated transgene expression compared to PC3 cells, which showed distributed localization throughout the cytoplasm. We show, for the first time, that treatment with the histone deacetylase 6 (HDAC6) inhibitor tubacin, which acetylates tubulin of microtubules in the cytoplasm, abolished quantum dot and polyplex sequestration at the perinuclear recycling compartment/microtubule organizing center (PNRC/MTOC) and increased polymer-mediated transgene expression by up to forty-fold compared to cells not treated with the HDAC6 inhibitor drug. Treatment with the class I and II HDAC inhibitor trichostatin A (TSA) demonstrated similar levels of transgene expression enhancement. These results indicate that mediators of intracellular trafficking can be employed to modulate nanoparticle fate and enhance the efficacy of nanoscale therapeutics in cells. Simultaneous use of high-efficacy polymers along with mediators of intracellular trafficking is an attractive synergistic strategy for enhancing polymer-mediated transgene expression.

1. Introduction

Nucleic acid based treatments, including gene and siRNA therapy, are attractive therapeutic options since they have the potential to overcome consequences of genetic mutations that are characteristic of many diseases, including cancer [1]. Retrovirus, adenovirus, and adeno-associated viral vectors have been widely investigated as vehicles for transgene delivery [2e4]. Viral vectors, however, suffer from low vector titer, high risk of insertional mutagenesis, high immunogenicities, susceptibility to lysis by human serum, restricted tropism for certain cell types, degradation, low titers and high production costs, and therefore, motivate the identification of non-viral methods for gene delivery [5e7].
Cationic polymers and lipids have been extensively employed for delivering nucleic acids to mammalian cells [8e14]. The cyto- toxicity of polymers initially investigated for gene delivery (e.g. polyethyleneimine) led to recent focus on biocompatible cationic polymers including those based on carbohydrates [15e17], genet- ically engineered polypeptides [18], naturally occurring polymers [19], and biodegradable polymers [20e23]. However, non-viral gene delivery systems typically do not possess the transduction efficacies observed with viral vectors [5], and have therefore met with limited success. Strategies that enhance transgene expression in mammalian cells can help realize the potential of polymer- mediated gene delivery.
We recently employed a parallel synthesis and screening approach in order to rapidly identify polymers that demonstrate high transgene expression and low cytotoxicity towards mamma- lian cells following delivery of plasmid DNA (pDNA) [24]. Here, the role of intracellular trafficking on differential efficacies of polymer- mediated transgene expression in closely related prostate cancer cell lines was investigated using a high-efficacy polymer from this library in addition to polyethyleneimine, a current standard for polymer-mediated transgene delivery. The influence of chemo- therapeutic mediators of intracellular trafficking was investigated as a strategy to enhance the efficacy of polymer-mediated transgene expression. In particular, tubacin, which inhibits histone deacetylase 6 (HDAC6) in the cytoplasm and therefore modulates intracellular transport of cargo, and trichostatin A, an HDAC inhibitor that acts in the cytoplasm as well as the nucleus, were evaluated.

2. Materials and methods

2.1. Cell culture

PC3 human prostate cancer cells were obtained from the American Type Culture Collection (ATCC, VA). The PC3ePSMA cell line [25], derived by transducing PC3 cells for stable expression of the Prostate Specific Membrane Antigen (PSMA) receptor, was a generous gift from Dr. Michel Sadelain (Memorial Sloan-Kettering Cancer Center, New York, NY). Cells were cultured in RPMI-1640 medium (HyClone®, UT) containing 10% heat-inactivated fetal bovine serum (FBS; HyClone®, UT) and 1% antibiotics (100 units/mL penicillin and 100 mg/mL streptomycin; HyClone®, UT).

2.2. Transgene expression in PC3 and PC3ePSMA cells

Transfection of PC3 and PC3ePSMA cells using polyplexes of pGL3 control vector (Promega Corp., Madison, WI) and 1,4C-1,4Bis or pEI-25 (25 kDa polyethyleneimine) polymers, were carried out as described previously [24]. The 1,4C-1,4Bis polymer was synthesized from the addition polymerization of 1,4-cyclohexanedimethanol diglycidyl ether and 1,4bis(3-aminopropyl) piperazine) monomers [24]. Cells were seeded at a density of 50,000 cells/well in 500 ml growth medium (RPMI-1640 medium with 10% FBS) in a 24-well plate and allowed to attach overnight. Polyplexes were prepared by incubating different amounts of polymers and 200 ng pGL3 plasmid DNA for 20 min at room temperature (approximately 22e25 ◦C) resulting in polymer:pDNA polyplex weight ratios of 1:1, 5:1, 10:1, 20:1, and 25:1. Cells were treated with polyplexes in antibiotics-containing RPMI-1640 medium for 6 h, following which, the medium was replaced with fresh serum-containing medium. Luciferase protein expression (relative luminescence units or RLU) was determined using a plate reader (Bio-Tek Synergy 2) using the luciferase activity assay (Promega) 48 h after transfection. The protein content in each well was determined using the Pierce® BCA Protein Assay Kit (Pierce Biotechnology, Rockford, IL). Transgene expression in PC3 and PC3ePSMA cells was calculated as RLU per milligram (mg) protein (RLU/mg). Cytotoxicity of the polyplexes was determined by measuring cell viability using MTT assay (ATCC, Manassas, VA) at a wavelength of 570 nm using the Bio-Tek plate reader [24].

2.3. Intracellular trafficking of nanoscale cargo (polyplexes and quantum dots) in PC3 and PC3ePSMA cells

Fluorescein labeled LabelIT® plasmid delivery control DNA (Mirus Bio Corpora- tion, Madison, WI) was employed in order to visualize the intracellular localization of polyplexes. Two thousand nanograms of LabelIT® pDNA were complexed with an equivalent amount of the 1,4C-1,4Bis polymer and incubated for 20 min at room temperature in order to form polyplexes; higher amounts of the fluorescently labeled plasmid DNA were required in order to visualize localization inside cells. Red-fluorescent quantum dots (655 nm QDs, 0.2 nM) were added to cells along with polyplexes in order to facilitate co-localization studies as described previously [26]. Following co-incubation with polyplexes and QDs for 6 h, serum-free media was replaced by serum-containing media for 48 h, in order to mimic transfection experiments. Cells were then washed with 1× PBS, mounted in fluoro-gel media
(Electron Microscopy Sciences, PA), and analyzed using a laser scanning Nikon C1 confocal microscope (Nikon Instruments Inc., Melville, NY). The LabelIT® pDNA was excited with a 488 nm argon laser for fluorescein emission at 515 nm, while QDs were excited at 445 nm and emission was recorded at 655 nm. Images were acquired using EZ-C1 FreeViewer analysis software (Gold Version 3.20 build 615, Nikon Corporation) at 100× objective with a z-step of 0.4 mm/slice and with dual-channel scanning at 512 × 512 pixels using PMTs. Images were then stacked in RGB color using Image Processing and Analysis in Java (ImageJ) 1.38X software; the average fluorescence intensity was used in image analyses for visualizing the intracellular co-localization of QDs and fluorescently labeled DNAepolymer polyplexes. Reported images are representative of at least three independent experiments.

2.4. Microtubule involvement in intracellular polyplex transport in PC3ePSMA cells

PC3ePSMA cells were incubated with or without the microtubule disruption agent nocodazole (40 mM; SigmaeAldrich) for 1 h followed by 6 h incubation with the polyplexes of LabelIT pDNA in serum-free media. Live cells were post-stained with 10 mg/ml of the nuclear dye 40 ,6-diamidino-2-phenylindole (DAPI; Invitrogen) for 1 h and washed with 1× PBS. Cells were then fixed in 4% paraformaldehyde for 15 min at room temperature, washed again with 1× PBS, and mounted in fluoro-gel media. Intracellular localization of the polyplexes was examined as described above.

2.5. Role of histone deacetylase inhibitors (HDACi) on intracellular trafficking of nanoscale cargo

Tubacin [27,28], a recently discovered inhibitor of HDAC6, a cytoplasmic histone deacetylase, and its inactive analog, niltubacin were gifts from Professor Stuart Schreiber at the Broad Institute in Boston, MA. PC3 and PC3ePSMA cells (50,000/ well) were co-incubated with QD655 (Invitrogen) and varying concentrations of tubacin (0e6 mM) for 48 h at 37 ◦C in order to investigate the effects of tubacin on trafficking of nanoscale cargo. PC3ePSMA cells were also incubated with QDs in presence of 4 mM niltubacin which was used as a negative control. QDs were used in the trafficking experiments due to the ease of imaging with these nanoparticles.

Following incubation, cell nuclei were stained with DAPI for 1 h and cells were washed with 1× PBS. Intracellular localization and redistribution of QDs in the perinuclear region were visualized at 48 h using Zeiss AxioObserver D1 inverted microscope (Carl Zeiss MicroImaging Inc., Germany). Phase contrast images of the cells and the corresponding fluorescence images were captured with an Axio- CamMR3 camera (Zeiss Inc.) connected to the microscope with a LD Plan-Neofluor 40×/0.6 (N.A.) objective and exported in TIFF format using AxioVision software (Zeiss Inc.).

PC3ePSMA cells (50,000/well) were incubated with 2000 ng of LabelIT® pDNA polyplexes using 1,4C-1,4Bis polymer:pDNA for 48 h in presence or absence of 4 mM tubacin or niltubacin. Cell nuclei were stained with 10 mg/ml DAPI for an hour. The cells were then washed, mounted and imaged using confocal microscopy as described previously. Differential interference contrast (DIC) images were captured concurrently using a transmitted light detector.

Quantitative analysis of confocal microscopy images was carried out in order to calculate the percentage of total cell area occupied by QDs and polyplexes in pres- ence or absence of tubacin and niltubacin using the ImageJ software. The tracing tool in ImageJ was used to map a) cellular boundary from phase contrast and DIC images and b) the endocytosed QDs and polyplexes from the fluorescence images. The pixel areas and intensities of the mapped regions were determined, and percentage of total cell area occupied by the QDs and polyplexes was calculated.

2.6. Enhancement of polymer-mediated transgene expression using histone deacetylase inhibitors (HDACi)

PC3 and PC3ePSMA cells were co-incubated with 10:1 and 25:1 weight ratios of 1,4C-1,4Bis polymer and pGL3 pDNA polyplexes and different doses of the HDAC6 inhibitor tubacin (0e6 mM) or niltubacin as described in the previous section. Luciferase protein expression in presence of tubacin and niltubacin was determined (RLU/mg) and compared with protein expression in case of cells treated with pGL3 plasmid alone, polymers alone and vehicle (DMSO alone). Transgene expression was reported as fold increase in RLU/mg relative to cells treated with the polyplex alone but not with tubacin or niltubacin.
PC3 and PC3ePSMA cells were treated with polyplexes in the presence and absence of different concentrations (0e1 mM) of trichostatin A (TSA; Sigma) a class I and class II HDAC inhibitor. Following incubation with 10:1 and 25:1 polyplex ratios for 48 h in presence of TSA, cells were trypsinized, resuspended in medium, and luciferase expression was determined as described previously. Transgene expression was reported as fold increase in RLU/mg relative to cells treated with polyplexes in the absence of TSA.

2.7. Statistical analyses

Values are expressed as the mean one standard deviation (S.D.). All experi- ments were carried out at least in triplicate unless mentioned otherwise. The significance of the difference between the control and each experimental test condition was analyzed by two-tailed, paired Student’s t-test.

3. Results and discussion

3.1. Differential intracellular trafficking of polyplexes and transgene expression in closely related prostate cancer cells

Previous studies on the parallel synthesis and screening of a library of cationic polymers resulted in the identification of a polymer, 1,4C-1,4Bis, which demonstrated significantly higher transgene expression compared to pEI-25 [24]. Fig. 1 compares luciferase expression in the two humanprostate cancer cell lines, PC3 and PC3ePSMA, using polymer:pGL3 control vector (pDNA) weight ratios ranging from 1:1 to 25:1. Luciferase expression (RLU/mg) was consistently higher in PC3 cells compared to PC3ePSMA cells at equivalent polymer:pDNA weight ratios for transfections using the 1,4C-1,4Bis polymer. Polyplexes based onpEI-25 did not demonstrate significant differences in luciferase expression between the two cell lines; only up to two-fold higher transfection of PC3 cells over PC3ePSMA cells was observed for most polymer:pDNA weight ratios investigated. Highest transgene expression levels were obtained for a polymer:pDNA weight ratio of 10:1 in both cell lines (Supporting Information Table S1). Transgene efficacy correlates with size of the polyplexes; polyplexes formed at the 10:1 ratio (150e200 nm) are well within the endocytic limit. However, use of 25:1 ratio resulted in larger polyplexes ranging from 240 to 260 nm which possibly decreased uptake efficiency and subsequent transgene expression at these ratios. The 1,4C-1,4Bis polymer was chosen for all subsequent analyses due to a combination of higher transgene expression and lower cytotoxicity compared to pEI-25 (Supporting Information Table S1 and Figure S1).

We have recently demonstrated that cellular phenotypic differ- ences had a dramatic consequence on the intracellular transport and localization of nanoparticles (quantum dots or QDs) in PC3 and PC3ePSMA cells [26]. While QDs localized as punctate structures in vesicles throughout the cytoplasm of PC3 cells, they were trafficked along microtubules primarily to a single perinuclear location, the perinuclear recycling compartment (PNRC), which is close to the microtubule organizing center (MTOC) in PC3ePSMA cells. Further characterization indicated no presence of aggresome-specific proteins, ubiquitin, HSP-70 or vimentin [29e31] at the PNRC/MTOC, indicating that this perinuclear compartment was most likely not an aggresome (Supporting Information Figure S2). An investigation into the intracellular fate of polyplexes indicated high co-localiza- tion of QDs (red-fluorescent) and polyplexes (green fluorescent) at the perinuclear location in PC3ePSMA cells as seen from the single yellow-colored spot in Fig. 2 top row, indicating that the delivered plasmid was largely sequestered as aggregates in the PNRC/MTOC. Diverse cargo including anti-PSMA antibodies [26], transferrin [26], 27 nm diameter quantum dots [26] and 150e250 nm polyplexes, meet a similar intracellular fate in PC3ePSMA cells, despite the differences in size (polyplex and QD size was determined using DLS; Supporting Information Table S2), indicating that PC3ePSMA cells employed this as a default trafficking pathway. Polyplexes, but not QDs, were found around the cell periphery in both cell types, presumably due to cationic polymerecell membrane binding.

Intracellular transport of diverse cargo, including plasmid DNA and their complexes, is mediated by molecular motors that travel on microtubules in the cytosol [32e38]. We verified that localization of polyplexes at the PNRC/MTOC was dependent on microtubule- based transport; disruption of microtubules using nocodazole led to an arrest of polyplexes close to cell periphery (Fig. 3), which was consistent with our previous observations with QD trafficking [26] and other observations in the literature [34,35].

Correlation of intracellular trafficking profiles with transgene expression indicated that sequestration of polyplexes as aggregates in the PNRC/MTOC may be partly responsible for the lower lucif- erase expression in PC3ePSMA cells compared to PC3 cells, particularly in case of the 1,4C-1,4Bis polymer. In contrast, poly- plexes, like QDs, were seen localized in vesicles throughout the cytoplasm of PC3 cells (Fig. 2 bottom row). It is possible that distribution of polyplexes in smaller vesicles throughout the cyto- plasm, as opposed to aggregation at a single large perinuclear location, can allow for greater endosomal escape, which in turn can enhance transgene expression.

3.2. HDAC inhibitors alter intracellular trafficking of nanoscale cargo

In light of the hypothesis that polyplex localization in the PNRC near the MTOC correlated with lower transgene expression in PC3ePSMA cells, we asked if altering the intracellular localization of nanoscale cargo, away from the MTOC in these cells can increase transgene expression levels. Histone deacetlylases (HDACs) are a family of enzymes that are mostly nuclear in their location, cleave acetyl groups from e-N-acetyl-lysines in target proteins, and regulate transcription, angiogenesis, cell motility, and a variety of other functions in cancer cells [39e41]. Cytoplasmic HDACs, in particular HDAC6, mediate intracellular trafficking of cargo to the MTOC by deacetylating a-tubulin and regulating dynein motor transport on microtubules (Fig. 4) [30,42]. Tubacin is a recently discovered small-molecule HDAC6 inhibitor (HDACi) which acts as an inhibitor of a-tubulin deacetylation [27]. Increased tubulin acetylation (or reduced deacetylation) results in microtubule stabilization and enhances intracellular transport due to greater recruitment of dynein and kinesin motors to microtubules [42]. PC3ePSMA cells treated with tubacin concentrations of 0e4 mM did not demonstrate any obvious changes in trafficking or localization of internalized QDs (Supporting Figure S3). However, at tubacin concentrations ≥ 4 mM, internalized QDs showed a diffused profile around the PNRC, indicating that the nanoscale cargo was delivered over a larger area in the cytoplasm with the possibility that some cargo was delivered outside of the PNRC (Fig. 5a). This concentra- tion range is consistent with previous reports which indicate that the half-maximum effective concentration (EC50) of tubacin- inhibited tubulin deacetylation was 2.5 mM [27]. Niltubacin, which lacks HDAC6i activity [27], did not alter QD trafficking behavior under similar conditions, indicating that the altered trafficking depended on HDAC6 inhibition activity (Fig. 5a). The effect of tubacin on QD trafficking in PC3 cells was not obviously different from untreated cells under the microscopy resolution employed (Supporting Information Figure S3).

Remarkable differences were also seen in polyplex trafficking following tubacin treatment in PC3ePSMA cells. Polyplexes by themselves were largely found localized at the PNRC/MTOC or at the cell periphery (Fig. 5b). Tubacin treatment (4 mM), however, completely abolished the sequestration of polyplexes in the PNRC/ MTOC; polyplexes were found distributed in vesicles throughout the cytoplasm in these cells, much similar to the intracellular localization profile observed in PC3 cells. On the other hand, poly- plexes were largely seen at the cell periphery and at the PNRC/ MTOC in cells treated with 4 mM niltubacin, indicating that the HDAC6 inhibition activity was necessary for abolishing polyplex transport to the PNRC/MTOC. It is not clear why niltubacin promotes arrest of polyplexes at the cell periphery; it is possible that the molecule possesses hitherto unexplored activity that may be responsible for this behavior. Tubacin concentrations above 6 mM were toxic to cells (not shown) and were therefore not employed. Quantitative image analyses indicated that the cell surface area covered by QDs was 2e3 fold higher in case of tubacin treated cells compared to untreated cells and niltubacin treated cells (Fig. 5c), which indicated the distribution of the nanoparticles over a larger cytoplasmic area. The average fluorescence intensity remained invariant in all three cases (Fig. 5c). Interestingly, both tubacin and niltubacin treatment enhanced the cytoplasmic coverage of polyplexes in PC3ePSMA cells compared to untreated cells; however, tubacin alone was able to abolish polyplex locali- zation at the PNRC/MTOC. Our current analysis cannot rule out the intriguing possibility that more nanoscale cargo (QDs or poly- plexes) is delivered to the perinuclear region following increased molecular motor-mediated transport as a consequence of tubulin acetylation with tubacin [42,43].

3.3. HDAC inhibitors enhance polymer-mediated transgene expression in cells

PC3ePSMA cells were treated with different concentrations of tubacin (0e6 mM) along with 10:1 and 25:1 weight ratios of 1,4C- 1,4Bis polymer and pGL3 plasmid-based polyplexes in order to investigate if tubacin-mediated differences in intracellular traf- ficking also influenced transgene expression. A forty-fold increase in transfection efficacy in PC3ePSMA cells was observed when poly- plexes were employed at 25:1 polymer:pDNA weight ratio (polyplex ratio) along with 4 mM tubacin compared to cells not treated with tubacin (Fig. 6a). Other tubacin concentrations showed up to twenty- fold (p < 0.005) higher transgene expression in PC3ePSMA cells compared to cells not treated with the HDAC6 inhibitor. Treatment with similar concentrations of niltubacin resulted in a statistically insignificant 1.5-fold enhancement of transgene expression (Supporting Information Figure S4) compared to untreated cells. This increase was negligible compared to that observed with tubacin, and therefore, implicated HDAC6 inhibition and tubulin deacetylation in enhancing transgene expression. Interestingly, only a two-fold enhancement in transgene expression was observed with tubacin when 10:1 polyplex ratio was used, indicating that the tubacin activity was dependent on basal levels of transgene expression, polyplex dose and composition, and possibly, size. Taken together, these results indicate that tubacin concentrations which result in inhibition of tubulin deacetylation and alter nanoparticle trafficking in PC3ePSMA cells, also resulted in high (40e80 fold) increase in transfection efficacies. It is important to point out that 1,4C-1,4Bis polymer demonstrated forty to eighty-fold higher transgene expression than pEI-25 in PC3ePSMA cells at a polyplex ratio of 25:1 [24]. This was further enhanced by up to forty-fold by using tubacin, indicating that transgene expression could be enhanced by 1600- fold in PC3ePSMA cells by using a synergistic combination of more effective polymeric transfection agent (1,4C-1,4Bis) and the HDAC6 inhibitor, tubacin. The efficacy of this combination treatment is significantly higher than what has been traditionally possible with pEI-25, a standard for polymeric gene delivery. Tubacin treatment (5 mM) resulted in up to a thirty-fold increase in transfection of PC3 cells compared to cells not treated with the HDAC6 inhibitor for the 25:1 polyplex ratio. Significant levels of transfection enhancement, approximately 8-, 16- and 20-fold, were observed in case of 2, 4, and 6 mM tubacin, respectively (Fig. 6b).

Trichostatin A (TsA), is a class I and II HDAC inhibitor that has both nuclear and cytoplasmic activity, and was also employed in order to determine if the enhancement in transgene expression was specific to tubacin or HDAC6 inhibition alone. As shown in Fig. 7a, approximately 35-fold increase in luciferase expression was observed in PC3ePSMA cells treated with 250 nM TsA and 25:1 polyplex ratio of 1,4C-1,4Bis:pGL3 plasmid, compared to cells treated with in the absence of the HDAC inhibitor. Enhancement of transfection efficiency in PC3 cells was also dependent on TsA dose (Fig. 7b); maximum relative luciferase expression in PC3 cells was found with 100 nM TsA for 25:1 polyplex ratio. However, TsA treatment led to only up to ten-fold enhancement of transgene expression in PC3 cells. The reasons for the differential enhance- ment between PC3 and PC3ePSMA cells with TsA are unclear at this point. Co-treatment of TSA and polyplexes resulted in negligible loss of PC3ePSMA cell viability, while PC3 cell viability was found to be between 60 and 70%; similar results were seen in case of tubacin (Supporting Information, Figures S5 and S6).

Taken together, the above results indicate that both cytoplasmic and nuclear HDAC inhibitors, tubacin and TsA were able to enhance transgene expression in the two human prostate cancer cell lines employed in the current investigation. Importantly, not just relative levels but also absolute levels of transgene expression (RLU/mg) were highest with tubacin treatment in both cell types investigated (Table 1). The level of enhancement was dependent on the poly- mer:pDNA ratio, the HDAC inhibitor, and the cell line employed.

Following uptake, nanoscale cargo, including polyplexes, is trafficked along the default degradative lysozomal pathway in most cells. Following endosomal escape, polyplexes must traffic through the cytoplasm to the nucleus and at some point the plasmid must be accessible to the nuclear machinery for transcription of the delivered transgene. While passive diffusion of plasmid DNA is negligible through the cytoplasm, our current results and those of others demonstrate that cytoplasmic trafficking of polyplexes/ plasmid DNA is dependent on microtubules and is mediated by molecular motors [33,34,38]. Furthermore, only a small fraction of plasmid DNA in the cytoplasm is imported into the nucleus; cyto- plasmic proteins that bind to specific sequences on the DNA sequence mediate active nuclear import [44]. The distribution and localization of plasmid DNA inside the nucleus also greatly influ- ences transcription and transgene expression [45]. The efficacy of non-viral gene delivery can be greatly enhanced by strategies that enhance both cytoplasmic trafficking and nuclear import and transcription machineries.

Tubacin selectively inhibits HDAC6 and is therefore largely cytoplasmic in its action. Tubacin treatment resulted in significant changes in intracellular trafficking of nanoscale cargo in PC3ePSMA cells and also led to increase in transgene expression in these cells. In the absence of tubacin, transgene expression was higher using 10:1 polyplex ratio than 25:1 polyplex ratio in PC3ePSMA cells. Tubacin treatment, however, resulted in highest transgene expression levels in case of the 25:1 polyplex ratio, but did not lead to similar levels of enhancements in case of transfections with 10:1 polyplex ratio (Table 1). It is possible that a combined effect of higher polyplex size (approximately 250 nm for 25:1 polyplex ratio compared to 150 nm for 10:1 polyplex ratio) and the effect of tubacin on polyplex/pDNA trafficking resulted in high transgene expression in this case. Results with PC3 cells were different in that while tubacin enhanced transgene expression in all cases, obvious differences in pDNA localization were not apparent these cells, although differences at the single-molecule level could not be detected using our current methods. Transgene expression remained highest in the case of 10:1 polyplexes even after tubacin treatment, in contrast to PC3ePSMA cells. These results indicate that ‘uniform’ distribution of polyplexes throughout the cytoplasm may be required for high basal levels of polymer-mediated trans- gene expression as seen in PC3 cells. On the other hand, seques- tration of polyplexes in a single compartment compromises transfection efficacy as in the case of PC3ePSMA cells leading to lower basal levels of transgene expression. Although mediators of intracellular trafficking such as tubacin can enhance transgene expression in both cases, polyplex size, intracellular localization profiles, and corresponding basal levels of transgene expression affect the overall enhancement observed. Similar trends were observed with TSA (Table 1).

The HDAC6i activity of tubacin is responsible for enhanced acetylation/stabilization of microtubules in cells, which has been implicated in increasing transgene expression levels in cells [33,46,47]. Enhanced recruitment of dynein and kinesin motor proteins by acetylated microtubules [42,43] following tubacin treatment can increase the transport of pDNA towards the nucleus and therefore increase transgene expression. It is thought that increased stability of microtubules following acetylation is partly responsible for increased affinity for molecular motors, which increases transport and transgene expression [46]. Additionally, faster minus-end transport can also reduce degradation of pDNA in the cytoplasm, which can contribute to increased transgene expression. An intriguing but hitherto unexplored possibility is that tubacin modulates the ‘tug-of-war’ dynamics [48] between both plus-end (kinesin) and minus-end (dynein) on microtubules by activating both molecular motors. Tug-of-war between vesicles containing nanoscale cargo (QDs or polyplexes) due to opposing motor activity may provide enough time for escape of additional polyplexes or dissociated plasmid DNA into the cytoplasm from endosomal compartments, leading to higher transgene expression. It is possible that we were able to visualize the consequences of some of these processes in case of PC3ePSMA cells due to the sequestration of polyplexes in a relatively large, micron-sized jux- tanuclear location. However, higher-resolution single-molecule studies will be necessary to further elucidate the role of tubacin in intracellular plasmid DNA transport and nuclear uptake in cells.

Histone Deacetylase Inhibitors (HDACi) promote the relaxation of DNA wrapped around core histones following acetylation of histones and enhance transgene expression by allowing the tran- scription apparatus to access DNA promoter regions[41,49,50]. Trichostatin A (TsA) possesses both, class I and II HDAC inhibition activity, and therefore can act in the cytoplasm and the nucleus. In the cytoplasm, TsA can inhibit HDAC6 and therefore has similar activity to that described for tubacin. In the nucleus, TsA is known to enhance transcription and transgene expression by acetylating histones. Acetylation of core histones reduces interactions with DNA and is associated with increased transcriptional activation. Recent studies have shown that TsA plays a role in repositioning pDNA towards transcriptionally active sites (e.g. acetylated histones) in the nucleus and therefore enhances transgene
expression [45,51]. Finally, other mechanisms, including promoter activation [40], may also contribute towards the observed enhancement in transgene expression by HDAC inhibitors.

4. Conclusions

In the present work, we investigated mediators of intracellular trafficking for enhancing polymer-mediated transgene expression. The histone deacetylase 6-inhibitor (HDAC6i) tubacin was able to enhance transgene expression in both PC3 and PC3ePSMA cells, indicating that microtubule stabilization and/or enhanced recruit- ment of molecular motors to microtubules can enhance the efficacy of polymer-mediated gene delivery. The use of tubacin, which is cytoplasmic in its action, implicates microtubule acetylation in enhancing polymer-mediated transgene expression. To our knowledge, this is the first report that demonstrates the use of a selective small-molecule inhibitor of cytoplasmic histone deace- tylase (HDAC6) for enhancing transgene expression in cells. The use of trichostatin A indicated that HDAC inhibitors with both nuclear and cytoplasmic activity could also be employed as enhancers of transgene expression. A synergistic approach that involves mate- rials chemistry (polymer design) along with mediators of intra- cellular trafficking can therefore result in high transgene expressions using polymers. These approaches can have significant implications for polymer gene delivery in clinical applications, which has traditionally lagged behind viral vectors and other non- viral methods, including lipofection and hydrodynamic injection, due to poorer transgene expression levels and toxicity concerns. Finally, our results with nanoparticles, both quantum dots and polyplexes, indicate that tubacin might be a promising molecule for enhancing and/or modulating the delivery of nanoscale therapeu- tics and imaging agents to a variety of cells.