Comparison of transfection chemical methods in PC12 cells
We compared transfection efficiencies obtained in PC12 cells with the lipopolyplex transfection reagent TransIT-LT1 (Mirus) and the cationic lipids Lipofectamine 2000 and LTX (Invitrogen). Cells were grown in conditions promoting proliferation including use of an enriched medium, but the transfections were carried out in a serum- and antibiotic- free environment. We performed these experiments with different amounts of pEGFP-C1 DNA (Clontech), a plasmid driving the expression of an enhanced green fluorescent protein (EGFP) under the control of the CMV promoter. The ratio DNA (μg): transfection reagent (μl) was 1:3 and we transfected increasing amounts of plasmid (0.25, 0.5, 0.75 and 1 μg).
Forty-eight hours after transfection, nuclei were stained with the viable Hoechst 33342 fluorescent dye and images were acquired on an Operetta System, which combines fluorescence microscopy in a multi-well format with automated image acquisition and quantitative analysis (Fig. 1). Data analysis for transfection efficiency was performed by using the Harmony® High-Content Imaging Software (Perkin Elmer), comparing the number of EGFP-positive cells (detected with an Alexa Fluor 488 filter) to the total number of cells (detected with an Hoechst 33342 filter). Mock transfection controls using the transfection reagent without DNA had an auto-fluorescence background in both the Hoechst 33342 or Alexa Fluor 488 channels comparable to un-transfected PC12 cells (data not shown).
As shown in Fig. 2a, the TransIT-LT1 Transfection Reagent (Mirus) was not effective in facilitating transfection of DNA in PC12 cells, even when a higher concentration of DNA was used. However, in parallel experiments performed on the HEY4 ovarian cancer cell line, TransIT-LT1 had a transfection efficiency of approximately 35 %, thereby indicating that efficiency depends on the cell type (data not shown).
On the contrary, when we performed the transfection experiments by using Lipofectamine LTX and Lipofectamine 2000 we managed to transfect DNA into PC12 cells, with efficiencies that ranged from 7 and 15 % respectively, with 0.25 μg DNA, to 35 and 46 % respectively, when 1 μg of DNA was used.
As expected, transfection efficiency for both reagents correlates with the amounts of DNA used. However, in comparing the transfection efficiency of the two cationic lipids, Lipofectamine LTX seems to perform better than Lipofectamine 2000 at low DNA amounts, while Lipofectamine 2000 outperforms Lipofectamine LTX at higher DNA amounts.
In particular, comparing our results with those by Lee and colleagues (2008), with Lipofectamine 2000 we reached 21 % transfection efficiency (with 0.5 μg of DNA in a 24-well plate) while Lee and collaborators reported 14 % efficiency in similar conditions (1 μg of DNA in a 12-well plate).
Cell viability was measured by Trypan Blue Staining 48 h after transfection (Fig. 2b). Mock-transfected cells had a viability comparable to the non-transfected cells with either of the three transfection reagents (approx. 95 %). For all methods, viability decreased as the DNA amounts increased. However, as clearly indicated in Fig. 2b, there is a variation of trend for different reagents. TransIT-LT1 had a milder impact on cell viability, reaching 92 % when 1 μg of DNA was used. On the contrary, Lipofectamine 2000 and Lipofectamine LTX reached 87–88 % viability in those conditions.
DNA electroporation in PC12 cells
As the transfection efficiency obtained with Lipofectamine 2000 was not sufficient for our purposes, we investigated if we could obtain a higher percentage of transfected PC12 cells with an electroporation method (Neon Transfection System, Invitrogen).
To optimize conditions, 0.5 μg of plasmid DNA and two different cell densities (6 × 104 or 1 × 105 cells/well) were used. Moreover, a range of voltage, pulse width and pulse number combinations were tested (Table 1).
Forty-eight hours after transfection, nuclei were stained with Hoechst 33342 and images were acquired on an Operetta System (Fig. 3). Data analysis of transfection efficiency was carried out as detailed above and revealed that the density of the cells in the suspension is one of the most important variables affecting transfection efficiency in our electroporation protocols (Fig. 4a). In fact, with higher cell density, any electroporation condition tested yields high transfection efficiency, ranging between 80 and 98 %. On the contrary, when a lower cell density was used, conditions can be separated into two classes, based on the effect they have on transfection efficiency: a high-efficiency class (84–91 %) and a low-efficiency class (20–66 %).
It has to be pointed out, however, that all conditions, except one (6 × 104 cells, 1,500 V, 20 ms, 1 pulse), yielded a transfection efficiency higher than that one obtained by the use of Lipofectamine 2000 (dashed line in Fig. 4a).
The method shows high reproducibility, as determined by comparing efficiencies from several replicate experiments.
Cell viability was measured by Trypan Blue Staining 48 h after transfection (Table 1). Mock-electroporated cells were manipulated as the electroporated cells but did not receive DNA nor an electrical pulse, and had a viability comparable to the non-electroporated cells (approx. 90 %). In the case of cell viability, there was no clear correlation with cell density, although in general cells at higher density performed better (Fig. 4b). Furthermore, different electroporation conditions worked better for the different cell densities, and a general trend could not be inferred from the graph in Fig. 4b. In general, the electroporation protocol appeared to be more pernicious to PC12 cells than the lipofection protocols, as only few conditions provided a cell viability comparable or higher than that one obtained by Lipofectamine LTX and 1 μg of DNA (dashed line in Fig. 4b).
In order to establish if electroporation in general is a better method than lipofection, or if the Neon System in particular has a high performance in transfecting PC12 cells, we performed electroporation with a different system, namely Gene Pulser Xcell (Bio-Rad). 8 μg of plasmid DNA and two different cell densities (6 × 105 or 1 × 106 cells/well) were used. The increase in DNA and cell amounts, compared to Neon System, is due to the fact that, while the Neon system is miniaturized to use electroporation tips and 10 μl transfection volumes, the Gene Pulser XCell requires specific 400 μl cuvettes. For each cell density we tested three different voltage and capacitance combinations (Table 2), in triplicate experiments. Namely, we tried 240 V/1,000 μF (Murphy et al. 2008), 250 V/960 μF (Yaron et al. 2001) and 300 V/500 μF (Lombardi et al. 2001). Analysis of the transfection efficiency and cell viability was carried out as described above. As visible in Fig. 4c and in Table 2, in our hands the Gene Pulser Xcell system outperformed lipofection methods, reaching efficiencies higher than the ones reported in the literature. However, when comparing Gene Pulser Xcell with Neon System, the latter remained the method of choice for the transfection of PC12 cells (Fig. 4a, c). As far as viability is concerned, though, Gene Pulser Xcell compared to lipofection (Fig. 4d) and was better than Neon System in the majority of conditions.
For our purposes, the best electroporation conditions were obtained with Neon transfection System, 1 × 105 cells/well and 3 pulses of 1,500 V and 10 ms each. These conditions, in fact, yielded 90 % transfection efficiency and 99 % viability. We have chosen these conditions for the subsequent experiments.
Cell differentiation and neurite analysis
With the aim of investigating whether the electroporated PC12 cells retain the ability to differentiate into neuron-like cells, we transfected the cells using the conditions reported above, and treated them with NGF 2.5S (75 ng/ml) starting 24 h after transfection (day 0) and for the subsequent 7 days (Fig. 5a).
Images were taken at 8 h and every 24 h after transfection, in bright field and with a green fluorescence filter (Fig. 5b). NGF-treated PC12 cells showed increase in neurite length and generation during time (Fig. 5b). The total number of cells and the number of cells presenting neurites was counted in the bright-field images. Differentiated PC12 cells progressively increased reaching 23 % of total cells 4 days after induction. At later times the percentage of differentiated cells remained stable (Fig. 5c).
EGFP expression was observed both in differentiated and non-differentiated cells. The intensity of EGFP signal increased over time. At the initial time point, the intracellular expression of EGFP in NGF-differentiated PC12 cells was located in the nucleus. At later time points, it was around cytoplasm and inside the nucleus. On the 7th day post-transfection, EGFP was seen throughout the entire cell, spreading out to the tips of the neurite extensions (not shown).
The number of transfected cells was counted in pictures taken with the GFP filter. Already 8 h after electroporation 20 % of cells showed EGFP signal. When NGF was added, 24 h after electroporation, 69 % of cells were EGFP-positive. The maximum transfection efficiency was reached 48 h after electroporation (86 %), and rapidly decreased to reach 33 % at days 6 and 7.
The discrepancy between the maximum transfection efficiency in these experiments and the results described in the previous paragraph (where 90 % of transfection efficiency was obtained) might reflect the change in serum percentages in the growth medium and the addition of NGF, 24 h after transfection.
Electroporation posed no impairment on differentiation ability of PC12 cells, since about half of the differentiated cells were green. On the contrary, our observations suggest that electroporation might even induce PC-12 cells to differentiate. It is remarkable in fact, that approximately 30 % of green cells were presenting neurite outgrowth, from day 4 onwards, while the percentage of differentiated cells in non-green PC-12 cells was about 20 %.
This stimulus to differentiate might explain why at days 0, 1 and 2, green differentiated cells were already present (6, 10 and 12 % of the total cells) while non-green differentiated cells reached comparable percentages (12 %) only at day 3.
Moreover, looking at the actual cell numbers, it appears that at day 2 non-green, non-differentiated cells started outnumbering green, non-differentiated cells, possibly indicating a higher proliferation rate of non-transfected cells, compared to non-transfected cells.