Close this search box.

Cellular transfection using rapid decrease in hydrostatic pressure – Scientific Reports

Certain cell types, particularly transformed lines, can be efficiently transfected using several physical methods including electroporation, lipofection, and calcium phosphate-mediated uptake. In contrast, many primary cell types including embryonic stem cells are more difficult to transfect. The use of dynamic hydrostatic pressure for transfection overcomes these difficulties. Briefly, in the method we describe here (Fig. 1A–C), borosilicate capillary segments were prepared, sterilized, and filled with single cell solutions in electroporation buffer whereupon they were sealed with sterilized petrolatum to allow rapid transmission of local hydrostatic pressure (see “Methods”). Capillaries were then placed into a sealed silicone oil-filled pressure vessel. Hydrostatic pressure was then increased to the desired static holding pressure at a rate of approximately 65 MPa/min and maintained for periods of 30–600 s, whereupon this pressure was slowly or abruptly returned to ambient pressure (Fig. 1D).

Figure 1
figure 1

Hydrostatic pressure application to mammalian cells. (A) High hydrostatic pressure generating device used for experimental pressures of 1–200 MPa. Pressure line with valve seat (black arrow) and pressure chamber (red arrow) are indicated. Total chamber capacity 10 mL. (B) Example of cell suspension isolated in borosilicate capillary. (C) Closeup of capillary petrolatum seal. Scale bars in figures (B) and (C) equal to 2 mm. (D) Schematic of the Pressure Jump-Poration. Phase 1: Sample is pressurized from ambient to the desired pressure at a rate of ~ 65 MPa/min; Phase 2: pressure is maintained at the desired static pressure for the indicated period (0.5–10 min); Phase 3: sample pressure is suddenly returned to ambient by opening valve. (E) Schematic of pressure-induced transfection of genetic material. (F) Detection of propidium iodine uptake in select cells at 60 MPa (borosilicate capillary) immediately following acute depressurization. Scale bar denotes 100 μm.

It was observed that sudden depressurization from a particular range of static holding pressures resulted in a small number of cells permeabilizing their outer cell membrane without overt cellular destruction (Fig. 1E). An example of this can be seen in Fig. 1F, in which the normally cell impermeant marker propidium iodide (MW 668.4) has been added to a concentration of 0.75 μm (0.5 μg/mL) to the resuspension solution. Immediately following depressurization (< 5 min) a minority of cells could be observed allowing uptake of propidium iodide suggesting such cells might also be capable of absorbing higher molecular weight entities such as plasmid DNA. As shown in Fig. 2A, following pressure-induced transfection in the presence of a ~ 9.2 kb DNA plasmid inducing puromycin resistance (Addgene # 62988), numbers of puromycin-resistant colonies were enhanced within a specific pressure range following puromycin selection an initial plating density of 1042 cells/cm2 (10,000 cells/well of standard 9.6 cm2 6-well TC plate). We observed that static pressures in the range of 60–80 MPa followed by sudden depressurization resulted in a marked increase in plasmid uptake. Duration at these static pressures (30, 60, 90, 120, 180, 300 or 600 s) did not appear to affect the ability of cells to take up plasmid DNA. While durations of static pressure hold outside of these durations (< 30 s or > 300 s) were not examined, they are likely to yield similar results given our previous observations. As such all further cell uptake experiments were held at a static pressure for 30 s followed by sudden depressurization. To differentiate true functional plasmid expression in viable cells from simple DNA uptake, such uptake was assessed via clonogenic assay of puromycin-resistant clones. This provides demonstration of faithful plasmid uptake together with continued cell survival, propagation and function following cell transfection. Importantly it also demonstrates that transfected plasmids were actively and accurately transcribed. Plasmid uptake was found to be significantly reduced if the static holding pressure was outside the range of 60–80 MPa (Fig. 2A).

Figure 2
figure 2

Efficacy of transfection of ES cells using pressure-jump-poration. (A) Transfection rate for the pressures indicated compared to standard electroporation conditions following transfection. NDN: (−) DNA control; DNE: (+) DNA, no electroporation; E.STD: (+) DNA, (+) electroporation; ND80: (−) DNA at P = 80 MPa; 0–100: (+) DNA, (+) pressure treatment; 70SL: (+) DNA, (+) pressure treatment, but with slow (3 min) pressure release. All other samples were held at the indicated pressure for 30 s followed by acute depressurization. For each pressure condition n = 9 independent experiments with three replicates within each experiment were performed, and n = 5 independent experiments with three replicates within each experiment for electroporation conditions. Results shown ± SD. (B) ES cells on gelatin at 24 h following pressure treatment at 60 MPa (no selection). Scale bar denotes 150 μm. (C) Clonal ES cell colony on fibroblast bed layer following pressure treatment at 60 MPa with subsequent puromycin selection for 5 days. All pressurization experiments were performed at a concentration of 4000 cells/μL with treated cells plated at 10,000 cells/well in a 6 well plate; electroporation standards were plated similarly. Scale bar denotes 200 μm. Results shown ± SD. *Denotes significant enhancement at p < 0.01 over electroporation. Transfection plasmid shown in (B) and (C) expresses dTomato/puromycin.

To determine the relative efficiency of PJP, results with ES cells were compared in identical solutions and DNA concentrations to an optimized electroporation standard (E STD: Biorad 240 V, 500 μF in 4 mm cuvette). As shown in Fig. 2 the values obtained compare favorably to the current standards. No transfection was observed in the absence of added puromycin-encoding DNA in either the absence (ND) or presence (ND80) of applied hydrostatic pressure; similar to that seen for electroporation (NDE). As shown in Fig. 2, even in the presence of added plasmid DNA (20 μg/mL), very low levels of transfection are observed at static pressures below 30 MPa. In addition, at optimal pressures for transfection in ES cells, low levels of transfection were observed under conditions where the hydrostatic pressure applied was reduced slowly (70SL). Though the data presented here are for ES cells resuspended in EmbryoMax electroporation buffer, results obtained using OptiMEM or serum-free basal media such as DMEM were not significantly different demonstrating similar transfection results (data not shown).

To assess post-treatment function, unselected ES cells were plated on gelatin and visualized 24 h following treatment with static pressures equal to 60 MPa (Fig. 2B). As indicated in the figure, dTomato transfected cells were capable of appropriately transcribing and translating transfected DNA markers during this period. Similarly pressure transfected ES cells plated onto DR4 multidrug resistant mitomycin-treated fibroblasts (Fig. 2C), demonstrated continued growth characteristics similar to those seen for other forms of gene transfection (electroporation, lipofection).

To further address the capabilities and characteristics of pressure transfection, experiments were conducted in a separate ES cell line expressing EYFP (citrine) from the ROSA26 locus33 in the cytoplasmic cellular compartment. An advantage of this line is that it allows real-time examination for diffusion of the fluorescent 27.1 kDa EYFP protein from the cytoplasmic compartment in the event of interruption of the plasma membrane, which will recover over the long term given proper resealing of the plasma membrane. As shown in Fig. 3A–C (red arrows), transfection of EYFP-expressing ES cells using PJP at different static pressures and plated onto supporting DR4 fibroblasts (no puro selection applied) demonstrated competent expression of dTomato from the 9.5 kb plasmid vector in a minority of cells within 24 h post-transfection.

Figure 3
figure 3

Expression of DNA constructs in mammalian cells. ES cells were transfected with plasmids containing several selection markers to assess expression competency. (AC) Fluorescent photomicrographs of pressure treated EYFP-expressing ES cells 24 h following treatment at the pressures indicated in the absence of selection. Pressurization experiments were performed at 4000 cells/μL with cells plated at 25,000 cells/well in a 24 well plate. Relative numbers of transfected, dTomato-expressing cells are indicated (red arrow). Transfected cells exhibit reduced levels of cytoplasmic citrine compared to non-transfected cells (green arrow). Scale bar denotes 100 μm. (DF) Beta-galactosidase expression in R1 transfected ES cells 48 h following pressure treatment. Pressurization experiments were performed at 4000 cells/μL with cells plated at 25,000 cells/well (D,E) or 10,000 cells/well (F) in a 24 well plate. Scale bar denotes 300 μm. (D) (+) DNA, (−) pressure treatment, (E) (−) DNA (+) 80 MPa pressure, (F) (+) DNA (+) pressure treatment. Beta-galactosidase positive cells are indicated (blue arrow), as are examples of dead cells following pressure treatment (box).

Consistent with these findings, the data in Fig. 2A show that the frequency of these events followed a general trend with respect to applied static hydrostatic pressure, with a reduction in the average intensity of EYFP expression within dTomato+ versus dTomato cell populations (Fig. 3B). As shown in Fig. 3D–F, similar experiments performed in ES R1 cells using a 12.3 kb Lac-Z expression plasmid demonstrated beta-galactosidase expression at 48 h following pressure treatment. Even plating at 2.5 fold higher concentrations (25,000 cells/well in a 1.9 cm2 24 well plate, Fig. 3D,E) failed to demonstrate any beta-galactosidase expressing cells in the absence of either pressure treatment + DNA (Fig. 3D), or pressure treatment minus added plasmid DNA (Fig. 3E). By contrast beta-galactosidase positive cells (blue arrows) could readily be observed even at 2.5 fold lower plating densities under appropriate conditions of pressure treatment + vector (Fig. 3F). As expected, pressure treatment did result in the induction of cell death within a subpopulation of cells (boxed). Thus, using two distinct embryonic stem cell lines and several different expression vectors we observed that upon sudden depressurization after maintaining appropriate levels of static pressure resulted in the uptake and expression of ectopic DNA in a subpopulation of surviving cells.

In order to better understand the features of this pressure responsivity in mammalian cells, parental and sublines of R1 or citrine-expressing ES cells were examined for their DNA transfection potential as a function of pressure as shown in Fig. 4A. While absolute differences in the pattern and level of DNA uptake were noted between lines, a similar pattern of responsivity across different primary cell lines was observed for static pressures between 60–80 MPa.

Figure 4
figure 4

Characteristics of pressure-induced DNA transfection. All pressurization experiments were performed at a concentration of 4000 cells/μL. (A) Transfection response profiles of six ES cell lines to pressure treatment with treated cells plated in a 6 well plate at a density of 10,000 cells/well. *Denotes significant enhancement at p < 0.01 over values seen at 100 MPa. Lines: 1—R1, 2—Citrine, 3—Casp3KO, 4—RipK1KO, 5—RIPK3KO, 6—Casp8KO. (B) Initial plating density dependence on numbers of resulting transformants in pressure-treated ES cells. Cells were plated in a 24 well plate at 2000 or 10,000 cells/well as indicated. *Denotes significant enhancement at p < 0.01 over values seen at 2000 cells/well. (C) Co-transfection incidence in transfected R1 ES cells treated at 70 MPa with different reporter plasmid ratios at 48 h post transfection: 0 μg dTomato:5 μg EGFP; 2.5 μg dTomato:2.5 μg EGFP; 3.75 μg EGFP:1.25 μg dTomato; 4 μg EGFP:1 μg dTomato. Cells were plated at 25,000 cells/well in a 24 well plate. For each pressure condition, n = 3 independent experiments with three replicates within each experiment were performed. Results shown ± SD.

As with other biophysical methods of cell transfection, successful uptake of ectopic molecules represents a balance between sufficient and excessive injury to the cell membrane. Integral to this is the level of support enabling injured cells to recover. Compared to other types of primary cells, stem cells exhibit strong density-dependent effects with respect to survival and growth, therefore transfection efficiency was examined as a function of cell plating density34. As shown in Fig. 4B, higher initial plating densities resulted in significantly higher apparent transfection efficiency compared to lower plating densities over the pressure range of 60–80 MPa. This effect is likely due to the higher level of autocrine support present at higher cell densities in ES cells, in turn promoting better survival of transfected ES cells; rather than a direct effect on transfection efficiency.

A common feature of transfection methods which act through interruption of the cell membrane is the creation or ‘poration’ of transient holes in the phospholipid bilayer, followed by resealing on the time scale of seconds to minutes whereupon the cell membrane recovers its intrinsic impermeability35. To assess the ability of PJP to allow multiple plasmid uptake, the ability of pressure treated R1 ES cells to uptake multiple plasmid reporters was examined. As shown in Fig. 4C, once cells are in a state to take up one plasmid, their propensity to take up an additional reporter appears to be high, as variance of the relative reporter within the ratios indicated did little to alter the overall rate of double positive cells. This suggests that the primary limitation to transfection lies in achieving the competent uptake state, and that once achieved, the odds of continued survival with further uptake remain high, at least on the time scale of plasmid uptake.

Despite the strong propensity of ES cells to uptake multiple plasmids upon being rendered labile following PJP; the overall efficiency of this process (like many biophysical forms of gene transfection) is relatively low (< 1/100 cells). In order to better understand the population dynamics with respect to cell permeabilization following PJP, ES cells were examined at three stages, before, immediately after PJP or 15 min after PJP. As shown in Supplemental Fig. 1A, ES cells were first incubated in 10 μM Calcein-AM (green) and 5.7 μM (2 μg/mL) DAPI (blue) for 20 min at 37 °C and the resulting Calcein-AM and DAPI positive populations quantified. Immediately following pressure jump transfection at 70 MPa the populations shown in ‘Stage 2’ were characterized. Cells in which pressure treatment does not induce membrane permeabilization remain Calcein-AM+/DAPI− (approximately 59% of cells following 1 pressure jump), while the cells undergoing permeabilization of their cell membrane become DAPI+ /Calcein-AM− due to leakage in the immediate post-treatment period. DAPI positive cells present prior to treatment are indistinguishable from pressure-jump permeabilized cells by these measures. As shown in the figure, this population (DAPI+) rises from ~ 2 to ~ 41% following pressure treatment. However following the incubation of cells at room temperature for 15 min a small number of DAPI+ cells (thus disrupted by the pressure jump process), regain their membrane integrity following this recovery period as shown by their impermeability to 3 μM (2 μg/mL) propidium iodide added following recovery. Examples of this can be seen (white circles) in the population shown in Supplemental Fig. 1B. A subgroup of this population (green circles) even appears to recover some degree of Calcein-AM signal following membrane closure over this period, albeit at greatly reduced levels compared to non-disrupted cells.

Given that several straightforward methods currently exist to efficiently transfect immortalized cell lines, we instead initially focused our efforts on more difficult to transfect cells such as ES cells. Following determination of the basic functional parameters regulating transfection of ES cells using dynamic pressure, we sought to examine the ability of PJP to transfect other cell types. Initially we anticipated the response pattern across cell types to be similar to other biophysical methods such as electroporation. However, as shown in Fig. 5A, although PJP-induced transfection of primary fibroblasts demonstrated a pattern similar to that seen for ES cells, attempts at transfection in mouse L-cells using PJP were unsuccessful. Confirmation that this effect was unrelated to the reporter plasmid utilized is demonstrated by successful transfection of both L-cells and primary fibroblasts using standard electroporation. Similar attempts to pressure transfect well-characterized immortalized cell lines HEK293T and Cos-7 (lines efficiently transfected by electroporation, calcium phosphate mediated transfection and lipofection) were similarly unsuccessful (data not shown).

Figure 5
figure 5

Properties of pressure-treated cells. (A) Differences in pressure-mediated transfection efficiency in immortalized versus primary cells. Shown are relative transfection efficiencies of primary fibroblasts (blue) vs. mouse L-cells (red) by electroporation and pressure-mediated transfection. Pressurization was performed at of 4000 cells/μL with cells plated at 10,000 cells/well in a 6 well plate; electroporation standards plated similarly. For each pressure condition, n = 3 independent experiments with three replicates within each experiment were performed. Results shown ± SD. (BH) Cellular features and morphology of pressure-treated R1 ES cells. Live citrine-expressing ES cells were treated with 1.8 μM (1 μg/mL) Hoechst 33342, 10 nM (5 ng/mL) TMRM and 0.75 μM (0.5 μg/mL) propidium iodide prior to pressure treatment. (B) Cells in the absence of pressure treatment; (C) Typical appearances of ES cells following 1 min of pressure treatment at 100 MPa (pressurization at 4000 cells/μL, cells were held for 1 h at 25,000 cells/well in the slide chamber). (D) A small subgroup of these cells become PI+ but retain cellular features. (E) A major portion of these PI+ cells go on to exhibit features of cellular destruction in the immediate (30 min) post-treatment period. (FH) ES cells following 5 min at 100 MPa. The majority of these cells demonstrate features shown in (F,G). (G) A portion of recovered cells exhibit features of reduced cellular volume. (H) The great majority of cell which become PI+ following treatment at 100 MPa for 5 min exhibit features of cellular degeneration. For (BH) scale bar indicate in (B) represents 10 μm. (IK) Electron photomicrographs of ES cells. (I) Following rapid depressurization at 80 MPa, a population of ES demonstrated the presence of intracellular voids proximal to the cell membrane (red arrows), frequently associated with protuberances of the cell membrane (blue arrows). ES cell held at ambient pressure is shown for comparison (J). By contrast ES cells subjected to slow pressure release at 80 MPa, (K) often demonstrated extensive extrusions (blue arrowheads). For figures (IK) scale bar represents distance of 1 μm.

While previously there has been extensive characterization of several cell types in response applied hydrostatic pressure, the current study differs in that high pressure treatment terminates in a sudden drop to ambient pressure. In order to better understand how this process affects cellular dynamics, several morphological and intracellular features were examined at the upper limit of pressure for transfection and ES cell viability. For these experiments we used citrine-tagged ES cells expressing monomeric EYFP from the Rosa26 locus33, which were incubated with Hoechst 33342, tetramethylrhodamine methyl ester (TMRM) and propidium iodide at final concentrations of 1.8 μM, 10 nM and 0.75 μM respectively. As shown in Fig. 5B, cells untreated by pressure transfection exhibited intact nuclei with granules of TMRM fluorescence, indicative of dye accumulation in active mitochondria with an intact membrane potential. In Fig. 5C, the samples were held at a static pressure of 100 MPa for 1 min followed by sudden depressurization. The ES cells continued to demonstrate TMRM fluorescence, however with some reduction in the number of distinct granules suggestive of mitochondrial fusion. In addition, a small number of cells (~ 1%) began to demonstrate features indicative of entry of the cell permeability marker propidium iodide to identify membrane disruption (Fig. 5D). Additionally, a significant number (~ 30%) of cells developed morphologies like those shown in Fig. 5E over 5–30 min post-treatment. Such cells presented with overt morphologic cell disruption with membrane-bound cell fragmentation reminiscent of apoptotic cell death. If the static pressure is maintained for longer periods (5 min at 100 MPa, Fig. 5F–H), then the majority of cells continue to present intact plasma and nuclear membranes, with cytoplasmic compartments exhibiting granules of TMRM fluorescence albeit with a small decrease in the average signal intensity. A fraction of these cells (Fig. 5G) exhibited features consistent with significant reduction in cellular volume. Extended pressure treatment resulted in an increased number of cells exhibiting overt features of cellular degeneration with very few structurally intact cells exhibiting propidium iodide entry (Fig. 5H).

Given the apparent reduction in cellular volume seen under conditions of PJP, cell size was examined, and volumes estimated based upon area measurements through the cell diameter from free-floating cells as determined by optical sectioning. For ES cells, we calculated an interpolated volume of 1536.4 ± 19 μm3 prior to the pressure cycling process and 1250.6 ± 24 μm3 after exposure to 100 MPa for 5 min followed by sudden depressurization (n > 60). This result is consistent with an average reduction of volume of 18.6% following pressure treatment. However many immortalized lines demonstrate little change in volume, for example, HEK293T cells revealed interpolated volumes of 2022 ± 22 μm3 prior to pressure cycling, and 2251 ± 19 μm3 following exposure to 100 MPa for 5 min followed by sudden depressurization (n > 40); an apparent increase of 10.2%. Thus, the volume change appears to depend on the cell type.

To examine in greater detail the structural features resulting from PJP, TEM imaging was performed on ES cells fixed 5 min following exposure to one of three treatments: (1) non-pressurized cells resuspended in EmbryoMax electroporation buffer (EB) and loaded into borosilicate capillaries cells in the presence of reporter DNA for 10 min; (2) ES cells resuspended in EB loaded into capillaries in the presence of reporter DNA for 10 min followed by pressurization to 80 MPa, maintained at this pressure for 1 min, followed by sudden depressurization; (3) ES cells resuspended in EB and loaded into capillaries in the presence of reporter DNA for 10 min followed by pressurization to 80 MPa, maintained at this static pressure for 1 min, followed by slow depressurization over the course of 3 min to ambient pressure. As shown in Fig. 5I, while several treated cells exhibited features of overt rupture, a sub-population of cells experiencing sudden depressurization presented with the appearance of distinct cellular voids proximal to the cell membrane (red arrows), often in conjunction with irregular protrusions of the cell membrane and associated cytoplasm (blue arrows) compared to untreated cells (Fig. 5J). Notably little disturbance of the nuclear membrane or material was observed under these conditions. Paradoxically, ES cells experiencing a slow reduction in static pressure frequently exhibited significantly greater disruption of cell constituents, with gross disruption of cytoplasmic, membrane and nuclear constituents (Fig. 5K).

Based upon the functional criteria of continued survival and growth, the results demonstrated that even sensitive lines such as ES cells could survive for short periods at 100 MPa with minimal reductions in survival; however a number of transformed cell lines appear resistant to pressure-induced transfection (Fig. 5A). Given that the cellular response to pressure is known to be a function of both absolute pressure and the length of time to which cells are exposed23,25,28, we examined the potential nature of such differences among transfection- positive and transfection-negative lines by examining the survival properties of cell lines at different pressures and exposure periods. For each cell line investigated, equivalent cell densities were maintained at a specified pressure and time, followed by sudden return to ambient pressure. To determine relative levels of cell survival, treated cells were then plated and cultured in the manner described above. As shown in Table 1, cell types such as ES cells and primary fibroblasts demonstrated lower upper pressure limits with respect to survival compared to cell lines that exhibit transfection resistance such as mouse L-cells, exhibiting greater survival at significantly higher static pressures. A similar pattern was observed for other transfection resistant cell lines such as HEK293T, T24 and Lovo cells. Brightfield photomicrographs of these cell lines at three different levels of pressure treatment are shown in Fig. 6.

Table 1 Differential survival of primary and transformed cells after pressure treatment. Pressurization was performed at a concentration of 4000 cells/μL with cells plated at 25,000 cells/well in a 6 well plate. For each of the pressure treatments listed, the average percentages of surviving cells (D = dead) were determined from the initial number plated 24 and 72 h post-treatment, 3 wells/conditions with n = 3 independent experiments/treatment. Results are indicated < ± 3.5%. Determinations were performed in triplicate for 3 independent replicates with general survival characteristics indicated for each of the cell line.
Figure 6
figure 6

Brightfield photomicrographs of primary and transformed cells after pressure treatment. Indicated below are images of (A) ES R1 cells 24 h post-treatment, (B) HEK293T cells 24 h post-treatment, (C) T24 cells 24 h post-treatment, and (D) L-cells 72 h post-treatment for each of the conditions indicated. Scale bar in (A) represents 100 μm for all images in (AD).