Transfection of cells using polyethylenimine (PEI) polymers is a widely used technique for gene insertion and exogenous protein expression that offers several advantages over virus or lipid-based transfection methods. PEI facilitates endocytosis of DNA into the cell, release of DNA from the endolysosomal system and protection from nucleases. Although PEI is extensively investigated for application in research and therapy, the mechanism of its release from the endosomal system has remained a point of debate. The most prevailing model states that PEI causes endosomal rupture, resulting in release of PEI-DNA complexes into the cytosol. Hence, the use of PEI transfection should be applied with great care in studies on the endosomal system. Here we study the effect of PEI transfections on the endolysosomal system. Using fluorescent and electron microscopy we find a decrease in early endosomes after PEI transfection. Adapting the PEI transfection protocol by including a chase time after initial uptake of the PEI-DNA polyplexes rescues this phenotype without affecting transfection efficiency. Our data result in an adapted protocol for PEI transfection that can be used in studies involving the endolysosomal system.

Exogenous expression of (tagged) proteins is a commonly used method in Life Science studies. For example, to localize proteins using fluorescence or electron microscopy when endogenous protein levels are too low for detection or when there are no specific antibodies available. Also functional assays, for example, luciferase assays or co-immunoprecipitations, often require over-expression of (tagged) proteins to obtain sufficient levels for quantitative measurements. The transfection method used to express the protein of interest should have minimal effect on the cells to allow unbiased measurements. However, transfection protocols pose a number of strains onto cells in order to deliver foreign DNA to the nucleus and to escape the cells natural defense mechanisms against incorporation of non-self DNA^[1]. Polyethylenimine (PEI) is a synthetic polymer agent for nucleic acid delivery in vitro and in vivo. Since its initial discovery as a transfection agent^[2] it has become one of the most widely used gene transfer agents. The reasons for PEI’s popularity are: i) PEI is not expensive and can be stored for several years. ii) PEI is able to transfect many different cell lines with high efficiency. iii) The protocol to use PEI is simple. iv) PEI can be chemically modified to adapt it for special transfection procedures. These advantages make PEI a favored reagent for gene transfections in biochemistry and cell biology research (Reviewed in ^[3]), as well as a research target for the application of gene therapy^[4][5].

The efficiency of PEI in gene delivery relies on its chemical characteristics. PEI is a cationic polymer that efficiently binds to nucleic acids via charged amino groups^[2]. Despite extensive research, however, the details on the mechanism by which PEI transfects cells is still unclear. The PEI-DNA complexes, called polyplexes, condense into particles with a cationic surface, which can interact with anionic proteoglycans that are abundantly present on the cell surface^[6][7][8]. These particles are endocytosed via clathrin-dependent as well as clathrin-independent endocytosis^[9]. After endocytosis, the polyplexes need to escape from the endocytic system to the cytosol in order to reach the nucleus. The generally accepted mechanism for endosomal release is the “proton sponge” hypothesis^[10].

Endocytosis encompasses the transfer of endocytosed cargo from early endosome (EEs) to late endosomes (LEs) known as endosomal maturation. During endosomal maturation, the pH inside endosomes is lowered by resident ATPase proton pumps that actively translocate protons into the endosomes. This causes the PEI polymers to be protonated, inhibiting acidification of the endosomal lumen. The proton sponge hypothesis poses that since there is no change in pH, the ATPase proton pumps will continue to translocate protons into the endosomes, resulting in an accumulation of protons inside endosomes causing an influx of water and chloride ions. This causes osmotic swelling of endosomes leading to disruption of their membrane and consequent release of the PEI-DNA polyplexes to the cytosol^[11][12]. After release into the cytosol, a portion of the PEI-DNA polyplexes enters the nucleus, either via the nuclear pore complex or after dismantling of the nuclear envelope during cell division^[13].

Studies involving the endolysosomal system are often combined with protein expression to label distinct compartments, for example with fluorescent Rab5 or RAB7 to label EEs or LEs, respectively^[14][15]. However, since PEI transfections may cause endosomal membrane rupture, this imposes a problem for studying the endolysosomal system in combination with PEI-mediated transfection. Here we test to what extent the endolysosomal system is affected by PEI-mediated transfections. We show that PEI transfections specifically affect EEs and provide a protocol to overcome these PEI-induced artefacts.

In this study, we aimed to establish the effect of PEI-mediated transfection on the endolysosomal system and optimize the PEI transfection protocol to circumvent PEI-induced artifacts.

Transfected PEI-DNA polyplexes accumulate in the endosomal system

Endocytosis starts with vesicles that bud off from the plasma membrane (PM) and fuse with an early endosome (EE)^[16]. From EEs, endocytosed cargo can be recycled to the PM via recycling endosomes (REs)^[17]or proceed to late endosomes (LEs) and lysosomes for degradation^[18][19][20]. To track PEI-DNA polyplexes throughout the endolysosomal system, we transfected HeLa cells overnight with a GFP construct using a well-established PEI-based transfection protocol^[21] with a DNA:PEI ratio of 1:5 (protocol A). After overnight transfection, cells were fixed and prepared for electron microscopy (EM) to assess the cellular localization of the PEI-DNA polyplexes, which by EM are visible as dense aggregates of approximately 60–100 nm^[22]. Using this protocol, we could readily identify PEI-DNA polyplexes in EEs, LEs and lysosomes (Fig. 1A). The polyplexes appeared as individual particles (Fig. 1A, arrows left panel), or larger accumulations consisting of multiple particles. The proton sponge hypothesis predicts that PEI-mediated transfection causes endosomal swelling prior to polyplex release. To study if PEI indeed causes an increased endosomal size, we measured the size of endosomes after PEI-mediated transfection and control cells. Endocytosed PEI polyplexes prevent the morphological characterization of endosome subtype by morphology, therefore all endosomes were measured. This revealed a small yet significant increase in endosome size after transfection, indicating that endosomal swelling indeed occurred (Fig. 1A, graph). These data show that endocytosed PEI-DNA polyplexes traverses the entire endolysosomal system to lysosomes and induce swelling of endosomal compartments. PEI-mediated membrane rupture could occur at both early and late stages of endosomal maturation.

PEI transfection leads to a reduced early endosomal population

To test whether PEI-mediated transfection affects the composition of the endolysosomal system we used markers for specific endosomal compartments and quantified their distribution after PEI-mediated transfection. HeLa cells were transfected overnight with GFP using protocol A, after which cells were fixed and labeled for the early endosomal marker EEA1 or the LE/lysosomal marker LAMP-1. In addition, cells were incubated with the fluorescent endocytic marker dextran for 2 h before fixation to mark the entire endolysosomal system. Subsequent analysis revealed that EEA1 and dextran fluorescence was decreased in the transfected GFP-positive cells compared to the non-transfected cells in the wild-type condition, whereas LAMP-1 was not affected (Fig. 1B). Quantification showed that EEA1 and dextran were reduced 2-fold in transfected cells compared to non-treated cells, while LAMP-1 did not change significantly (Fig. 1B, graph). Decreased EEA1 fluorescence indicates a reduction in the number of EEs, or a reduction in recruitment of EEA1 to endosomes. By contrast, LAMP-1-positive LEs/lysosomes were not affected by the transfection. Reduced levels of dextran indicate that either endocytosis is slowed in transfected cells or disruption of endosomal membranes causes leaking of fluorescent dextran out of the compartment. Together these results indicate that the endolysosomal system is disrupted after PEI-mediated transfection and primarily affects EEs.

Lysosome integrity is not affected after PEI-mediated transfection

According to the proton sponge hypothesis, PEI-DNA polyplexes escape the endolysosomal pathway through membrane rupture. Several studies suggest that permeabilization occurs at the lysosome^[11][12][10], while our data indicate PEI-mediated transfection affects endosomes. Permeabilization of the lysosomal membrane can lead to release of lysosomal enzymes from the lysosomal lumen to the cytoplasm, which ultimately can cause cell death^[23][24]. To study lysosomal integrity after PEI-mediated transfection, we used a galectin3 based assay that detects lysosomal membrane permeabilization^[25]. In short, the stably transfected mAzami-green (mAG)-conjugated galectin3 protein (described in^[26]) is localized to the cytoplasm and nucleus in steady state conditions. However, upon disruption of the lysosomal membrane, galectin3 translocates into the lysosome and binds to β-galactosides in the glycocalix. This causes the appearance of fluorescent mAG-galectin3 puncta, which correspond to damaged lysosomes. We used stably transfected mAG-galectin3 HeLa cells to test lysosomal membrane permeabilization (LMP) after PEI-mediated transfection of a mCherry-construct using protocol A. We also included a positive control where we induced lysosomal leakage by using a lipid reagent. Both in control cells and cells transfected with protocol A, the mAG-galectin3 localization was primarily cytosolic. In contrast, clearly visible puncta were observed in the positive control (Fig. 1C). Quantification of mAG-galectin3 puncta per cell confirmed that PEI transfection does not increase the number or size of mAG-galectin3 puncta (Fig. 1C, graph). We conclude from these data that PEI transfection does not cause significant lysosomal membrane permeabilization.

The early endosome population recovers after a pulse-chase PEI transfection protocol

Our findings show that after overnight PEI transfection the population of EEA1-marked EEs is decreased (Fig. 1B). Since EEs are critical compartments for sorting and signaling of a wide variety of proteins, distortion of EE function is a highly unwanted side-effect in studies that require transfection. We reasoned that the impact of PEI on EE function might be restored over time and therefore developed a pulse-chase transfection protocol to increase the recovery time after transfection. To this end, cells were transfected for 8 h, washed to remove surplus PEI-DNA polyplexes and chased overnight with fresh medium. We refer to this transfection protocol as protocol B. Cells were then fixed and prepared for analysis by EM or IF microscopy. By EM, PEI-DNA polyplexes were observed throughout the endolysosomal system (Fig. 1D), similar as was observed in protocol A. Analysis of the endosomal size revealed a similar amount of endosomal swelling as protocol A (Fig. 1D, graph). However, quantification of the EEA1 signal by IF showed that protocol B restored EEA1 fluorescence area to control levels (Fig. 1E). These data show that the disruption of the endosomal population is restored by the chase period in protocol B.

The pulse-chase protocol has identical transfection efficiency

A prerequisite for use of our adapted pulse-chase transfection protocol B is an efficient transfection yield. To test this we used FACS analysis to determine GFP expression after protocol A or B (Fig. 1F). This showed that the transfection efficiency of HeLa cells was identical in the original (A) and adapted (B) protocol (Fig. 1F, graph). Together our data show that the effect of PEI-mediated transfection on the endolysosomal system is temporary and that cells are able to restore their endosomal system. Our adapted transfection protocol incorporates time for restoring endosome function while maintaining high expression levels of the transfected protein.

In this study we use fluorescence and electron microscopy to study the effect of PEI transfection on the endolysosomal system, using HeLa cells as a model system. We show that a widely used transfection protocol for PEI^[21] caused a significant decrease in EEA1-positive EE numbers. By contrast, LAMP-1-positive LE/lysosome numbers and integrity were unaffected. Importantly, adapting the standard overnight transfection protocol to a pulse-chase protocol restored the number of EEs to wild-type levels while retaining a high level of transfection activity. From these data, we conclude that PEI-mediated transfections particularly affect EEs but that this effect is temporary and can be reversed. We, therefore, recommend an adapted transfection protocol for studies that require an intact endolysosomal system in PEI transfected cells.

Escape from the endolysosomal system is an essential step for successful transfection^[27]. For cationic polymers such as PEI, the proton sponge model for endosomal escape indicates that rupture of endosomal membranes is needed for transfer of DNA from the endosomal lumen to the cytoplasm. By EM we found PEI-DNA polyplexes in early and late endosomes as well as lysosomes, which is in agreement with earlier observations^[10][12]. Notably, despite the presence of PEI-DNA polyplexes in lysosomes, we observed no change in LE/lysosome number and no lysosomal membrane permeabilization could be detected. This indicates that lysosomes are not affected by PEI, implicating that PEI-DNA polyplexes that reach lysosomes will be degraded rather than transferred to the cytosol. An explanation for this is that lysosomes have heavily glycosylated limiting membranes that protect them from hydrolase activity^[19][20]and may protect them from rupture.

We revealed a significant decrease in EEA1 fluorescence signal after PEI-mediated transfection. One possible explanation is that the number of EEA1-positive EEs is decreased. If osmotic rupture disrupts endosomal membranes, it is likely that damaged endosomes are cleared from the cytosol by autophagy, a mechanism that is known to clear damaged endosomes and lysosomes^[28][29]. Alternatively, or additionally, disruption of the endosomal membrane could prevent recruitment of effector proteins such as EEA1 to EEs. In addition to decreased levels of EEA1, we found a significant decrease in dextran levels after PEI-mediated transfection. This can be explained by either a decrease in endocytosis or by leakage of dextran from damaged endosomes to the cytosol. Since PEI-DNA polyplexes are endocytosed themselves, and we could detect polyplexes in EEs after transfection (Fig. A), it is likely that the latter explanation causes the decrease in dextran fluorescence. Dextran leaking into the cytosol is diluted, resulting in loss of fluorescent signal and a decrease in fluorescent puncta.

Other commonly used transfection methods that require endosomal escape are the use of lipid-based reagents or calcium phosphate precipitates (CCPs). CCPs induce endosomal damage, upregulating autophagy resulting in major alterations of the endolysosomal system after transfection^[28]. CCP-based transfection protocols could therefore also potentially profit from a protocol that includes a chase period. By contrast, the destabilizing effect of the lipid complexes on the endolysosomal system is less damaging since they do not cause osmotic swelling or membrane rupture, but back-fuse with the limiting membrane of endosomal compartments^[30][31][32], causing only minor disruptive effects on endosomal membrane integrity. However, in studies that require an intact endolysosomal system the use of any transfection method that requires endosomal escape could potentially benefit from a protocol that includes a chase period to allow restoration of endosome numbers.

In summary, our data show that PEI-induced alterations to the endolysosomal system are limited to the early endocytic compartments and transient in nature. Application of an adapted protocol that includes a chase period is an easy tool to circumvent this disruption.

By EM we show that endocytosed PEI-DNA polyplexes are found in early and late endosomes, as well as lysosomes.

PEI-mediated transfection causes a minor swelling of endosomes.

PEI-mediated transfection disrupts the endolysosomal system of HeLa cells and primarily affects EEs.

PEI-mediated transfection does not cause significant lysosomal membrane permeabilization in HeLa cells.

The disruptive effect of PEI-mediated transfection on the endolysosomal system is reversible.

Adapting the transfection protocol to incorporate a chase period results in the restoration of endosome numbers while maintaining high expression levels of the transfected protein.

Cell culture

Hela cells (ATCC clone ccl-2) were cultured in Dulbecco’s modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated Fetal Bovine Serum (FBS), 2 mM L-Glutamin, 100 U/mL penicillin and 100 µg/mL streptomycin (complete DMEM) at 37°C and 5% CO2.

Antibodies and reagents

Mouse-anti-human EEA1 (BD Transduction Lab), mouse-anti-human LAMP-1 CD107a (BD Pharmingen), Fluorescent secondary antibodies were obtained from Invitrogen, Fluorescent 10.000 MW dextran-Alexa Fluor 568 (Invitrogen), linear 40 kDa polyethylenimine (PEI) (Polysciences inc.), effectene transfection reagent (Qiagen).

Transfection protocols

HeLa cells were transfected with Pcdna3.2-GFP cDNA using linear 40 kDa polyethylenimine (Polysciences inc.) with a DNA:PEI ratio of 1:5 using protocol described in^[21]. In the standard protocol, the transfection mix was added to the HeLa cells for 16 h. In the adapted protocol cells were washed 3 times and chased for 16 h with 37°C complete DMEM after being subjected to the PEI-DNA mixture for 8 h. Effectene transfection reagent (Qiagen) was used as a positive control for lysosomal leakage assay using a 5 times higher concentration of effectene than the manufacturer’s protocol.

Immunofluorescence microscopy

HeLa cells grown on sterile glass coverslips were washed with ice-cold PBS and fixed with 4% paraformaldehyde (PFA) in PBS for 20 min at RT (room temperature). Then, cells were permeabilized using 0,1% Triton-X100 in PBS for 5 min and blocked for 15 min using PBS supplemented with 1% BSA. Cells were labeled with primary antibodies diluted in blocking buffer at RT for 1 h, washed, and labeled with fluorescent secondary antibodies for 30 min in the dark. After labeling the cells were washed and mounted using Prolong Gold antifade reagent with DAPI. Cells were imaged on a Deltavision wide field microscope using a 100X/1.4NA immersion objective. Widefield pictures were deconvolved using Softworx software and analyzed using FIJI.

Electron microscopy

Cells grown on 6 cm dishes were fixed in 2% wt/vol PFA, 2.5% wt/vol GA in Na-cacodylate buffer (Karnovsky fixative) for 2 h at RT. Subsequently, the Karnovsky fixative was replaced with 0.1 M Na-cacodylate buffer, pH 7.4. Post-fixation before Epon embedding was performed with 1% wt/vol OsO4, 1.5% wt/vol K3Fe(III)(CN)6 in 0.065 M Na-cacodylate for 2 h at 4°C. Subsequently, the cells were stained with 0.5% uranyl acetate for 1 h at 4°C, dehydrated with ethanol, and embedded in Epon. Ultrathin sections were stained with uranyl acetate and lead citrate. Images were taken on a TECNAI T12 electron microscope.

Lysosomal membrane permeabilization assay

The galectin3-based assay detects lysosomal membrane permeabilization as described in^[25]. HeLa cells were stably transfected with the mAzami-green (mAG)-conjugated galectin3 construct (described in^[26], Addgene Plasmid #62734) using lentiviral transduction. In steady state conditions, mAG-Gal3 is localized to the cytoplasm and nucleus. Upon lysosomal membrane permeabilization (LMP) galectin3 is able to bind N-acetyllactosamine-containing glycans present on the inner membrane of lysosomes, leading to an accumulation selectively in damaged lysosomes. Effectene transfection reagent (Qiagen) was used as a positive control for lysosomal leakage assay using 5 times higher concentration of Effectene than the manufacturer’s protocol. Of note, recommended use of Effectene transfection reagent does not induce significant lysosomal leakage.

FACS analysis

Transfection efficiency was measured by FACS using the GFP signal after PEI transfection of PCDNA3.2-GFP in HeLa cells using 20.000 cells analyzed on a BD CANTO-II FACS.

We thank Ann De Mazière and George Posthuma for assistance with the electron micrographs and René Scriwanek for preparation of EM figures.

The authors declare no conflicts of interest.

Not applicable.

No fraudulence is committed in performing these experiments or during processing of the data. We understand that in the case of fraudulence, the study can be retracted by ScienceMatters.