HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 20, 18021-18028, May 17, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/20/18021    most recent M111257200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Zuhorn, I. S. Articles by Hoekstra, D. Search for Related Content PubMed PubMed Citation Articles by Zuhorn, I. S. Articles by Hoekstra, D. Social Bookmarking                      What's this?

Lipoplex-mediated Transfection of Mammalian Cells Occurs through the Cholesterol-dependent Clathrin-mediated Pathway of Endocytosis^*

Inge S. Zuhorn, Ruby Kalicharan^§, and Dick Hoekstra^¶

From the  Department of Membrane Cell Biology, University of Groningen, A. Deusinglaan 1, Groningen 9713 AV, and the ^§ Laboratory for Cell Biology and Electron Microscopy, University of Groningen, Oostersingel 69, Groningen 9713 EZ, The Netherlands

Received for publication, November 27, 2001, and in revised form, January 28, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

Synthetic amphiphiles are widely used as a carrier system. However, to match transfection efficiencies as obtained for viral vectors, further insight is required into the properties of lipoplexes that dictate transfection efficiency, including the mechanism of delivery. Although endocytosis is often referred to as the pathway of lipoplex entry and transfection, its precise nature has been poorly defined. Here, we demonstrate that lipoplex-mediated transfection is inhibited by more than 80%, when plasma membrane cholesterol is depleted with methyl--cyclodextrin. Cholesterol replenishment restores the transfection capacity. Investigation of the cellular distribution of lipoplexes after cholesterol depletion revealed an exclusive inhibition of internalization, whereas cell-association remained unaffected. These data strongly support the notion that complex internalization, rather than the direct translocation of plasmid across the plasma membrane, is a prerequisite for accomplishing effective lipoplex-mediated transfection. We demonstrate that internalized lipoplexes colocalize with transferrin in early endocytic compartments and that lipoplex internalization is inhibited in potassium-depleted cells and in cells overexpressing dominant negative Eps15 mutants. In conjunction with the notion that caveolae-mediated internalization can be excluded, we conclude that efficient lipoplex-mediated transfection requires complex internalization via the cholesterol-dependent clathrin-mediated pathway of endocytosis.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

Currently, several carrier systems, including those based on synthetic cationic amphiphiles, are exploited for delivery of DNA constructs into cells for cell biological or therapeutic purposes. Compared with viral vectors, the transfection efficiency with most of the amphiphilic carriers ("lipoplexes") is still relatively low. However, because the latter offer considerable advantages over the former in terms of biological inertness, health risks, and large scale production, efforts are ongoing to improve their effectiveness of delivery. To achieve this goal it will be imperative to carefully define their mechanism of cellular entry. In fact, the mechanism of uptake of cationic amphiphilic gene carriers by cells is still a matter of debate. Early work suggested that cationic amphiphile-DNA complexes could enter the cell via fusion with the plasma membrane (1). Although attractive given its membranous nature, lipid mixing assays did not reveal a correlation between fusion events of lipoplexes with cellular membranes and their transfection efficiency (2-5). Next to fusion, a mechanism that involves internalization via endocytosis has received most support thus far (6-8). The evidence is often based on the application of metabolic inhibitors of endocytosis like chloroquin, monensin, and NH₄Cl. However, the precise effect of a certain metabolic inhibitor is often difficult to interpret because both attenuation and diminution of transfection efficiency have been reported, while using one and the same inhibitor (6, 9-13). An additional complication of the use of several of these inhibitors is that some of these compounds perturb processes following the actual internalization of the DNA construct or plasmid into cells.

The goal of the present work was to more accurately define the involvement per se and nature of the endocytic process in accomplishing cationic amphiphile-mediated DNA delivery. To this end, transfection was determined following cholesterol depletion of the plasma membrane with methyl--cyclodextrin (MCD)¹ (14-16). As has been recently shown, cholesterol depletion, as accomplished in this manner, results in the inhibition of clathrin-mediated endocytosis, whereas non-clathrin-mediated endocytosis is much less affected. It was demonstrated that the decrease in clathrin-mediated endocytosis is caused by the disability of coated pits to invaginate and detach from the plasma membrane (17, 18).

Here, we have determined the transfection efficiency of COS7 cells, using lipoplexes composed of the cationic amphiphile SAINT-2 and dioleoylphosphatidylethanolamine (DOPE) (5, 19, 20), following depletion of plasma membrane cholesterol with MCD. Because MCD treatment of cells, next to an interference of coated pit organization, may potentially interfere with the internalization via caveolae as well, the effect of two additional drugs, filipin III and chlorpromazine, were tested to differentiate between the involvement of either type of endocytosis (non-coated pit (caveolae) versus coated pit (clathrin-dependent)). In addition, colocalization studies with transferrin, which is internalized via the clathrin-coated endocytic pathway, were carried out. Moreover, the effects of potassium depletion and expression of dominant negative mutants of Eps15 (epidermal growth factor receptor pathway substrate clone 15) on lipoplex internalization were investigated, both treatments modulating the pathway of clathrin-mediated endocytosis (21-23).

Our data demonstrate that SAINT-2/DOPE-mediated transfection of cells requires prior internalization and that effective transfection depends on cellular entry of the complex via clathrin-mediated endocytosis.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

Lipoplex Preparation and Transfection-- Lipoplexes were prepared in serum-free cell culture medium. An equal volume of a 2 µg/ml plasmid DNA solution (pEGFP-N1, CLONTECH) was added to a 30 µM solution of SAINT-2/DOPE (1:1) liposomes, resulting in the formation of lipoplexes with a molar charge ratio of 2.5:1 (cationic lipid:DNA). SAINT-2 was synthesized as described in detail elsewhere (24); DOPE was purchased from Avanti Polar Lipids (Birmingham, AL). One day before transfection, COS7 cells were seeded into 6-wells plates at 3 × 10⁵ cells per well. Prior to transfection, cells were depleted of cholesterol (see below) and/or washed once with HBSS (Calbiochem). 1 ml of the lipoplex solution (15 nmol of lipid, 1 µg of pDNA) was added per well and incubated for 4 h at 37 °C. Then the transfection medium was replaced by cell culture medium. After 24 h the medium was refreshed, whereas after another 24 h the cells were screened for GFP-reporter gene expression by FACS-analysis (Elite, Coulter; _ex. 488 nm, _em. 530 nm; 5000 events).

Cholesterol Depletion of COS7 Cells-- COS7 cells were incubated with a 10 mM MCD solution in serum-free cell culture medium for 1 h at 37 °C. Lovastatin (1 µg/ml), an inhibitor of 3-hydroxy-3-methylglutaryl-CoA reductase, was added, during both cholesterol depletion and subsequent transfection, to prevent refill of the depleted plasma membrane cholesterol pool by newly synthesized cholesterol. Alternatively, plasma membrane cholesterol was bound by the pentaene macrolide antibiotic filipin III (1 µg/ml, 1 h, 37 °C), which was kept present during the entire transfection.

Cholesterol Replenishment of COS7 Cells-- To investigate the reversibility of the effect of cholesterol depletion with MCD on the transfection efficiency with lipoplexes, we replenished cellular cholesterol prior to transfection according to two protocols. First, following cholesterol depletion, replenishment of the cellular cholesterol pool was allowed to occur via de novo biosynthesis by incubating the cells overnight with complete (serum-containing) cell culture medium prior to transfection. Alternatively, cholesterol was repleted exogenously by incubation with a cyclodextrin-cholesterol complex (CDCHOL) in the presence of lovastatin for 1.5 h at 37 °C. CDCHOL was made as described by Klein et al. (25). Briefly, 125 mg of MCD was dissolved in 2.5 ml of millipore water. 3.75 mg of cholesterol (in chloroform) was added in small aliquots to the MCD solution under continuous stirring in a water bath at 75 °C. After complete dissolution of the cholesterol, the solution was quickly frozen in liquid nitrogen and freeze-dried overnight. The remaining powder was dissolved in 0.94 ml of HBSS to make a CDCHOL stock solution containing 4 mg/ml cholesterol. Cells were incubated with CDCHOL (400 µg/ml cholesterol) by diluting the stock solution in serum-free cell culture medium.

Quantification of Cellular Internalization of Lipoplexes-- To quantify the amount of cell-associated and internalized lipoplexes in control and cholesterol-depleted COS7 cells, we included ¹⁴C-labeled SAINT-2, prepared as described elsewhere (26) (specific activity: 8 × 10⁵ dpm/ml), in the initial liposome formulation. After 1-4 h of incubation with lipoplexes cells were treated with CellScrub buffer, according to the manufacturer's instructions, to remove surface-associated lipoplexes. In a control experiment (not shown), we determined, using this procedure, that 85-90% of the cell-associated lipoplex fraction could be removed following an incubation at 4 °C. The fractions of associated (CellScrub buffer supernatants) and internalized lipoplexes (cellular fraction, collected by trypsinization) were counted in a scintillation counter.

Fluorescence Microscopy of Transfected COS7 Cells-- To follow the effect of cholesterol depletion on the internalization of lipoplexes by COS7 cells, 0.5 mol% of the fluorescent lipid N-lissamine rhodamine B phosphatidylethanolamine (N-Rh-PE) was included in the initial liposome formulation. After 1 h of incubation with lipoplexes, the cells (grown on coverslips) were washed twice with HBSS and examined with an Olympus fluorescence microscope. To distinguish between extracellularly and intracellularly localized lipoplexes, the samples were incubated in a 0.2% trypan blue solution in HBSS (10 min, RT) to quench extracellular fluorescence. To determine the efficiency of the inhibitory effect of filipin III treatment on caveolar functioning, we followed its effect on the internalization of Bodipy-LacCer, a sphingolipid that is internalized via caveolae (27). To this end, COS7 cells were incubated with filipin III for 1 h at 37 °C, followed by an incubation with 3 µM Bodipy-LacCer (Molecular Probes, Eugene, OR) in serum-free medium in the presence of filipin III (30 min at 10 °C). The cells were subsequently washed with cold HBSS and transferred to 37 °C for 5 min, followed by "back exchange" (six washes for 10 min in 5% bovine serum albumin (Sigma Chemical Co., St. Louis, MO) in HBSS at 10 °C) to remove extracellular lipid. Fluorescence images were taken with an Olympus camera. The amount of intracellular fluorescence was quantified using computer software (Scionimage, Scion Corp.). 7-10 fields of ~10 cells/field/condition were analyzed.

Electron Microscopy of Transfected COS7 Cells-- After 1 h of transfection, cells were first fixed for 30 min at RT in 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4. The cells were rinsed trice in the same buffer with 6.8% sucrose and subsequently postfixed in 1% OsO₄/1.5% K₄Fe(CN)₆ followed by an en bloc staining with 1% uranyl acetate in water. For visualization of clathrin-coated vesicles, cells were fixed for 1 h at RT in 1.5% glutaraldehyde in 0.1 M cacodylate buffer containing 1% sucrose, pH 7.4. Subsequently, cells were stained with 2% tannic acid and 1% uranyl acetate. After rinsing followed by dehydration in graded alcohol series, the cells were finally embedded in Epon 812 and polymerized 4 days at 45 °C. Ultrathin sections (60 nm) were cut and stained with uranyl acetate and lead citrate. The sections were examined using a Philips CM 100 electron microscope operating at 60 kV, and micrographs were taken.

Colocalization Study of Lipoplexes and Transferrin-- COS7 cells grown on coverslips were incubated with N-Rh-PE-labeled (0.25 mol%) lipoplexes (30 min, 37 °C), extensively washed with HBSS, and subsequently incubated with fluorescein-labeled transferrin (Molecular Probes) for 5 min at 37 °C. Following the incubation with transferrin, the cells were washed with HBSS (4 °C) and kept on ice until examination by confocal microscopy (Leica TCS SP2, Germany; N-Rh-PE-lipoplexes: _ex 543 nm, _em 600-700 nm; fluorescein-transferrin: _ex 488 nm, _em 500-530 nm). Settings were chosen such that no red signal was visible in the green channel and vice versa.

Potassium Depletion of COS7 Cells-- COS7 cells were rinsed with K⁺-free buffer (140 mM NaCl, 20 mM Hepes, 1 mM CaCl₂, 1 mM MgCl₂, 1 mg/ml D-glucose, pH 7.4) and subsequently rinsed for 5 min with hypotonic buffer (K⁺-free buffer diluted 1:1 with distilled water). Cells were then washed three times with K⁺-free buffer and incubated with N-Rh-PE-labeled lipoplexes in K⁺-free buffer for 1 h at 37 °C. Extracellular fluorescence was quenched by incubation in trypan blue prior to examination by fluorescence microscopy, as described above. As a control, cells were incubated with lipoplexes in a similar buffer containing 10 mM KCl. The amount of intracellular fluorescence was quantified using Scionimage. 7-10 fields of ~10 cells/field/condition were analyzed.

Expression of Eps15 Mutants in COS7 Cells-- EH21, DIII, and D32 GFP-constructs were a kind gift from Alexandre Benmerah and Alice Dautry-Varsat (Institut Pasteur, Paris, France). COS7 cells were transfected with the above-mentioned constructs as described under "Lipoplex Preparation and Transfection." Two days after transfection, cells were incubated for 1 h at 37 °C with N-Rh-PE-labeled lipoplex. After quenching of extracellular fluorescence with trypan blue, the amount of internalized lipoplex was measured by fluorescence microscopy and quantified using Scionimage (7-10 fields of ~10 cells/field/condition).

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

Cholesterol Depletion Inhibits Lipoplex-mediated Transfection-- As shown previously (5, 20) COS7 cells are readily transfected with SAINT-2/DOPE lipoplexes, giving rise to transfection efficiencies as high as 80-90%. When the cells had been depleted of cholesterol, as accomplished by treatment with 10 mM MCD for 1 h prior to transfection, the efficiency, expressed as a percentage of GFP reporter gene-expressing cells, decreased by more than 80% (Fig. 1a). Lovastatin, an inhibitor of 3-hydroxy-3-methylglutaryl-CoA reductase, was included in the incubation medium during the incubation with lipoplexes to prevent de novo synthesis of cholesterol and re-supply of the extracted pool. Transfection in the presence of lovastatin per se, i.e. without depletion of plasma membrane cholesterol with MCD, resulted only in a minor decrease in transfection efficiency (~10%, data not shown), implying that the inhibition of transfection efficiency specifically relates to cholesterol depletion. Depletion of cellular plasma membranes of cholesterol in this manner has been specifically associated with an interference of clathrin-mediated endocytosis (17, 18). Because the disappearance of cholesterol-rich caveolae cannot be excluded, the next set of experiments were aimed at obtaining further insight into the mechanism by which lipoplexes were internalized by cultured cells.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   a, removal of cholesterol inhibits cellular transfection with SAINT-2/DOPE lipoplexes. COS7 cells were treated with 10 mM MCD for 1 h to deplete the pool of plasma membrane cholesterol. Subsequently, the cells were transfected in the presence of lovastatin as described under "Materials and Methods" (chol). Alternatively, following depletion, replenishment of the depleted cholesterol pool was carried out by exogenous addition of a cholesterol-cyclodextrin complex in the presence of lovastatin (+chol, exo) or by incubating the cells overnight, to allow recovery via endogenous synthesis (+chol, endo), after which the cells were transfected as described. Experiments were performed three times, in duplicate. Error bars represent standard deviation between the means of the three experiments. The transfection efficiency of control cells was set as 100%. b, perturbation of clathrin- rather than caveolae-localized cholesterol inhibits transfection with SAINT-2/DOPE lipoplexes. Cellular cholesterol pools, associated with clathrin (coated endocytosis) and caveolae (non-coated endocytosis) were modulated by treatment of COS7 cells with chlorpromazine (10 µg/ml, CPM), filipin III (1 µg/ml, FIII), or cytochalasin D (10 µg/ml, cytD), respectively. Following treatment the cells were transfected as described. Experiments were performed three times in duplicate. Error bars represent standard deviation between the means of the three experiments. The transfection efficiency of control cells was set as 100%.

Cholesterol Depletion Inhibits Lipoplex Internalization, Not Binding-- To examine whether cholesterol depletion might have affected the extent of interaction of lipoplexes with the cells, which could consequently have led to a diminished internalization and, presumably, transfection, we compared the amount of cell surface-associated radiolabeled lipoplexes in cholesterol-depleted and untreated cells. As shown in Fig. 2, irrespective of MCD treatment there was no difference in the total cell-associated fractions of lipoplexes. However, after subsequent treatment with CellScrub buffer, the amount of internalized lipoplexes (see "Materials and Methods") in cholesterol-depleted cells was substantially reduced. Thus, over an incubation period of 4 h, an inhibition of complex internalization of ~60-70% was seen after cholesterol removal.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Cholesterol depletion inhibits lipoplex internalization, not total cell-association. Plasma membrane-localized cholesterol in COS7 cells was depleted with MCD (squares; control cells are represented by triangles). The cells were subsequently incubated with SAINT-2/DOPE complexes, labeled with a trace amount of 14C-SAINT-2. Following the incubation times as indicated the cells were extensively washed with CellScrub buffer as described. The remaining cell-associated radioactivity is considered to represent internalized complex (open symbols). The sum of cell-associated radioactivity and that recovered in the supernatant represents the total cell-associated fraction (closed symbols). Experiments were performed three times, in duplicate. Error bars represent the standard deviation between the means of the three experiments. The internalized fraction (dpm/µg of protein) in control cells after 4 h of incubation was set as 100%.

To further characterize the effect of cholesterol depletion on cellular processing of the lipoplexes, we monitored their fate by fluorescence microscopy, using N-Rh-PE-labeled lipoplexes. As shown in Fig. 3, following cholesterol depletion of COS7 cells with MCD, a typical chain-like accumulation of lipoplexes at the cell surface became visible. Also note that the morphology of cells depleted of cholesterol was changed from a round to a more jagged shape, as noted before (18). After incubation with trypan blue, which quenches accessible rhodamine fluorescence and does not penetrate viable cells, the typical chain-like staining at the cell surface disappeared, consistent with their surface localization. As anticipated, compared with control cells, cholesterol-depleted cells showed a considerable decrease in intracellularly accumulated fluorescence (Fig. 3, C versus D), which is entirely consistent with the data obtained when quantifying the intracellular fraction, as determined by radiolabeling (Fig. 2). Moreover, electron micrographs of cholesterol-depleted COS7 cells further confirmed the strong reduction in intracellularly localized complexes, implying that cholesterol depletion inhibited complex internalization. Thus as demonstrated in Fig. 3 (E and F), the cytoplasm of untreated cells revealed the presence of numerous endosomal structures containing electron-dense internalized material (presumably representing internalized lipoplexes), whereas most cholesterol-depleted cells were virtually devoid of such structures. When cells were investigated early after the onset of transfection (10 min) electron-dense material was detectable in coated vesicles (Fig. 4, B, F, and G), whereas in non-transfected cells the coated vesicles appeared empty (A, D, and E). Strikingly, in transfected cells the coated vesicles were more abundant and often localized in clusters compared with their appearance in non-transfected cells (cf. Fig. 4, A and B). Cells that were depleted of cellular cholesterol prior to incubation with lipoplex showed flat-coated pits (Fig. 4C), consistent with data presented by other laboratories (17, 18).

View larger version (133K): [in this window] [in a new window]   Fig. 3.   Cholesterol depletion causes lipoplex accumulation at the cell surface and a diminishment in complex internalization. Control and cholesterol-depleted COS7 cells were incubated for 1 h with N-Rh-PE-labeled lipoplexes. Note the scattered cell-associated appearance of complexes with control cells (A), whereas cholesterol-depleted cells show a typical chain-like appearance of complexes, closely associated with the cell periphery (B). After quenching of the extracellular fluorescence with trypan blue, the chain-like staining of lipoplexes in cholesterol-depleted cells disappeared, revealing a reduced internalization compared with control cells (C and D). The reduced intracellular accumulation of complexes following cholesterol depletion was confirmed by electron microscopic analysis. Note the strongly enhanced accumulation of electron dense material within compartments that presumably represent endosomal compartments, when comparing control cells (E) versus cholesterol-depleted cells (F). Inserts in pictures A, C, and D represent corresponding phase contrast images. Bars in images E and F indicate 5 µm.

View larger version (122K): [in this window] [in a new window]   Fig. 4.   Lipoplex internalization results in the appearance of electron dense material in coated vesicles, whereas MCD treatment precludes the pinching off of coated pits. In control cells (A, D, E) coated vesicles are conspicuously devoid of any electron-dense material. By contrast, in transfected cells (B, F, G) coated vesicles often appear in clusters, are more abundant than in control cells, and contain electron-dense material. In MCD-treated cells (C) flat-coated pits could be visualized. The scale bar in A also applies to B and C and represents 1355 nm. The scale bar in D also applies to E, F, and G and represents 205 nm. Arrows indicate coated pits and vesicles.

In summary, the data demonstrate that cholesterol depletion following treatment of the cells with MCD resulted in an extensive decrease (60-70%) in the internalization of lipoplexes, without affecting the total cell-association fraction. As a consequence, a dramatic decrease in transfection efficiency (approximately 80%) was observed. These data imply that to accomplish productive transfection, the plasmid requires prior internalization, rather than that it gains direct cytosolic excess by crossing the plasma membrane.

Re-supply of Cholesterol Restores Transfection Efficiency-- To investigate whether the transfection efficiency after cholesterol depletion with MCD was lowered due to the absence of cholesterol in the plasma membrane, we next examined whether the effect of cholesterol depletion on transfection efficiency was reversible. To this end we took advantage of the fact that cholesterol can be transferred to cells via preformed cyclodextrin-cholesterol complexes because of the natural preference of cholesterol to partition into a phospholipid-rich environment (e.g. the plasma membrane) (28). As shown in Fig. 1a, by incubating cholesterol-depleted cells with a preformed cyclodextrin-cholesterol complex in the presence of lovastatin we were able to restore the level of transfection efficiency up to 85% of that obtained for non-treated cells. Alternatively, restoration of plasma membrane cholesterol content via de novo synthesis, following an overnight incubation of the cells in serum-containing medium (endogenous), also led to an essentially complete recovery of the transfection efficiency.

Lipoplexes and Transferrin Colocalize in Early Endocytic Compartments-- In light of previous studies on the effect of cholesterol depletion and re-supply on endocytosis, showing an inhibition and recovery of clathrin-mediated endocytosis of ligands like transferrin (Tf) and epidermal growth factor, respectively (17, 18), the data obtained thus far would similarly support a mechanism that would involve clathrin-mediated endocytosis as a critical step in the mechanism of lipoplex-mediated transfection. To obtain further support for such a mechanism, we therefore verified in the present system the processing of Tf and determined whether the lipoplexes colocalized with this ligand in early endocytic compartments. As shown in Fig. 5 (A and B), cholesterol depletion inhibits internalization of Tf uptake in COS7 cells, as clearly revealed by the distinct (fluorescent) membrane boundary, apparent in cholesterol-extracted cells (Fig. 5B), indicating that Tf internalization, but not binding, was inhibited. Hence these data imply that cholesterol depletion inhibits clathrin-mediated endocytosis in COS7 cells. To examine the fate of the lipoplexes relative to that of Tf, we first incubated the cells with N-Rh-PE-labeled lipoplexes for 30 min. Subsequently, the cells were labeled with fluorescein-labeled transferrin for 5 min. Fig. 5C shows the merged image of confocal microscopy scans for N-Rh-PE-lipoplex and fluorescein-transferrin. Because transferrin (~80 kDa) is much smaller than the average lipoplex (around 200 nm) Fig. 5D is included to facilitate visualization of the small yellow dots, indicating colocalization of transferrin and lipoplex.

View larger version (67K): [in this window] [in a new window]   Fig. 5.   Lipoplex internalization, like endocytosis of Tf occurs, through clathrin-mediated endocytosis. COS7 cells were incubated with fluorescein-labeled transferrin for 5 min at 37 °C, and the ligand's intracellular distribution was examined by confocal microscopy, as described under "Materials and Methods." Note the presence of numerous small vesicles in the cytoplasm of the cell, representing early endosomal compartments and a more intensely labeled and enlarged spot, which presumably represents the perinuclear localized recycling compartment (A). MCD treatment of COS7 cells results in an uptake block for transferrin, as reflected by a diffuse surface staining and, in the appropriate plane of focus, a clear and distinct plasma membrane labeling (B). In C, the cells had been incubated for 30 min with N-Rh-PE-labeled lipoplexes, which was followed by an incubation for 5 min with fluorescein-labeled Tf, at 37 °C. Enlargement of at random areas (D) reveal yellow dots, showing Tf being (co-)localized in lipoplex-containing compartments.

Is it realistic to assume that complexes with sizes as large as up to 200 nm can be accommodated in clathrin-coated vesicles? Although coated vesicles in neuronal synapses typically have a size of 70-90 nm, coated vesicles that mediate receptor uptake can reach sizes up to 200 nm (29-31). Indeed, in cells that were incubated with lipoplexes, coated vesicles are seen with sizes of at least 100-200 nm, as revealed upon electron microscopic examination (Fig. 4, F and G). In fact, it has been reported that particles as large as bacteria (Escherichia coli), when opsonized, can be internalized via the pathway of clathrin-mediated endocytosis (32, 33), suggesting that clathrin-mediated endocytosis of particles exceeding diameters of 200 nm is not unprecedented and that the process may involve a considerable degree of size flexibility of the coated vesicles. Nevertheless, although cultured cells like COS7 cells (i.e. non-macrophages) are not known to display substantial phagocytic activity, a mechanism responsible for internalization of particles by macrophages, we determined the possible involvement of phagocytosis in lipoplex internalization. To this end, COS7 cells were incubated with the actin-disrupting agent cytochalasin D (10 µg/ml, 30 min, 37 °C), a well-established inhibitor of phagocytosis (34, 35) prior to transfection. As is shown in Fig. 1b, disruption of the actin filament network with cytochalasin D does not inhibit transfection efficiency with lipoplex, demonstrating an actin-independent process. Moreover, phagocytosis is known to be clathrin-independent (36), implying that the data are incompatible with a significant contribution of phagocytosis to the mechanism of lipoplex internalization by cells, leading to productive transfection.

Presumably, lipoplexes bind (nonspecifically) to cellular receptors, which are or become clustered in coated regions, followed by internalization. Commonly, internalization of a ligand-receptor complex via coated vesicles is a very fast process, and a complete cycle, from coated pit formation until vesicle trafficking and uncoating, can occur within 1 min. Compared with such kinetics, the endocytic processing of lipoplexes turns out to be a relatively slow process, given that only after 4-5 h of significant accumulation of N-Rh-PE-labeled complex in late endosomal/lysosomal compartments can be detected, as demonstrated elsewhere (26, 37). In analogy with the size-dependent rate of internalization of virions (38), this slow internalization of lipoplexes could be explained by restraints of the lipoplexes to enter coated vesicles. Possibly, the rate-limiting steps include the rate and extent by and to which receptors nonspecifically associate with lipoplexes. Not unlikely, it could be speculated that a minimal degree of tightness with the coated region is required to effectively invaginate the relatively large complexes (i.e. compared with single ligands). In addition, it is possible that after subsequent binding to receptors, the entire complex has to be recruited into coated regions, such a lateral displacement or clustering representing another rate-limiting step in overall internalization. In this context it is interesting to note that clustering of polyplexes on the cell surface prior to internalization has been reported as a key step in the mechanism of polyplex transfection (39).

In late endosomal/lysosomal compartments, Tf is never detected, whereas lipoplexes were not detected in the Tf-containing recycling compartment. Indeed, because colocalization was only detectable early after internalization, we conclude that these data would thus be entirely consistent with a clathrin-mediated pathway of lipoplex internalization. To further strengthen this conclusion we next examined the effect on lipoplex-mediated transfection of potassium-depletion and overexpression of Eps15 constructs, both of which specifically inhibit clathrin-mediate endocytosis.

Lipoplex-mediated Transfection Relies on Complex Entry via Clathrin-mediated Endocytosis-- Potassium depletion is a well-established procedure to arrest coated pit formation (21). We therefore investigated its effect on lipoplex internalization. As can be seen in Fig. 6 a highly effective inhibition of lipoplex uptake in potassium-depleted cells compared with non-depleted cells could be detected. The percentage of inhibition of lipoplex internalization was 88.8 ± 4.0%, as determined by image analysis of the fluorescently labeled cells. Importantly, after quenching of extracellular fluorescence with trypan blue, no blue cell nuclei could be detected. Accordingly, potential toxic effects of K⁺ depletion, which could have caused a decreased internalization, are excluded.

View larger version (47K): [in this window] [in a new window]   Fig. 6.   Potassium depletion of COS7 cells results in inhibition of lipoplex internalization. After K+ depletion of COS7 cells and subsequent incubation with N-Rh-PE-labeled lipoplex for 1 h at 37 °C, the internalization of lipoplex is largely inhibited compared with the internalization in non-depleted cells (compare left and right images).

Finally we investigated the effect of overexpressing dominant negative mutants of Eps15, EH21-GFP and DIII-GFP, which have been reported to inhibit clathrin-mediated endocytosis (22, 23), on lipoplex internalization. Eps15 plays a role in the docking of the adaptor protein AP2 onto the plasma membrane. Interaction of AP2 with clathrin then results in the formation of coated pits at the plasma membrane. Expression of EH21-GFP and DIII-GFP, which encode the AP2 binding sites of Eps15 but not the domains for correct coated pit targeting of Eps15, was described to inhibit the endocytosis of transferrin, whereas the GFP fusion protein that lacks AP2 binding sites (D32-GFP) did not affect transferrin endocytosis. Similar effects were obtained for the internalization of lipoplex in cells expressing the Eps15 mutants. Expression of EH21-GFP and DIII-GFP in COS7 cells resulted in an inhibition of lipoplex uptake of 55.4% ± 5.2 and 69.5% ± 1.7, respectively. Note that cells expressing the control D32-GFP fusion protein show essentially as much internalized lipoplex as untransfected cells (Fig. 7B, compare arrow and arrowhead), whereas cells expressing the DIII-GFP fusion protein show a strongly reduced uptake of lipoplex compared with untransfected cells. (Fig. 7D, compare arrow and arrowhead).

View larger version (74K): [in this window] [in a new window]   Fig. 7.   Expression of the dominant negative Eps15 mutant DIII in COS7 cells inhibits lipoplex internalization. COS7 cells were transfected with GFP-tagged control D32 (A) and the mutant DIII (C) construct of Eps15. After 2 days, the cells were incubated with N-Rh-PE-labeled lipoplex for 1 h at 37 °C. The left and right images show the same field of cells. Expression of the DIII-GFP construct in COS7 cells results in an inhibition of lipoplex uptake (D) compared with cells expressing the control D32-GFP construct (B), as visualized by the changes in Rh fluorescence (B versus D). Thus, note that lipoplex internalization in D32-positive cells (A, GFP) is similar to that in non-transfected cells in the same population (arrow versus arrowhead, respectively, in B). By contrast, in the DIII mutants (C), the internalization of Rh-labeled complexes is clearly reduced compared with non-transfected cells in the same population (arrow versus arrowhead, respectively, in D).

Thus these data are entirely consistent with a requirement for complex internalization along the pathway of cholesterol-dependent clathrin-mediated endocytosis in the process of lipoplex-mediated transfection. To analyze whether difference between transfection efficiencies for the same lipoplexes may correlate with cell-specific differences in endocytic capacity, the following experiments were carried out.

The Endocytic Capacity of Cells for Lipoplexes May Codetermine Transfection Efficiency-- Although COS7 cells are easily transfectable with lipoplexes, A2780 cells show a diminished transfection efficiency compared with COS7 cells. To establish whether this difference can be attributed to a lower endocytic capacity of A2780 cells compared with COS7 cells, we determined the extent of lipoplex internalization after 1 and 4 h of incubation with lipoplexes and the ensuing transfection efficiency. As can be calculated from the data presented in Fig. 2, the endocytic capacity of COS7 cells, expressed as the percentage of internalized lipoplex relative to cell-associated lipoplex, is 20 ± 2%. As can be seen in Fig. 8, the endocytic capacity of A2780 cells is clearly less than that of COS7 cells (compare panels A and B). In addition, a longer incubation time with lipoplex (4 versus 1 h) results in a higher amount of internalized lipoplex for both cell lines. Subsequently, the transfection efficiencies obtained for COS7 and A2780 cells were determined by fluorescence-activated cell sorting measurement of GFP-positive cells and visualized by fluorescence microscopy (Fig. 9). Obviously, the higher uptake of lipoplex after 4 h of incubation compared with that after 1 h of incubation results in a higher transfection efficiency. For A2780 these values, determined after 48 h, were 13.1 ± 1.3% (1 h) and 27.6 ± 0.2% (4 h), and for COS7 cells 24.6 ± 0.1% (1 h) and 74.1 ± 3.0% (4 h). Moreover, the higher endocytic capacity of COS7 cells compared with A2780 cells results in a more efficient transfection. However, it should be noted that, although uptake of lipoplexes is a prerequisite for obtaining transfection, the extent of internalized lipoplexes need not necessarily correlate with transfection efficiency. After all, a crucial step in the overall transfection process is the endosomal release of lipoplex and/or DNA into the cytosol. Thus even if a substantial uptake of lipoplexes into cells occurs, failure of significant endosomal escape of the plasmid prior to arrival of lipoplexes into lysosomes, where degradation occurs, will preclude significant transfection (cf. Ref. 26). It is relevant, therefore, to take this consideration into account when comparing transfection efficiencies of different cell lines. Evidently, besides the efficiency of lipoplex internalization, the endosomal escape and processing of genetic cargo into the nucleus may differ between cells.

View larger version (67K): [in this window] [in a new window]   Fig. 8.   Lipoplex internalization in A2780 and COS7 cells after 1 and 4 h of incubation with lipoplexes. A2780 and COS7 cells were incubated with N-Rh-PE-labeled lipoplexes for 1 and 4 h. Subsequently, the cells were examined by fluorescence microscopy before and after incubation with trypan blue, to distinguish between total cell-associated and internalized lipoplex, respectively. Panel A shows fluorescent images of A2780 cells; in panel B fluorescent images of COS7 cells are depicted.

View larger version (145K): [in this window] [in a new window]   Fig. 9.   Transfection efficiencies of A2780 and COS7 cells after 1 and 4 h of incubation with lipoplexes. Cells were incubated with lipoplexes for 1 and 4 h and after 2 days the number of GFP-positive cells was visualized by fluorescence microscopy.

Next to an inability for clathrin-coated pits to detach from the cell surface, MCD treatment of cells may also result in the removal of caveolae from the cell surface, thereby inhibiting caveolae-mediated endocytosis. Although, to the best of our knowledge, Tf is not significantly internalized via caveolae, we nevertheless carried out the following experiments to investigate this possibility to estimate the potential contribution of caveolae-mediated internalization of lipoplexes.

Involvement of Caveolae-mediated Internalization Is Insignificant in Lipoplex-mediated Transfection-- To distinguish between the relative involvement of a clathrin-mediated, versus a caveolae-mediated, internalization pathway for lipoplexes as a critical step in the mechanism leading to transfection, we investigated the effects of two additional drugs, filipin III and chlorpromazine. Filipin III is a sterol-binding pentaene macrolide antibiotic, which is able to insert into lipid membranes that contain cholesterol. Filipin selectively inhibits caveolae invagination by the formation of cholesterol precipitates over these structures while leaving coated pits unaffected (40-43). The cationic amphiphilic drug chlorpromazine causes clathrin to localize and accumulate in late endosomes, thereby inhibiting coated pit endocytosis (44, 45). Preincubation of the cells with filipin III (FIII) and the presence of FIII during transfection caused an inhibition of approximately 20% in the efficiency of SAINT-2/DOPE-mediated transfection (Fig. 1b). On the other hand, incubation with chlorpromazine prior to and during transfection resulted in a drop in transfection efficiency of more than 50%. In a strict sense, these data could imply that a (minor) fraction of the lipoplexes could be internalized via caveolae. To obtain further insight, we therefore directly determined the effect of FIII treatment on lipoplex internalization per se. Thus using N-Rh-PE-labeled complexes and by determining the internalized pools following trypan-blue quenching in control and FIII-treated cells, we found that the extent of lipoplex internalization was inhibited by less than 10% (not shown). On the other hand, at similar conditions and following FIII treatment, the internalization of Bodipy-LacCer, a fluorescently tagged sphingolipid that is specifically internalized via caveolae following its insertion into the plasma membrane (27), was nearly completely inhibited (99.3% ± 0.1, not shown). In passing, elsewhere (26) we have shown that, eventually, lipoplexes accumulate into lysosomal compartments, as revealed by colocalization with the marker lysotracker red, the end point of an endocytic rather than a caveolae-mediated internalization pathway (33). Accordingly, taking these observations and considerations into account and in light of the much more pronounced inhibition seen upon chlorpromazine treatment, we thus conclude that clathrin-dependent endocytosis appears to represent the major entry pathway for lipoplex internalization by COS7 cells.

Summarizing, by using biochemical and genetic approaches to inhibit endocytosis, and by monitoring simultaneously clathrin-mediated endocytosis of Tf, this study demonstrates that clathrin-coated endocytosis represents a major pathway by which cationic amphiphile-DNA complexes acquire cellular entry, subsequently leading to productive cellular transfection. Thus, in a direct manner, it could be shown that the mere cell surface association of lipoplexes (at 37 °C) does not suffice to bring about significant cellular transfection. Additional work (not shown) with other cell lines, including HepG2 and Chinese hamster ovary cells, revealed that SAINT-2/DOPE-mediated delivery of plasmid DNA or oligonucleotides to these cells similarly resulted in an effective inhibition of transfection, following cholesterol depletion. These observations are particularly relevant in light of the fact that HepG2 cells lack caveolae (46), thus further supporting the notion that, rather than caveolae, clathrin-coated pits mediate the entry of lipoplexes into cells. These data thus imply that clathrin-dependent endocytosis is a critical step in the overall pathway of lipoplex-mediated transfection in vitro.

	ACKNOWLEDGEMENTS

Dr. Jan Willem Kok and Freark Dijk are acknowledged for assistance with confocal and electron microscopy, respectively, as is Dr. Sven van IJzendoorn for helpful discussion. Drs. Alexandre Benmerah and Alice Dautry-Varsat (Institut Pasteur, Paris, France) are thanked for generously providing the Eps15 constructs.

	FOOTNOTES

^* This work was supported by the Netherlands Foundation for Chemical Research (to C. W.)/Netherlands Technology Foundation (to S. T. W.) (349-40001).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed. Tel.: 31-050-363-8168; Fax: 31-50-363-2728; E-mail: d.hoekstra@med.rug.nl.

Published, JBC Papers in Press, March 1, 2002, DOI 10.1074/jbc.M111257200

	ABBREVIATIONS

The abbreviations used are: MCD, methyl--cyclodextrin; DOPE, dioleoylphosphatidylethanolamine; Eps15, epidermal growth factor receptor pathway substrate clone 15); HBSS, Hanks' balance salt solution; GFP, green fluorescence protein; CDCHOL, cyclodextrin-cholesterol complex; N-Rh-PE, N-lissamine rhodamine B phosphatidylethanolamine; Bodipy-LacCer, Bodipy-lactosylceramide; RT, room temperature; Tf, transferrin; FIII, filipin III; SAINT, synthetic amphiphile international.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				A. W. Wong, S. J. Scales, and D. E. Reilly DNA Internalized via Caveolae Requires Microtubule-dependent, Rab7-independent Transport to the Late Endocytic Pathway for Delivery to the Nucleus J. Biol. Chem., August 3, 2007; 282(31): 22953 - 22963. [Abstract] [Full Text] [PDF]

				I. A. Khalil, K. Kogure, H. Akita, and H. Harashima Uptake pathways and subsequent intracellular trafficking in nonviral gene delivery. Pharmacol. Rev., March 1, 2006; 58(1): 32 - 45. [Abstract] [Full Text] [PDF]

				I. A. Khalil, K. Kogure, S. Futaki, and H. Harashima High Density of Octaarginine Stimulates Macropinocytosis Leading to Efficient Intracellular Trafficking for Gene Expression J. Biol. Chem., February 10, 2006; 281(6): 3544 - 3551. [Abstract] [Full Text] [PDF]

S. P. Gordon, S. Berezhna, D. Scherfeld, N. Kahya, and P. Schwille Characterization of Interaction between Cationic Lipid-Oligonucleotide Complexes and Cellular Membrane Lipids Using Confocal Imaging and Fluorescence Correlation Spectroscopy Biophys. J., January 1, 2005; 88(1): 305 - 316. [Abstract]