Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Aug 1.
Published in final edited form as: Biochim Biophys Acta. 2011 Jun 6;1810(8):752–758. doi: 10.1016/j.bbagen.2011.05.013

A HA2-Fusion Tag Limits the Endosomal Release of its Protein Cargo Despite Causing Endosomal Lysis †

Ya-Jung Lee 1, Gregory Johnson 1, Grantham C Peltier 1, Jean-Philippe Pellois 1,*
PMCID: PMC3128663  NIHMSID: NIHMS303236  PMID: 21664431

Abstract

BACKGROUND

Protein transduction domains (PTDs) can be fused to a protein to render it cell-permeable. The delivery efficiencies of PTDs are, however, often poor because PTD-protein conjugates cannot escape from endosomes. A potential solution this problem consists in adding HA2 analogs to the PTD-protein construct as these peptides can cause endosomal lysis upon acidification of the endosomal lumen. To date, however, the utility of HA2-based PTDs has not been clearly established.

METHODS

We investigate the biophysical and cellular properties of the glutamate-rich HA2 analog E5 fused to the model protein TAT-mCherry.

RESULTS

E5-TAT-mCherry causes the release of fluorescent dextrans trapped with the protein inside endosomes. Yet, E5-TAT-mCherry itself is not released in the cytosol of cells, indicating that the protein remained trapped inside endosomes even after endosomal lysis takes place. Cytosolic delivery of the protein could be achieved, however, by insertion of a disulfide bond between E5 and its cargo.

CONCLUSIONS

These results show that E5 causes the retention of its fused protein inside endosomes even after lysis takes place.

GENERAL SIGNIFICANCE

These data establish that HA2 analogs might not be useful PTDs unless cleavable linkers are engineered between PTD and protein cargo.

Keywords: protein delivery, HA2 peptide, endosomal escape, protein transduction domain, lysis

1. Introduction

The delivery of proteins into cells is potentially of great utility in medicine, cell biology and biotechnology. A method to achieve protein delivery into cells consists in attaching a so-called protein transduction domain (PTD) to a protein cargo. This is an attractive approach because PTDs are peptides that can be genetically encoded and a cell-permeable PTD-protein fusion can be produced recombinantly without the need of further chemical modifications. PTDs typically consist of short peptide sequences that have the ability to transport macromolecular cargo into cells. A prototypical PTD is TAT (GRKKRRQRRR), a peptide originally identified from the HIV Trans-Activator of Transcription protein.[1, 2] Although several mechanisms are involved in TAT-mediated transduction, endocytosis is a predominant pathway for the cellular uptake of macromolecules conjugated to TAT. [3-5] Following internalization, TAT-protein conjugates appear able to escape from the lumen of endosomes into the cell’s cytosolic space. However, this second step in the delivery process is not efficient. As a result, a large fraction of the internalized bioactive macromolecules remains trapped inside endocytic organelles and little or no biological effect is achieved.[4, 6] Peptides that can lyse lipid bilayers at acidic pH have been proposed as a solution to this problem.[4, 7] This strategy is based on the idea that the lytic activity of these reagents can be triggered inside endosomes as acidification of the endosomal lumen occurs.[8] In principle, these peptides can therefore cause the lysis of endosomes without causing the disruption of other cellular membranes. An example of an endosomolytic peptide is the HA2 fusion peptide, a peptide derived from the N-terminus of the HA2 subunit of influenza hemagglutinin (HA).[9-11] HA2 and HA2 mutants lyse membranes in a pH-dependent manner.[10, 11] These peptides insert into lipid bilayers upon protonation of their glutamate residues and form pores into membranes.[12, 13] HA2 and HA2 analogues have been used to enhance the endosomal release and delivery of DNA and RNA particles as well as proteins encapsulated in liposomes.[14-17] In addition, the chimeric peptide HA2-TAT has been shown to improve the delivery of proteins when added to the incubation media of live cells.[4, 18] These reports highlight the usefulness of these reagents as additives that improve the delivery of proteins into cultured cells. Yet, when considering the delivery of proteins for in vivo therapeutic applications, one expects that HA2 and protein of interest would have to be conjugated to one another for successful delivery. The value of HA2 and HA2 analogs as protein delivery fusion tags remains, however, unclear. Addition of HA2 to the N-terminus of a polyarginine-p53 construct has for instance been shown to increase the antitumor potency of the protein.[19] However, the hydrophobicity of HA2 is often problematic and it is therefore unclear how this peptide might affect the biophysical and cellular properties of a protein fused to it. In addition, HA2 has a strong affinity for lipid bilayers and the question of whether a HA2-protein conjugate can be released from the lipid bilayer of endosomes after lysis takes place has not been answered. To test the hypothesis that HA2-based fusion tags increase the delivery of a protein into the cytosolic space of live cells, we investigate the biophysical and cellular properties of the model protein E5-TAT-mCherry. E5 is a HA2 analog known to be more soluble and more endosomolytic than wild-type HA2.[13-15, 18, 20, 21] E5 (GLFEAIAEFIENGWEGLIEGWYG) consists of the wild-type HA2 sequence (residues 1-23, strain X31) modified with the mutations M17L, D19E, G4E and G8E. M17L and D19E are introduced to avoid methionine oxidation and aspartyl isomerization while G4E and G8E are introduced to increase the pH-responsiveness and lysis activity of the peptide.[15, 20, 21] E5 was introduced at the N-terminus of the protein because HA2 peptides typically need to have a free N-terminal glycine to display a lytic activity.[10] TAT was included in the constructs to promote the endocytosis of mCherry into live cells.[22] Finally, mCherry was chosen as a model protein cargo because this protein is monomeric, soluble, non-cytotoxic, and without a cellular activity. Our rationale was therefore that the effects of the E5 fusion tag would be easy to identify since the protein cargo itself should not interfere with our assays. mCherry is also fluorescent and the trafficking of this protein inside live cells can be monitored by fluorescence microscopy. Finally, mCherry remains folded and fluorescent at acidic pH (pKa <4.5).[23] The lytic activity of E5-TAT-mCherry at acidic pH, in vitro or inside endocytic organelles, can therefore be measured without having denaturation of the protein affect the results of these experiments.

2. Material and methods

2.1 Cloning and expression of HA2-TAT-mCherry and E5-TAT-mCherry

pTXB1-SUMO-E5-TAT-mCherry was obtained by mutagenesis of previously described pTXB1-SUMO-HA2-TAT-mCherry using the QuikChange site-directed mutagenesis kit and the oligonucleotides 5′-GGT GGT GGT CTG TTC GAA GCT ATC GCT GAA TTC ATC GAA AAC GGT TGG-3′ and 5′-CCA CCA CCA GAC AAG CTT CGA TAG CGA CTT AAG TAG-3′.[24] The clone pTXB1-mCherry-Cys was obtained by mutagenesis of the previously reported pTXB1-mCherry using the QuikChange site-directed mutagenesis kit and the oligonucleotides 5′-CGG CAA AAA GTG CAA AGT TGG CTG C-3′ and 5′-GCA GCC AAC TTT GCA CTT TTT CTT GCC G-3′.[6] The plasmids were transformed into BL21-DE3 cells and protein expression was induced with 0.5 mM IPTG at 37°C for 3 hr. Cells were harvested and resuspended in lysis buffer containing 20 mM Tris-Cl (pH 7.5) and 200 mM NaCl. After cell lysis by sonication and high-speed centrifugation at 15000 rpm for 30 min, the soluble fraction was applied to chitin resin pre-equilibrated with lysis buffer and incubated overnight at 4°C (the proteins contain a C-terminal intein-chitin binding domain purification tag). The resin was washed with 8 column volumes of lysis buffer. The proteins were cleaved from the resin by incubation of the beads with 1 column volume of buffer supplemented with 100 mM DTT for 24 hr at room temperature. This cleavage step yields a cleaved protein with a C-terminal carboxylate. The SUMO tag was removed with the SUMO protease with the molar ratio of 1: 25 (SUMO protease: protein). It is important to note that the SUMO tag was introduced in our cloning strategy to obtain an N-terminal glycine residue of the E5 sequence after SUMO cleavage (a possible N-terminal methionine by expression of E5-TAT-mCherry without an N-terminal tag). The proteins were purified by cation-exchange chromatography in 50 mM HEPES at pH 7 (HiTrap SP HP, GE Healthcare). The final products are E5-TAT-mCherry and mCherry-Cys. Protein sequence analysis by automated Edman chemistry was performed to further confirm that the N-terminus of E5 was properly generated. The N-terminal sequence was identified as GLFEAIAEFI with 99% confidence.

2.2. Synthesis of E5-TAT-S-S-mCherry

The peptide E5-TAT-Cys (GLF EAI AEF IEN GWE GLI EGW YGC G) was purchased as a crude cleavage product from RayBiotech (Norcross, GA) and purified by reverse phase HPLC. The correct identity of the peptide was confirmed by MALDI-TOF mass spectrometry (AXIMA-CFR, Shimazu, Kyoto); expected mass: 4138.8 Da, observed mass: 4142.8 Da. E5-TAT-Cys (2 mM) was added to mCherry-Cys (500 μM) in a pH 7.5 buffer of 25 mM HEPES and 100 mM NaCl. The solution was exposed to air and gently mixed overnight. The reaction progress was monitored by SDS-PAGE using a loading buffer lacking reducing agents. The product E5-TAT-S-S-mCherry was purified from mCherry-Cys and its dimerized form (mCherry-Cys)2 by cation-exchange chromatography with 50 mM HEPES at pH 7 (HiTrap SP HP, GE Healthcare). Pure E5-TAT-S-S-mCherry was analyzed by SDS-PAGE (Supplemental Information Fig. S1)

2.3. Membrane binding Assays

The phospholipids 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), 1,2-dipalmitoyl-sn-glycero-3-phospho-L-serine (DPPS), sphingomyelin (brain, porcine)(SM), and cholesterol were purchased from Avanti Lipids. Neutral multi lamellar vesicles (MLVs) were prepared with a molar ratio of 50:30:20 (DPPC:Chol:SM) by addition of the lipids in chloroform to a glass vial. For negatively charged MLVs, stock DPPS was added to the mixture for a total of 9 mol% DPPS in the final mixture. Chloroform was evaporated under a nitrogen stream and the vials were placed in a vacuum desiccator overnight. HEPES buffer (50mM, pH 7.0) was added to the dried lipids to give a total lipid concentration of 1mM. The vials were capped again after filling with nitrogen and allowed to incubate in a shallow 42°C water bath for 20min. Each vial was then subjected to ten freeze/thaw cycles using liquid nitrogen and a 42°C water bath, after which the vials were cooled to room temperature immediately used in the binding experiments. E5-TAT-mCherry or TAT-mCherry were incubated with MLVs in either acetic acid (50 mM pH 5.0) or HEPES buffer (50 mM, pH 7.0) to a final protein concentration of 1 μM in 200 μL.[23] The concentration of lipid in the outer bilayer of MLVs available for binding was estimated to be 50 μM according to reported protocols.[25] Samples were incubated for 20 min and centrifuged for 2 min at 14,000 rpm. The supernatant was saved while the pellet from this sample was then resuspended in 200 μl of fresh buffer. It is important to note that when E5-TAT-mCherry is incubated in the absence of MLVs, no pellet forms and the protein concentration in solution does not change after centrifugation (the protein therefore does not precipitate at either pH 5.0 or 7.0 in the absence of MLVs; data not shown). These steps were repeated to obtain pellets that had been washed one or two times along with the corresponding supernatants. Each sample was placed in a well of a multi-well plate and imaged using a Typhoon Trio Variable Mode Imager (GE Life Sciences). A green excitation laser at 532 nm and an emission filter at 610nm were used to detect the fluorescence of mCherry. The Image Quant software was used for the densitometric analysis of the fluorescence signal present in each well.

2.4. Microscopy Assays

HeLa cells were seeded on 8-well chamber glass slide at 3.0 × 104 cells/well in Dulbecco’s modified Eagle’s medium supplemented with 10% FBS and incubated for 24 h at 37°C in a humidified atmosphere containing 5% CO2. Cells were then incubated with L-15 medium (no FBS added) and placed on an inverted epifluorescence microscope (Model IX81, Olympus, Center Valley, PA) equipped with a heating stage maintained at 37°C. Images were collected using a Rolera-MGI Plus back-illuminated EMCCD camera (Qimaging, Surrey, BC, Canada). Images were acquired using bright field imaging and three standard fluorescence filter sets: CFP (Ex = 436±20 nm / Em= 480±40 nm), Texas Red (Ex = 560±40 nm / Em= 630±75 nm), and FITC (Ex = 482±35 nm / Em= 536±40 nm). Cells were treated with E5-TAT-mCherry (5 μM) in L-15 medium containing 2.5 mg/mL of 70 kDa or 10 kDa FITC-dextran. After incubation at 37°C for 15 or 60 minutes, the cells were washed 5 times with PBS containing heparin (1 mg/ml) and the medium was replaced with fresh L-15. Heparin, a negatively charged and soluble glycosaminoglycan known to inhibit the binding of TAT to membrane bound heparan sulfate proteoglycans, was used to wash the proteins associated with the plasma membrane.[26, 27] The integrity of the plasma membrane of the cells was determined by addition of the cell-impermeable DNA stain SYTOX® Blue. Cells with a blue fluorescent nucleus were considered compromised or dead.). To inhibit acidification of the endolysosomal organelles, cells were first pretreated with bafilomycin (200 nM, Sigma, MO) for 30 minutes, and then with E5-TAT-mCherry and 70 kDa FITC-dextran while keeping bafilomycin present. Cells containing a cytosolic distribution of FITC-dextran were counted manually in images obtained using a 10 X objective. Cells containing a diffuse distribution of green fluorescence with an intensity of at least 3 times greater than that of background signal (signal outside of cells of interest) were counted as positive (cells positive for SYTOX® Blue were not included). The total number of cells in each image was determined using the corresponding bright field image. For each samples, 10 images were acquired in different areas of the dish, assuring that a minimum of 1000 cells were examined. Each experiment was reproduced at least 3 times. The values reported represent the average percentages of cells containing a diffuse distribution of FITC-dextran and the corresponding standard deviations.

3. Results

3.1. E5-TAT-mCherry lyses membranes in a pH-dependent manner

The lytic activity of E5-TAT-mCherry was characterized by an in vitro erythrocyte lysis assay (Fig. 1). We reported that E5-TAT lyse red blood cells in a pH-dependent manner.[13] We were therefore interested in determining whether addition of mCherry at the C-terminus of this peptide would affect its hemolytic activity. E5-TAT-mCherry was incubated with human RBCs in PBS with a pH adjusted to different values. The release of hemoglobin from lysed red cells was then measured spectrophotometrically at 450 nm. The control proteins mCherry and TAT-mCherry did not cause any hemolysis at any of the conditions tested (up to 20 μM, pH 4.0-7.5). In contrast, E5-TAT-mCherry caused hemolysis below pH 6.0. However, the hemolytic activity of E5-TAT-mCherry was reduced in comparison to E5-TAT (Fig. 1B). Together, these results therefore suggest that addition of a protein at the C-terminus of E5 can significantly reduce the lytic activity of the peptides. The molecular basis for this effect is unclear. Nonetheless, these experiments indicate that E5-TAT-mCherry can potentially lyse acidified endosomes.

Figure 1.

Figure 1

Open in a new tab

E5-TAT-mCherry lyses membranes in a pH-dependent manner. A) Hemolytic activities of TAT-mCherry and E5-TAT-mCherry as a function of pH. B) Hemolytic activities of E5-TAT and E5-TAT-mCherry at pH 7.0 and 4.0 as a function of peptide or protein concentration and for 1.25% RBC suspensions.

3.2. E5-TAT-mCherry associates with membranes in a pH-dependent manner

E5 is designed to bind to the lipid bilayer of endosomes to cause endosomal lysis. We were therefore interested in investigating the membrane-binding properties of E5-TAT-mCherry. To address this issue, the association of E5-TAT-mCherry with model membranes was characterized. Multi lamellar vesicles (MLVs) containing phosphatidyl cholines (PC), choline sphingomyelins (SM), phosphatidylserines (PS), and cholesterol were prepared as models for the endosomal lipid bilayer. Because PS is thought to be restricted to cytosolic leaflet of endosomes, MLVs lacking PS (−PS) were prepared as models of the luminal leaflet of endosomes (the leaflet E5-TAT-mCherry might interact with prior to lysis). Yet, lysis can cause the loss of this lipid asymmetry and one can hypothesize that E5-TAT-mCherry could be exposed to the PS during during or after lysis. MLVs containing PS (+PS) were therefore used as a model for lysed endosomal lipid bilayers. E5-TAT-mCherry was incubated with the different MLVs for 10 min at pH 7.0 (non-lytic pH) or at pH 5.0 (lytic pH). TAT-mCherry was used as a control to determine the effect of E5. Each sample was centrifuged at high-speed to obtain a pellet that contains the protein associated with MLVs and a supernatant that contain the soluble protein. The fluorescence of the pellets and supernatants were then analyzed as a measure of the protein content of the membrane-bound and soluble protein fractions, respectively. As shown in Figure 2 A, the binding of E5-TAT-mCherry to membranes is greater at pH 5.0 than at pH 7.0. This effect is not observed for TAT-mCherry, confirming that E5 is responsible for this pH dependence. The binding of E5-TAT-mCherry to bilayers containing PS is also greater at pH 5.0 than the binding of E5-TAT-mCherry to bilayers lacking it. Interestingly, this effect is also observed for TAT-mCherry at both pH 7.0 and pH 5.0, suggesting that the TAT moiety of E5-TAT-mCherry participates in the protein’s binding to PS containing membranes. This is expected since the positively charged TAT has been shown to bind electrostically with negatively charged lipid such as PS.[28, 29] Together, these data suggest that E5-TAT-mCherry binds to lipid bilayers and that this binding is dependent on both pH and nature of the lipids present.

Figure 2.

Figure 2

Open in a new tab

E5-TAT-mCherry binds to membranes in a pH-dependent manner. A) Binding of E5-TAT-mCherry and TAT-mCherry to neutral (−PS) or negatively charged (+PS) MLVs as a function of pH. Binding was assessed by separating the soluble protein fraction (supernatant, S) from the membrane-bound protein fraction (pellet, P) by centrifugation and by measuring the relative protein fluorescence intensity of each fraction. The average of three experiments and the corresponding standard deviations are represented. The protein to lipid ratio was kept constant in all experiments. B) Effect of pH on the dissociation of E5-TAT-mCherry from a membrane. E5-TAT-mCherry was bound to negatively charged MLVs at pH 5.0. The pellet containing E5-TAT-mCherry bound to MLVs and obtained after centrifugation was then washed in fresh buffer at either pH 5.0 or 7.0. The protein released into solution was separated from the protein remaining bound to the MLVs by a second centrifugation step (wash 1). These steps were then repeated (wash 2). The supernatants (S) and pellets (P) obtained after each step were placed in a multi-well plate and imaged on a fluorescence scanner. The fluorescence image obtained is represented as an inverted monochrome.

Based on the previous results, one can extrapolate that the protein should bind to the lipid bilayers of endosomes as the endosomal lumen acidifies. The data, however, do not reveal whether this binding would be reversed once endosomes lyse. One can envision that, upon lysis, the membrane-bound protein is exposed to the cytosol. The protein could therefore dissociate from the membrane because of a dilution effect. In addition, the protein could remain exposed to an acidic pH of the endosome or become exposed to the neutral pH of the cytosol. To test these different scenarios, the pellets obtained after binding of E5-TAT-mCherry to MLVs at pH 5.0 were resuspended in fresh buffer at either pH 5.0 or pH 7.0. The samples were centrifuged again to separate the proteins remaining bound to MLVs from the proteins released in solution. As shown in Figure 2 B, E5-TAT-mCherry did not dissociate from MLVs when the MLVs were washed at pH 5.0. On the other hand, E5-TAT-mCherry could be partially released in solution when the washed were performed at pH 7.0. These results therefore suggest that the protein can dissociate from a membrane upon dilution. Yet, the equilibrium between membrane-bound and soluble states is strongly shifted toward the membrane-bound state when the pH is acidic.

3.3. Characterization of the endosomolytic and cytosolic release activities of E5-TAT-mCherry

The cytosolic delivery of E5-TAT-mCherry was examined by incubating the protein with HeLa cells. Based on the cytotoxicity data and on the principle that protein delivery should ideally be achieved without causing significant levels of cell death, a protein concentration of 5 μM was chosen for the incubation (Supplemental Information Fig. S2). The protein TAT-mCherry was used for comparison and to establish the effects of E5. Fluorescently labeled dextrans (10 and 70 kDa FITC-dextran) were also added during incubation as markers of pinocytosis. After 1 hour incubation, TAT-mCherry and E5-TAT-mCherry co-localized with FITC-dextrans in a punctate distribution consistent with these molecules being localized within endocytic organelles. The proteins did not appear to diffuse into the cytosolic space of the cells examined in any of the samples tested. However, in approximately 5% of the cells incubated with E5-TAT-mCherry, the fluorescent dextrans had a cytosolic distribution as well as a punctate distribution (Fig. 3 and 4). These cells were not stained by SYTOX Blue, indicating that the permeability of the plasma membrane of these cells was not compromised post-incubation. The cytosolic distribution of dextrans was not observed when E5-TAT-mCherry was not added to the incubation media or when TAT-mCherry was used. These results therefore indicate that the E5 moiety in E5-TAT-mCherry is involved in mediating the transport of the dextrans into the cytosol of cells.

Figure 3.

Figure 3

Open in a new tab

Fluorescence microscopy of HeLa cells incubated with TAT-mCherry (5 μM), E5-TAT-mCherry (5 μM), 10 kDa FITC-Dextran (FITC-Dx), E5-TAT, and bafilomycin. Images were obtained with a 10 X objective. The fluorescence signals are represented in inverted monochrome images or in an overlay image (mCherry is pseudo-colored red and FITC is pseudo-colored green). The fluorescence signal of E5-TAT-mCherry associated with cells corresponds to a punctate endocytic distribution not discernable at 10 X but discernable at 100 X (see Fig. 4). The fluorescence signal of FITC-Dx observed is that of molecules delivered into the cytosolic space of cells (the fluorescence signal of FITC is quenched inside acidic endocytic organelles and, at 10 X, this signal is not detectable under the imaging conditions used). Typical cells with a cytosolic FITC-Dx signal are highlighted with black arrows and their percentage in each sample is reported. The cells represented were not stained by SYTOX Blue, indicating that the plasma membrane of these cells is not compromised (not shown).

Figure 4.

Figure 4

Open in a new tab

Fluorescence microscopy of HeLa cells incubated with E5-TAT-mCherry, TAT-mCherry, mCherry, and E5-TAT-S-S-mCherry. Images were obtained with a 100 X objective. A 70 kDa FITC-Dextran (FITC-Dx) was used during incubation to detect the endosomal release of endocytosed molecules (as detected by its diffuse cytoplasmic distribution accompanied by nuclear exclusion). Endosomal release of TAT-mCherry and mCherry was induced by addition of the endosomolytic peptide E5-TAT in the incubation media. E5-TAT-mCherry shows a punctate distribution whether E5-TAT is present or not (cells incubated with E5-TAT-mCherry but without E5-TAT are represented here). The cells were not stained by SYTOX Blue, indicating that their plasma membrane is not compromised (not shown).

An interesting observation made during the microscopy analysis was that E5-TAT-mCherry appeared to remain associated with endocytic organelles even in cells in which FITC-dextran had redistributed in the cytosolic space (Fig. 4 and Supporting Information Fig. S3). The radius of gyration of 70 kDa FITC-dextran is approximately 18 nm while the radius of the 32 kDa E5-TAT-mCherry is expected to be less than 10 nm.[30] If the larger FITC-dextran can escape while E5-TAT-mCherry cannot, it is therefore not the size of the protein that limits its ability to reach the cytosolic space. E5-TAT-mCherry showed a diffuse fluorescence distribution when directly microinjected into the cytoplasm of HeLa, indicating that the protein could in principle diffuse into the cytosol after escape from endosomes (Supplemental Information Fig. S4). To investigate the endosomal retention phenomenon further, we first attempted to increase the delivery of E5-TAT-mCherry by increasing the concentration of the protein. However, as previously mentioned, the toxicity of the protein above 5 μM led to extensive cell-death. As a result, much brightly fluorescent cellular debris was present in the samples, making the imaging impractical and inconclusive. To get around this problem, E5-TAT was added to the incubation mixture. We showed that E5-TAT could improve the cytosolic delivery of proteins and dextrans in a simple co-incubation protocol.[13, 18] Our goal was therefore to determine whether E5-TAT-mCherry could be released into the cytosol of cells if endosomes were rendered more leaky by E5-TAT. Consistent with our previous report, addition of E5-TAT to the E5-TAT-mCherry and FITC-dextran incubation mixture greatly increased the population of cells containing FITC-dextran in their cytosolic space (approximately 60% of cells, Fig. 3). Addition of the vacuolar H+-ATPase inhibitor bafilomycin inhibited the endosomal release of FITC-dextran. This indicates that acidification of the lumen of endosomes is required for the cytosolic delivery of the dextran and that E5-TAT is mediating this process by causing endosomal release as opposed to plasma membrane permeabilization. Despite this observed increase in the number of cells containing FITC-dextran in their cytosolic space, addition of E5-TAT did not increase the cytosolic delivery of the protein. In contrast, addition of E5-TAT led to the cytosolic delivery of mCherry and TAT-mCherry (Fig. 4). Together these results indicate that E5-TAT-mCherry fails to reach the cytosolic space because of the E5 moiety. These results also suggest that E5-TAT-mCherry does not reach the cytosol because it remains associated with endosomes. This happens even when endosomal lysis takes place (as shown by the release of FITC-dextran) and whether endosomal lysis is mediated by E5-TAT-mCherry itself or by the addition of E5-TAT.

3.4. Introduction of a disulfide bond between E5-TAT and mCherry allows release of the protein from endosomes into the cytosolic space

Since E5-TAT-mCherry did not appear to escape from endosomes because of E5, we tested whether the introduction of a disulfide bond between E5-TAT and mCherry could solve this problem. We showed that a disulfide bond between TAT and mCherry is not reduced within the endocytic compartments of HeLa cells under the conditions of our experiments.[6] Nonetheless, endosomes rendered leaky by E5-TAT might become permeable to cytosolic reducing agents such as glutathione and that the disulfide bond between E5-TAT and mCherry might be reduced (if a 70 kDa FITC-dextran can exit endosomes, glutathione should be able to enter these organelles). If true, mCherry might then be released from E5-TAT and diffuse into the cytosolic space of cells. To test this idea, the delivery of E5-TAT-S-S-mCherry, a construct in which E5-TAT is linked to mCherry through a single disulfide bond, was examined. When E5-TAT-S-S-mCherry and FITC-dextran were incubated with HeLa cells, cells containing a cytosolic distribution of FITC-dextran also contained a cytosolic distribution of the protein (Fig. 4). The cytoplasmic and nuclear distribution of the protein also corresponded to that of microinjected mCherry but not of that of E5-TAT-mCherry, indicating that the disulfide bond between E5-TAT and mCherry was cleaved in these cells. Unlike E5-TAT-mCherry, the mCherry moiety of E5-TAT-S-S-mCherry is therefore able to diffuse away from endosomes in cells in which endosomal lysis can be observed (e.i. FITC-dextran escape). Overall, these data further confirm that E5-TAT-mCherry remains associated to lysed endosomes because of a retention mechanism that can be attributed to the E5 moiety.

4. Discussion

Proteins fused to PTDs, while efficiently internalized by endocytic mechanisms, often remain trapped inside endocytic organelles during delivery. Addition of an endosomolytic peptide such as E5 to a PTD might provide a solution to this problem. Our data indeed show that E5-TAT-mCherry lyses endosomes in cellulo, as observed by the release of endocytosed dextrans into the cytosol of cells. However, our assays reveal that E5 does not increase the cytosolic delivery of its protein cargo. Instead, the protein remains associated with endocytic organelles even after endosomal lysis takes place. A potential explanation for this phenomenon might be that the endosomolytic activity of E5-TAT-mCherry was not high enough to achieve sufficient cytosolic release of the protein. Yet, addition of E5-TAT during incubation greatly increased the release of FITC-dextran from endosomes but did not increase the release of E5-TAT-mCherry into the cytosol. In contrast, mCherry and TAT-mCherry were delivered in the cytosol of cells under the same conditions. Together, these results establish that E5 is responsible for the endosomal retention observed. Moreover, these data also suggest that enhancing the lytic activity of the fusion tag would not be sufficient to achieve cytosolic delivery.

One could envision that the endosomal retention observed for E5-TAT-mCherry is due to the strong affinity the E5 tag displays for lipid bilayers. Based on the data obtained with MLVs, E5-TAT-mCherry is tightly associated with lipid bilayers at pH 5.0. However, when returning the pH to 7.0, the protein dissociates from the membrane and it is partially released in solution. Together, these results suggest that the pH of an acidic endosome does not return to a neutral pH when lysis is mediated by E5-TAT-mCherry. The protein therefore remains bound to the endosomal membrane under these conditions. While this hypothesis remains to be tested, our results indicate a possible solution to this problem. The data obtained with E5-TAT-S-S-mCherry suggest that introducing a cleavable linker between E5 and its protein cargo can lead to successful cytosolic delivery. Yet, such construct is not genetically encodable and the formation of disulfide bonds between proteins and a delivery peptide is not always possible or convenient. E5-TAT fused to a genetically encodable linker that can be cleaved inside cells would instead be optimal for the design of future cell-permeable proteins. Because the hemolytic activity of E5-TAT is greater than that of E5-TAT fused to mCherry, an ideal linker would also be one that is cleaved inside endosomes. Linker cleavage would therefore induce an increase in the endosomolytic activity of E5-TAT but also release an unperturbed protein that would be able to escape into the cytosol. The design of such systems will be the object of further studies.

Supplementary Material

01

*Research Highlights.

> We investigate the cytosolic delivery of the protein mCherry attached to the endosomolytic peptide E5. > E5 causes the lysis of endosomes upon acidification of the endosomal lumen. > mCherry fused to E5 cannot escape from lysed endosomes. > mCherry attached to E5 through a disulfide bond can escape from lysed endosomes. > E5 does not increase endosomal escape of a protein cargo unless a cleavable linker is present.

Acknowledgments

We thank Professor Donald Pettigrew for his help and valuable discussions. Protein sequencing was performed at the Laboratory for Protein Chemistry of Texas A&M University and we thank Dr Lawrence Dangott for his help. This work was supported by Awards Number R01GM087227 and R01GM087981 from the National Institute of General Medical Sciences and the Norman Ackerman Advanced Research Program.

Abbreviations

PTD

protein transduction domain

DPPC

1,2-dipalmitoyl-sn-glycero-3-phosphocholine

PC

phosphatidyl cholines

DPPS

1,2-dipalmitoyl-sn-glycero-3-phospho-L-serine

PS

phosphatidylserines

SM

sphingomyelin

MLV

multi lamellar vesicle

Dx

dextran

FITC

fluorescein isothiocyanate

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • [1].Wadia JS, Dowdy SF. Protein transduction technology. Curr. Opin. Biotechnol. 2002;13:52–56. doi: 10.1016/s0958-1669(02)00284-7. [DOI] [PubMed] [Google Scholar]
  • [2].Zorko M, Langel U. Cell-penetrating peptides: mechanism and kinetics of cargo delivery. Adv Drug Deliv Rev. 2005;57:529–545. doi: 10.1016/j.addr.2004.10.010. [DOI] [PubMed] [Google Scholar]
  • [3].Amand HL, Fant K, Norden B, Esbjorner EK. Stimulated endocytosis in penetratin uptake: effect of arginine and lysine. Biochem Biophys Res Commun. 2008;371:621–625. doi: 10.1016/j.bbrc.2008.04.039. [DOI] [PubMed] [Google Scholar]
  • [4].Wadia JS, Stan RV, Dowdy SF. Transducible TAT-HA fusogenic peptide enhances escape of TAT-fusion proteins after lipid raft macropinocytosis. Nat. Med. 2004;10:310–315. doi: 10.1038/nm996. [DOI] [PubMed] [Google Scholar]
  • [5].Kaplan IM, Wadia JS, Dowdy SF. Cationic TAT peptide transduction domain enters cells by macropinocytosis. J Control Release. 2005;102:247–253. doi: 10.1016/j.jconrel.2004.10.018. [DOI] [PubMed] [Google Scholar]
  • [6].Lee YJ, Datta S, Pellois JP. Real-time fluorescence detection of protein transduction into live cells. J Am Chem Soc. 2008;130:2398–2399. doi: 10.1021/ja7102026. [DOI] [PubMed] [Google Scholar]
  • [7].El-Sayed A, Futaki S, Harashima H. Delivery of macromolecules using arginine-rich cell-penetrating peptides: ways to overcome endosomal entrapment. AAPS J. 2009;11:13–22. doi: 10.1208/s12248-008-9071-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Forgac M. Vacuolar ATPases: rotary proton pumps in physiology and pathophysiology. Nat Rev Mol Cell Biol. 2007;8:917–929. doi: 10.1038/nrm2272. [DOI] [PubMed] [Google Scholar]
  • [9].Smith AE, Helenius A. How viruses enter animal cells. Science. 2004;304:237–242. doi: 10.1126/science.1094823. [DOI] [PubMed] [Google Scholar]
  • [10].Wharton SA, Martin SR, Ruigrok RW, Skehel JJ, Wiley DC. Membrane fusion by peptide analogues of influenza virus haemagglutinin. J Gen Virol. 1988;69(Pt 8):1847–1857. doi: 10.1099/0022-1317-69-8-1847. [DOI] [PubMed] [Google Scholar]
  • [11].Tamm LK, Han X. Viral fusion peptides: a tool set to disrupt and connect biological membranes. Biosci Rep. 2000;20:501–518. doi: 10.1023/a:1010406920417. [DOI] [PubMed] [Google Scholar]
  • [12].Zhelev DV, Stoicheva N, Scherrer P, Needham D. Interaction of synthetic HA2 influenza fusion peptide analog with model membranes. Biophys J. 2001;81:285–304. doi: 10.1016/S0006-3495(01)75699-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Lee YJ, Johnson G, Pellois JP. Modeling of the endosomolytic activity of HA2-TAT peptides with red blood cells and ghosts. Biochemistry. 2010;49:7854–7866. doi: 10.1021/bi1008408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [14].Wagner E, Plank C, Zatloukal K, Cotten M, Birnstiel ML. Influenza virus hemagglutinin HA-2 N-terminal fusogenic peptides augment gene transfer by transferrin-polylysine-DNA complexes: toward a synthetic virus-like gene-transfer vehicle. Proc. Natl. Acad. Sci. U.S.A. 1992;89:7934–7938. doi: 10.1073/pnas.89.17.7934. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Plank C, Oberhauser B, Mechtler K, Koch C, Wagner E. The influence of endosome-disruptive peptides on gene transfer using synthetic virus-like gene transfer systems. J. Biol. Chem. 1994;269:12918–12924. [PubMed] [Google Scholar]
  • [16].Mastrobattista E, Koning GA, van Bloois L, Filipe AC, Jiskoot W, Storm G. Functional characterization of an endosome-disruptive peptide and its application in cytosolic delivery of immunoliposome-entrapped proteins. J Biol Chem. 2002;277:27135–27143. doi: 10.1074/jbc.M200429200. [DOI] [PubMed] [Google Scholar]
  • [17].Lundberg P, El-Andaloussi S, Sutlu T, Johansson H, Langel U. Delivery of short interfering RNA using endosomolytic cell-penetrating peptides. FASEB J. 2007;21:2664–2671. doi: 10.1096/fj.06-6502com. [DOI] [PubMed] [Google Scholar]
  • [18].Lee YJ, Erazo-Oliveras A, Pellois JP. Delivery of macromolecules into live cells by simple co-incubation with a peptide. Chembiochem. 2010;11:325–330. doi: 10.1002/cbic.200900527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Michiue H, Tomizawa K, Wei FY, Matsushita M, Lu YF, Ichikawa T, Tamiya T, Date I, Matsui H. The NH2 terminus of influenza virus hemagglutinin-2 subunit peptides enhances the antitumor potency of polyarginine-mediated p53 protein transduction. J. Biol. Chem. 2005;280:8285–8289. doi: 10.1074/jbc.M412430200. [DOI] [PubMed] [Google Scholar]
  • [20].Murata M, Takahashi S, Kagiwada S, Suzuki A, Ohnishi S. pH-dependent membrane fusion and vesiculation of phospholipid large unilamellar vesicles induced by amphiphilic anionic and cationic peptides. Biochemistry. 1992;31:1986–1992. doi: 10.1021/bi00122a013. [DOI] [PubMed] [Google Scholar]
  • [21].Esbjörner EK, Oglecka K, Lincoln P, Gräslund A, Nordén B. Membrane Binding of pH-Sensitive Influenza Fusion Peptides. Positioning, Configuration, and Induced Leakage in a Lipid Vesicle Model. Biochemistry. 2007 doi: 10.1021/bi701075y. [DOI] [PubMed] [Google Scholar]
  • [22].Murata M, Kagiwada S, Hishida R, Ishiguro R, Ohnishi S, Takahashi S. Modification of the N-terminus of membrane fusion-active peptides blocks the fusion activity. Biochem Biophys Res Commun. 1991;179:1050–1055. doi: 10.1016/0006-291x(91)91925-3. [DOI] [PubMed] [Google Scholar]
  • [23].Shaner NC, Campbell RE, Steinbach PA, Giepmans BN, Palmer AE, Tsien RY. Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat. Biotechnol. 2004;22:1567–1572. doi: 10.1038/nbt1037. [DOI] [PubMed] [Google Scholar]
  • [24].Lee YJ, Erazo-Oliveras A, Pellois JP. Delivery of Macromolecules into Live Cells by Simple Co-Incubation with a Peptide. Chembiochem. 2009 doi: 10.1002/cbic.200900527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [25].Yang ST, Zaitseva E, Chernomordik LV, Melikov K. Cell-penetrating peptide induces leaky fusion of liposomes containing late endosome-specific anionic lipid. Biophys J. 2010;99:2525–2533. doi: 10.1016/j.bpj.2010.08.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [26].Hakansson S, Jacobs A, Caffrey M. Heparin binding by the HIV-1 tat protein transduction domain. Protein Sci. 2001;10:2138–2139. doi: 10.1110/ps.23401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [27].Rusnati M, Coltrini D, Oreste P, Zoppetti G, Albini A, Noonan D, d’Adda di Fagagna F, Giacca M, Presta M. Interaction of HIV-1 Tat protein with heparin. Role of the backbone structure, sulfation, and size. J Biol Chem. 1997;272:11313–11320. doi: 10.1074/jbc.272.17.11313. [DOI] [PubMed] [Google Scholar]
  • [28].Thoren PE, Persson D, Esbjorner EK, Goksor M, Lincoln P, Norden B. Membrane binding and translocation of cell-penetrating peptides. Biochemistry. 2004;43:3471–3489. doi: 10.1021/bi0360049. [DOI] [PubMed] [Google Scholar]
  • [29].Tiriveedhi V, Butko P. A fluorescence spectroscopy study on the interactions of the TAT-PTD peptide with model lipid membranes. Biochemistry. 2007;46:3888–3895. doi: 10.1021/bi602527t. [DOI] [PubMed] [Google Scholar]
  • [30].Lenart P, Rabut G, Daigle N, Hand AR, Terasaki M, Ellenberg J. Nuclear envelope breakdown in starfish oocytes proceeds by partial NPC disassembly followed by a rapidly spreading fenestration of nuclear membranes. J Cell Biol. 2003;160:1055–1068. doi: 10.1083/jcb.200211076. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS303236-supplement-01.pdf (2.2MB, pdf)

RESOURCES