| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
,,
* Department of Neurochemistry, Arrhenius Laboratories, Stockholm University, Stockholm, Sweden;
Institute of Molecular and Cell Biology, University of Tartu, Tartu, Estonia; and
Estonian Biocentre, Tartu, Estonia
1 Correspondence: Department of Neurochemistry, Stockholm University, S-106 91 Stockholm, Sweden. E-mail: caroline{at}neurochem.su.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: protein transduction domains • perforin • granzyme B • granulysin • cellular wound healing
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4)
. Resealing even occurs in liposomes as the unfavorable energetic state, caused by the exposure of hydrophobic backbone to the hydrophilic solute, results in an increase in free energy promoting resealing (3)
.
The debate on how cell-penetrating peptides (CPPs) or protein transduction domains (PTDs) translocate the cell membrane is still ongoing and highly active. Several studies have suggested macropinocytosis as the possible uptake mechanism (5
, 6)
. Evidence exists for other uptake pathways, however, such as clathrin-dependent endocytosis (7)
, or for the idea that different forms of endocytosis function simultaneously (8
, 9)
and by blocking 1 pathway, other pathways become more active. A study by Fretz et al. (10)
shows that the uptake is concentration dependent and that the peptides use 2 pathways; endocytosis at low concentrations and direct penetration at high concentrations. Moreover, Tünnemann et al. (11)
found the uptake to be divided into at least 2 functionally distinct uptake mechanisms, dependent on the size of cargo.
Previously, we showed that CPPs are rapidly degraded in both mammalian and nonmammalian cells (12
, 13)
. This parameter is important to take into account because the peptides outside the cell seem to be degraded before the uptake is equilibrated. Furthermore, our data indicated that the peptides translocate to different pools in the cell, 1 where the peptides remain intact and another where the peptides are degraded rapidly, suggesting that at least 2 uptake processes are involved. When CPPs bind the cell surface, there will most likely be an increased local concentration of peptides at the plasma membrane, which could induce endocytosis. Concurrently, the local increase in peptide concentration could cause a mass imbalance followed by a direct translocation (14)
or could disrupt the plasma membrane locally, thus inducing membrane resealing and activating the membrane repair response. Because the calcium influx triggers the resealing within seconds, this would explain why no leakage of cytoplasmic contents on CPP uptake is detected, even though there seems to be a membrane disturbance caused by the peptides. Herce et al. (15)
suggested that peptides translocate the membrane by inserting charged side chains into the plasma membrane, which induces the formation of a transient pore, and that the peptides diffuse onto the pore walls while carrying attached phospholipids with them during the translocation. The temporary nature of the transient pore would explain why no leakage is seen. We propose that the higher disturbances in plasma membrane caused by accumulation of peptides are masked and resealed by lysosomal patching induced by the calcium influx and MRR mechanism.
Recently, the MRR was shown to be involved in target cell apoptosis in which granules containing 3 proteins—perforin, granulysin, and granzymes—are exocytosed from natural killer cells and cytotoxic T cells (16)
. The delivery of granzymes to the cytoplasm was found to be necessary for the induction of apoptosis and depended on perforin. Although granzymes could be endocytosed on their own, perforin was needed for induction of apoptosis by granzymes. Two sequences responsible for energy-independent binding and cellular uptake of granzyme B were identified by Bird et al. (17)
. A reduction of cytotoxicity and uptake was found when mutating these sequences, and analogously with CPP uptake, heparin and cells deficient in cell surface sulfate or glycosaminoglycans showed decreased granzyme B uptake. Shorter sequences of the third protein involved in target cell apoptosis, namely granulysin, which is cytolytic against microbes and tumors but not red blood cells, were found to retain their lytic activity (18)
.
In this study, we have synthesized peptides from perforin (19)
, granulysin, and granzyme B proteins and compared them with 2 known CPPs, namely model amphipathic peptide (MAP) (20)
and penetratin (21)
, to investigate whether their internalization into cells induce a calcium influx, triggering the MRR and mobilization of cellular vesicles to the plasma membrane. We compared the effects of the novel CPPs with the full-length protein perforin and studied the MRR by detecting translocation of lysosomes to the plasma membrane using antibodies against the lysosomal-associated membrane protein CD107b (LAMP-2), exocytosis of the lysosomal enzyme β-hexosaminidase (22)
, and intracellular calcium measurements. Taken together, the data suggest that the MRR serves to mask the pore-forming effect of CPPs, giving the distinct possibility that uptake can indeed occur through common membrane disturbance.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Peptide synthesis
Peptides (Table 1
) were synthesized in a stepwise manner in a 1 mmol scale on a peptide synthesizer (model 431A; Applied Biosystems, Foster City, CA, USA) using t-Boc strategy of solid-phase peptide synthesis as described previously (13)
.
|
Cell culture
HeLa or CHO-K1 cells were cultured in Dulbecco modified Eagle medium (DMEM) or DMEM/F-12 with Glutamax-I (Gibco) supplemented with 10% FBS, 1% nonessential amino acids, 1% sodium pyruvate, 100 µg/ml streptomycin, and 100 U/ml penicillin.
FACS analysis
HeLa cells seeded in 6-well plates 1 day before experiment (250,000 cells/well) were washed with serum-free (SF) medium 2x and incubated with 1 ml peptide solution in SF medium (3 µM) for 1 h at 37°C. The cells were washed 2x with SF medium and incubated with 400 µl trypsin for 5 min at 37°C, followed by addition of 1 ml medium with serum and transferred to 1.5 ml Eppendorf tubes. The cells were centrifuged for 10 s, 10,000 rpm, and the supernatant was discarded. After the addition of 0.005% trypan blue solution in PBS, the cells were resuspended and kept on ice before analysis by FACS. Untreated cells were used as controls.
Electron microscopy
pGrB with a cysteine at N terminus instead of fluorescein was tagged with nanogold (Monomaleimido Nanogold, d 1.4 nm; Nanoprobes Inc., Yaphank, NY, USA) by forming a covalent bond between the thiol group of peptide and the maleimide group of label at a molar ratio 2:1 in 50% methanol. The sample was mixed at 300 rpm at 30°C for 60 min. Methanol was removed and conjugate concentrated by speed-vac to reach 50–100 µM concentration. Cells were incubated with 0.5 or 2.5 µM conjugate in culture medium at 37°C for 15 min or 1 h. The control cells were incubated with cationized nanogold (positively charged nanogold, 1.4 nm; Nanoprobes Inc.) under identical conditions. The specimens were prepared for electron microscopy as described previously (9)
. Ultrathin sections of cells embedded in epoxy resin were analyzed by transmission electron microscopy (JEM-100S; Jeol, Akishima, Japan).
Lactate dehydrogenase (LDH) leakage assay
The assay was performed using the Promega CytoTox-ONETM assay (Promega, Madison, WI, USA). HeLa or CHO-K1 cells were seeded in 96-well plates in 100 µl medium (20,000 cells/well) and used for experiments performed in triplicates one day after seeding. Prior to the incubation with 100 µl peptide at 37°C, cells were washed 2x with HEPES-Krebs-Ringer (HKR) buffer (125 mM NaCl, 5 mM KCl, 1.2 mM MgSO4·7H2O, 1 mM CaCl2·2H2O, 1.2 mM KH2PO4, 25 mM HEPES, 6 mM glucose, pH 7.4) supplemented with 1 g/L D-glucose. After 25 min incubation, samples of 80 µl were added to 80 µl of CytoTox-ONE reagent in a black 96-well-plate and incubated at 21°C for 10 min. Fluorescence was measured at 560/590 nm. Untreated cells were defined as no leakage, and total LDH release by lysing cells with 0.1% Triton X-100 was defined as 100% leakage.
Intracellular calcium measurements
HeLa or CHO-K1 cells were seeded in 96-well plates in 100 µl medium (20,000 cells/well) and used for experiments performed in triplicates one day after seeding. FURA-2 AM (2 µM) was added to the cell medium, and cells were incubated for 30 min at 37°C. The medium was replaced by HKR buffer, and cells were incubated for additional 15 min. The ratio of 340(Ca-bound FURA-2)/380(FURA-2) nm (Lm1/Lm2) excitatory wavelengths was measured at 510 nm in FlexStation II (Molecular Devices Corp., Sunnyvale, CA, USA) with 4 s interval 17 s before and 5 or 30 min after the addition of penetratin (1, 3, 5, 10, 20 µM), MAP (1, 3, 5, 10 µM), or pGrB (1, 3, 10, 20, 40 µM). Untreated cells and cells treated with 10 µM ionomycin were used as controls.
β-Hexosaminidase efflux assay
Measurements of β-hexosaminidase were made essentially as described by Howl et al. (22)
. Cells grown for 2 days in 24-well plates were washed 2x with HKR buffer and peptides at the indicated concentrations, in 200 µl of the same buffer were added. After incubation, 10 µl of cell-exposed buffer was transferred into 96-well plates and incubated with 50 µl 1 mM 4-methylumbelliferyl N-acetyl-β-D-glucoseamide in 0.1 M sodium citrate buffer, pH 4.5, for 1 h at 37°C. Subsequently 150 µl of K2CO3 buffer (pH 10.5) was added, and β-hexosaminidase activity was determined fluorometrically at 365/445 nm (em/ex).
Translocation of LAMP-2 to plasma membrane
HeLa cells were seeded in 24-well plates (50,000 cells/well) on glass coverslips 2 days before experiment. Cells were washed 2x with SF medium and incubated with peptides (0.5–20 µM) or perforin (0.1–0.8 µg/ml) in SF medium for 30 min at 37°C. Treated cells were washed 2x with PBS and fixed with 3% paraformaldehyde in phosphate buffer (0.1 M, pH 7.4) for 30 min and washed 2x with PBS. The control cells were permeabilized with methanol at –20°C for 15 min to expose all intracellular LAMP-2 antigens. The sites of unspecific binding were blocked with 10% nonfat dry milk in PBS for 1 h; LAMP-2 was stained with monoclonal antibody (H4B4, DSHB, 1:100) for 1 h; and Alexa Fluor 488 (or 555)-conjugated goat anti-mouse antibody (Invitrogen, 1:400) for 30 min at room temperature in dark. The solution of secondary antibody was supplemented with 1 µg/ml propidium iodide to assess the intactness of plasma membrane. The coverslips were washed, mounted on glass slides with Fluoromount G (Electron Microscopy Sciences, Hatfield, PA, USA), and analyzed with fluorescence microscopy. Images were recorded using an Olympus BX61 microscope equipped with CCD camera DP70 or Olympus FV1000 confocal microscope (Olympus, Tokyo, Japan). Obtained images were processed with Adobe Photoshop 7.0 (Adobe Systems, San Jose, CA, USA).
Analysis of CPP uptake by HPLC
HeLa or CHO-K1 cells seeded in 24-well plates in 500 µl medium (100,000 cells/well) were used for experiments performed in triplicates 1 day after seeding. Cells for uptake experiments were washed 2x with HKR buffer containing different calcium concentrations and incubated with 200 µl peptide (2 µM) at 37°C. After 1 h, the incubation buffer containing the peptides was removed. The cell lysate was analyzed with HPLC as described previously (13
, 20)
.
Statistical analysis
Data are expressed as means ± SE and were processed by SoftMax Pro 4.8 and GraphPad Prism 4. Statistical analysis was made by ANOVA followed by Dunett’s multiple comparison test (P<0.05).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(c–h) were taken from the full-length proteins and synthesized to investigate whether the peptides alone were able to induce calcium influx and trigger the MRR. It has previously been shown that shorter cationic sequences from perforin, granulysin, and granzyme B contribute to cytotoxicity and/or uptake of the respective proteins (17
, 19
, 23)
. Four potential sequences, perforin 82–98 (pPrF82), perforin 338–354 (pPrF338), granzyme B 89–104 (pGrB), and granulsyin 68–82 (pGrL) were synthesized, labeled with carboxyfluorescein at the N terminus, and assessed for their cell-penetrating properties and membrane effect. Two more peptides, perforin 1–34 (pPrF1a) and perforin 1–16 (pPrF1b), which previously were shown to have the same characteristics as the native protein, were synthesized (19)
. They were not taken up into cells, however, and no calcium influx could be detected (data not shown).
The uptake of the new peptides was studied by flow cytometry and compared with two known CPPs, penetratin and MAP (Fig. 1A
). Cells were incubated with 3 µM of each peptide for 1 h followed by trypsination. Trypan blue was added to the analysis buffer to quench any membrane bound peptide still in the membrane. All peptides were taken up, although the new peptides were not as efficient as penetratin or MAP. The uptake assay of peptides was also performed at 4°C, followed by trypsination and trypan blue quenching (data not shown). A lower amount of the peptides were still translocated. For trypsin to work, however, the trypsination was performed at 37°C for 5 min, which could be enough time for the peptides to translocate. The confocal laser scanning microscopy confirmed the internalization of the novel CPPs, which localized in HeLa cells in punctuate structures of diffusely stained cytoplasm (Fig. 1B-D
). An LDH leakage test was performed to confirm that the uptake did not cause any leakage of cytoplasmic compounds (data not shown).
|
To detail the distribution of peptides on the plasma membrane and inside the cell, cysteine rather than fluorescein was coupled to the N terminus of pGrB. The thiol group of cysteine was conjugated to a nanogold particle (1.4 nm) with an active maleimide group. Cells were incubated for 15 min (data not shown) or 1 h with the gold-tagged peptides and visualized by electron microscopy (Fig. 2
). The black dots, corresponding to pGrB, bound to the plasma membrane at specific loci at 0.5 µM, preferentially at protrusions but also at flat areas of the plasma membrane (Fig. 2A
). The peptide interacted with the membrane in a cooperative manner, forming spherical structures with the electron-dense background, which contained several peptide molecules (black dots in Fig. 2
). Both MAP and penetratin formed similar structures on the plasma membrane of HeLa cells, as visualized through the use of the respective conjugates with nanogold (data not shown). The peptide assemblies were able to insert into the plasma membrane and translocate to cortical cytoplasm but not deeper toward the cell center (Fig. 2B
). At 2.5 µM concentration, pGrB accumulated in the plasma membrane, and massive translocation to cortical cytoplasm could be seen (Fig. 2C
). Despite the accumulation of pGrB, however, the barrier function of the plasma membrane was not impaired; no propidium iodide influx was detected by fluorescence microscopy, and the intracellular ultrastructures remained completely normal and unchanged (data not shown). The peptide still seemed to interfere with the ordered packing of membranous phospholipids because the plasma membrane became less distinct. In addition, the peptide seemed to increase the fluidity of the plasma membrane, leading to vesicular uptake of pGrB (Fig. 2D
). Positively charged nanogold, in contrast, was very rarely detected on the plasma membrane or intracellular vesicles and only as single particles. It never assembled into specific structures containing many nanogold particles, confirming that the spherical structures detected were typical for pGrB only (data not shown).
|
To assess whether the peptides induced calcium influx, HeLa (Fig. 3A
) or CHO-K1 (Fig. 3B
) cells were preloaded with calcium indicator FURA-2 AM and then treated with peptides. Calcium influx was measured for 5 or 30 min before and after addition of peptides at different concentrations, with untreated cells and ionomycin as controls. For clarification, only 2 peptide concentrations are shown in Fig. 3
. Because the signal response was weak in CHO-K1 cells, only MAP, the most efficient peptide, was analyzed (Fig. 3B
). During the first 5 min, a transient and concentration-dependent influx of calcium was seen in HeLa cells with penetratin (Fig. 3C
) and pGrB (data not shown), although the calcium level did not return to the basal level. MAP also induced a transient influx of calcium during the first 5 min, but then the calcium level increased once more and was sustained throughout the measurement (Fig. 3D
). The influx was not as high as with the positive control ionomycin, however, even though LDH leakage can be seen already at 5 µM for MAP, whereas pGrB and penetratin show no leakage up to 100 µM (data not shown). The kinetics indicate that MAP-induced influx follows at least a second-order reaction with an EC50 value of 
1.9 µM in both HeLa and CHO-K1 cells because plotting the time to half maximum against the inverted concentration of peptides yielded a straight line for MAP. Unfortunately, it was not possible to deduce what type of process penetratin or pGrB followed.
|
Influx of calcium ions into cells induces several activities, including lysosomal exocytosis and plasma MRR. To quantify the lysosomal exocytosis caused by the CPPs, the influence of penetratin and MAP on the activity of the lysosomal protein β-hexosaminidase in cell supernatant was assessed. After 5 min, incubation with HeLa (Fig. 4A, C
) or CHO-K1 cells (Fig. 4B, D
), an increase in the enzyme activity could be seen already at 100 nM concentration of penetratin (Fig. 4A, B
) and MAP (Fig. 4C, D
). The β-hexosaminidase activity increased with increasing peptide concentration and, as expected, MAP was more potent in promoting lysosomal exocytosis. The measurements showed that the process is in fact complicated, however, and the result depends on several factors. The induced enzyme activity decreased with time; after 45 min incubation, the activity had declined and approached baseline (data not shown). Probably, CPP degradation time dependently decreased the membrane disturbance, and in addition, the fraction of lysosomes available for plasma membrane docking is not known.
|
To examine whether the extracellular calcium concentration affected the β-hexosaminidase outflow, cells were incubated in buffer containing peptides and 5 mM calcium (Fig. 4)
. Here, extracellular enzyme activity significantly increased after treatment with both peptides in CHO-K1 cells (Fig. 4B, D
) and penetratin in HeLa cells (Fig. 4A
), but no significant increase over the basal condition could be seen with MAP in HeLa cells (Fig. 4C
).
If the peptides cause a calcium influx sufficient to trigger the MRR, lysosomes should be translocated to the plasma membrane. The fusion of lysosomal membranes with the plasma membrane can be detected by the presence of lysosomal membrane protein LAMP-2 on the plasma membrane. Incubation of HeLa cells with recombinant perforin at low concentration did not trigger the plasma membrane repair (Fig. 5A
), whereas 0.4 µg/ml perforin induced exposure of LAMP-2 to the plasma membrane (Fig. 5B
). High concentration of perforin, on the contrary, led to only slight translocation of lysosomes toward cell periphery with less marked targeting of the plasma membrane but more staining of cell center (Fig. 5C
), as shown earlier by Keefe et al. (16)
. Despite the influx of calcium ions into cytosol and the MRR, the plasma membrane retained its barrier function; no uptake of propidium iodide by cells was detected at the used perforin concentrations (data not shown). Still, not all LAMP-2 antigens become accessible to antibodies at the maximal MRR because the removal of membrane lipids led to a significant increase in staining of cells (Fig. 5D
).
|
The CPPs exerted an analogous effect, inducing appearance of LAMP-2 in the plasma membrane without inducing propidium iodide uptake (data not shown). MAP, known to interfere with plasma membrane integrity at low concentrations (24)
, stimulated targeting of LAMP-2 to the plasma membrane in some cells even at 0.5 µM (Fig. 6A
). Raising the concentration of MAP to 1 µM led to plasma membrane accumulation of lysosomal protein in all cells (Fig. 6B
). Analogously to perforin, higher concentrations of MAP induced less translocation of LAMP-2 to the plasma membrane (Fig. 6C
) and more staining of the cell center.
|
Penetratin is considered to be of low cytotoxicity; therefore, we assessed its effect on LAMP-2 translocation. In line with MAP, penetratin induced appearance of lysosomal protein at the cell surface of cells, although at markedly higher concentrations (Fig. 6D-F
). LAMP-2 translocation of nonpermeabilized cells remained undetectable on the plasma membrane up to 3 µM and was negligible at 5 µM (Fig. 6D
). Starting from 10 µM concentration, however, induction of plasma membrane repair was clearly detectable (Fig. 6E, F
).
Because calcium influx was seen after treatment with peptides, it was important to investigate whether the concentration of calcium in the medium surrounding the cells affected the uptake. Cells were incubated with 2 µM fluorescently labeled peptides in buffer with 1.0, 2.5, 5.0 mM or no calcium for 1 h and analyzed in HPLC after treatment with diazotized 2-nitroaniline (Fig. 7
), as described previously (13)
. Penetratin and pGrB showed a significant increase in uptake when incubated in buffer without calcium, which was expected because there was no calcium to activate the MRR. In the presence of excess calcium, the uptake was reduced significantly in a concentration-dependent manner and was corroborated by flow cytometry (data not shown). Again, MAP, unlike the other peptides, was not significantly affected by the calcium concentration. In addition, the uptake of pPrF82 and pPrF338 was found to be calcium dependent (data not shown), concordant with penetratin and pGrB.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Lysosomes are believed to be the major contributor to plasma membrane resealing (25)
, and microinjected antibodies against LAMP-1 cytoplasmic domain have been shown to inhibit resealing and exocytosis of lysosomes because of aggregation (26)
. Thus, it is possible to measure the MRR by detecting intralumenal lysosomal proteins on the plasma membrane, such as LAMP-2 along with transient influx of calcium ions into cells, or by measuring the release of β-hexosaminidase into the extracellular medium caused by the exocytosis (22)
.
In target cell apoptosis, natural killer cells and cytotoxic T cells exocytose granules containing 3 proteins: perforin (19)
, granulysin (18)
, and granzyme B (17)
. Well inside the cytoplasm, apoptosis is induced by a group of serine proteases named granzymes or another protein named granulysin, which in addition to inducing apoptosis, has antimicrobial characteristics. Keefe et al. (16)
thoroughly investigated the actions of perforin and granzyme B during target cell apoptosis. It was found that perforin bound to the plasma membrane and formed pores at high concentrations, causing necrosis. At physiologically relevant concentrations, it triggered the MRR followed by granzyme B-induced apoptosis. Moreover, granzyme B was taken up by cells without perforin but could not induce apoptosis on its own. In summary, perforin activated the MRR and facilitated the delivery of granules in different cell lines and released granzyme B to the cytoplasm to induce apoptosis.
The uptake mechanism of CPPs strongly resembles the action of perforin, granzyme B, and granulysin in that they bind the plasma membrane or, more specifically, proteoglycans (6)
, which promotes actin rearrangement and uptake. Therefore, we wanted to compare the characteristics of perforin, granzyme B, and granulysin to two well-known CPPs: MAP, which is considered to be cytotoxic at low concentrations, and penetratin, which is not toxic at low concentrations. The first goal was to identify sequences with CPP-like motifs responsible for the uptake of each protein. Potential sequences (Table 1)
were analyzed for their cell-penetrating ability (Fig. 1)
and compared with penetratin and MAP. FACS data showed that the new peptides—pPrF82, pPrF338, pGrL68, and pGrB—were taken up into cells but not as efficiently as penetratin or MAP. Thus, these sequences might be responsible for the internalization of each protein.
The LDH leakage is an indication of membrane damages. Even though the LDH molecule is large, the test correlates with leakage of even smaller molecules, such as deoxyglucose (27
, 28)
. None of the peptides showed any LDH leakage even though both granulysin and perforin originally are lytic proteins (data not shown). The exception was MAP, which started to show leakage at 5 µM (data not shown) but still showed calcium influx (Fig. 3)
and LAMP-2 translocation (Fig. 6)
at lower concentrations. Electron micrographs of HeLa cells incubated with pGrB showed peptides bound to the plasma membrane but no induction of vesicular uptake at low concentrations (Fig. 2)
. CPPs seem to bind outgrowths on the plasma membrane as well as locally distorting the plasma membrane, as reported previously (29)
. Vesicular-bound peptides can be seen in cells at the higher concentration only, although most of them are still at the plasma membrane. It is rather difficult to unambiguously distinguish peptides bound to the plasma membrane from internalized peptides because the plasma membrane becomes less distinct on interaction. The changes in membrane morphology together with the nonexisting leakage of endogenous molecules suggest that the MRR compensates the disturbances caused by the peptides with exocytosis of lysosomes. The only leakage possible seems to be an ion leakage, not out of the cell but into the cytoplasm. The specific structures of pGrB detected in electron micrographs are rather similar to temporary pores modeled for Tat peptide (15)
and transient pore-like structures induced in model membranes (30
, 31)
, corroborating the possibility for limited influx of calcium ions into cytoplasm. As reported previously for perforin, which induced a transient calcium influx (16)
, all peptides tested induced a concentration-dependent calcium influx that lasted for 5 min. The exception was MAP, which after the first 5 min, showed an irreversible increase of cytosolic calcium (Fig. 3D
), although not as high as the positive control ionomycin. The calcium influx was not only concentration dependent but also followed CPP efficiency, meaning that MAP, being more efficient, induced calcium influx at lower concentrations than penetratin, and penetratin at lower concentrations than pGrB, although still in a concentration-dependent manner. It should be noted, however, that MAP is toxic and starts to induce leakage at 5 µM. Nevertheless, calcium influx and LAMP-2 translocation already occured at lower concentrations of MAP, and both pGrB and penetratin showed no toxicity over 100 µM. Plotting the time to half maximum against the inverted concentration of peptides yielded a straight line for MAP (r2=0.96), whereas for penetratin, neither a straight line with the natural logarithm (r2=0.89) nor the inverted concentration (r2=0.84) of the peptide was significant. Thus, the inflow of calcium could be second order with respect to MAP but less straightforward for penetratin or pGrB.
Quantification of the release of β-hexosaminidase, a sensitive indication of lysosomal exocytosis, showed significant increase of extracellular activity at peptide concentrations that are much lower than those needed for LDH detection (Fig. 4)
. β-Hexosaminidase release at low peptide concentration probably occurs through kiss-and-run mechanisms, whereas higher concentrations or longer time is necessary for membrane fusion that would result in LAMP-2 exposure in plasma membrane. Analogously with other forms of exocytosis, lysosomes are subject to kiss and run, which on stronger stimulation leads to fusion of vesicles with plasma membrane (32
, 33)
. Plotting the increase in intracellular calcium levels caused by the peptides (in percentage of maximum) against β-hexosaminidase activity in the supernatant (in percentage of basal) yielded a straight line (with a slope of 0.42 and x intercept –42.35 and y intercept 100, r2=0.99). This result showed that these two parameters are well correlated and could indicate that the increased intracellular calcium concentration caused by the peptides is responsible for the stimulated β-hexosamindase release.
To visualize that the MRR triggered by peptides actually occurs and induces mobilization of lysosomes to the plasma membrane as the native protein perforin (Fig. 5)
, antibodies against an intracellular lysosomal protein, LAMP-2, were used to detect resealing. Both MAP and penetratin induced the membrane repair mechanism in the same manner as the recombinant protein perforin (Figs. 5
and 6
). Microscopy photos illustrate the biphasic dependence between the peptide uptake and translocation of lysosomes to the plasma membrane, and by increasing the concentration, the translocation first increased and later decreased.
Because the peptides induced an influx of calcium ions into the cytoplasm, triggering the MRR, it was important to know whether the uptake depended on the calcium concentration in the extracellular environment, as previously shown (34)
. If the peptides trigger the resealing response, the uptake should most probably increase if no calcium is present outside the cells because the response cannot be activated. Moreover, if the calcium concentration is increased, it might lead to a decrease in CPP uptake because the plasma membrane is repaired much more efficiently. Our speculation was supported by the increase in peptide uptake when no calcium in the extracellular buffer was present and a concentration-dependent decrease in uptake with increased calcium concentrations in the extracellular buffer (Fig. 7)
. Furthermore, FACS data showed a decrease in CPP uptake in the medium with elevated calcium concentration (data not shown). Nevertheless, it should be mentioned that the decrease in uptake of CPPs could also result from the increment of positive calcium ions in the buffer competing with peptides for electrostatic interactions with the proteoglycans or membrane stability correlated with less LAMP-2 externalization. It has also been shown that extracellular calcium in itself can increase the stability of membranes (35)
.
Studies have indicated endocytosis together with direct penetration as uptake mechanisms for CPPs. The direct penetration mechanism has remained elusive, however, and been thought not to involve membrane damage because no indication of membrane disruption has been seen at relevant concentrations of peptide. We present evidence that can explain why membrane disturbances associated with direct penetration are difficult to visualize: The MRR, mobilizing vesicles within seconds to patch any broken membrane, could be the answer.
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 2, 2008. Accepted for publication August 14, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-helical amphipathic model peptide with the potential to deliver polar compounds into the cell interior non-endocytically. Biochim. Biophys. Acta 1414,127-139[Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), MAP (
), or pGrB (
) at different concentrations for 5 min. The calcium influx is presented as percentage of ionomycin. C, D) HeLa cells were treated with penetratin (C) and MAP (D) at different concentrations for 30 min. For clarification, only 2 peptide concentrations are shown and the calcium influx is presented as percentage of untreated cells during the first 5 min. Symbols correspond to ionomycin (
), buffer (
), 5 or 3 µM (
), 20 or 5 µM (
) penetratin and MAP, respectively. Calcium influx was measured by FlexStation II with untreated cells or ionomycin as controls. Data are representative of at least 3 independent experiments.
