Conceived and designed the experiments: EES JPP. Performed the experiments: DS NM GAJ AEO JL. Analyzed the data: DS NM JPP. Contributed reagents/materials/analysis tools: DS NM GAJ AEO JL. Wrote the manuscript: DS JPP.
The authors have declared that no competing interests exist.
Cell-penetrating peptides (CPPs) can transport macromolecular cargos into live cells. However, the cellular delivery efficiency of these reagents is often suboptimal because CPP-cargo conjugates typically remain trapped inside endosomes. Interestingly, irradiation of fluorescently labeled CPPs with light increases the release of the peptide and its cargos into the cytosol. However, the mechanism of this phenomenon is not clear. Here we investigate the molecular basis of the photo-induced endosomolytic activity of the prototypical CPPs TAT labeled to the fluorophore 5(6)-carboxytetramethylrhodamine (TMR).
We report that TMR-TAT acts as a photosensitizer that can destroy membranes. TMR-TAT escapes from endosomes after exposure to moderate light doses. However, this is also accompanied by loss of plasma membrane integrity, membrane blebbing, and cell-death. In addition, the peptide causes the destruction of cells when applied extracellularly and also triggers the photohemolysis of red blood cells. These photolytic and photocytotoxic effects were inhibited by hydrophobic singlet oxygen quenchers but not by hydrophilic quenchers.
Together, these results suggest that TAT can convert an innocuous fluorophore such as TMR into a potent photolytic agent. This effect involves the targeting of the fluorophore to cellular membranes and the production of singlet oxygen within the hydrophobic environment of the membranes. Our findings may be relevant for the design of reagents with photo-induced endosomolytic activity. The photocytotoxicity exhibited by TMR-TAT also suggests that CPP-chromophore conjugates could aid the development of novel Photodynamic Therapy agents.
Cell-penetrating peptides (CPPs) are used to deliver molecular cargos into live cells
Here we examined the photolytic activity of TAT labeled with the fluorophore 5(6)-carboxytetramethylrhodamine (TMR). We show that TMR-TAT causes rapid and efficient endosomal membrane disruption in the presence of light. Remarkably, however, this is accompanied by the rapid destruction of cells. These effects involve the targeting of TMR by TAT to cellular membranes and the formation of singlet oxygen within membranes. Interestingly, TMR itself has a low singlet oxygen quantum yield and displays no photolytic activity and no photocytotoxicity. Therefore, our data indicate that TAT can convert an innocuous chromophore into a potent photolytic agent.
To test the photoendosomolytic activity of CPP-fluorophore conjugates, four peptides conjugated to the fluorophore 5(6)-carboxytetramethylrhodamine (TMR) were examined: the CPP TAT (GRKKRRQRRR), retro-inverso TAT (riTAT, rrrqrrkkrgy), the CPP R9 (RRRRRRRRR), and K9 (KKKKKKKKK). TMR was chosen as a model fluorophore because it is not toxic to cells in both the absence and presence of light. It is also commonly used for live cell imaging and it is synthetically readily accessible. Finally, the fluorescence of TMR is not as pH-dependent as that of fluorescein. Therefore, TMR is not affected by the acidic pH of endosomes to the extent that fluorescein is. Peptides were synthesized by solid-phase peptide synthesis using standard F-moc chemistry and purified by HPLC. In TMR-TAT and TMR-R9, the amino terminus of the peptide was directly coupled to TMR. riTAT, a peptide in which the TAT sequence is reversed and the constituent amino acids have a D rather than a L-configuration, was prepared to obtain an analogue of TAT that would be resistant to degradation by cellular proteases. TMR was introduced at the N-terminus of the peptide. TMR-K9 is comprised of 9 positively charged lysine residues and serves as a control. This peptide is endocytosed by cells and therefore it is localized inside endosomes like TAT or R9. However, K9 does not have the cell penetration activity attributed to TAT and R9.
TMR-TAT (TMR-GRKKRRQRRRG-NH2) expected mass: 1865.0 Da, observed mass: 1866.1 Da. TMR-R9 (TMR-RRRRRRRRRG-NH2) expected mass: 1893.2 Da, observed mass: 1894.4 Da. TMR-riTAT (TMR-rrrqrrkkrgy-OH) expected mass: 1973.3 Da, observed mass: 1975.3 Da. TMR-K9 (TMR-KKKKKKKKK-NH2) expected mass: 1583.0 Da, observed mass: 1583.3 Da.
HeLa (human cervical adenocarcinoma), COS-7 (SV40 transformed African green monkey kidney fibroblast-like cell line) were obtained from ATCC. COLO 316 (human ovarian carcinoma) were obtained from Robert Burghardt (Department of Veterinary Integrative Biosciences, Texas A&M University)
Whole blood was purchased from Gulf Coast Regional Blood Center (Houston, TX). Erythrocytes were centrifuged for 5 min at 1500 g and the erythrocyte pellet was resuspended with PBS. This was repeated three times to remove plasma and buffy coat. The erythrocytes (50% volume in PBS) were diluted in PBS to a final concentration of 0.25%. Indicated concentration of peptide was added to the medium and the samples were added to the wells of an 8-well chamber glass slide (Nunc). Cells were typically allowed to settle to the bottom of the dish for 5 minutes prior to imaging so as to obtain a layer of cells in the focal plane.
Cells were placed on an inverted epifluorescence microscope (Model IX81, Olympus, Center Valley, PA) equipped with a heating stage maintained at 37°C. The microscope is configured with a spinning disk unit to perform both confocal and wide-field fluorescence microscopy. Images were collected using a Rolera-MGI Plus back-illuminated EMCCD camera (Qimaging, Surrey, BC, Canada). Images were acquired using bright field imaging and two standard fluorescence filter sets: Texas Red (Ex = 560±20 nm/Em = 630±35 nm), and CFP (Ex = 436±10 nm/Em = 480±20 nm). The excitation light was from a 100 W halogen lamp (Olympus USH 1030C) passed through the filter cubes and 40 or 100× objective lenses. Neutral density filters (ND 1, 2, 3, or 4 on the instrument, corresponding to 100, 25, 12.5, or 5% transmittance) and different exposure times were used to modulate the amount of light samples were exposed to. The bright field and fluorescence intensities of cells and ghosts were measured using the SlideBook 4.2 software (Olympus, Center Valley, PA). To determine the percentage of cells stained by SYTOX® Blue, cells were imaged with a 20× objective by phase contrast. Ten to twenty images were acquired for each experiment. The total number of cells in a given image was determined from the phase contrast image while the number of dead cells was determined by identifying cells containing a blue fluorescent nucleus stained by SYTOX® Blue. Cell viability was determined by establishing a ratio of dead cells/total number of cells for each sample (at least 1000 cells were counted in each experiments and each experiments were repeated 3 times).
Irradiances at the specimen were 100 mW/cm2 when no neutral density filter and no objective were present in the light path (and, for instance 5 mW/cm2 when ND4 was inserted). Irradiances were measured using a monochromic photometer (model 840-c, Newport, Irvine, CA). Irradiation area has a diameter of 1.3 cm without objective but the light beam is focused into an area of 3×10−3 cm2 by the 100× objective. Irradiances can therefore be approximated to be at 21 or 420 W/cm2 with ND4 or ND1, respectively. Irradiances provided in the figure legends are based on these calculations. To confirm that the irradiances measured on the microscope were accurate, experiments were also reproduced, when possible, using light illumination with a collimated light source from Oriel (Stratford, CT) equipped with a 500 W Hg lamp. Selective irradiation at 560 nm was performed using an analytical line filter (Oriel, 9.4 nm bandwidth). In this case, irradiances of the light beam (diameter of 2.5 cm) could be measured more precisely using the monochromic photometer.
The effect of light irradiation on cells that have endocytosed TMR-TAT was first investigated. HeLa cells were incubated in the presence of TMR-TAT (3 µM) for 1 hour at 37°C. After washing the cells, the sample was placed on a microscope and the cellular distribution of the peptide was assessed by fluorescence imaging. The peptide initially showed a punctate intracellular distribution consistent with their accumulation within endocytic organelles (TAT typically requires higher concentrations than the one used here to penetrate the cytosolic space of cells directly) (
A) Light irradiation causes escape of TMR-TAT from endocytic organelles into the cytosol. Hela cells were incubated with TMR-TAT (3 µM), washed, and irradiated at 560±20 nm through a 100× objective on a wide-field microscope. Images were acquired in a time-lapse experiment with a light excitation of 300 ms and an interval of 2 seconds. The time displayed on the images represents the total light exposure time. The TMR fluorescence signal is represented as inverted monochrome (black = fluorescent signal, white = no signal). The TMR signal, initially in a punctate distribution, can be seen to diffuse away from individual endocytic organelles upon irradiation (black arrows). The perimeter of the nucleus is highlighted by a dashed line in the last image and the signal from TMR-TAT presumably accumulated at nucleoli is indicated with white arrows. B) Photosensitization of TMR-TAT or TMR-R9 endocytosed by cells causes plasma membrane damage and cell death. In contrast, cells containing TMR-K9 remain viable and the punctate distribution of TMR-K9 is not affected by the light irradiation under the conditions tested. The fluorescence signals of TMR and SYTOX® Blue are represented as inverted monochrome or pseudo-colored red and green in the overlay image, respectively.
A) Irradiation of TMR-TAT endocytosed by cells causes plasma membrane blebbing and permeabilization (SYTOX® Blue staining) within 20 seconds after 2 sec irradiation at 560 nm(the radiant exposure is approximately 40 J/cm2). Cell destruction is observed in more than 95% of the cells irradiated at 560 nm. In contrast, no photocytoxicity is observed when cells treated with TMR-K9 (10 µM) or TMR (3 µM) and TAT (3 µM) are irradiated under similar conditions. Cells incubated with TMR-TAT but irradiated at 430 nm are not destroyed (excitation wavelength of SYTOX® Blue but not of TMR, radiant exposure was also approximately 40 J/cm2). The histogram represents the average percentage of cells stained by SYTOX® Blue after peptide and light treatment (the number of cells examined is at least 3000) and the error bars represent the standard deviation (experiments were reproduced at least 3 times). B) The cytotoxic effect of TMR-TAT is limited to only irradiated areas. COLO 316 cells were incubated with TMR-TAT (3 µM) for 1 h. The cells were washed with fresh media and endocytosis of TMR-TAT was confirmed by fluorescence microscopy. The cells within the circled area were exposed to light at 560 nm. Morphological changes and SYTOX® Blue staining (inverted monochrome or pseudo-colored cyan) are only observable in the irradiated area.
TMR-R9 and TMR-riTAT had similar photoendosomolytic and photocytotoxic activities as TMR-TAT, indicating that an arginine-rich composition might be sufficient for these activities (
The previous experiments suggest that TMR and TAT have to be conjugated to one another to achieve the light-induced endosomal release and the observed cell-death (
HeLa cells were incubated with TMR-TAT or TMR-riTAT (3 µM) for 1 h, washed, and incubated at 37°C for an additional 1, 4 or 8 hours. Irradiation of cells incubated with TMR-riTAT led to the cytosolic distribution of the TMR fluorescence signal, plasma membrane blebbing (as seen in the bright field image) and permeabilization (represented in histogram) under all tested conditions. In contrast, these photo-induced effects are dramatically reduced for TMR-TAT as the time between peptide incubation and irradiation is increased. The histogram represents the average percentage of cells stained by SYTOX® Blue after peptide and light treatment (the number of cells examined is at least 3000) and the error bars represent the standard deviation (experiments were reproduced at least 3 times).
The disruption of lysosomes by photosensitizers or lysosomotropic compounds has been shown to induce cell-death
A) HeLa cells were incubated with TMR-TAT (3 µM) at 4°C to block endocytosis of the peptide. Images represent the bright field image and the fluorescence image of SYTOX® Blue (inverted monochrome). B) The effects of photosensitization of hematoporphyrin (HP) on cell morphology are similar to those obtained by photosensitization of TMR-TAT. HP was incubated with HeLa cells for 1 hour, washed with fresh L-15 media, and irradiated at 560 nm. Black arrows in A) and B) highlight the membrane blebbing observed during light exposure.
To further confirm that the fluorescent peptides are acting on membranes when irradiated, the effect of TMR-TAT, TMR-R9, and TMR-riTAT on the membrane of red blood cells (RBCs) was examined. RBCs do not typically endocytose extracellular material to significant levels
A) RBCs incubated with TMR-TAT (2 µM) have either a concave or crenated morphology initially. Irradiation of the sample at 560 nm causes formation of spherical cell ghosts (highlighted with white arrows in inserts) that lose their contrast in bright field images as the cells lyse. B) Percentage of lysed RBCs as a function of the compounds present in the media. TMR and TMR containing peptides were used at 2 µM. All experiments were performed by irradiating the samples at 560 nm for 400 msec (∼8 J/cm2). The data in the histogram represents the average of 4 experiments and the error bars correspond to the standard deviation. No lysis was observed for any of the samples in the absence of light.
To determine whether oxygen is involved in the light-induced lysis of RBCs mediated by TMR-TAT, oxygen was removed from the cell culture by degassing the incubation media with nitrogen. Under these conditions, the photohemolysis of RBCs exposed to TMR-TAT was greatly reduced (
TMR-TAT (3 µM) was incubated with crocetin (50 µM) and HeLa cells for 1 hour. Cells were washed with fresh L-15, placed on the microscope, and irradiated at 560 nm for the times indicated. The left panel represents cells imaged with the 100× and illustrates that endosomal release of TMR-TAT appears reduced by crocetin but not abolished. However, endosomal release is not accompanied by membrane blebbing or SYTOX® Blue staining (no signal detected, not shown) as seen when crocetin is omitted (right panel, cells are imaged with 20× objective).
Upon light irradiation TMR-TAT rapidly and efficiently disrupts the membrane of endosomes. However, in our hands, this activity is also accompanied by the simultaneous destruction of cells. The peptide therefore played the role of a photosensitizer, a molecule able to kill cells upon light irradiation. While multiple cellular events might take place during the photosensitization of the compound, the damage to cellular membranes other than that of endosomes appears to lead to this photocytotoxic effect. The peptide was for instance able to lyse the plasma membrane of HeLa cells as well as the membrane of red blood cells upon light irradiation. The photocytotoxicity of the peptide also involves the formation of singlet oxygen upon light irradiation. Singlet oxygen is short-lived (lifetime <4 µs) and diffuses across only very short distances (singlet oxygen would travel a distance of ∼220 nm in water during three lifetimes and presumably much less when reacting with biomolecules inside a cell)
The peptide had many desirable properties: photo-induced endosomal release required only short illumination with light of moderate intensities (as opposed to long irradiation time with intense lasers), this phenomenon happened in all irradiated cells, and a large fraction of the material endocytosed was released into the cytosolic space. On the other hand, in our assays, photo-induced endosomal release could not be achieved without cell-death occurring either simultaneously or with a short delay. The photocytoxicity associated with TMR-TAT constitutes a significant problem when one considers using this peptide in the context of photo-induced delivery of macromolecules into live cells. Addition of crocetin to the media might help solve this problem as crocetin inhibited the photocytotoxicity of TMR-TAT. However, the efficiency of endosomal release was also reduced by this treatment. In order to maximize the potential of this delivery approach, a future challenge will be to identify efficient photo-endosomolytic peptides that do not cause cell-death. Future investigations into the exact mechanisms by which TMR-TAT functions should help in the rational design of such compounds.
Despite being a problem for delivery applications, the photocytotoxicity of TMR-TAT might have interesting applications in Photodynamic therapy
Toxicity of TMR-TAT toward HeLa cells in the absence of light. TMR-TAT was incubated at the concentration indicated with cells for 1 hour in L-15 at 37°C. Cells were washed with fresh L-15 and incubated for an additional 4 h. Cells were then treated with L-15 containing DAPI and SYTOX® green. DAPI stains the nucleus of all cells while SYTOX® green only stains the nucleus of dead cells. Cells were imaged by fluorescence microscopy using the DAPI and FITC filters to detect DAPI and SYTOX® green, respectively. For each experiments, five representative images were acquired using the 20× objective and the percentage of dead cells were calculated from the ratio of cells stained by SYTOX® green divided by the number of cells stained by DAPI. The reported data is the average of 3 experiments (3×5 images, ≥1000 cells/experiments) and the error bar represents the standard deviation.
(TIF)
Photosensitization of TMR-riTAT endocytosed by HeLa cells. HeLa cells were incubated with TMR-riTAT (3 µM) for 1 h and washed with fresh L-15 media. Cells were then incubated with L-15 containing 1 µM SYTOX® Blue to detect cells with compromised plasma membranes. Cells were observed using a 100× objective using bright field and fluorescence imaging (RFP filter to detect TMR-riTAT, pseudocolored red, and CFP filter to detect SYTOX® Blue, pseudo colored blue). The images are the overlay of TMR and bright field images and the insert images are the overlay of SYTOX® Blue and bright field images. At a low exposure dose, TMR-riTAT is distributed in a punctate manner within cells and cells are impermeable to SYTOX® Blue. As with TMR-TAT, TMR-riTAT is however quickly redistributed thought the cell as light exposure is increased. As this take place, the cell shrinks and membrane blebs form. The nuclei of cells are also stained by SYTOX® Blue, indicating that the plasma membrane integrity is compromised.
(TIF)
Comparison of the fluorescence intensity of endosomes containing TMR-TAT (incubation at 3 µM) or TMR-K9 (incubation at 10 µM) in the images presented in
(TIF)
Light irradiation alone does not cause cell-death. TMR-TAT induces cell-death upon light irradiation in HeLa or COS-7 cells. A) The conditions of light irradiation used in
(TIF)
A) Photohemolysis of RBCs with TMR-TAT. Short light exposure tat 560 nm to RBCs incubated with TMR-TAT (3 µM) causes lysis and formation of ghost cells as observed by bright field imaging. Lysis of more than 90% of the cells can be achieved when light exposure is increased. Initially, the ghosts formed appear to have a constant diameter as shown in the images corresponding to exposure at 20 J/cm2 and 60 J/cm2. As light exposure is increased, the ghosts shrink to a much smaller diameter (120 J/cm2 image). This shrinkage was not observed when the ghosts formed by irradiation with 60 J/cm2 of light were incubated without additional light irradiation (data not shown). These results suggest that light irradiation causes damages to membranes well after lysis as occurred. B) TMR-TAT does not appear to penetrate RBCs. RBCs were incubated with TMR-TAT (10 µM) in PBS for 1 hour. The RBCs were then spun down at low speed and the supernatant was removed from the pelleted cells. Cells were rapidly washed with cold PBS (4°C) and spun down twice. Images are the bright field and TMR confocal fluorescence images before and after washing of the RBCs. During incubation, the interior of RBCs display a dark contrast when compared to the fluorescent peptide present in solution. After washing the cells, no appreciable TMR fluorescence could be detected. It is important to note that RBCs have a weak autofluorescence signal in the TMR channel. The contrast in the image represented was therefore adjusted to display a signal that would be above this autofluorescence background. In addition, light irradiation of the washed cells did not lead to photohemolysis. Together, these results suggest that TMR-TAT does not penetrate RBCs to a large extent.
(TIF)
We thank Professor Paul Straight and Michael Polymenis for valuable discussions.