Potentiation of Chlorin e6 Photodynamic Activity in Vitro with Peptide

These results highlight the utility of designing peptides as vehicles for regulating ... For a more comprehensive list of citations to this article, u...
1 downloads 0 Views 648KB Size
Download PDF
982

Bioconjugate Chem. 1999, 10, 982−992

Potentiation of Chlorin e6 Photodynamic Activity in Vitro with Peptide-Based Intracellular Vehicles Stuart K. Bisland, Devender Singh,† and Jean Garie´py* Department of Medical Biophysics, University of Toronto and the Division of Molecular and Structural Biology, Ontario Cancer Institute, Princess Margaret Hospital, 610 University Avenue, Toronto, Ontario, Canada M5G 2M9. Received February 23, 1999; Revised Manuscript Received August 20, 1999

Photodynamic therapy (PDT) is a targeted treatment modality where photosensitizers accumulate into cells and are selectively activated by light leading to the production of toxic species and cell death. Focusing the action of photosensitizers to a unique intracellular target may enhance their cytotoxicity. In this study, we demonstrate that the routing of the porphyrin-based photosensitizer chlorin e6, to the nucleus of cells can significantly alter its toxicity profile. The cellular localization of chlorin e6 was achieved by coupling the chromophore during solid-phase synthesis to a nucleus-directed linear peptide (Ce6-peptide) or a branched peptide (Ce6-loligomer) composed of eight identical arms displaying the sequence of the Ce6-peptide. These constructs incorporated signals guiding their cytoplasmic uptake and nuclear localization. Ce6-peptide and Ce6-loligomer displayed an enhanced photodynamic activity compared to unconjugated chlorin e6, lowering the observed CD50 values for CHO and RIF-1 cells by 1 or more orders of magnitude. The intracellular accumulation of Ce6-peptide and Ce6-loligomer was assessed by electron and confocal microscopy as well as by flow cytometry. Constructs were internalized by cells within an hour and by 6 h, the release of active oxygen species could be observed within the nucleus of cells pretreated with Ce6-loligomer. These results highlight the utility of designing peptides as vehicles for regulating the intracellular distribution of photosensitizers such as chlorin e6 in order to maximize their efficacy in PDT.

INTRODUCTION

The dynamic use of light to selectively kill proliferating cells is termed photodynamic therapy (PDT). PDT represents a noninvasive procedure requiring the use of a photosensitizer (a photosensitive compound) to generate cytotoxic oxygen radicals within a diseased tissue. This approach has recently been popularized by the introduction of Photofrin, a porphyrin derivative, presently being evaluated in clinical trials (1). Photofrin, a clinically effective first-generation photosensitizer, is characterized by a low absorbance in the red region of the spectrum and displays side effects such as prolonged skin photosensitivity (2). Recent animal studies found chlorin e6, a related porphyrin-based agent, to be a useful photosensitizer in vivo, with low toxicity and preferential tumor localization (3-5). The efficiency of photosensitizers in situ is ultimately dependent on their selective uptake into tissues and on their cellular accumulation within tissues. This finding is supported by the fact that singlet oxygen (1O2)-induced photodamage incurred upon porphyrin activation is usually localized to within 0.1 µm of its site of release (6). The toxicity of photodynamic drugs could potentially be enhanced by delivering high amounts of a photosensitizer into subcellular organelles such as the nucleus where nucleic acids represent target molecules sensitive to photodamage (7). The photodynamic potential of unbound chlorin e6 is limited as a result of its tendency to localize largely at the plasma membrane (8). Bovine serum albumin (BSA) conjugates incorporating chlorin * To whom correspondence should be addressed at the Ontario Cancer Institute. Phone: (416) 946-2967. Fax: (416) 9466529. E-mail: [email protected]. † Present address: Gemma Biotechnology, 620 University Avenue, Toronto, Canada M5G 2M9.

e6 and nuclear localization sequences have recently been assembled to guide their cellular localization (11). These PDT-labeled conjugates have an enhanced phototoxicity in relation to free chlorin e6 (9-12). In particular, BSA conjugates incorporating insulin have been reported to target the insulin receptor in an effort to create cellspecific constructs (11). The composition of these BSA constructs, however, remains approximate and had to be inferred from the average ratio of each component present in these conjugates as estimated from polyacrylamide gel electrophoresis and by absorbance measurements based on the incorporation of chlorin e6 molecules (11). The engineering and characterization of defined, optimal delivery agents based on the chemical crosslinking of multiple components remain a challenging and difficult task. Our group has recently designed branched peptides that act as multitasking intracellular vehicles (13-15). This first generation of vehicles incorporates eight identical peptide arms coding for two functional domains, namely a pentalysine domain acting as a cytoplasmic transport sequence (CTS)1 and the simian virus SV40 large T antigen nuclear localization signal (NLS) which 1 Abbreviations: A124SV40LTa (124-135), residues 124-135 of the SV40 large T antigen with a threonine to alanine substitution at position 124; C6-peptide, a linear peptide composed of a chlorin e6 group linked to a nuclear localization signal and a pentalysine cytoplasmic translocation domain; C6loligomer, a loligomer harbouring 8 N-terminal arms that comprise of a chlorin e6 group followed by a nuclear localization signal and a pentalysine cytoplasmic translocation domain; CTS, cytoplasmic translocation signal; loligomer, a squid-like branched peptide construct that incorporates cell penetration and intracellular localization signals; NLS, nuclear localization signal.

10.1021/bc990020u CCC: $18.00 © 1999 American Chemical Society Published on Web 10/16/1999

Intracellular Delivery of a PDT Drug

guide their nuclear uptake (13-15). These squid-like, intracellular vehicles are referred to as loligomers (derived from the Latin root loligo referring to members of the squid family). This study describes the assembly and application of a nucleus-directed linear peptide and a loligomer incorporating the photosensitizer chlorin e6. The chromophore was introduced during solid-phase peptide synthesis and the resulting constructs display a photodynamic activity that exceeded the toxicity of free chlorin e6. MATERIALS AND METHODS

Synthesis of Chlorin e6 Peptides and Their Derivatives. Peptides were assembled using tert-butoxycarbonyl (t-Boc) chemistry and phenylacetamidomethyl (PAM) resin supports as described elsewhere (13, 14). Chlorin e6 was coupled to the N-terminus of the peptide and the loligomer by first activating its carboxylic groups with 0.5 molar equivalent of dicyclohexylcarbodiimide and N-methyl-2-pyrrolidone at 23 °C for 40 min. The activated chlorin e6 (4 molar excess in relation to the number of N-terminal amino groups) was then added to the reaction vessel containing the peptide resins. The coupling reaction was left to proceed in the dark for 24 h at room temperature. Both chlorin e6-containing peptides were desalted using Sephadex G10 (Ce6-peptide) or G25 size-exclusion columns (Ce6-loligomer), respectively. Columns were preequilibrated in phosphate-buffered saline (PBS; pH 7.2). Ce6-peptide was further purified by reversed-phase HPLC (Lichrosorb C18) using a linear gradient of acetonitrile in water (20 to 60%) containing 0.1% (v/v) trifluoroacetic acid. The peptide eluted at 40% acetonitrile:water. The mass of Ce6-peptide was established by ionization mass spectroscopy. (MH+ ion: calculated, 2889; observed, 2888). The amino acid composition of both peptides were confirmed by amino acid analysis. The calculated and expected (parentheses) number of residues comprising each peptide construct are listed in accordance with the three-letter code for each amino acid. Ce6-peptide: Arg, 1.0 (1); Asp, 0.9 (1); Glu, 1.0 (1); Gly, 1.9 (2); Lys, 8.3 (9); Pro, 3.0 (3); Ala, 0.9 (1); Tyr, 1.0 (1); Val, 1.0 (1). Ce6-loligomer: Arg, 8.0 (8); Asp, 6.6 (8); Glu, 9.8 (8); Gly, 19.8 (17); Lys, 81.4 (79); Pro, 22.5 (24); Ala, 8.3 (8); Tyr, 2.4 (1); Val, 8.3 (8). The absorption spectra of either chlorin e6, Ce6-peptide, or Ce6-loligomer displayed a characteristic Soret band (maximum at ∼401 nm) and Q-band (maxima at 654, 656 and 664 nm), respectively (data not shown). Using chlorin e6 as a standard, extinction coefficients were calculated at 401 nm [Ce6-peptide (17 000 M-1 cm-1) and Ce6loligomer (129 000 M-1 cm-1)] and at 664 nm [chlorin e6 (10 000 M-1 cm-1), Ce6-peptide (12 000 M-1 cm-1), and Ce6-loligomer (40 000 M-1 cm-1)]. The calculated number of chlorin e6 molecules attached to Ce6-loligomer based on amino acid analysis and spectral properties (410 nm) was 7.6 or 0.95 mol of chlorin e6/mol of loligomer arm (7.6/8). A cysteine residue was included within the C-terminus (analytical arm) of each peptide to allow the attachment of either a maleimido-containing fluorescent probe or a biotin moiety. Texas red or fluorescein were incorporated by dispensing 1 mg of Ce6-peptide or Ce6loligomer in 1 mL of PBS, pH 7.4 into a glass vial and by adding dropwise a solution of 0.1 mg of either Texas red C2 maleimide or fluorescein-5-maleimide dissolved in 100 µL of dimethylformamide (Molecular Probes, Eugene, OR). The coupling reactions were left to proceed with constant stirring for 4 h at 4 °C in the dark. The molar ratio of each incorporated fluorochrome was 3-fold (3.3 times; 49.9 ( 12.1 particles/µm2) higher than the level of Ce6-peptide-biotin (15.3 ( 5.4 particles/µm2). In contrast, the amount of Ce6loligomer-biotin in the cytoplasm of RIF-1 cells after 1 and 6 h incubations was 0.5 (12.8 ( 1.9 vs 27.9 ( 9.3 particles/µm2) and 0.7 (46.1 ( 16.4 vs 34.3 ( 4.9particles/ µm2) times that of Ce6-peptide-biotin, respectively. Analysis for individual cells demonstrated that after an incubation period of 1 or 6 h, Ce6-peptide-biotin preferentially localized within the cytoplasm of RIF-1 cells with less than half of the total amount of biotinylated peptide within the cell being present in the nucleus (0.4, 12.4 ( 2.2 vs 27.9 ( 9.3 particles/µm2, and 0.3, 15.3 ( 5.4 vs 46.1 ( 16.4 particles/µm2, respectively). Meanwhile, the amount of Ce6-loligomer-biotin within the nucleus of RIF-1 cells was 1.5 times (19.2 ( 1.9 vs 12.8 ( 1.9

particles/µm2) and 1.4 times (49.9 ( 12.1 vs 34.3 ( 4.9 particles/µm2) the amount in the cytoplasm after a 1 or 6 h incubation, respectively. These results confirm the enhanced nuclear localization properties of Ce6-loligomer-biotin compared with Ce6-peptide-biotin. Photodynamic Action of Nucleus-Targeted Chlorin e6-Peptide Constructs. The viability of CHO and RIF-1 cells was determined as a function of exposure to increasing concentrations of chlorin e6 or the two peptide constructs using an established cell viability assay (WST-1 tetrazolium salt assay). Dose response curves (Figure 6) indicate that both constructs were moderately cytotoxic toward CHO and RIF-1 cells in the absence of photoirradiation (Figure 6, panels a and b). CD50 values determined for chlorin e6 and the related peptide conjugates following photoirradiation (Figure 6, panels c and d), revealed that the photodynamic activity of Ce6loligomer against CHO cells (CD50; 20 nM) was 400 times higher than that of chlorin e6 alone (CD50; 8 µM), whereas the activity of Ce6-peptide (CD50; 200 nM) was 40 times that of chlorin e6. The photodynamic activity of Ce6-

988 Bioconjugate Chem., Vol. 10, No. 6, 1999

Bisland et al.

Figure 5. Electron micrographs (negative prints) detailing the endocytic uptake and migration of Ce6-loligomer-biotin (0.1 µM) in RIF-1 cells. The loligomer construct was initially internalized through the plasma membrane by absorptive endocytosis (panel a; arrow) and subsequently encapsulated within a membrane-bound endocytic vesicle (panel b). Within 1 h of exposure to Ce6-loligomerbiotin, RIF-1 cells contained large cytoplasmic vesicles densely packed with gold particle-associated Ce6-loligomer-biotin (panel c). The construct accumulated within regions of the nucleus (n) after 6 h (panel d). Scale bar represents 0.88 µm (panels a and b), 0.94 µm (panel c) or 1.85 µm (panel d). Table 1. Quantification of Gold Particles in Electron Micrographs of RIF-1 Cells Exposed to Biotinylated Ce6-Peptide and Ce6-Loligomer peptide construct

biotinylated Ce6-peptidea

biotinylated Ce6-loligomera

time (h) 1 6 1 6 nucleus (N) 12.4 ( 2.2 15.3 ( 5.4 19.2 ( 1.9 49.9 ( 12.1 cytoplasm (C) 27.9 ( 9.3 46.1 ( 16.4 12.8 ( 1.9 34.3 ( 4.9 total cellular 40.4 ( 11.5 61.4 ( 21.8 32 ( 3.8 84.2 ( 17 uptake (N + C) N/C ratio 0.4 0.3 1.5 1.4 a Biotinylated peptide constructs were labeled with streptavidinconjugated gold particles. Values are expressed as the number of gold particles per square micron area (µm2) of cell.

loligomer against RIF-1 cells (CD50; 1.6 µM) was less than that for CHO cells although significantly enhanced (∼40fold) in relation to chlorin e6 (CD50; 63 µM). The photo-

toxicity of Ce6-peptide toward RIF-1 cells (CD50; 6.3 µM) as in the case of CHO cells, was less than the effects of Ce6-loligomer and showed a 10-fold killing enhancement relative to chlorin e6. The enhanced phototoxicity of our Ce6 peptide conjugates thus depends in part on the nature of the cell line targeted. Generation of Reactive Oxygen Species within the Nucleus of RIF-1 Cells by Chlorin e6-Loligomer Construct. To establish the precise subcellular sites of photoactivation/photooxidation of chlorin e6 and to confirm that the conjugation of chlorin e6 to the loligomer did not interfere with its activity, RIF-1 cells which had previously been treated with Ce6-loligomer-TxRd (0.01 µM) for 6 h were subsequently treated with CM-H2 DFDA (5 µM). CM-H2 DFDA passively diffuses into cells, where the acetates are cleaved by intracellular esterases to form reduced 2′,7′-dichlorofluorescin. The subsequent produc-

Intracellular Delivery of a PDT Drug

Bioconjugate Chem., Vol. 10, No. 6, 1999 989

Figure 6. Cell viability curves for CHO and RIF-1 cells incubated (6 h) in the presence of increasing concentrations of chlorin e6 (9), Ce6-peptide (2), or Ce6-loligomer (b). Cell viability was determined for cells exposed (+) or not (-) to photoirradiation. Cell viability was reported as the percentage of surviving cells measured using the WST-1 assay (ref 17; see Materials and Methods). The term drug concentration represents the molar concentration of chlorin e6 either free or attached to the peptide constructs.

tion of active oxygen species induced by photoirradiation oxidizes 2′,7′-dichlorofluorescin to a highly fluorescent product, dichlorofluorescein (28). The thiol-reactive chloromethyl groups of CM-H2 DFDA potentially interact with intracellular thiols to form adducts that prolong its retention within cells. This event allows the observation of reactive oxygen for several seconds after chlorin e6 activation (Figure 7). Cells were either treated in the presence (Figure 7f) or absence (Figure 7b) of AcrO (5 µM) to indicate the position of the cell nucleus. The fluorescent signal generated by Ce6-loligomer-TxRd was recorded prior to the activation of chlorin e6 (Figure 7, panels a and e) as well as after chlorin e6 activation (Figure 7, panels c and g). CM-H2 DFDA-associated fluorescence was undetectable prior to the activation of chlorin e6 (Figure 7, panels b and f). Photoirradiation of the cells and the subsequent activation of intracellular chlorin e6 resulted in a delay of ∼10-20 s before the generation of reactive oxygen species within RIF-1 cells, an event that was confirmed by a strong fluorescence burst which persisted for 1-2 min. The release of reactive oxygen species colocalizes exactly with the site of Ce6loligomer accumulation (see Figure 7, panels a and d, or Figure 7, panels e and h). AcrO staining provided evidence of its release in the nucleus of these cells (see Figure 7, panels f and h). DISCUSSION

Photodynamic therapy has emerged within the past decade as a promising treatment modality for certain malignancies, offering the possibility of guided therapies which restrict phototoxicity to the tumor site. PDT is currently used in the treatment of early and late stage cancers including bladder, esophageal, lung, gastric, and

cervical cancers and is either in use or being evaluated in clinical trials for early stage oesophageal and lung cancer (30). The phototoxicity of photosensitizers is associated with the production of singlet oxygen species (1O2), which has a lifetime of several microseconds in aqueous environments and a limited cellular diffusion [∼20 nm (31, 32)]. The sphere of influence of 1O2 is thus very small, and only intracellular regions with both a high 1O2 concentration and a high photosensitizer concentration are susceptible to photooxidative damage. Therefore, the location of the photosensitizer within cells governs the areas subjected to its photodynamic action. Potential targets of 1O2-mediated toxicity include cellular membranes which may undergo lipid peroxidation, protein cross-linking and loss of ionic homeostasis (33), lysosomes which can release acidic hydrolases into the cytosol (34), mitochondria as well as the endoplasmic reticulum (35).1O2 is also known to cause single-strand breaks and alkali-labile lesions in DNA (36) as well as inactivate enzymes involved in DNA repair (37). Consequently, the nucleus represents an attractive target for photooxidative damage. Unfortunately, photosensitizers tend to accumulate to the perinuclear region of the cell cytoplasm. As a result, DNA is not usually subjected to photoinduced toxicity. The realization that certain sites within a cell display enhanced sensitivity to photooxidative damage justifies the design of delivery strategies which target photosensitizers to these sensitive sites. One strategy would be to use intracellular vehicles such as linear and branched peptides (13-15) to shuttle photosensitizers to specific cellular loci. The results of this study demonstrate that peptide constructs can be used to traffic the photodynamic drug chlorin e6 to the nucleus of cells and, in doing so, can potentiate its cytotoxicity.

990 Bioconjugate Chem., Vol. 10, No. 6, 1999

Bisland et al.

Figure 7. Light activation of chlorin e6 in RIF-1 cells as monitored by fluorescence microscopy. Photoinduced generation of reactive oxygen species within two individual RIF-1 cells (cell 1, panels a-d; cell 2, panels e-h) following their treatment with Ce6-loligomerTxRd for 6 h. Panels a, b, e, and f depict fluorescence images recorded prior to light exposure and panels c, d, g, and h are images taken following light activation. Cell 2 was stained with AcrO to identify areas within the cell containing DNA (panel f). CM-H2DCFDA was used to visualize the generation of reactive oxygen species in cells following the activation of chlorin e6 (panels d and h). The fluorescent signal of CM-H2DCFDA colocalizes with the signal of Ce6-loligomer-TxRd in both cells.

Chlorin e6 was coupled to either a nucleus-directed loligomer (Ce6-loligomer; Figure 1C) or a peptide corresponding to one arm of this construct (Ce6-peptide; Figure 1B) in an effort to guide the intracellular uptake and nucleus routing of this photosensitizer. The endocytosis and subsequent localization of these constructs within cell nuclei could be visualized using fluorescently tagged constructs with confocal microscopy and by following the migration of biotinylated constructs inside cells using electron microscopy. The time course and extent of intracellular accumulation of fluorescently tagged constructs were also assessed by flow cytometry. Cellular uptake studies were conducted with CHO and RIF-1 cells maintained in suspension. Loligomers have been shown in the past to be internalized rapidly into eukaryotic cells by absorptive endocytosis, with their maximal uptake occurring between 6 and 8 h of incubation (13, 14). In contrast, unconjugated chlorin e6 is taken up by passive membrane diffusion (38, 39). The interaction between photosensitizers and cells is ultimately governed by their chemical properties. Negatively charged hydrophobic photosensitizers such as chlorin e6 tend to concentrate at the surface of membranes with minimal electrostatic drive to actually traverse them (7). In contrast, cationic dyes such as rhodamines and kryptocyanines, are subject to electrochemical gradients which promote their interaction with anionic sites and subsequent translocation across cellular membranes, in particular, across mitochondrial membranes (40). Ce6-loligomer and Ce6-peptide were rapidly imported inside CHO cells (Figures 2 and 3). Furthermore, Ce6-peptide uptake into CHO cells was considerably greater than the level of import previously observed for a similar linear peptide which lacked chlorin e6 (13), suggesting a role for the porphyrin group in promoting the internalization of the construct. Another factor which may influence intracellular uptake is the net negative charge of the cell membrane which varies depending on cell type and cell function (41).

All cells (CHO and RIF-1) were targeted by the constructs as confirmed by flow cytometry (results not shown) and by confocal microscopy (Figure 3). However, the distribution of biotinylated constructs inside cells as revealed by electron microscopy indicate that nuclear uptake is only significant for the Ce6-loligomer constructs (Figure 4, panels c and d). Electron microscopy also provided evidence suggesting that biotinylated variants of both peptide (Figure 4, panels a and c) and loligomer (Figure 4, panels b and d) constructs enter the cytosol of target cells in a similar fashion. To reach the cell nucleus, these constructs must first escape their vesicular compartments with their NLS sequences still intact (13-15). The precise mechanism underlying this translocation event of constructs across endosomal membranes is presently unknown. Zauner et al. (42) suggested that polylysine may under certain circumstances promote lysis of internal vesicles. The distribution of biotinylated Ce6peptide and Ce6-loligomer constructs inside cells was monitored with colloidal gold particles conjugated to steptavidin. The constructs were released from endocytic vesicles into the cytoplasm and subsequently routed to the nucleus (Figure 5). Dual labeling of loligomers with distinct fluorochromes attached to the N-terminus and C-terminus arms have indicated that both probes remain at least partly associated following their release from vesicular compartments (Sheldon and Garie´py, unpublished results) and that 80% of the intracellular pool of loligomer is retained within CHO cells (14). These results suggest that the degradation of signals on loligomers is not extensive within the time scale of our measurements (hours) and that cycling of such constructs back to the cell surface does not represent a dominant cellular pathway. This result was confirmed in the present study where the fluorescence signal associated with both chlorin e6 linked to the N-terminus and a fluorescein label attached to the C-terminal end of Ce6-loligomer-FL5 were shown to co-localize in cells over a period of at least 24 h (Figure 3). The branching design of loligomers with

Intracellular Delivery of a PDT Drug

multiple SV40 NLS sequences was also shown to enhance nuclear localization (ref 13; see also Figures 3-5 and Table 1). The quantitation of gold particles present in the cytosol and nucleus of RIF-1 cells indicated that the amount of Ce6-loligomer in their nuclei at 6 h was more than three times that of Ce6-peptide (Table 1). Once inside the nucleus, biotinylated loligomer constructs appear to localize in areas of high DNA content. It has previously been documented that cationic photosensitizers (e.g., methylene blue) have a greater affinity for nucleic acid interactions compared with noncationic photosensitizers (43). The incorporation of eight chlorin e6 molecules into a single loligomer construct (Ce6-loligomer) was found to enhance its toxicity by 400-fold in CHO cells and by 40fold in RIF-1 cells when compared on a mole-to-mole basis with chlorin e6 alone (Figure 6). Similarly, the phototoxicity of the single arm construct Ce6-peptide, although modest in relation to Ce6-loligomer, was greater than the toxicity level observed for chlorin e6. A comparison of CD50 values for Ce6-loligomer with values obtained for loligomer without the photosensitizer (14) reveals a shift in CD50 values of almost 4 log units. These results point out that the level of enhanced phototoxicity observed using our chlorin e6-containing peptide constructs is dependent on the cell line targeted. The present study supports results published recently by Akhlynina et al. (11) describing the construction and delivery of macromolecular conjugates composed of insulin, chlorin e6, and NLS peptides chemically cross-linked to bovine serum albumin into cells. The insulin-containing BSA constructs were shown to enhance the photodynamic activity of chlorin e6 by >2000-fold. A comparison of the levels of enhancement in phototoxicity observed between our compounds and these BSA conjugates is difficult to establish since the basal toxicity of the insulin-containing BSA constructs in the absence of photoirradiation was not reported and the cell lines tested by Akhlynina’s group were different from the ones used in our studies. The BSA conjugates incorporating insulin do target the insulin receptor and as such may represent constructs affording a more guided delivery. However, this past study did not provide phototoxicity data of these constructs toward a control cell line that did not express the insulin receptor. Akhlynina et al. (11) suggested that the enhanced PDT activity of their constructs could be partly linked to the number of NLS peptides attached to the conjugates, rather than the number of chlorin e6 molecules. In the present study, the introduction of multiple NLS sequences in the structure of Ce6-loligomer also resulted in the enhanced nuclear localization potential and cytotoxicity profile of this construct in relation to Ce6-peptide. From the perspective of designing intracellular vehicles, the simplicity and directed assembly of Ce6-peptide and Ce6-loligomer can be contrasted with the preparation of the reported BSA conjugates (11). More precisely, the average ratio of components in these BSA conjugates had to be inferred from molecular weight estimates derived from polyacrylamide gel electrophoresis and optical density measurements of chlorin e6. The present study demonstrates that peptide-based vehicles such as Ce6peptide and Ce6-loligomer represent simple and easily constructed synthetic assemblies able to deliver PDT agents into specific areas of a cell which display heightened photosensitivity in order to improve the photodynamic efficiency of these agents. The mechanism of Ce6-loligomer-mediated toxicity was shown to involve the generation of reactive oxygen species, confirming that the activation of chlorin e6 is not

Bioconjugate Chem., Vol. 10, No. 6, 1999 991

hindered upon its attachment to the loligomer (Figure 7). Although it was demonstrated that Ce6-loligomer reaches the cytosol and accumulates in the nucleus of CHO and RIF-1 cells (Figures 3-5 and Table 1), it remains unclear if the enhanced phototoxicity observed is predominantly due to singlet oxygen events occurring in the nucleus of cells as a result of nuclear localization or a simple consequence of the improved cellular uptake of chlorin e6 when attached to this peptide construct. The sites of production of reactive singlet oxygen species as reported using CM-H2-DFDA (Figure 7) suggest a diffuse distribution of the reactive 1O2 species throughout the cytosol and nucleus of RIF-1 cells. Preliminary results from flow cytometry studies also confirmed that PDTinduced cell death following treatment with uncoupled chlorin e6 or related constructs occurred via an apoptotic pathway for CHO and RIF-1 cell lines (results not shown). Morphological signs of apoptotic cell death were apparent as cytoplasmic blebbing and shrunken nuclei. In summary, the cytotoxicity of photosensitizers such as chlorin e6, in the context of photodynamic therapy, is dependent on the intracellular localization of these agents. The incorporation of chlorin e6 into peptide-based intracellular vehicles will dictate its toxicity. In particular, the routing of chlorin e6 to the cytoplasm and nucleus of CHO and RIF-1 cells using intracellular peptide shuttles can alter its phototoxicity. ACKNOWLEDGMENT

The authors thank Drs. Lothar Lilge and Eric LaCasse for their assistance with experimental apparatus and protocol, Battista Calvieri, Steven Doyle, and Trudy Franklin for their kind help with electron and confocal microscopy, Juliet Sheldon for her technical assistance in operating the flow cytometer and Jim Ferguson for synthesizing peptides. This work was supported by a grant from the National Cancer Institute of Canada with funds from the Canadian Cancer Society. LITERATURE CITED (1) Boyle, R. W., and Dolphin, D. (1996) Structure and biodistribution relationships of photodynamic sensitizers. Photochem. Photobiol. 64, 469-485. (2) Bellnier, D. A., Ho, Y.-K., Pandey, R. K., Missert, J. R., and Dougherty, T. J. (1989) Distribution and elimination of Photofrin II in mice. Photobiol. 50, 221-228. (3) Kostenich, G. A., Zhuravkin, I. N., Furmanchuk, A. V., and Zhavrid, E. A. (1991) Photodynamic therapy with chlorin e6. A morphologic study of tumor damage efficiency in experiment. J. Photochem. Photobiol., B 11, 307-318. (4) Kostenich, G. A., Zhuravkin, I. N., Furmanchuk, A. V., and Zhavrid, E. A. (1993) Sensitivity of different rat tumour strains to photodynamic treatment with chlorin e6. J. Photochem. Photobiol., B 17, 187-194. (5) Kostenich, G. A., Zhuravkin, I. N., and Zhavrid, E. A. (1994) Experimental grounds for using chlorin e6 in the photodynamic therapy of malignant tumours. J. Photochem. Photobiol., B 22, 211-217. (6) Henderson, B. W., and Dougherty, T. J. (1992) How does photodynamic therapy work? J. Photochem. Photobiol. 55, 145-157. (7) Peng, Q., Moan, J., and Nesland, J. M. (1996) Correlation of subcellular and intratumoral photosensitizer localization with ultrastructural features after photodynamic therapy. Ultrastruct. Pathol. 20, 109-129. (8) Bachor, R., Shea, C. R., Gillies, R., and Hasan, T. (1991) Photosensitized destruction of human bladder carcinoma cells treated with chlorin e6-conjugated microspheres. Proc. Natl. Acad. Sci. U.S.A 88, 1580-1584. (9) Akhlynina, T. V., Rosenkranz, A. A., Jans, D. A., Gulak, P. V., Serebryakova, N. V., and Sobolev, A. S. (1993) The use of

992 Bioconjugate Chem., Vol. 10, No. 6, 1999 internalizable derivatives of chlorin e6 for increasing its photosensitizing activity. Photochem. Photobiol. 58, 45-48. (10) Akhlynina, T. V., Rosenkranz, A. A., Jans, D. A., and Sobolev, A. S. (1995) Insulin-mediated intracellular targeting enhances the photodynamic activity of chlorin e6. Cancer Res. 55, 1014-1019. (11) Akhlynina, T. V., Jans, D. A., Rosenkranz, A. A., Statsyuk, N. V., Balashova, I. Y., Toth, G., Pavo, I., Rubin, A. B., and Sobolev, A. S. (1997) Nuclear targeting of chlorin e6 enhances its photosensitizing activity. J. Biol. Chem. 272, 2032820331. (12) Sobolev, A. S., Akhlynina, T. V., Yachmenev, S. V., Rosenkranz, A. A., and Severin, E. S. (1992) Internalizable insulinBSA-chlorin e6 conjugate is a more effective photosensitizer than chlorin e6 alone. Biochem. Int. 26, 445-450. (13) Sheldon, K., Liu, D., Ferguson, J., and Garie´py, J. (1995) Loligomers: Design of de novo peptide-based intracellular vehicles. Proc. Natl. Acad. Sci. U.S.A 92, 2056-2060. (14) Singh, D., Kiarash R., Kawamura K., LaCasse E. C., and Garie´py J. (1998) Penetration and intracellular routing of nucleus-directed peptide-based shuttles (Loligomers) in eukaryotic cells. Biochemistry 37, 5798-5809. (15) Singh, D., Bisland, S. K., Kawamura, K., and Garie´py, J. (1999) Peptide-based intracellular shuttle able to facilitate gene transfer in mammalian cells. Bioconjugate Chem. 10, 745-754. (16) McBurney, M. W., and Whitmore, G. F. (1974) Isolation and biochemical characterization of folate deficient mutants of chinese hamster cells. Cell 2, 173-182. (17) Schmid, I., Krall, W. A., Uittenbogaart, C. H., Braun, J., and Giorgi, J. V. (1992) Sensitive method for measuring apoptosis and cell surface phenotype in human thymocytes by flow cytometry. Cytometry 13, 204-208. (18) Mosmann, T. J. (1983) Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. Immunol. Methods 65, 55-63. (19) Kalderon, D., Richardson, W. D., Markham, A. F., and Smith A. E. (1984) Sequence requirements for nuclear location of simian virus 40 large-T antigen. Nature 311, 33-38. (20) Goldfarb, D., Garie´py, J., Schoolnik, G., and Kornberg, R. D. (1986) Synthetic peptides as nuclear localization signals. Nature 322, 641-644. (21) Lanford, R. E., Kanda, P., and Kennedy, R. C. (1986) Induction of nuclear transport with a synthetic peptide homologous to the SV40 T antigen transport signal. Cell 46, 575-582. (22) Rihs, H. P., Jans, D. A., Fan, H., and Peters, R. (1991) The rate of nuclear cytoplasmic protein transport is determined by the casein kinase II site flanking the nuclear localization sequence of the SV40 T-antigen. EMBO J. 10, 633-639. (23) Jans, D. A., Ackermann, M. J., Bischoff, J. R., Beach, D. H., and Peters, R. (1991) p34cdc2-mediated phosphorylation at T124 inhibits nuclear import of SV-40 T antigen proteins. J. Cell Biol. 115, 1203-1212. (24) Ryser, H. J.-P., and Shen, W. C. (1978) Conjugation of methotrexate to poly (L-lysine) increases drug transport and overcomes drug resistance in cultured cells. Proc. Natl. Acad. Sci. U.S.A. 75, 3867-3870. (25) Shen, W.-C., and Ryser, H. J.-P. (1978) Conjugation of poly (L-lysine) to albumin and horseradish peroxidaase: a novel method of enhancing the cellular uptake of proteins. Proc. Natl. Acad. Sci. U.S.A. 75, 1872-1876.

Bisland et al. (26) Shen, W. C., and Ryser, H. J.-P. (1979) Poly (L-lysine) and poly (D-lysine) conjugates of methotrexate: different inhibitory effect on drug resistant cells. Mol. Pharmacol. 16, 614622. (27) Arnold, L. J., Jr., Dagan, A., Gutheil, J., and Kaplan, N. O. (1979) Antineoplastic activity of poly (L-lysine) with some acsites tumor cells. Proc. Natl. Acad. Sci. U.S.A. 76, 32463250. (28) Tam, J. P. (1988) Synthetic peptide vaccine design: synthesis and properties of a high-density multiple antigenic peptide system. Proc. Natl. Acad. Sci. U.S.A. 85, 5409-5413. (29) Patel, A. K., Hallet, M. B., and Campbell, A. K. (1987) Threshold responses in production of reactive oxygen metabolites in individual neutrophils detected by flow cytometry and microfluorimetry. Biochem. J. 248, 173-180. (30) Schuitmaker, J. J., Baas, P., van Leengoed, H. L. L. M., van der M., Star, W. M., and van Zandwijk, N. (1996) Photodynamic therapy: a promising new modality for the treatment of cancer. J. Photochem. Photobiol., B 34, 3-12. (31) Moan, J. (1990) On the diffusion length of singlet oxygen in cells and tissue. J. Photochem. Photobiol., B 6, 343-344. (32) Moan, J., and Berg, K. (1991) The photodegradation of porphyrins in cells can be used to estimate the lifetime of singlet oxygen. Photochem. Photobiol. 53, 549-553. (33) Kessel, D. (1977) Effects of photoactivated porphyrins at the cell surface of L1210 cells. Biochemistry 16, 3443-3449. (34) Jori, G., and Spikes, J. D. (1984) Photobiochemistry of porphyrins. In Topics in Photomedicine (K. C. Smith, Ed.) pp 183-318. (35) Moan, J., Johannessen, J. V., Christensen, T., Espevik, T., and Mcghie, J. B. (1982) Porphyrin-sensitized photoactivation of human cells in vitro. Am. J. Pathol. 109, 184-192. (36) Gomer, C. J. (1980) DNA damage and repair in CHO cells following hematoporphyrin photoradiation. Cancer Lett. 11, 161-167. (37) Boegheim, J. P., Scholte, H., Dubbelman, T. M. A. R., Beems, E., Raap, A. K., and van Steveninck, J. (1987) Photodynamic effects of hematoporphyrin-derivative on enzyme activities of murine L929 fibroblasts. J. Photochem. Photobiol. B. Biol. 1, 61-73. (38) Spikes, J. D. (1990) Chlorins as photosensitizers in biology and medicine. J. Photochem. Photobiol., B 6, 259-274. (39) Schmidt-Erfurth, U., Diddens, H., Birngruber, R., and Hasan, T. (1997) Photodynamic targeting of human retinoblastoma cells using covalent low-density lipoprotein conjugates. Br. J. Cancer 75, 54-61. (40) Oseroff, A. R., Ohuoha, D., Ara, G., McAuliffe, D., Foley, J., and Cincotta, L. (1986) Intramitochondrial dyes allow selective in vitro photolysis of carcinoma cells. Proc. Natl. Acad. Sci. U.S.A 83, 9729-9733. (41) Bauer, J. (1987) Electrophoretic separation of cells. J. Chromatogr. 418, 359-383. (42) Zauner, W., Kichler, A., Schmidt, W., Mechtler, K., and Wagner, E. (1997) Glycerol and polylysine synergize in their ability to rupture vesicular membranes: a mechanism for increased transferrin-polylysine-mediated gene transfer. Exp. Cell Res. 232, 137-145. (43) Gomer, C. J., Rucker, N., Ferrario, A., and Wong, S. (1989) Properties and applications of photodynamic therapy. Radiation Res. 120, 1-18.

BC990020U