SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Topical or systemic administration of 5-aminolevulinic acid (ALA) and its esters results in increased production and accumulation of protoporphyrin IX (PpIX) in cancerous lesions allowing effective application of photodynamic therapy (PDT). The large concentrations of exogenous ALA practically required to bypass the negative feedback control exerted by heme on enzymatic ALA synthesis and the strong dimerization propensity of ALA are shortcomings of the otherwise attractive PpIX biosynthesis. To circumvent these limitations and possibly enhance the phototoxicity of PpIX by adjuvant chemotherapy, covalent bonding of PpIX with a drug carrier, β-cyclodextrin (βCD) was implemented. The resulting PpIX + βCD product had both carboxylic termini of PpIX connected to the CD. PpIX + βCD was water soluble, was found to preferentially localize in mitochondria rather than in lysosomes both in MCF7 and DU145 cell lines while its phototoxiciy was comparable to that of PpIX. Moreover, PpIX + βCD effectively solubilized the breast cancer drug tamoxifen metabolite N-desmethyltamoxifen (NDMTAM) in water. The PpIX + βCD/NDMTAM complex was readily internalized by both cell lines employed. Furthermore, the multimodal action of PpIX + βCD was demonstrated in MCF7 cells: while it retains the phototoxic profile of PpIX and its fluorescence for imaging purposes, PpIX + βCD can efficiently transport tamoxifen citrate intracellularly and confer cell death through a synergy of photo- and chemotoxicity.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Photodynamic therapy of cancer (PDT) [1, 2] provides cancer treatment through the synergy of three essential, yet individually nonchemotoxic components: (1) The photosensitizer (PS), a light activated drug, (2) light of the appropriate wavelength to activate the PS and (3) oxygen which is the terminal generator of toxic species [3, 4]. Most of the PSs used in PDT are porphyrins, tetrapyrrole derivatives that stimulate production of cytotoxic reactive oxygen species in tissues illuminated with visible light. Porphyrins are known for their insolubility in aqueous media due to strong aggregation phenomena, the driving force being the overall thermodynamic gain due to ππ-stacking and hydrophobic interactions [5, 6]. Aggregation is manifested by widening and splitting of the intense Soret band around ca 400 nm in the absorption spectra, as well as deviations from Beer's law and the quenching of fluorescence with increase in concentration. For example, cationic porphyrins [7] show concentration-, temperature-, pH- and ionic strength–dependent aggregation. Protoporphyrin IX (PpIX) is an endogenous porphyrin in heme biosynthesis from 5-aminolevulinic acid (ALA). PpIX is the penultimate step prior to heme production following complexation with Fe(II). Exogenous administration of ALA and its esters [8] practically bypasses the negative feedback control exerted by heme on enzymatic ALA synthesis, thus making ALA a very efficient PpIX prodrug widely used for clinical PDT, following either topical or systemic application [9]. Nevertheless it was previously shown that exogenous application of PpIX to cells resulted in very efficient photocytotoxicity [10]. One serious limitation in the exogenous use of PpIX much alike many other PSs is its low water solubility, and consequently the difficulty in preparing pharmaceutical formulations. Alternative strategies have been sought for improving the delivery characteristics of the PSs including liposomal formulations, oil dispersions/micelles, polymeric nanoparticles, or polymer-PS conjugates [11]. Moreover, it seems that PS amphiphilicity could be an advantage: molecules should be hydrophilic enough to travel through the blood stream but also lipophilic enough to bind to the receptors on cancer cell membranes [12]. PpIX (Scheme Scheme 1) is a characteristic example of an amphiphilic, water insoluble porphyrin [13]. At pH 0–3 PpIX appears in monomeric form due to repulsion between charged units. At pH > 8 PpIX aggregates mostly as dimers via axial ππ interactions and lateral hydrophobic forces. At intermediate pH (3–7) aggregation to larger entities is favored due to additionally formed H-bonds among the propionic acid tails [13]. In the crystalline state the unit cell of PpIX dimethyl ester is composed of two molecules, the vinyl groups being nearly coplanar with the pyrrole rings [14].

Cyclodextrins (CDs) are an extensively studied family of cyclic oligosaccharides composed of α-d-glucopyranose units (α-, β- και γ-CDs with 6, 7 or 8 units respectively) [15]. In aqueous media CDs act as hosts that encapsulate hydrophobic molecules such as drugs in their cavities [16]. Inclusion complex formation is usually associated with increase in aqueous solubility, stability and bioavailability of the drug guest. CDs additionally are known to facilitate the crossing of biological barriers of various bioactive molecules [17].

To enhance the aqueous solubility of porphyrins, cyclodextrin–porphyrin conjugates have been prepared, either as enzyme mimics [18, 19] or as PSs [20, 21] and only in few examples as drug delivery systems [22]. Whereas multisubstituted CD–porphyrin conjugates are reportedly water soluble, mono-CD-porphyrin conjugates display limited aqueous solubility [21, 23].

The aim of the present work was to demonstrate successful conjugation of PpIX to β-CD via a simple and reproducible procedure affording a bimodal product, PpIX + βCD, endowed with satisfactory water solubility, ability to host drugs and transport them through cell membranes, and capacity to target subcellular compartments. The drug selected herein as a paradigm of PpIX + βCD-mediated intracellular transport was N-desmethyltamoxifen (NDMTAM), a chemotherapeutic agent mainly applicable to estrogen receptor–positive breast cancer.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

NDMTAM hydrochloride, 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT), paraformaldehyde, dimethyl sulfoxide (DMSO), O-(7-azabenzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate (HATU), dry dimethylformamide (DMF) and the dialysis membranes (cellulose tubing, benzoylated, MWCO 1.2 kD) were purchased from Sigma-Aldrich. Deuterium oxide (D2O), deuterated dimethylsulfoxide (DMSO-d6), deuterated chloroform (CDCl3) and deuterated methanol (CD3OD) were purchased from Deutero GmbH. βCD was a product of CycloLab. βCD and derivatives were dried by heating at 70°C under vacuum for 20 h before reaction. N,N-diisopropylethylamine (DIPEA) was used without purification. PpIX was purchased from Porphyrin Systems (Appen, Germany). 1D and 2D NMR spectra were performed on a Bruker Avance 500 MHz spectrometer. RPMI 1640 without phenol red, fetal bovine serum (FBS), penicillin/streptomycin, l-glutamine, PBS, Mitotracker® Green FM, LysoTracker® Green DND-26 and trypsin/versene were purchased from Invitrogen Ltd., (Paisley, UK).

Fluorescein-N-NDMTAM (NDMTAM-FITC) was prepared as described previously [24]. Mono(6-p-toluenesulfonyl)-β-cyclodextrin (βCD) [25] was converted into mono(6-azido-6-deoxy)-βCD [26] and subsequently to mono(6-amino-6-deoxy)-βCD [27] according to previously published methods.

Synthesis

PpIX (PpIX, 0.020 g, 0.0355 mmol) was dissolved in dry DMF (1 mL) and allowed to stir at 0°C for 30 min under an argon atmosphere. HATU (0.103 g, 0.272 mmol) was slowly added at 0°C and stirred for 1 h. The mixture was allowed to attain room temperature and then DIPEA (33 μL, 0.191 mmol) was added, followed by gradual addition of mono(6-amino-6-deoxy)-βCD (0.040 mg, 0.0355 mmol) during 1 h with more DMF (1 mL). The reaction mixture was allowed to stir at 20–23°C for 3 days while protected from ambient light. Subsequently the solvent was evaporated under vacuum to dryness and washed exhaustively with methanol to remove unreacted PpIX (checked by TLC on aluminum-backed silica gel plates eluted with isopropanol/ethyl acetate/water/ammonia, 5:3:3:1, v/v, visualized under a UV lamp and subsequent heating following spraying with 10% sulfuric acid in ethanol). The material was then dialyzed against deionized water ([2] days) and then chromatographed using size exclusion chromatography (Sephadex G50) and distilled water as eluent. The first fraction collected gave a dark purple solid (30%) that contained PpIX-CD (90%) and PpIX-2CD (10%) in 9:1 ratio, as indicated by mass spectra (vide infra). This solid, named PpIX + βCD, was used for the cell experiments described subsequently. 1H ΝΜR (DMSO, 298 K, 500 ΜHz): δ 10.40-10.31 (4H, H5, H10, H15, H20), 8.55 (m, 2H, H3a, H8a), 6.49-6.24 (2d, 4H, H3b, H8b), 5.71 (br s, OH2, OH3), 4.84 (br s, 9H, CD-H1), 4.60-4,17 (m, OH6, 13a, 17a), 4.03-2.46 (2 br s, CD-H6, 12a, 18a, 2a, 7a, CD-H3, -H5, -H6A), 3.31 (br s, CD-H4, CD-H2), 2.81 (br s, 13b, 17b), −3.85 (m, 2H, NH) ppm. 13C NMR (DMSO, 298 K, 125 ΜHz): δ 129.44 (C3a, C8a), 120.70 (C3b, C8b), 100.8 (CD-C1), 96.21 [5, 10, 15, 20], 80.22 (CD-C4), 72.34, 71.46, 71.84 (CD-C3, CD-C5, CD-C2), 59.29 (CD-C6), 50.74 (CD-C6A), 36.2 (C13b, C17b), 21.99 (C13a, C17a), 12.32, 10.83 (C2a, C7a, C12a, C18a) ppm. UV/Vis (DMSO): 405, 504, 540, 571, 630 nm; emission spectrum (λex = 405 nm): 632, 680, 698 nm; UV/Vis (H2O): 388, 510, 545, 575 nm; emission spectrum (λex = 388nm): 626, 691 nm. Mass Spectrum (MALDI-TOF) m/q: Found, 1660.7 (100%, base peak, [M + Η]+), 2794.2 (11%, [PpIX-2CD + H]+). Calculated for C76H101N5O36 (PpIX-CD), 1659.62; calculated for C118H172N6O69 (PpIX-2CD): 2793.01.

Binding studies of NDMTAM. HCl with parent βCD by NMR spectroscopy

Initially, the stoichiometry of the complex βCD/NDMTAM.HCl was determined with the continuous variation plots method by mixing variable concentrations (0–2 mm) of the substances to a constant final volume and recording of the 1H NMR spectral changes of βCD cavity protons. Plots of ΔδΗ3Χ[βCD] vs mole fraction (Figure S8) gave a stoichiometry of 1:1. Furthermore, titration of a 2 mm solution of βCD in deuterium oxide with portions of solid NDMTAM.HCl was performed; following 10–15 min sonication, the 1H NMR spectra were acquired and the chemical shift changes (strong shielding) of βCD cavity protons, H3 and H5, were recorded vs molar ratio [NDMTAM]/[βCD]. After addition of ca 0.5 equiv. solid NDMTAM.HCl the onset of cloudiness could be observed. The ratio of solubilized [NDMTAM.HCl]/[βCD] was obtained from integration. The data (Δδ Η3 vs molar ratio) were plotted and fitted to an equation suitable for 1:1 stoichiometry [28] using GraphPad Prism from which the binding constant was estimated.

Cell culture

Cells used in this study were the DU145 human prostate carcinoma and MCF7 human breast adenocarcinoma cell lines. All cells were grown at 37°C in a humidified, 5% CO2 atmosphere, in RPMI 1640 media supplemented with 10% FBS and penicillin (50 IU mL−1)/streptomycin (50 μg mL−1). Cells were inoculated into 96-well plates (2 × 104 cells per 100 μL media per well) 24 h prior to phototoxicity experiments or onto coverslips housed in 35 mm Petri dishes (1 × 105 cells per 2 mL media per dish) 24 h prior to confocal microscopy imaging. For cell experiments 15 μL of 10 mm PpIX + βCD (stock) was initially diluted into 360 μL of PBS (400 μm PpIX + βCD—4% DMSO). This was further diluted with cell media to a final PpIX + βCD concentration of 7 μm (or 20 μm for imaging) resulting in DMSO content added to cells <0.3%. Accordingly, 15 μL of 10 mm PpIX in DMSO was diluted with cell media down to 7 μm (or 20 μm for imaging), again yielding a final DMSO concentration <0.3%.

Confocal microscopy

Cells seeded onto coverslips were incubated with: (1) PpIX (stock in DMSO, final concentration 20 μm) or PpIX + βCD (stock in PBS-4% DMSO, final concentration 20 μm) for 3 h. Both the parent (PpIX) and the daughter (PpIX + βCD) molecules were ultrasonicated for 1 h and then syringe filtered (0.22 μm) before the administration to cells. Half an hour prior to imaging, the subcellular organelle probes (Mitotracker® Green FM 200 nm, Lysotracker® Green DND-26 200 nm) were added to the appropriate cell groups. (2) PpIX + βCD complexed with NDMTAM-FITC [24]: A 2 mm solution of PpIX + βCD (4% DMSO—96% PBS, [stock in PBS-4% DMSO]) and NDMTAM-FITC at equimolar ratio was left under stirring overnight. Upon stirring the solution lost all turbidity and became clear, indicating that NDMTAM-FITC had been entirely taken up by the PpIX + βCD host. The solution immediately before being introduced to cells was subjected to ultrasonication for 1 h and cells were subsequently incubated for 3 h at a final PpIX + βCD concentration 20 μm.

In all cases, the coverslips were next washed twice with PBS and placed over the 63× oil immersion quartz objective (NA 1.3) of a Biorad MRC 1024 scanning confocal microscope, in physiological saline. Intracellular PpIX fluorescence was excited using the 568 nm line of an argon–krypton laser (3–10% of total laser power) whereas the internalized probes or NDMTAM-FITC were excited by the 488 nm line of the same laser (3% of total laser power). PpIX fluorescence was collected with the use of a long-pass filter at ≥585 nm, whereas FITC-type fluorescence was collected through a bandpass filter centered at 522 (±35) nm. During image acquisition a Kalman level 3 (three iterations per image) smoothing routine was applied each time to eliminate spurious signal. The contribution of cell autofluorescence was checked each time in untreated cells and was, in all cases, found to be negligible.

Light irradiation protocol

In 96-well plates the 16 (4 × 4) central wells were used for irradiation. Cells were incubated half with PpIX and half with PpIX + βCD (7 μm) in complete media for 3 h, then immediately prior to irradiation washed twice with PBS and complete media (without phenol red) restored. Light-positive controls without PpIX or PpIX + βCD, as well as dark PpIX and PpIX + βCD controls were included in each experiment. Light was provided by a halogen light source (50 W) through a Fresnel lens. The light transversed a long-pass filter (Schott RG610; Schott UK Ltd., Stafford, UK) and irradiated the cells through the plate underside. RG610 has a cut off at 610 nm, allowing irradiation above 610 nm i.e. at the 635 nm Q-band of PpIX; the average light fluence rate at the cell level was maintained at 15 mW cm−2 throughout the experiments as determined by a LI-185A Quantum /Radiometer/ Photometer (LI-COR Instruments, Lincoln, NE). The light dose was each time determined by irradiation duration. The light source power jitter was in all cases less than 5%.

Cell viability assessment

Twenty-four hours following irradiation, the mitochondrial redox function (relating to cytotoxicity at that time frame) of all cell groups, including nonirradiated controls was assessed by the MTT assay. This was carried out by adding 100 μL of complete media containing 1 mg mL−1 MTT to cells and incubating at 37°C in a 5% CO2 humidified atmosphere for 2 h. MTT media were subsequently removed from all cells and the resulting formazan was solubilized with 100 μL DMSO per well. The plates were then shaken for 10 min at 100 rpm in a Stuart SI500 orbital shaker, and the endpoint absorbance measurement at 562 nm was performed in an Infinite M200 plate reader (Tecan group Ltd., Männedorf, Switzerland). Blank values measured in wells with DMSO and no cells were in all cases subtracted.

Synergistic effect of PpIX + βCD and tamoxifen citrate in vitro

Optimal PpIX complexation with tamoxifen citrate (TAM-CIT, prepared from TAM; Sigma) was performed in an analogous fashion to NDMTAM. Briefly, 15 μL of 10 mm PpIX + βCD (stock) was initially diluted into 360 μL of cell media (400 μm PpIX + βCD—4% DMSO). Equimolar TAM-CIT was added and the solution was left stirring overnight. Upon complete complexation the resulting solution was further diluted with cell media to a final PpIX + βCD concentration of 7 μm and added to the appropriate MCF7 cell groups for 3 h. Half the cells were incubated with 7 μm empty PpIX + βCD vector prepared in the same experimental conditions as the complex. The cells were then irradiated as per the irradiation regimen described above at 4 and 8 J cm−2. The toxicity was assessed in all cell groups including dark and light controls at 48 h following irradiation by standard MTT assays.

Statistics

All experiments were repeated at least three times independently. The graph error bars represent one standard deviation for at least four independent values.

Results and Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

The PpIX to βCD conjugation, performed under HATU-mediated coupling conditions for amide bond formation (Scheme Scheme 1), reproducibly afforded a dark purple solid product; conversely, traditional N-hydroxybenzotriazole and dicyclohexylcarbodiime coupling reagents, instead of HATU, could not facilitate conjugation. The product was subjected to size exclusion chromatography to yield the water soluble fraction that had both porphyrin and cyclodextrin content, according to thin layer chromatography.

The 1H NMR spectrum in DMSO-d6 (Figs. S1, S2) allowed satisfactory characterization of the product: the carboxyl group protons of the parent PpIX (Figure S3) were now absent, suggesting that reaction had occurred at the carboxyl sites. Parts of the PpIX signals due to the exocyclic double bonds (ca 6.5 and 8.6 ppm) were observed along with those of the meso-positions (ca 10.3 ppm) and of the pyrrole endocyclic NH (ca −3.8 ppm). On the other hand, the anomeric proton signals of βCD (ca 5.0 ppm) were reasonably separated from those of the βCD primary (ca 4.5 ppm) and secondary (ca 5.8 ppm) hydroxyl groups. The nearly 1 ppm spread of the signals due to the CD-H6,6′, revealed by the HSQC spectrum (Figure S2), indicated that the cyclodextrin macrocycle had been modified at its primary side thus its molecular symmetry had been reduced. Integration of the porphyrin and CD regions gave a ratio PpIX/βCD ≈1.3. The IR spectrum of the product verified coexistence of the two macrocyclic molecules and the shift of the PpIX carbonyl stretching vibrations from ca 1695– ca 1730 cm−1 (Figure S4). The MALDI-TOF mass spectrum confirmed PpIX conjugation to βCD by a two-point attachment, yielding the rigid amido ester of PpIX-βCD (Scheme Scheme 1) (base peak, 100%) but also revealed presence of a doubly CD-bearing product, PpIX-2βCD (11%). The ratio of the two conjugates, as inferred from the MS spectrum was ca 9:1 while the corresponding value obtained from 1H NMR integration is ca 8:2. It is interesting to note that the proximity of the hydroxyl groups of the βCD primary side to the point of initial amide bond formation probably facilitated the second linking via the ester bond to the same βCD, thus favoring the intramolecular over the intermolecular reaction, i.e. attachment of a second βCD to PpIX. As a result, the PpIX-2βCD conjugate is the minor component. The product will be referred as PpIX + βCD in the following sections.

The 1H NMR spectrum in D2O exhibited very broad signals indicating extensive aggregation in aqueous solution presumably due to the high concentration used (mm), required by the method (Figure S5). Aqueous solubility was, however, achieved at biologically relevant concentrations (μm range) as UV–Vis and fluorescence spectra of PpIX + βCD were successfully obtained in both aqueous and DMSO solutions, while PpIX alone was soluble only in DMSO. The absorbance intensity of the product in H2O (Fig. 1A) despite the nearly double concentration was half of that in DMSO (Fig. 1B).

image

Figure 1. Absorption spectra of (A) PpIX + βCD in Η2Ο (1.7 × 10−6 m) and (B) in DMSO (black solid trace, 10−6 m) in comparison with PpIX alone (red dashed trace, 10−6 m). Fluorescence spectra of (C, D) PpIX + βCD in water (1.7 × 10−6 m), λex = 388 and 510 nm respectively; (E, F) of PpIX + βCD in DMSO (black solid trace, 10−6 m), λex = 406 and 505 nm, respectively, in comparison with PpIX alone (red dashed trace, 10−6 m), λex = 407 and 507 nm respectively.

Download figure to PowerPoint

This suggested that some aggregation occurred even at the low μm range. In addition, broadening of the Soret band in Η2Ο relative to DMSO also indicated a degree of aggregation. Consequently, the emission of the product in Η2Ο (Fig. 1C,D) was nearly five times lower than in DMSO (Fig. 1E,F) On the other hand, comparison of the absorption spectrum of PpIX + βCD in DMSO with that of parent PpIX (Fig. 1B) revealed a small but noticeable increase in the Soret bandwidth in the product even though the extinction coefficient was similar. Finally, comparison of the emission spectra of PpIX + βCD with that of parent PpIX in DMSO (Fig. 1E,F) revealed a nearly 20% quenching of the fluorescence at ca 632 nm in the product, and the appearance of enhanced intensity of the band at ca 680 nm, absent in the emission spectra of PpIX + βCD in water.

The product structure (both of major and minor components) suggests that it retains the amphiphilic character of the parent PpIX: the CD region is apparently a heavily hydrated part, whereas the porphyrin region with its vinyl groups is highly lipophilic. This feature evidently promotes some residual aggregation of the conjugate in water despite the aqueous solubility. DLS experiments at 20 μm concentration in RPMI 1640 medium with 4% DMSO (as used for cell experiments, vide infra) indicated formation of nanoparticles sized ca 6.5 nm on average, corresponding to aggregation of a few molecules based on molecular size (estimated ca 2.5 nm for PpIX-βCD and ca 2.9 nm for PpIX-2βCD, Chem3D model measurements). The presence of larger aggregates in very small amounts could be observed in the size vs intensity diagrams (Figure S6). The inclusion capability of the CD moiety in the product, that is, the accessibility of the cavity by a potential guest molecule was subsequently investigated.

N-Desmethyltamoxifen, one of the primary active metabolites of tamoxifen, was chosen as a suitable and strategic guest to be solubilized by the conjugate PpIX + βCD in water. When one equivalent of NDMTAM.HCl was added into a 2.5 mm solution of the conjugate in D2O-4% DMSO-d6 gradual solubilization was observed suggesting possible formation of a water soluble inclusion complex. Indeed the 1Η NMR spectrum of the resulting solution showed broad peaks for PpIX + βCD and NDMTAM alike, indicating that NDMTAM had been incorporated in the aggregate formation revealing a final envelope of peaks instead of discrete resonances (Figure S7). The process was indirectly quantified by studying the complex formation between the parent βCD and NDMTAM.HCl in water. The stoichiometry of the complex, as determined by continuous variation plots (Figure S8A) was found to be clearly 1:1. On the other hand, 1H NMR titrations of a βCD solution with portions of solid NDMTAM.HCl confirmed inclusion complex formation in the fast exchange regime in the NMR time scale by the very strong shielding (Δδ = δfree − δcomplx) of only the cavity protons (Figure S8B) suggesting strong binding. The onset of slight cloudiness observed after addition of ca 0.5 equivalents of NDMTAM.HCl indicated that the solubility of the resulting complex should be below 2 mm. Treatment of the Δδ observed with a suitable equation for 1:1 complexation in fast exchange [28] gave an estimate of the binding constant. The best fit values were found around 7000 (±1500) m−1 (R2 = 0.98, evidently due to use of unavoidably approximate concentrations), nevertheless indicating strong binding indeed. Furthermore, 2D ROESY spectra revealed intense intermolecular dipolar interactions between the βCD cavity protons with all NDMTAM phenyl groups in the aqueous environment (Figure S9). NDMTAM is a tritopic guest molecule, thus on statistical grounds it is expected to display enhanced binding to βCD: the data above confirm that the participation of all aromatic rings in the inclusion process results in binding considerably higher of that of a typical phenyl-monosubstituted guest [16] and indicate that even in the low concentrations used for the cell experiments (vide infra) NTMTAM could be efficiently transported by the prepared PpIX + βCD conjugate. Few data on tamoxifen and CDs are found in the literature: complexation between tamoxifen citrate and βCD or 2,3-di-O-hexanoyl-βCD [29] was previously studied with solid-state methods. The complex with the amphiphilic 2,3-di-O-hexanoyl-βCD was found to display augmented anticancer activity tested on the MCF7 breast cancer cells. Elsewhere, the pharmacokinetics of tamoxifen and tamoxifen citrate following oral or intravenous administration were improved upon formulation with hydroxubutenyl-βCD [30].

This study showed that the βCD cavity can host the phenyl groups of tamoxifen and thus the PpIX + βCD conjugate could efficaciously solubilize and transport NDMTAM via a similar action.

Cell experiments

PpIX and PpIX + βCD subcellular localization in confocal microscopy

As incubation was in all cases carried out in serum-containing media, there was a concern that partitioning of the compounds between the cells and media could become a limiting factor in the cellular uptake, especially in the case of PpIX + βCD that is profoundly more water soluble than PpIX, leading to a lower intracellular PpIX + βCD uptake. Both PpIX and PpIX + βCD were, however, internalized by both cell lines quite efficiently, without apparent quantitative discrepancies, judging by their corresponding fluorescence intensities (Figs. 2 and 3). This can be ascribed to the amphiphilic nature of the PpIX + βCD product that preserves some hydrophobicity in part of the molecule, although the actual structure of the in situ formed aggregates is not known. The subcellular distributions of PpIX and PpIX + βCD are shown in Figs. 2 and 3.

image

Figure 2. Representative confocal microscopy images following DU145 and MCF7 cell 3 h incubation with 20 μm concentration of PpIX or PpIX + βCD. The cells were in all cases coincubated with Mitotracker® Green FM (200 nm) for 30 min. PpIX fluorescence was excited at 568 nm and collected at ≥585 nm, whereas Mitotracker® Green FM fluorescence was excited at 488 nm and collected at 522 (±35) nm. In the overlay images yellow denotes good green and red fluorescence colocalization.

Download figure to PowerPoint

image

Figure 3. Representative confocal microscopy images following DU145 and MCF7 cell 3 h incubation with 20 μm concentration of PpIX or PpIX + βCD. The cells were in all cases coincubated with LysoTracker® Green DND-26 (200 nm) for 30 min. PpIX fluorescence was excited at 568 nm and collected at ≥585 nm, whereas LysoTracker® Green DND-26 fluorescence was excited at 488 nm and collected at 522 (±35) nm. In the overlay images yellow denotes good green and red fluorescence colocalization.

Download figure to PowerPoint

In all cases PpIX-type (red) fluorescence partly localized in cell membranes and cell synapses. Moreover, a rather punctuate cytosolic component is evident, which upon prolonged and more intense irradiation became diffuse presumably upon photoinduced changes to organelle structures and concomitant dye release; in that context imaging durations as brief as possible and low laser intensities were kept in all occasions. No notable nuclear fluorescence was observed for PpIX or for PpIX + βCD. These results are in good accordance with our previous PpIX findings [10]. Coincubation with Mitotracker® Green FM (Fig. 2) revealed a significant mitochondrial accumulation for both the parent and daughter species and in general there was no notable change of that trend following PpIX conjugation with βCD. Although mitochondrial localization was evident for both compounds in accordance with previous reports on PpIX [31], we could not verify a notable lysosomal component expected upon exogenous PpIX administration [31] for either PpIX or PpIX + βCD and in either of the cell lines studied (Fig. 3). In fact, in all the occasions of Fig. 3 the LysoTracker® Green DND-26 green fluorescence was quite limited in comparison with the PpIX-associated red fluorescence. In these conditions it is very difficult and thus precarious to assign lysosomal localization; however, if there were any, it would not be extensive and certainly not comparable to the corresponding plasmalemmal or mitochondrial deposits.

PpIX and PpIX + β-CD photocytotoxicity

The dark toxicity of PpIX and PpIX + βCD is shown in Fig. 4. It can be seen that cell incubation with 7 μm PpIX for 3 h confers a residual toxicity of around 25%, whereas the corresponding dark toxicity of PpIX + βCD is 15%. Although these values are not considerably diverse within experimental errors, there is a certain trend of dark toxicity reduction for PpIX + βCD in both lines. Heuristically, from our extended experience from handling the compounds we believe that the PpIX dark toxicity is mainly derived from the formation of aggregates. In the case of PpIX + βCD the possibility of aggregate formation (especially π-stacking) is largely diminished due to steric hindrance by the βCD. In any case the selected concentrations of both compounds and especially PpIX + βCD exhibit minimal dark toxicity. In our previous study on exogenously administered PpIX [10], we discovered a higher concentration tolerance (20 μm); however, it has to be noted that we then used the dimethyl ester instead of the free acid of PpIX used herein, while also the cell line employed was different (PAM 212 keratinocytes).

image

Figure 4. Toxicity and phototoxicity of PpIX and PpIX + βCD: (A). Dark toxicity 24 h following 3 h cell incubation with 7 μm PpIX and PpIX + βCD. (B). Phototoxicity 24 h following 3 h cell incubation with 7 μm PpIX and PpIX + βCD and irradiation through a Schott RG610 long-pass filter. Spheres and open circles: MCF7 cells; open and filled squares: DU145 cells. Solid lines: PpIX; dashed lines: PpIX + βCD.

Download figure to PowerPoint

In Fig. 4B, the photocytotoxicities of PpIX and PpIX + βCD are shown for both cell lines studied. Both compounds were found to be phototoxic, albeit perhaps to a different extend in the two cell lines. In that context the LD50 values for PpIX were 10 and 7 J cm−2 for MCF7 and DU145 cells, respectively, whereas the corresponding values for PpIX + βCD were 10 and 6 J cm−2. It can also be concluded from Fig. 4B that essentially there was no discrepancy between the photocytotoxic capacities of PpIX and PpIX + βCD; in fact, if there is a slight divergence, this is in favor of PpIX + βCD, i.e. PpIX + βCD is less chemotoxic and slightly more phototoxic. These results show that PpIX conjugation to βCD had no adverse affects on its phototoxicity while it reduced its chemical toxicity and provided the possibility of drug transportation in its βCD cavity. It additionally has to be noted that at 15 J cm−2, the cell viability was close to zero in all cases and certainly below 10%. In our previous study of exogenously administered PpIX to PAM 212 keratinocytes [10] the photodynamic action was more profound again in that case, the incubation concentration was considerably higher (20 μm), the incubation time longer (5 h) but also different cell type and PpIX moiety (dimethyl ester) was employed.

Cell internalization of the PpIX + βCD/NDMTAM-FITC complex

The main incentive for the implementation of PpIX + βCD was the possibility of inclusion of a guest molecule in the CD cavity to endow the parent PpIX with a bimodal action. In this context we complexed FITC-labeled NDMTAM [24] (NDMTAM-FITC) with PpIX + βCD, and incubated our cell lines with the water soluble complex to study the intracellular fate of the two moieties. The results of this study are shown in the representative images of Fig. 5.

image

Figure 5. Representative confocal microscopy images following DU145 and MCF7 cell 3 h incubation with 20 μμ PpIX + βCD in complex with NDΤΑΜ-FITC: PpIX + βCD fluorescence was excited at 568 nm and collected at ≥585 nm, whereas NDΤΑΜ-FITC fluorescence was excited at 488 nm and collected at 522 (±35) nm. In the overlay images yellow denotes good green and red fluorescence colocalization.

Download figure to PowerPoint

In both cell lines the two fluorescent species are shown to largely colocalize subcellularly. The minor discrepancies in unmatched red fluorescence especially observed in cell membranes could be attributed to empty PpIX + βCD vector (Figs. 2 and 3). Conversely, all NDMTAM-FITC carried by PpIX + βCD seems not to profoundly localize at cell membranes, perhaps suggesting that the guest molecule could possibly determine the intracellular fate of the complex. These results nevertheless, unambiguously demonstrate that the robustly bound constituents of the complex (as shown by the ROESY spectra, Figure S9) do not take separate subcellular routes; in that context PpIX + βCD apart from its photodynamic capacity is an efficient intracellular transporter of suitable bioactive moieties.

Bimodal photo- and chemotoxic action of PpIX + βCD complex with TAM-CIT in MCF7 cells

The synergistic effect of PpIX + βCD phototoxicity and TAM-CIT guest chemotoxicity at 48 h following irradiation is shown in Fig. 6. The dark toxicity of PpIX + βCD is ca 86% as compared with media controls which is consistent with the dark toxicity values in Fig. 4A. The dark toxicity of the complex was found to be ca 50% which is a combination of the dark toxicity of PpIX + βCD and the chemotoxic effect of TAM-CIT. Irradiation of the PpIX + βCD only MCF7 cell groups with 4 J cm−2 red light conferred a 30% phototoxicity, however, irradiation of the complex with the same light dose yielded a 70% synergistic cytotoxicity. The corresponding values for 8 J cm−2 irradiation were 67% phototoxicity and 85% bimodal toxicity. The above results clearly demonstrate a profound synergy between the phototoxicity and the adjuvant chemotoxicity of the PpIX + βCD-TAM complex, substantially enhancing the PDT effect of the parent PpIX or the empty PpIX + βCD molecule.

image

Figure 6. Bimodal action of PpIX+βCD complexed with TAM-CIT in MCF7 cells. Toxicity 48 h following 3 h cell incubation with 7 μm PpIX + βCD and PpIX + βCD complexed with TAM-CIT and irradiation through a Schott RG610 long-pass filter at 4 and 8 J cm−2.

Download figure to PowerPoint

image

Scheme Scheme 1. Preparation of the PpIX + βCD product.

Download figure to PowerPoint

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

A novel PpIX + βCD product has been developed with a reproducible procedure. The product is a bicomponent composition, with one (major) and two (minor) βCD cavities connected to the PpIX carboxyl groups. The product was shown to display good aqueous solubility and the ability to traverse cell membranes, localizing primarily in the mitochondria and also transporting guest molecules (NDMTAM) intracellularly. PpIX + βCD displayed satisfactory dark toxicity and phototoxicity equivalent of that of the parent PpIX. The demonstrated product properties along with its ease and reproducibility of preparation and potential for up scaling represent an important bimodal system combining delivery of a chemotherapeutic agent with a well-documented photodynamic action. Furthermore, the multimodal action of PpIX + βCD was demonstrated: While it retains the ptototoxic profile of PpIX and its fluorescence for imaging purposes, PpIX + βCD can efficiently transport chemotoxins (or other drugs) into cells and confer cell death through a synergy of photo- and chemotoxicity.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Funding by the Marie Curie Initial Training Networks (FP7-People-ITN-2008 CYCLON), Project no. 237962 “CYCLON” is gratefully acknowledged. A scholarship to C. A. by NCSR “Demokritos” is gratefully acknowledged. We also thank Mr. A. R. Goncalves of our group for preparing tamoxifen citrate and NDMTAM.HCl. Istituto di Ricerche Chimiche e Biochimiche “G. Ronzoni” Milano, Italy, is also thanked for the MALDI-TOF measurements.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    Agostinis, P., K. Berg, K. A. Cengel, T. H. Foster, A. W. Girotti, S. O. Gollnick, S. M. Hahn, M. R. Hamblin, A. Juzeniene, D. Kessel, M. Korbelik, J. Moan, P. Mroz, D. Nowis, J. Piette, B. C. Wilson and J. Golab (2011) Photodynamic therapy of cancer: an update. CA Cancer J. Clin. 61, 250281.
  • 2
    MacRobert, A. J. and T. Theodossiou (2005) Photodynamic therapy of cancer. In Encyclopedia of Modern Optics, Vol. 1. (Edited by R. D. Guenther), pp. 5362. Elsevier, Amsterdam.
  • 3
    Dolmans, D. E., D. Fukumura and R. K. Jain (2003) Photodynamic therapy for cancer. Nat. Rev. Cancer 3, 380387.
  • 4
    Dougherty, T. J., C. J. Gomer, B. W. Henderson, G. Jori, D. Kessel, M. Korbelik, J. Moan and Q. Peng (1998) Photodynamic therapy. J. Natl. Cancer Inst. 90, 889905.
  • 5
    Hunter, C. A. and M. K. J. Sanders (1990) The nature of π-π interactions. J. Am. Chem. Soc. 112, 55255534.
  • 6
    Siggel, U., U. Bindig, C. Endisch, T. Komatsu, E. Tsuchida, E. Voigt and J.-H. Fuhrhop (1996) Photophysical and photochemical properties of porphyrin aggregates. Ber. Bunsenges. Phys. Chem. 100, 20702075.
  • 7
    Kano, K., K. Fukuda, H. Wakami, R. Nishiyabu and R. F. Pasternak (2000) Factors influencing self-aggregation tendencies of cationic porphyrins in aqueous solution. J. Am. Chem. Soc. 122, 74947502.
  • 8
    Gaullier, J. M., K. Berg, Q. Peng, H. Anholt, P. K. Selbo, L. W. Ma and J. Moan (1997) Use of 5-aminolevulinic acid esters to improve photodynamic therapy on cells in culture. Cancer Res. 57, 14811486.
  • 9
    Peng, Q., T. Warloe, K. Berg, J. Moan, M. Kongshaoug, K. E. Giercksky and J. M. Nesland (1997) 5-Aminolevulinic acid-based photodynamic therapy. Clinical research and future challenges. Cancer Res. 79, 22822308.
  • 10
    Theodossiou, T. and A. J. MacRobert (2002) Comparison of the photodynamic effect of exogenous photoprotoporphyrin and protoporphyrin IX on PAM 212 murine keratinocytes. Photochem. Photobiol. 76, 530537.
  • 11
    Niamien Konan, Y., R. Gurny and E. Allemann (2002) State of the art in the delivery of photosensitizers for photodynamic therapy. J. Photochem. Photobiol., B 66, 89106.
  • 12
    Allison, R. R. and C. H. Sibata (2010) Oncologic photodynamic therapy photosensitizers: a clinical review. Photodiagn. Photodyn. Ther. 7, 6175.
  • 13
    Monsu Scolaro, L., M. Castriciano, A. Romeo, S. Patane, E. Cefalı and M. Allegrini (2002) Aggregation behavior of protoporphyrin IX in aqueous solutions: clear evidence of vesicle formation. J. Phys. Chem. B 106, 24532459.
  • 14
    Caughey, W. S. and J. A. Ibers (1977) Crystal and molecular structure of the free base porphyrin, protoporphyrin IX dimethyl ester. J. Am. Chem. Soc. 99, 66396645.
  • 15
    Szejtli, J. (1988) Cyclodextrin Technology. Kluwer Academic, Dordreht.
  • 16
    Rekharsky, M. V. and Y. Inoue (1998) Complexation thermodynamics of cyclodextrins. Chem. Rev. 98, 18751917.
  • 17
    Martin Del Valle, E. M. M. (2004) Cyclodextrins and their uses: a review. Proc. Biochem. 39, 10331046.
  • 18
    Breslow, R., X. Zhang, R. Xu, M. Maletic and R. Merger (1996) Selective catalytic oxidation of substrates that bind to metalloporphyrin enzyme mimics carrying two or four cyclodextrin groups and related metalosalens. J. Am. Chem. Soc. 118, 1167811679.
  • 19
    Woggon, W.-D., A. Schlatter and H. Wang (2008) b-cyclodextrin-linked Ru complexes for oxidations and reductions. Adv. Inorg. Chem. 60, 3158.
  • 20
    Nuno Silva, J., A. M. G. Silva, J. P. Tome, A. O. Ribeiro, M. R. M. Domingues, J. A. S. Cavaleiro, A. M. S. Silva, G. M. P. M. S. Neves, A. C. Tome, O. A. Osvaldo, A. Serra, F. Bosca, P. Filipe, R. Santuse and P. Morliere (2008) Photophysical properties of a photocytotoxic fluorinated chlorin conjugated to four b-cyclodextrins. Photochem. Photobiol. Sci. 7, 834843.
  • 21
    Puglisi, A., R. Purrello, E. Rizzarelli, S. Sortino and G. Vecchio (2007) Spectroscopic and self-association behavior of a porphyrin-b-cyclodextrin conjugate. New J. Chem. 31, 14991506.
  • 22
    Kralova, J., Z. Kejik, T. Briza, P. Poukova, A. Kral, P. Martasek and V. Kral (2010) Porphyrin-cyclodextrin conjugates as a nanosystem for versatile drug delivery and multimodal cancer therapy. J. Med. Chem. 53, 128138.
  • 23
    Kiba, T., H. Suzuki, K. Hosokawa, H. Kobayashi, S. Baba, T. Kakuchi and S.-I. Sato (2009) Supramolecular J-aggregate assembly of a covalently linked zinc porphyrin-b-cyclodextrin conjugate in a water/ethanol binary mixture. J. Phys. Chem. B 113, 1156011563.
  • 24
    Theodossiou, T. A., K. Yannakopoulou, C. Aggelidou and J. S. Hothersall (2012) Tamoxifen subcellular localization; observation of cell-specific cytotoxicity enhancement by inhibition of mitochondrial ETC complexes I and III. Photochem. Photobiol. 88, 10161022.
  • 25
    Brady, B., N. Lynam, T. O'Sullivan, C. Ahern and R. Darcy (2000) Preparation of 6-monotosyl-β-cyclodextrin. Org. Synth. 77, 220223.
  • 26
    Petter, R. C., J. S. Salek, C. T. Sikorski, G. Kumaravel and F.-T. Lin (1990) Cooperative binding by agregated mono-6-(alkylamino)-β-cyclodextrins. J. Am. Chem. Soc. 112, 38603868.
  • 27
    CRC handbook of Chemistry and Physics (1988). CRC Press, Inc., Boca Raton, FL.
  • 28
    Maffeo, D., L. Leondiadis, I. M. Mavridis and K. Yannakopoulou K (2006) Positive effect of natural and negatively charged cyclodextrins on the stabilization of penicillins towards β-lactamase degradation due to inclusion and external guest-host association. An NMR and MS study. Org. Biomol. Chem. 4(7), 12971304.
  • 29
    Bilensoy, E., D. L. M. Sen and A. Hincal (2007) Complexation behavior of antiestrogen drug tamoxifen citrate with natural and modified b-cyclodextrins. J. Incl. Phen. Macrocycl. Chem. 57, 651655.
  • 30
    Buchanan, C. M., N. L. Buchanan, K. J. Edgar, J. L. Little, M. O. Malcolm, K. M. Ruble, V. J. Wacher and M. F. Wempe (2007) Pharmacokinetics of tamoxifen after intravenous and oral dosing of tamoxifen-hydroxybutenyl-b-cyclodextrin formulations. J. Pharm. Sci. 96, 644660.
  • 31
    Sandberg, S. and I. Romslo (1981) Phototoxicity of protoporphyrin as related to its subcellular localization in mice livers after short-term feeding with griseofulvin. Biochem. J. 198, 6774.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
php12127-sup-0001-FigS1-S9.docWord document565K

Figure S1. 1Η ΝΜR (500 MHz, DMSO-d6, 25°C) of PpIX + βCD product.

Figure S2. 2D HSQC NMR spectrum of PpIX + βCD product (DMSO-d6, 25°C).

Figure S3. 1H NMR spectra of PpIX in DMSO-d6, 25°C.

Figure S4. Overlay of IR spectra: PpIX (blue trace) and PpIX + βCD (red trace).

Figure S5. 1Η ΝΜR spectrum of PpIX + βCD in D2O.

Figure S6. DLS diagrams showing particle size distribution in RPMI 1640 solution-4% DMSO (20 μm, 22°C) of PpIX + βCD (a) before and (b) after filtration through a 0.22 μm disk.

Figure S7. Overlay of 1Η ΝΜR NMR spectra of PpIX + βCD/NDMTAM.HCl in D2O, PpIX + βCD in D2O and NDMTAM.HCl in DMSO-d6.

Figure S8. A. Continuous variation (Job) plots of NDTAM.HCl/βCD complex in D2O. B. Dual display of 1H NMR spectra of βCD alone and in the presence of NDMTAM.HCl.

Figure S9. 2D ROESY spectra NDMTAM.HCl/βCD (1:1) complex in D2O.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.