1.

Introduction

Nanostructures such as semiconductor quantum dots (QDs) emerged in the fields of biology and medicine as promising imaging and therapeutic agents.^1–6 Lately, it was proposed that QDs could be used in the photodynamic therapy (PDT) of cancer.^5,6 Unique optical properties of QDs such as size-tunable photoluminescence (PL) spectrum, high PL quantum yield and long lifetime as well as high extinction coefficient and broad absorption spectrum make them advantageous energy donors for conventional photosensitizers (PS) used in PDT.^3–5,7 Moreover, due to extremely large two-photon (TP) absorption cross-section of QDs^8–10 it becomes possible to use TP excitation in a QD (donor)–PS (acceptor) system for indirect excitation of PS, which usually have a low TP absorption cross-section.^11–14 TP excitation is usually provided within the near infrared (NIR) spectral region, which coincides with the tissue optical window. NIR irradiation can penetrate deeper into the tissue compared to visible light, allowing more profound examination and treatment of tumor. In addition, since TP excitation is confined within the focal volume of laser beam, selective activation of PS is achieved, which effectively enhances spatial resolution and produces less damage to surrounding healthy tissue.¹⁵ To our knowledge, there are only a few studies to date focusing on TP excitation of QD–PS systems,^16–18 which show that TP excited QDs can efficiently donate energy to electrostatically^16,17 or covalently¹⁸ attached PS. However, covalent coupling of QD–PS complexes requires additional chemical procedures which are not always feasible, while QD–PS complexes formed via electrostatic interaction might be less stable in biological media. In our study, for the first time, we have exploited the hydrophobic interaction to design QD–PS complex. Furthermore, we have used biocompatible QDs with lipid-based coating, which not only guarantees biocompatibility and stability of QDs for future applications in biological medium, but also serve as binding sites for amphiphilic PS ensuring its close localization to QD core and subsequently efficient Forster resonance energy transfer (FRET). As a PS, chlorin $e_6$ (${\mathrm{Ce}}_6$) was chosen. It is a typical light-activated drug for PDT with a high singlet oxygen generation quantum yield. However, a small TP absorption cross-section limits its application in TP PDT.

We have already shown, using single-photon (SP) excitation, that lipid-coated CdSe/ZnS QD form a stable, noncovalent complex with PS ${\mathrm{Ce}}_6$.^19,20 ${\mathrm{Ce}}_6$ molecules penetrate into the lipid coating of QD and localize close enough to QD shell/core for the FRET to occur. Here, we demonstrate that TP excited QD can work as efficient energy donors for ${\mathrm{Ce}}_6$ as well.

2.

Materials and Methods

${\mathrm{Ce}}_6$ tetrasulfonic acid was purchased from Frontier Scientific Inc. (USA). CdSe/ZnS QDs (${\lambda }_{\mathrm{em}}=625\pm 5\mathrm{nm}$, core diameter—$7.1\pm 0.5\mathrm{nm}$,²¹ hydrodynamic diameter—$22.8\pm 3\mathrm{nm}$) coated by phospholipid layer with PEG and terminal carboxyl groups were obtained from eBioscience Inc. (USA). All materials were used without further purification. ${\mathrm{Ce}}_6$ was dissolved in a small amount of 0.2 M NaOH solution and further diluted with a phosphate buffer (PB) (pH 7) to concentrations varying from 0.02 to 0.3 μM. Stock solution of QDs (10 μM) was diluted with PB to 0.02 μM working concentration. $\mathrm{QD}-{\mathrm{Ce}}_6$ solutions were made by adding 5 μL of concentrated ${\mathrm{Ce}}_6$ solution to the already prepared 0.02 μM solution of QDs, obtaining $\mathrm{QD}:{\mathrm{Ce}}_6$ molar ratios from $1:1$ up to $1:15$. Dilution effect for $\mathrm{QD}-{\mathrm{Ce}}_6$ solution after addition of ${\mathrm{Ce}}_6$ was encountered by adding 5 μL of PB to the control QD solution.

SP absorption spectra were recorded with Cary 50 UV–Vis spectrophotometer (Varian Inc., Palo Alto, California). SP excited PL spectra were measured with Cary Eclipse spectrophotometer (Varian Inc., USA). TP excitation of samples was produced by laser system PHAROS (Light Conversion, Lithuania) operating at 1030 nm wavelength with pulse duration of 280 fs and 1 kHz repetition rate. The average laser power of focused beam entering the sample was approximately 200 mW (pulse energy 200 μJ; peak pulse power 0.7 GW). TP induced PL signal was collected at a 90-deg angle and guided to spectrophotometer by an optical fiber. All experiments were performed at room temperature (RT).

TP absorption cross-sections of samples were calculated by a comparative method²² using Rhodamine 6G TP absorption cross-section ${\sigma }_2^{(r)}({\lambda }_2)$ as a reference [${\sigma }_2^{(r)}(\lambda =1030\mathrm{nm})=9.8$ Goeppert-Mayer (GM) in methanol].²³ SP and TP induced PL intensities of samples were measured at their corresponding PL band maxima $F_1^{(s)}({\lambda }^{(s)})$ and $F_2^{(s)}({\lambda }^{(s)})$ respectively. In both cases ${\lambda }^{(s)}=627\mathrm{nm}$. The SP extinction coefficient at the excitation wavelength of the samples ${\epsilon }_1^{(s)}({\lambda }_1)$, (${\lambda }_1=515\mathrm{nm}$ ) was evaluated from the absorption spectra. The Rhodamine 6G SP and TP induced PL intensities at their corresponding band maxima $F_1^{(r)}({\lambda }^{(r)})$ and $F_2^{(r)}({\lambda }^{(r)})$, respectively, $({\lambda }^{(r)}=556\mathrm{nm})$ and SP extinction coefficient ${\epsilon }_1^{(r)}({\lambda }_1),{\lambda }_1=515\mathrm{nm}$ were evaluated in the same way as for investigated samples. TP absorption cross-section can then be calculated as follows:

By this method, neither the parameters of the excitation light (pulse energy, pulse duration, spatial, and temporal intensity distribution), nor the parameters of the detection setup (detector characteristics and refractive indexes of solvents), nor PL quantum yield need to be known.

The efficiency $E$ of FRET between QDs and ${\mathrm{Ce}}_6$ molecules was calculated from the decrease of QD (donors) PL in the presence of ${\mathrm{Ce}}_6$ molecules (acceptors):

3.

Results and Discussion

SP absorption and PL spectra of pure ${\mathrm{Ce}}_6$ and QDs aqueous solutions are shown in Fig. 1. Absorption spectrum of ${\mathrm{Ce}}_6$ consists of the intensive Soret band at 405 nm and less intensive $Q$ bands, with the most intensive $Q(I)$ in the red spectral region at 654 nm. In aqueous solution the maximum of ${\mathrm{Ce}}_6$ fluorescence band is at 660 nm. Absorbance of QDs gradually decreases from UV to the red spectral region, ending with the last excitonic band at 615 nm. QDs have an intensive, narrow, and symmetrical PL band with a peak at 627 nm. It can be clearly seen that the PL band of QDs overlaps with the absorption spectrum of ${\mathrm{Ce}}_6$ (Fig. 1). This satisfies the main requirement for FRET between QDs and ${\mathrm{Ce}}_6$ to occur. Our previous study showed that addition of ${\mathrm{Ce}}_6$ to QD aqueous solution results in stable $\mathrm{QD}-{\mathrm{Ce}}_6$ complex formation, driven by the hydrophobic interaction between nonpolar moiety of ${\mathrm{Ce}}_6$ and lipids of QD coating.^19,20 Here, PL spectra after SP excitation of QDs solution mixed with different amounts of ${\mathrm{Ce}}_6$ are shown in Fig. 2(a). Localization of ${\mathrm{Ce}}_6$ molecules in the hydrophobic interior of the lipid part of QDs results in shift of ${\mathrm{Ce}}_6$ fluorescence band from 660 to 670 nm. Analogous red shift of ${\mathrm{Ce}}_6$ fluorescence maximum to 670 nm has been reported when ${\mathrm{Ce}}_6$ molecules were incorporated into lipid bilayer.^24,25 An increasing ${\mathrm{Ce}}_6$ concentration significantly quenches PL intensity of QDs, while the fluorescence intensity of bound ${\mathrm{Ce}}_6$ molecules at 670 nm simultaneously increases [Figs. 2(a) and 3]. A significant increase in fluorescence intensity of ${\mathrm{Ce}}_6$ is observed up to the $\mathrm{QD}:{\mathrm{Ce}}_6$ molar ratio of $1:5$ however, further rise in ${\mathrm{Ce}}_6$ concentration leads only to negligible changes in ${\mathrm{Ce}}_6$ fluorescence intensity [Figs. 2(a) and 3(b)]. Meanwhile, the PL intensity of QDs noticeably decreases even at higher concentrations of ${\mathrm{Ce}}_6$ (${\mathrm{Ce}}_6/\mathrm{QD}>5$) [Figs. 2(a) and 3(a)]. This might be explained by the self-quenching of bound ${\mathrm{Ce}}_6$ molecules on the surface of QDs at relatively high concentrations of ${\mathrm{Ce}}_6$. Moreover, the fluorescence intensity of pure ${\mathrm{Ce}}_6$ under SP (at 515 nm) excitation is much lower than that bound to QDs [Fig. 2(a)], as can be clearly seen in inset of Fig. 3(b), where the intensity of pure ${\mathrm{Ce}}_6$ fluorescence maximum (660 nm) is shown against the concentration. In addition, in our previous experiments we have shown that the fluorescence excitation spectra measured at 670 nm, contain a significant contribution of QDs spectrum.¹⁹ Also, the average PL decay time of QDs decreases upon $\mathrm{QD}-{\mathrm{Ce}}_6$ complex formation.¹⁹ All spectral features listed above indicate FRET occurring between QDs and bound ${\mathrm{Ce}}_6$ molecules. The efficiency of FRET in $\mathrm{QD}-{\mathrm{Ce}}_6$ complex at $\mathrm{QD}:{\mathrm{Ce}}_6$ molar ratios of $1:1$ and $1:5$ was found to be 45% and 82%, respectively. Similar PL measurements for pure QDs, ${\mathrm{Ce}}_6$, and $\mathrm{QD}-{\mathrm{Ce}}_6$ aqueous solutions at different $\mathrm{QD}:{\mathrm{Ce}}_6$ molar ratios were conducted using TP excitation at 1030 nm [Fig. 2(b)]. The intensive PL band of QDs observed at 627 nm confirms that QDs can be effectively excited using TP excitation. QDs are known to have a large TP absorption cross-section, which ranges from 75 to 47,000 GM, depending on their chemical structure, size, and excitation wavelength.⁹ Here, TP absorption cross-section for QDs was calculated according to Eq. (1) $F_2^{(s)}({\lambda }^{(s)})=1600\text{a.u.}$; $F_2^{(r)}({\lambda }^{(r)})=1500\text{a.u.}$; $F_1^{(r)}({\lambda }^{(r)})=9600\text{a.u.}$; $F_1^{(s)}({\lambda }^{(s)})=330\text{a.u.}$; ${\epsilon }_1^{(s)}({\lambda }_1)=724720\mathrm{l}·\mathrm{mol}-1·{\mathrm{cm}}^{-1}{\epsilon }_1^{(r)}$; $({\lambda }_1)=65890\mathrm{l}·{\mathrm{mol}}^{-1}·{\mathrm{cm}}^{-1}$ and found to be 3325 GM, which is in agreement with the previous reports.^8–10

Fig. 1

Normalized SP absorption and PL spectra of pure QDs and ${\mathrm{Ce}}_6$ aqueous solutions (PB pH 7). The arrow indicates a wavelength used for SP at 515 nm and corresponding TP at 1030 nm excitation of the samples.

Fig. 2

PL spectra of pure QDs (0.02 μM), ${\mathrm{Ce}}_6$ (0.1 μM), and mixed $\mathrm{QD}-{\mathrm{Ce}}_6$ aqueous solutions at different $\mathrm{QD}:{\mathrm{Ce}}_6$ molar ratios ($1:1$ to $1:15$) under excitation of (a) SP 515 nm and (b) TP 1030 nm. Spectra were normalized to the PL peak of QDs at 627 nm.

Fig. 3

Changes in the PL intensities of $\mathrm{QD}-{\mathrm{Ce}}_6$ solutions upon increasing $\mathrm{QD}:{\mathrm{Ce}}_6$ molar ratio from $1:1$ to $1:15$ under SP and TP excitation (a) at QD maximum at 627 nm (values are normalized to the initial intensity of QDs without ${\mathrm{Ce}}_6$) and (b) at bound ${\mathrm{Ce}}_6$ maximum at 670 nm. Inset shows the intensity changes of pure ${\mathrm{Ce}}_6$ solution at fluorescence maximum at 660 nm by increasing the concentration of ${\mathrm{Ce}}_6$ from 0.02 to 0.3 μM under SP and TP excitation wavelength.

Meanwhile, only a negligible emission signal can be observed for pure ${\mathrm{Ce}}_6$ solution under TP excitation, since ${\mathrm{Ce}}_6$ TP absorption cross-section is significantly lower (13 GM $(F_2^{(s)}({\lambda }^{(s)})=110\text{a.u.};F_1^{(s)}({\lambda }^{(s)})=57\text{a.u.};{\epsilon }_1^{(s)}({\lambda }_1)=6830\mathrm{l}·{\mathrm{mol}}^{-1}·{\mathrm{cm}}^{-1})$ ). For organic dyes and porphyrin type molecules TP absorption cross-section are usually in the order of tens of GM units or less.^{11–14,22,23} Thus, effective application of such organic molecules in TP PDT is limited. Higher TP absorption cross-section for QDs is obtained at shorter wavelengths as 800 nm.¹⁰ However, 1030 nm wavelength light is more suitable for FRET investigation in $\mathrm{QD}-{\mathrm{Ce}}_6$ complex, since only QD is excited via TP excitation.

Upon TP excitation similar spectral changes in mixed $\mathrm{QD}-{\mathrm{Ce}}_6$ solutions are observed as in the case of SP excitation [Figs. 2(b) and 3]. In mixed $\mathrm{QD}-{\mathrm{Ce}}_6$ solution PL intensity of QDs significantly decreases compared to pure QDs and an intensive fluorescence peak at 670 nm appears. No significant differences in PL changes of QDs between SP and TP excitation were observed suggesting that energy transfer pathway to bound ${\mathrm{Ce}}_6$ molecules is independent on the excitation way [Fig. 3(a)].

Fluorescence intensity of ${\mathrm{Ce}}_6$ at 670 nm, under TP excitation at 1030 nm of $\mathrm{QD}-{\mathrm{Ce}}_6$ solution increases until $\mathrm{QD}:{\mathrm{Ce}}_6$ molar ratio of $1:5$, and begins to decrease for higher concentrations of ${\mathrm{Ce}}_6$ [Fig. 3(b)]. This suggests that up to five ${\mathrm{Ce}}_6$ molecules per single QD can localize within the lipid part of QDs coating without loss in fluorescence intensity. Further increase in concentration negatively acts on PL of both the QDs and ${\mathrm{Ce}}_6$, without any enhancements on energy transfer efficiency or PL intensity. Such decrease was not observed under SP excitation at 515 nm most probably because it is compensated for by the fluorescence signal from direct excitation of free ${\mathrm{Ce}}_6$. However, when SP excitation at 465 nm is used, where absorption of ${\mathrm{Ce}}_6$ is minimal, a similar decrease can be observed (data not shown). As mentioned previously, it can be explained by fluorescence self-quenching of bound ${\mathrm{Ce}}_6$ molecules due to their possible aggregation on the surface of QDs at higher ${\mathrm{Ce}}_6$ concentrations. Under TP excitation at 1030 nm, the fluorescence of ${\mathrm{Ce}}_6$ can be selectively observed only due to FRET from TP excited QDs to bound ${\mathrm{Ce}}_6$ molecules. FRET efficiency under TP excitation from QD to ${\mathrm{Ce}}_6$ in $\mathrm{QD}-{\mathrm{Ce}}_6$ complex at $\mathrm{QD}:{\mathrm{Ce}}_6$ molar ratio of $1:1$ was found to be 49% and increased to the maximum 82% at $\mathrm{QD}:{\mathrm{Ce}}_6$ molar ratio $1:5$. FRET efficiencies under TP excitation coincide with those obtained under SP excitation. It can be assumed that TP excitation can also be used to effectively excite other amphiphilic PS that can hydrophobically interact with the lipid coating of QDs.²⁰ It is noteworthy to mention that FRET efficiency values were obtained at RT; however, during the biomedical applications (experiments in cell cultures and experimental animals) $\mathrm{QD}-{\mathrm{Ce}}_6$ complexes will be exposed to higher temperatures. The increase of the temperature to 50°C causes decrease in FRET efficiency approximately by 10% (data not shown). It can be considered as negligible and $\mathrm{QD}-{\mathrm{Ce}}_6$ complex can be used for biomedical applications.

4.

Conclusions

For the first time, we demonstrated that $\mathrm{QD}-{\mathrm{Ce}}_6$ complex self-assembled due to the hydrophobic interaction between nonpolar moiety of ${\mathrm{Ce}}_6$ molecules and lipids of QDs coating can be effectively excited via TP irradiation at 1030 nm, which spectrally coincides with the transparency window of biological tissue. Indirect excitation of ${\mathrm{Ce}}_6$ by energy transfer from TP excited QDs, overcomes difficulties in direct TP excitation of ${\mathrm{Ce}}_6$, thus allowing one to utilize therapeutic action of ${\mathrm{Ce}}_6$ under TP irradiation. The highest efficiency of FRET within $\mathrm{QD}-{\mathrm{Ce}}_6$ complex under TP excitation reached up to 82% when multiple ${\mathrm{Ce}}_6$ molecules are bound to QD, which is significantly higher than that obtained by other similar QD–PS systems to date.^16–18 High FRET efficiency within $\mathrm{QD}-{\mathrm{Ce}}_6$ complex exploits full PDT potential of ${\mathrm{Ce}}_6$ at relatively small concentrations of PS that is very important when concerning dark toxicity and clearance issues of $\mathrm{QD}-{\mathrm{Ce}}_6$ complex. These promising results constitute a first step toward the application of such type QD–PS complexes alongside TP irradiation in PDT.