- Email: [email protected]
Fluorescence lifetime imaging in PDT. An overview
ARTICLE IN PRESS
Medical Laser Application 20 (2005) 125–129 www.elsevier.de/mla
Fluorescence lifetime imaging in PDT. An overview Angelika Ru¨ck, Frank Dolp, Christian Hu¨lshoff, Carmen Hauser, Claudia Scalfi-Happ Institut fu¨r Lasertechnologien in der Medizin und Meßtechnik, Helmholtzstr. 12, 89081 Ulm, Germany Received 11 March 2005; accepted 14 March 2005
Abstract Various problems arising during molecular imaging of different fluoroprobes and metabolites used in photodynamic therapy (PDT) could be circumvented by focusing on time-resolved detection. For this, an interesting new method seems to be time-correlated single photon counting, where a time-to-amplitude converter determines the temporal position and a scanning interface connected to the scanning unit of a laser microscope determines the spatial location of a signal. In combination with spectral resolved detection (spectral lifetime imaging) the set-up achieves the features of highly sophisticated lifetime imaging systems. The potential of time-resolved detection is evident for the precursor 5-ALA, where the photosensitizing porphyrins are synthesized in the tumor cells. Protoporphyrin IX, which is produced in mitochondria, induces high phototoxicity but also different metabolites could be involved. During cell differentiation various enzymes, which are active in varying degrees control the ALA metabolism. It is therefore of main interest to localize 5-ALA metabolites during cell differentiation and cell growth. Subcellular differentiation of those metabolites without extensive extraction procedures is not trivial, because of highly overlapping spectral properties. Measuring the fluorescence lifetime on a subcellular level could be a successful alternative (fluorescence lifetime imaging, FLIM). As it will be discussed FLIM and SLIM in principal are powerful techniques to improve selectivity of fluorescence guided diagnosis and in fact PDT. Usage in clinics will depend at least on the availability of easy and fast detection systems, which can be realized in different ways. For excitation, relatively cheap laser diodes or simple custom-made Ti:Saphir laser systems are now on the market. r 2005 Elsevier GmbH. All rights reserved. Keywords: Laser scanning microscopy; Pulsed diode lasers; TCSPC; FLIM; SLIM; PDT; 5-ALA metabolites
Introduction Conventionally in imaging, fluorescence fluorescence intensity between fluorophores characteristics. Also
photodynamic therapy (PDT) intensity is measured. However, is not sufficient to distinguish with highly overlapping spectral differences in the fluorescence
Corresponding author.
E-mail address: [email protected] (A. Ru¨ck). 1615-1615/$ - see front matter r 2005 Elsevier GmbH. All rights reserved. doi:10.1016/j.mla.2005.03.009
intensities due to quenching by other molecules, aggregation, energy transfer, etc. can be difficult to quantify or interpret. In contrast to this, fluorescence lifetime imaging (FLIM) uses the fluorescence decay time which is the inverse of the sum of the rate parameters for all depopulation processes [1]. The fluorescence decay time, also commonly used as lifetime t depends on the intrinsic characteristics of the fluorophore itself and on the local environment, such as viscosity, pH, refractive index, aggregation, as well as
ARTICLE IN PRESS 126
A. Ru¨ck et al. / Medical Laser Application 20 (2005) 125–129
interactions with other molecules. Thus, imaging of the fluorescence lifetime by FLIM can be used to probe the surroundings of a fluorophore. The time-resolved information is obtained either in the time domain by exciting the sample with a short pulsed laser and observing the decay of the fluorescence intensity with time-correlated single photon counting (TCSPC), gating or a streak camera, or in the frequency domain, where the fluorescence decay is calculated from the phase shift of the fluorescence. Different species of tumor-localizing porphyrins, which are of importance in PDT could be distinguished by time-resolved methods, i.e. monomeric and aggregated porphyrin molecules as well as ionic species located at different cellular sites [2–4]. A picosecond laser line-scanning microscope combined with a subnanosecond-gated image intensifier and a chargecoupled device (CCD) was used to study the fluorescence lifetime of the photosensitizers disulphonated aluminum phthalocyanine (AlPcS2), pyridinium zinc(II)phthalocyanine (ZnPPC) and meta-tetra(hydroxyphenyl)chlorine (m-THPC) [5]. FLIM of hematoporphyrin derivative (HpD) in tumor bearing mice, using a time-gated video camera system revealed that the fluorescence decay was appreciably slower in the tumor than in the healthy tissues nearby [6]. The uptake of photosensitizer aggregates and their monomerization inside tumor cells could also be demonstrated by means of FLIM using TCSPC techniques [7]. For time-resolved techniques the use of mode-locked Ti:Saphir laser systems are wide-spread [8]. In addition, new developments in diode laser technology suggest the usage of such lasers [9]. Spectrally resolved FLIM microscopy was performed by a pulsed laser diode emitting at 635 nm and TCSPC techniques to measure the fluorescence lifetime of fluorophor labeled oligonucleotides in living cells [10]. Only recently, we coupled a short-pulsed diode laser emitting at 398 nm to a confocal laser scanning microscope for detecting FLIM of photosensitizers with TCSPC techniques [11]. With this equipment the time-resolved characteristics of the mitochondrial marker Rhodamine 123 and 5-ALA (5aminolevulinic-acid) as well as 5-ALAhe (5-aminolevulinic-acid-hexylester) induced protoporphyrine IX (PPIX), were observed [12]. 5-ALA, a precursor of photosensitizing porphyrins, is approved for the treatment of actinic keratosis, basal cell carcinomas, as well as fluorescence guided diagnosis of bladder cancer and is now widely used in PDT [13]. So far, PPIX, which is synthesized in mitochondria seems to be the only porphyrin, which is of importance during 5-ALA PDT. However, other metabolites from 5-ALA, like Uroporphyrin or Coproporphyrin could be involved. Interestingly, different enzymes control the ALA metabolism, which are active in varying degrees during cell differentiation [14]. Subcellular determina-
tion of 5-ALA metabolites without extensive extraction procedures is however not trivial, because of highly overlapping spectral properties. Therefore, 5-ALA is a good example to demonstrate the superiority of timeresolved methods to discriminate different metabolites of photosensitizers or even different molecular species, which could improve significantly selectivity of fluorescence guided diagnosis and therapy.
Materials and methods In order to investigate the time-resolved subcellular fluorescence characteristics of 5-ALA induced metabolites a short pulsed diode laser emitting at 398 nm (LDH 400, PicoQuant GmbH, Berlin, Germany) was coupled to a laser scanning microscope (LSM410, Carl Zeiss, Germany), additionally to the standard laser sources of the LSM410, an argon-ion laser emitting light at 488 and 514 nm and a HeNe laser emitting at 633 nm. The details of the experimental setup are described in Refs. [11,15]. Briefly, FLIM was performed using the singlechannel TCSPC module SPC-730, connected to the ultra-fast Hamamatsu PMH-100 (Becker&Hickl GmbH, Berlin, Germany) which was attached to the LSM410. Alternatively, time-resolved detection by spectral lifetime imaging (SLIM) was performed with the TCSPC module 830/PML/16 (Becker&Hickl GmbH, Berlin, Germany) using the ultra-fast Hamamatsu R5900-01-L16 and the spectrograph MS125 (LOT-Oriel), with 600 lines/mm diffraction grating in front of the detector. With this equipment a spectral resolution of about 25 nm was achievable. As described in Ref. [16], the TCSPC module receives the timing pulse and the scan clock signals (frame sync, line sync and pixel clock) from the scanning unit of the microscope. For each photon, the TCSPC module determines the location within the scanning area, the time of the photon with respect to the laser pulse sequence and the detector channel number. Every pixel of the fluorescence lifetime image was achieved by a software binning of 4 4 pixels of the LSM image. From these set of data fluorescence lifetime images could be calculated using the SPCImage Version 2.6.2 software (Becker&Hickl GmbH). The fluorescence lifetime image is represented in pseudocolors. 5-ALA was obtained from Fluka (Neu-Ulm, Germany). A 100 mM stock solution of 5-ALA was prepared in aqua bidest, the solution was subsequently neutralized with NaOH. To demonstrate production of different metabolites, cell cultures were incubated with 1 mM 5-ALA for 4 h. Only freshly prepared solutions were used for incubation. Protoporphyrin IX disodiumsalt (Porphyrin Products, USA) was dissolved in 0.1 M HCl, neutralized with 0.1 M NaOH and diluted with
ARTICLE IN PRESS A. Ru¨ck et al. / Medical Laser Application 20 (2005) 125–129
PBS. Cells were incubated with 1 mM for 4 h. Uroporphyrin I and III were dissolved in PBS and neutralized with 1 M NaOH. Cells were incubated with 10 mM for 24 h. Cultures of RR 1022 epithelial cells were grown in M199 medium (Gibco, UK) supplemented with 10% fetal calf serum (FCS) and antibiotics (penicillin, streptomycin) at 37 1C and 5% CO2. The cells were seeded on glass bottom microwell dishes, coverglass 0.16–0.19 mm (MatTek, USA) at a density of 75–100 cells/mm2 and were allowed to grow for 36 h. The medium was replaced and cells were incubated in
127
the exponential growth phase. Microscopic observation was performed immediately after removing the incubation medium and rinsing twice with indicator free medium at 37 1C.
Results With the described detection setups the time-resolved subcellular fluorescence characteristics of 5-ALA metabolites, Uroporphyrin I and III and PPIX were studied. Fig. 1 demonstrates the fluorescence lifetime image of
Fig. 1. FLIM of RR 1022 cells, incubated with 5-ALA, fluorescence emission 4590 nm, monoexponential fitting.
Fig. 2. FLIM of 5-ALA metabolites, Uroporphyrin I, Uropophyrin III and PPIX.
ARTICLE IN PRESS 128
A. Ru¨ck et al. / Medical Laser Application 20 (2005) 125–129
5-ALA, after an incubation time of 4 h. The lifetime was calculated pixel by pixel by a monoexpontential curve fitting. The blue curve in the diagram correlates with the detected photons, the green curve is the system response and the red curve results from the fitting. The colorcoding of the lifetime image is represented in the histogram. The lifetimes differed significantly within the cells. In principal three different lifetime areas could be detected, one with a long lifetime (blue color), which correlated with the plasma membrane and the outer cytoplasm, one with a middle lifetime (green color) attributed to the mitochondria and cytoplasm of the cells and one with a short lifetime (red color). From photobleaching experiments the areas with the longest lifetimes exhibited the same bleaching kinetics, whereas the compound with the short lifetime was different (data not shown). It seems, that different 5-ALA metabolites (photoproducts or different porphyrins) could be detected within the cells. To support our hypothesis comparative results obtained for 5-ALA and the ‘‘pure’’ porphyrins Uroporphyrin I, Uroporphyrin III and PPIX, are demonstrated in Fig. 2. In all cases the same colorcoding was used. Now, the lifetime was calculated by a biexponential curve fitting. Binning was set to 1, that means, that the lifetime was calculated as the mean from a 3 3 pixel area. The histograms of the mean lifetimes are also presented in Fig. 2. In the case of 5-ALA metabolites at least two different mean lifetime regions could be identified, one with a long lifetime which correlates to the plasma membrane and the outer cytoplasm (represented in blue color), one with a shorter lifetime attributed mainly to the cytoplasm of the cells (green color). Some cells even exhibited much shorter lifetimes (orange color). When comparing the lifetime image of 5-ALA with Uroporphyrin I, Uroporphyrin III, as well as PPIX, it seems, that the long lifetime could be attributed to PPIX, whereas the shorter lifetimes contain to some extend Uroporporphyrin I. Some of the Uroporphyrin III seems to be metabolized to PPIX. The aim of our study was to show, whether different 5-ALA metabolites are produced within the cells, depending on cell cycle and cell differentiation. Because it is known, that phototoxic efficiency of the porphyrins varies significantly, a clear discrimination would help to understand the sometimes unwanted different cellular response of tumor cells. For this, FLIM is a powerful method and a challenge to improve PDT.
Discussion This work describes time-resolved fluorescence techniques and applications for PDT. The advantage of
using TCSPC and pulsed laser excitation to improve selectivity of fluorescence guided diagnosis is due to the unlimited dynamic range associated with photon counting techniques and the easy visualization of fluorescence decays. The disadvantage is a slow detection time, because each photon has to be timed individually. For the practical usage, a fast streak camera with very high temporal resolution would be perfect, however still expensive. The potential of time-resolved detection is evident for the precursor 5-ALA, where the photosensitizing porphyrins are synthesized in the tumor cells. PPIX, which is produced in mitochondria, induces high phototoxicity but also different isomers of Uroporphyrin and Coproporhyrin, which are localized in the cytoplasm of the cells could be important. During cell differentiation various enzymes, which are active in varying degrees control the ALA metabolism. It is therefore of main interest to localize 5-ALA metabolites during cell differentiation and cell growth. Obviously, non-invasive methods should be preferred. Whereas spectral differentiation is very difficult, due to highly overlapping emission spectra of the porphyrins involved, we demonstrated on a cellular level that FLIM is able to distinguish the compounds. This technique achieves the features of highly sophisticated lifetime imaging systems for molecular studies in PDT. Whether usage in the clinical practice becomes acceptable depends at least on the availability of easy and fast detection systems with resonable prices. On the excitation site, relatively cheap laser diodes or simple custommade Ti:Saphir laser systems are now on the market.
Acknowledgements This work is carried out by the Ministry of Economic Affairs of the state of Baden-Wu¨rttemberg by means of the Landesstiftung Baden-Wu¨rttemberg, order 44332.62-ILM/15.
Zusammenfassung Zeitaufgelo¨ste Fluoreszenz und PDT Im Rahmen der photodynamischen Therapie ist die laserinduzierte Fluoreszenzdiagnostik ein wichtiges Hilfsmittel, um Tumorbegrenzungen oder Tumorreste sichtbar zu machen und so eine vollsta¨ndige Resektion zu gewa¨hrleisten. Das Problem jedoch ist neben der Sensitivita¨t eine ha¨ufig zu geringe Selektivita¨t, die besonders im Bereich von Hirntumoren zu unbefriedigenden Ergebnissen fu¨hrt. Um die Selektivita¨t zu verbessern, bieten sich zeitaufgelo¨ste Methoden an, mit deren Hilfe es mo¨glich wird, verschiedene Fluorophore und Metaboliten eindeutig zu unterscheiden. Neu und
ARTICLE IN PRESS A. Ru¨ck et al. / Medical Laser Application 20 (2005) 125–129
interessant ist in diesem Sinne die zeitkorrelierte Einzelphotonenza¨hlung, auch time-correlated single photon counting (TCSPC) genannt, die in Kombination mit einer Scanningeinheit und kurzgepulster Anregung die bildgebende Darstellung der Fluoreszenzabklingzeit (sogenanntes fluorescence lifetime imaging (FLIM)) ermo¨glicht. An Hand der fu¨r die Klinik zugelassenen 5-Aminolevulinsa¨ure (5-ALA) la¨sst sich das Potential zeitaufgelo¨ster Fluoreszenzmessungen besonders gut demonstrieren. Nach der Applikation von 5-ALA entstehen in den Zellen unterschiedliche Porphyrine, die alle mehr oder weniger stark phototoxisch sind. Protoporphyrin IX, das in den Mitochondrien synthetisiert wird, ist dabei das wirksamste Porphyrin. Es ko¨nnen jedoch auch andere Metaboliten, wie Uroporphyrin und Koproporphyrin entstehen. Die Synthese dieser Porphyrine wird durch Enzyme gesteuert, die in Abha¨ngigkeit des Diffenzierungsgrades der Zelle mehr oder weniger aktiv sind. Wie im Rahmen dieser Arbeit diskutiert wird, ko¨nnen durch FLIM die verschiedenen Metaboliten mit subzellula¨rer Auflo¨sung unterschieden werden. Zeitaufgelo¨ste Methoden ko¨nnen damit allgemein die Selektivita¨t der Fluoreszenzdiagnostik erho¨hen und damit die PDT signifikant verbessern. r 2005 Elsevier GmbH. All rights reserved. Schlu¨sselwo¨rter: PDT; 5-ALA; Laserscanning Mikroskopie; Gepulste Diodenlaser; TCSPC; Zeitaufgelo¨ste Fluoreszenz (FLIM); Spektral- und zeitaufgelo¨ste Fluoreszenz (SLIM)
References [1] Phillips D. Luminescence lifetimes in biological systems. Analyst 1994;119:543–50. [2] Schneckenburger H, Ko¨nig K, Kunzi-Rapp K, WestphalFro¨sch C, Ru¨ck A. Time-resolved in-vivo fluorescence of photosensitizing porphyrins. J Photochem Photobiol B: Biol 1993;21:143–7. [3] Schneckenburger H, Gschwend MH, Sailer R, Ru¨ck A, Strauß WSL. Time-resolved pH dependent fluorescence of hydrophilic porphyrins in solution and in cultivated cells. J Photochem Photobiol B: Biol 1995;27:251–5. [4] Strauß WSL, Sailer R, Schneckenburger H, Akgu¨n N, Gottfried V, Chetwer L, Kimel S. Study of the photodynamic efficacy of naturally occurring porphyrins in
[5]
[6]
[7]
[8] [9] [10]
[11]
[12]
[13]
[14]
[15]
[16]
129
endothelial cells in vitro and microvasculature in vivo. J Photochem Photobiol B: Biol 1997;39:176–84. Connelly JP, Botchway SW, Kunz L, Pattison D, Parker AW, MacRobert AJ. Time-resolved fluorescence imaging of photosensitiser distributions in mammalian cells using a picosecond laser line-scanning microscope. J Photochem Photobiol A: Chemistry 2001;142:169–75. Cubeddu R, Canti G, Pifferi A, Taroni P, Valentini G. Fluorescence lifetime imaging of experimental tumors in hematoporphyrin derivative-sensitized mice. Photochem Photobiol 1997;66:229–36. Kelbauskas L, Dietel W. Internalization of aggregated photosensitizers by tumor cells: subcellular time-resolved fluorescence spectroscopy on derivatives of pyropheophorbide-a ethers and chlorin e6 under femtosecond oneand two-photon excitations. Photochem Photobiol 2002; 76:686–94. Ko¨nig K. Multiphoton microscopy in life science. J Microscopy 2000;200:83–104. Wahl M, Kaenders W. Blue diode lasers light up in pulsed and tuneable output systems. EuroPhotonics 1999;4:52–4. Knemeyer JP, Herten DP, Sauer M. Detection and identification of single molecules in living cells using spectrally resolved fluorescence lifetime imaging microscopy. Anal Chem 2003;75:2147–53. Kress M, Meier T, Steiner R, Dolp F, Erdmann R, Ortmann U, Ru¨ck A. Time- resolved microspectrofluorometry and fluorescence lifetime imaging of photosensitizers using ps pulsed diode lasers in laser scanning microscopes. J Biomed Opt 2003;8(1):26–32. Ru¨ck A, Dolp F, Scalfi-Happ C, Steiner R, Beil M. Timeresolved microspectrofluorometry and fluorescence lifetime imaging using ps pulsed diode lasers and TCSPC techniques in laser scanning microscopes. Proc SPIE 2003;5139:166–72. Bissonette R, Bergeron A, Liu Y. Large surface photodynamic therapy with aminolevulinic acid: treatment of actinic keratoses and beyond. J Drugs Dermatol 2004; 3(1):26–31. Ickowicz Schwartz D, Gozlan Y, Greenbaum I, Babushkina T, Katcoff DJ, Malik Z. Differentiation-dependent photodynamic therapy regulated by porphobilinogen deaminase in B16 melanoma. Br J Cancer 2004;90(9): 1833–41. Ru¨ck A, Dolp F, Hu¨lshoff C, Hauser C, Scalfi-Happ C. FLIM and SLIM for molecular imaging in PDT. Proc SPIE 2005:5700–32 in press. Becker W, Bergmann A, Biscotti G, Koenig K, Riemann I, Kelbauskas L, Biskup C. High-speed FLIM data acquisition by time-correlated single photon counting. Proc SPIE 2004;5323:27–35.
Medical Laser Application 20 (2005) 125–129 www.elsevier.de/mla
Fluorescence lifetime imaging in PDT. An overview Angelika Ru¨ck, Frank Dolp, Christian Hu¨lshoff, Carmen Hauser, Claudia Scalfi-Happ Institut fu¨r Lasertechnologien in der Medizin und Meßtechnik, Helmholtzstr. 12, 89081 Ulm, Germany Received 11 March 2005; accepted 14 March 2005
Abstract Various problems arising during molecular imaging of different fluoroprobes and metabolites used in photodynamic therapy (PDT) could be circumvented by focusing on time-resolved detection. For this, an interesting new method seems to be time-correlated single photon counting, where a time-to-amplitude converter determines the temporal position and a scanning interface connected to the scanning unit of a laser microscope determines the spatial location of a signal. In combination with spectral resolved detection (spectral lifetime imaging) the set-up achieves the features of highly sophisticated lifetime imaging systems. The potential of time-resolved detection is evident for the precursor 5-ALA, where the photosensitizing porphyrins are synthesized in the tumor cells. Protoporphyrin IX, which is produced in mitochondria, induces high phototoxicity but also different metabolites could be involved. During cell differentiation various enzymes, which are active in varying degrees control the ALA metabolism. It is therefore of main interest to localize 5-ALA metabolites during cell differentiation and cell growth. Subcellular differentiation of those metabolites without extensive extraction procedures is not trivial, because of highly overlapping spectral properties. Measuring the fluorescence lifetime on a subcellular level could be a successful alternative (fluorescence lifetime imaging, FLIM). As it will be discussed FLIM and SLIM in principal are powerful techniques to improve selectivity of fluorescence guided diagnosis and in fact PDT. Usage in clinics will depend at least on the availability of easy and fast detection systems, which can be realized in different ways. For excitation, relatively cheap laser diodes or simple custom-made Ti:Saphir laser systems are now on the market. r 2005 Elsevier GmbH. All rights reserved. Keywords: Laser scanning microscopy; Pulsed diode lasers; TCSPC; FLIM; SLIM; PDT; 5-ALA metabolites
Introduction Conventionally in imaging, fluorescence fluorescence intensity between fluorophores characteristics. Also
photodynamic therapy (PDT) intensity is measured. However, is not sufficient to distinguish with highly overlapping spectral differences in the fluorescence
Corresponding author.
E-mail address: [email protected] (A. Ru¨ck). 1615-1615/$ - see front matter r 2005 Elsevier GmbH. All rights reserved. doi:10.1016/j.mla.2005.03.009
intensities due to quenching by other molecules, aggregation, energy transfer, etc. can be difficult to quantify or interpret. In contrast to this, fluorescence lifetime imaging (FLIM) uses the fluorescence decay time which is the inverse of the sum of the rate parameters for all depopulation processes [1]. The fluorescence decay time, also commonly used as lifetime t depends on the intrinsic characteristics of the fluorophore itself and on the local environment, such as viscosity, pH, refractive index, aggregation, as well as
ARTICLE IN PRESS 126
A. Ru¨ck et al. / Medical Laser Application 20 (2005) 125–129
interactions with other molecules. Thus, imaging of the fluorescence lifetime by FLIM can be used to probe the surroundings of a fluorophore. The time-resolved information is obtained either in the time domain by exciting the sample with a short pulsed laser and observing the decay of the fluorescence intensity with time-correlated single photon counting (TCSPC), gating or a streak camera, or in the frequency domain, where the fluorescence decay is calculated from the phase shift of the fluorescence. Different species of tumor-localizing porphyrins, which are of importance in PDT could be distinguished by time-resolved methods, i.e. monomeric and aggregated porphyrin molecules as well as ionic species located at different cellular sites [2–4]. A picosecond laser line-scanning microscope combined with a subnanosecond-gated image intensifier and a chargecoupled device (CCD) was used to study the fluorescence lifetime of the photosensitizers disulphonated aluminum phthalocyanine (AlPcS2), pyridinium zinc(II)phthalocyanine (ZnPPC) and meta-tetra(hydroxyphenyl)chlorine (m-THPC) [5]. FLIM of hematoporphyrin derivative (HpD) in tumor bearing mice, using a time-gated video camera system revealed that the fluorescence decay was appreciably slower in the tumor than in the healthy tissues nearby [6]. The uptake of photosensitizer aggregates and their monomerization inside tumor cells could also be demonstrated by means of FLIM using TCSPC techniques [7]. For time-resolved techniques the use of mode-locked Ti:Saphir laser systems are wide-spread [8]. In addition, new developments in diode laser technology suggest the usage of such lasers [9]. Spectrally resolved FLIM microscopy was performed by a pulsed laser diode emitting at 635 nm and TCSPC techniques to measure the fluorescence lifetime of fluorophor labeled oligonucleotides in living cells [10]. Only recently, we coupled a short-pulsed diode laser emitting at 398 nm to a confocal laser scanning microscope for detecting FLIM of photosensitizers with TCSPC techniques [11]. With this equipment the time-resolved characteristics of the mitochondrial marker Rhodamine 123 and 5-ALA (5aminolevulinic-acid) as well as 5-ALAhe (5-aminolevulinic-acid-hexylester) induced protoporphyrine IX (PPIX), were observed [12]. 5-ALA, a precursor of photosensitizing porphyrins, is approved for the treatment of actinic keratosis, basal cell carcinomas, as well as fluorescence guided diagnosis of bladder cancer and is now widely used in PDT [13]. So far, PPIX, which is synthesized in mitochondria seems to be the only porphyrin, which is of importance during 5-ALA PDT. However, other metabolites from 5-ALA, like Uroporphyrin or Coproporphyrin could be involved. Interestingly, different enzymes control the ALA metabolism, which are active in varying degrees during cell differentiation [14]. Subcellular determina-
tion of 5-ALA metabolites without extensive extraction procedures is however not trivial, because of highly overlapping spectral properties. Therefore, 5-ALA is a good example to demonstrate the superiority of timeresolved methods to discriminate different metabolites of photosensitizers or even different molecular species, which could improve significantly selectivity of fluorescence guided diagnosis and therapy.
Materials and methods In order to investigate the time-resolved subcellular fluorescence characteristics of 5-ALA induced metabolites a short pulsed diode laser emitting at 398 nm (LDH 400, PicoQuant GmbH, Berlin, Germany) was coupled to a laser scanning microscope (LSM410, Carl Zeiss, Germany), additionally to the standard laser sources of the LSM410, an argon-ion laser emitting light at 488 and 514 nm and a HeNe laser emitting at 633 nm. The details of the experimental setup are described in Refs. [11,15]. Briefly, FLIM was performed using the singlechannel TCSPC module SPC-730, connected to the ultra-fast Hamamatsu PMH-100 (Becker&Hickl GmbH, Berlin, Germany) which was attached to the LSM410. Alternatively, time-resolved detection by spectral lifetime imaging (SLIM) was performed with the TCSPC module 830/PML/16 (Becker&Hickl GmbH, Berlin, Germany) using the ultra-fast Hamamatsu R5900-01-L16 and the spectrograph MS125 (LOT-Oriel), with 600 lines/mm diffraction grating in front of the detector. With this equipment a spectral resolution of about 25 nm was achievable. As described in Ref. [16], the TCSPC module receives the timing pulse and the scan clock signals (frame sync, line sync and pixel clock) from the scanning unit of the microscope. For each photon, the TCSPC module determines the location within the scanning area, the time of the photon with respect to the laser pulse sequence and the detector channel number. Every pixel of the fluorescence lifetime image was achieved by a software binning of 4 4 pixels of the LSM image. From these set of data fluorescence lifetime images could be calculated using the SPCImage Version 2.6.2 software (Becker&Hickl GmbH). The fluorescence lifetime image is represented in pseudocolors. 5-ALA was obtained from Fluka (Neu-Ulm, Germany). A 100 mM stock solution of 5-ALA was prepared in aqua bidest, the solution was subsequently neutralized with NaOH. To demonstrate production of different metabolites, cell cultures were incubated with 1 mM 5-ALA for 4 h. Only freshly prepared solutions were used for incubation. Protoporphyrin IX disodiumsalt (Porphyrin Products, USA) was dissolved in 0.1 M HCl, neutralized with 0.1 M NaOH and diluted with
ARTICLE IN PRESS A. Ru¨ck et al. / Medical Laser Application 20 (2005) 125–129
PBS. Cells were incubated with 1 mM for 4 h. Uroporphyrin I and III were dissolved in PBS and neutralized with 1 M NaOH. Cells were incubated with 10 mM for 24 h. Cultures of RR 1022 epithelial cells were grown in M199 medium (Gibco, UK) supplemented with 10% fetal calf serum (FCS) and antibiotics (penicillin, streptomycin) at 37 1C and 5% CO2. The cells were seeded on glass bottom microwell dishes, coverglass 0.16–0.19 mm (MatTek, USA) at a density of 75–100 cells/mm2 and were allowed to grow for 36 h. The medium was replaced and cells were incubated in
127
the exponential growth phase. Microscopic observation was performed immediately after removing the incubation medium and rinsing twice with indicator free medium at 37 1C.
Results With the described detection setups the time-resolved subcellular fluorescence characteristics of 5-ALA metabolites, Uroporphyrin I and III and PPIX were studied. Fig. 1 demonstrates the fluorescence lifetime image of
Fig. 1. FLIM of RR 1022 cells, incubated with 5-ALA, fluorescence emission 4590 nm, monoexponential fitting.
Fig. 2. FLIM of 5-ALA metabolites, Uroporphyrin I, Uropophyrin III and PPIX.
ARTICLE IN PRESS 128
A. Ru¨ck et al. / Medical Laser Application 20 (2005) 125–129
5-ALA, after an incubation time of 4 h. The lifetime was calculated pixel by pixel by a monoexpontential curve fitting. The blue curve in the diagram correlates with the detected photons, the green curve is the system response and the red curve results from the fitting. The colorcoding of the lifetime image is represented in the histogram. The lifetimes differed significantly within the cells. In principal three different lifetime areas could be detected, one with a long lifetime (blue color), which correlated with the plasma membrane and the outer cytoplasm, one with a middle lifetime (green color) attributed to the mitochondria and cytoplasm of the cells and one with a short lifetime (red color). From photobleaching experiments the areas with the longest lifetimes exhibited the same bleaching kinetics, whereas the compound with the short lifetime was different (data not shown). It seems, that different 5-ALA metabolites (photoproducts or different porphyrins) could be detected within the cells. To support our hypothesis comparative results obtained for 5-ALA and the ‘‘pure’’ porphyrins Uroporphyrin I, Uroporphyrin III and PPIX, are demonstrated in Fig. 2. In all cases the same colorcoding was used. Now, the lifetime was calculated by a biexponential curve fitting. Binning was set to 1, that means, that the lifetime was calculated as the mean from a 3 3 pixel area. The histograms of the mean lifetimes are also presented in Fig. 2. In the case of 5-ALA metabolites at least two different mean lifetime regions could be identified, one with a long lifetime which correlates to the plasma membrane and the outer cytoplasm (represented in blue color), one with a shorter lifetime attributed mainly to the cytoplasm of the cells (green color). Some cells even exhibited much shorter lifetimes (orange color). When comparing the lifetime image of 5-ALA with Uroporphyrin I, Uroporphyrin III, as well as PPIX, it seems, that the long lifetime could be attributed to PPIX, whereas the shorter lifetimes contain to some extend Uroporporphyrin I. Some of the Uroporphyrin III seems to be metabolized to PPIX. The aim of our study was to show, whether different 5-ALA metabolites are produced within the cells, depending on cell cycle and cell differentiation. Because it is known, that phototoxic efficiency of the porphyrins varies significantly, a clear discrimination would help to understand the sometimes unwanted different cellular response of tumor cells. For this, FLIM is a powerful method and a challenge to improve PDT.
Discussion This work describes time-resolved fluorescence techniques and applications for PDT. The advantage of
using TCSPC and pulsed laser excitation to improve selectivity of fluorescence guided diagnosis is due to the unlimited dynamic range associated with photon counting techniques and the easy visualization of fluorescence decays. The disadvantage is a slow detection time, because each photon has to be timed individually. For the practical usage, a fast streak camera with very high temporal resolution would be perfect, however still expensive. The potential of time-resolved detection is evident for the precursor 5-ALA, where the photosensitizing porphyrins are synthesized in the tumor cells. PPIX, which is produced in mitochondria, induces high phototoxicity but also different isomers of Uroporphyrin and Coproporhyrin, which are localized in the cytoplasm of the cells could be important. During cell differentiation various enzymes, which are active in varying degrees control the ALA metabolism. It is therefore of main interest to localize 5-ALA metabolites during cell differentiation and cell growth. Obviously, non-invasive methods should be preferred. Whereas spectral differentiation is very difficult, due to highly overlapping emission spectra of the porphyrins involved, we demonstrated on a cellular level that FLIM is able to distinguish the compounds. This technique achieves the features of highly sophisticated lifetime imaging systems for molecular studies in PDT. Whether usage in the clinical practice becomes acceptable depends at least on the availability of easy and fast detection systems with resonable prices. On the excitation site, relatively cheap laser diodes or simple custommade Ti:Saphir laser systems are now on the market.
Acknowledgements This work is carried out by the Ministry of Economic Affairs of the state of Baden-Wu¨rttemberg by means of the Landesstiftung Baden-Wu¨rttemberg, order 44332.62-ILM/15.
Zusammenfassung Zeitaufgelo¨ste Fluoreszenz und PDT Im Rahmen der photodynamischen Therapie ist die laserinduzierte Fluoreszenzdiagnostik ein wichtiges Hilfsmittel, um Tumorbegrenzungen oder Tumorreste sichtbar zu machen und so eine vollsta¨ndige Resektion zu gewa¨hrleisten. Das Problem jedoch ist neben der Sensitivita¨t eine ha¨ufig zu geringe Selektivita¨t, die besonders im Bereich von Hirntumoren zu unbefriedigenden Ergebnissen fu¨hrt. Um die Selektivita¨t zu verbessern, bieten sich zeitaufgelo¨ste Methoden an, mit deren Hilfe es mo¨glich wird, verschiedene Fluorophore und Metaboliten eindeutig zu unterscheiden. Neu und
ARTICLE IN PRESS A. Ru¨ck et al. / Medical Laser Application 20 (2005) 125–129
interessant ist in diesem Sinne die zeitkorrelierte Einzelphotonenza¨hlung, auch time-correlated single photon counting (TCSPC) genannt, die in Kombination mit einer Scanningeinheit und kurzgepulster Anregung die bildgebende Darstellung der Fluoreszenzabklingzeit (sogenanntes fluorescence lifetime imaging (FLIM)) ermo¨glicht. An Hand der fu¨r die Klinik zugelassenen 5-Aminolevulinsa¨ure (5-ALA) la¨sst sich das Potential zeitaufgelo¨ster Fluoreszenzmessungen besonders gut demonstrieren. Nach der Applikation von 5-ALA entstehen in den Zellen unterschiedliche Porphyrine, die alle mehr oder weniger stark phototoxisch sind. Protoporphyrin IX, das in den Mitochondrien synthetisiert wird, ist dabei das wirksamste Porphyrin. Es ko¨nnen jedoch auch andere Metaboliten, wie Uroporphyrin und Koproporphyrin entstehen. Die Synthese dieser Porphyrine wird durch Enzyme gesteuert, die in Abha¨ngigkeit des Diffenzierungsgrades der Zelle mehr oder weniger aktiv sind. Wie im Rahmen dieser Arbeit diskutiert wird, ko¨nnen durch FLIM die verschiedenen Metaboliten mit subzellula¨rer Auflo¨sung unterschieden werden. Zeitaufgelo¨ste Methoden ko¨nnen damit allgemein die Selektivita¨t der Fluoreszenzdiagnostik erho¨hen und damit die PDT signifikant verbessern. r 2005 Elsevier GmbH. All rights reserved. Schlu¨sselwo¨rter: PDT; 5-ALA; Laserscanning Mikroskopie; Gepulste Diodenlaser; TCSPC; Zeitaufgelo¨ste Fluoreszenz (FLIM); Spektral- und zeitaufgelo¨ste Fluoreszenz (SLIM)
References [1] Phillips D. Luminescence lifetimes in biological systems. Analyst 1994;119:543–50. [2] Schneckenburger H, Ko¨nig K, Kunzi-Rapp K, WestphalFro¨sch C, Ru¨ck A. Time-resolved in-vivo fluorescence of photosensitizing porphyrins. J Photochem Photobiol B: Biol 1993;21:143–7. [3] Schneckenburger H, Gschwend MH, Sailer R, Ru¨ck A, Strauß WSL. Time-resolved pH dependent fluorescence of hydrophilic porphyrins in solution and in cultivated cells. J Photochem Photobiol B: Biol 1995;27:251–5. [4] Strauß WSL, Sailer R, Schneckenburger H, Akgu¨n N, Gottfried V, Chetwer L, Kimel S. Study of the photodynamic efficacy of naturally occurring porphyrins in
[5]
[6]
[7]
[8] [9] [10]
[11]
[12]
[13]
[14]
[15]
[16]
129
endothelial cells in vitro and microvasculature in vivo. J Photochem Photobiol B: Biol 1997;39:176–84. Connelly JP, Botchway SW, Kunz L, Pattison D, Parker AW, MacRobert AJ. Time-resolved fluorescence imaging of photosensitiser distributions in mammalian cells using a picosecond laser line-scanning microscope. J Photochem Photobiol A: Chemistry 2001;142:169–75. Cubeddu R, Canti G, Pifferi A, Taroni P, Valentini G. Fluorescence lifetime imaging of experimental tumors in hematoporphyrin derivative-sensitized mice. Photochem Photobiol 1997;66:229–36. Kelbauskas L, Dietel W. Internalization of aggregated photosensitizers by tumor cells: subcellular time-resolved fluorescence spectroscopy on derivatives of pyropheophorbide-a ethers and chlorin e6 under femtosecond oneand two-photon excitations. Photochem Photobiol 2002; 76:686–94. Ko¨nig K. Multiphoton microscopy in life science. J Microscopy 2000;200:83–104. Wahl M, Kaenders W. Blue diode lasers light up in pulsed and tuneable output systems. EuroPhotonics 1999;4:52–4. Knemeyer JP, Herten DP, Sauer M. Detection and identification of single molecules in living cells using spectrally resolved fluorescence lifetime imaging microscopy. Anal Chem 2003;75:2147–53. Kress M, Meier T, Steiner R, Dolp F, Erdmann R, Ortmann U, Ru¨ck A. Time- resolved microspectrofluorometry and fluorescence lifetime imaging of photosensitizers using ps pulsed diode lasers in laser scanning microscopes. J Biomed Opt 2003;8(1):26–32. Ru¨ck A, Dolp F, Scalfi-Happ C, Steiner R, Beil M. Timeresolved microspectrofluorometry and fluorescence lifetime imaging using ps pulsed diode lasers and TCSPC techniques in laser scanning microscopes. Proc SPIE 2003;5139:166–72. Bissonette R, Bergeron A, Liu Y. Large surface photodynamic therapy with aminolevulinic acid: treatment of actinic keratoses and beyond. J Drugs Dermatol 2004; 3(1):26–31. Ickowicz Schwartz D, Gozlan Y, Greenbaum I, Babushkina T, Katcoff DJ, Malik Z. Differentiation-dependent photodynamic therapy regulated by porphobilinogen deaminase in B16 melanoma. Br J Cancer 2004;90(9): 1833–41. Ru¨ck A, Dolp F, Hu¨lshoff C, Hauser C, Scalfi-Happ C. FLIM and SLIM for molecular imaging in PDT. Proc SPIE 2005:5700–32 in press. Becker W, Bergmann A, Biscotti G, Koenig K, Riemann I, Kelbauskas L, Biskup C. High-speed FLIM data acquisition by time-correlated single photon counting. Proc SPIE 2004;5323:27–35.