Monitoring Cellular Metabolism with Fluorescence Lifetime of

Mar 20, 2009 - Part of the “Hiroshi Masuhara Festschrift”. , * To whom correspondence should be addressed. Current address: Institute of Biophoton...
0 downloads 0 Views 5MB Size
11532

J. Phys. Chem. C 2009, 113, 11532–11540

Monitoring Cellular Metabolism with Fluorescence Lifetime of Reduced Nicotinamide Adenine Dinucleotide† Vladimir V. Ghukasyan‡ and Fu-Jen Kao*,§ Institute of Biophotonics, National Yang-Ming UniVersity, 155, Li-Nong Street, Section 2, Taipei 11221, Taiwan, and Department of Photonics, National Sun Yat-sen UniVersity, Kaohsiung 80424, Taiwan ReceiVed: December 11, 2008; ReVised Manuscript ReceiVed: January 25, 2009

Formulation of oxidative phosphorylation and its first observation by means of fluorescence spectroscopy in the 1960s led to the acceptance of bioenergetics as a new field of studies. The new discipline grew fast with the increasing number of papers, related to the energy generation in mitochondria, advancement of the instrumentation, and improvement of observation techniques. As such, fluorescence lifetime imaging microscopy (FLIM) has gained popularity as a sensitive technique to monitor the functional/conformational states of nicotinamide adenine dinucleotide reduced (NADH)sone of the main compounds of oxidative phosphorylation. We hereby review the development and current application of cellular metabolism observation via NADH FLIM, illustrating it with the examples of both physiological (cell density, apoptosis, necrosis) and pathological states (inhibition of the electron transfer chain). Introduction As a minimum invasive technique, autofluorescence spectroscopy has found its application in the characterization of tissues1-6 and diagnosis of skin, cervix, bladder, breast, stomach, and other cancers, Alzheimer disease, etc.7-13 The major distinguishing power arises from the change of the nanoenvironment, quantities, and functional/conformational states of the endogenous fluorophoressmainly represented with aromatic aminoacids Trp, Tyr, and Phe, vitamin derivatives retinol, riboflavin, and reduced nicotinamide adenine dinucleotide (NADH), as well as pyridolamine cross-links in elastin and some collagens.5,6,14-18 A number of techniques have been applied to exhibit these changes and characterize the correlation between the fluorescence properties of these compounds and pathological processes in tissues.19-21 Introduction of multiphoton microscopy22-25 has allowed the overcoming of some difficulties in autofluorescence spectroscopy based on single-photon excitation, such as photobleaching,26,27 photodamage,28,29 significant light scattering, and absorption in tissues.30-32 Within the working spectrum range of the most advantageous multiphoton Ti:Sapphire laser (700-1000 nm), the two-photon excitation cross sections are maximal at 700-750 nm.33 The majority of information derived at this spectrum for the most tissues originates from NADH and collagens with the former observed from intracellular and the latter from extracellular space. Previously known as diphosphopyridine nucleotide or pyridine nucleotide, NADH has a single-photon excitation peak at ∼340 nm, with the fluorescence spectrum centered at ∼460 nm. Identical to those of NADH, excitation and emission spectra are exhibited also by nicotinamide adenine dinucleotide phos†

Part of the “Hiroshi Masuhara Festschrift”. * To whom correspondence should be addressed. Current address: Institute of Biophotonics, National Yang-Ming University, 155, Li-Nong Street, Section 2, Taipei 11221, Taiwan. E-mail: [email protected]. Tel.: +886 2 28267336. Fax: +886 2 28235460. ‡ National Yang-Ming University. § National Sun Yat-sen University.

phate (NADPH), and thus cellular autofluorescence excited in the region of ∼340 nm is often termed NAD(P)H fluorescence, as the two species cannot be distinguished via their fluorescence excitation and emission. This species differs from NADH by an additional phosphate group and is normally assumed to be constant with respect to metabolic perturbations; moreover, the intensity of the NADPH is much lower than that of NADH. Its impact is usually considered insignificant.34 NADH serves as a coenzyme and a principal electron donor within the cell for both oxidative phosphorylation (aerobic respiration) and glycolysis (anaerobic respiration). The molecule exists in two functional forms: free and bound, whereas the latter is associated mostly with the dehydrogenases of so-called complex Isone of four mitochondrial membrane protein complexes, which mediate electron transfer from NADH to O2 and use this flow to pump the hydrogen protons to the mitochondrial intermembrane space from the matrix. This gradient of protons and electrical potential, termed proton-motive force, is utilized to synthesize new adenosine triphosphate (ATP) molecules at ATP synthase via adenosine diphosphate (ADP) phosphorylation.35 Thus bound forms can be associated with the energy generation in the form of ATP, and the relative quantities of free and bound species can give an insight to the metabolical state of a cell. Moreover, it has been pointed out that the reaction velocity of a given intracellular NADH-linked dehydrogenase depends on the concentration of locally available NADH, i.e., the local concentration of free NADH.1 The discovery of the fluorescence properties of pyridine nucleotides by Theorell and Chance36 in the early 1950s led to a whole new field of studies: an insightful application of fluorescence spectroscopy to the processes in live mitochondria in vitro and, as the technologies improved, in vivo. The discovery led to an avalanche in mitochondrial bioenergetics studies mainly via accumulation of the fluorescence signal and its further filtering. Dual-wavelength spectrometry was used in isolated mitochondrial preparations to first associate the absorption and fluorescence of NADH and flavin adenine dinucleotide (FAD) with the mitochondrial respiratory chain. The first results

10.1021/jp810931u CCC: $40.75  2009 American Chemical Society Published on Web 03/20/2009

Fluorescence of Reduced NADH

J. Phys. Chem. C, Vol. 113, No. 27, 2009 11533

Figure 1. NADH free/bound ratio mapping at different segments of a HeLa cell colony in a culture (taken on the fifth day of cell culture growth). The color coding reveals lower values of the a1/a2 ratio for the cells at the center (left) and higher values for the cells at the edge (right) of a colony possibly indicating higher and lower metabolic activity, correspondingly. The scale bar is 100 µm. Data acquisition time: 700 s.

in this discipline were summarized by Chance and Williams, laying the grounds for the basics on mitochondrial function.20 However, these methods had complications. The substantial overlapping of the endogenous spectra makes it difficult to distinguish between individual components. Additionally, autofluorescence spectroscopy techniques do not provide a good contrast between free and enzyme-bound forms of NADH and thus do not give an insight into detailed molecular dynamics of the endogenous fluorophores. These issues were addressed with another technique of high sensitivity and nanoresolutionsfluorescence lifetime imaging microscopy (FLIM). FLIM is based on the measurement of the average time fluorescent molecules spent in the excited states. Observations of the changes in the fluorescence lifetime parameters give an insight into a range of dynamic processes, such as conformation, interaction with solvent and other molecules in the system, concentration of ions in the surrounding media, and changes in refractive index, viscosity, and pH of the environment, etc.37 The first demonstration of FLIM applicability to NADH imaging has been demonstrated by Lakowicz et al. by imaging the free and malate-dehydrogenase-bound forms of NADH in solution,35 whereas the first application of fluorescence decay measurements from NADH via pulse fluorometry and its power to distinguish free and bound species date back to late 1970s.38-41 In contrast to spectroscopy, which poorly distinguishes between the functional/conformational states of NADH (free vs bound), these exhibit a significant difference in terms of their fluorescence decays. Free NADH in aqueous solution at room temperature exhibits a biexponential fluorescence decay with fluorescence lifetime components of ∼0.3 and ∼0.7 ns and a mean fluorescence lifetime of ∼0.4 ns.22,38,41 Protein-bound NADH also exhibits a biexponential fluorescence decay, and the shorter-lifetime component can be comparable to that of the long-lifetime component of free NADH. The fluorescence lifetime of proteinbound NADH depends on the enzyme to which it is bound, and this suggests that it can be probed by the lifetime of this fluorophore. The main reason for the difference between the functional states of the coenzyme is self-quenching of the nicotinamide by the adenine moiety in the free form.40 Therefore, the common practice of fitting a biexponential decay profile to

a mixture of free and protein-bound NADH exhibiting four decay components (or more, with multiple bound proteins) is a simplification of the underlying dynamics, 42 which, however, has proven its applicability and coherence with the results of the biochemical analysis. Aside from the biochemical analysis, which requires extraction of pyridine nucleotides, fluorescence lifetime is the only technique currently providing a noninvasive way to assess the free/bound ratio of NADH, whereas the short lifetime at the range of (∼300-700 ps) is usually attributed to the free, and the long lifetime (∼2500-3000 ps) to the bound, form of NADH.35 NADH FLIM has been demonstrated in a range of applications, such as sensing and diagnostics. Thus, FLIM has enriched previous observation of NADH fluorescence intensity increase at high glucose concentration in cell environment as it is reduced along with the gradual oxidation of the glucose molecules.43,44 In the experiment conducted, addition of 30 mM of glucose in vitro to the fibroblasts and adipocytes resulted in an overall decrease of average lifetime due to the decrease of the relevant amplitude of bound forms and thus increase of free/bound ratio. One of the fastest growing fields of NADH FLIM is cancer diagnostics. One of the hallmarks of carcinogenesis is a shift from cellular oxidative phosphorylation to cellular glycolysis for ATP production (the Warburg effect),8 which makes NADH FLIM a method of choice to observe this dynamics as the shift to glycolysis should result in a relative increase of the free species. Thus, a difference in fluorescence lifetimes measured from the NADH in normal, cancer metastatic, and cancer nonmetastatic cells has been demonstrated for human and rat cells, whereas the best contrast was observed for the nonmetastatic cells, and small difference without well-expressed trends between normal and metastatic cells.45 The capacity of the NADH FLIM has also been demonstrated in vivo with comparison characterization between low-grade precancer (mild to moderate dysplasia) and high-grade precancers (severe dysplasia and carcinoma in situ). The imaging, made on the hamster cheek pouch model of oral carcinogenesis, revealed a complicated dynamics of NADH lifetime dynamics. Thus, statistically significant decrease of the bound NADH lifetime and its relative amplitude is observed for the low-grade precancers. However, as the cells progress to high-grade pre-

11534

J. Phys. Chem. C, Vol. 113, No. 27, 2009

Ghukasyan and Kao

cancers, the lifetime of bound forms of NADH and its relative amplitude increased back, becoming similar to that of normal cells.46 The results obtained show that the Warburg effect applicability for the distribution of lifetimes and relative amplitudes might not be straightforward, and further studies are required to reveal the fine correlation between the changes in cells’ biochemistry undergoing the cancerogenic pathway and the dynamics of NADH lifetimes. With the wide spectrum of NADH FLIM studies being performed currently, its capacities as metabolism level indicator is still underexplored from one side, and proper characterizations of normal physiological states in terms of fluorescence lifetime dynamics as well as the latter’s correlation to basic biochemical events still have to be done. We hereby illustrate the capacity of FLIM NADH to properly characterize the metabolic state of normal and pathological states of a cell with the application of the technique to reveal the dynamics of NADH lifetime at different cell density, cell death pathway, and inhibition of mitochondrial electron transport chain (ETC). Experimental Section Imaging. Time-domain FLIM was performed on a modified two-photon laser scanning microscope (FV300 with the IX71 inverted microscope, Olympus Co.) as described in details previously.47 Images were acquired with a 60× 1.45NA PlanApochromat oil objective lens (Olympus Co.). In this study, samples were excited at 750 nm (two-photon) by a mode-locked Ti:Sapphire Mira F-900 laser, pumped by a solid-state frequencydoubled 532 nm Verdi laser (both from Coherent Inc., CA). The scanning speed of the FV300 can be controlled externally by a function generator (AFG310, Tektronix Inc., Beaverton, OR) to optimize the image acquisition. Fluorescence photons were detected in a nondescanned mode by a cooled photoncounting photomultiplier (H7422P-40, Hamamatsu Photonics K.K., Hamamatsu, Japan). Time-resolved detection was conducted by the single-photon-counting SPC-830 PC board (Becker & Hickl GmbH, Berlin, Germany).48-57 A band-pass filter of 447 ( 30 nm (Semrock, NY) was used to match the spectral characteristics of NADH’s autofluorescence.22 Additional short-pass and IR cutoff filters were used to reject the excitation light at 750 nm. The average laser power used at the focal plane of the objective was kept within ∼4 mW, which was lower than the reported laser power of two-photon damage and was found optimal for the prevention of photobleaching.27-29 For the observation of metabolic changes at normal physiological conditions, the morphology of the cells was monitored during the measurements to make sure that cells were not affected by the laser power applied during the lifetime data collection period. For the same set of experiments, little to no changes in photon count rate have been detected during the measurement indicating insignificant if any photobleaching for the given parameters. All the images were taken at 256 × 256 pixels resolution with the acquisition time in the range of 700-900 s for accumulating enough photon count statistics at the given laser power for further data analysis. Data Analysis. Data was analyzed with the commercially available SPCImage v2.8 software package (Becker & Hickl GmbH, Berlin, Germany) via a mathematical convolution of a model function with the instrument response function (IRF) and fitting of the model function to the experiment data. To calculate the lifetime from the composite decays of NADH, we convolved an IRF, Iinstr, with a double-exponential model function, defined

Figure 2. Dynamics of NADH average fluorescence lifetime after the treatment of 143B cells with 1 mM STS. The average lifetime increases within the first 15 min after addition to 3359 ps with gradual decrease for all subsequent time-lapse measurements down to 1720 ps on the given series.

in eq 1, with offset correction for the ambient light and/or dark noise I0, to obtain calculated lifetime decay function Ic(t) in eq 2.

F(t) ) a1e-t/τ1 + a2e-t/τ2 Ic(t) )

∫-∞∞ Iinstr(t){I0 + F(t)} dt

(1) (2)

Here a1e- t/τ1 and a2e- t/τ2 represent the contributed fluorescence decays from short- and long-lifetime compounds of NADH, respectively, τ1 and τ2 are their corresponding lifetimes; a1 and a2 are the corresponding relative amplitudes. Iinstr was measured experimentally from the PPLN crystal at 370 nm (the second harmonic of 740 nm from the Ti:Sapphire laser). The decay fwhm thus obtained was equal to ∼330 ps.

Fluorescence of Reduced NADH

J. Phys. Chem. C, Vol. 113, No. 27, 2009 11535

Figure 3. Distribution of the χR2 of NADH lifetime fitted with the single- and double-exponential model at the time-lapse observation of 143B cells, treated with 1 mM STS. The data, collected immediately after the treatment (0 min), could be well fitted with a single-exponential model with no further improvement of the χR2 at double-exponential fitting. The single-exponential model still fits well to the measurement at the 15th minute after treatment; however, a tail to the longer range exhibits an increased number of pixels, badly fitted. For all the subsequent measurements, the χR2 of the single-exponential fitting is growing gradually with the tail extending further, whereas the double-exponential fitting is improving, thus showing the increasing number of pixels exhibiting double-exponential decay.

The average lifetime was calculated as an amplitude-weighted of the two lifetime components:

τa )

a1τ1 + a2τ2 a1 + a2

(3)

The model parameters (i.e., ai and τi) were derived by the fitting the calculated decay Ic(t), defined in eq 2, to the actual data Ia(t) through minimizing the goodness-of-fit χR2 function defined in eq 3 using the Levenberg-Marquardt search algorithm. n

χR2 ) [

∑ [Ia(t) - Ic(t)]2/Ia(t)]/(n - p)

(3a)

k)1

Here n is the number of the data (time) points (equal to 256 in this study), and p is the number of the model parameters. τa, or their weighted average, can be used to indicate the status of metabolism, as in the case of apoptosis (Figures 2 and 3).58 To reflect the metabolism change, however, we found that the ratio of a1 and a2 is best in indicating the free and protein-bound states of NADH. All the measurements were taken at physi-

ological conditions (37 °C, 5% CO2) maintained by the microscope stage incubator (H-201, Oko-Laboratory, Naples, Italy). Sample Preparation. Cells were grown on 22 mm acidwashed round glass coverslips, placed in 6-well plates filled with growth media. For the imaging sessions, the coverslips were mounted on metallic rings and covered with a plastic chamber, suited to hold 1 mL of growth media. Results and Discussion Observation of Physiological States. With the capacity to report on the cellular metabolic state noninvasively, the NADH FLIM is able to properly characterize both physiological and pathological states of live cells. This property of the technique is rather important to understand the coherence of the NADH lifetime dynamics with the biochemical processes in the cell and to be able to distinguish the lifetime perturbations in norm from that specifically caused by pathology. In particular, all conditions of the experimental setup that might affect the energy metabolism should be assessed in terms of its influence on fluorescence lifetime.

11536

J. Phys. Chem. C, Vol. 113, No. 27, 2009

The influence of the cell density on the energy metabolism has been studied earlier.59 Several approaches have been taken and several hypotheses ruled out to explain the outcomes of the measurements. Thus, lifetime has been assessed for the different stages of cell culture growth by plating the cells in the concentrations, corresponding to the early, midlogarithmic, and confluent points on the cellular growth curve. The researchers have observed a significant decrease of the average fluorescence lifetime of both free and bound NADH species from the initial to confluent concentrations.16 At this, the ratio of free to protein-bound NADH (a1:a2) increased by more than a factor of 2 from the early to confluent phase of the growth curve. Such an effect has been explained by the authors with the decrease of the oxygen levels in cells, thus limiting the oxidative phosphorylation and shift of the energy metabolism to cytoplasm glycolysis. Similar effects are observed upon the addition of the mitochondrial ETC inhibitor potassium cyanide (KCN), which inhibits complex IV of the ETC, thus interrupting the electrons flow from NADH to the O2 and mimicking hypoxia. In our studies a different approach has been taken. By plating of the cells at small inoculum (2 × 103/cm2) and further daily measurements of the NADH lifetime and free/ bound ratio, different results have been obtained as the cell culture grew. At least five measurements have been taken each day. Right upon plating the cells on coverslips, they exhibited a relatively large statistical variance in terms of the values of both lifetimes (free τ1 and τ2 bound) and their fractions ratio (a1:a2). During the initial stage of recovery and restoration of normal metabolic levels, the free/bound ratio increased a little from 3.56 ( 0.35 measured 12 h upon plating to 3.8 ( 0.25, recorded on the third day. Further proceeding to the exponential growth results in gradual decrease of the ratio daily down to 3.1 ( 0.14 measured on the sixth day. Upon reaching the confluency, the cell culture growth slope decreases, and a gradual increase back to higher values (3.25 ( 0.17 and 3.69 ( 0.28 on the seventh and eighth days, respectively) is obtained. Surprisingly, the results differed for the NADH signal measured from the cells in the center and at the edge of cell colonies. At the edge of such colonies, the free/bound ratio was higher than the value recorded from the cells in the center (3.59 ( 0.49 at the edge versus 3.11 ( 0.14 in the center measured on the sixth day of the culture growth) (Figure 1). These results imply that the energy metabolism dynamics at different cell densities along cell culture growth cannot be solely attributed to the lack of oxygen. We assume that the hypoxic condition might be the reason for the slight increase of the free/bound ratio observed at the confluency, whereas the decrease of the ratio during the exponential growth might result from the different cell cycle stage prevailing at initial, exponential, and confluency stages. The well-expressed difference in the metabolic state between the cells squeezed in the center and well spread at the edge of cell colonies leads us to the hypothesis that the dynamics observed might be also dependent on the cell morphology and thus cytoskeleton state.57 Apoptosis and Necrosis. Noninvasive detection of cell death is important for the assessment of a range of cancer therapies. In photodynamic therapy (PDT), a combination of topically or systematically applied photosensitizer drug with further delivery of light power is applied to kill target cells. Despite the photosensitizers’ target organelles varying depending on the exact chemical agent, a massive destruction to any of the cellular organelles through the rapid production of reactive oxygen species (ROS), mainly singlet oxygen, at excitation with light leads to cellular death through either apoptosis or necrosis. This

Ghukasyan and Kao

Figure 4. Dynamics of the free NADH relative fraction (a1) at the time-lapse observation of 143B cells, treated with 1 mM STS. Taken from the 30th minute after treatment, the a1 exhibits gradual increase toward higher values. Despite the fact that the original values are not reached, the dynamics are well expressed. The scale bar is 100 µm.

type of therapy has been approved by the U.S. Food and Drug Administration (FDA) for the treatment of a number of cancers, in particular, late-stage esophageal cancer and nonhyperkeratotic actinic keratosis. Successful clinical trials have been conducted also for brain tumors, breast cancers, etc. The outcome of the therapy requires optimization, especially in dosimetry of the photosensitizer, as it can result in different cell death pathways. The latter might be critical for PDT treatment: as opposed to the least harmful for the organism “normal” programmed cell death (apoptosis), cells dying of extremely high doses of ROS via pathological necrosis burst their contents to the environment, causing inflammatory reactionsin some cases, massive. One of the first events both at apoptosis and necrosis is loss of mitochondrial function; thus, observation of fluorescent participants to the oxidative phosphorylation might be a straightforward way to monitor the efficiency of therapy. In the earlier works of Pogue addressing this problem, a difference in NADH fluorescence lifetime, measured in vivo from mouse leg muscle, treated with verteporfin (a lipid formulation of benzoporfirin derivative monoacid ring “a”), exhibited an average lifetime of 1.5 ( 0.2 ns, whereas untreated cells had a characteristic 2.1 ( 0.5 ns.60 The observers stressed that at the light delivery no significant change of the lifetime within experimental error has been observed, whereas the total fluorescence intensity, measured at 450 ( 25 nm (NADH band), has been reported as decreasing. Similar effect has been observed in vitro by Wang et al.58 for cells undergoing necrosis, whereas different dynamics of NADH lifetime was detected for

Fluorescence of Reduced NADH

J. Phys. Chem. C, Vol. 113, No. 27, 2009 11537

Figure 5. Dynamics of the (a) free NADH relative fraction (a1) and (b) NADH average lifetime after HeLa cells treatment with 1 mM H2O2. No dramatic decrease similar to that observed upon treatment with the apoptosis inducer is observed. Slight decrease of the peak value from 80% to 70% is observed at some of the time-lapse images. However, no trend could be traced due to the gradual decrease of the fluorescence intensity. The same is true for the average lifetime (b). The lifetime is changing slightly around ∼1500 ps. For the sake of color-coding, the intensities of the images have been adjusted to be visible. The scale bar is 100 µm.

apoptosis. In this work the difference between different cell death pathways has been assessed on cervical carcinoma (HeLa) and osteosarcoma (143B) cell cultures. Cells were treated with staurosporine (STS), a well-known apoptosis inducer, and hydrogen peroxide (H2O2), which is known to cause necrosis at high concentrations. Upon the addition of 1 µM STS, cells exhibit a single-exponential decay with the rapid and dramatic increase of the average lifetime from 1.36 ( 0.24 ns to 3.6 ( 0.56 ns for HeLa and from 1.26 ( 0.49 ns to 3.45 ( 0.76 ns for 143B cells (Figure 2). For both a dramatic increase of the fluorescence has been observed as well. Here we demonstrate that the dynamics observed were caused by the changes of the free and bound species relative fractions. The lifetime shift along with data fitted well with a singleexponential model suggesting that the free NADH fraction is diminished to a great extent immediately after the STS addition. At this, an interesting dynamics has been observed (Figures 2-4): the χ2Rof the data, fitted with the single-exponential model, gradually rose with time, exhibiting restoration of the free NADH pool, whereas the plot of the χ2R distribution on the image

revealed a growing tail of the parameter to the higher values (Figure 3). Correspondingly, the a1 exhibited gradual increase (Figure 4), thus moving the average lifetime to the lower values at the time-lapse imaging. For the sample, depicted in Figure 4, a1 reaches the 55% value on the 10th hour of observations. Despite the fact that even after several hours the initial values of the average lifetime and a1 ratio were not observed, the effect was diminished to a great extent. The same was observed for the NADH fluorescence intensity: right after the initial increase further slow decrease has been observed. At the same time, very little change in the coenzyme’s lifetime has been observed in necrosis tests. At 1 mM H2O2 the average lifetime measured from the HeLa cell culture was changing from 1.36 ( 0.24 ns to 1.51 ( 0.51 ns, whereas all the following time-lapse measurements revealed the same result with only the intensity of fluorescence dropping (Figure 5). It should be noted, however, that earlier studies of the NADH fluorescence intensity at necrosis in yeasts revealed a rapid increase immediately upon the death induction, only then followed by a gradual quenching of the cell autofluorescence.61 This implies that there may exist

11538

J. Phys. Chem. C, Vol. 113, No. 27, 2009

Ghukasyan and Kao

Figure 6. Influence of the ETC inhibitors on the dynamics of NADH fluorescence lifetime. Influence of 20 mM KCN on (a) HeLa and (b) WI-38 cells. (c) Influence of 1 µM rotenone on HeLa cells. The scale bar is 100 µm.

some fast dynamics of NADH autofluorescence lifetime at necrosis as well, which, however, could not be detected due to the long accumulation times and is a subject of further studies. In any case, the approach clearly demonstrates that FLIM is well suited for distinguishing between the cell death pathways. Moreover, the ability to distinguish between the free and bound species of NADH supposedly gives an insight into the energy metabolism dynamics in both cases. Pathological States. We have also explored the applicability of the autofluorescence lifetime imaging as a diagnostics tool of the mitochondrial respiratory chain defects, which are the main cause for a range of diseases, such as myopathies and neurodegenerative disorders. Two different inhibitors have been used for the given demonstration: complex I, the immediate binding site of the NADH has been inhibited by rotenone; potassium cyanide (KCN) has been used to interrupt the electrons flow to the final O2 via the inhibition of complex IV. Cells, cervical carcinoma HeLa, and fibroblasts WI-38 have been plated on coverslips at 3 × 104 cells per 24 mm L coverslip one day prior to measurements. For the observation the coverslips have been mounted on a metal ring chamber with the further addition of DMEM. For the given experiment, after taking the control images, the normal growth media was replaced with 1 mL of DMEM containing 20 mM potassium

cyanide KCN (Figure 6a, b) and 1 µM (2R,6aS,12aS)1,2,6,6a,12,12a-hexahydro-2-isopropenyl-8,9-dimethoxychromeno[3,4-b]furo(2,3-h)chromen-6-one (rotenone) (Figure 6c) with further time-lapsed observation under physiological conditions with the interval of 15 min. The measurements revealed different patterns of the NADH fluorescence lifetime dynamics at the inhibition of complex I and complex IV. In both cases first a rapid and for some cells a dramatic increase in the free/bound ratio is observed, with further decrease over a period of time. Here different dynamics is observed for KCN and rotenone. Upon treatment with KCN, the decrease goes back to the original values, whereas in the case of rotenone the free/bound ratio shows just small deviations around a single value, higher, however, than the original one in the nontreated cells. Despite the fact that the details of the inhibitors’ influence on the cell metabolism as to the dosedependence and adaptive (short-term) and long-term outcomes of the ETC inhibition, etc., are the subject of further studies, the dynamics shown demonstrate that the perturbations to the single members of the ETC can be observed and that some reparation mechanism of the inhibition exists. Future Developments. We believe that the NADH FLIM has further capacity still underexplored. Thus, a proper characterization of normal physiological states by means of lifetime

Fluorescence of Reduced NADH dynamics still has to be carried out to understand fully and correlate the fluorescence decays of NADH with the biochemistry of the energy metabolism. As such some additional roles of NADH can be pointed out, such as regulation of the carboxylterminal binding protein CtBP, a known participant to the development, cell-cycle regulation, transformation, and transcriptional pathways.60 Combination of the fluorescence lifetime and polarization microscopy may further extend the sensitivity of the technique, providing the information on the changes in viscosity, size, and conformation of the molecules of interest. Thus, this anisotropy capacity may reveal the bound species, which exhibits shorter lifetime and is otherwise indistinguishable.34,48 Many enzymes bind to NADH in the metabolic pathway, and favored metabolic pathways shift with cancer progression, resulting in the Warburg effect. The change in the distribution of NADH binding sites suggests increased significance of the bound NADH observation as a metabolic indicator. Here application of the fluorescence lifetime anisotropy imaging will provide the increased sensitivity and the power to distinguish NADH bound to different enzymes, along with the changes in viscosity that often accompany the hypoxic conditions.38 Additionally, many newly developed techniques in the category of nanobiospectroscopy are expected to further elucidate the roles and mechanisms related to NADH.62 Summary Since the majority of the NADH activity is related to mitochondria, observation of the mitochondrial NADH redox state has a potential for detecting the changes in the cellular metabolic state and thus properly characterizing normal and pathological states. The technique was able to reveal fine changes of the cellular metabolic activity at different cell densities. Moreover, drastic changes in NADH lifetime at apoptosis as opposed to necrosis make the technique a possible tool for the dosimetry adjustments and efficiency of therapy, involving tumor termination. And finally, the technique has a big potential in the diagnosis of a range of diseasesswe demonstrate that the alteration of complexes I and IV of the ETC can be detected and possibly differentiated by NADH FLIM. Acknowledgment. We thank Professor Chi-Hung Lin and Ms. Pei-Yun Ho, Institute of Microbiology and Immunology, National Yang-Ming University, for the kind provision of WI38 cells. We also thank Professor Yau-Huei-Wei and Mr. ShiBei Wu, Institute of Biochemistry, National Yang-Ming University, for providing us the HeLa cells. We also acknowledge the work of Ms. Tatyana Buryakina in metabolic mapping of cell structure growth with FLIM. This work is generously supported by the National Science Council under grants NSC 97-3112-B-010-006 and NSC 96-2112-M-010-001 and the the Ministry of Education under the “Aim for Top University” project. References and Notes (1) Williamson, D. H.; Lund, P.; Krebs, H. A. Biochem. J. 1967, 103 (2), 514–527. (2) Mayevsky, A.; Zarchin, N.; Kaplan, H.; Haveri, J.; Haselgroove, J.; Chance, B. Brain Res. 1983, 276, 95–107. (3) Wu, Y.; Xi, P.; Qu, J. Y.; Cheung, T. H.; Yu, M. Y. Opt. Exp. 2004, 12 (14), 3218–3223. (4) Wu, Y.; Qu, J. Y. Opt. Lett. 2006, 31 (12), 1833–1835. (5) Kunz, W. S.; Gellerich, F. N. Biochem. Med. Metab. Biol. 1993, 50, 103–110.

J. Phys. Chem. C, Vol. 113, No. 27, 2009 11539 (6) Kunz, W. S.; Kunz, W. Biochim. Biophys. Acta 1995, 841, 237– 246. (7) Gulledge, C. J.; Dewhirst, M. W. Anticancer Res. 1996, 162, 741– 749. (8) Warburg, O. The Metabolism of Tumors; Constabel: London, 1930. (9) Benavides, J.; Chang, S.; Park, S.; Richards-Kortum, R.; Mackinnon, N.; MacAulay, C.; Milbourne, A.; Malpica, A.; Follen, M. Opt. Exp. 2003, 11 (10), 1223–1236. (10) Zaakab, D.; Steppb, H.; Baumgartnerb, R.; Schneedeab, P.; Waidelichab, R.; Frimbergerab, D.; Hartmannc, A.; Knchelc, R.; Hofstetterab, A.; Hohlab, A. Urology 2002, 60 (6), 1029–1033. (11) Tadrous, P. J.; Siegel, J.; French, P. M.; Shousha, S.; Lalani, E.N.; Stamp, G. W. J. Pathol. 2003, 199, 309–317. (12) Mayinger, B.; Jordan, M.; Horbach, T.; Horner, P.; Gerlach, C.; Mueller, S.; Hohenberger, W.; Hahn, E. G. Gastrointest. Endosc. 2004, 59 (2), 191–198. (13) Hanlon, E. B.; Itzkan, I.; Dasari, R. R.; Feld, M. S.; Ferrante, R. J.; McKee, A. C.; Lathi, D.; Kowall, N. W. Photochem. Photobiol. 1999, 70 (2), 236–42. (14) Sato, K.; Nishina, Y.; Shiga, K.; Tanaka, F. J. Photochem. Photobiol., B. 2003, 70, 67–73. (15) Reinert, K. C.; Dunbar, R. L.; Gao, W.; Chen, G.; Ebner, T. J. J. Neurophysiol. 2004, 92, 199–211. (16) Bird, D.; Yan, L.; Vrotsos, K. M.; Eliceiri, K. W.; Vaughan, E. M.; Keely, P. J.; White, J. G.; Ramanujam, N. Cancer Res. 2005, 65 (19), 8766– 8773. (17) Foster, K. A.; Beaver, C. J.; Turner, D. A. Neuroscience , 132, 645–657. (18) Berger, F.; Ramirez-Hernandez, M. H.; Ziegler, M. Trends Biochem. Sci. 2004, 29, 111–118, B. (19) Chance, B.; Cohen, P.; Jobsis, F.; Schoener, B. Science 1962, 137, 499–508. (20) Chance, B.; Williams, G. R. J. Biol. Chem. 1955, 217, 395–407. (21) Kann, O.; Kovacs, R.; Njunting, M.; Behrens, C. J.; Otahal, J.; Lehmann, T. N.; Gabriel, S.; Heinemann, U. Brain 2005, 128, 2396–2407. (22) Huang, S. H.; Heikal, A. A.; Webb, W. W Biophys. J. 2002, 82, 2811–2825. (23) Zipfel, W. R.; Williams, R. M.; Webb, W. W. Nat. Biotechnol. 2003, 21 (11), 1369–77. (24) Gratton, E.; Breusegem, S.; Sutin, J.; Ruan, Q.; Barry, N. J. Biomed. Opt. 2003, 8 (3), 381–390. (25) Xu, C.; Webb, W. W. Multiphoton excitation of molecular fluorophores and nonlinear laser microscopy. In Topics in Fluorescence Spectroscopy; Plenum Press: New York, 1997; Vol. 5. (26) Patterson, G. H.; Piston, D. W Biophys. J. 2000, 78, 2159–2162. (27) Tiede, L. M.; Nichols, M. G. Photochem. Photobiol. 2006, 82, 656– 664. (28) Konig, K.; Becker, W.; Fischer, P.; Riemann, I.; Halbhuber, K. J. Opt. Lett. 1999, 24, 113–115. (29) Chen, I.-H.; Chu, S.-W.; Sun, C.-K.; Cheng, P.-C.; Lin, B.-L. Opt. Quant. Electr. 2002, 34, 1251–1266. (30) Schonle, A.; Hell, S. W. Opt. Lett. 1998, 23, 325–327. (31) Liu, Y.; Cheng, D. K; Sonek, G. J.; Berns, M. W.; Chapman, C. F.; Tromberg, B. J Biophys. J. 1995, 68, 2137–2144. (32) Conchello, J. A.; Lichtman, J. W. Nat. Methods 2005, 2, 920–931. (33) Zipfel, W. R.; Williams, R. M.; Christie, R.; Nikitin, A. Yu.; Hyman, B. T.; Webb, W. W. PNAS 2003, 100 (12), 7075–7080. (34) Vishwasrao, H. D.; Heikal, A. A.; Kasischke, K. A.; Webb, W. W. J. Biol. Chem. 2005, 280, 25119–25126. (35) Lakowicz, J. R.; Szmacinski, H.; Nowaczyk, K.; Johnson, M. L. PNAS 1992, 89 (4), 1271–1275. (36) Chance, B.; Baltscheffsky, M Biochem. J. 1958, 68 (2), 283–295. (37) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; Plenum Press: New York, 1999. (38) Gafni, A.; Brand, L. Biochemistry 1976, 15, 3165–3171. (39) Brochon, J. C.; Wahl, P.; Jullon, J. M.; Inatsuba, M. Biochemistry 1976, 15, 3259–3265. (40) Visser, A. J. W. G.; Hoek, A. V. J. Photochem. Photobiol. 1981, 33, 35–40. (41) Wakita, M.; Nishimura, G.; Tamura, M. J. Biochem. (Tokyo) 1995, 118, 1151–1160. (42) Niesner, R.; Peker, B.; Schlu¨sche, P.; Gericke, K.-H. Chem. Phys. Chem. 2004, 5, 1141–1149. (43) Evans, N. D.; Gnudi, L.; Rolinski, O. J.; Birch, D. J. S.; Pickup, J. C. Diabetes Technol. Theor. 2003, 5, 807–816. (44) Evans, N. D.; Gnudi, L.; Rolinski, O. J.; Birch, D. J. S.; Pickup, J. C. J. Photochem. Photobiol. 2005, 80, 122–129. (45) Pradhan, A.; Pal, P.; Durocher, G.; Villeneuve, L.; Balassy, A.; Babai, F.; Gaboury, L.; Blanchard, L J. Photochem. Photobiol. B. 1995, 31, 101–112. (46) Skala, M. C.; Riching, K. M.; Bird, D. K.; Gendron-Fitzpatrick, A.; Eickhoff, J.; Eliceiri, K. W.; Keely, P. J.; Ramanujam, N. J. Biomed. Opt. 2007, 12 (2), 024014.

11540

J. Phys. Chem. C, Vol. 113, No. 27, 2009

(47) Ghukasyan, V.; Hsu, Y. Y.; Kung, S. H.; Kao, F. J. J. Biomed. Opt. 2007, 12, 024016. (48) O’Connor, D. V.; Phillips, D. Time-Correlated Single-Photon Counting; Academic Press: London, 1984. (49) Gadella T. W. J., Jr. Fluorescence Lifetime Imaging Microscopy (FLIM): Instrumentation and Applications. In Fluorescent and Luminescent Probes for Biological ActiVity. A Practical Guide to Technology for QuantitatiVe Real-Time Analysis; Academic Press: New York, 1999. (50) Dale, R. E.; Eisinger, J.; Blumberg, W. E Biophys. J. 1979, 26, 161–194. (51) Majumdara, K.; Hickersonb, R.; Nollerb, H. F.; Clegg, R. M. J. Mol. Biol. 2005, 351 (5), 1123–1145. (52) Becker, W.; Bergmann, A.; Hink, M. A.; Konig, K.; Benndorf, K.; Biskup, C. Microsc. Res. Technol. 2004, 63, 58–66. (53) Becker, W. AdVanced Time-Correlated Single Photon Counting Techniques; Springer: Berlin, Heidelberg, New York, 2005. (54) Becker, W.; Bergmann, A.; Haustein, E.; Petrasek, Z.; Schwille, P.; Biskup, C.; Anhut, T.; Riemann, I.; Konig, K. Proc. SPIE-Int. Soc. Opt. Eng. 2005, 5700, 144–151.

Ghukasyan and Kao (55) Becker, W.; Bergmann, A.; Biskup, C.; Klebauskas, L.; Zimmer, T.; Klocker, N.; Benndorf, K. Proc. SPIE-Int. Soc. Opt. Eng. 2003, 4963, 175–184. (56) Becker, W.; Bergmann, A.; Wabnitz, H.; Grosenick, D.; Liebert, A. Proc. SPIE-Int. Soc. Opt. Eng. 2001, 4431, 249–254. (57) Duncan, R. R.; Bergmann, A.; Cousin, M. A.; Apps, D. K.; Shipston, M. J. J. Microsc. 2004, 215 (1), 1–12. (58) Wang, H. W.; Ghukasyan, V.; Chen, C. T.; Wei, Y. H.; Guo, H. W.; Yu, J. S.; Kao, F. J J. Biomed. Opt. 2008, 13 (5), 054011-1-054011-9. (59) Bereiter-Hahn, J.; Munnich, A; Woiteneck, P. Cell Struct. Funct. 1998, 23, 85–93. (60) Pogue, B. W.; Pitts, J. D.; Mycek, M.-A.; Sloboda, R. D.; Wilmot, C. M.; Brandsema, J. F.; O’Hara, J. A. Photochem. Photobiol. 2001, 74 (6), 817–824. (61) Liang, J.; Wu, W. L.; Liu, Z. H.; Mei, Y. H.; Cai, R. X.; Shen, P. Spectrochim. Acta, Part A 2007, 67, 355–35. (62) Masuhara, H.; Kawata, S.; Tokunaga, F. Nano Biophotonics: Science and Technology; Elsevier: Amsterdam, 2007.

JP810931U