OSA's Digital Library

Virtual Journal for Biomedical Optics

Virtual Journal for Biomedical Optics

| EXPLORING THE INTERFACE OF LIGHT AND BIOMEDICINE

  • Editors: Andrew Dunn and Anthony Durkin
  • Vol. 7, Iss. 10 — Oct. 5, 2012
« Show journal navigation

flatFLIM: enhancing the dynamic range of frequency domain FLIM

Klaus C. Schuermann and Hernán E. Grecco  »View Author Affiliations


Optics Express, Vol. 20, Issue 18, pp. 20730-20741 (2012)
http://dx.doi.org/10.1364/OE.20.020730


View Full Text Article

Acrobat PDF (3059 KB)





Browse Journals / Lookup Meetings

Browse by Journal and Year


   


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools

Share
Citations

Abstract

Fluorescence Lifetime Imaging Microscopy (FLIM) is a quantitative technique to probe the nanoenvironment of fluorescent molecules. It is the most robust way to quantify Förster Resonance Energy Transfer (FRET) as it allows reliable differentiation between concentration changes and quenching. In this way, molecular interactions can be imaged in single living cells. The most common wide-field implementation is homodyne Frequency Domain (FD) FLIM, which determines the fluorescence lifetime by measuring the phase and modulation changes of the fluorescence in each pixel upon excitation with a light source modulated at a high frequency. The fluorescence lifetimes are derived from a stack of images acquired at different phase shifts between excitation and detection. In this work we describe a simple method to enhance the dynamic range of FD-FLIM based on precompensating the expected fluorescence modulation by varying the laser power through the phase stack. We show theoretically and experimentally that most of the dynamic range of the camera can be recovered to quantify cells with different intensities. This improvement can be added to any FD-FLIM setup with minimal modifications, enhancing the throughput of information content.

© 2012 OSA

1. Introduction

Since the first optical observations of micrometer sized organisms, microscopy has been a fundamental tool for cellular studies. From its original incarnation as a simple magnifying glass, many techniques have been developed to exploit different properties of the interaction of light with matter allowing us to quantify the molecular processes that drive the cellular dynamics [1

1. H. C. Ishikawa-Ankerhold, R. Ankerhold, and G. P. C. Drummen, “Advanced fluorescence microscopy techniques-FRAP, FLIP, FLAP, FRET and FLIM.” Molecules (Basel, Switzerland) 17, 4047–132 (2012). [CrossRef]

, 2

2. R. S. Fischer, Y. Wu, P. Kanchanawong, H. Shroff, and C. M. Waterman, “Microscopy in 3D: a biologist’s toolbox.” Trends Cell Biol. 21, 682–91 (2011). [CrossRef] [PubMed]

]. One such technique is Fluorescence Lifetime Imaging Microscopy (FLIM) which allows mapping the fluorescence lifetime properties of fluorophores in single living cells [3

3. F. S. Wouters, P. J. Verveer, and P. I. H. Bastiaens, “Imaging biochemistry inside cells,” Trends Cell Biol. 11, 203–211 (2001). [CrossRef] [PubMed]

]. The fluorescence lifetime can be used as a reporter of the nanoscale environment of a fluorophore as it is affected by conditions such as pH [4

4. S. Ogikubo, T. Nakabayashi, T. Adachi, M. S. Islam, T. Yoshizawa, M. Kinjo, and N. Ohta, “Intracellular pH sensing using autofluorescence lifetime microscopy.” J. Phys. Chem. B 115, 10385–90 (2011). [CrossRef] [PubMed]

], polarity and viscosity[5

5. M. K. Kuimova, G. Yahioglu, J. A. Levitt, and K. Suhling, “Molecular rotor measures viscosity of live cells via fluorescence lifetime imaging” J. Am. Chem. Soc. 130, 6672–3 (2008). [CrossRef] [PubMed]

] and is sensitive to excited state interactions such as Förster Resonance Energy Transfer (FRET) [3

3. F. S. Wouters, P. J. Verveer, and P. I. H. Bastiaens, “Imaging biochemistry inside cells,” Trends Cell Biol. 11, 203–211 (2001). [CrossRef] [PubMed]

, 6

6. T. Förster “Energy migration and fluorescence - 1946” J. Biomed. Opt. 17, 011002–10 (2012). [CrossRef] [PubMed]

].

FRET is a photophysical process by which energy is transferred in a non-radiative manner between nearby (1–10 nm) fluorophores. This length scale is appropriate to monitor key biochemical processes such as molecular conformation and association [7

7. E. A. Jares-Erijman and T. M. Jovin, “FRET imaging.” Nat. Biotechnol. 21, 1387–95 (2003). [CrossRef] [PubMed]

]. While there are multiple methods to measure FRET, FLIM has proven to be one of the most reliable and the only one that allows quantification of the relative concentrations of interacting species. The most common implementation of wide-field FLIM is an intensifier-based frequency-domain (FD) technique [8

8. T. W. J. Gadella, T. M. Jovin, and R. M. Clegg, “Fluorescence lifetime imaging microscopy (FLIM) Spatial resolution of microstructures on the nanosecond time scale,” Biophys. Chem. 48, 221–239 (1993). [CrossRef]

]. In this method the excitation is modulated, which in turn, yields a modulated fluorescence emission superimposed to a concentration dependent constant value. The emission is dephased and demodulated with respect to the excitation due to the time needed for the excited molecule to decay by emitting a photon. Therefore, by measuring the dephase and demodulation the fluorescence lifetime can be obtained.

To measure fluorescence lifetimes in the nanosecond scale typical of biological probes, modulation frequencies from 10 to 100 MHz are required. An intensified camera modulated at the same frequency as the excitation is used to demodulate such signals (Fig. 1(a)). The relative phase between the excitation light modulation and the detector gain modulation is then shifted to acquire a series of images (phase stack) from which the dephase and demodulation can be determined from the intensity profile in each pixel (Fig. 1(b), red markers).

Fig. 1 (a) FLIM setup (b) Principle of intensity compensated FLIM

An important drawback of this technique is the narrowing of the range of concentrations that can be imaged in the same field of view. While in fluorescence microscopy the full intensity dynamic range of the camera can be exploited, in FD-FLIM saturation and underexposure within the FLIM stack effectively reduce the concentration dynamic range. This can lead to an impairment in the analysis of biological systems, when dealing with cells of different brightness in the same field of view. Different methods have been demonstrated to enhance the concentration dynamic range in scanning confocal [9

9. R. A. Hoebe, C. H. Van Oven, T. W. J. Gadella, P.B. Dhonukshe, C. J. F. Van Noorden, and E. M. M. Manders, “Controlled light-exposure microscopy reduces photobleaching and phototoxicity in fluorescence live-cell imaging” Nat. Biotechnol. 25, 249–53 (2007). [CrossRef] [PubMed]

] and wide-field microscopy [10

10. W. Caarls, B. Rieger, A. H. B. De Vries, D. J. Arndt-Jovin, and T. M. Jovin, “Minimizing light exposure with the programmable array microscope” J. Microsc. (Oxford, U. K.) 241, 101–10 (2011). [CrossRef]

]. In these methods, a pre-image of the sample is first acquired to estimate the spatial intensity distribution. This information is then used to image the sample with a compensating spatially modulated pattern. In FD-FLIM, the reduction of the dynamic range arises from the different intensities expected when light emission and detector modulation gain are in phase and counterphase. Here we describe a simple method named flatFLIM that enhances the dynamic range in FD-FLIM measurements. It is based on recovering the lost dynamic range by changing the laser intensity during the phase-stepping (Fig.1(b), blue bars) to counter the modulation (Fig. 1(b), blue markers). We show that using flatFLIM most of the dynamic range of the camera can be recovered for measuring cells with distinct expression levels.

2. Theory

2.1. Dynamic range of FD-FLIM

In frequency domain FLIM, the detected fluorescence intensity (F) in each pixel can be expressed as a function of the relative phase between excitation and detection (ϕi):
F(ϕi)=IC(1+m0mτcos(ϕiϕ0ϕτ)),
(1)
where I is proportional to the illumination intensity, C is proportional to the concentration of fluorophore in the sample, m0 and ϕ0 are the modulation and phase of the excitation at the sample. mτ = (τ2ω2 + 1)−1/2 and ϕτ = arctan(τω) are demodulation and dephase introduced by the fluorophore, which might differ pixel-to-pixel, with ω as the angular frequency.

The range in which F is informative is determined by the background level (BG) and the saturation value (SAT) of the detector.
BG<F(ϕi)<SAT
(2)

As the maximum and minimum values for F are I C(1 + m0mτ) and I C(1 − m0mτ),respectively, the concentration range is given by:
BGI(1m0mτ)<C<SATI(1+m0mτ)
(3)

The dynamic range (D), defined as the ratio between the maximum and minimum suitable concentration value, is then:
D=1m0mτ1+m0mτSATBG,
(4)
The first term is always smaller than one and therefore reduces the dynamic range of the detector (SAT/BG) for FLIM measurements. This decrease varies with the demodulation, and thus depends on ωτ, it becomes smaller towards higher frequencies and longer lifetimes. In a system with 80 MHz and a lifetime of 3 ns, the dynamic range is reduced by a factor of 0.29, for systems with 40 and 20 MHz it already reduces to 0.11 and 0.03, respectively. Reducing τ has a similar effect, thus reducing the lifetime to 1.5 ns lowers the accessible dynamic range to 0.11 at 80 MHz.

In flatFLIM, we recover the lost dynamic range by varying the intensity of the illumination as we step through the phases to pre-compensate for the changes in detected intensity (Fig. 1, blue markers). The detected emission (Eq. 1) then becomes:
F(ϕi)=I(ϕi)C(1+m0mτcos(ϕiϕ0ϕτ)),
(5)

In this case, the accessible dynamic range becomes:
Dflat=Min[F]Max[F]SATBG
(6)

If I(ϕ) is chosen to be the exact inverse of the modulation term, the complete dynamic range is recovered. However, the exact inverse of the modulation term is not known a priori - knowing it will yield its measurement irrelevant and unnecessary. Moreover, in the general case the fluorescence lifetime in each pixel and with it the required compensation function will differ. The challenge is therefore to choose a functional shape for I(ϕ) that proves most effective within the expected range of fluorescence lifetimes using incomplete a priori information.

2.2. Compensation function

The difficulty and the limiting factor of the effectiveness of flatFLIM is finding the correct compensation function for each specific condition. In case of a known single lifetime, finding the compensation function is simply:
I(ϕi)=I0(1+m0mλcos(ϕiϕ0ϕλ))1
(7)
where λ denotes the compensation fluorescence lifetime. The best dynamic range enhancement is expected when the assumed fluorescence lifetime is exactly the lifetime of the sample, thus countering exactly the modulating effects and yielding an identical intensity for all images throughout the phase-sweep. Assuming a reference modulation of 1, Fig. 2(a) shows the maximum attainable dynamic range enhancement (Dflat/DFLIM) for a range of compensated lifetimes and fluorophore lifetimes between 0.5 ns and 6 ns calculated from Eq. 6 using F(ϕ) from Eq.5 and the compensation function I(ϕ) from Eq. 7. The color indicates the dynamic range enhancement with a linear scale between 0 and 1 and a logarithmic scale above 1 to clearly visualize the break-even point, where the dynamic range enhancement of flatFLIM vanishes.

Fig. 2 Dynamic range enhancement of a single fluorophore with only one lifetime component, (a) Simulation of maximum dynamic range enhancement for a variety of lifetimes compensated with a range of assumed single lifetimes (sinusoidal excitation, M = 1), (b) Dynamic range enhancement for Alexa 488 (+) in water for various compensation lifetimes as compared to the corresponding simulated curve (red), (c) Simulation of maximum dynamic range enhancement with two fluorophores (2 ns and 4 ns) mixed, axes describe the fraction of 4 ns fluorophore, (d) Simulation of a FRET system with 3 ns lifetime as a function of FRET efficiency

The graph shows that it is generally safer to compensate with a higher assumed lifetime since the dynamic range enhancement asymptotically subsides towards 1 (no enhancement) when the compensation lifetime is much longer than the typical lifetime of the sample. Naturally, the dynamic range enhancement is larger for shorter lifetimes due to their inherently larger modulation and therefore lower dynamic range, and becomes lower for higher lifetimes. However, also the effect of wrongly estimating the lifetime increases concurrently for lower lifetimes. For a mixture of two fluorophores of 2 ns and 4 ns lifetimes, the dynamic range enhancement was calculated as a function of the fraction of the longer component (Fig. 2(c)). We observe that also in this system it is safer to compensate with a higher fraction of long lifetime component. In the case of a FRET experiment with a donor lifetime of 3 ns and varying FRET efficiencies it is also always safe to compensate with a longer lifetime (Fig. 2(d)). This information is valuable for assessment of transfer efficiency in cells where only certain regions might be showing FRET, compensating with the donor lifetime will yield dynamic range enhancement without detrimental effects.

3. Materials and methods

3.1. FLIM instrumentation

All FLIM measurements were performed using a previously described in-house developed system [11

11. H. E. Grecco, P. Roda-Navarro, S. Fengler, and P. I. H. Bastiaens, “High-Throughput quantification of posttranslational modifications In Situ by CA-FLIM.” Methods Enzymol. 500, 37–58 (2011). [CrossRef] [PubMed]

]. Briefly, the illumination scheme consists of a multiline Argon laser (Coherent Innova 300C) running in light-track mode modulated at 79.2 MHz using an acousto-optic modulator (AOM, Intra Action SWM-804AE1-1). Wavelength and power selection is achieved using an acousto-optic tunable filter (AA opto-electronic AOTF.nC). The laser is then coupled into a fully motorized microscope (Olympus IX81) using a multimode fiber, which is vibrated to average out the speckle pattern. Images are acquired using a high rate image intensifier (LaV-ision PicoStar HR) coupled to a 12 bit CCD camera (LaVision Imager QE). Both AOM and image intensifier are modulated with coupled frequency generators (National Instruments PXI-5404). The whole setup is controlled by a program developed in-house using LabVIEW 2010 (National Instruments).

3.2. FLIM measurements and analysis

The samples were excited using the 488 nm line of the argon laser. The emission was collected through a 20x air objective (solutions) or 40x air objective (cells) and spectrally filtered using a YFP filterset (dichroic: 505 nm, emission filter: 535/30 nm). Before each experiment, the phase and modulation of the system was calibrated using an aluminum foil located at the sample plane. A mirror mounted in a filter cube was used to routinely characterize the system’s phase and modulation drift as previously described [11

11. H. E. Grecco, P. Roda-Navarro, S. Fengler, and P. I. H. Bastiaens, “High-Throughput quantification of posttranslational modifications In Situ by CA-FLIM.” Methods Enzymol. 500, 37–58 (2011). [CrossRef] [PubMed]

]. In order to reduce the effect of bleaching for the calculation of phase and modulation lifetimes, the images were taken in permuted phase order [12

12. E. B. van Munster and T. W. J. Gadella, “Suppression of photobleaching-induced artifacts in frequency-domain FLIM by permutation of the recording order.” Cytometry, Part A 58, 185–94 (2004). [CrossRef]

]. FLIM datasets were analyzed using singular value decomposition to find the phase and modulation in each pixel as previously described [13

13. P. J. Verveer and P. I. H. Bastiaens, “Evaluation of global analysis algorithms for single frequency,” J. Microsc. (Oxford, U. K.) 209, 1–7 (2003). [CrossRef]

]. For flatFLIM, the phase-stacks were rescaled by the inverse of the compensation function and then analyzed as standard FLIM datasets.

3.3. Cell preparation

HeLa and MCF-7 cells were used for the single lifetime and FRET experiments, respectively. The cells were grown in 8-well LabTek chambers in DMEM (PAN Biotech GmbH) supplemented with 10 % heat-inactivated fetal calf serum (FCS) (PAN Biotech GmbH) and 2 mM L-glutamine under 37 °C and 5 % CO2. They were transfected according to standard procedures with 6 μl Fugene 6 (Roche Applied Science) and 2 μg DNA in 200 μl serum free medium spread over the eight wells and allowed 24 h for transfection, protein expression and fluorophore maturation. The HeLa cells were measured directly in imaging medium. The MCF-7 cells were starved in serum free medium for 4 h, stimulated with EGF for 5’ and then fixed with 3 % paraformaldehyde. Following that they were washed with TBS buffer three times and imaged in PBS buffer. After the donor-only measurement we added a generic phosphotyrosine antibody (PY72) labelled with Cy3.5 and measured after 20 min.

4. Results

4.1. Dynamic range enhancement

In order to calculate the dynamic range enhancement, we measured FLIM and flatFLIM for a range of different concentrations of Alexa 488 solution (60 pM to 3000 pM) using different compensation lifetimes between 0.5 ns and 8 ns. By using a non uniform illumination we obtain images with an intensity gradient. The resulting phase-stacks were thresholded for under/overexposure using the condition that pixels are only properly exposed when the minimum across the stack does not fall below 150 counts and the maximum does not exceed 4050 counts (Fig. 3(a)).

Fig. 3 (a) FLIM and flatFLIM images, over/underexposed regions are marked red/blue, for a variety of concentrations for a compensation lifetime of 4.1 ns (b) Percentage of usable pixels over the concentration range, flatFLIM 4.1 ns (gray area), 3 ns (blue, solid), 6 ns (blue, dashed), FLIM (green area)

For each combination of compensation lifetime and concentration, the number of properly exposed pixels was calculated. The exposure time for the complete concentration series was adjusted in order to obtain fully over- or underexposed images for the lowest and highest concentrations. In order to calculate the dynamic range, we determined the ratio of the highest and lowest concentration at which 50 % of the pixels were resolvable (Fig. 3(b)). We observe a dynamic range of 7 for FLIM and a compensation lifetime dependent dynamic range for flat-FLIM. The dynamic range enhancement was calculated as the ratio between the two quantities (Fig. 2(b), black crosses) as compared to the simulated curve.

4.2. Photon dose and photobleaching

The light dose (L) and the average intensity of a pixel across the phase-stack with N images for FLIM can be expressed as:
LFLIM=NI
(8a)
<FFLIM>=1Nn=1NF(ϕi)~ICN
(8b)
while for flatFLIM and assuming a perfectly matched compensation lifetime it is:
LflatFLIM=NI0[nI(ϕi)]1
(9a)
<FflatFLIM>~I0CN
(9b)

We observe that the photon dose needed for a single flatFLIM stack is larger than for a single FLIM stack if the same average intensity is desired. However, multiple FLIM stacks are required to obtain a similar concentration dynamic range resulting in a better photon economy in flatFLIM (Fig. 4(a)). This concept is shown in Fig. 4(a) for a lifetime of 3.44 ns and a resulting dynamic range enhancement of 3. In FD-FLIM three images at different laser powers (0.33 A.U., 1 A.U, 1.66 A.U.) are needed to image concentration gradient. In contrast, flatFLIM requires only one image with a slightly higher average laser power (1.15 A.U.).

Fig. 4 Effect of photobleaching in the fluorescence lifetime determination: (a) dynamic range enhancement over FLIM (black) and increase in light dose for same SNR plotted versus the fluorescence lifetime (green) and comparison to the requirements for FLIM for the same dynamic range at 3.4ns lifetime and thus 3x dynamic range enhancement (b) effect of the reference phase on the lifetime calculation upon a bleaching of 10 % and a simulated lifetime of 3 ns; effect of different amounts of photobleaching for a simulated lifetime of 3 ns on the calculated phase and modulation lifetimes for FLIM and flatFLIM

Photobleaching introduces bias in the recovered fluorescence lifetime when the phase-stack is acquired in sequential order. This problem can be partially avoided acquiring the phase-stack twice in opposite directions, as previously shown. Alternatively, acquiring the phase-stack in a permuted order has proven very reliable without the need of acquiring more images [12

12. E. B. van Munster and T. W. J. Gadella, “Suppression of photobleaching-induced artifacts in frequency-domain FLIM by permutation of the recording order.” Cytometry, Part A 58, 185–94 (2004). [CrossRef]

].

To investigate the effects of photobleaching, we simulated data with different levels of photobleaching for sequential and permuted phase orders. We generated FLIM data with 12 phase steps, reordered them in the case of permuted order and applied the bleaching. Then we calculated the phase and modulation lifetime from the data using Fourier component analysis. Firstly, we analyzed the effect of the reference phase ϕ0 on a single lifetime (3 ns) and single bleaching (P = 10 %). The bleaching is defined as the intensity loss after recording a stack of images -for flatFLIM the increased dose is taken into account. The reference phase has a larger impact on the flatFLIM experiments than on the FLIM measurements due to the laser power varying throughout the experiment (Fig. 4(b)). The same trend is visible for different levels of bleaching (Fig. 4(b)), when the maximum lifetime deviation across the reference phases is plotted. The deviations are higher than for FLIM scanned in permuted order but stay well below 1 % for realistic values of photobleaching (10 - 15 % intensity loss). For scanning in linear order the effects are much larger and in the same range as those for FLIM (not shown).

4.3. flatFLIM in cells

We then tested flatFLIM in cells expressing mCitrine. We imaged the same field of view first with FLIM and then with flatFLIM with the same exposure time and a laser power yielding the same average pixel intensities. The typical bleaching after a FLIM sequence was 1 %. As expected, we observe that certain regions are over- or underexposed in the FLIM dataset but not in the flatFLIM dataset (Fig. 5). Moreover, we observe that the fluorescence lifetime distribution of the pixels observable with both methods is maintained indicating that there is no significant bias due to photobleaching.

Fig. 5 (a) FLIM and flatFLIM images of HeLa cells transfected with EYFP showing over (red) and under (blue) exposed regions as well as modulation lifetime and phase lifetime images (b) Intensity projection across the yellow line in both FLIM and flatFLIM images

4.4. flatFLIM to image FRET

Fig. 6 Phase lifetime images (FLIM and flatFLIM) of MCF-7 cells with EGFR-YFP receptor before and after addition of Cy3.5 labelled PY72 antibody (smaller images: under/overexposed regions within the cell)

5. Conclusion

We have shown that the flatFLIM enhances the concentration dynamic range in FD-FLIM by adjusting the laser power to compensate for the expected intensity variation. Choosing the best compensation function seems to be a circular problem as the fluorescence lifetime of the sample is needed. However, we have shown both, theoretically and experimentally, that choosing a compensation fluorescence lifetime higher than the actual lifetime always yields an enhanced dynamic range. This enhancement vanishes when the compensation lifetime is chosen an order of magnitude above, an unlikely situation given that the range of fluorescence lifetimes is usually known in most biological applications.

We have also shown that most of the dynamic range of the camera can be restored. However, the dynamic range enhancement is slightly shifted and lower as compared to the prediction, likely due to the effect of other harmonics not considered in the model. This does not affect the applicability of flatFLIM as the effect is not significant as compared to the uncertainty introduced by the lack of a known lifetime. We have shown theoretically and experimentally that the effect of bleaching is comparable between flatFLIM and FD-FLIM and that it can be reduced to the same degree by using a pseudo random phase order.

In comparison with FD-FLIM, a single flatFLIM stack requires a higher light dose than a single FD-FLIM stack. However, the enhancement in concentration dynamic range greatly exceeds the additional light dose. In other words, to be able to image the same concentration range with FD-FLIM as with flatFLIM, multiple sequences, acquired at different powers, are needed adding up to a higher light dose thus leading to more bleaching. Therefore, like CLEM [9

9. R. A. Hoebe, C. H. Van Oven, T. W. J. Gadella, P.B. Dhonukshe, C. J. F. Van Noorden, and E. M. M. Manders, “Controlled light-exposure microscopy reduces photobleaching and phototoxicity in fluorescence live-cell imaging” Nat. Biotechnol. 25, 249–53 (2007). [CrossRef] [PubMed]

] and MLE-PAM [10

10. W. Caarls, B. Rieger, A. H. B. De Vries, D. J. Arndt-Jovin, and T. M. Jovin, “Minimizing light exposure with the programmable array microscope” J. Microsc. (Oxford, U. K.) 241, 101–10 (2011). [CrossRef]

], flatFLIM reduces photo-damage and enhances the dynamic range. The first two operate by equalizing the intensity across spatial dimension and flatFLIM across the phase stack. For FLIM applications, MLE-PAM and flatFLIM could be combined since the two techniques are independent and complement each other without interference. An additional benefit of such combination is the enhanced lateral resolution that might counter the typical loss of resolution due to the intensifier.

An important advantage of this method is that it can be implemented with only minor software changes in microscopes equipped with a computer controlled light source. In our setup, we have used the same AOTF that we have previously used to select the laser line and its intensity. This implementation does not change the total acquisition time significantly as the AOTF can change the intensity in less than 50 μs when analog control is used. The analysis is the same as in standard FD-FLIM with only a preprocessing step in which each image of the phase-stack is multiplied by the inverse of the corresponding compensation intensity.

If using an AOTF is not a suitable option, other slower implementations are possible. For example, a filter wheel with properly selected OD filters can be added to the excitation pathway. The main drawback is that the limited number of filters will discretize the compensation function and thereby lead to a partial dynamic range enhancement. Alternatively, the exposure time can be used as a compensation mechanisms, taking into account in the analysis that the dark noise will differ across the stack. Care must be taken that the phase and modulation of the light source is not affected by the intensity change as would probably be the case for modulated light emitting diodes. Since the intensity of the laser is important for the reconstruction of the data, a careful calibration of the laser power is required. However, it is relatively easy to calibrate with high precision so that the errors introduced by the laser power are greatly exceeded by those coming from the image intensifier[14

14. B. Q. Spring and R. M. Clegg, “Image analysis for denoising full-field frequency-domain fluorescence lifetime images” J. Microsc. (Oxford, U. K.) 235, 221–237 (2009). [CrossRef]

].

flatFLIM can be especially appropriate for screening applications [15

15. A. Esposito, C. P. Dohm, M. Bähr, and F. S. Wouters, “Unsupervised fluorescence lifetime imaging microscopy for high content and high throughput screening.” Mol. Cell. Proteomics 6, 1446–54 (2007). [CrossRef] [PubMed]

, 16

16. H. E. Grecco, P. Roda-Navarro, A. Girod, J. Hou, T. Frahm, D. C. Truxius, R. Pepperkok, A. Squire, and P. I. H. Bastiaens, “In situ analysis of tyrosine phosphorylation networks by FLIM on cell arrays.” Nat. Methods 7, 467–72 (2010). [CrossRef] [PubMed]

], when low magnification objectives are used, as the likelihood of observing cells with very different expression levels within the same field of view is higher. An increased dynamic range thereby increases the effective throughput (i.e. number of properly imaged cells per image) without the need of additional images acquired at lower or higher intensities. Finally, not only the throughput but most importantly the information content is increased. We have shown in the past that we can recognize the presence of positive feedback loops by correlating the fraction of phosphorylated protein with its concentration as determined from the intensity [16

16. H. E. Grecco, P. Roda-Navarro, A. Girod, J. Hou, T. Frahm, D. C. Truxius, R. Pepperkok, A. Squire, and P. I. H. Bastiaens, “In situ analysis of tyrosine phosphorylation networks by FLIM on cell arrays.” Nat. Methods 7, 467–72 (2010). [CrossRef] [PubMed]

]. Variability in expression level is a source of information and by increasing the observable dynamic range in a single picture, flatFLIM expands the applicability of this concept.

Acknowledgments

This work was supported by the Deutsche Forschungsgemeinschaft ( DFG GR 3848/1-1) and the Max-Planck Gesellschaft (MPG). The authors would like to thank Prof. Dr. Philippe I.H. Bastiaens, Dr. Peter J. Verveer and Dr. Astrid Krämer for critical reading of the manuscript.

References and links

1.

H. C. Ishikawa-Ankerhold, R. Ankerhold, and G. P. C. Drummen, “Advanced fluorescence microscopy techniques-FRAP, FLIP, FLAP, FRET and FLIM.” Molecules (Basel, Switzerland) 17, 4047–132 (2012). [CrossRef]

2.

R. S. Fischer, Y. Wu, P. Kanchanawong, H. Shroff, and C. M. Waterman, “Microscopy in 3D: a biologist’s toolbox.” Trends Cell Biol. 21, 682–91 (2011). [CrossRef] [PubMed]

3.

F. S. Wouters, P. J. Verveer, and P. I. H. Bastiaens, “Imaging biochemistry inside cells,” Trends Cell Biol. 11, 203–211 (2001). [CrossRef] [PubMed]

4.

S. Ogikubo, T. Nakabayashi, T. Adachi, M. S. Islam, T. Yoshizawa, M. Kinjo, and N. Ohta, “Intracellular pH sensing using autofluorescence lifetime microscopy.” J. Phys. Chem. B 115, 10385–90 (2011). [CrossRef] [PubMed]

5.

M. K. Kuimova, G. Yahioglu, J. A. Levitt, and K. Suhling, “Molecular rotor measures viscosity of live cells via fluorescence lifetime imaging” J. Am. Chem. Soc. 130, 6672–3 (2008). [CrossRef] [PubMed]

6.

T. Förster “Energy migration and fluorescence - 1946” J. Biomed. Opt. 17, 011002–10 (2012). [CrossRef] [PubMed]

7.

E. A. Jares-Erijman and T. M. Jovin, “FRET imaging.” Nat. Biotechnol. 21, 1387–95 (2003). [CrossRef] [PubMed]

8.

T. W. J. Gadella, T. M. Jovin, and R. M. Clegg, “Fluorescence lifetime imaging microscopy (FLIM) Spatial resolution of microstructures on the nanosecond time scale,” Biophys. Chem. 48, 221–239 (1993). [CrossRef]

9.

R. A. Hoebe, C. H. Van Oven, T. W. J. Gadella, P.B. Dhonukshe, C. J. F. Van Noorden, and E. M. M. Manders, “Controlled light-exposure microscopy reduces photobleaching and phototoxicity in fluorescence live-cell imaging” Nat. Biotechnol. 25, 249–53 (2007). [CrossRef] [PubMed]

10.

W. Caarls, B. Rieger, A. H. B. De Vries, D. J. Arndt-Jovin, and T. M. Jovin, “Minimizing light exposure with the programmable array microscope” J. Microsc. (Oxford, U. K.) 241, 101–10 (2011). [CrossRef]

11.

H. E. Grecco, P. Roda-Navarro, S. Fengler, and P. I. H. Bastiaens, “High-Throughput quantification of posttranslational modifications In Situ by CA-FLIM.” Methods Enzymol. 500, 37–58 (2011). [CrossRef] [PubMed]

12.

E. B. van Munster and T. W. J. Gadella, “Suppression of photobleaching-induced artifacts in frequency-domain FLIM by permutation of the recording order.” Cytometry, Part A 58, 185–94 (2004). [CrossRef]

13.

P. J. Verveer and P. I. H. Bastiaens, “Evaluation of global analysis algorithms for single frequency,” J. Microsc. (Oxford, U. K.) 209, 1–7 (2003). [CrossRef]

14.

B. Q. Spring and R. M. Clegg, “Image analysis for denoising full-field frequency-domain fluorescence lifetime images” J. Microsc. (Oxford, U. K.) 235, 221–237 (2009). [CrossRef]

15.

A. Esposito, C. P. Dohm, M. Bähr, and F. S. Wouters, “Unsupervised fluorescence lifetime imaging microscopy for high content and high throughput screening.” Mol. Cell. Proteomics 6, 1446–54 (2007). [CrossRef] [PubMed]

16.

H. E. Grecco, P. Roda-Navarro, A. Girod, J. Hou, T. Frahm, D. C. Truxius, R. Pepperkok, A. Squire, and P. I. H. Bastiaens, “In situ analysis of tyrosine phosphorylation networks by FLIM on cell arrays.” Nat. Methods 7, 467–72 (2010). [CrossRef] [PubMed]

OCIS Codes
(170.1420) Medical optics and biotechnology : Biology
(170.2520) Medical optics and biotechnology : Fluorescence microscopy
(170.3650) Medical optics and biotechnology : Lifetime-based sensing
(170.6280) Medical optics and biotechnology : Spectroscopy, fluorescence and luminescence

ToC Category:
Medical Optics and Biotechnology

History
Original Manuscript: June 18, 2012
Revised Manuscript: July 25, 2012
Manuscript Accepted: August 1, 2012
Published: August 24, 2012

Virtual Issues
Vol. 7, Iss. 10 Virtual Journal for Biomedical Optics

Citation
Klaus C. Schuermann and Hernán E. Grecco, "flatFLIM: enhancing the dynamic range of frequency domain FLIM," Opt. Express 20, 20730-20741 (2012)
http://www.opticsinfobase.org/vjbo/abstract.cfm?URI=oe-20-18-20730


Sort:  Author  |  Year  |  Journal  |  Reset  

References

  1. H. C. Ishikawa-Ankerhold, R. Ankerhold, and G. P. C. Drummen, “Advanced fluorescence microscopy techniques-FRAP, FLIP, FLAP, FRET and FLIM.” Molecules (Basel, Switzerland)17, 4047–132 (2012). [CrossRef]
  2. R. S. Fischer, Y. Wu, P. Kanchanawong, H. Shroff, and C. M. Waterman, “Microscopy in 3D: a biologist’s toolbox.” Trends Cell Biol.21, 682–91 (2011). [CrossRef] [PubMed]
  3. F. S. Wouters, P. J. Verveer, and P. I. H. Bastiaens, “Imaging biochemistry inside cells,” Trends Cell Biol.11, 203–211 (2001). [CrossRef] [PubMed]
  4. S. Ogikubo, T. Nakabayashi, T. Adachi, M. S. Islam, T. Yoshizawa, M. Kinjo, and N. Ohta, “Intracellular pH sensing using autofluorescence lifetime microscopy.” J. Phys. Chem. B115, 10385–90 (2011). [CrossRef] [PubMed]
  5. M. K. Kuimova, G. Yahioglu, J. A. Levitt, and K. Suhling, “Molecular rotor measures viscosity of live cells via fluorescence lifetime imaging” J. Am. Chem. Soc.130, 6672–3 (2008). [CrossRef] [PubMed]
  6. T. Förster “Energy migration and fluorescence - 1946” J. Biomed. Opt.17, 011002–10 (2012). [CrossRef] [PubMed]
  7. E. A. Jares-Erijman and T. M. Jovin, “FRET imaging.” Nat. Biotechnol.21, 1387–95 (2003). [CrossRef] [PubMed]
  8. T. W. J. Gadella, T. M. Jovin, and R. M. Clegg, “Fluorescence lifetime imaging microscopy (FLIM) Spatial resolution of microstructures on the nanosecond time scale,” Biophys. Chem.48, 221–239 (1993). [CrossRef]
  9. R. A. Hoebe, C. H. Van Oven, T. W. J. Gadella, P.B. Dhonukshe, C. J. F. Van Noorden, and E. M. M. Manders, “Controlled light-exposure microscopy reduces photobleaching and phototoxicity in fluorescence live-cell imaging” Nat. Biotechnol.25, 249–53 (2007). [CrossRef] [PubMed]
  10. W. Caarls, B. Rieger, A. H. B. De Vries, D. J. Arndt-Jovin, and T. M. Jovin, “Minimizing light exposure with the programmable array microscope” J. Microsc. (Oxford, U. K.)241, 101–10 (2011). [CrossRef]
  11. H. E. Grecco, P. Roda-Navarro, S. Fengler, and P. I. H. Bastiaens, “High-Throughput quantification of posttranslational modifications In Situ by CA-FLIM.” Methods Enzymol.500, 37–58 (2011). [CrossRef] [PubMed]
  12. E. B. van Munster and T. W. J. Gadella, “Suppression of photobleaching-induced artifacts in frequency-domain FLIM by permutation of the recording order.” Cytometry, Part A58, 185–94 (2004). [CrossRef]
  13. P. J. Verveer and P. I. H. Bastiaens, “Evaluation of global analysis algorithms for single frequency,” J. Microsc. (Oxford, U. K.)209, 1–7 (2003). [CrossRef]
  14. B. Q. Spring and R. M. Clegg, “Image analysis for denoising full-field frequency-domain fluorescence lifetime images” J. Microsc. (Oxford, U. K.)235, 221–237 (2009). [CrossRef]
  15. A. Esposito, C. P. Dohm, M. Bähr, and F. S. Wouters, “Unsupervised fluorescence lifetime imaging microscopy for high content and high throughput screening.” Mol. Cell. Proteomics6, 1446–54 (2007). [CrossRef] [PubMed]
  16. H. E. Grecco, P. Roda-Navarro, A. Girod, J. Hou, T. Frahm, D. C. Truxius, R. Pepperkok, A. Squire, and P. I. H. Bastiaens, “In situ analysis of tyrosine phosphorylation networks by FLIM on cell arrays.” Nat. Methods7, 467–72 (2010). [CrossRef] [PubMed]

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.


« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited