DCS-120 FLIM System Detects FMN in Live Cells
Wolfgang
Becker, Lukas Braun, Becker & Hickl GmbH
Abstract: It has long been suspected that FAD metabolic
FLIM data are biased by a fluorescence decay component from FMN. We therefore
checked for possible traces of FMN fluorescence in FLIM data from our DCS-120
metabolic FLIM system. Indeed, we found a decay component with a lifetime
around 4.8 ns which cannot be attributed to FAD but very well to FMN. The
amplitude of the FMN component was 3.9% in normal cells and 8.8% in cancer
cells.
The fluorescence
decay functions of NAD(P)H (nicotinamide adenine
(pyridine) dinucleotide) and
FAD (flavin adenine dinucleotide) contain information on the metabolic state
of the cells. Bound and unbound NAD(P)H and FAD have different lifetimes, and
the relative amount of bound and unbound forms varies with the metabolic state.
Therefore the apparent lifetime or, better, the ratio of the amplitudes of the decay
components, a1/a2, can be used as an indicator of the metabolic state [1].
Although FLIM
results are largely conclusive for NADH [2,3,5,6], this is not the case for
FAD. In contrast to NADH, where the fast decay component comes from the free
form, for FAD the slow component is associated to the free form. Therefore, the
lifetime effect of a change in the free/bound ratio for FAD and NADH goes in
opposite directions. Often in FAD FLIM data the signal from FAD is not cleanly
separated from that of NADH. The result then depends on the relative amounts of
FAD and NADH. This can modify or even reverse the observed lifetime effect.
However, also for perfectly recorded data the FAD results are often
inconclusive. It has long been suspected that the FAD signal contains
fluorescence from FMN (flavin mononucleotide) [4]. We therefore looked for possible traces
of FMN fluorescence in NADH/FAD data recorded with our DCS-120 metabolic FLIM
system [1, 2].
Fig. 2 shows NAD(P)H and FAD FLIM images of
freshly excised human epithelial bladder cells [6]. The images were recorded by
multiplexed excitation at 375 nm (NAD(PH) and 410 nm (FAD). NAD(P)H
was detected from 420 to 470 nm, FAD from 490 to 580 nm. Please see [1,2] for technical
details. Both images show the amplitude of the fast decay component, a1
[1]. The
sample contains mostly normal cells (area A), but there are also a few tumorous
cells with an enhanced fraction of free NADH (area B).
![](https://www.becker-hickl.com/wp-content/uploads/appnotes/16740/fmn-dcs-02-Dateien/image001.jpg)
Fig.
1: FLIM images of human epithelium bladder cells. NAD(P)H (left) and FAD
(right). Amplitude, a1, of fast decay component
To check for possible FMN contribution in
the FAD data, an ROI was chosen which contains only cells that qualify 'normal'
in the NADH channel. Please see Fig. 2. The decay curve of the combined pixels
within the ROI is shown on the right. Triple-exponential analysis clearly
delivers a decay component of 4.83 ns, present with an amplitude of 3.9 %.
A lifetime of 4.83 ns cannot be attributed to FAD, but very well to FMN.
Fig. 3 shows the same data, but the ROI was
set on cells which are classified as 'cancer' in the NADH image. The lifetime
of the third component is 4.38 ns, not much different than in Fig. 2.
However, the amplitude has increased to 8.84 %.
![](https://www.becker-hickl.com/wp-content/uploads/appnotes/16740/fmn-dcs-02-Dateien/image002.jpg)
Fig. 2: Left: FAD Image of human epithelial bladder cells. The region of
interest includes healthy cells only. Right: Decay curve of combined pixels in
ROI, and triple-exponential fit. The third component, with t3 = 4.835 ns
and a3 = 3.88%, is compatible with FMN.
![](https://www.becker-hickl.com/wp-content/uploads/appnotes/16740/fmn-dcs-02-Dateien/image003.jpg)
Fig. 3: Same data as in Fig. 2, but ROI on cells which are probably tumor
ells. The third component is t3 = 4.383 ns and a3 = 8.84%.
Fig. 4 shows an attempt to extract the
amplitude of the FMN component on a pixel-by-pixel basis. Of course, this poses
a challenge to the data analysis. Triple-exponential analysis has to be used, the
number of photons is lower than in a region of interest, and the slow component
of FAD, t2, and the FMN component, t3, differ only by a
factor of two in lifetime. Moreover, the amplitude of the FMN component is low
compared to the amplitudes of the FAD components. Nevertheless, with a binning
factor of 8 in SPCImage NG (17x17 pixels, the entire image has 512x512 pixels)
and MLE analysis we obtained a very satisfactory result. As expected, the
amplitude of the FMN component, a3, is about 4% in the normal cells,
and about 8% in the tumor cells. This does not mean that the absolute amount of
FMN in the tumor cell is higher. The higher a1 can also come from
the lower intensity of the tumor cell. Whether the lower intensity is a result
of lower FAD concentration or lower redox ratio in the tumor cell is not known.
In the second case, a3 may be another expression of the redox ratio.
Whether it can be used to distinguish cells of different metabolic state should
be subject of further investigation.
![](https://www.becker-hickl.com/wp-content/uploads/appnotes/16740/fmn-dcs-02-Dateien/image004.jpg)
Fig. 4: Triple-exponential analysis of the image in the FAD channel.
Colour shows amplitude, a3, of FMN component.
The presence of a decay component from FMN
may have implications to the use of FAD data for cancer diagnosis, especially
if the amplitude, a1, the amplitude ratio, a1/a2 or the FLIRR ratio are used as
cancer indicators. Certainly, an amplitude contribution of 3.8 % or even
8.84 % does not render a1 or a1/a2 of FAD
or the FLIRR useless. It can, however, introduce unpredictable shifts in these
parameters. The results then can no longer be considered quantitative. Kalinina
et al. therefore suggested new definitions of the FLIRR based on
triple-exponential analysis of the FAD [4].
References
1. W. Becker, The bh TCSPC Handbook, 9th edition (2021)
2. W. Becker, A. Bergmann, L. Braun, Metabolic Imaging with the DCS-120
Confocal FLIM System: Simultaneous FLIM of NAD(P)H and FAD, Application note,
available on www.becker-hickl.com (2018)
3. Becker Wolfgang, Suarez-Ibarrola Rodrigo, Miernik Arkadiusz, Braun
Lukas, Metabolic Imaging by Simultaneous FLIM of NAD(P)H and FAD. Current
Directions in Biomedical Engineering 5(1), 1-3 (2019)
4. S. Kalinina, C. Freymueller, N. Naskar, B. von Einem, K. Reess, R.
Sroka, A. Rueck, Bioenergetic Alterations of Metabolic Redox Coenzymes as NADH,
FAD and FMN by Means of Fluorescence Lifetime Imaging Techniques. Int. J. Mol.
Sci. 22, 5952, 1-15 (2021)
5. M. M. Lukina, V. V. Dudenkova, N. I. Ignatovaa, I. N. Druzhkova, L.
E. Shimolina, E. V. Zagaynovaa, M. V. Shirmanova, Metabolic cofactors NAD(P)H
and FAD as potential indicators of cancer cell response to chemotherapy with
paclitaxel. BBA General Subjects 1862, 1693-1700 (2018)
6. Rodrigo Suarez-Ibarrola, Lukas Braun, Philippe Fabian Pohlmann,
Wolfgang Becker, Axel Bergmann, Christian Gratzke, Arkadiusz Miernik, Konrad
Wilhelm, Metabolic Imaging of Urothelial Carcinoma by Simultaneous
Autofluorescence Lifetime Imaging (FLIM) of NAD(P)H and FAD. Clinical
Genitourinary Cancer (2020)
Contact:
Wolfgang Becker
Becker & Hickl GmbH
Berlin, Germany
Email: becker@becker-hickl.com