Multiphoton Microscopy for Bacterial Imaging: A Label-free

Solution Resting on Endogenous Two-photon Fluorescence

Cédric Delmon

, Erwan Ferrandon

, Emilie Chouzenoux

, Audrey Prorot

, Sophie Alain

and Claire Lefort

Peirene, GRESE (EA 4330), University of Limoges, Limoges, France

XLIM Research Institute, UMR CNRS 7252, University of Limoges, Limoges, France

Center for Visual Computing, CentraleSupélec, Inria Saclay, Université Paris-Saclay, Gif-sur-Yvette, France

RESINFIT, UMR INSERM 1092, University of Limoges, Bacteriology-Virology-Hygiene Department,

University Hospital Center, France

Keywords: Multiphoton Microscopy, Bacterial Imaging, Spectral Characterization, Two-photon Excitation, Fluorescence

Emission, Metabolic Indicators, Point-Spread-Function, Computational Restoration, Block Distributed

Majorize-minimize Memory Gradient.

Abstract: We demonstrate the interest of multiphoton microscopy (MPM) for imaging bacteria without any labelling

process. Six families of bacteria are tested: Escherichia coli, Staphylococcus epidermidis, Proteus vulgaris,

Pseudomonas fluorescens, Bacillus subtilis and Clostridium perfringens. For each of these bacteria, the image

of a cell is recorded through a multiphoton microscope thus revealing the 3D shape of these bacteria. For the

first time, the images of such bacteria are recorded without any labelling solution. A protocol of controlling

the image produced is led thanks to a standard staining protocol with carboxy fluorescein diacetate (CFDA)

for E. Coli and Staphylococcus epidermidis. Similar object shapes with or without labelling are produced,

thus validating the label-free images generated by MPM. Then, the two-photon excitation spectra are

measured for each of these bacteriaand the emission spectra delivered by E. Coli and Bacillus subtilis are

shown. The origin of the two-photon fluorescence (TPF) emission of the bacteria thanks to the nonlinear

imaging solution is discussed regarding to the TPF excitation and emission spectra of metabolic indicators.

1 INTRODUCTION

Currently, no label free solution devoted to bacterial

imaging exists and labelling bacteria remains a

technical challenge. Several techniques have been

used for bacterial detection and identification, such as

conventional laboratory-based culture media,

polymerase chain reaction (PCR), immunological

techniques, and Raman spectroscopy (Yoon 2021).

However, these methods are often expensive, time-

consuming, they need complex procedures, or can

lead to false positive/ negative results. With the recent

technical advances of bacteria detection using

fluorescence-based dyes, it is now possible to assess

simultaneously the physiological states of cells with

two or three different fluorescent dyes that target

specific biomolecules and physiological processes

https://orcid.org/0000-0002-5421-2477

https://orcid.org/0000-0002-7685-2061

(Yoon, 2021, Wilkinson 2018). The use of flow

cytometry multiparameter analysis in conjunction

with fluorescent dyes provides consistent information

about physiological state of cells at a single level.

However, these methods need complex protocol

adaptation and complementary investigative

techniques. Above all, further advances are required

for “in situ” methods improvement in order to assess

microbial viability without coloration steps. To this

purpose, other more direct technical solutions still

have to be developed. Therefore, the need in new

solutions resting on microscopic tools for the optical

characterization of bacteria is high.

In this publication, for the first time in the field of

biomedical microscopy and in the field of

microbiology bacteria, we propose to merge

multiphoton microscopy (MPM) (Larson, 2011) and

Delmon, C., Ferrandon, E., Chouzenoux, E., Prorot, A., Alain, S. and Lefort, C.

Multiphoton Microscopy for Bacterial Imaging: A Label-free Solution Resting on Endogenous Two-photon Fluorescence.

DOI: 10.5220/0010979600003121

In Proceedings of the 10th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2022), pages 76-83

ISBN: 978-989-758-554-8; ISSN: 2184-4364

bacterial imaging, in order to reveal the image of such

a biological object through a nonlinear optical

process. Several set of experiments usually involved

for biological, medical or vegetal samples such as

cells, tissues, of whole organs (Gobel 2007, Plotnikov

2006, Hortholary 2021) are now newly presented for

nonlinear characterisation of bacetria. We are testing

six families of bacteria: Escherichia coli (E. Coli),

Staphylococcus epidermidis, Proteus vulgaris,

Pseudomonas fluorescens, Bacillus subtilis and

Clostridium perfringens. For each of these bacteria,

morphologic characteristics are gathered into Table 1,

showing the diversity of the sample chosen: each

GRAM is represented (GRAM + and GRAM –),

bacilus and coccus shapes, sporulant or not and

aeroby or anaeroby. For such large diversity of

bacteria, a preliminary test is led: a standard protocol

of fluorescence imaging through a nonlinear process

of two-photon fluorescence is involved. Thus, two-

photon fluorescence images of these bacteria are

produced. The first interesting result obtained

concerned the ability of this strategy to deliver an

image for each of the 6 bacteria selected. By

consequence, a second set of experiment was led. For

each bacetria, a recording of the two-photon

excitation spectra highlights large regions of

absorptions, and similar spectral shapes and regions

of emission are recorded, as illustrated by the

emission spectra of E. Coli and Bacillus subtilis. A

3D image of Clostridium Perfringens is proposed

thanks to the application of our instrumental and

computational pipeline FAMOUS resting on the

estimation of the point-spread-function (PSF) of the

system and the deblurring/denoising of the image

(Lefort 2021). Finally, we raise a discussion about the

origin of the endogenous fluorescence observed and

about the interest of MPM devoted to the optical

characterization of bacteria. More precisely, the

question raised by the presence of fluorescence

emission in the same emission range and biphotonic

excitation range for all of the bacteria presenting such

different characteristics led to the investigation about

the presence of similar substances for all these

biological objects. Our conclusions are therefore

oriented towards metabolic indicators, substances

existing for all of the living bacteria.

2 EXPERIMENTAL SETUP

The experimental setup involved in this study rests on

a two-side contribution including optical &

computational engineering on the one hand and

bacterial process engineering on the other hand. First

the bacteria are chosen thanks to their specificities for

a complete representation of the different kinds of

bacteria existing. Then, the instrumental and

computational solution resting on MPM is presented

and our original solution FAMOUS devoted to the

image restoration is presented. We note here that for

all the images presented in this publication, false

colours have been used. Indeed, the recording of the

fluorescence signal emitted by the bacteria

endogenously or after a fluorescence labelling

process is led by a PMT which detects a signal

intensity level, discretely coded.

2.1 Bacteria

2.1.1 Six Bacteria Chosen

The experimental characterization protocol for

imaging bacteria is tested with six bacteria from

different families. Table 1 gathers the properties of

the related bacteria (Prescott 2018).

Table 1: Characteristics of the bacteria used.

E. coli Staphylococcus

epidermidis

Proteus

vulgaris

Pseudomonas

fluorescens

Bacillus

subtilis

Clostridium

perfringens

GRAM - + - - + +

Shape Bacillus Coccus Bacillus Bacillus Bacillus Bacillus

Sporulation No No No No Yes Yes

Aerobic Yes Yes Strict

Anaerobic Facultative Facultative

In order to cover the great complexity of bacterial

structure and physiological features, a large variety of

biological models such as gram-negative, gram-

positive and spore forming bacteria were used in this

study.

2.1.2 Protocol of Preparation

The six bacteria studied (E. coli, Staphylococcus

epidermidis, Pseudomonas fluorescens, Proteus

vulgaris, Bacillus subtilis, Clostridium perfringens)

were cultured in a TSB liquid culture medium

(Trypto-Casein-Soy BK046HA, Biokar). 1.2mL of

bacteria culture in TSB liquid culture medium was

removed and centrifuged at 5400rpm for 15 minutes

at room temperature. The pellets were taken up in a

volume of 4 mL of PBS and their optical density was

measured for λ = 600 nm. Observations were made at

optical densities ranging from 0.1 to 1.6. The bacteria

were cultured for 2 to 3 days in an oven at 37°C and

stored at 4 ° C after resuspension in PBS.

Multiphoton Microscopy for Bacterial Imaging: A Label-free Solution Resting on Endogenous Two-photon Fluorescence

2.2 The Instrumental and

Computational Pipeline FAMOUS

The raw images of bacteria are recorded thanks to a

multiphoton microscope. The endogenous

fluorescence recorded in this situation exposes MPM

to nonnegligible levels of blur and noise especially in

the case of endogenous fluorescence recordings, this

method requires ad hoc computational processing

strategy to produce 3D images with an adapted visual

quality. In this part, we present the specificities of the

pipeline FAMOUS, combining MPM and our

computational strategy.

2.2.1 Experimental Setup

Multiphoton Microscope. The experimental setup

involved in this set of experiments rests on a standard

multiphoton microscope, associated with an original

computational strategy of image recovery. Figure 1.

presents the resulting experimental setup. The whole

instrumental and computational chain has been

presented in details in (Lefort 2021). A commercial

multiphoton microscope from Olympus (reference

BX61WI) was involved for the whole setup. The laser

excitation had a tunable central wavelength between

650 and 1080 nm. For the imaging experiments, the

central wavelength of the laser excitation, a mode-

locked titanium-doped sapphire laser (Ti: Sa) was

fixed at 740 and 840 nm. When the evaluation of the

two-photon excitation spectra was led, the central

wavelength of excitation was tuned between 680 and

1040 nm with scanning steps of 20 nm. During this

specific procedure, the laser average power was

followed and kept constant thanks to optical densities

all along the recordings. The microscope stand was

composed with a scanning device with two

galvanometric mirrors. The detection system is

composed by a set of two dichroic mirrors and two

photomultipliers tubes (PMTs), each of them being

coupled with a bandpass filter.

Recording of the Instrumental Response Function.

The main experiments of bacterial imaging involved

a single PMT, excepted when the instrumental

contribution to the image has to be characterized (see

below section 2.2.3 Computational solution). In that

case, the second PMT is requisitioned. For this

specific part of the experiment, the biological sample

is mixed with standardized object, composed by

fluorescent microsphere having a diameter of 200 nm.

The fluorescence emission is detected between 575

and 630 nm by the second PMT. The emission

efficiency of these objects is significantly higher than

the endogenous fluorescence emitted by the bacteria.

The sensitivity PMT is consequently adjusted.

Starting from the recordings of the fluorescence

emission of the microspheres, the estimation of the

instrumental response function, also named the 3D

Point-Spread-Function (PSF) is recorded. Two

detection spectral ranges are specifically recorded:

between 420 nm and 500 nm on the one hand, and

between 575 nm and 630 nm on the other hand. The

emission of fluorescence is characterized thanks to a

punctual adaptation of the detection channel. Indeed,

a homemade system of emission detection is

positioned thanks to a fibered spectrometer. Thus, the

emission spectra have been delivered for few

bacteria.

Figure 1: Experimental setup for recording the multiphoton

images of bacteria, associated with a computational strategy

for 3D image recovery.

PHOTOPTICS 2022 - 10th International Conference on Photonics, Optics and Laser Technology

2.2.2 Principle of MPM

MPM is an optical solution devoted to the production

of images which rests on a principle of nonlinear

optics. TPF, a nonlinear process which results from

the third-order of susceptibility of the fluorophore

(3)

, is the main process involved in MPM for

generating the detectable optical signal. Alternative

sources of contrast exists, such as second harmonic

generation (SHG) and three-photon fluorescence

(ThPF) for example. SHG is an instantaneous

phenomenon of nonlinear coherent light scattering

which requires drastic conditions of phase-matching

while combining the simultaneous interaction of two

photons with the structure probed. In our case, the six

bacteria tested are good candidates for two-photon

excitation and do not seem to be at the origin of the

generation of a second harmonic signal. The interest

of MPM for bacteria rests on two facts. First, the

multiphoton process is involved for wavelengths in

the near-infrared range, a spectral range far from the

one-photon absorption and emission ranges and lowly

energetic compared to linear fluorescence processes.

Thus, the bacterial integrity is preserved during the

image recording and the spectral absorption ranges

can be scanned without any risk of spectral

superposition between the excitation and the emission

windows. Thus, the endogenous fluorescence emitted

from bacteria can be recorded. Then, the TPF solution

offers an optical sectioning inherent to its principle

resting on a mechanism weakly likely consequently

occurring exclusively at the focal point of the system

where the photon density is enough for producing a

two-photon absorption process. Parasite light, usually

emitted from the upper and lower optical plans are de

facto highly reduced, contrary to one-photon

processes producing photon in these plans sometime

mechanically removed thanks to a more or less

opened pinhole (confocal microscopy).

2.2.3 Computational Solution

3D image recovery rests on the estimation of the PSF

model thanks to the algorithm FIGARO (Chouzenoux

2019). A multivariable Gaussian fitting process

delivers a mathematical model of the transfer function

of the instrument thanks to the recording of the signal

coming from the fluorescent microspheres. The

instrumental contribution at the image is thus

characterized. Then, the 3D image recovery can take

the contribution of the instrument to the image in

consideration and specifically remove its role. In our

case, for the 3D image recovery, the algorithm from

(Chouzenoux 2013) in its accelerated version, is

involved. The block distributed majorize-minimize

memory gradient (BD3MG) minimizes a least-

squares criterion, regularized with a smoothed total-

variation term reducing noisy artefacts and a

quadratic penalty to constrain the range of the pixel

intensities.

3 LABEL-FREE IMAGES OF

BACTERIA

3.1 3D Endogenous Images of Bacteria

The illustration of bacterial imaging is led with E.

Coli. No labelling process is involved in this part.

Figure 2 illustrates the image recording of E. Coli.

Figure 2A presents the large field of view including

regions of interest where bacteria look isolated; in the

white square, at least two bacteria are identified.

Figure 2B shows the 3D point of view of the bacteria

present into the white square, revealing a potential

Figure 2: Endogenous fluorescence of E. Coli. A. A 2D

record of multiphoton images of E. Coli label free. B. A 3D

reconstruction of the image of E. Coli.

Multiphoton Microscopy for Bacterial Imaging: A Label-free Solution Resting on Endogenous Two-photon Fluorescence

superimposition of two bacteria on each other. The

Bacillus shape expected for E. Coli is observed.

3.2 Control Process of the Object

Shape Detected at the Image

A control set of experiment were led thanks to the

observations resting on labeled bacteria. The

objective was to confirm that the endogenous

fluorescence and the resulting image shape of objects

were emitted from bacteria. Two bacteria were tested

with this protocol. We labeled the bacteria E. coli and

Staphylococcus epidermidis with CFDA (carboxy

fluorescein diacetate), an enzymatic activity indicator

most often used in flow cytometry

. CFDA, a non-

fluorescent molecule, is lysed by the esterase of

bacteria and converted to CF (a fluorescent form)

once in the cytoplasm of the cell, allowing living cells

to fluoresce with enzymatic activity. 10 μL of CFDA

were added to 1 ml of bacterial solutions in PBS. The

solution was placed in an oven at 37°C for 30 minutes

and then rinsed with PBS. The purpose of these

markings is to find bacteria more easily under the

microscope and thus confirm their shape and

distribution on the slide in order to then compare the

results with the absence of marking. Figure 3.

illustrates the images of E. coli recorded with the

labelling with CFDA.

Figure 3: Image of a E. coli cells recorded with the labelling

with CFDA. Excitation wavelength: 740 nm and detection

wavelength fixed between 420 and 500 nm.

Similar information at the image are resulting from

this set of experiment, thus validating the endogenous

images of bacteria presented from the endogenous

fluorescence (Figure 2A). As expected, the level of

contrast is greater with the labelling process involving

a substance efficiently fluorescent. Moreover, the

protocol of labelling bacteria does not produce a

labelling of all cells, since CFDA only stain

enzymatically active bacteria. The quantity of

bacteria revealed at the image looks thus smaller than

without labelling (Figure 2A).

3.3 3D Representation of Bacteria

through the Pipeline FAMOUS

The Clostridium perfringens where imaged in

experimental conditions, fitting with the technical

recommendation required for the correct application

of the instrumental and computational pipeline

FAMOUS: fluorescent microspheres with a fixed

diameter of 200 nm were mixed with the culture of

Clostridium perfringens.

Stacks of images were recorded with an imaging

step of 0.1 µm all along the imaging depth of 8 µm.

Figure 4 shows the superposition of two images: the

raw acquisition recorded in red resulting in more or

less condensed dots and the restored image revealed

in green at the image. Figure 4A represents an

extraction from the 80 2D stacks of the recorded

image of Clostridium Perfringens. Figure 4B

illustrates a 3D reconstruction of the Clostridium

perfringens gathering raw and reconstructed signals.

In the present situation, the stack of the raw images

was recorded with a reduced average power

compared to standard 2D recordings. Indeed, the

duration of a 3D image recording is 80 times longer.

Thus, for a too high average power level deposited on

bacteria, the resulting energy on the sample produces

a movement of the bacteria at the image under the

local temperature increase (Lefort 2015).

4 SPECTRAL ANALYSIS OF

BACTERIA

The origin of two-photon emission fluorescence from

bacteria is investigated through the study of the two-

photon excitation spectra and the spectral range of

emission.

4.1 Two-Photon Excitation Spectra

We characterized the two-photon excitation spectra

of each strain of bacteria. In that aim, the central

wavelength of the laser excitation is tuned each 20

nm, between 680 and 1080 nm and the laser average

PHOTOPTICS 2022 - 10th International Conference on Photonics, Optics and Laser Technology

Figure 4: Image of Clostridium perfringens label free cells.

Red dots: the endogenous emission of fluorescence from

the bacteria. In yellow appears the superimposition of raw

and reconstructed signals. Green: reconstructed signal of

image through the pipeline FAMOUS with a contribution

of red decreased compared to green.

power is kept constant all along the recordings. A ten

of images were recorded for each of the central

wavelengths, each of them containing many

individual bacteria. A ten of bacteria were selected for

each image and their fluorescence intensity were then

recorded in function with the central wavelength of

excitation. Figure 5 gathers the normalized intensity

of the two-photon excitation spectra of each of the six

bacteria.

The emission range of detection is between 420

and 500 nm. The protocol presented at the previous

part is involved. The observation of the fluorescence

excitation ranges of each of the six strain tested

highlights a similar shape, all of them being contained

between 700 and 800 nm.

Figure 5: Two-photon excitation spectra of each of the six

bacteria recorded between 680 and 1080 nm each 20 nm for

a constant average power of 20 mW upon the bacteria.

4.2 Emission Spectra

For two bacteria, we had recorded the two-photon

emission spectra. Figure 6 highlights the emission

spectra for E. Coli and Bacillus Subtilis.

Figure 6: Fluorescence emission spectra for E. Coli and

Bacillus Subtilis.

The emission range of these two bacteria covers the

visible range with a maximum value at 400 nm and

with a full width at the half maximum of 120 nm. We

suspect the fluorescence spectra to be larger in the red

range. But our system, resting on a multiphoton

system, has a dichroic mirror involved for the

separation of the beam between excitation starting at

680 nm and emission fixed with a central wavelength

Multiphoton Microscopy for Bacterial Imaging: A Label-free Solution Resting on Endogenous Two-photon Fluorescence

of 650 nm. The decreasing slope of the fluorescence

emission spectra around 650 nm is due to the dichroic

mirror.

5 HYPOTHESIS ABOUT THE

ORIGIN OF THE

ENDOGENOUS EMISSION OF

FLUORESCENCE

The main question raised by this work concerns the

question of optical characterization of bacteria.

Bacteria contain many fluorescent cellular

components as protein tryptophan, a few other

aromatic amino acids, nucleic acids, and some

coenzymes (Surre 2018, Torno 2013). The spectra of

these autofluorescence cover most of the spectral

range because different endogenous fluorophores

emit at different wavelengths of the electromagnetic

spectrum (Surre 2018). Among the various

endogenous fluorescent molecules, the influence of

flavin adenine dinucleotide (FAD) and nicotinamide-

adenine dinucleotide (NAD) are considered to be

major. Indeed, the fluorescence of reduced

nicotinamide-adenine dinucleotide (NADH) is

commonly used as metabolic biomarker of cell due to

its key role in the conversion of energy. Moreover,

numerous optical studies that exist in literature

present the two-photon excitation and emission

spectra of fluorescence of NADH (Qin 2021,

Hortholary 2021, Georgakoudi 2012, Lakowicz

2006). All of these works highlight a two-photon

excitation range for NADH located between 700 and

800 nm and an emission range, cantered at 450 nm

covers the range between 400 and 600 nm. Moreover,

no distinct shape of bacteria can be visually identified

at the image. Knowing that all of the bacteria contain

metabolic indicators, this cluster of clues leads us to

reasonably think that the fluorescence emission we

have detected is originated from these chemical

substances instead of membrane proteins. Therefore,

for living bacteria the presence of fluorescence must

be produced for all the bacteria presenting a

metabolic activity.

6 CONCLUSIONS

These series of experiments bring new elements of

answer to the question raised in introduction

concerning the difficulties to visualize the bacteria,

their presence, to see their shape, their composition.

resting on two scientific facts, initially known but

usually treated independently until the present works.

Indeed, the endogenous fluorescence of metabolic

indicators is not a new scientific result, whether it be

by a one-photon or a two-photon fluorescence

process. The presence of metabolic indicators in

bacteria is also known. In the present set of

experiments, the interest of a two-photon excitation

strategy lies in the separation of excitation and

emission ranges, with several hundreds of nanometers

(450 nm for emission and 750 nm for the maximum

of excitation). Thus, the endogenous fluorescence

initially known as “optical noise” at the image for a

one-photon fluorescence process is now easily

separable from the excitation beam. Furthermore, our

computational solution, adapted for the low intensity

levels of detected signal resulting in a highly

corrupted image allows to reveal for the first time the

3D optical image of a bacteria in its current

environment without any labelling process. The

construction of the image of the bacteria label free is

thus accessible and gives a visual representation of

the target.

For future works, one question emerges from all

these experiments. The image of bacteria can be

produced through endogenous fluorescence, which is

an important information, especially when

considering the characterization of bacteria density

often led with flow-cytometry through a one-photon

process involving labelling procedure. But by

evidence, the human eye cannot recognize or

differentiate two bacteria from one another when they

originate from two different strains. In the case of a

need concerning the identification of the bacteria, an

assistance of the most recent computational strategies

resting on artificial intelligence has to be deployed.

ACKNOWLEDGEMENTS

We especially thanks Pr Jean-Christophe Pesquet for

his enlightened opinion about this work and the GDR

Imabio from CNRS for his financial support for

Erwan Ferrandon.

REFERENCES

Chalvidal, M., and Chouzenoux, E., (2020) Block

distributed 3MG algorithm and its application to 3D

image restoration IEEE Int. Conf. on Image Processing

(ICIP 2020), Virtual Conf., 20259472

Chouzenoux E., Lau T. T. K., Lefort C. Pesquet J.-C. (2019)

Optimal multivariate Gaussian fitting with applications

PHOTOPTICS 2022 - 10th International Conference on Photonics, Optics and Laser Technology

to PSF modeling in two-photon microscopy imaging. J.

Math. Imaging Vis. 61 1037–50

Chouzenoux, E., Jezierska, A, Pesquet, J.-C. and Talbot, H.

(2013) A majorize-minimize subspace approach for l2-

l0 image regularization. SIAM J. Imaging Sci. 6 563–

91.

Georgakoudi, I., Quinn, K.P. (2012). Optical imaging using

endogenous contrast to assess metabolic state, Annual

Review of Biomedical Engineering ,14, 351-367

Gobel, W., Kampa, B. M. and Helmchen, F. (2007)

Imaging cellular network dynamics in three dimensions

using fast 3D laser scanning. Nat. Methods 4 73–79

Hortholary, T., Carrion, C., Chouzenoux, E., Pesquet, J.-C.

and Lefort, C. (2021) Multiplex-multiphoton

microscopy and computational strategy for biomedical

imaging. Microscopy Research and Technics 84 (7),

1553-1562.

Jin C., Kong L., Dana H., Xie H., Cao L., Jin G., Dai Q.

(2020) Advances in point spread function engineering

for functional imaging of neural circuits in vivo. J.

Phys. D: Appl. Phys. 53 383001

Lakowicz, J. R., (2006). Principles of Fluorescence

Spectroscopy, Springer, Third Edition

Larson, A. M. (2011). Multiphoton microscopy Nat.

Photonics 5 1

Lefort, C. Moreau, D., Lévêque, P., O’Connor, R. (2015)

Optical measurement of temperature in tissue culture

surfaces under infrared laser light excitation at 800 nm

using a fluorescent dye. 2015 International Conference

on Photonics, Optics and Laser Technology

(PHOTOPTICS), IEEE, 41-46

Lefort, C., Chalvidal, M., Paretné A., Blanquet, V.,

Massias, H., Magnol, L., Chouzenoux, E. (2021)

FAMOUS: a fast instrumental and computational

pipeline for multiphoton microscopy applied to 3D

imaging of muscle ultrastructure. J. Phys. D: Appl.

Phys. 54, 274005.

Plotnikov, S. V., Millard, A. C., Campagnola, P. J. and

Mohler, W. A. (2006). Characterization of the myosin-

based source for second-harmonic generation from

muscle sarcomeres. Biophys. J. 90 693–703

Prescott, L. M., Willey, J. M., Sherwood, L. M.,

Woolverton, C. J., Coyette, J., Joseleau, J.-P., Perraud,

R. “Prescott's Microbiology”, 5

edition, De Boeck

supérieur (2018)

Qin, Y., Xia, Y. (2021). Simultaneous Two-Photon

Fluorescence Microscopy of NADH and FAD Using

Pixel-to-Pixel Wavelength-Switching. Frontiers in

Physics, 9, 642302

Surre, J. Saint-Ruf, C., Collin, V., Orenga, S. (2018) Strong

increase in the autofluorescence of cells signals

struggle for survival. Scientific reports (8) 12088

Torno K., Wright, B. K., Jones, M. J., Digman, M. A.,

Gratton, E., Phillips, M. (2013) Real-time analysis of

metabolic activity within Lactobacillus acidophilus by

phasor Fluorescence Lifetime Imaging Microscopy of

NADH. Curr Microbiol (2013) 66:365–367

Wilkinson, M. G., (2018) Flow cytometry as a potential

method of measuring bacterial viability in probiotic

products. Trends in Food Science and Technology (78)

1-10

Yoon, S. A., Park, S. Y., Cha, Y., Gopala, L., Lee, M. H.

(2021) Strategies of Detecting Bacteria Using

Fluorescence-Based Dyes Frontiers in chemistry (9)

7443923

Multiphoton Microscopy for Bacterial Imaging: A Label-free Solution Resting on Endogenous Two-photon Fluorescence