Visualization of contacts with Fluorescence Resonance Energy Transfer using confocal microscopy

(1)

Bachelor Thesis Chemistry

Visualization of contacts with Fluorescence

Resonance Energy Transfer using confocal

microscopy

by

Nicole Oudhof

4

th

_{of July 2017}

Student Number 10780645

Research Institute Supervisor

Van ’t Hoff Institute for Molecular Sciences Prof. Dr. A.M. Brouwer

Research Group Daily Supervisor

(2)

(3)

Abstract

Mechanical contact is the direct reason for the friction we encounter. All surfaces have a certain roughness, thus if two surfaces are close together only a small area comes into actual contact. The main problem with visualizing these contact areas is that they can be on microscopic scale length even if the surfaces are on a way bigger scale. This report will look into Fluorescence Resonance Energy Transfer (FRET) for the visualization of contact between objects. FRET is a quantum mechanical phenomenon which occurs between an excited donor molecule and an acceptor molecule in the ground state. The donor molecules can transfer energy to the acceptor molecule by long-range dipole-dipole interactions between the two types of molecules. The aim of this report is to visualize mechanical contacts on nanoscale with FRET between a flat surface and a plastic sphere using confocal microscopy. To achieve this goal, glass cover slips and PMMA spheres were immobilized with fluorescent dye molecules. By studying the emission of the two dyes of a FRET pair with two photodetectors on a confocal microscope I attempt to quantify the FRET between the objects. The lifetimes of the dyes were also analyzed to determine the amount of FRET. The intensity in the acceptor channel was higher in presence of the donor, which indicates that FRET occurs. The lifetime analysis also suggested that FRET occurs, but the amount cannot be quantified at this stage.

(4)

Samenvatting

Wrijving wordt veroorzaakt door het contact tussen twee oppervlakken. Doordat alle opper-vlakken tot op zekere hoogte ruw zijn, is er maar een klein deel in daadwerkelijk contact. Het probleem met het zichtbaar maken van deze contactoppervlakken is de microscopische schaal waarop het oontact plaatsvindt. Het doel van dit onderzoek is om contactoppervlakken zichtbaar te maken met Fluoresence Resonance Energy Transfer (FRET), gebruikmakend van een confocale microscoop. FRET is een quantum mechanisch fenomeen dat plaatsvindt tussen een ge¨exciteerd donor molecuul en een acceptor molecuul in de grondtoestand. De donormoleculen kunnen ener-gie overdragen op acceptormoleculen met behulp van dipool-dipool interacties. Een confocale microscoop kan met een betere resolutie afbeeldingen maken door een heel klein gaatje voor de lichtdetector. Licht van andere plaatsen in het monster kan hierdoor niet bij de detector komen. Om het doel te behalen zijn er fluorescente moleculen vastgezet op glazen plaatjes en kleine plastic bolletjes. Er is geprobeerd om de hoeveelheid FRET die plaatsvindt tussen de twee oppervlakken te bepalen door te kijken naar de emissie van de donor en acceptor in twee verschillende detectors met een confocale microscoop. Ook is er naar de levensduur van de aangeslagen toestand van de donor gekeken. Het vastzetten van de fluorescente moleculen zorgde voor problemen, maar uiteindelijk is het gelukt om uniform bedekte plaatjes te krijgen. Op de plastic bolleetjes zijn de moleculen succesvol vastgezet. De metingen van de intensiteit van de emissie van de fluorescente moleculen gesuggereerd dat FRET plaatsvindt, doordat er een hogere intensiteit in de lichtdetec-tor van de acceplichtdetec-tormoleculen was dan in afwezigheid van de donormoleculen. De levensduur van de donor was korter geworden op bepaalde plekken, wat ook suggereert dat FRET plaatsvindt. Andere metingen moeten nog gedaan worden om de hoeveelheid te kunnen bepalen.

(5)

1 Introduction

Friction slows us all down every move we make. One-third of the fuel energy used in passenger cars is to overcome the friction in the engine, brakes and other parts of the car.1If we understand friction better on nanoscale, we may change our devices to have less friction and thus use less energy. This could be a big step into the reduction of CO2 emission.

Figure 1: Schematic drawing of contact points between two surfaces on a microscopic scale.2

Mechanical contact is the direct reason for the friction we encounter. All surfaces have a certain roughness, thus if two surfaces are close together only a small area comes into actual contact (Figure 1). The main problem with visualizing these contact areas is that they can be on microscopic scale length even if the surfaces are on a way bigger scale.3_{Previous studies}

have been conducted with viscosity-sensitive fluorescent probes attached on a flat surface. A plastic sphere was pressed on this surface and the fluorescence was measured.4Because the viscos-ity of a solvent is lower at the contact points, strong fluorescence was observed in these areas. A contact area was obtained that agreed well with the classical Hertz theory, which is an indica-tion that the true mechanical contact is obtained. Other studies were carried out with fluorescent liquid, where the amount of fluorescence is a scale for the distance between two surfaces.

In this report Fluorescence Resonance Energy Transfer (FRET) will be used as another method to visualize the

con-tact points. This method is mostly used by biophysicists to study protein-protein interactions and dynamics of structures. It is a quantum mechanical phenomenon which occurs between an excited donor molecule (D) and an acceptor molecule (A) in the ground state (Figure 2).5_{The donor molecules}

can transfer energy to the acceptor molecule by long-range dipole-dipole interactions between the two types of molecules. This makes it a non-radiative energy transfer. Therefore, fluorescence resonance energy transfer is a misleading name for the phenomenon. Other names such as resonance energy transfer and F¨orster resonance energy transfer named after the German scientist Theodor F¨orster are preferred. The name is even more misleading because of the fact that fluorescence is not a requirement for FRET to occur.

There are three criteria to let FRET occur: (i) The emission spectra of the donor should overlap with the absorption band of the acceptor. This makes it possible to transfer the energy from D to A. (ii) The distance between D and A should be around the F¨orster distance (R0), which is the distance

where 50% of the energy gets transferred from D to A. Typical distances are around 5-10 nm. It is best to have distances smaller than R0, but with a bigger distances it is still possible only with low

efficiency’s. (iii) The lifetime of the D should be long enough for FRET to occur. The rate of FRET is dependent on the distance (r) between D and A, and is proportional to r−6, which makes it a great method to measure distances between donor and acceptor molecules. Donor and accceptor molecules that meet the requirements as listed above are called a FRET pair.

Figure 2: Jab lo´nski diagram representation of FRET. Abs, absorption; knr, rate of nonradiative

(7)

Figure 3: The principle of a confocal microscope.7 Data for this study were collected using confocal microscopy.

This is an optical imaging technique that increases the resolution of the microscope images.7 This is achieved by the addition of a spatial pinhole in front of the light detector. Out of focus light will get eliminated, because it cannot go through the pinhole. Figure 3 shows the principle of a confocal microscope. A dichroic mirror is used that reflects the excitation wavelength, but trans-mits the emitted wavelengths. With an objective lens the light can be focused on the sample and the pinhole. The dotted lines represent fluorescence from other points in the sample, that is rejected by the detector aperture as shown. Images obtained with the confocal microscope show the amount of fluorescence in the sample.

The aim of this report is to visualize mechanical contacts with Fluorescence Resonance Energy Transfer between a flat surface and a plastic sphere using confocal microscopy. To achieve this goal glass cover slips and PMMA spheres are im-mobilized with fluorescent dye molecules. It is attempted to determine the amount of FRET that occurs between the two objects is tried to determine by looking at the emission of the two dyes of a FRET pair with two photodetectors on a confocal

microscope. The emission of the dyes was attempted to be separated, to only have the emission of one dye in each photodetector. This is done with different sets of emission filters. The detector with only the emission of the donor dye is called the donor channel and the one with only the acceptor dye is the acceptor channel. Two different FRET pairs will be tried; Fluorescein 27 and Rhodamine B, and Rhodamine 110 and Rhodamine 101 (Figure 4). If FRET would take place between the cover slip and the sphere, the images from both detectors should be different. On the dark places in the donor channel, there should be a high intensity in the acceptor channel and vice versa. And thus a darker spot in the donor channel will indicate that the two areas come into contact at that spot. The lifetimes of the dyes were also analyzed to determine the amount of FRET. The lifetime of the donor should become shorter when FRET occurs and the acceptor lifetime should show a rise in the acceptor channel. To make sure that only the contact points will show, a uniform monolayer of the dye is a necessity. Cl Cl HO O OH O O (a) Fluorescein 27 O OH O N N (b) Rhodamine B O OH O NH H₂N (c) Rhodamine 110 O OH O N N (d) Rhodamine 101

Figure 4: The chemical structures of the FRET fluorophores.

2 Experimental section

2.1 Equipment

Steady state fluorescence was measured on a Spex Fluorolog 3-22. Fluorescence lifetime images were measured with a MicroTime 200 confocal microscope (PicoQuant GmbH) with a 100 1.4 NA objective (UplanSApo, Olympus) mounted on a piezo-scanning stage (Physik Instruments GmbH). A detection pinhole with a diameter of 75 µm was used. The laser used is supercontinuum NKT Photonics wavelength 488 nm and reprate 25 MHz. In the first detector to only detect the donor (in this case fluorescein) 535/50x and 510/20 filters were used. In the second detector to only detect rhodamine b 665/45 and LP635 were used.For R110 and R101 the donor detector had a filter set containing 535/50x and 510/20 filters and the acceptor filter was 593LP.

(8)

2.2 Modification of glass

Cleaning4

Cover slips (High Precision 170±5µm, No. 1.5H, 22 x 22 mm) and spheres were washed in 0.3% Extran AP12 solution by sonication for 30 min, followed by sonicated in Milli Q for 10 min and in ethanol for 30 min. The cover slips were dried in a stream of nitrogen and further cleaned in an ozone photoreactor for 2 hours.

Silanization4

Clean cover slips and spheres were silanized with 2% (volume) (3-Aminopropyl)triethoxysilane (APTES) in 96% ethanol in which 1% (volume) of water was added (Scheme 1). The pH of this solution was adjusted to around 6 with acetic acid. The spheres or a teflon rack with cover slips were kept for 20-25 min in this solution while stirring. Afterwards, the coverslips were sonicated in ethanol for 30 min, dried in a stream of nitrogen and kept in the oven at 110◦C overnight.

Scheme 1: The reaction of the silanization of the glass surface with APTES. Immobilization of dyes4

Dye (0.04 mmol, 1 eq), (Benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexafluorophosphate (BOP) (53 mg, 0.12 mmol, 3 eq), N-hydroxybenzotriazole (HOBt)(18.4 mg, 0.14 mmol, 3.5 eq) and diisopropylethylamine (DIPEA) (41.8 µL, 0.24 mmol, 5.3 eq) were added to the silanized cover slips in a teflon rack and/or glass spheres in dry DMF (60 mL). The reaction mixture was stirred at room temperature for 3 hours under nitrogen atmosphere. Afterwards, the cover slips and spheres were washed with acetone, DCM and ethanol, followed by drying under a stream of nitrogen.

2.3 Modification of PMMA spheres

For the modification of poly(methyl methacrylate) (PMMA) spheres with these methods it is necessary to use crosslinked PMMA spheres. Otherwise the spheres will dissolve and crack in the solvents. Cleaning8

The spheres (1 mm, crosslinked PMMA) were washed in a mixture of 2-propanol:water = 1:1 for 10 min with sonication. Afterwards they were dried in a stream of nitrogen.

Aminolysis8

The cleaned PMMA spheres were stirred in a 1 M ethylenediamine solution in DMSO for 20-25 min. Then washed with 2-propanol and dried in a stream of nitrogen. See Scheme 2 for a schematic diagram of the aminolysis.

O O O NH

NH₂

1 M ethylenediamine DMSO

(9)

Immobilization of dyes

Dye (0.04 mmol, 1 eq), (Benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexafluorophosphate (BOP) (53 mg, 0.12 mmol, 3 eq), N-hydroxybenzotriazole (HOBt)(18.4 mg, 0.14 mmol, 3.5 eq) and diisopropylethylamine (DIPEA) (41.8 µL, 0.24 mmol, 5.3 eq) were added to the aminolyzed PMMA spheres in dry DMF (60 mL). The reaction mixture was stirred at room temperature for 3 hours under nitrogen atmosphere. Afterwards, the cover slips and spheres were washed with acetone, DCM and sonicated in ethanol for 10 min and dried under a stream of nitrogen.

3 Results and Discussion

3.1 Silanization of glass

To verify that the glass surface of the cover slips were silanized with APTES, the water contact angle was determined. This is a measure of the hydrophobicity of the surface. A water droplet is dropped on a solid surface and the angle between the droplet and the liquid-solid interface and the liquid-vapor interface will be determined.9 Figure 5 shows an illustration with different contact angles of a droplet on a surface. With a small contact angle the droplet is almost flat and has the most contact with the surface. At the same time, this contact is smaller with a larger contact angle. Hence, when the interactions between the droplet and the surface are favorable, the contact angle will be small. When the interactions are unfavorable the contact angle will be large. In this experiment, the contact angle was measured with water as the droplet. Therefore, a larger contact angle means that the surface is more hydrophobic, which is the case if APTES is functionalized on the glass surface. Table 1 shows the water contact angles of a clean cover slip and of a silanized cover slip with APTES. The contact angle with APTES is in agreement with literature.10_{Since the angle became larger after silanization,}

the method works.

Table 1: Water contact angles of clean and silanized cover slips. Type w.c.a (◦₎

Clean 14 ±1 APTES 37 ±2

Figure 5: Illustration of a droplet on a solid face with different contact angles.9

3.2 Immobilization of glass with fluorescent dyes

Fluorescein 27 and Rhodamine B

The first FRET pair that was tried was Fluorescein 27 and Rhodamine B of which the structures are shown in Figure 4. Both dyes were immobilized on glass cover slips according the procedure described in the experimental section only with different reaction times. This is a known procedure, however every dye has a different reactivity towards the reagent. That is the reason that the every dye has an other optimal reaction time. Different reaction times were tried to find the optimal time to get bright and uniformly immobilized cover slips. The figures show different batches of the cover slips. Afterwards, the cover slips were measured with a spectrofluorometer.

Figure 6a and 7a show the emission spectra of Fluorescein 27 immobilized cover slips with different reaction times, when excited with a excitation wavelength of 480 nm and a slit width of 5 nm. In

(10)

Figure 6a it seems like around 30 min would be the optimal time, but in Figure 7a reaction times of 40-50 min give the highest emission. Thus to find the right time, the experiment should be done again. However, time did not allow to carry these out.

The shifts of the maxima can be seen in the normalized emission spectra (Figures 6b and 7b). In 6b there is a clear trend in the shifts of the maximum. They shift to longer wavelengths with a longer reaction time. This could be caused by the formation of aggregates or by an error of the machine. As cover slips are measured in stead of a solution, a bigger error can be expected. The angle of excitation of the cover slip can vary, which could also have caused this shift. This is most likely also the case for the shift for 20 min in Figure 7b.

(a) (b)

Figure 6: Emission spectrum (a) and normalized emission spectrum (b) of Fluorescein 27

immobilized on a glass cover slip with different reaction times. An excitation wavelength of 480 nm was used and a slit width of 5 nm.

(a) (b)

Figure 7: Emission spectrum (a) and normalized emission spectrum (b) of Fluorescein 27

The same cover slips as measured in Figure 7 were measured on the confocal microscope. Figure 8 provides the images obtained by these measurements when excited with 475 nm. The cover slip in Figure 8a is nonuniformly covered, seen by the obvious spots. Figures 8c-d show quite uniformly covered because the intensity is almost the same everywhere in the measured sample. Longer reaction times obtain brighter cover slips. However, these intensities are too low for the measurement of FRET. Even longer reaction times could result in brighter cover slips. This is very promising to maybe obtain

(11)

cover slips with higher intensities with longer reaction times. Due to time limitations, another FRET pair is tried.

(a) 20 min (b) 30 min (c) 40 min (d) 50 min

Figure 8: Images made with the confocal microscope of cover slips after the immobilization of Fluorescein with different reaction times and cleaning methods. Excited with 488 nm. Different reaction times were used. High intensity of fluorescence corresponds to high pixel values in the images.

The low intensities may be caused by a wrong pH of the solution. For the immobilization a peptide coupling reagent is used. For the peptide coupling to take place between the amine functionalized on the surfaces and the carboxylic acid group of fluorescein, the carboxylic acid should be deprotonated. See SI1 for the mechanism of the peptide couplings reaction. However, there is also an alcohol group that could get deprotonated, that cannot couple to the surface. This could be the an explanation for the disappointing intensities obtain with the confocal microscope. Another reason could be the formation of aggregates, because the alcohol and carboxylic acid group could react towards an ester in solution. No monolayer can be formed anymore if this happens. However, the images taken with the confocal microscope show up quite uniform. If aggregates are formed, they do not get attached to the surface apparently. It is known from literature that the quantum yield of fluorescein dyes becomes higher in a slightly basic solution.11 _{When measuring the sample, no basic solvent was present, hence}

the quantum yield could be too low.

Figures 9a and 10a show the emission spectra of cover slips immobilized with Rhodamine B with different reaction times. The figures show different batches of immobilized cover slips. In 9a a reaction time of 10 min seems to be optimal, but 10a has the highest emission after 60 min. There is no trend in the amount of emission with the different reaction times. Reactions times of longer than 60 min should be tried again. Since the immobilization of Fluorescein did not work out sufficiently enough, another FRET pair was tried. There was no need to optimize the reaction time of Rhodamine B. Figures 9b and 10b show the normalized emission spectra. Only the 5 and 20 min curves in Figure 9b show a shift in their maximum, which is most likely caused by an error of the machine due to an other angle of excitation. No trend in the shifts of the maxima can be found. To get a more quantitative measure of the amount of fluorophores attached on the surface, absorbance measurements should be done.

(12)

(a) (b)

Figure 9: Emission spectrum (a) and normalized emission spectrum (b) of Rhodamine B

(a) (b)

Figure 10: Emission spectrum (a) and normalized emission spectrum (b) of Rhodamine B

Rhodamine 110 and Rhodamine 101

Next, Rhodamine 110 (R110) and Rhodamine 101 (R101) were tried as a FRET pair. Both dyes were immobilized on a glass cover slips following the procedure described in the experimental section with different reaction times to find the optimal reaction time. The cover slips were measured under the confocal microscope afterwards.

Figure 11 shows images taken with the confocal microscope of cover slips immobilized with R110 with different reaction times and cleaning methods. Reaction times of 35 and 75 min result in non uniformly covered cover slips, seen by the spots (Figures 11a and 11b). After longer reaction times they seemed to become more uniform (Figures 11c and 11d), yet these were still not reproducible. When a reaction time of 3h was tried with sonication in ethanol for 10 min instead of rinsing, quite uniformly immobilized cover slips were obtained (Figure 11e). However, these cover slips show a lower intensity than with just rinsing with ethanol. It seems to be that not all dye molecules are chemically attached to the surface. The dye molecules that are not chemically attached are not removed by rinsing with solvent, hence sonication is needed. However, cover slips with a decreased intensity are produced. This suggests that there are other interactions between the surface and the dye molecules.

(13)

These are most likely electrostatic interactions between the positive APTES surface and the anionic dye molecules.12

(a) 35 min (b) 75 min (c) 105 min (d) 180 min (e) 180 min with sonication in ethanol

Figure 11: Images made with the confocal microscope of cover slips after the immobilization of R110 with different reaction times and cleaning methods. Excited with 560 nm. Different reaction times and cleaning methods were tried. High intensity corresponds to high pixel values in the images.

Figure 12 shows images taken with the confocal microscope of cover slips immobilized with R101 with different reaction times and cleaning methods. Reaction times of 30 and 60 min (Figures 12a and 12b) result in nonuniformly covered cover slips, seen by the spots. After longer reaction times they seemed to become more uniform (Figures 12c and 12d), yet these were still not reproducible. When a reaction time of 3h was used with sonication in ethanol instead of rinsing, quite uniformly immobilized cover slips were obtained (Figure 12e). However, with a lower intensity, the same as happened with the immobilization of R110.

(a) 30 min (b) 60 min (c) 100 min (d) 155 min (e) 180 min with sonication in ethanol

Figure 12: Images made with the confocal microscope of cover slips after the immobilization of R101 with different reaction times and cleaning methods. Excited with 560 nm. Different reaction times and cleaning methods were tried. High intensity corresponds to high pixel values in the images.

3.3 Immobilization on PMMA

Figure 13: Schematic representation of the method to perform pressure on the PMMA sphere.

PMMA spheres were labeled with the fluorescent dyes. First, non-crosslinked PMMA spheres were immobilized, but these dissolved in DMSO and DMF, which are the solvents that are used with the immobi-lization. From literature it is known that this can happen.13_{To make the}

immobilization possible, crosslinked PMMA spheres need to be used. Figure 14 shows images taken with the confocal microscope of immo-bilized PMMA spheres. They were pressed on a clean cover slip with a force of 3.6 N, as shown in Figure 13. The weight on top of sphere always pushes with an angle, so that one part of the sphere is closer to the sur-face than the other part. It is expected that if the dye is immobilized on the sphere the parts of the bead that pressed on the cover slips will show fluorescence. In all the images there are brighter areas as expected, which is most likely caused by the immobilization of the spheres. However, there is no solvent used between the two surfaces, which makes it possible that the intensities are caused by refraction of light that could occur at the contact points. More experiments should be done to determine if this is

(14)

the sphere and the cover slip to avoid the effects of refraction of light. DMSO has a refractive index close by the ones of PMMA and glass, which should avoid the refraction effects. Nothing can be said about the full immobilization of the sphere, because of the roughness not all areas can be seen on the microscope images. Another methods should be used to determine this. However, these images give an indication that the spheres are actually labeled with the dyes.

(a) Rhodamine B (b) R110 (c) R110 (d) R101

Figure 14: PMMA spheres with different immobilized dyes pressed with a force of 3.6 N on a clean cover slip, measured with the confocal microscope. High intensity corresponds to high pixel values in the images. a-c were excited with 488 nm and d with 560 nm.

3.4 FRET experiments

Fluorescein 27 and Rhodamine B

Figure 15: Absorption and emission spectra of Fluorescein (green) and Rhodamine B (red). The blue dotted line shows the overlap of the emission of Fluorescein and the absorption of Rhodamine.14 The first FRET pair used was Fluorescein 27 (donor) and Rhodamine B (acceptor). This FRET pair has already been used in previous studies, because there is good overlap of the emission spectrum of Fluorescein with the absorption spectrum of Rhodamine B (Figure 15) and the emission of both dyes are quit easily separated with filters.15 _{Besides, both dyes have large molar absorption coefficients}

and high quantum yields. The FRET pair has a R0 of 49 ˚A.16

If FRET would take place, it is expected that only on the contact points there is intensity in the acceptor channel. All the places where the distance is more than two times the F¨orster distance, no intensity is expected in this channel. For the donor channel it is the other way around. It is expected that on the contact points there is no intensity and on the other places there will be.

In the following images the green image represents the intensities in the donor channel and the red image shows the intensities in the acceptor channel. An overlay of the two is also presented.

Figure 16 presents images taken with the confocal microscope of Fluorescein immobilized on a glass sphere and Rhodamine B immobilized on the cover slip. No extra pressure was exhibited on the glass sphere. It is apparent from these images that the intensity in both channels is very low and the same places have a higher intensity. This can be caused by emission of the acceptor dye in the donor channel and vice versa, which is called bleedthrough. Another explanation could be direct excitation of the acceptor. Due to the low intensities it was decided to work with another FRET pair. Additionally, PMMA beads were tried to get a bigger contact area.

(15)

(a) Donor channel (b) Overlay (c) Acceptor channel

Figure 16: Microscope images taken with a confocal microscope of Fluorescein 27 (donor)

immobilized on a glass sphere and Rhodamine B (acceptor) on cover slip. An excitation wavelength of 475 nm was used. High intensity corresponds to high pixel values in the images.

Rhodamine 110 and Rhodamine 101

Figure 17: Absorption and emission spectra of Rhodamine 110 (green) and Rhodamine 101 (red).17

The next FRET pair that was used was R110 and R101 (see Figure 4 for the chemical structures). R101 has additional rigid rings in comparison with R110 which causes a shift of the absorption and emission maxima to higher wavelengths.18 _{Figure 17 shows the absorption and emission spectra of}

R110 and R101. It can be seen that R110 functions as donor in this FRET pair and R101 as the acceptor. There is overlap between the absorption band of R101 and the emission band of R110, however the emission will be difficult to separate completely. The calcuted value for R0is calculated

to be 56 ˚A.17

Figure 18 and Figure 19 show images taken with the confocal microscope of R110 on the cover slip and R101 on the PMMA sphere when excited with 488 nm. The sphere was pressed with 3.6 N onto the cover slip. In Figure 18 the intensity is quite high in both channels. However, in the acceptor channel most places are quite bright, which cannot be explained by FRET. A reason could be that emission from the donor dye is also detected in the acceptor channel. The dark spots in the donor channel are on the same place as the spots in the acceptor channel. This can also be explained by the bleedthrough. From these images it cannot be concluded that FRET occurs. The acceptor and donor bleedthrough should be determined to give a real indication about the amount of FRET.

In the acceptor channel of Figure 19 is nice that there are dark spots in the donor channel, while the rest of the channel is quite uniform. However, the intensity of the donor is very low and the places where it is dark in the donor channel, it is also dark in the acceptor channel. This is most likely caused by bleedthrough. The intensity in the acceptor channel is higher in presence of the donor, this gives an indication that FRET occurs.

(16)

Figure 18: Microscope images taken with a confocal microscope of R110 (donor) immobilized on a cover slip and R101 (acceptor) on a PMMA sphere. A force of 3.6 N was applied on the sphere. An excitation wavelength of 488 nm was used. High intensity corresponds to high pixel values in the images.

Figure 19: Microscope images taken with a confocal microscope of R110 (donor) immobilized on a cover slip and R101 (acceptor) on a PMMA sphere. A force of 3.6 N was applied on the sphere. An excitation wavelength of 488 nm was used. High intensity corresponds to high pixel values in the images.

Figure 20 and Figure 21 present images taken with the confocal microscope with R101 (acceptor) on the cover slip and the PMMA sphere immobilized with R110 (donor). An additional FRET efficiency image is included which is generated by FRET and Colocalization analyzer ImageJ Plug-in. For this analysis some extra images are needed, which are provided in SI2. The higher the index in the FRET index images, the more colocalization and FRET takes place at those spots. For this Plug-in the donor and acceptor channel bleedthrough is compensated. This means that the same part of the sphere needs to be found twice, which makes it hard to analyze. The Plug-in does not calculate the distance between the surfaces or the energy efficiency. For further analysis this should be done with another plug-in or analysis method.

From the FRET efficiency images can be thought that FRET occurs. However, not exactly the same part of the bead was found twice. The bead was glued on the holder, but still it was challenging to find the same area. This can give wrong information about the colocalization and thus the index can be a bit off. Another method should be thought of to obtain the same image of the bead more easily. If this works, the FRET efficiency image could give more information if FRET really occurs. Due to the irregularities of the place on the bead, it cannot be said that FRET occurs in this system. The intensity is again higher in the acceptor channel in presence of the donor, which indicates that FRET occurs. This amount can not be quantified due to limitations.

(17)

(a) Donor channel (b) Overlay (c) Acceptor channel (d) FRET efficiency

Figure 20: Microscope images taken with a confocal microscope of R101 (acceptor) immobilized on a cover slip and R110 (donor) on a PMMA sphere. A force of 3.6 N was applied on the sphere. An excitation wavelength of 488 nm was used. (a-c) High intensity corresponds to high pixel values in the images (d) high pixel values indicate that FRET occurs at these points.

(a) Donor channel (b) Overlay (c) Acceptor channel (d) FRET efficiency

Figure 21: Microscope images taken with a confocal microscope of R101 (acceptor) immobilized on a cover slip and R110 (donor) on a PMMA sphere. A force of 3.6 N was applied on the sphere. An excitation wavelength of 488 nm was used. (a-c) High intensity corresponds to high pixel values in the images (d) high pixel values indicate that FRET occurs at these points.

In the previous images, the two surfaces did not have solvent between them. This makes effects of light refraction possible at the contact points. To rule this out, a solvent with a similar refractive index to the glass and PMMA can be used between the surface. An example of such a solvent is DMSO. Another reason to try DMSO was to see if this would enhance the quantum yields and the amount of FRET observed. Structural flexibility of the groups on the nitrogen atoms can reduce the quantum yield.19 _{This can cause a decrease in the quantum yield of R110 in the presents of a}

solvent and thus a lower intensity. The orientation of the immobilized dyes could also change in the presence of a solvent, which could also have an effect on the amount of FRET. The F¨orster distance is proportional to the quantum yield and orientation factor, hence when these transition the spots of FRET could change in size.

Figure 22 presents the images of a PMMA sphere with R101 onto it pressed onto a cover slip immobilized with R110 while both are in contact with DMSO. The images do not drastically change, thus the other images do not only show the effects of refraction. The intensity in the acceptor channel increased in comparison with the systems without DMSO. However, in the donor channel the intensity is still very low. The size of the spots did also not change drastically, but they seem to be a bit smaller. Thus the F¨orster distance could be changed by the contact of DMSO, but not very drastically.

(18)

Figure 22: Microscope images taken with a confocal microscope of R110 (donor) immobilized on a cover slip and R101 (acceptor) on a PMMA sphere. These were both in contact with DMSO. A force of 3.6 N was applied on the sphere. An excitation wavelength of 488 nm was used. High intensity corresponds to high pixel values in the images.

The other analysis method was to look at the lifetimes of the species in the spots. If FRET takes place, the lifetime of the donor will decrease, due to an extra dissipation pathway for the excited state energy of the donor.20 The distance can be quantified by the reduction of the lifetime of the donor. The main advantage of this analysis method in comparison to the first one is that no corrections have to be made for bleedthrough.

Figure 23 shows fluorescence lifetime images taken with the confocal microscope of the lifetimes of the species in the donor and acceptor channel. This is the same place as in Figure 18 only the lifetimes are presented here. In a few points the lifetimes are decreased in the donor channel (Figure 23a), which indicates that FRET occurs. However, the amount cannot be quantified. To have a quantification, the scanning time per pixel should be increased. This will give more precise lifetimes to calculate the reduction of the donor correctly. The lifetime of R101 (acceptor) in methanol is 4.9 ns21_{and of R110 (donor) the lifetime is 4.2 ns.}22_{The average lifetime in the acceptor channel (Figure}

23b) is higher than in the donor channel in the whole sample, which is expected since the lifetime of the acceptor is higher. It is remarkable that the lifetime is higher throughout the whole sample, since it should only be present on the contact points. Thereby, in both channels the lifetimes are lower than in methanol. Most likely the absence of solvent caused this decrease in lifetime.

Figures 23c and 23d show the fluorescence intensity and the average lifetime in a cross section of the microscopic images. In Figure 23c the intensity and lifetime decrease at the same spots. However, in Figure 23d around 7 µm the intensity is high, but the lifetime decreased. This is a rather remarkable result. The intensity in the acceptor channel is around three times higher than in the donor channel which is directly excited. This is an indication that FRET should occur.

In order to verify the legitimacy of the approach, contact area size in the measurements was compared to the Hertz contact area size. The classical Hertz theory describes the radius a of the contact area between a sphere with radius R and a flat surface pressed against each other with force F.4 _{The equation below describes the this. In the equation, v is the Poisson ratio of the material}

of the sphere and E is the Young’s modulus. From literature it is known that the Poisson ratio of PMMA is 0.3523 _{and the Young’s modulus is around 2.0 GPa.}4 _{PMMA spheres with a radius of 1}

mm were used, which results in a value of a around 84 µm. This does not correspond with the size of the areas in Figures 19, 20 and 22. This areas have a diameter of around 80 µm, but this could be caused by the angle of pressure.

a3=3R(1 − v

2₎

(19)

(a) Donor channel (b) Acceptor channel

(c) Donor channel cross section (d) Acceptor channel cross section

Figure 23: (a-b) Fluorescence lifetime images taken with a confocal microscope of the contact between R110 (donor) immobilized on a cover slip and R101 (acceptor) on a PMMA sphere. A force of 3.6 N was applied on the sphere. An excitation wavelength of 488 nm was used. The pixel value gives the lifetime of the species in ns. (c-d) Cross sections in the images in (a-b). The average lifetime and the fluorescence intensity are plotted.

4 Conclusion

The aim of the present research was to visualize mechanical contacts with the use of Fluorescence Resonance Energy Transfer between a flat surface and a plastic sphere. Glass cover slips were suc-cessfully silanized with APTES proven by the water contact angle, but with the immobilization of dye more struggles were encountered. Sonication removed the irregularities, hence quite uniformly immobilized cover slips were obtained seen on the confocal microscope. PMMA spheres also seemed to be immobilized with the dyes. The intensity in the acceptor channel was higher in presence of the donor, which indicates that FRET occurs. The lifetime analysis also suggests that FRET occurs, but the amount cannot be quantified due to time limitations.

5 Outlook

An even better understanding of the immobilization of the dyes with the peptide coupling reagent should be achieved, to form a bright, uniform monolayer of the dyes. First of all, another molecule could be silanized on the glass surfaces, such as another silane or a silatrane. Silatranes have stronger donor–acceptor interactions between nitrogen and silicon atoms that make them chemically more stable to hydrolysis than trialkoxysilanes.24_{It has been suggested that this will prevent the formation}

of polymer cluster and thus form a more uniform basis for the dye immobilization. Secondly, what could be tried to enhance the immobilization of the dyes is to first synthesize the dye with APTES,

(20)

another silane or silatranes and after this immobilization on the glass surface. Scheme 3 shows an example of this idea with R110. It seems to be that the peptide coupling is the step with the most complications and those could be avoided with this method.

O NH O HN H₂N Si O O O OH OH SiO2 O Si O Si O O O HN SiO₂ NH₂ NH O O O O HN H₂N NH

Scheme 3: A reaction equation of silanized R110 on glass.

It would also be good to determine the surface density of the immobilized dyes on the glass surfaces quantitatively. This could be done with absorption spectroscopy, where the amount of absorption of the cover slip gives an indication for the density. From this method can however not be concluded if the cover slip is uniformly immobilized. For the cover slips Atomic Force Microscopy (AFM) could also be measured to see if there are aggregates formed, which is a sign that the cover slip is not uniformly covered.

A method for the surface density of the spheres could also be done with absorption. This method uses the fact that a APTES surface has a positive charge.12_{If an anionic dye molecule is in solution}

and the absorption is measured with and without the sphere in there, an indication can be made about the amount of dye molecules attracted by the APTES. This can give an indication about the amount of APTES on the surface of the glass sphere. Again, nothing can be concluded about the uniformity of the APTES layer.

A method should be developed for the determination the uniformity of the surface of the spheres after aminolysis and immobilization of the dye. If very smooth spheres are available, AFM could be tried to check for aggregation. However, an attachment should be glued on the sphere and this glue should not dissolve in the solvents that are used to modify the spheres. The reaction time for the immobilization should also be investigated, because in this report the same reaction time was used for the glass and PMMA surfaces. However, PMMA surfaces could reaction differently due to other amino compounds functionalized on the surfaces.

Lastly, the analysis for the FRET images should be looked into more. There were some programs found to analyze the images taken with the confocal microscope. A decision should be made of which analysis method will be used; with the emission or lifetimes. The measurement should be adapted to the method. For the emission analysis, it should be tried to decrease the bleedthrough and direct excitation of the acceptor. The measure time per pixel should be increased with the lifetime analysis. The amount of FRET and the energy efficiency’s should be tried to be determined.

(21)

Acknowledgements

I would like to thank prof. A.M. Brouwer for the opportunity to work in his group, Dina for the daily guidance throughout the project. I would also like to thank Teun for sharing his knowledge about LaTeX and helping with all the small problems. Dongdong is also acknowledged for the help with printing and Danny, Tom and Rosa for the feedback on my report. Prof. dr. D. Bonn is thanked for his work as my second corrector. Last but not least, I would like to thank the whole Molecular Photonics group and Rosa for the nice lunches and cookie/coffee times together.

(22)

References

[1] Holmberg, K.; Andersson, P.; Erdemir, A. Tribology International 2012, 47, 221–234.

[2] Cutnell, J.; Johnson, K.; Stadler, S. Physics, 10th ed.; Wiley: United States of America, 2014; pp 94–100. [3] Bowden, F.; Tabor, D. Proc. R. Soc. A 1939, 169, 391–413.

[4] Suhina, T.; Weber, B.; Carpentier, C. E.; Lorincz, K.; Schall, P.; Bonn, D.; Brouwer, A. M. Angew. Chem., Int. Ed. 2015, 54, 3688–3691.

[5] Lakowicz, J. R. Principles of fluorescence spectroscopy, 3rd ed.; Springer: Singapore, 2006; pp 443–476.

[6] Gryczynski, Z.; Gryczynski, I.; Lakowicz, J. R. In Molecular Imaging; Periasamy, A., Day, R. N., Eds.; American Physiological Society: San Diego, 2005; pp 21–56.

[7] Valeur, B.; Berberan-Santos, M. Molecular Fluorescence: Principles and Applications; Wiley: Weinheim, 2001; pp 354–355.

[8] Brown, L.; Koerner, T.; Horton, J. H.; Oleschuk, R. D. Lab Chip 2006, 6, 66–73.

[9] Yuan, Y.; Lee, T. R. In Surface Science Techniques; Bracco, G., Holst, B., Eds.; Springer Berlin Heidelberg: Berlin, Heidelberg, 2013; pp 3–34.

[10] Asenath Smith, E.; Chen, W. Langmuir 2008, 24, 12405–12409.

[11] W¨urth, C.; Grabolle, M.; Pauli, J.; Spieles, M.; Resch-Genger, U. Anal. Chem. 2011, 83, 3431–3439, PMID: 21473570.

[12] Qin, M.; Hou, S.; Wang, L.; Feng, X.; Wang, R.; Yang, Y.; Wang, C.; Yu, L.; Shao, B.; Qiao, M. Colloids Surf., B: Biointerfaces 2007, 60, 243–249.

[13] Evchuk, I. Y.; Musii, R. I.; Makitra, R. G.; Pristanskii, R. E. Russ. J. Appl. Chem. 2005, 78, 1576–1580. [14] Kuscu, M.; Kiraz, A. G.; Akan, O. B. Sci. Rep. 2015, 5, 7831–7837.

[15] Wanichacheva, N.; Hanmeng, O.; Kraithong, S.; Sukrat, K. J. Photochem. Photobiol., A 2014, 278, 75–81. [16] Cardullo, R. A.; Agrawal, S.; Flores, C.; Zamecnik, P. C.; Wolf, D. E. Proc. Natl. Acad. Sci. U.S.A. 1988, 85,

8790–8794.

[17] Synak, A.; Bojarski, P.; Grobelna, B.; Ku lak, L.; Lewkowicz, A. J. Phys. Chem. C 2013, 117, 11385–11392. [18] Liu, J.; Diwu, Z.; Leung, W.-Y.; Lu, Y.; Patch, B.; Haugland, R. P. Tetrahedron Lett. 2003, 44, 4355–4359. [19] Vogel, M.; Rettig, W.; Sens, R.; Drexhage, K. H. Chem. Phys. Lett. 1988, 147, 452–460.

[20] Sun, Y.; Rombola, C.; Jyothikumar, V.; Periasamy, A. Cytometry Part A 2013, 83, 780–793.

[21] Beaumont, P. C.; Johnson, D. G.; Parsons, B. J. J. Chem. Soc. Faraday Trans.C 1993, 89, 4185–4191. [22] Tellinghuisen, J.; Goodwin, P. M.; Ambrose, W. P.; Martin, J. C.; Keller, R. A. Anal. Chem. 1994, 66, 64–72. [23] Jee, A.-Y.; Lee, M. Polym. Test. 2010, 29, 95–99.

[24] Shlyakhtenko, L. S.; Gall, A. A.; Filonov, A.; Cerovac, Z.; Lushnikov, A.; Lyubchenko, Y. L. Ultramicroscopy 2003, 97, 279–287.

(23)

Supporting information

SI1 Mechanism of peptide coupling

For the immobilization of dye, peptide coupling reagents BOP and HOBt are used also known as Castro’s reagent. The condensation reaction for the formation of the amide bond is an equilibrium, which lies to the free amine and carboxylic acid that have undergone an acid-base reaction.25_Because

of this, the acid should first be activated by attachment of a good leaving group on the acyl carbon. With these reagents an activated ester is formed, because the electrophilicity of the carbon centre of the carbonyl is increased by the electron withdrawing groups on the alcohol. The acid is first deprotonated with DIPEA reacts with BOP to generate OBt (minus charge) and an activated acylphosphonium species. HOBt reacts with the activated acid to produce a reactive Bt ester, which can react with the amine species. The driving force of this reaction is the generation of the corresponding oxide (Scheme 4). Castro’s reagent is very effective, but generates hexamethylphosphoric triamide (HMPA), which is extremely toxic.

R OH O N N H R O O P N N N O N N N P N N N O N N N O R O N N N O P O N N N R O O N N N N H R' H R O O N N N N H _H R' N N N O R O N R' H

(24)

SI2 Extra images for the Plugin

(a) (b) (c)

Figure 24: Images taken with the confocal microscope of immobilized cover slips. (a) and (b) donor channel with only R110 immobilized on a sphere, that is pressed on a clean cv with a force of 3N. n excitation wavelength of 488 nm was used. (c) acceptor channel with a cover slip immobilized with R101. An excitation wavelength of 560 nm was used.

Visualization of contacts with Fluorescence Resonance Energy Transfer using confocal microscopy

Bachelor Thesis Chemistry