Single-molecule FRET on dual-labelled DNA Holliday junction

Single-molecule FRET on

dual-labelled DNA Holliday

junction

THESIS

submitted in partial fulfillment of the requirements for the degree of

BACHELOR OF SCIENCE

in PHYSICS

Author : Olivier M. van Capelleveen

Student ID : s1856502

Supervisor : Zohre Eskandari Alughare MSc Prof. dr. Michel Orrit 2ndcorrector : Prof. dr. ir. John van Noort

Single-molecule FRET on

dual-labelled DNA Holliday

junction

Olivier M. van Capelleveen

Huygens-Kamerlingh Onnes Laboratory, Leiden University P.O. Box 9500, 2300 RA Leiden, The Netherlands

July 2, 2019

Abstract

Single-molecule FRET is a method used to measure distances in the nanometer range. It makes use of fluorescence of a donor and acceptor attached to the molecule to measure that distance. FRET can be used to study conformational dynamics of molecules, which was the aim of this

experiment. The studied molecule is the Holliday junction (HJ), a synthetic molecule made up of four cross-linked DNA strands. It moves

such that when two fluorescent dyes are attached to two arms the distance between those arms can be measured using FRET, and so the movement of the HJ can be described. The measurements are made in room temperature and measured in presence or absence of MgCl₂. For

lack of time very few measurements could be done, leading to no spectacular results. There is a possibility FRET is measured, but because of possible leakage from the signals to either of the two detectors, no clear

conclusions could be drawn. As well as intensity being unreliable, the lifetime measurements made, can be considered unreliable, since of the

five measurements three came out as higher than the regular lifetime, which is the exact opposite of what is to be expected.

Contents

1 Introduction 7

1.1 Fluorescence 7

1.2 FRET 8

2 Methods 13

2.1 Single-molecule immobilization on glass 13

2.1.1 Materials 13

2.1.2 Cover slip silanization 14

2.1.3 Single-molecule Holliday junction immobilization on

glass 14

2.2 Confocal microscope setup 15

3 Results 17

3.1 Data analysis 17

3.1.1 Scanning images of dual-labelled DNA Holliday

junc-tion 17

3.1.2 Room temperature time trace fluorescence intensity 21 3.1.3 Room temperature fluorescence life time 25 3.2 Determining FRET efficiency based on fluorescence

inten-sity and donor lifetime 26

4 Discussion and conclusion 29

4.1 Discussion 29

Chapter

1

Introduction

This experiment consisted of two major parts. One is the immobilization of a single molecule to be measured on a glass cover slip. The other part is the concept of FRET, the measurement technique used on the single molecule.

1.1

Fluorescence

Fluorescence is the process of light emission by a molecule after excitation by a light source or some other form of electromagnetic radiation.1 It has important applications in the studying of molecules. Its lifetime has been used to study the structures of molecules,2and fluorescence spectroscopy has given much insight into the conformational dynamics of numerous molecules. In figure 1.1 the process of FRET, one of the ways fluorescence spectroscopy can be applied, is shown.

Figure 1.1: Process of FRET. The donor is excited and passes its energy on to the acceptor through resonance energy transfer. The acceptor in turn fluoresces.3

1.2

FRET

F ¨orster resonance energy transfer, or fluorescence resonance energy trans-fer, commonly written as FRET, is a method used to measure small dis-tances; either between molecules, so called intermolecular, or within a molecule, so called intramolecular. The distances are usually in the 10-80 ˚A range,4, 5that’s why FRET is known as a ”microscopic ruler”. In this paper an experiment using this measurement technique on the synthetic molecule known as the Holliday junction is described. The Holliday junc-tion is a molecule that consists of four DNA strands linked in a cross-shape as seen in figure 1.2.

Figure 1.2: Holliday junction labelled with donor and acceptor dyes on the top left and right respectively and biotin used for binding to the glass surface on the bottom.

The reason the Holliday junction is used is the fact that much is known about it and it moves in quite a predictable way. Even at room tempera-ture under the right conditions the Holliday junction moves such that it is possible to measure the conformational dynamics of the molecule using FRET. The plan for future research is to use FRET to look at the confor-mational dynamics of a protein. The measurements made on the Holliday junction are a proof of concept in preparation for that experiment. Since this DNA molecule moves in a way that is relatively predictable, if these results coincide with what is know about its movements, the same method can be used on that protein.

The distance is measured using the so-called FRET efficiency (E), it can be written as in equation 1.1 in relation to the distance r.6

1.2 FRET 9

E = 1

1+ (_Rr

0)

6 (1.1)

Figure 1.3: E-r graph with on the y-axis the FRET-efficiency E, and on the x-axis the distance r in nm. The value for R0 = 3.84 nm is the numerically calculated

value and circled in the graph.7

In this equation R0 is the F ¨orster distance, defined as the distance for

which E = 50%. It can also be expressed by equation 1.2,3 but in this case the value is numerically calculated. See figure 1.3. r is the distance to be measured, the distance between the donor and acceptor dyes that are attached to the molecule(s).

R0=0.211(κ 2_Q D n4 J(λ)) 1 6 (in ˚A) (1.2)

In eq. 1.2 QD is the quantum yield of the donor without presence of the

medium, J(λ) is the overlap integral between the normalized emission

spectrum of the donor (FD) and the extinction coefficient of the acceptor

(εA). J(λ) = R∞ 0 FD(λ)εA(λ)λ4dλ R∞ 0 FD(λ)dλ (1.3) When measuring FRET two dye molecules are attached to whichever molecule or molecules you want to measure the distance between. One of those is called a donor, the other one is the acceptor. The way FRET works is as follows. A laser is shone at the objective containing the sam-ple holder and excites the donor. Depending on the distance some of that energy then transfers through to the acceptor through non-radiative en-ergy transfer. The fraction of enen-ergy transferred is known as the FRET-efficiency. Both the donor and the acceptor then fluoresce with a certain intensity. Those intensities are linked to the FRET efficiency and thereby to the distance between donor and acceptor. There are a few requirements for the donor and acceptor to be able to be used for this purpose. First of all, the emission spectrum of the donor must overlap with the absorption spectrum of the acceptor, making sure the energy can transfer from donor to acceptor. Secondly, the absorption spectrum of the acceptor shouldn’t overlap with the excitation energy used, so both absorption spectra must be sufficiently separated. Thirdly, both dyes used must have a high quan-tum yield, meaning that most absorbed energy is emitted as a photon, and does not leave the molecule as heat. A fourth condition is the orien-tation of the molecules. When the dipoles of the donor and acceptor are perpendicular with respect to one another FRET is much less efficient. A particularly good combination of donor and acceptor, and the one used for this experiment, are ATTO 488 dye as donor and ATTO 594 as acceptor. In figure 1.4 you can see the emission and absorption spectra of these dyes.8, 9

1.2 FRET 11

Figure 1.4: A: Absorption spectrum of donor dye ATTO 488. B: Emission spec-trum of donor dye ATTO 488. C: Absorption specspec-trum of acceptor dye ATTO 594. D: Emission spectrum of acceptor dye ATTO 594.

The FRET efficiency (E) can be calculated in a few different ways, two of which will be discussed. In principle this is done to assure that the calculated value is the correct one. One way is to use the fluorescence intensities. The equation is,1, 10

E = IA

ID +IA

(1.4) where IA is the intensity of the acceptor and ID is the intensity of the

donor. The percentage total intensity that comes from the acceptor is the intensity coming from FRET when properly corrected for background and leakage of the light to the ’wrong’ detector, so-called bleedthrough. If all the intensity comes from the acceptor, in other words ID =0, then E =1.

When FRET does not occur IA =0 making E=0.

The other method makes use of the lifetime of the donor dye.1, 10

E=1−τDA

τD (1.5)

Here τDAand τD are the lifetime of the donor in presence and absence

of acceptor respectively. Using the known time between the pulses of the laser and the time between incoming photons at the detectors determined with TCSPC, over several hundred or thousand pulses eventually an ex-ponential becomes visible. When an exex-ponential function is fitted to this

data the lifetime can be calculated. This lifetime can then be used to calcu-late E.

Chapter

2

Methods

Before beginning measurements on the Holliday junction, measurements were first made on the fluorescence donor, ATTO 488. This was done to make sure the method used for single-molecule immobilization on glass was correct and the confocal microscope setup worked correctly.

2.1

Single-molecule immobilization on glass

For single-molecule immobilization on glass several different techniques can be used. At first several methods using (3-Mercaptopropyl)trimethoxy-silane (MPTS), for silanization, and HEPES buffer with pH = 7, tris(2carboxyethyl)phosphine (TCEP), NeutrAvidin (NA), succinimidyl 4(p -maleimidophenyl)butyrate (SMPB) and Biotin-PEG-SH, for functionaliza-tion of the glass, were tried. Experimentally was determined another method must be used. The following description is of the method used for the Holliday junction and the final measurements on the fluorescence donor.

2.1.1

Materials

Ethanol for cleaning, methanol, acetic acid glacial and (3-aminopropyltri)-ethoxysilane (APTES, bought from Sigma-Aldrich) were used for the amino-silanization. The buffer used was a 10 mM PBS buffer with pH = 7.4 con-taining 2.7 mM KCl, 137 mM NaCl and MgCl₂ concentrations of either 0 mM or 100 mM. Streptavidin (SA), Biotin-PEG-NHS (MW = 3400 Da, Laysan Bio, inc.), methoxy-PEG-NHS (MW = 2000 Da, Laysan Bio, inc.) were the main materials used for the functionalization.

2.1.2

Cover slip silanization

The glass cover slips used had ø= 25 mm, thickness No. 1 (VWR interna-tional). They were first cleaned and silanized to prepare for the attachment of the molecule. They were sonicated, first in water and then in ethanol for 20 minutes each. After drying them with a nitrogen gas flow they were put in the UV-ozone cleaner (model 42-220, Jelight, Irvine, CA) for 30 min-utes for further cleaning and preparing for the silanization. They were subsequently immersed in a methanol solution with a gentle stirrer for 30 minutes. The methanol solution contained 5% glacial acetic acid and 1% APTES. The process was completed very carefully behind a fume hood. After washing the glass cover slips with methanol and drying them with a nitrogen gas flow, they were put in the oven at 65oC for 3 hours. Then they were washed with methanol and finally stored under vacuum for further use.

2.1.3

Single-molecule Holliday junction immobilization on

glass

On the day of the measurement the cover slips were incubated in a phos-phate buffer solution (10 mM, pH = 7.4) with 0.05 mg mL−1 biotin-PEG-NHS and 5 mg mL−1 methoxy-PEG-NHS for 1.5 hours. The biotin-PEG-NHS is a bi-functional linker that connects the APTES on the glass surface via the NHS group to the SA via the biotin. Methoxy-PEG-NHS binds with its NHS-group in the same way and is used as a competition for the biotin-PEG-NHS to create gaps between the SA molecules so the Holli-day junction molecules will not bind too close to one another. After that they were incubated in SA (100 nM) for 30 minutes. SA can bind to mul-tiple biotin, so on one end it has the biotin from the biotin-PEG-NHS and on the other side it binds to the biotin on The Holliday junction. Finally the glass cover slips were incubated for another 30 minutes in varying nanomolar concentrations of the Holliday junction. The concentrations that were examined using confocal measurements were 0.5 nM, 0.25 nM and 0.125 nM. Based on those measurements the best single-molecule im-mobilization was obtained for 0.125 nM. See figure 2.1. The immobiliza-tion using 0.125 nM was done using two different concentraimmobiliza-tions MgCl₂ in the phosphate buffer. This was done to slow down the movement of the Holliday junction to make sure FRET could be measured.11

2.2 Confocal microscope setup 15

Figure 2.1: Scheme of the sample design for single-molecule immobilization of the Holliday junction on glass. On the bottom a layer of APTES. Then a layer of both methoxy-PEG-NHS and biotin-PEG-NHS to bind the layer of SA, which in its turn binds to the biotin on the Holliday junction.

2.2

Confocal microscope setup

The measurements were made using the confocal microscope setup seen in figure 2.2. The setup contains two lasers. For excitation of the donor the 485 nm laser (LDH-P-C-485, PicoQuant) on the right is used. The other laser is there for alignment of the APD meant for the acceptor emission . The beam goes through an optical fiber and clean-up filter (ZET488/10X, Chroma) and after a beam splitter it’s deflected onto an oil-based objec-tive (100x, NA = 1.4, noil = 1.518, ULPSAPO super apochromat,

Olym-pus) and sample holder. There the beam excites the donor at which point FRET either does occur or not and the fluorescence signals from donor and acceptor return back through the objective, pass the beam splitter. After filtering away the laser light with a 488 notch filter (ZET488NF, Chroma) and a 500 long-pass filter (500 LP ET, Chroma) the fluorescence emission falls onto a pinhole. From there the light passes a fiber into the detector box. In the detector box there are two separate detectors, one for photons coming from the acceptor, one for the donor. The incoming photons are separated into emission from the donor and emission from the acceptor using a dichroic mirror (575dcxr, Chroma) detected by one of the APDs (both SPCM-AQRH-15, PerkinElmer) and registered using TCSPC. The program SymPhoTime, by PicoQuant was used for the data acquisition and analysis.

Figure 2.2: This is the confocal microscope setup used for the FRET measure-ments. Two lasers were used, one for excitation at 485 nm and one of 636 nm for alignment of the acceptor detector. After going through the telescope the beam reaches a diameter of 0.9 cm and passes a 488 nm clean-up filter. After exciting the donor on the sample, the excitation goes back through the objective and passes a 488 nm notch filter and 500 nm long-pass filter to filter away any residue laser light. The beam is then focused on a pinhole from where it travels to two APDs separated by a 575 nm dichroic mirror. TCSPC then counts the photons coming to the detectors and passes that information on to the computer.

Using equation 1.4 the FRET-efficiency can be calculated. With that knowledge, the value of R0 for that specific combination of donor and

acceptor and a previous measurement of the donor without an acceptor nearby, or the known lifetime, the distance between them can be calcu-lated.

Chapter

3

Results

3.1

Data analysis

3.1.1

Scanning images of dual-labelled DNA Holliday

junc-tion

The images and time traces are obtained by integration of all photons com-ing to either detector durcom-ing one recorded time frame. No corrections are made for the background or bleedthrough, leakage from fluorescence emission into the wrong detector, since those factors are negligible.

As mentioned before, measurements were made on two different sam-ples. All the measurements were made with a small amount of PBS on the sample, either containing 100 mM of MgCl₂or 0 mM, meaning that there are four possible measurement conditions of which three were actualized. In figure 3.1 a 20 µm x 20 µm area of the glass surface obtained with a pixel size of 70 nm can be seen. This sample was prepared without the use of MgCl₂, neither during immobilization nor during measuring. The excitation energy was 5.8 µW, with an acquisition time of 10 ms.

Figure 3.1: Scanning image of size 20 µm x 20 µm from the glass cover slip prepared with 125 pM dual-labelled Holliday junction solution using no MgCl₂ during immobilization or measuring, with pixel size 70 nm and excitation energy 5.8 µW. The colour indicates the estimated lifetime and the intensity indicates the photon intensity on the detector. A: Donor fluorescence image. B: Acceptor fluorescence image. C: Superposition of A. and B. Marked with 1 and 2 are points whose time trace is found in the next section. In each image points 1 and 2 are displayed enlarged.

There are many spots on the surface that point to single molecules on the surface of the glass, but there are slightly more points visible on the donor channel pointing to FRET occurring with low efficiency only, or not at all. Several time trace measurements were made, but in only one of those there appeared to be FRET. See figures 3.4 and 3.5 in the next section for the time traces of points 1 and 2 respectively.

3.1 Data analysis 19

The second sample was prepared using MgCl₂during the immobiliza-tion as well as during the measurements. A scanning image with size 20 µm x 20 µm and pixel size 70 nm can be seen in figure 3.2. The power used was 1.7 µW with acquisition time 10 ms.

Figure 3.2: Scanning image of size 20 µm x 20 µm from the glass cover slip pre-pared with 125 pM dual-labelled Holliday junction and 100 mM MgCl₂ solution for immobilization, and measured in PBS containing 100 mM MgCl₂. With pixel size 70 nm and excitation energy 1.7 µW. The colour indicates the estimated life-time and the intensity indicates the photon intensity on the detector. A: Donor fluorescence image. B: Acceptor fluorescence image. C: Superposition of A. and B.

Using this technique leads to the DNA aggregating while still in solu-tion, making it so that barely a single molecule attach to the surface. This sample is therefore unusable when it comes to single-molecule FRET mea-surements.

The third and last sample combines the previously mentioned tech-niques, i.e. the preparation of the first, and the measurement of the sec-ond.This means that during immobilization no MgCl₂was used but when measuring a small volume of PBS containing 100 mM MgCl₂ was de-posited on the sample. This yielded better results. In figure 3.3 a 20 µm x 20 µm sized area with pixel size 70 nm can be seen obtained with power 4.3 µm and acquisition time 10 ms.

Figure 3.3: Scanning image of size 20 µm x 20 µm from the glass cover slip prepared with 125 pM dual-labelled Holliday junction for immobilization, and measured in PBS containing 100 mM MgCl₂. With pixel size 70 nm and excita-tion energy 4.3 µW. The colour indicates the estimated lifetime and the intensity indicates the photon intensity on the detector. A: Donor fluorescence image. B: Acceptor fluorescence image. C: Superposition of A. and B. The cirles marked 3 and 4 indicate points the time trace of which can be seen in 3.6 and 3.7 respec-tively. Again in each image an enlarged image around the point is displayed.

Measurements on this sample gave similar results in the aspect of in-tensity as the first sample, as can be seen in the time traces in figures 3.6, 3.7 and 3.8, where the third time trace was made from a point on a different area of the same sample.

3.1 Data analysis 21

3.1.2

Room temperature time trace fluorescence intensity

After an image is taken it’s possible to take a time trace of any point on the surface, meaning the point is illuminated by the laser and the photon count from that excitation is measured. In figure 3.4 a time trace for point 1 on figure 3.1 can be seen. The molecule is excited at 5.8 µW for 60 seconds.

Figure 3.4: Time trace of point 1 as shown by figure 3.1 when illuminated with 5.8 µW for 60 seconds. In the top right corner a zoomed in image is shown for the acceptor and donor detector respectively.

Both acceptor and detector are either emitting light at the same time or just the background is visible. This may be because FRET occurred. Af-ter 12 seconds neither emit any light, indicating that both fluorescent dyes bleached, meaning they can no longer be excited. The high acceptor emis-sion may also be caused by the acceptor being exited by the laser directly.

In figure 3.5 a time trace for point 2 as shown in figure 3.1 can be seen. Again laser power is at 5.8 µW.

Figure 3.5: Time trace of point 2 as shown by figure 3.1 when illuminated with 5.8 µW for 40 seconds. In the top a zoomed in image is shown for the acceptor and donor detector respectively.

Again when the acceptor emits a small increase in donor emission can be seen. But here you can also clearly see that the donor emits for 4 seconds in the beginning and FRET does not occur. Also around 15 seconds there is a sharp peak for both acceptor and donor, possibly an anomaly due to laser fluctuations.

3.1 Data analysis 23

When looking at the sample measured while incubated in the 100 mM MgCl₂ PBS a similar time trace can be found. In figure 3.6 the time trace of point 4 can be seen. Excitation power is 1.7 µW.

Figure 3.6: Time trace of point 3 as shown by figure 3.3 when illuminated with 1.7 µW for 50 seconds. In the top right a zoomed in image is shown for the acceptor and donor detector respectively.

Again both acceptor and donor emit simultaneously for 4 seconds, but between seconds 8 and 11 just the acceptor is emitting, albeit at much lower intensity, but the fact that the donor doesn’t emit any light - besides just background - could mean FRET occurred or that the acceptor got ex-cited by the laser so no FRET took place. When looking at the intensity it shows the acceptor emitting with the same intensity in the first burst as in figure 3.5.

In figure 3.7 the time trace for point 4 can be seen. Again with excitation power 1.7 µW.

Figure 3.7: Time trace of point 4 as shown by figure 3.3 when illuminated with 1.7 µW for 50 seconds. In the top a zoomed in image is shown for the acceptor and donor detector respectively.

Here the acceptor emission is again at the same intensity as previously, but only for a short time. After 1.5 seconds the acceptor is bleached and only a peak due to laser fluctuation can be seen at 33 seconds. The fact that there seems to be no donor emission again points to the acceptor pos-sibly being excited directly or leakage from other sources into the acceptor channel.

3.1 Data analysis 25

In figure 3.8 a time trace from the same sample as points 3 and 4 can be seen. Let’s call it point 5, illuminated with 2.9 µW. Here both acceptor and donor emit for a much longer time than before.

Figure 3.8: Time trace of point 5, similar to and on the same sample as points 3 and 4 but from a different area not shown here. The laser power is at 2.9 µW.

During the time that both emit the difference between both intensities increases slightly possibly indicating FRET occurring more efficiently as time goes on, until both bleach after 20 seconds of emitting. The fluctua-tion of the total intensity IA + ID makes it harder to determine the

FRET-efficiency, if FRET occurred.

3.1.3

Room temperature fluorescence life time

Besides looking at the intensity to determine the FRET-efficiency it’s also possible to look at the lifetime of the donor and use 1.5. When FRET occurs the lifetime of the donor decreases with increase of E. The donor dye used, ATTO 488 normally has a lifetime of τD = 4.1 ns.8 For three of the five

points, 1, 2 and 5, the lifetime of the donor came out as higher than 4.1 ns. Even though for points 3 and 4 the donor lifetime was under 4.1 ns, τD3 =

3.81 ns τD4 = 3.84. The fact the lifetimes for points 1, 2 and 5 were over 4.1

ns, makes these measurement highly unreliable and cannot be trusted to calculate the FRET efficiency correctly.

3.2

Determining FRET efficiency based on

fluo-rescence intensity and donor lifetime

When using the lifetime to determine E, doing so based on points 1, 2 and 5 makes no sense since then a negative value for E would come out. For points 3 and 4 E can be calculated. These values for E hold no importance though, because of the unreliability of the lifetime measurements.

Eτ 3 =1− 3.81 4.1 =0.07=7% Eτ 4 =1− 3.84 4.1 =0.06=6%

Another problem is that just one value for E gives us nothing in terms of the actual motion of the molecule. Only when different values are reached can you begin to describe the movement of the molecule. The fact that both of these efficiencies are very close together is very good, but also obvious since the lifetimes were almost identical as well.

When looking at the intensities more can be said about measurements made for points 1, 2 and 5. In figure 3.9 you can see FRET-efficiency (E) over time for points 1 and 2 calculated using equation 1.4.

For point 1 there are roughly three distinct phases with different val-ues for E. What becomes immediately obvious is that either both donor and acceptor emit simultaneously, or neither emit anything above their respective background. E_1.1I =0.65 then E_1.2I =0.8 and finally. E_1.3I =0.7

Also, it can be seen that there is a lot more spread in E when neither emit, this is no problem though because no FRET occurs there, so the value of E has no meaning at that time.

3.2 Determining FRET efficiency based on fluorescence intensity and donor lifetime 27

Figure 3.9: E-t diagrams for points 1 (left) and 2 (right). Top row: time traces also found in figures 3.4 and 3.5. Bottom row: E calculated using equation 1.4 and the time traces.

For Point 2 the same problem arises when neither donor nor acceptor emits (for points 3, 4 and the same effect can be seen), the spread is even bigger than for point 1. Here are two even more distinct phases.

E_2.1I =0.2

When only the donor emits above the background and the acceptor does not. In the second phase

E_2.2I =0.90

In figure 3.10 the E over time diagrams can be seen for point 3, 4 and 5 with their respective time traces for reference.

Figure 3.10: E-t diagrams for points 3 (left), 4 (middle) and 5 (right). Top row: time traces also found in figures 3.6, 3.7, and 3.8. Bottom row: E calculated using equation 1.4 and the time traces.

For point 3 two phases can be distinguished. E_3.1I =E_3.2I =0.8

Even though the acceptor intensity in the second part is much lower, the value of E remains the same. For point 4,

E₄I =0.9 and for point 5 there are two phases with E_5.1I =0.6 and

E_5.2I =0.7 respectively.

Chapter

4

Discussion and conclusion

4.1

Discussion

In figure 1.4 it can be seen that when using a 575 dichroic mirror there is a portion of both donor and acceptor emission going into the wrong de-tector, so-called crosstalk or bleedthrough. This is a possible cause of why the acceptor signal is so much higher than the donor signal. A substantial portion of the donor emission passes the dichroic mirror increasing the acceptor signal while decreasing the donor signal.

All measurements were done on just two samples. So when the sample prepared without MgCl₂was measured under the condition of MgCl₂ in the PBS, that sample had already been measured under the other condi-tion. This sample was made three days previous at the time of measuring, which only increased the probability the sample had been contaminated, and decreased the accuracy of the results.

For future research it would be best to decrease the concentration of Holliday junction on the sample, and not use the PBS containing MgCl₂ during immobilization. PBS containing MgCl₂can be used while measur-ing, for this seems to have a positive effect on the lifetime measurements, while mostly leaving the intensity unchanged. Measuring on the same day as the sample preparation would most likely also benefit the accuracy of the results. To decrease the background and leakage, an extra band-pass filter of 525 nm can be put before the donor detector to separate the donor emission. A long-pass filter with slightly higher cut-on wavelength than the dichroic mirror can be used to eliminate the donor emission from the acceptor detector. This will also slightly decrease the acceptor signal, but is still beneficial since the relative contamination becomes much smaller. Measuring each point using the same intensity would also make it much

easier to compare the time traces of each point, which now is not the case.

4.2

Conclusion

The lifetime measurements were affected the most by the different condi-tions. They remain unreliable for both samples, but nonetheless, use of MgCl₂made it at least possible to measure the lifetime decreasing. Given the unreliability of the lifetime measurements, it comes as no surprise that for both points 3 and 4 the values for E found using the lifetime of the donor were vastly different from the ones found using the intensities of donor and acceptor. The fact that all values of EI lie around E = 80% (when the acceptor intensity was higher than the donor intensity) indicates that the real value of E for one of the possible states possibly lies in that re-gion as well. This would mean that, using equation 1.1 and R0= 3.84 nm,7

r = 3 nm when the arms of the Holliday junction are close together. But since there was no one measurement made under perfect conditions no value can be derived from that statement. When zooming in even more on areas of high intensity it can be seen even more clearly that acceptor and donor rise and fall almost exclusively together. When FRET occurs you would expect to see the donor decrease in intensity when the accep-tor increases. This is not the case and leads to the conclusion that FRET did not occur. The question then remains why the intensity of the acceptor de-tector is so much higher than the intensity found in the donor dede-tector. When all corrections mentioned before are made to the setup and sam-ple measurement it should lead to correct results and may finally lead to the measuring of FRET, and eventually to learning more about molecular conformational dynamics.

Bibliography

1_{B. Valeur, Molecular Fluorescence: Methods and Applications, vol. 8. 2001.} 2_{S. Kawato, K. Kinosita, and A. Ikegami, “Dynamic Structure of Lipid}

Bi-layers Studied by Nanosecond Fluorescence Techniques,” Biochemistry, vol. 16, no. 11, pp. 2319–2324, 1977.

3_{D. K. Sasmal, L. E. Pulido, S. Kasal, and J. Huang, “Single-molecule}

flu-orescence resonance energy transfer in molecular biology,” Nanoscale, vol. 8, no. 48, pp. 19928–19944, 2016.

4_{T. F ¨orster, “Modern Quantum Chemistry,” in Modern Quantum}

Chem-istry (O. Sinanoglu, ed.), pp. 93–137, New York, USA: Academic Press, 1965.

5_{L. Stryer and R. P. Haugland, “Energy transfer: a spectroscopic ruler,”}

Proceedings of the National Academy of Sciences, vol. 58, no. 2, pp. 719–726, 1967.

6_{T. F ¨orster, “Energiewanderung und Fluoreszenz,” Die}

Naturwis-senschaften, vol. 33, no. 6, pp. 166–175, 1946.

7_{Z. Eskandari Alughare, Private communication. Leiden University, 2019.} 8_{ATTO-TEC, “ATTO 488.” https://www.atto-tec.com/product_info.}

php?info=p99_atto-488.html, 2018. Retrieved on 2019-06-25.

9_{ATTO-TEC, “ATTO 594.” https://www.atto-tec.com/product_info.}

php?info=p106_atto-594.html, 2018. Retrieved on 2019-06-25.

10_{J. R. Lakowicz, Principles of Fluorescence Spectroscopy. New York, USA:}

11_{A. Gietl, P. Holzmeister, D. Grohmann, and P. Tinnefeld, “DNA origami}

as biocompatible surface to match single-molecule and ensemble exper-iments,” Nucleic Acids Research, vol. 40, no. 14, pp. 1–10, 2012.