In this thesis I used FCS to study tracer diffusion in heterogeneous systems of synthetic nature. In the first part of my thesis using the combination of the laser scanning confocal microscopy (LSCM) and fluorescencecorrelationspectroscopy (FCS) I studied in situ the dynamics of phase separation in the polymer blend polystyrene/poly(methyl phenyl siloxane) (PS/PMPS) at the macroscopic and microscopic length scales, respectively. LSCM was used to monitor the process of phase separation in real time providing images of droplet coalescence and growth during the intermediate and late stages. Measuring the small tracer (terrylene dye) diffusion in the PMPS phase of the phase separated blend and by comparing the diffusion of the same tracers in pure bulk PMPS, additional information on the purity of phases was obtained using FCS.
primary expression levels independent from secondary cell-to-cell transmission based on bulk cellular GFP levels such as by Flow Cytometry (MFI) or also western blotting and will require more sophisticated strategies. A comprehensive analysis of CD63-GFP tagged EVs derived from transiently transfected cells, previously shown to be functionally taken up by a range of cell lines [ 34 ], confirmed largely retained composition in terms of content, morphology and physicochemical properties as compared to native vesicles. This is well in line with the fact that we detected no dramatic increase of CD63 in the EVs by MS-proteomics upon transient overexpression in the cells, resulting in an average of ~30 molecules per vesicle as determined by FCS. Due to the direct quantification of single fluor- escent molecules before and after vesicle lysis and the possibility to use the identical fluorophore for virtually every fluorescent cargo as an internal reference, we propose that the quantification by FluorescenceCorrelationSpectroscopy is a reliable method for single molecule-single vesicle quantification. This is inher- ently limited to the quantification of fluorescent mole- cules and does not account for the possible contribution of a fraction of non-fluorescent fusion proteins, such as due to GFP quenching by, e.g. acidic pH. With orthogonal methods emerging that have the potential of leveraging single vesicle quantification to the single molecule-single vesicle level [ 33 ], a systematic comparison with respect to the limitations of the different technologies will be warranted to approach absolute quantification in a near future.
Following the concept of fluorescencecorrelationspectroscopy (FCS) [Magde et al., 1972], the autocorrelation function of the number and brightness of fluorescently labeled molecules within a small observation volume is determined by the dynamics that govern the fluorescence fluctuations. In the following years, FCS developed into an increasingly versatile tool to study not only molecular mobility [Aragón and Pecora, 1976,Koppel et al., 1976,Fahey et al., 1977], but also bi-molecular interactions [Kinjo and Rigler, 1995,Schwille et al., 1996, Rauer et al., 1996, Schwille et al., 1997]. Thus, kinetic parameters governing biochemical reactions can be extracted from the autocorrelation function. Focused laser excitation and tightly confined confocal detection allowed the study of individual molecules with FCS [Rigler et al., 1993, Eigen and Rigler, 1994]. More advanced FCS variants that scan the confocal detection in the sample [Berland et al., 1996] proved useful to elucidate not only mobility in the membrane [Benda et al., 2003, Ries and Schwille, 2006, Petrášek and Schwille, 2008], but also the partitioning into different membrane phases [Ries et al., 2009a] and the binding to membrane-localized receptors [Ries et al., 2009b]. Simultaneous to the advent of wide-field total internal reflection fluorescence (TIRF) microscopy [Ax- elrod, 1981], the pioneering works of Thompson and colleagues highlighted the potential of total internal reflection fluorescencecorrelationspectroscopy (TIR-FCS) to investigate surface binding [Thompson et al., 1981, Thompson, 1982, Thompson and Axelrod, 1983]. Studies on lipid mobility of the basal membrane of cells and of supported lipid bilay- ers (SLBs) [Tamm and McConnell, 1985] profited from the surface-selectivity of TIRF microscopy in combination with FCS [Thompson et al., 1993, Ohsugi et al., 2006]. Mod- ern electron-multiplying charge-coupled device (EMCCD) camera technology enabled the multiplexing of FCS acquisition in multiple detection modalities [Krieger et al., 2015], including TIRF microscopy [Burkhardt and Schwille, 2006, Kannan et al., 2006, Kannan et al., 2007, Sankaran et al., 2009]. Image correlationspectroscopy (ICS) [Petersen et al., 1993,Wiseman, 2013], a closely related method, extracts molecular mobilities and has been used to investigate the binding of membrane-diffusing receptors to larger protein complexes as well as of ligands to microtubules [Brandão et al., 2014].
Since a large use of hydrogels is being made in biological environments, however, deeper knowledge of the system’s behavior in water is needed. In addition to the basis of a structure theory, the systematic engineering of hydrogels for sensing applications also has the potential to hugely benefit from a systematic understanding of the structure, dynamics and duration limits of the gels themselves. To this end, the present fluorescencecorrelationspectroscopy (FCS) study revolves around PNIPAAm anchored hydrogels, with multiple goals. These include primarily probing the gels’ behavior in aqueous environments at different temperatures, encompassing the thermal-dependence, and providing a more coherent picture of the structure-dynamics relationship at the nanometer scale. Hence, the present work focus on the analysis of the swelling ratio R S of and tracer diffusion in PNIPAAm hydrogels under different physicochemical conditions. These two quantities, which are obtained from the same experimental technique under equivalent conditions, represent space averaged quantities in the micrometer and submicrometer scales.
Mario Schneider 1 , Stefan Walta 2 , Chris Cadek 1 , Walter Richtering 2 & Dieter Willbold 1,3 The amyloid-beta peptide (Aβ) plays a major role in the progression of Alzheimer’s disease. Due to its high toxicity, the 42 amino acid long isoform Aβ42 has become of considerable interest. The Aβ42 monomer is prone to aggregation down to the nanomolar range which makes conventional structural methods such as NMR or X-ray crystallography infeasible. Conformational information, however, will be helpful to understand the different aggregation pathways reported in the literature and will allow to identify potential conditions that favour aggregation-incompetent conformations. In this study, we applied fluorescencecorrelationspectroscopy (FCS) to investigate the unfolding of Alexa Fluor 488 labelled monomeric Aβ42 using guanidine hydrochloride as a denaturant. We show that our Aβ42 pre-treatment and the low-nanomolar concentrations, typically used for FCS measurements, strongly favour the presence of monomers. Our results reveal that there is an unfolding/folding behaviour of monomeric Aβ42. The existence of a cooperative unfolding curve suggests the presence of structural elements with a Gibbs free energy of unfolding of about 2.8 kcal/mol.
This work focusses on protein-membrane interactions within the coagulation cascade, particularly, the binding of FVIII to membranes. Investigations on PS-vesicles as model systems for platelets showed that binding of inactivated FVIII to membranes is strongly associated with electrostatic interactions and therefore increases with the concentration of charge (i.e. phosphatidylserine) in the membrane . For activated FVIII however, binding and particularly, its dependence on the PS-content, has not yet been investi- gated. Likewise, there is not yet quantitative support for regulatory processes such as the scenario of competitive binding of FVIII and annexin as depicted in the coverfigure. FluorescenceCorrelationSpectroscopy is a powerful tool for the investigation of protein-membrane interactions. It measures diffusion times and can discriminate freely diffusing proteins from proteins bound to vesicles as model membranes based on the size-dependence of the diffusion time [7, 8]. Hence, binding isotherms can be obtained from consecutive titration experiments. Compared to other conventional techniques used for binding experiments, such as ELISA, which all rely on one component bound to a surface, an advantage of FluorescenceCorrelationSpectroscopy is that it allows for mea- surements in solution. However, if desired, experiments can be performed on supported lipid bilayers as described in chapter 12. In this thesis protein-membrane binding within coagulation was investigated with FCS using both, experiments in solution as well as on supported membranes.
Thermosensitive liposomes (TSLs) whose phase-transition temperature (Tm) lies slightly above body temperature are ideal candidates for controlled drug release via local hyperthermia. Recent studies, however, have revealed disruptive shifts in the transition temperature in mouse plasma, which are attributed to undefined interactions with blood proteins. Here, we study the effects of four major plasma proteins – serum albumin (SA), transferrin (Tf), apolipoprotein A1 (ApoA1) and fibrinogen (Fib) – on the temperature- dependent release of fluorescein di-β-D-galactopyranoside (FDG) from TSLs. The amount of fluorescein released was quantified by fluorescencecorrelationspectroscopy (FCS) after hydrolysis of FDG with β-galactosidase (-Gal). This approach is more sensitive and thus superior to previous release assays, as it is impervious to the confounding effects of Triton on conventional fluorescence measurements. The assay determines the molar release ratio, i.e. the number of molecules released per liposome. We show that shifts in the Tm of release do not reflect protein affinities for the liposomes derived from adsorption isotherms. We confirm a remarkable shift in induced release towards lower temperatures in the presence of mouse plasma. In contrast, exposure to rat or human plasma, or fetal bovine serum (FBS), has no effect on the release profile.
The structural properties of halloysite/biopolymer aqueous mixtures were firstly investigated by means of combining diﬀerent techniques, including small-angle neutron scattering (SANS), electric birefringence (EBR) and fluorescencecorrelationspectroscopy (FCS). Among the biopolymers, non-ionic hydroxypropylcellulose and polyelectrolytes (anionic alginate and cationic chitosan) were selected. On this basis, the specific supramolecular interactions were correlated to the structural behavior of the halloysite/biopolymer mixtures. SANS data were analyzed in order to investigate the influence of the biopolymer adsorption on the halloysite gyration radius. In addition, a morphological description of the biopolymer-coated halloysite nanotubes (HNTs) was obtained by the simulation of SANS curves. EBR experiments evidenced that the orientation dynamics of the nanotubes in the electric field is influenced by the specific interactions with the polymers. Namely, both variations of the polymer charge and/or wrapping mechanisms strongly aﬀected the HNT alignment process and, consequently, the rotational mobility of the nanotubes. FCS measurements with fluorescently labeled biopolymers allowed us to study the aqueous dynamic behavior of ionic biopolymers after their adsorption onto the HNT surfaces. The combination of EBR and FCS results revealed that the adsorption process reduces the mobility in water of both components. These eﬀects are strongly enhanced by HNT/polyelectrolyte electrostatic interactions and wrapping processes occurring in the halloysite/chitosan mixture. The attained findings can be useful for designing halloysite/polymer hybrids with controlled structural properties.
In the recent years, the fluorescencecorrelationspectroscopy has emerged as a powerful tool for investigation of the diffusion of fluorescent molecules, macromolecules or nanoparticles in various environments. FCS is a single molecule spectroscopic technique able to address extremely low concentrations (nano molar) of fluorescence species. Furthermore the method offers an extremely small observation volume (<1µm 3 ) that provides a great potential to locally access the systems which require high spatial resolution. Not surprisingly therefore, during the last decade, FCS has become one of the best tools to evaluate the diffusion and transport properties in complex systems. Mainly for traditional reasons, however, so far the utilization of the FCS has been limited mostly to biological studies, i.e. aqueous environments. [1, 10, 48, 49, 55-58] Only recently FCS was successfully applied to study the size and conformation of macromolecules in organic solvents [4, 9] , adsorbed polymers, [14-16] grafted gel,  diffusion in polymer solutions, [3, 5, 7, 11, 13] and thin polymer films.  Furthermore it was shown that FCS can follows the process of radical polymerization of styrene over an extensive conversion range  and address the swelling of crosslinked polymer microbeads in organic solvents. 
The two amino acid substitutions F64L and S65T are responsible for convenient use of EGFP in biological applications. The S65T and F64L mutations result a 35 fold increase in fluorescence over wt-GFP when excited at 488 nm and this proves that the second mutation is critical for the maximal fluorescence (Cormack et al., 1996). EYFP contains substitutions such as S65G, V68L, S72A and T203Y and is the most red shifted avGFP. An improved version of YFP, Venus have been presented recently (Nagai et al., 2000). As a result of well known folding mutations, F64L/M153T/V163A/S175G, the folding of Venus is dramatically increased. Venus also contains a novel mutation F46L which greatly facilitated the maturation of YFP by accelerating the oxidation step of the chromophore formation. Oxidation of chromophore is considered to be a rate limiting step of maturation. Beyond that, the protein became more tolerant to acid-sensitivity and quenching by chloride ion (Nagai et al., 2000). In this aspect, we focus on F64L mutation which is an important constituent of EGFP that gives maximal fluorescence. So we introduced super-EYFP (SEYFP) in fluorescencecorrelationspectroscopy measurements together with EYFP and EGFP. SEYFP was created with the help of multiple site mutagenesis by the introduction of four folding mutations simultaneously into EYFP (EYFP - F64L/M153T/V163A/S175G) which facilitates the folding of YFP at 37 0 C (Nagai et al., 2000).
Figure 15: Normalized cross-correlation functions with lag time for GS-II (circles) and GS-II upon addition of PHEMA-b-PGlcNAcEMA (triangles). Two-component fits are indicated by solid curves. The dashed lines display the particular inflection points.
problems can be neglected using 2fFCS as an improved version of the standard confocal FCS setup. 27,29 In this work, the protein binding was analyzed using FITC-labeled GS-II at a molar concentration of 17 nmol L −1 . The unlabeled block glycopolymer was used at a molar concentration of 0.25 mmol L −1 . Complete diagrams with auto and cross-correlation functions and their fits, for GS-II and PHEMA-b-PGlcNAcEMA respectively, are presented in the SI (Figures 25 and 26). CCFs were fitted using a two-component fit model (Figure 15, solid curves), which accounts for the presence of residual FITC (not chemically attached to GS-II). Indeed, the GS-II stock solution was purified by SEC (Figure 23 in the SI), but a certain fraction of 30% of free dye still remained. Figure 15 shows the normalized and averaged CCFs plotted against the lag time of freely diffusing GS-II (circles) and GS-II upon addition of PHEMA-b-PGlcNAcEMA (triangles). The dashed lines in Figure 15 indicate that the inflection points of the CCFs are significantly different. This corresponds to a change in the diffusion time meaning that, after addition of PHEMA-b-PGlcNAcEMA, GS-II was bound to the diblock glycopolymer. The hydrodynamic radii obtained from the fits are in good agreement with DLS and TEM analysis: 6.8 ± 0.1 nm for GS-II and 15.4 ± 0.2 nm for GS-II upon addition of PHEMA-b-PGlcNAcEMA. The slightly larger radius provided by TEM may be caused by flattening of the soft micelles on the target. We conclude that little or no unbound GS-II remained in solution after addition of diblock glycopolymer, which hints to strong multivalent glycan-mediated protein adsorption hence.
was used (filtered through a MilliQ purification system, resistivity 18.2 MΩ∙cm) without any buffers. However, only for the study of antibody diffusion (4 th Chapter, section III) acetate buffer was used to swell the gels. The temperature-dependent swelling ratio, ( ), was determined as the ratio between the fully swollen thickness and the dry thickness. The dry thickness has been measured by a step profiler (KLA-Tencor Stylus P-16+ profilometer) as mentioned previously, while the thickness of the fully swollen hydrogels was determined using the FCS setup. By shifting the microscope objective, the position of the FCS observation volume was scanned in z-direction (normal to the film plane) with a step of 1 μm and the average fluorescence intensity signal that is proportional to the local density of the tracers was recorded as a function of the focus position (z-scan). A typical z-scan (Fig.3.1) depicts 2 transition regions, representing the hydrogel/water and glass/hydrogel interfaces. The distance between these regions represents the thickness of the fully swollen gel. The monomer concentrations (monomer volume fractions, ( ) ( )) for the studied hydrogels (HG) are shown in Tables 4.1 and 4.2(4 th
detection. These distortions become more pronounced with increasing distance of the laser focus from the cover glass surface. As a result, a larger detection volume and thus an apparently smaller diffusion coefficient are obtained in conventional FCS. As discussed above, the diffusion coefficients determined in 2fFCS should not be biased by the refractive- index mismatch. However, in many cases, one has only to work in aqueous buffer solutions with a slightly different refractive index. In this study, the microgel suspensions became partly turbid, viscous, and jelly-like after compression, indicating that the refractive index is considerably larger than in water. Therefore, the influence of the measurement position on the diffusion coefficient had to be investigated first to be sure that our 2fFCS setup is still not affected by optical aberrations. For this purpose, experiments were conducted on dense-packed homogeneous microgel matrixes (100:0 and 0:100 of soft:stiff microgels) at depths of 20, 40, and 60 µm apart from the cover glass surface. These measurements show that the tracer diffusivity in homogeneous packings of solely soft and solely stiff microgels does not depend on the measurement depth, as depicted in Figure 73. In addition, the standard deviations of manifold repeated measurements are just between ± 2% and ± 5% in most cases. From these two limiting cases, it is reasonable to conclude that the diffusion coefficient is also independent of the measurement depth in our inhomogeneous microgel packings. For this reason, all further measurements of this work were performed at a distance of 20 µm from the cover glass surface, which assures to get a high fluorescence signal due to just little scattering on the one hand and to be far enough apart from the cover glass surface on the other hand.
a) In a first step it was necessary to invent an extended theoretical model, which took into account size and labelling effects in 2fFCS data analysis, because impact of non-negligible particle size with different fluorescent labelling could not be neglected in FCS and 2fFCS measurements. Various labelling distributions and their influence on correlation func- tions were modelled and discussed. Performed DLS and 2fFCS measurements on large uniformly labelled latex beads of different size gave excellent agreement between DLS and 2fFCS measurements when the extended model was applied to 2fFCS data analysis. The experimentally determined values for the hydrodynamic radius as calculated with the ex- tended model were in perfect agreement with the specifications of the bead manufacturer as well as with the results from the DLS measurements. The proposed models for differ- ent labelling geometries allow utilization of 2fFCS (and also standard FCS) as a powerful complementary method to investigate hydrodynamic radii of extended diffusing objects at very low concentrations in colloid and polymer science, e.g. Ref. [98–105] . Nevertheless,
The membrane proximal external region (MPER) studied in this work corresponds to amino acids 662 to 673 of the envelope glycoprotein gp41 of HIV-1, modified with a cysteine, which is targeted for fluorescent labeling by ATTO488 with a maleimide moiety. This peptide forms an α-helix and inserts into the membrane under an angle, due to its distribution of hydrophobic amino acids [Sun et al., 2008]. The MPER is part of the viral envelope complex, which plays a major role in the virus entry into the host [Chan and Kim, 1998, Castagna et al., 2005, Montero et al., 2008]. As such, gp41 as a whole, but also MPER itself, have become important targets for HIV-1 drugs [Gardner and Farzan, 2017, Kelsoe and Haynes, 2017, Montero et al., 2008]. The MPER peptide does not only have a strong medical significance, but is also interesting for establishing FCS as a tool to study the interaction of biomolecules with lipid monolayers. Fluorescence imaging and FCS show no evidence for aggregation of MPER at the lipid monolayer, supposedly because this short peptide does not exhibit a higher order structure which may unfold when it encounters the hydrophobic interface. Moreover, MPER is known to insert its α-helix into the membrane [Sun et al., 2008], which at least in theory makes its interaction conceptually different from the interaction of CtxB with lipid monolayers.
DLS was found not to be suitable to study coalescence in our systems because the estimated diameters of the nanodroplets were influenced by the diluent and the dilution. Zeta potential and FCS measurements, respectively, showed that there is neither a change in the surface coverage of the nanodroplets nor a change in their concentration and fluorescence brightness during the solvent evaporation process. This points to an absence of coalescence. We used DC FCCS on a mixture of two polymeric emulsions, each of them containing polymer labeled with a different dye. This allowed quantification of the magnitude of coalescence, which was found to be insignificant. It cannot account for the observed particle size distribution. The DC FCCS measurements are supported by FRET measurements on labeled particles, which also show nearly no coalescence. The combination of techniques used in this study shows that the typical large size distribution of nanoparticles prepared by emulsion-solvent evaporation is very likely due to the process itself producing droplets with large size distribution. Thus, further efforts towards the fabrication of monodisperse nanoparticles by the emulsion-solvent evaporation method have to be continued in this direction.
Due to significant improvement in the field of optics during the last decades micro- scopes are nowadays able to detect photons emitted by single molecules. Such sensitive microscopes opened the way for fluorescencecorrelationspectroscopy (FCS). The basic principle of FCS is the excitation of fluorescing molecules with a laser beam followed by the detection of the emitted light. To realize the detection of only a few molecules at a time the excitation volume is reduced by focusing the laser beam and using the same optical path for excitation and emission light. Due to the Stokes shift the collected light has a higher wave length and can be separated by a dichroic filter and further be detected by highly sensitive avalanche photo diodes (APDs). With this setup nanomolar solutions of substances to be analyzed are measured for a period of time. Either by software or using a hardware correlator, data is analyzed such that any measured fluorescence is compared to any data point with a temporal pitch, yielding a distribution of likelihood to see fluorescence at a given time after a former event.
Fig. S.5.5 2fFCS measurement on 10-kDa dextran labeled with Alexa Fluor 647 in a PAAM hydrogel matrix with cross-linker to monomer ratio of 1:60 and a PAAM concentration of 50 gL ʹ1 at 25 °C. A single particle model including triplet state relaxation is used to fit the data (smooth lines). Fluorescence excitation was achieved with O ex = 470nm at an excitation power of 2 µW. The detection signal was filtered by a HC 687/70
Figure 4.2 Confocal fluorescence microscopy images of a core–shell microgel particle as sketched in
Scheme 1. The images are taken at the equator plan of the core of a core–shell particle lying on the lower cover slide of the sample cell. a) – d): particle in pure water at different temperatures; the core diameter of about 60 m is retained upon heating and cooling. e): particle in water/MeOH mixture with 40 mol% of MeOH. Adding MeOH lead to deswelling of the core and the shell, but the shell is more swollen than the core. The shell thickness influences the deswelling of the core and leads to a deformation of the core in regions where the shell is thin (lower-left part of the particle); vice versa, this deformation also deforms the shell in these regions. No effect of adhesion between particle and cover slide on the deswelling of the core is visible. The scale bar in Panel e) equally applies to all panels.
with the absence of contact between MutS and the donor dye when the protein binds the mismatch with Phe-36 stacked on the T base in the lower strand ( Figure 1 C, Model II). The high r D population will be addressed further below. We were interested in determining if the unkinked G A :T D DNA is free or MutS-bound, and thus measured the DNA diffusion times by subensemble FluorescenceCorrelationSpectroscopy (seFCS) of bursts selected for this FRET species from the data trace (see Supplementary Data Section 2.4 ). The obtained apparent diffusion times are displayed in the bar diagrams of Figure 4 A and B (left panel) for G A :T D DNA without and with MutS, respectively. The apparent diffusion times of both FRET species are clearly longer in the presence of MutS, which reveals that also unkinked DNA is in complex with MutS ( Supplementary Figure S1 ). This population corresponds to either a non-speciﬁc DNA–MutS complex (MutS bound to any position on DNA) or to a speciﬁc unkinked DNA–MutS complex (MutS bound at the mismatch site but without inducing a kink). The latter species has been proposed earlier, based on AFM and FRET analysis of immobilized DNA ( 36 , 37 ); however, it remains to be shown if MutS is in direct contact with the mismatch in the unkinked complex, and whether this complex is a bona ﬁde intermediate in the mismatch rec- ognition pathway.