Opponent: Prof. Jérôme Wenger,
Supervisor: Prof. Jerker Widengren, ; Dr. Haichun Liu
Various approaches are used in fluorescence-based biological research to enhance the signal-to-background ratio (SBR). One general approach is to move to NIR wavelengths for excitation, where the background from cellular autofluorescence (AF) is minimal. The use of NIR fluorophores offers several additional advantages, including higher penetration depth and lower phototoxicity from lower photon energy. However, the low sensitivity of standard detectors in the NIR range is a major factor hindering the widespread use of NIR fluorophores. The low quantum yield and shorter lifetime of NIR fluorophores further exacerbate this challenge. Some detectors, like superconducting nanowire single-photon detectors (SNSPDs), are designed for improved operation in the NIR range. However, they are associated with significant costs. In Paper II, a dual-channel detection possibility with a single SNSPD was demonstrated to further add to their potential. Lock-in amplification is another commonly used method for enhancing the SBR. However, lock-in detection also amplifies any unintended laser light from scattering or reflection since it is modulated at the same reference frequency. Contributions from scattered laser light can introduce a significant background in tissue imaging where emission filters often cannot easily suppress this background. In Paper III, the nonlinear properties of lanthanide-doped upconversion nanoparticles (UCNPs) were exploited as frequency mixers to generate new frequency components, which could then be filtered out by a Fast Fourier Transform of the emission signal. Moreover, additional low-frequency beating signals could be generated by modulated excitation with more than one base modulation frequency. These results open for background-free imaging based on UCNPs and using low-speed detectors, including cameras. Although the two aforementioned studies focused on the advantages of excitation and detection within the NIR spectral range, thereby aiming to remove AF, the benefits of label-free studies based on AF itself should not be overlooked. The amino acid tyrosine is one of the abundant but dim sources of AF in the human body. Its emission can also contain valuable information about the local environment, particularly via its blinking properties reflecting its photophysical state transitions. This information has remained inaccessible by use of established methods such as fluorescence correlation spectroscopy (FCS). FCS analyses fluorescence intensity fluctuations from brightly fluorescent molecules in low concentrations and requires single-molecule detection (SMD) conditions, making FCS studies of dim tyrosine-containing molecules extremely difficult. Instead, the transient state (TRAST) monitoring technique can be used, allowing the exploration of their photophysics and extraction of related environmental information. TRAST, not requiring SMD conditions, provides the flexibility to be applied to dim samples where low fluorescence intensity signals can be compensated by increasing their concentration. In Paper I, TRAST measurements on tyrosine were demonstrated and applied to follow the conformational state of the tyrosine-containing protein Calmodulin. Exploiting AF is particularly valuable because it offers a label-free option, avoiding perturbations that may follow from fluorophore labeling. Yet, labeling with external fluorophores can also add further information to fluctuation-based methods, such as TRAST and FCS. In Paper IV, studying the co-enzyme Q10 on fluorescein-labeled unilamellar vesicles allowed direct observation of proton exchange kinetics and of proton collecting antenna (PCA) effects of Q10 in the vesicle membrane. In this case, although Q10 is itself dimly AF, PCA effects were better observed through changes in protonation relaxation of the pH-sensitive fluorescein fluorophore label upon variation of the Q10 concentrations in the vesicle membranes.