Dengue Virus Capsid-Protein Dynamics in Live Infected Cells Studied by Pair Correlation Analysis

Abstract

It has become increasingly evident that unveiling the mechanisms of virus entry, assembly, and virion release is fundamental for identifying means for preventing viral spread and controlling viral disease. Due to virus mobility and structural and/or functional heterogeneity among viral particles, high spatiotemporal resolution single-virus/single-particle techniques are required to capture the behavior of viral particles inside infected cells.

In this chapter, we present fluorescence imaging analysis methods for studying the mobility of fluorescently labeled dengue virus (DENV) proteins in live infected cells. Some of the most recent Fluorescence Fluctuation Spectroscopy (FFS) methods will be presented and, in particular, the pair Correlation Functions (pCF) approach will be discussed. The pCF method does not require individual molecule isolation, as in a particle-tracking experiment, to capture single viral protein behavior. In this regard, image acquisition is followed by the spatiotemporal cross-correlation function at increasing time delays, yielding a quantitative view of single-particle mobility in intact live infected cells.

We provide a general overview and a practical guidance for the implementation of advanced FFS techniques, and the pair Correlation Functions analysis, as quantitative tools to reveal insights into previously unreported DENV mechanisms. We expect this protocol report will serve as an incentive for further applying correlation imaging studies in virology research.

Appendix

1.1 The (Very) Basics of Fluorescence Fluctuation Spectroscopy Techniques: History and New Challenges

The pCF approach described in this chapter belongs to a broader family of fluctuation-based methods known as Fluorescence Fluctuation Spectroscopy (FFS) and called Fluorescence Correlation Spectroscopy (FCS). FFS has grown in the last decades because of its multiple applications to study molecular dynamics in a wide variety of biological, chemical, and physical systems [24, 32, 33]. The original version of FFS , namely the point FFS approach, was introduced for the first time in 1972 by Magde, Elson, and Webb [34]. In their work, Magde and coworkers focused a laser beam on a solution containing a complex formed by both a drug and DNA. They demonstrated that, by analyzing the fluctuations of the collected fluorescence intensity originated by the variation in the emission quantum yield of the fluorophores when bound to DNA, it was possible to determine the concentration of each component. Later, in 1974, the same group described the capabilities of the method to measure the diffusion coefficient and kinetic constants of fluorescent molecules in solution [35]. These pioneering experiments triggered a variety of new techniques based on the intensity fluctuation principle. One of the main advantages of these approaches is that it is not required to externally perturb the system. Instead, the FFS uses the spontaneous fluctuations at equilibrium of the system under study, making them suitable for the study of intracellular dynamics in live organisms.

Several excellent reviews have been written describing the theory of FFS-based techniques and the reader is directed to these sources for an in-depth theoretical discussion [24, 28, 29]. Summarized below are the basic principles and underlying statistical analyses used to extract molecular dynamics values from fluorescence fluctuation measurements and correlation analysis. Briefly, FFS techniques have proved to be powerful tools to determine, with extreme sensitivity, molecular diffusion properties, local concentration, molecular interactions, and conformational changes down to the level of single molecules [24, 25, 27].

The simplest FCS approach consists of illuminating a sample with a focused laser beam in a fixed position of the sample and measuring the collected fluorescence as a function of time (Fig. 6). Intensity fluctuations can originate from molecules moving in and out of the observation volume of a confocal or two-photon microscope, or from intensity fluctuations due to photophysical processes such as blinking, changes in orientation, or conformational transitions. In any case, as originally described by [34, 35], molecular dynamics can be obtained by computing the time auto-correlation function (ACF) or G(τ) of the fluorescence intensity as:

$$ G\left(\tau \right)=\frac{<F(t)F\left(t+\tau \right)>}{<F(t){>}^2}-1 $$

Fig. 6

Schematic representation of a single-point FCS experiment. (a) Molecules diffusing in and out of the observation volume. (b) Intensity traces over time. Molecular diffusion produces fluctuations on the detected fluorescence intensity. Aggregates (schematized as big red circles) diffuse slower than monomers (schematized as small green circles), so the duration of red fluctuations are longer in time. Increasing the molecular concentration (blue circles) the fluctuations have smaller amplitudes. (c) The temporal auto-correlation function (also called ACF) computed over different time delays, τ. This case is equivalent to the pCF at zero distance, pCF(0), because no scanning is performed. The amplitude is proportional to γ/<N> where <N> is the average number of molecules in the observation volume and γ a geometrical factor which depends on the 3D shape of the microscope observation volume

where F(t) is the fluorescence intensity collected at time t, and τ is the time lag. The brackets <> indicate sum over all measured time points. In the mathematical framework of pCF, the ACF is the analogous of the pCF(0), the pair correlation function computed at zero distance. In the case of intensity variations due to free pure diffusion, the average duration of the fluctuations is determined by the mean transit time of molecules to traverse the observation volume. The transit time (or correlation time τ_c) depends on the diffusion coefficient D and by the lateral waist size of the observation volume ω₀ by τ_c = ω₀²/nD, where n is a geometrical factor equals to 4 for confocal or 8 for two-photon systems [24, 36]. The average amplitude of the fluctuations depends inversely on the average number of molecules in the observation volume [24].

The shape and size of the observation volume is given by the microscope settings, the microscope objective, the excitation wavelength, and other factors. It has been shown [29, 37] that its shape is well described by a 3D Gaussian function for a confocal microscope or a 2D Gaussian-Lorentzian function for a two-photon setup. However, many chemical, physical, and biological processes occur in large spatial scales and cannot be completely studied by observing a single “point” in the cell interior. A protein, for example, can move along the whole cytoplasm, bordering membranes or confining at specific regions.

To overcome this limitation, it is possible to sequentially perform point FCS experiments at different positions of the sample (i.e., in a scanned line) computing ACF curves along the pixels of the scanned lines. This approach is suitable for having quantitative information at different regions of the sample, i.e., the diffusion coefficient at different compartments of the cell or at different positions of the same compartment. In a commercial microscope, this experiment can be simply performed by periodically moving the excitation beam in a line of N pixels over the sample while molecules diffuse. Some of these molecules will be illuminated at many points as they follow a movement like the scanned line, while other molecules will follow different paths and then will be only illuminated during a few pixels over the line. Computing the ACF at subsequent points along a line has expanded the capabilities of the point FCS method to capture the dynamic in the cell interior. However, this approach does not allow us to relate what is happening between different regions, such as molecular translocation or nucleocytoplasmic shuttling of viral proteins. To overcome this limitation, Gratton and collaborators have introduced the concept of the pair Correlation Function (pCF) whereby correlating the fluorescence intensity fluctuations at a pair of specific points in a scanned line returns the time it takes a molecule to move (diffusing or not) between the two locations. As previously noted in this chapter, the pCF approach combines the advantages of FFS and SPT methodologies returning the time needed for each molecule to be found in each point in space. If a barrier to diffusion is present, a longer time will be needed for the same molecule to be found at a position across the barrier, as already demonstrated in the case of molecular transport across the NE [16]. Thus, FFS techniques and more specifically the pCF approach are extremely appropriate to study molecular shuttling across the NE and transport across other intracellular components [17, 21]. For further and deeper discussion on the discussed techniques, readers are referred to specialized literature [15, 17,18,19, 24].

Finally, we expect this work will help to stimulate the use of advanced FFS techniques, in particular the pCF, for dissecting molecular processes for viral infection, which is a fundamental step in identifying control measures.