Читать книгу Diatom Microscopy - Группа авторов - Страница 20

Super-resolution optical microscopy is based on the photophysical and photochemical properties of specific fluorophores. Super-resolution optical microscopy can be divided into two main categories: 1) Methods where the nonlinear responses of fluorophores are used to shrink the point spread function in order to enhance image resolution. These methods include stimulated emission depletion (STED) microscopy [1.26], ground state depletion (GSD) microscopy [1.25], and reversible saturable optical fluorescence transitions (RESOLFT) microscopy [1.24]; 2) Methods where the stochastic emission of single fluorophores is used to register the location of each emitter at distinct times, thereby making it possible to distinguish two nearby fluorophores in time, These methods include photo-activated localization microscopy (PALM) [1.2], stochastic optical reconstruction microscopy (STORM) [1.58], and super-resolution optical fluctuation imaging (SOFI) [1.13].

Due to the nature of light diffraction, the image resolution of confocal microscopy is limited to roughly 250 nm, which makes it difficult to visualize silica-embedded proteins (ranging from ten to several hundred nanometers) via live-cell imaging. Super-resolution optical microscopy based on photoswitchable fluorescent proteins makes it possible to determine with a high degree of precision the location of proteins at the single-molecule level. Fusion proteins with silaffin-3 (tpSil3) are embedded in biosilica and permanently entrapped inside the valve region during biosilica formation [1.50]. PALM super-resolution optical imaging involves the expression of fusion proteins with photoswitchable fluorescent proteins, which can prevent the problem of silica limiting antibody access to the protein(s) of interest in STORM. Genetic transformation facilitates live-cell imaging of proteins associated with the cell wall, thereby making it possible to use the expression of GFP fusion proteins in Thalassiosira pseudonana to localize distinct regions of biosilica. Gröger et al. [1.22] used the inherently high labeling density of PALM to study proteins in diatom biosilica.

Prior to imaging, it is necessary to select suitable super-resolution probes for biosilica embedded fusion-proteins. Six fluorescent proteins (PATagRFP, PAmCherry1, PA-GFP, mEOS3.2, Dendra2, and Dronpa) have been activated efficiently in cytosol; however, only Dendra2, mEOS3.2, and Dronpa have been activated when embedded in biosilica. Fusing three fluorescent proteins to tpSil3 for PALM imaging has made it possible to achieve average localization precision of roughly 25 nm. Images are reconstructed (stacked) from more than 1000 frames to facilitate localization. Note that the production of pure biosilica requires the extraction of chloroplasts using a detergent-based buffer to eliminate their autofluorescence, which can hinder in vivo single-molecule localization microscopy. Figure 1.18a presents a PALM image of Dendra2 in the biosilica of the valve region of Thalassiosira pseudonana. Figure 1.18d shows the focal plane in the girdle band region. As shown in Figure 1.18b, the average full width at half maximum (FWHM) of 76.0 nm ± 4.6 nm (± S.D.) was derived from multiple line scans perpendicular to the biosilica cylinder (Figure 1.18a). The 76 nm measurement can be further deconvoluted to 53 nm ± 3 nm by taking into account the localization precision and the linker length of 3 nm between tpSil3 and the chromophore of the fluorescent protein. Fourier ring correlation analysis (Figure 1.18c) was used to estimate image resolution [1.43], the results of which (74.7 nm) were in good agreement with those in Figure 1.18b. Total internal reflection fluorescence (TIRF) imaging has also been used to determine the locations of tpSil3-Dendra2, tpSil3–mEOS3.2 and tpSil3–Dronpa embedded in biosilica (Figures 1.18d,e,f; left side). This provides a representative comparison of the enhanced resolution/contrast shown in Figure 1.18d,e,f (right side). Note that tpSil3-Dendra2, tpSil3-mEOS3.2 and tpSil3–Dronpa are shown in the outer region of the fultoportulae basal chamber instead of within the external tubes of the fultoportulae. The appearance of large circular non-fluorescent gaps correlates with the structural elements of fultoportulae observed in SEM images (insets of Figures 1.18d,e,f).

Figure 1.18 PALM analysis of tpSil3. (a) Comparison of epifluorescence image and reconstructed super-resolution image of tpSil3-Dendra2 with z-focus on the girdle band area of the diatom. The position of the line scan is highlighted. (b) Line scan through the silica cell wall showing the fluorescence intensity profiles obtained using the two imaging modalities. The FWHM of the PALM image is indicated. (c) Fourier ring correlation applied to super-resolution image in A revealing an effective resolution of 74.7nm. Comparison of epifluorescence images and reconstructed super-resolution image of Dendra2 (d), mEOS3.2 (e), and Dronpa (f) fused to tpSil3 with thew z-focus on the valve region of the diatom. Enlarged details of the fultoportulae to the right. For comparison, we also present an SEM image at the same scale corresponding to the enlarged details shown in the fluorescence image. Scale bars = 1μm and 100nm in the zoomed images. From [1.22] with permission of Springer Nature.