The three-dimensional distribution of αA-crystalline in rat lenses and its possible relation to transparency

Abstract

Lens transparency depends on the accumulation of massive quantities (600-800 mg/ml) of twelve primary crystallines and two truncated crystallines in highly elongated "fiber" cells. Despite numerous studies, major unanswered questions are how this heterogeneous group of proteins becomes organized to bestow the lens with its unique optical properties and how it changes during cataract formation. Using novel methods based on conical tomography and labeling with antibody/gold conjugates, we have profiled the 3D-distribution of the αA-crystalline in rat lenses at ∼2 nm resolutions and three-dimensions. Analysis of tomograms calculated from lenses labeled with anti-αA-crystalline and gold particles (∼3 nm and ∼7 nm diameter) revealed geometric patterns shaped as lines, isosceles triangles and polyhedrons. A Gaussian distribution centered at ∼7.5 nm fitted the distances between the ∼3 nm diameter gold conjugates. A Gaussian distribution centered at ∼14 nm fitted the Euclidian distances between the smaller and the larger gold particles and another Gaussian at 21-24 nm the distances between the larger particles. Independent of their diameters, tethers of 14-17 nm in length connected files of gold particles to thin filaments or clusters to ∼15 nm diameter "beads." We used the information gathered from tomograms of labeled lenses to determine the distribution of the αA-crystalline in unlabeled lenses. We found that αA-crystalline monomers spaced ∼7 nm or αA-crystalline dimers spaced ∼15 nm center-to-center apart decorated thin filaments of the lens cytoskeleton. It thus seems likely that lost or gain of long-range order determines the 3D-structure of the fiber cell and possible also cataract formation.

Figure 1. Projected Structure.

Panel A shows a low magnification view of three cortical fiber cells from the equatorial region of a rat lens. The plasma membranes appear as electron-dense bands. In contrast, the cytoplasm appears unstructured with occasional clusters of electron-dense particles. At this magnification, the cytoskeleton assemblies referred as “beaded” filaments are not present. Panel B shows a higher magnification view of a region of plasma membrane surface to demonstrate the characteristic pentalamellar structure of gap junctions. Bar: A = 0.2 µm, B = 40 nm.

Figure 2. Flow Chart of the Experimental Protocol.

Figure 3. Distribution and Classification of Gold Conjugates.

Panel A is a Z-projection of a tomogram computed at maximum intensity and presented in reverse contrast to highlight the distribution of gold conjugates in the tomogram. Depending on their diameters and peak intensities, the conjugates were classified as yellow (>4 nm diameter OR peak intensity >200) or blue (>2 nm in diameter OR peak intensity >150). Panel B shows the gold conjugates contained in the volume. The color-coded lines represent the Euclidean distances connecting yellow (red lines), yellow-to-blue (green lines) and blue conjugates (brown). Bar A–B: 0.1 µm.

Figure 4. Measurement of the Euclidean Distances between Conjugates.

Panel A shows a small region of a tomogram with six yellow (∼7 nm diameter) and nineteen blue (∼3 nm diameter) conjugates. The color-coded lines represent the Euclidian distances between yellow (red), yellow-to-blue (green) and blue (brown) conjugates. Note that independent of their diameters, the conjugates occupy the vertices of isosceles triangles. Panels B&C show histograms of the Euclidian distances between 107 yellow and 225 blue conjugates. Panel B shows that two Gaussian distributions centered at ∼14 nm and 21–24 nm fit the Euclidian distances between yellow (red) and yellow-to-blue conjugates (green). Panel C shows that the single Gaussian distribution centered at ∼7.5 nm fits the Euclidian distances between blue conjugates. In B and C, the x-axis plots distance in nm and the y-axis percentage.

Figure 5. Tethers.

Panel A is a low magnification view of a fiber cell labeled with anti-αA-crystalline/gold complexes (white discs). At this magnification, the larger conjugates (∼7 nm diameter) appear randomly scattered in the volume and the tethers that connect them to protein assemblies are not visible. To identify these tethers, small volumes (square, 128 128 47 pixels) around a central gold conjugate were cut and segmented using the watershed transform (see Methods). Panel B is a view of the square in A. It was rotated to find out the direction that visualized these tethers. Three tethers (blue) connected the gold conjugate (green) to protein assemblies shaped as thin filaments and spherical particles (red). Panel C is a higher magnification view of the thin filament colored red in B. The arrow indicates place where the tether (blue) joins the thin filament. Bar: A = 70 nm; B = 12 nm; C = 5 nm.

Figure 6. Thin Filament and “Bead” Assemblies.

Panel A shows a thin filament decorated with three small gold conjugates (brackets) and a larger gold conjugate labels a single spherical particle (the “bead”). The inset shows a “bead” assembly labeled with four small conjugates at the vertices of a square. Panel B shows a single-pixels slice of a “bead” assembly connected to a large gold conjugate. The brackets indicate evenly spaced particles in the “bead.” Panel C shows a view of a rendered volume showing a single small gold conjugate tethered to a “bead” assembly. Panel D shows a large and a small gold conjugate tethered to a single “bead” assembly. The arrow points to the region where a tether joins the “bead.” To visualize these tethers, the image was computed at high intensity. Bar: A–C = 10 nm.

Figure 7. Analysis of Unlabeled Cells.

Panel A is a single-pixel slice showing matrices comprised of filaments (green) and ∼15 nm diameter particles, referred as “bead” assemblies (red). The region inside the square (256 256 47 pixels) was segmented to reveal the repeating particles that decorate the filaments and comprise the “beads.” Bars A = 45 nm. Panels B–E show selected steps of the analysis. Panel B shows the volume in the square before segmentation. Panel C shows small (blue) and large (yellow) protein particles generated by segmentation. The irregular regions colored brown represent “aggregates comprised of particles >7 nm diameter. Panel D shows the same volume after removing the “aggregates.” Panel E shows the map after measuring the Euclidian distances between blue and yellow protein particles. Lines color-coded according to the dimensions of the particles connected the centers of mass of blue-to-blue, yellow-to-blue and yellow-to-yellow particles (see Table 2). Bar: 60 nm.

Figure 8. Model.

Panel A underscores the fact that the dimers of the αA-crystalline particles (yellow) skip a position in the filament and become spaced at twice the distance (∼14 nm instead of ∼7 nm). At the right side, a kinked file of “real” yellow conjugates reflects this condition. Panel B shows two neighboring filaments decorated with monomers of αA-crystalline. In addition of being spaced at ∼7 nm apart, the αA-crystalline monomers occupy the vertices of isosceles triangles. At the right side, distributions of “real” blue conjugates reflect this condition. Panel C shows how the association of three filaments decorated with monomers at the minimum distance form tetrahedrons or pyramids. At the right side, blue and the yellow conjugates reflect this condition. Panel D shows a larger tetrahedron (the “bead”) formed by the assembly of smaller αA-crystalline particles. At the right side, a “real” particle reflects this condition. Panel E shows a region where these “beads” associate to form the larger “aggregates.” The right side panel shows a view of this type of “aggregate” where the “beads” adopt cobblestone patterns in the interior.