Whole Mount Imaging to Visualize and Quantify Peripheral Lens Structure, Cell Morphology, and Organization

The ocular lens is a flexible, transparent tissue situated at the anterior region of the eye that functions to fine-focus light onto the retina. The ability of the lens to function can be attributed, in part, to its intricate architecture and organization^1,2,3,4,5,6. Surrounding the lens tissue is the capsule, a basement membrane essential for maintaining lens structure and biomechanical properties^7,8,9. The lens itself is entirely cellular, consisting of two cell types: epithelial and fiber cells. The epithelial layer consists of a monolayer of cuboidal cells that cover the anterior hemisphere of the lens¹⁰. Throughout life, the epithelial cells proliferate and migrate along the lens capsule toward the lens equator. Anterior epithelial cells are quiescent and cobble-stone in cross-section, and near the lens equator, epithelial cells proliferate and start to undergo the differentiation process into new fiber cells^11,12. Equatorial epithelial cells transform from randomly packed cells into organized meridional rows with hexagon-shaped cells. Hexagonal cell shape is maintained on the basal side of these differentiating cells while the apical side constricts and anchors at the lens fulcrum or modiolus^4,13,14,15. As the equatorial epithelial cells start to elongate into newly formed fiber cells, the apical tips of the cells migrate along the apical surface of anterior epithelial cells toward the anterior pole while the basal tips move along the lens capsule toward the posterior pole. New generations of fiber cells overlay previous generations of cells, creating spherical shells of fibers. During the cell elongation and maturation process, fiber cells substantially alter their morphology^11,12,16. These fiber cells form the bulk of the lens mass^{11,12,16,17,18}.

The molecular mechanisms that contribute to establishing intricate lens microstructures, cell morphology, and unique cellular organization are not entirely known. Moreover, the contribution of the lens capsule and cell structure to overall lens function (transparency, lens shape change) is unclear. However, these relationships are being elucidated using new imaging methodology and quantitative assessments of lens structural and cellular features^{2,4,19,20,21,22}. New protocols to image whole lenses that allow for high spatial resolution visualization of the lens capsule, epithelial cells, and peripheral fiber cells have been developed. This includes methodology to quantify capsule thickness, cell size, cell nucleus size and circularity, meridional row order, fiber cell packing, and fiber cell widths. These visualization and image quantification methods allow in-depth morphometric examination and have advantages over other visualization methods (imaging of flat mounts or tissue sections) by preserving overall 3D tissue structure. These methods have permitted for the testing of novel hypotheses and will enable continued advancement in understanding of lens cell pattern development and function.

For the following experiments, we use wild-type and Rosa26-tdTomato mice tandem dimer-Tomato (B6.129(Cg)-Gt(ROSA) (tdTomato)²³ (Jackson Laboratories) in the C57BL/6J background between the ages of 6 and 10 weeks, of both sexes. The tdTomato mice allow for visualization of cellular plasma membranes in live lenses via expression of tdTomato protein fused to the N-terminal 8 amino acids of a mutated MARCKS protein that targets the plasma membrane via N-terminal myristylation and internal cysteine-palmitoylation sites²³. We also use NMIIA^{E1841K/E1841K} mice²⁴ obtained originally from Dr. Robert Adelstein (National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD). As described previously²⁰, NMIIA^{E1841K/E1841K} mice in FvBN/129SvEv/C57Bl6 background that has loss of CP49 beaded intermediate filament protein (maintains mature fiber cell morphology and whole lens biomechanics), are backcrossed with C57BL6/J wild-type mice. We screened the offspring for the presence of the wild-type CP49 allele.

Confocal imaging was performed on a laser-scanning confocal fluorescence microscope with a 20x (NA = 0.8, working distance = 0.55mm, 1x zoom), a 40x (NA = 1.3 oil objective, working distance = 0.2mm, 1x zoom), or a 63x (NA = 1.4 oil objective, working distance = 0.19mm, 1x zoom) magnification. All images were acquired using a pinhole size, which is a determinant of optical section thickness, to 1 Airy Unit (the resultant optical thicknesses are stated in figure legends). Images were processed on Zen Software. Images were exported to .tif format and then imported into FIJI ImageJ Software (imageJ.net).

Mice are housed in the University of Delaware animal facility, maintained in a pathogen-free environment. All animal procedures, including euthanasia by CO₂ inhalation, were conducted in accordance with approved animal protocols by the University of Delaware Institutional Animal Care and Use Committee (IACUC).

1. Whole lens mount preparation and imaging

1. Fixation of lenses for whole mount imaging
   1. Following euthanasia, enucleate eyes and dissect lenses as previously described²⁵. Following dissection, transfer lenses immediately into fresh 1x phosphate buffered saline (PBS; 1.1 mM KH₂PO₄, 155 mM, NaCl, 3.0 mM Na₂HPO₄-7H₂O; pH 7.4) at room temperature.
      NOTE: Cellular morphology may be altered if lenses are stored in PBS for an extended period of time, therefore, it is recommended to fix immediately within ~10 min of dissection.
   2. For whole-mount imaging of the lens anterior region, fix whole lenses by immersing in 0.5 mL of freshly made 4% paraformaldehyde (PFA) in 1x PBS in a microcentrifuge tube at room temperature. After 30 min, wash the lenses 3x (5 min per wash) with 1x PBS. Proceed to step 1.2 or store in 1x PBS at 4 °C. Fixed lenses can be stored for up to 5 days.
   3. For whole-mount imaging of the lens equatorial region, fix whole lenses by immersing in 0.5 mL of freshly made 4% PFA in 1x PBS in a microcentrifuge tube at room temperature. After 1 h, wash lenses 3x (5 min per wash) with 1x PBS. Proceed to step 1.3 or store in 1x PBS at 4 °C. Fixed lenses can be stored for up to 5 days.
2. Whole mount of anterior lens region (Fixed or live)
   1. For whole-mount imaging of fixed lenses, proceed to step 1.2.3.
   2. For whole-mount imaging of live lenses, transfer lenses into a well of a 48-well plate containing 1 mL of Medium 199 (phenol red-free) containing 1% antibiotic/antimycotic. Incubate lenses at 37 °C and 5% CO₂ until imaging. Before imaging, incubate lenses in a solution containing fluorescent-labeled Wheat Germ Agglutinin (WGA-640, 1:500) and Hoechst 33342 (1:500) in Medium 199 for at least 10 min. Proceed to step 1.5.
   3. To stain lens capsule, F-actin, and nuclei of fixed lenses, place lenses in a 500 µL solution containing WGA-640 (1:500), rhodamine-phalloidin (1:50), and Hoechst 33342 (1:500) in permeabilization/blocking buffer (1x PBS containing 0.3% Triton, 0.3% bovine serum albumin (Fraction V) and 3% goat serum) in a microcentrifuge tube. Stain lenses at 4 °C overnight.
   4. After overnight incubation, wash the lenses 3x (5 min per wash) with 1 mL of 1x PBS. Proceed to imaging lenses.
   5. To stabilize the fixed lens for confocal imaging, create lens immobilization divots in agarose on an imaging dish as previously described²¹ (Figure 1).
      1. Heat and gently mix 2% agarose in PBS using a microwave till the solution is liquified. Pipette 250 µL of liquefied 2% agarose into a glass bottom dish (Figure 1A) and flatten the agarose across the dish using a flexible plastic coverslip (Figure 1B). Once the agarose is cooled and fully solidified, remove the coverslip using fine-tip forceps (Figure 1C) and use a 3mm biopsy punch to create a hole in agarose at the center of the dish (Figure 1D).
      2. Remove excess agarose using a delicate task wipe (Figure 1E). Keep agarose mold hydrated with 1x PBS and maintain at 4 °C until use. Agarose molds have been successfully stored for up to 1 week.
   6. Using embryo forceps, gently transfer the live or fixed lens into the divot in the agarose (Figure 1F), containing ~2 mL of 1x PBS (fixed lens) or phenol red-free Medium 199 (live lens), and then place the dish on an inverted microscope stage. To confirm that the lens is situated with the anterior region facing the objective, visualize nuclei staining. If no nuclei are observed, the posterior region may be facing the objective.
   7. To invert the lens, use curved forceps and gently rotate the lens ~180° so that the anterior region faces the objective. Acquire images using a confocal microscope.
   8. For visualization of lens capsules, acquire z-stack images using a 40x objective with a step size of 0.3 µm. Acquire the first image before the surface of the lens capsule (indicated by WGA staining) and the last image after the apical surface of the epithelial cells. To visualize epithelial cells, acquire z-stack images using a 63x objective with a step size of 0.3 µm for visualization of lens epithelial cells.
      NOTE: These microscope parameters allow for adequate three dimensional (3D) reconstruction of images which is essential for image quantification of capsule thickness or epithelial cell area. It is also possible to image sutures in live lens tdTomato lenses by imaging past the epithelial cell monolayer as previously described⁴.
3. Lens equatorial epithelial and fiber cell staining
   1. Place fixed lenses in a 0.5 mL solution of rhodamine-phalloidin (1:300), WGA-640 (1:250), and Hoechst 33342 (1:500) in permeabilization/blocking solution (3% BSA, 3% goat serum and 0.3% Triton) within a microcentrifuge tube. Maintain at 4 °C overnight.
   2. After overnight incubation, wash lenses 3x in 1mL 1x PBS (5 min per wash).
   3. Create lens immobilization agarose wedges as previously described^4,10.
      1. Briefly, create agarose wedges in glass bottomed dishes (FD35-100, WPI) by pouring ~5-6 mL of molten 2% agarose in 1x PBS into dishes (Figure 2A). Once the agarose solidifies (Figure 2B), create a triangular divot with a sharp blade (Figure 2C). Remove the agarose wedge and place 1 mL of 1x PBS into the agarose dish (Figure 2D).
      2. Aspirate any residual agarose that may be left behind from cutting the agarose. Multiple wedges can be created per dish to fit multiple different sizes of lenses (not shown). To store the agarose mold, place 1mL of 1x PBS onto the agarose mold and keep at 4 °C. Wedges can be kept for up to 1 week.
   4. Using curved tweezers, place the lens within an agarose wedge containing 1mL of 1x PBS and adjust so that the equatorial region of the lens is facing down onto microscope glass above the confocal objective (Figure 2E-F). To confirm that the equatorial region is in focus, visualize nuclei and ensure that nuclei are aligned in rows at the equator. The equatorial epithelial cells can also be identified as they are irregularly packed and shaped. As well, the meridional row cells are precisely aligned and hexagonal shaped, indicated by F-actin staining at the cell membranes.
   5. If nuclei in the field of view are randomly packed, the anterior side of the lens is facing the objective. If nuclei cannot be observed, then the posterior side of the lens is most likely facing the objective. If the lens is not sitting on its equator, re-orient the lens, and use the curved tweezer to rotate the lens until the precisely aligned nuclei at the lens equator is observed.
   6. Image the lens equatorial epithelial and fiber cells using a laser-scanning confocal microscope (20x objective, NA = 0.8, step size 0.5 µm and/or 40x oil objective, NA = 1.3, step size 0.4 µm). Rotate the images before z-stack collection, in which equatorial epithelial cells are on the top, followed by meridional row and fiber cells right below.

2. Image analysis methodology

1. Lens capsule thickness measurement
   1. Using z-stack images obtained with a 40x objective (0.3 µm step size) of either live tdTomato lenses labelled with WGA or fixed lenses labelled with WGA and rho-phalloidin, obtain an optical 2D projection in the XZ view of the 3D reconstruction and save in .tif format. Process images using Zen software.
   2. Open XZ slice (.tif) images in FIJI ImageJ software.
   3. To measure lens capsule thickness, use the straight line function (depicted in Figure 3A,B). Set the line width to 50 pixels by selecting Edit > Options > Line width. This line width was empirically determined to provide a high signal-to-noise ratio with the microscope settings. Optimal line width may differ on other microscopes/microscope settings or when using lenses from other species. Next, draw a line from the capsule's top surface to beneath the epithelial cells' basal region.
   4. Save the line as a region of interest by pressing Ctrl + T.
   5. Separate channels into tabs for image, color, and split channels.
   6. Measure the intensity along the line for the capsule channel by pressing Ctrl + K.
   7. In a spreadsheet, plot intensity values as a function of distance and determine intensity peaks for the capsule and F-actin, which represent the capsule surface and basal region of epithelial cells, respectively. On the x-axis is the line distance.
   8. Next, calculate capsule thickness by determining the distance between capsule and epithelial fluorescent intensity peaks (Figure 3C, D).
2. Cell area analysis
   1. Using z-stack images obtained using a 63x objective lens (NA = 1.4; 0.3 µm step size) on live tdTomato lenses or fixed lenses labelled with phalloidin, analyze an XY view slice from a z-stack confocal image corresponding to where the middle (lateral) region of epithelial cells is in focus. Note the lateral membranes are visible using the tdTomato membrane dye or by visualizing phalloidin which are present at the lateral cell membranes.
      NOTE: Due to the curvature of the lens (as seen in an XZ view; Figure 4A-C), not all epithelial cells will be in focus in the image. The middle region of the epithelium corresponds to the region where both the cell lateral membranes (Figure 4D-E) and nuclei (Figure 4G-I) are in focus.
   2. Export the image as a .tif and open it in FIJI ImageJ.
   3. To set the scale in FIJI ImageJ for analysis, use the line tool to create a line that is the length of a scale bar from the confocal image.
   4. Record the pixel length of the line and the length of the scale bar. Go to Analyze on the toolbar menu and select Set Scale. Input the number of pixels that is the length of the line in Distance in Pixels and in the Known Distance input the known length of the scale bar.
   5. Using the tdTomato or rhodamine-phalloidin staining as an indication of cell boundaries, manually outline a population of cells that are in focus within the image using the polygonal tool. Only trace cells where lateral membranes are visible (Figure 4E) and avoid tracing partially visible cells (i.e., at the edge of the images). Save ROI by pressing Ctrl+T. Go to Edit on the toolbar menu and select Clear Outside. Now measure total cell area (ROI) by pressing Ctrl + M in FIJI ImageJ.
   6. Calculate the average cell area by dividing the ROI area by the total cell number. A simple way to determine the number of cells is by counting the number of nuclei. Go to the nuclei channel (blue). On the ROI menu, select the saved ROI outline of cells (Figure 4G). Clear outside of the ROI by going to Edit and then clicking Clear outside (Figure 4H). Using the Multi-point tool, count the nuclei by clicking on individual nuclei.
3. Cellular nuclear area and shape analysis
   1. For nuclei area and shape analysis, analyze an XY plane view optical section from a z-stack confocal image corresponding to where the middle (lateral) region of tdTomato epithelial cells are in focus. Analyze a z-stack slice from a confocal image where the nuclear area is the largest (Figure 5A); this represents the mid portion of the nuclei.
      NOTE: Note that due to the curvature of the lens, not all nuclei will be in focus, as can be seen in the XZ view (Figure 4G and Figure 5A).
   2. Select a subpopulation of nuclei that are in focus and save a ROI.
   3. Use the Clear Outside function. Within the ROI, using the Freehand selection tool (fourth menu button from the left), carefully trace the borders of the nuclei (Figure 5B). Save each nuclear trace into the ROI window by pressing Ctrl+T. Only outline nuclei where the complete nuclei can be seen, and the nuclei are not touching each other.
   4. Once all the nuclei are outlined, measure individual cells' nuclear area and shape (i.e., circularity) by pressing Ctrl+M.
   5. Copy and paste data into a spreadsheet and calculate the average nuclear area and circularity (Figure 5C).
4. Meridional row epithelial packing
   1. To measure meridional row disorder, acquire images using a 20x objective (0.5 µm step size).
   2. Identify the lens fulcrum/modiolus on images. The fulcrum is the region where the apical tips of elongating epithelial cells constrict to form an anchor point during initial fiber cell differentiation and elongation at the equator. Identify the fulcrum based on bright F-actin intensity (indicated by arrowhead in XZ view in Figure 6A; indicated by a red dashed line in Figure 6B) and change in cellular organization in a single optical section at XY view.
      NOTE: In addition, F-actin staining of both the basal and apical regions of epithelial cells are apparent above the fulcrum, whereas only the basal region of the elongating cells are apparent below the fulcrum (Figure 6A). The epithelial cells above the fulcrum line are irregularly packed and tend to form rosettes, whereas the fiber cells below the fulcrum are arranged in parallel rows, indicated by bright F-actin staining at the membrane (Figure 6B).
   3. Once the fulcrum is identified in 2-month-old mice, select the single optical section ~4.5-5 µm peripheral to the fulcrum (toward the lens capsule) in the XY plane. This distance is selected based on the observation that all the nuclei of the meridional row cells in 2-month-old mice are in focus when they are ~5 µm peripheral to the fulcrum. Save the single optical section (.tif). Open the image on FIJI.
      NOTE: This analysis has only been performed on 2-month-old mice and therefore, it cannot be concluded whether this distance changes with age.
   4. Before performing image analysis, set the scale to µm in ImageJ using step 2.2.2.
   5. Manually outline the entire meridional row regions using the freehand line tool (region of interest/ ROI) by identifying nuclear alignment, as shown in Figure 7A. Save ROI by pressing Ctrl+T. Go to Edit on the toolbar menu and select Clear Outside. Measure total cell area (ROI) by pressing Ctrl + M in FIJI ImageJ.
   6. Identify any regions of disorder indicated by F-actin staining by outlining using the freehand line tool in FIJI ImageJ. Criteria for disordered regions include branching of rows, irregular packing, and misalignment of rows (Figure 7B), as shown previously²⁰.
   7. Measure the outlined disordered area/patched by pressing Ctrl + M in FIJI ImageJ. Sum the total disordered area in a spreadsheet.
   8. Divide the total disordered area by the ROI area, then multiply by 100 to get a percent disordered area. If no disorder is observed, place a value of 0% for the disordered area.
5. Meridional row number of neighboring cells
   1. To measure number of neighboring cells, use F-actin-stained images acquired with 40x oil objective (0.4 µm step size).
   2. Identify an optical section (XY plane) at the basal region of the meridional row cells inward from the lens capsule where F-actin is enriched around the entire perimeter of the meridional row cells, all meridional row cells are in focus and are on the same plane.
   3. Count the number of adjacent cells each meridional row cell has (Figure 8). Measure the average percentage of cells with six adjacent cells in a spreadsheet.
      NOTE: Determine the number of adjacent cells with a 20x objective. However, it is significantly easier to observe the hexagon shape with a 40x objective.
6. Image analysis of equatorial fiber cells
   1. Take z-stack confocal images of rhodamine phalloidin stained lenses with a 20x objective (0.5 µm step size).
   2. Identify the fulcrum as indicated in step 2.4.2. Once the fulcrum is identified, select a single optical section. For standardization purposes and to compare between lenses, quantify fiber cell widths ~10 µm inward from the fulcrum in the XY plane (Figure 9A).
   3. Export raw images to FIJI ImageJ. In FIJI ImageJ, draw a line (usually ~300-400 µm long) across several adjacent fiber cells to measure the distance between F-actin-stained peaks (line scan analysis; Figure 9A; pink line).
   4. Obtain the fluorescent intensity over the line scan distance in FIJI by pressing Ctrl+K. Next, export data into the spreadsheet to calculate the interpeak distance (example shown in Figure 9A-B). This corresponds to fiber cell widths.

Anterior lens capsule, epithelial cell area, and nuclear area
To analyze lens capsule thickness, we stained lens capsules, in either live or fixed lenses, with WGA. We identified lens epithelial cells by labeling membranes with tdTomato in live lenses (Figure 2A), or via rhodamine-phalloidin staining for F-actin at the cell membranes in fixed lenses (Figure 2B). In an orthogonal (XZ) projection, staining for WGA and tdTomato/rhodamine-phalloidin allows us to perform peak-to-peak line scan analysis of fluorescence intensity. The major peak in the WGA channel indicates capsular surface whereas the major peak in the tdTomato/rhodamine-phalloidin channel indicates the basal region of epithelial cells. By calculating the distance between these peaks, we can obtain capsular thickness. The line scan analysis shows that capsules from a 9-week-old mice live lens had a thickness of 11.2 µm, and capsules from a 9-week-old mice fixed lens had a thickness of 12.5 µm. These observed capsule thicknesses are representative of previous findings^2,4.

Whole-mount imaging of tdTomato-labeled transgenic mouse lenses (or rhodamine-phalloidin stained lenses; not shown) allows for live visualization of epithelial cell morphology. The orthogonal (XZ) projection provides a side view of the lens epithelial cell, while the planar (XY) view at the lateral regions allows for visualization of the epithelial polygonal shape. In healthy lenses, we do not observe any gaps between cells. We can calculate the average cell area by tracing a population of cells in an image and dividing the area by the number of cells within the ROI. The number of cells is determined by counting the number of Hoechst-stained nuclei in a given ROI. The analysis demonstrates an average cell area of 260 µm² which is in keeping with previous studies^2,4.

Hoechst staining of nuclei also allows for examination of lens epithelial cell nuclear morphometry in epithelial cells. The orthogonal (XZ) projections allow for a side view of nuclei. The planar (XY) view demonstrates the circular/ellipsoid shapes of the nuclei. Tracing nuclei allows for calculating individual cells' nuclear area, and other shape parameters such as circularity. The analysis demonstrates an average nuclear area of 64 µm² with an average circularity of 0.8. Circularity values close to 1 indicate a perfect circle, whereas values approaching 0 indicate a more elongated morphology.

Equatorial epithelial cell packing, fiber cell hexagonal packing, and fiber cell widths
The planar (XY) view of the lens equatorial region demonstrates hexagon shaped and regularly packed lens epithelial cells which converge at the fulcrum. The fulcrum is where the apical tips of elongating epithelial cells constrict to form an anchor point during initial fiber cell differentiation and elongation at the equator^4,13,14,15 (Figure 6B, indicated by a red dashed line). The fulcrum can be localized based on increased F-actin staining forming a continuous line separating the equatorial epithelial cells and fiber cells (Figure 6B, red dashed line). The F-actin staining at the cellular membranes also demonstrates changes in cell shape, in which cells below the fulcrum are precisely aligned and arranged in parallel rows (Figure 6). Cell nuclei are also aligned beneath the fulcrum.

To visualize lens meridional row epithelial cells and fiber cells, an image 4.5-5 µm peripheral to the fulcrum (toward the lens capsule) in the XY plane, where the basal region of meridional row cells are in view, is selected as described previously²⁰. To measure meridional row disorder, a region of interest (ROI) is outlined (Figure 7A; yellow box). The area of the ROI in the wildtype lens image is 15,833 µm². As there is no observable disorder, the area of disorder percentage is 0. The critical role that non-muscle myosin IIA (NMIIA) plays in cell hexagonal packing using mice with an NMIIA-E1841K mutation has been previously described²⁰. Figure 7A shows a representative NMIIA^{E1841K/E1841K} lens equatorial image to demonstrate meridional row cell disorder. The meridional row ROI was 20,757µm2. Next, the total area of disordered patches was traced. The total area of disorder was 3,185 µm². The calculated percentage of disorder was determined to be 15.3% (disordered area x 100/total ROI; Figure 7B). This percent disordered area is within the range in a previous study²⁰.

Next, an examination of hexagonal packing in wild-type meridional row cells was demonstrated²⁰. Because F-actin is enriched around the entire perimeter of meridional row cells and at all six vertices of the basal regions of cell membranes in wild-type (i.e., NMIIA^+/+) lenses, F-actin staining was used to assess cell shapes and packing organization. In representative image 1, cells were labeled and the number of adjacent cells around each cell was counted. In image 1, all ten cells have six adjacent cells, which suggests that these cells are arranged in a honeycomb packing organization (Figure 8). In contrast, eight out of 10 cells (80% of cells) have six adjacent cells in image 2 that indicate that the cells are irregularly packed (Figure 8).

Finally, to measure the fiber cell width, the peripheral fiber cells located ~10 µm inward from the fulcrum in fixed wild-type lenses labeled with rhodamine-phalloidin were examined (Figure 9A)^2,4. Of note, it is also possible to measure fiber cell widths using live tdTomato mice, however, the signal from lenses that are heterozygous for tdTomato tends to be weak at the equatorial regions. Therefore, using mice that are homozygous for tdTomato is recommended as they have been found to have stronger fluorescence (not shown). The distance between the F-actin-stained cell boundaries using line-scan analysis was measured as described previously to indicate fiber cell width^2,4 (Figure 9B). This analysis revealed that the average interpeak distance in wild-type lenses is 11.45 ± 2.11 µm (N=117 fiber cells from 4 different mouse lens images, Figure 9B).

Figure 1: Steps to create agarose divot to immobilize lens during imaging. (A) Using a glass bottom dish, pipette liquefied 2% agarose. (B) Flatten with a flexible cover slip and (C) when cooled, remove using fine tip forceps. (D) For the divot, create a hole using a 3 mm biopsy punch. (E) Aspirate residual agaros, rinse with PBS, wipe surface clean using a lint free tissue. (F) Carefully mount whole lens within divot. Please click here to view a larger version of this figure.

Figure 2: Steps to create agarose divot for imaging lens equatorial epithelial and fiber cells. (A) Pour 2% molten agarose in the tissue culture dish. (B) Cool the agarose at roomtemperature until it solidifies completely. (C) With a sharp blade, create a triangulardivot. (D) Put 1 mL of 1x PBS in the tissue culture dish. (E) Place the dissected lens inthe wedge with (F) the lens propped up on its equator wedged between the agarose walls (red arrow). Please click here to view a larger version of this figure.

Figure 3: Determination of lens capsule thickness. Sagittal (X, Z plane view) optical sections from reconstructions of confocal z-stacks of (A) live and (B) fixed lens capsules. Fluorescent intensity of line scan of (C) live and (D) fixed lenses demonstrate a single WGA (green) peak that corresponds to the top surface of the capsule and the basal region of epithelial cells (red) adjacent to the capsule. The distance between the two peaks is measured to quantify capsule thickness. Images acquired using a 40x oil objective with a 1 airy unit pinhole resulting in optical sections of 1.0 μm and 1.2 μm in the tdTomato/Rhodamine-Phalloidin and WGA channels, respectively. Please click here to view a larger version of this figure.

Figure 4: Quantification of epithelial cell area. (A) X, Z plane view of the lens epithelial cells marked with (B) tdTomato for cell membranes and (C) Hoechst for nuclei. (D) X, Y plane view of the middle region of lens epithelial cells. (E) tdTomato signal is used as a guide to define a region of interest corresponding to a group of cells that are in focus. (F) The area of the defined region is determined. (G) Hoechst staining for nuclei is used to determine the number of cells within the defined region. (H) The number of nuclei is counted using (I) FIJI ImageJ's multipoint tool. To calculate the average cell area, divide the total area of the defined region of interest by the total number of cells. Images acquired using a 63x oil objective with a 1 airy unit pinhole resulting in optical sections of 0.7 μm and 1.0 μm in the Hoechst and tdTomato channels, respectively. Please click here to view a larger version of this figure.

Figure 5: Determination of nuclear area and shape. (A) XY plane view of the middle region of nuclei. (B) A region of interest where nuclei are in focus was defined. Nuclei within the ROI are outlined. (C) The area and circularity of nuclei of individual cells were tabulated and averages for area and circularity were calculated. Images acquired using a 63x oil objective with a 1 airy unit pinhole resulting in an optical section of 0.7 μm in the Hoechst channel. Please click here to view a larger version of this figure.

Figure 6: Identification of the lens fulcrum. F-actin and nuclei can be used to identify the fulcrum and meridional rows. (A) XZ view shows enriched phalloidin staining for F-actin which corresponds to the fulcrum. (B) Single optical XY plane view section with the lens fulcrum in focus. Hoechst staining for nuclei shows that the nuclei above the fulcrum are irregularly packed whereas the nuclei below the fulcrum are precisely aligned (Top right). Phalloidin staining for F-actin shows that the cells above the fulcrum are irregularly shaped whereas the cells below the fulcrum are packed in parallel rows (Bottom left). The red dashed line shows the location of the fulcrum. Images acquired using a 20x objective with a 1 airy unit pinhole resulting in optical sections of 1.5 μm, 2.0 μm and 2.2 μm in the Hoechst, Rhodamine-Phalloidin, and WGA-640 channels, respectively. Please click here to view a larger version of this figure.

Figure 7: Determination of meridional row disorder. (A) Single XY plane view optical section of the wildtype and NMIIA^{E1841K/E1841K} meridional row cells ~5 µm away from the fulcrum (toward the lens capsule). The entire meridional row cells are outlined in blue based on nuclear alignment. F-actin (red) staining in wildtype lens shows that meridional row cells are precisely aligned with no signs of disorder (0%). A representative ordered region in wildtype is outlined (orange). In lenses with disordered cells, we outline the disordered regions (yellow). (B) High magnification of ordered region from wildtype (orange box in A). (C) High magnification of disordered areas showing different types of disorder including (I) branching of rows, (II) irregular cell shape and loss of honeycomb packing, and (III) misalignment of rows. Images from NMIIA^{E1841K/E1841K} lens show 15.3% meridional row cells disorder. Images acquired using a 20x objective with a 1 airy unit pinhole resulting in optical sections of 1.5 μm, and 2.0 μm in the Hoechst and Rhodamine-Phalloidin channels, respectively. Please click here to view a larger version of this figure.

Figure 8: Analysis of lens meridional row cell honeycomb packing. (A) Single optical XY sections of meridional row cells stained with rhodamine-phalloidin for F-actin. Low magnification images (Top panel), show the cells being evaluated (numbered in pink). Region outlined in yellow is enlarged (Bottom panel). The yellow roman numerals are the counts of adjacent cells. Image 1 shows cells that are all hexagon in shape, each with 6 adjacent cells. Image 2 has irregularities with cell number 1 and 5 having 5 and 7 adjacent cells, respectively. (B) Data is recorded and tabulated with the percentage of hexagonal cells calculated. Images acquired using a 40x objective with a 1 airy unit pinhole resulting in an optical section of 1.0 μm in the Rhodamine-Phalloidin channel. Please click here to view a larger version of this figure.

Figure 9: Analysis of fiber cell widths. (A) Single optical XY view section of the lens fiber cells ~10 µm in from the fulcrum. Cells were stained with rhodamine-phalloidin for F-actin visualization at the cell membranes. A line (pink; dash) is drawn over a number of fiber cells. (B) Representative line scan of F-actin intensities as a function of distance. The interpeak distance represents the fiber cell width. Images acquired using a 40x objective with a 1 airy unit pinhole resulting in an optical section of 1.0 μm in the Rhodamine-Phalloidin channels. Please click here to view a larger version of this figure.

The protocols described enable high spatial resolution visualization of peripheral lens structures and cells at the anterior and equatorial regions of the lens. In this study, methods for the visualization of lens peripheral structures using intact (live or fixed) lenses where the overall 3D lens architecture is preserved were shown. Additionally, simple methods for morphometric quantitative analysis using publicly available FIJI ImageJ software were provided. The whole mount visualization and quantification methods has been used in previous studies. These methods allowed us for understanding the response of the anterior capsule and cells to lens shape change or aging^2,4. These methods have also been used to examine equatorial fiber cell expansion due to lens shape change or aging^2,4 and to determine NMIIA's role in establishing precise hexagonal shapes and peripheral fiber cell organization²⁰.

Whole-mount imaging allows for high spatial resolution en face imaging of lens structure. At the anterior lens region, whole-mount imaging is advantageous to flat-mount imaging by preventing damage that may alter capsule structure integrity and/or epithelial cellular morphology. Additionally, the interface between epithelial and fiber cells is preserved. This method provides advantages over the visualization of lens sections as en face imaging affords greater spatial resolution and permits analysis of epithelial cell area and nuclear area/shape the selected region (i.e., mid-lateral, as shown here) or other regions of cells. Furthermore, the imaging methods allow for quantifying morphological features at specific points along the lens's equatorial regions, enabling visualization of hexagonal shapes, meridional row cell packing, fulcrum, and fiber cell widths, which would not be readily achievable via imaging-stained lens sections due to lack of spatial resolution and tissue distortion that can occur during sectioning.

The outlined protocols also allow for the imaging of live lenses to track specific lens structures and cells over time. The live lens imaging protocol was instrumental in a previous study, where repeated measures analysis of capsule thickness and epithelial cell area from a subpopulation of cells from the same lens before and following lens compression to induce lens flattening was performed⁴. It was determined that lens compression induced a decrease in capsule thickness and an increase in epithelial cell area. Due to substantial variation in capsule thickness and epithelial cell area between individual lenses and the magnitude of effects on these parameters caused by lens compression, it would be difficult to detect differences if an independent measurement design was used (i.e., comparing individual non-compressed lenses versus individual compressed lenses). Using the live imaging methods, whole-mount imaging on live mouse lenses that endogenously express LifeACT-GFP²⁶ to visualize F-actin in epithelial cells²², which enables tracking of F-actin reorganization in live epithelial cells, have been conducted.

While the outlined protocols were developed for imaging mouse lenses and may be adapted to visualize lens structures in other species, the staining protocols to visualize lens both anterior and equatorial peripheral structures in other rodent lenses (rat, guinea pig) as well as in other mammalian lenses (cow, macaque, and human; data not shown). Visualization of structures in larger lenses requires longer fixation (4%PFA, on ice, 4 h), blocking (1 h), and staining (overnight, 4 °C). Of note, the imaging in different species was only conducted on fixed lenses. Live imaging of lenses from other species may be challenging as the live imaging protocol uses lenses from transgenic mice that express fluorogenic proteins. Therefore, this imaging may preclude imaging lenses of other species where genetic modifications are not commonplace. However, we have been able to conduct live visualization of lens epithelial cell membranes or F-actin by prestaining mouse lenses with the lipophilic probe, FM4-64 or an F-actin binding probe, SiR-Jasplakinolide (SiR-actin), respectively (not shown). It may be possible to use such probes to image live lenses from other species. However, caution must be taken when using such probes to avoid off-target effects; for example, SiR-Jasplakinolide can lead to altered F-actin dynamics²⁷. Another limitation of the staining methods, whether on live or fixed lenses, is that such probes have limited penetration. Therefore, imaging is limited to peripheral regions. It is possible to image inner regions (i.e., deep fiber cells) using lenses from tdTomato transgenic mice⁴, which again relies on the transgenic expression of fluorogenic proteins. The expression also must be high enough for bright signals in deep fiber cells.

Nevertheless, the protocols outlined have enabled effective morphological quantitation of peripheral lens features^2,4,20. The quantification methodologies are effective and use open-source, publicly available FIJI ImageJ software. While manual tracing tools were used to identify regions of interest, it may be possible to automate the identification of regions of interest. Automation may be as simple as applying fluorescent thresholding to enhance contrast for FIJI ImageJ-based identification of particular structures (i.e., nuclei boundaries), or more complex machine learning algorithms to detect cell shape differences (i.e., irregular epithelial or fiber cell shapes). More complex analysis may require either specialized plugins or imaging software. Specialized plugins or software may also allow for obtaining 3D volumetric morphological measures from acquired z-stacks. Overall, while the quantification methods described are easily accessible, cost-effective, and suitable for many useful outcome measures^2,4,20, the imaging protocol platform, in combination with sophisticated software analysis, could provide a wealth of additional morphological measurements for future studies.

The lens is an intricate biological tissue with specialized functions that rely upon location and depth-dependent geometries of the cells and their associated structures. Using the imaging protocols and quantification methods demonstrated here will allow for a greater understanding of how lens structures and the complex organization of the lens are established. Furthermore, the imaging protocols and quantification methods will help delineate the relationships between lens structure and functions.