Structural Specializations Emerging Late in Mouse Lens Fiber Cell Differentiation

Abstract

purpose. To describe a previously uncharacterized structural specialization in the mouse lens fiber cell and to delineate its emergence relative to lens development and fiber cell differentiation.

methods. Lens fixation efficiency was explored using ¹⁴C-formaldehyde and autoradiography. Lens fiber cell architecture was examined by scanning electron microscopy and by DiI labeling of methacrylate sections in lenses ranging from 2 weeks to 8 months.

results. Scanning electron microscopy identified an elaborate structural specialization that emerges late in fiber cell differentiation, largely after the cell has lost its nucleus. These elaborations project from the short side of the cell, are regularly spaced throughout the central region of the cell and are aligned with similar structures in adjacent cells. The structures are not found in fiber cells of lenses younger than two weeks of age, nor in the fiber cells that initially differentiate before that time.

conclusions. Fiber cells that arise later than 2 weeks of age undergo a structural differentiation program that is different from that of cells that arise earlier in development. This program includes the assembly of a series of regularly spaced, complex, lateral projections from the fiber cell that align themselves with similar structures in adjacent cells. Most if not all of the structural specialization occurs in cells that have lost their nuclei and organelles, suggesting that this component of fiber cell differentiation may not require ongoing transcription/translation.

The ocular lens has the unusual requirement of optical clarity. The features critical to clarity are only partly understood, but the extreme degree of structural differentiation of fiber cells, and the long-range order with which they are arranged are both considered essential. ^{1 2 3 4 5} The importance of these features is deduced in part from the degree to which these elements have been conserved among vertebrate lenses, but also the consequences to optical clarity when these features are disrupted. Among the many features of fiber cell differentiation that are highly conserved are (1) extraordinary elongation; (2) the emergence of unique interdigitations between fiber cells, commonly called “ball-and-socket” interactions (BS interactions), and/or “lateral protrusions” ⁶ ; (3) the long-range order assumed by fiber cells; and (4) the elimination of membranous organelles. ^{7 8 9 10 11 12 13 14}  

Also specific to the lens fiber cell is the beaded filament, a structurally and biochemically unique cytoskeletal element. ^{15 16 17 18 19 20 21} Efforts to define the function of the BF have included the generation of “loss of function” mice, in which one or both of the BF proteins have been knocked out. ^{22 23 24} The resultant lens phenotype has been described as exhibiting an increase in light scatter detectable by slit lamp examination starting at about 3 weeks, with a continued worsening of the optical qualities over time. ^{22 23 24} The light scattering is derived principally from a laminar region in the mid to deep cortex. Exactly how the loss of the BF protein causes increased light scatter is not fully understood, although a loss of cell shape and long range order among fiber cells has been implicated by Sandilands et al. ²⁴ Mutations in human beaded filament proteins have been implicated in opacifications that emerge in late childhood/early adolescence, a timing that is analogous to that in the mouse model. ^{25 26} The work reported herein was initiated in an effort to define precisely how dysfunctions in the beaded filament network cause light scatter. To accomplish this, the normal progression of structural differentiation in the wild-type mouse had to be established. 

The fiber cell undergoes many changes late in the process of fiber cell maturation, some of which appear to be occurring around the time fiber cells undergo organelle loss, a relatively sudden and coordinated event in fiber cell maturation. ¹⁰ In seeking to establish cause-and-effect relationships between BF proteins, changes that occur in BF proteins, the emergence of light scattering centers, changes in cell architecture, fiber cell long-range order, and other major watersheds in fiber cell maturation, such as organelle loss, we have begun efforts to refine the timing of these events more precisely. As a starting point we sought to define clearly the structural changes undergone by wild-type mouse lens fiber cells, with specific respect to the loss of fiber cell organelles. In gathering the data we noted the emergence of a complex structural specialization in the lens fiber cell relatively late in its maturation. These structures emerge only in fiber cells “born” after about postnatal day (P)14 and occur regularly along the middle region of the fiber cell. The periodic placement of these structures is well-aligned between fiber cells over a distance that spans many fiber cells, suggesting cell–cell coordination/interaction in the assembly of these structures. Remarkably, much of the assembly of these structures appears to occur after the fiber cell has lost its nucleus and other organelles. 

Materials and Methods

Tissue Processing

All animal procedures were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

C57bl6 mice (Jackson Laboratories, Bar Harbor, ME) of specified ages were euthanatized with dry carbon dioxide. Eyes were enucleated, and openings were made in the posterior of the globe. The globe was immersed in 2.5% glutaraldehyde 2% formaldehyde in 0.1 M sodium cacodylate buffer (GF fix) at room temperature, on a shaker table. In initial studies lenses were allowed to fix overnight. However, subsequent dissection of lenses suggested that even after 24 hours in fixative, aldehyde fixation was not occurring in the central lens. To test this, some lenses were immersed in GF fix, spiked with ¹⁴C-formaldehyde. Lenses of different ages were incubated for 1, 2, 4, and 24 hours, washed free of unbound ¹⁴C-formaldehyde, then processed into methacrylate for sectioning. To test whether fixative penetration was more effective in lenses that were split, a parallel set of lenses were fixed 30 to 60 minutes and split in half along the anterior–posterior axis. These were then immersed in ¹⁴C formaldehyde for 2 hours, and processed with whole lenses. Ten-micrometer sections were harvested and exposed to autoradiographic film (BioMax MS; Eastman Kodak, Rochester, NY) for 1 and 7 days at −80°C, to determine the degree of penetration of the radioactive formaldehyde. 

Scanning Electron Microscopy

For scanning electron microscopy, the mouse lenses were removed to GF fix, incubated for 30 minutes to 1 hour, and split in half along the midsagittal axis. Fixation was allowed to continue overnight. The tissues were dehydrated through acetone and critical point dried. The lenses were split into quarters along the anterior–posterior axis, yielding a freshly fractured surface that exposed a complete radius. 

DiI Membrane Labeling

To be able to identify both fiber cell nuclei and fiber cell membrane shape clearly, we fixed the lenses as described, dehydrated them through ethanol, and infiltrated them with glycol methacrylate resin. Sections were cut at approximately 2-μm thickness, dried on glass slides, and incubated in the lipophilic dye DiI (0.01 mg/mL in ethanol). After they were rinsed, the slides were photographed (model E800 microscope; Nikon, Tokyo, Japan). 

Results

Lens Fixation

Initial efforts to split or subdivide lenses fixed by immersion in conventional electron microscopy fixatives suggested that the central lens region was not adequately fixed, even after 24 hours. This was a significant concern, as it suggested that there was a fixation gradient that paralleled the gradient of fiber cell differentiation. This raised the possibility that changes in fiber cell structure (or immunoreactivity) that are commonly seen in lens studies might be a function of, or affected by, time to fixation. Thus, proteins (or cell structures) in surface cells may be effectively fixed and preserved in a matter of minutes, whereas those in deeper cells may not be fixed for many hours, during which time they may be subject to proteolysis or other changes. To determine the relative rates and degree of fixation, we immersed lenses in fixative spiked with ¹⁴C-formaldehyde, and then assessed fixative penetration by using autoradiography. 

Figure 1shows autoradiograms that reveal the relative degree of formaldehyde penetration as a function of time. Figure 1aestablishes that, after 1 hour, formaldehyde penetration showed a steep gradient: the outer edge was far more heavily labeled than the rest of the lens. At 2, 4, and 24 hours (Figs. 1b 1c 1d) , the gradient was still very steep, with sharp differences between lens surface and lens center, though the area of heavy labeling became increasingly larger with time. 

Autoradiography relays information about the relative degree of fixation. The absence of signal in the central lens does not necessarily mean that no fixation has occurred, only that it is vastly greater in the cortex. Extended exposure of tissues to film, however, did not produce an obvious increase in density in the nuclear regions that could be distinguished from encroaching exposure from the cortical regions. This, along with our experiences in handling the nuclear regions after immersion fixation, suggests that little or no fixation occurs in the central lens at these early time points. Thus, at the very least, immersion fixation of mouse lens results in a dramatic gradient in fixation, and this asymmetry does not equalize, even after 24 hours. 

In this study our interest resided in an examination of structural changes that spanned from surface to deep cortex. We therefore explored the utility of splitting the lens after a brief fixation. The prefixation gave just enough structural stability to the softer cortex that it did not compress or shear dramatically during the process of splitting. Once the lens was divided in half, fixative had access to both the outer surface and the full cut surface. To assess whether fixative penetrated into the deepest parts of the split lens, split lenses were exposed to ¹⁴C-formaldehyde for 2 hours, rinsed free of unbound probe, then processed for autoradiography. Sections taken from the mid regions of the split lenses and exposed to film are shown in Figure 1e . These showed that relatively uniform fixation was achieved across the full thickness of the split lens, in less than 3 hours, thus decreasing, though not eliminating, the potential for degradative artifact and altering the direction of the fixation gradient. 

Scanning Electron Microscopy

To map out the progression of structural changes that occur in the mature wild-type mouse lens, we examined 3-month-old lenses using SEM. Figure 2ashows an overview of a quartered lens spanning from surface to the center, or bow, region at the top of the image. The boxed region is shown at higher magnification in Figure 2b . The image shows that a dramatic change in fiber cell architecture occurs with a relatively sudden onset approximately 45 to 50 cells from the surface (arrow). A higher magnification of the region at which these specializations emerge (Fig. 2c)shows that the transition from a simpler fiber cell architecture to the more elaborate one occurs over the span of approximately 10 cells. 

A closer look at these structural elaborations is shown in Figure 3 . The newly differentiated fiber cell has a cross section that is a flattened hexagon and that shows the progressive emergence of what has been described as lateral interdigitations in the primate lens. ⁶ These interdigitations have many of the size and regularity features of what has been described alternatively as ball-and-socket (BS) interdigitations. For the present we will refer to them as ball and socket, but differences between “ball and socket” and “lateral interdigitations” may be real and prove significant. Figure 3ashows well-developed BSs, present in two parallel rows, projecting from the lateral edges of the fiber cell, with individual BS approximately 0.9 μm in diameter. At the juncture indicated by the solid arrow in Figure 2c , fiber cells suddenly developed regular, paddlelike protrusions from the lateral surfaces that were much larger (∼2–3 μm) than the well-documented BS interdigitations (∼0.9 μm). Figure 3bshows a fiber cell artificially elevated from adjacent fiber cells, revealing very clearly the nature and regularity of these paddlelike structures and how each was, in turn, covered with the smaller BS. Figure 3creveals the complexity of the lateral interactions between fiber cells bearing these paddlelike structures. This image shows that not only was there complementarity between cells at the level of the BS (small box), but also at the larger level of the paddlelike processes (large box). 

The location of these paddles was well-synchronized between adjacent cells not only within a generation, but between generations. Figure 4shows the lateral edges of approximately 40 fiber cells, revealing how these paddles “stack” from cell to cell, creating the appearance of mostly parallel columns (see for examples cells labeled with an asterisk). The section in Figure 4b , taken from a 6-month-old animal, shows that these columns or stacks of paddlelike extensions, tended to merge and diverge over a distance of 20 to 30 cells. 

In addition to exhibiting short side–long side asymmetry in their distribution, these structures were also much better developed in the mid or equatorial region of the fiber cell than at either end, thus showing anterior-posterior asymmetry as well. This can be seen in the low-magnification overview in Figure 2awhere the columns of aligned paddles created a texture that was rich in the equatorial region, but diminished toward the anterior and posterior poles. 

As seen in Figures 2b and 2c , the paddles emerged relatively suddenly approximately 45 to 50 cells from the surface. However, they did not persist to the lens center, showing instead a gradual diminution in size over the course of a few hundred cells, as can be seen in Figures 2a and 2b . Thus, while “onset” was abrupt and occurred over a span of 5 to 10 cells, the transition back to a simpler fiber cell profile was a very gradual one, occurring over a distance that spans well over a hundred cells, until finally there was little to no evidence of these paddles. 

The observation that these structures were evident only in the deeper cortex and lacking from the deeper regions suggests that they are either a transient structure, emerging late in fiber cell differentiation, then disappearing still later, or are a structure that is not formed in fiber cells with an early birth date, but are formed in fiber cells with a later birth date. To explore this notion, we examined both younger and older lenses. Figure 5shows sections of a P14 lens. An overview spanning from nucleus (N) to bow is shown in Figure 5a , and a higher-magnification view of the outer cortex, reaching to the equatorial surface, is shown in Figure 5b . The area encompassed in Figure 5bincludes the 45 to 50 cells region where, in the 3-month lens, the paddlelike structures emerge and mature. Figure 5cpresents a closer look at this region, showing the BS structures, but no evidence of forming columns of paddlelike projections. 

Examination of P21 and P28 lenses reveals the first emergence of the paddles at about P21. Figure 6ashows a lower-magnification view of the P21 lens, with the first evidence of the columns of paddlelike structures making an appearance (arrow). In Figure 6b , a higher-magnification view of the emerging columns (arrows) is shown. 

To assess whether the fiber cells with a birth date before P14 form these paddlelike structures, we measured the diameter of the P14 fixed lens, and examined that same region in a series of older lenses. Figure 7presents an overview of an 8-month-old lens. Superimposed on this image is the radius of a representative P14 lens, derived from tissues prepared for SEM. At no point did those cells with a “birth date” of P14 or earlier form the prominent paddlelike structures. No adjustment was made for the compression-compaction that occurs in the fiber cells of the older lens; thus, the 1250-μm diameter of the P14 lens, probably represents an even smaller diameter in the older lenses. 

Lipophilic Labeling of Fiber Cell Membranes

It is evident from the SEM overviews that the emergence of the paddle-like structures occurs somewhat suddenly, at approximately 45 to 50 cells from the bow surface. This is about the same point at which fiber cells are reported to lose their nuclei and organelles. We sought to correlate the timing of these two events. To count the number of nucleated fiber cells more accurately, we processed lenses into methacrylate and stained with the lipophilic dye DiI, an approach that clearly delineates fiber cell membranes, and counterstained them with DAPI to highlight the presence of nuclei and remnants of degenerating nuclei (Fig. 8 , small arrows). The last nucleated cells were found in the 45- to 50-cell range in the DiI-stained sections, compared with age-matched SEM sections in which cell counts could also be made, but nuclei are generally not evident. Defining precisely when the paddlelike structures emerged is somewhat difficult, because they emerged over the span of approximately 10 cells, but the emergence appeared to begin just as the last fiber cells were losing nuclei. The development of these structures continues in the anucleate cell. We also noted that the columns of paddles were evident in the DiI sections when the plane of section was grazing through such columns (Fig. 8 ; larger arrows), providing a different mechanism of visualization. 

Discussion

Many outstanding ultrastructural studies have been conducted in the lens (see for example Refs. ⁷ , ¹¹ , ¹² , ^{27 28 29 30 31} ). These studies have revealed a level of structural complexity that is quite striking and one that changes continuously across the age spectrum of fiber cells. The best studied of these changes, the enormous elongation, and the emergence of the BS, occurs before the loss of organelles, though there is abundant evidence of continued structural change subsequent to this point. ^{27 32 33} We focused on one set of structural specializations that have been noted, but not systematically described in the lens (see, for example, Refs. ⁵ , ¹⁴ , ²⁴ , ³⁴ , ³⁵ ). These structures are notable for both the magnitude and complexity of the of the structural change, the degree to which this change in one cell must be coordinated with cells on all sides, their differential expression during development, the timing of their emergence in lens fiber cell differentiation, and the asymmetry with which they are expressed in a given fiber cell. 

The function of these structures remains to be established. The hypothesis that the majority of light scatter in the normal lens results from the interaction of light with the fiber cell membrane ^{36 37} suggests that these elaborations may be detrimental to optical quality. This may be true and may well explain why these structures are absent from the more central regions of the lens and are not expressed in those first generations of fiber cells that will occupy the central lens through the life of the animal. 

The pattern of developmental expression of the paddlelike structures, specifically that cells formed early in development execute a program of structural differentiation that differs from that executed by cells appearing later in development, is not a new concept in lens biology (see, for example, studies on δ-crystallin expression ³⁸ or γS-crystallin expression ³⁹ ), though what regulates this program is unknown. Similarly, structural and organizational differences have been noted between the primary fiber cells formed in the first few days of embryonic lens development, and secondary fiber cells. ³³ Specifically, the primary fibers are more irregular in shape and more poorly organized. However, the structures that we observed did not emerge until much later (∼P21) in development than the primary fiber cells. By this time, many generations of secondary fiber cells had formed, exhibiting the uniform, flattened hexagonal profile and long-range order that characterizes secondary fibers. Only after this pattern of secondary fiber cell formation was long established did new fiber cells begin to form these paddlelike structures as part of their differentiation program. To our knowledge, this is the first characterization of such a dramatic difference in the structural program executed by early and late secondary lens fibers. 

Fiber cells lose their organelles abruptly at approximately 50 cells from the equatorial surface, a loss that suggests an end to protein synthesis at that time. ^{7 9 11 40} The absence of protein synthesis, though difficult to establish experimentally, is supported by experiments performed in chick lenses by Shestopalov and Bassnett. ⁴¹  

Comparison of protein profiles between nuclear and cortical lens samples, as well as analysis of a series of lens fractions harvested sequentially from cortex to nucleus, ⁴² reveal a widespread and progressive degradation in the overall protein profile. The changes that occur in the organelle-free cell have been reasonably characterized as “aging,” an unavoidable consequence of the loss of the repair and renewal mechanisms that cease to function at the loss of membranous organelles. Thus, it is somewhat surprising to find that most, if not all, of the rather dramatic structural differentiation described here occurs after the fiber cells have lost their organelles. Although progressive changes in fiber cell architecture have been described well into the nucleus, ^{12 13 27 29 30 32} these in general have been much more gradual, and presumed to represent the slow decline that results from a lack of synthetic ability. The structures described herein are differentially expressed in development, are asymmetric with respect to both the long and short axes of the fiber cell, and assemble in a well-orchestrated manner exhibiting regularity in distribution along the fiber cell, and between fiber cells. The complexity of these structures, as well as their differential expression and asymmetry in the fiber cell, suggests that a process other than a degenerative one may be unfolding. This notion raises the intriguing possibility that at least some of the changes in older fiber cells may represent a preprogramed series of changes, perhaps mediated by the posttranslational modifications assumed to be “degradation” or aging. Evidence of this possibility remains to be gathered. More recent studies suggest that there may be a macromolecular diffusion pathway between the deeper fiber cells, thus creating the possibility that newly synthesized proteins are present in these cells. ^{43 44}  

The complexity of the lateral membranes, deriving from both the BS and the paddle-like structures has consequences for the interpretation of studies that use substrates such as frozen sections. Such studies examine a thin piece of tissue, which nonetheless has a meaningful “volume.” Specifically, the amount of membrane located on the “short side” of the fiber cell cross section is much greater than that on the broader or longer side in the volume of a tissue section. For studies such as immunocytochemistry, this imbalance could lead to the mistaken conclusion that there is more antigen per unit membrane on the short side than on the long side, when in fact there may be the same level of antigen per square area, but simply more total membrane in the volume examined. 

Finally, the demonstration that routine immersion fixation protocols result in sharply different rates and degrees of fixation that parallel the fiber cell maturation gradient has the potential to produce misleading results as well, with cells at the periphery experiencing a much more rapid and thorough fixation and preservation than cells deeper in the cortex or nucleus. 

 Supported by National Eye Institute Grants EY08747 (PGF) and P30 EY012576. Work was conducted in a facility constructed with support from Research Facilities Improvement Program Grant Number C06 RR-12088-01 from the National Center for Research Resources, National Institutes of Health.

 Submitted for publication January 29, 2007; revised March 5, 2007; accepted May 16, 2007.

 Disclosure: T. Blankenship, None; L. Bradshaw, None; B. Shibata, None; P. FitzGeral d, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Autoradiograms of ¹⁴C-formaldehyde-fixed tissues. Whole eyes were fixed as described, with fixation times of (a) 1 hour. Arrows: margin of the lens, a, and of the retina-choroid-scleral complex, b; (b) 2, (c) 4 , and (d) 24 hours. (e) Autoradiogram of an isolated lens, split after 30 minutes of fixation, then allowed to fix another 2 hours. Sections were taken from the mid region of this half lens.

Figure 2.

 

Scanning electron micrographs of a 3-month lens. (a) Overview. N, nucleus; bow region is located at the top. (b) Higher magnification of boxed areain (a). Arrow: the point at which the paddlelike structures became evident in this image. (c) Higher magnification of the boxed area in (b). Solid arrow: zone of transition from fiber cells with a relatively simple profile to those bearing the paddlelike structures. Hollow arrows: nuclei of elongating fiber cells. Bar: (a) 500 μm; (b) 100 μm; (c) 50 μm.

Figure 3.

 

Scanning electron micrograph of 3-month lens. Longitudinal view of a small region of young fiber cells. The free lateral edge of cells shows clearly the well-developed BS occurring in two parallel rows on the lateral edge of the fiber cell. The far side of that same cell (arrow) shows the complimentary nature of the BS interdigitations. (b) A single fiber cell bearing the paddle-like specializations artifactually separated from the surrounding fiber cells, revealing clearly the nature of the paddlelike structures on the lateral edges of the fiber cell. Each paddle is, in turn, covered with the well-characterized BS. (c) Higher magnification showing the complexity of the lateral interactions between cells bearing the paddlelike structures. Small box: smaller scale interdigitations of the BS; large box: larger scale interdigitations of the paddles. Bar: (a) 5 μm; (b) 10 μm; (c) 3 μm.

Figure 4.

 

Scanning electron micrograph of 3-month (a) and 6-month (b) lenses. The degree of coordination in the positioning of the paddles on the fiber cell’s lateral edges is shown. The tissue fracture plane in (a) shows a lateral view of a stack of approximately 40 fiber cells, and the high degree of synchrony in the location of the paddles from cell to cell. (b) A lateral view of approximately 40 fiber cells, displaying how the columns formed by the paddles tend to merge and diverge, with some wandering over longer distances (arrows). Bar: 10 μm; (b) 20 μm.

Figure 5.

 

SEM of 2-week lens. (a) Low-magnification overview that spans from nucleus (N) to bow. No evidence of the columns that are indicative of paddle structures can be seen anywhere in the field. (b) Medium-magnification view that includes the 40- to 50-cell region in which the paddles are very evident in the 3-month lens. (c) Higher-magnification view of 2-week lens. While lateral protrusions are evident, the paddle-like structures are not yet present. Bar: (a) 200 μm; (b) 100 μm; (c) 50 μm.

Figure 6.

 

(a) SEM of the cortical region of a P21 lens. At this age, the first evidence of paddle formation can be seen, evidenced by the columns indicated by the arrow. (b) Higher-magnification view of the region where the paddles are just beginning to emerge. Arrows: two nascent, barely evident columns. Bar: (a) 100 μm; (b) 20 μm.

Figure 7.

 

Scanning electron microscope overview of 8-month lens. Arrow: the columns formed by stacked paddles. Black vertical line: relative radius of the P14 lens. Bar, 500 μm.

Figure 8.

 

Methacrylate section of 3-month lens, stained with DiI (red) to highlight membranes and DAPI (blue) to highlight nuclei. Small arrows: Nuclear remnants; large arrows: area where scanning electron microscopy shows evidence of the columns. Bar, 50 μm.