Fiber Cell Denucleation in the Primate Lens
Steven Bassnett
Purpose. To determine the morphologic and biochemical events preceding the breakdown of
fiber cell nuclei in the primate lens.
Methods. Monkey lens slices were labeled with fluorescent probes and optically sectioned using
a confocal microscope. The distribution of nuclear histones was visualized by immunofluorescence. DNA and cellular membranes were imaged simultaneously by staining with SYTO 17
and 3,3'-dihexyloxacarbocyanine iodide, respectively. The condition of fiber cell DNA during
differentiation was determined by an in situ DNA fragmentation assay. The assay was adapted
to allow the detection of DNA fragments with 3'-OH or 3'-PO4 termini.
Results. Monkey lens fiber nuclei passed through distinct stages before disintegrating. In the
outer cell layers, the nuclei were large, smooth, and oval-shaped with prominent nucleoli.
Deeper in the lens, they had a flattened profile with whorls of membranous material and
nucleic acid accumulated at one end. At this point, histone immunofluorescence was reduced
and the nucleoli had a characteristic, spoked appearance. At the border of the organellefree zone, the intracellular membranes (including the nuclear envelope) disappeared, and
particulate material was released from the nuclei into the cytoplasm. This material was stained
by SYTO-17 and the DNA fragmentation assay, indicating that it contained fragmented DNA
with 3'-OH termini.
Conclusions. The denucleation process in the primate lens differs from that described recently
in the embryonic chicken lens. In particular, the extrusion of nuclear material and persistence
of DNA-rich particles in the fiber cytoplasm are novel features. One similarity between the
denucleation process in these species is the appearance of 3'-OH ends in the DNA after the
loss of the nuclear membrane. Invest Ophthalmol Vis Sci. 1997; 38:1678-1687.
.Lens fiber cell differentiation is characterized by the
sudden, coincident loss of cytoplasmic organelles, including nuclei. 1 In the chicken lens, this process begins in the central fiber cells at approximately day 12
of development. Thereafter, a wave of organelle loss
sweeps outward across the more superficial cells. The
conditions that trigger or propagate this wave are
poorly understood, as is the biochemistry of organelle
disassembly. By the time of hatching, only the superficial fiber cells contain organelles, and these are largely
excluded from the light path, lying in the shadow of
the iris. The region of the lens composed of fiber cells
From the Cataract Research Center, Department of Ophthalmology and Visual
Sciences, St. Louis, Missouri.
Supported by National Institutes of Health grants RO1 EY09852 (SB) and
EY02687 (Core Grant for Vision Research) and by an unrestricted grant from
Research to Prevent Blindness, Inc.
Submitted November 4, 1996; revised February 4, 1997; accepted March 12, 1997.
I'roprielary interest category: N.
Reprint requests: Steven Bassnett, Cataract Research Center, Department of
Ophthalmology and Visual Sciences, Campus Box 8096, 660 South Euclid Avenue,
Washington University School of Medicine, St. Louis, MO 63110-1093.
1678
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
that have lost their organelles is easily visualized in
the chicken and monkey lenses 1 " 3 and has been
termed the organelle-free zone (OFZ) .4 Recently, we
have focused our attention on the biochemical and
morphologic transformations that chicken lens fiber
nuclei undergo at the border of the OFZ.4 Just before
nuclear degradation, in the chicken lens, the nuclei
lose their regular leaf-like appearance and become
irregular in profile with marginalized chromatin. Subsequendy, the nuclei collapse, holes appear in the nuclear lamina, and the nuclear envelope disintegrates.
At this point, the naked chromatin appears to be degraded by endogenous nucleases, causing fragmentation of the DNA and the concomitant appearance of
3'-OH DNA termini. Nuclear debris, containing degraded DNA, lamin B2, and histone proteins, persists
deep into the OFZ. A number of nuclease activities
have been identified in lens extracts, 5 " 13 and some of
these may play important roles in the denucleation
process.
The denucleation process has been intensively
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
Copyright © Association for Research in Vision and Ophthalmology
1679
Fiber Cell Denucleation in the Primate Lens
studied in the embryonic chicken lens, but it is important to determine whether a similar sequence of
events occurs during lens fiber cell differentiation in
higher mammals. In the current study, we have extended our earlier work on the chicken lens to the
monkey lens and find that, although it differs in some
respects, there are also fundamental similarities in the
denucleation process in these species.
METHODS
Tissue Preparation
All procedures used in this study complied with the
ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Whole globes were dissected
from young adult rhesus monkeys that had been killed
for vaccine testing. The globes were wrapped in gauze,
packed on wet ice, and delivered by overnight express
mail to the laboratory. On arrival, the globes were
opened by making an incision through the sclera adjacent to the optic nerve. The incision was extended
forward and around the circumference of the eyeball,
allowing the posterior globe to be removed and
exposing the back of the lens and attached vitreous
humor. Fine forceps were used to peel the ciliary epithelium from the overlying choroid and sclera,
thereby freeing the lens from the eye. Each lens, with
the ciliary epithelium and vitreous humor still
attached, was fixed for 30 minutes in 4% paraformaldehyde-phosphate-buffered saline (PBS, pH 7.4) at
room temperature and then bisected along the optic
axis. The two hemispheres were bisected, dividing the
lens into quarters. These were returned to the fixative
for a further 30 minutes. Because only those fiber cells
near the surface of the lens contain organelles, the
core of the lens (including the embryonic, fetal, and
adult nuclei) was dissected from each lens quadrant
and discarded. The hollow lens quadrants were returned to the fixative for 1 hour. The quadrants were
then embedded in 4% agar-PBS and sliced at 200 /zm
on a Vibratome (TPI; St. Louis, MO).
Immunofluorescence
Immunofluorescence was performed as described previously.3'4 Briefly, lens slices were permeabilized with
0.1% Triton X-100-PBS, blocked with goat serum,
and incubated with a 1:500 dilution of primary antibody overnight at 4°C. We used a monoclonal antibody that recognizes an epitope present in histone
HI, H2A, H2B, H3 and H4 (clone H11-4P; Boehringer
Mannheim, Indianapolis, IN). After a wash in PBS,
slices were incubated for 2 hours in fluorescein-conjugated goat antimouse immunoglobulin G antibody diluted 1:500 (Jackson Immunoresearch, West Grove,
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
PA). Slices were washed in PBS, mounted with a coverslip, and viewed.
DNA and Membrane Staining
DNA and cellular membranes were colocalized by
washing fixed lens slices in Tris-buffered saline (TBS;
10 mM Tris, 150 mM NaCl; pH 7.4) for 5 minutes and
incubating the slices for 40 minutes in TBS containing
1 ^M SYTO 17 (Molecular Probes, Eugene, OR) and
1 //g/ml 3,3'-dihexyloxacarbocyanine iodide (DiOCfi;
Molecular Probes) to stain DNA and cellular membranes, respectively.4 SYTO dyes are a family of cell
permeant nucleic acid stains, excited in the visible
spectrum, with comparable affinities for DNA and
RNA. Nuclear staining properties of SYTO dyes have
recently been used to differentiate apoptotic from
nonapoptotic thymocytes.14 SYTO 17 has spectral
properties similar to the Texas red fluorophore and
can be discriminated from the fluorescein-like DiOC 6
fluorescence in dual-labeled preparations.
After staining, the lens slices were washed in TBS
for 30 minutes and viewed with a confocal microscope
using the 488-nm laser line and a 515- to 565-nm bandpass filter for DiOC 6 fluorescence and the 568-nm
laser line and a 590-nm long-pass filter for SYTO 17
fluorescence.
In Situ Terminal Deoxynucleotidyl Transferase
Labeling
Lens slices were permeabilized for 30 minutes in 0.1%
Triton X-100-PBS at room temperature and then
washed for 20 minutes in PBS. The terminal deoxynucleotidyl transferase (TdT)-labeling reaction was
performed, using an in situ cell-death detection kit
(Boehringer Mannheim) according to the manufacturer's instructions, except that a washing step (gentle
agitation in 0.1% sodium dodecyl sulfate solution at
60°C for 30 minutes) was included after the labeling
reaction to remove unincorporated fluorescent nucleotides. 4 In the TdT assay, fluorescein-labeled nucleotides are incorporated at the 3'-OH ends of DNA molecules by the action of the enzyme TdT. If cell nuclei
are labeled by this assay, it indicates that the nuclear
DNA contains significant numbers of 3'-OH ends. In
each experiment, some lens slices were incubated in
labeling mixture lacking TdT, as a negative control.
Permeabilized slices were also sometimes pretreated with deoxyribonuclease I (DNase I, 50 U/ml
in 10 mM Tris, pH 7.6; Boehringer Mannheim) for
30 minutes at 37°C, as a positive control. DNase I is
an endonuclease that cleaves DNA into fragments with
3'-OH ends. Because there is evidence for the presence of DNase II-like nuclease activity in the chicken
lens, 13 we pretreated some lens slices with calf intestinal alkaline phosphatase (CIAP; Gibco, Bethesda,
MD). The standard TdT procedure only labels DNA
1680
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
l. Confocal micrographs of lens fiber cells from the bow region of the adult monkey
lens showing the distribution of nucleic acids (A,C) and cellular membranes (B,D). Lens
slices were stained with SYTO 17 (to stain nucleic acids) and DiOCg (to stain membranes)
and are viewed with a confocal microscope (see Methods). Fiber cell nuclei appear to pass
through three distinct stages (labeled I to III in A) before disappearing at the border of
the OFZ. Stage I fiber cell nuclei are smooth and oval shaped with one or more prominent
nucleoli. Stage II fiber cells are more elongated with accumulations of membrane and
extruded chromatin at one end of the nucleus. Stage III fiber cell nuclei are present at the
border of the OFZ. These nuclei still contain DNA and are bounded by an intact nuclear
envelope. In a high-magnification view of stage II lens fiber cell nuclei, multilamellar structures are indicated by arrows (C,D) and are positively stained by SYTO 17 (C) and DiOCr,
(D), indicating that they are comprised of nucleic acid and membranous material, respectively. DiOQi = 3,3'-dihexyloxacarbocyanine iodide; OFZ = organelle-free zone; Ep = lens
epithelium. Scale bars (A,B) = 25 ^m; (C,D) = 10 fim.
FIGURE
fragments that have 3'-OH termini. A DNase II-like
enzyme generates 3'-PCX, termini15 that would not be
detected by the standard assay. However, pretreatment
of lens slices with CIAP converts 3'-PO4 termini to 3'OH termini, with the latter detectable by the TdT
assay.4 Treatment with CIAP was carried out on permeabilized slices incubated for 1 hour at 50°C in 100 /zl of
dephosphorylation buffer (supplied with the enzyme)
containing 1 //I (26 U) of CIAP. As a positive control
for the CIAP treatment, some lens slices were first
incubated for 30 minutes at 37°C with micrococcal
nuclease (Sigma, St. Louis, MO), an enzyme that produces DNA fragments with 3'-PO^ termini. Slices were
viewed on a confocal microscope, using the 488-nm
laser line and a 515-nm long-pass filter.
equipped with an argon—krypton laser. In some experiments, a series of optical sections were obtained at
different focal planes within the lens slice and vised
to generate three-dimensional stereo images of the
tissue. For this purpose, 150 optical sections were collected at intervals of 0.26 [im in the z-axis. The images
were obtained using a X63 oil-immersion planapochromat objective lens {Carl Zeiss; numeric aperture =
1.4) and a confocal pinhole setting of 21. This combination of objective lens and pinhole aperture gave a
calculated z-axis resolution of 0.8 //m. The final stereo
projections were created using three-dimensional visualization software supplied with the confocal microscope and an angular separation of 5° for the projected image stacks.
Confocal Microscopy
Slices were viewed on a laser scanning microscope
(Model LSM410; Carl Zeiss, Thornwood, NY)
RESULTS
On arrival at the laboratory, the monkey lenses were
completely transparent, with a faint, yellow nuclear
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
1681
Fiber Cell Denucleation in the Primate Lens
FIGURE 2. Confocal micrographs of stage II fiber cell nuclei stained with SYTO 17. Stage II
nuclei have prominent nucleoli with spokes of less-intense fluorescence. (A) These dark
spokes give the nucleoli a "sand dollar" appearance. At the ends of the nuclei, extruded
DNA (A,B, airoius) can have (A) tangled, (B) looped or, lamellar configurations, N =
nucleoli. Scale bars = 5 /xm.
coloration. Slices of the bow region of the lens revealed a thin layer (~100 cells deep) of nucleated
fiber cells. Fiber cells deeper than this had lost their
nuclei and other organelles2 and, thus, had formed
part of the OFZ (Fig. 1A, IB). Lens slices were incubated in DiOCc and SYTO 17 and viewed with a confocal microscope (Fig. 1). The SYTO 17 fluorescence
highlighted the nuclei of epithelial and fiber cells
(Fig. 1A). The nucleoli were particularly prominent
in the cortical fiber cells. There was also a faint background staining of the cytoplasm of epithelial and
fiber cells with SYTO 17, which may have been caused
by the presence of mitochondrial DNA and cytoplasmic RNAs or by nonspecific binding to crystallin
proteins. The background with both probes notably
increased in the deeper layers of the lens. The DiOC6
fluorescence was particularly strong in the epithelial
cells (Fig. IB), which have the highest concentration
of intracellular organelles.3 In the fiber cells, DiOC6
stained the nuclear membranes and cytoplasmic structures, presumably elements of the endoplasmic reticulum (ER) or mitochondria. A particularly striking feature of the DiOCfi fluorescence was the presence of
whorls of membrane adjacent to the elongated nuclei
of fiber cells in the deep bow region (Fig. IB). On
the basis of the SYTO 17 and DiOC6 staining patterns,
three distinct stages (I, II, and III) in nuclear morphology could be discriminated (Figs. 1A, IB). Stage I nuclei were found in the most superficial layers of the
monkey lens cortex. These nuclei were oval, when
viewed in the midsagittal plane, with one or more
large, prominent nucleoli, and a well-defined nuclear
envelope. Stage II nuclei were found in fiber cells, 30
or more cells deep in the cortex. These nuclei had a
much thinner cross section when viewed in the midsagittal plane, although their axial length was similar
to that of stage I nuclei. Stage II nuclei usually had
whorls of DiOC6-stained membranous material closely
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
associated with one end of the nucleus. Stage III nuclei
were located between the stage II nuclei and the border of the OFZ. These nuclei were smaller than the
stage II nuclei; they lacked the associated membranous material, and their nucleoli were absent or disorganized. A few cells deeper into the cortex than the
stage III nuclei lies the boundary of the OFZ. At this
point, the nuclei disintegrated, with the concomitant
disappearance of all other cytoplasmic organelles.
A striking feature of stage II nuclei was the presence of multilamellar membrane whorls at one end of
the nuclei (Figs. 1C, ID). Interestingly, die membrane
whorls appeared to be contiguous with the nuclear
envelope. SYTO 17 stained the contents of the nucleus
proper and the membranous structures with similar
intensity, suggesting that they probably contain nucleic acid of nuclear origin. The configuration of the
membranous structures of stage II fiber nuclei was
quite variable (further examples, Fig. 2). The SYTO
17 fluorescence of stage II nuclei highlighted extruded DNA with a range of morphologies including
"tangled" profiles (Fig. 2A) and simple loops and
whorls (Fig. 2B). These images also reveal another
interesting feature of nuclear substructure. In stage
II nuclei, the nucleoli often contained spokes of less
intense fluorescence, resulting in a "sand dollar" appearance (Fig. 2A) similar to that seen previously at
the electron microscope level.lb
At the border of the OFZ, the chromatin of stage
III nuclei underwent dramatic transformations. The
SYTO 17 fluorescence revealed that the nucleic acids
(presumably DNA) became heterogeneously dispersed within the nucleus. Stage III fiber cells still
contained abundant organelles (evidenced by the
DiOCo staining of discrete structures in the fiber cell
cytoplasm), and an intact nuclear membrane (Fig.
3A). One or two cells deeper into the lens, however,
the fiber cell cytoplasm was largely featureless, al-
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
1682
FIGURE 3. Confocal micrographs of stage III nuclei stained with SYTO 17 (A) and DiOCfi
(B). Stage 111 nuclei are found immediately outside the border of the OFZ. They have an
intact nuclear membrane (B, anoru) and contain heterogeneously distributed chromatin.
DiOCc = 3,3'-dihexyloxacarbocyanine iodide; OFZ = organelle-free zone. Scale bars = 10
though small foci of SYTO 17 fluorescence are discernible, scattered throughout the fiber cell cytoplasm. These structures are not stained by DiOCG, and
are presumably not membrane bound.
It was extremely rare to find a nucleus that had
been fixed at the instant of disintegration, suggesting
that this final step may be relatively rapid. However,
one such nucleus is shown in Figure 4. Two or three
cell widths deeper than the last membrane-bound
stage III nucleus is a nuclear remnant. In this cell,
the nuclear envelope has collapsed, with disorganized
remnants collecting at one end of the nuclear space.
The chromatin has disintegrated into a cloud of particles, some of which have remained within the original
nuclear space, with others scattered throughout the
fiber cell cytoplasm. The SYTO 17-stained material in
the nuclear space may represent the remains of the
nucleolus.
At the time of organelle loss, there was a concomitant change in the physical and chemical nature of
the cytoplasm, such that the background staining with
both fluorescent probes was increased. Despite this,
however, it was possible to observe the fate of the
particles generated at the time of nuclear disintegration in fiber cells of the deep cortex of the lens (Fig.
5). Immediately inside the OFZ, the fiber cell cytoplasm contained scattered SYTO 17-stained particles.
These particles appeared to be randomly organized
and were not closely associated with the fiber cell
plasma membranes. The original position of the fiber
cell nuclei could be discerned as a reduction in the
cytoplasmic background fluorescence for both fluorescent probes, suggesting that cytoplasmic proteins
had not invaded the vacant nuclear space. Many of
these nuclear "ghosts" contained small, irregular bodies that were stained by SYTO 17. These were probably
FIGURE 4. Confocal micrographs of a disintegrating nucleus at the border of the OFZ stained
with SYTO 17 (A) and DiOC6 (B). The disintegrating nucleus (A,B> arrozos) is situated
immediately inside the OFZ, just two or three cell layers deeper than the nearby, intact
stage III nucleus. As the nucleus disintegrates, clusters of particles can be seen in the space
previously occupied by the chromatin. Particles are also visible in the cytoplasm (arroxoheads),
The remnants of the nuclear membrane have collapsed and collected at one end of the
nucleus (B, arrow). DiOC0 = 3,3'-dihexyloxacarbocyanine iodide; OFZ = organelle-free
zone. Scale bars = 5 //m.
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
1683
Fiber Cell Denucleation in the Primate Lens
FIGURES. Confocal micrographs of fiber cells in the deep cortex stained with SYTO 17 (A,C)
and DiOCc (B,D). (A,B) Approximately 50 cells inside the border of the OFZ. Small nucleic
acid-containing particles are scattered throughout the fiber cell cytoplasm (A, arrowheads).
The original position of the fiber cell nuclei can be discerned by a reduction in the background fluorescence. These nuclear "ghosts" often contain larger particulate structures (A,
arrow) that by virtue of their size and position probably represent remnants of the nucleoli.
Neither the cytoplasmic nor nuclear particles are stained by DiOCr,, indicating that they are
not membrane bound. (C,D) Several hundred cell layers within the OFZ, the particles finally
disappear. DiOCfi = 3,3'-dihexyloxacarbocyanine iodide; OFZ = organelle-free zone. Scale
bars = 10 /mi.
remnants of nucleoli. The nuclear ghosts and cytoplasmic particles persisted for many hundreds of cell
widths into the cortex before eventually disappearing
(Fig. 5C, 5D), leaving behind a featureless cytoplasm.
It was also of interest to determine the fate of the
major proteinaceous component of the chromatin,
the histones. We used a monoclonal antibody raised
against a shared epitope of the histone protein family
for immunofluorescence studies. Figure 6 shows a
stereo pair of images of histone immunofluorescence
in the bow region of a monkey lens slice. The image
stack was collected through a tissue depth of approximately 40 jum. The positions of stages I, II, and III
nuclei are indicated. The histone immunofluorescence was strongest in the superficial layers of the
lens, in stage I nuclei. Thereafter, the fluorescence
appeared to decline rapidly such that, in stage III nuclei, near the border of the OFZ, histone immunofluorescence was no longer detectable. A few, small, positively stained structures were often observed in the
cytoplasm of stage II and stage III fiber cells (Fig. 7).
These were distinct from the smaller and far more
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
6. Three-dimensional stereo images of the bow region of a lens slice after immunofluorescent localization of
histone proteins. This stereo pair was collected through a
depth of approximately 40 /j.m and constructed from a stack
of 150 individual optical sections. Note that histone immunofluorescence is strongest near the surface of the lens in
stage I nuclei. In the deeper cells, the fluorescence is markedly reduced and is completely undetectable in cells near
the border of the organelle-free zone. Positively stained bodies are evident in the cytoplasm of stage I fiber cells (arroxo).
Scale bar = 25 fj,m.
FIGURE
1684
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
FIGURE 7. Confocal micrographs of TdT-labeled lens slices. (A) Nuclear remnants {arrow)
and scattered cytoplasmic particles are labeled by the TdT assay, indicating the presence of
fragmented DNA with free 3'-OH ends. Note that all the labeled structures lie within the
OFZ. (B) Fragmented DNA with 3'-PO<i ends can also be detected, using a modified TdT
assay. This lens slice was incubated in micrococcal nuclease to produce 3'-PO4 ends and
treated with alkaline phosphatase to convert the 3'-PO4 to 3'-OH. The slice was then labeled
using the standard TdT assay. Note that after this treatment, all nuclei are labeled, demonstrating the efficacy of the technique. Stage I nuclei of newly differentiatedfibers{arrowheads)
and nuclear remnants in the OFZ {arrows) are labeled, although the latter are less strongly
labeled than with the conventional TdT technique (compare with Fig. 7A). (C) If slices
were treated with alkaline phosphatase before TdT labeling, the labeling pattern was indistinguishable from the conventional TdT assay (compare with Fig. 7A). In both cases, labeling
was restricted to the OFZ, in which nuclear remnants {arrow) and scattered cytoplasmic
particles were labeled. That alkaline phosphatase treatment did not result in any additional
labeling of stage I, II and III nuclei suggests that DNA fragments with 3'-PO4 ends do not
accumulate during nuclear degradation. (D) Shows a comparable region to that of (A) and
(C). In this case, however, the lens slice was incubated in a TdT-labeling mixture from which
the TdT enzyme was omitted. Under these conditions, no labeled structures were evident,
demonstrating the specificity of the labeling reaction. I, II, III, = position of stage I, II, and
III nuclei; TdT = terminal deoxynucleotidyl transferase; OFZ = organelle-free zone. Scale
bars (A,B) = 50 /zm; (C,D) = 25 /mi.
numerous SYTO 17-stained particles found in the cytoplasm of fiber cells within the OFZ.
The condition of the chromatin was also visualized
during fiber cell differentiation, by a TdT-labeling
assay. When slices were pretreated with DNase I, to
induce DNA fragmentation, all nuclei were labeled
(data not shown), indicating that the primate lens
does not contain any endogenous inhibitors of the
labeling reaction. If untreated lens slices were used in
the TdT-labeling assay, strongly fluorescent particles
of degraded DNA were detected at the border of the
OFZ and in deeper layers. There was no detectable
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
labeling of stage I, II, or III nuclei. Most of the labeling
in the OFZ consisted of small particles scattered
throughout the fiber cell cytoplasm but, in cells immediately within the OFZ, some larger structures were
labeled (Fig. 7A). These may correspond to the nucleolar remnants seen in the SYTO 17-stained preparations. The TdT-labeled particles were readily identified in the cytoplasm of fiber cells several hundreds
of cell layers within the OFZ (although this is not
apparent in Figure 7A, because of the orientation of
the optical section). Reports of recent studies have
suggested that a DNase Il-like enzyme activity may play
Fiber Cell Denucleation in the Primate Lens
a role in the degradation of fiber cell chromatin.13
DNA degradation related to the activity of such an
enzyme would not be detected by a standard TdTlabeling assay, because this only labels DNA fragments
that have 3'-OH ends. DNase II digestion of DNA results in DNA fragments with 3'-PO4 ends. To circumvent this difficulty, we pretreated monkey lens slices
with CIAP to convert any 3'-PO4 that might have been
present into 3'-OH. As a positive control for this treatment, we first incubated some lens slices in micrococcal nuclease (an enzyme known to produce DNA fragments with 3'-PO4 ends). Figure 7B shows the equatorial region of a lens slice that was treated with
micrococcal nuclease and CIAP before TdT labeling.
Under these circumstances, all nuclei (stages I, II, and
III) were labeled, as were the nuclear debris within the
OFZ. Interestingly, the positively labeled cytoplasmic
particles in the OFZ were less intensely fluorescent
after micrococcal nuclease-CIAP treatment than they
were in untreated lens slices. This may indicate that
the partially degraded DNA of the cytoplasmic particles was particularly vulnerable to nuclease treatment.
Overdigestion with micrococcal nuclease completely
eliminated the labeling of the cytoplasmic particles
(data not shown). Stage I, II, and III nuclei showed
variations in labeling intensity after micrococcal
nuclease pretreatment (with the stage I nuclei the
most intensely labeled). Whether this reflected overdigestion and subsequent loss of DNA from the stage
II and III nuclei or a reduced labeling efficiency in
those nuclei is not clear. If the lens slices were treated
with CIAP without micrococcal pretreatment, only cytoplasmic particles within the OFZ were labeled (Fig.
7C). Under these conditions, the labeling was indistinguishable from that of non-CIAP-treated lenses. The
apparent increase in abundance of cytoplasmic particles in Figure 7C is caused by the greater magnification of this image. The specificity of the labeling reaction was confirmed by performing the labeling assay in
the absence of the TdT enzyme. Omitting the enzyme
abolished the cytoplasmic labeling completely (Fig.
7D).
DISCUSSION
There have been many morphologic studies on lens
fiber cell differentiation, beginning with the pioneering work of Rabl at the turn of the century.17 Electron
microscopists, in particular, have described the differentiation process in great detail, focusing their attention on the denucleation process in primary and secondary fiber cells.1618"20 In the mammalian lens, Kuwabara and Imaizumi have suggested that the contents
of the secondary fiber cell nuclei become gradually
indistinguishable in staining characteristics from the
surrounding cytoplasm and that, after vesiculation of
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
1685
the nuclear membrane, nucleolar remnants are released into the cytoplasm.16 During denucleation of
primary fiber cells in mice, osmiophilic particles appear in the nucleus and in the cytoplasm of the fiber
cells.20 These particles may represent nucleosomes
and appear to be physically extruded into the extracellular space of the lens where they may be further degraded.20 In the current study, we showed that the
denucleation process in the primate lens may include
elements of both of these models. Histone protein
immunofluorescence is markedly reduced in the stage
II and III nuclei, perhaps explaining the gradual "fading" of the nucleus observed by Kuwabara and Imaizumi in EM preparations after heavy metal staining.16
At the border of the OFZ, we observed the sudden
loss of the nuclear membrane and the release into
the cytoplasm of a cloud of DNA-containing particles.
Particles were detected using either the SYTO 17- or
the TdT-labeling assay. Both assays identified similar
numbers of particles broadly dispersed throughout
the cytoplasm. Thus, although we did not double-label
these preparations, it seems reasonable to conclude
that the two labeling techniques identified a single
population of particles containing degraded fragments of DNA with free 3'-OH ends. At the level of
resolution of the confocal microscope, it was not possible to determine whether these particles represented
the putative nucleosomes seen by others.20
We assume that the reduction in histone immunofluorescence in stage II and III nuclei reflects the loss
of these proteins from the fiber cell nuclei during
differentiation. We also observed small, histone-positive structures in the cytoplasm of fiber cells in this
region. It is tempting to speculate that the decrease
in nuclear immunofluorescence is directly related to
the appearance of these cytoplasmic structures. If the
decrease in immunofluorescence is related to the loss
of histones from the maturing fiber nuclei, we would
expect those nuclei to be transcriptionally inactivated.
Interestingly, the position of the last transcriptionally
active fiber cell nucleus has not been precisely determined in any lens. However, it is also possible that the
decrease in nuclear histone immunofluorescence is
caused by posttranslational modification of the histone proteins. For example, acetylation, phosphorylation or other posttranslational modifications may
mask the epitope recognized by the monoclonal histone antibody and give the appearance of a reduced
nuclear histone content.
We found no evidence of an association between
the cytoplasmic particles and the fiber cell membranes, nor did we observe an accumulation of particles at the tips of the fibers or in the extracellular
space, as observed in the rodent lens.20 In the lens
slices used here, the particles were evenly distributed
within the cytoplasmic volume. This is a surprising
1686
Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9
observation because, if the particles represent aggregations of degraded chromatin that were released
from the centrally located nuclei, we might have expected the particles to be concentrated near the center of the fibers. The rate of diffusion of a large particle through the extremely viscous fiber cell cytoplasm
is likely to be very slow. The uniform distribution of
particles within the cytoplasm may be evidence therefore of a facilitated transport process for disseminating
the particles in the cytoplasmic volume.
One morphologic feature of the denucleation
process in the primate lens that appears to be unique
is the presence of whorls of membrane adjacent to
stage II nuclei. The SYTO 17 staining demonstrates
that the membrane whorls often encapsulate DNA
that appears to be extruded from the nucleus proper.
The DNA in these structures was not labeled by the
TdT assay, indicating that the DNA do contain abundant strand scissions. The structures were not observed in stage III fiber cells; they are presumably transient phenomena, the role of which remains obscure.
The identity of the nuclease enzyme (s) responsible for degrading the fiber cell DNA remains unknown, although several candidate activities have been
identified.5"13 In contrast with recent reports, we
found no evidence for the involvement of a DNase IIlike enzyme (that is, an endonuclease that cleaves
DNA into fragments with 3'-PO4 ends). Our data suggest the involvement of a DNase I-like activity in monkey lens fiber denucleation, leading to the production
of DNA fragments with 3'-OH ends. Results of recent
molecular studies suggest that the fiber cell endonuclease is probably not DNase I (EC 3-1-21-1) itself,
because the mRNA for this enzyme is not detectable
in the lens by reverse transcription-polymerase chain
reaction. 21
We have previously used the approaches employed here to study denucleation in the chicken lens
model, and it is useful to compare those data with
the current observations on the monkey lens. During
chicken lens fiber denucleation, rents appear in the
nuclear lamina and the nuclear envelope disintegrates. Subsequently, the nuclei collapse and the condensed naked chromatin (still containing lamin B2
and histone proteins) is digested by a DNase I-like
nuclease (s) that cleaves the DNA into fragments with
3'-OH ends. 4 The most striking difference between
the denucleation process in the chicken and primate
lenses is that, in the chicken, the process appears to
involve condensation of the nucleus. 412 A single nuclear remnant remains for several days in the cytoplasm of the central lens fibers before finally disappearing. In contrast, in the primate lens, the process
appears to be one of disintegration, in which particles
of nuclear DNA are scattered throughout the fiber
cell cytoplasm and persist for weeks or even months.
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
Whether these phenomenological differences reflect
important mechanistic differences in the denucleation process is unknown. It is also worth noting that
the chicken studies were performed on embryonic
lenses, whereas the current work was performed on
adult lenses. However, in both cases, the denucleation
of secondary fiber cells was examined.
During fiber denucleation in the chicken lens,
DNA degradation was only detectable in nuclei that
lacked intact nuclear membranes. 4 We have argued,
previously, that the nuclease (s) responsible may ordinarily be resident in the cytoplasm of the fiber cells,
only invading the nucleus after the integrity of the
nuclear membrane has been lost.4 The same argument
can be made about the primate lens, in which the
final degradation of the chromatin also appears to
occur in cells in which the nuclear membrane has
disintegrated. Another striking similarity between the
denucleation process in these species is the rapidity
with which overt DNA damage occurs. Although it is
possible that low levels of degraded DNA are present
in the superficial cells, it is only relatively late in the
denucleation process that significant numbers of 3'OH DNA ends accumulate. Thus, the action of lens
nucleases per se is unlikely to initiate fiber denucleation and the factors, that trigger and control this process remain to be determined.
Key Words
confocal microscope, DNA fragmentation, fiber differentiation, lens, terminal deoxynucleotidyl transferase
Acknowledgments
The author thanks Drs. Danijela Mataic and David C. Beebe,
of Washington University, for their helpful suggestions during the course of this work, and Dr. Sam Zigler of the National Eye Institute for providing the monkey lenses.
References
1. Bassnett S, Beebe DC. Coincident loss of mitochondria and nuclei during lens fiber cell differentiation.
DevDyn. 1992; 195:85-93.
2. Bassnett S. Mitochondrial dynamics in differentiating
fiber cells of the mammalian lens. Curr Eye Res.
1992;11:1227-1232.
3. Bassnett S. The fate of the Golgi apparatus and the
endoplasmic reticulum during lens fiber cell differentiation. Invest Ophthalmol Vis Sci. 1995; 36:1793-1803.
4. Bassnett S, Mataic D. Chromatin degradation in differentiating fiber cells of the eye lens. / Cell Biol. 1997;
137:37-49.
5. Arruti C, Chaudun E, De Maria A, Courtois Y, Counis
MF. Characterization of eye-lens DNases: Long term
persistence of activity in post apoptotic lens fiber cells.
Cell Death Differ. 1995; 2:47-56.
6. Chaudun E, Arruti C, Courtois Y, et al. DNA strand
breakage during physiological apoptosis of the embry-
1687
Fiber Cell Denucleation in the Primate Lens
7.
8.
9.
10.
11.
12.
onic chick lens: Free 3'OH end single strand breaks
do not accumulate even in the presence of a cationindependent deoxyribonuclease. / Cell Physiol. 1994;
158:354-364.
Counis MF Chaudun E, Allinquant B, et al. The lens:
A model for chromatin degradation studies in terminally differentiating cells. Int JBiochem. 1989; 21:235242.
Counis MF, Chaudun E, Courtois Y, Allinquant B.
Lens fiber differentiation correlated with activation of
two different DNAases in lens embryonic cells. Cell
Differ Dev. 1989;27:137-146.
Counis MF, Chaudun E, Courtois Y, Allinquant B.
DNAase activities in embryonic chicken lens: In epithelial cells or in differentiating fibers where chromatin is progressively cleaved. Biol Cell. 1991; 72:231-238.
Muel AS, Chaudun E, Courtois Y, Modak SP, Counis
MF. Nuclear endogenous Ca2+-dependent endodeoxyribonuclease in differentiating chick embryonic lens
fibers. / Cell Physiol. 1986; 127:167-174.
Muel AS, Laurent M, Chaudun E, et al. Increased
sensitivity of various genes to endogenous DNase activity in terminal differentiating chick lens fibers. Mutat
Res. 1989; 219:157-164.
Sanwal M, Muel AS, Chaudun E, Courtois Y, Counis
MF. Chromatin condensation and terminal differenti-
Downloaded From: http://iovs.arvojournals.org/ on 06/18/2017
13.
14.
15.
16.
17.
18.
19.
20.
21.
ation process in embryonic chicken lens in vivo and
in vitro. Exp Cell Res. 1986; 167:429-439.
Torriglia A, Chaudun E, Chany-Fournier F, Jeanny
JC, Courtois Y, Counis MF. Involvement of DNase II
in nuclear degeneration during lens cell differentiation. JBiol Chem. 1995; 270:28579-28585.
Frey T. Nucleic acid dyes for detection of apoptosis
in live cells. Cytometry. 1995;21:265-274.
Liao TH. The subunit structure and active site sequence of porcine spleen deoxyribonuclease. / Biol
Chem. 1985; 260:10708-10713.
Kuwabara T, Imaizumi M. Denucleation process of
the lens. Invest Ophthalmol. 1974; 13:973-981.
Rabl Cv Uber den bau und die entwicklung der linse.
Ill Teil: Die linse der saugethiere. Ruckblick und
schluss. ZwissZool. 1899;67:1-138.
Kuwabara T. The maturation of the lens cell: a morphologic study. Exp Eye Res. 1975; 20:427-443.
Srinivasan BD, Iwamoto T. Electron microscopy of
rabbit lens nucleoli. Exp Eye Res. 1973; 16:9-18.
Vrensen GFJM, Graw J, De Wolf A. Nuclear breakdown during terminal differentiation of primary lens
fibres in mice: A transmission electron microscopy
study. Exp Eye Res. 1991;52:647-659.
Hess JH, Fitzgerald P. Lack of DNase I mRNA sequences in murine lenses. Mol Vis. 1996; 2:8.