Age-related changes in the protein
concentration gradient and the crystallin
polypeptides of the lens
M. Kazem Mostafapour and Connie A. Schwartz
With the use of a freeze-sectioning technique, the existence of a centripetal protein gradient in
the rabbit lens is demonstrated. The steepness of this gradient (the difference in the protein
concentration between periphery and the center) varies with age: it increases from a value of
40% in 6- to 8-month-old lenses to 60% by the age of 4 to 5 years and then begins to decline at
an average rate of about 5% per year, reaching a value below 40% in 9-year-old lenses. The
decline in protein concentration gradient is reciprocated by a gain in the water content.
Electrophoretic evidence shows that the decline in the protein gradient is preceded and accompanied by postsynthetic modification of lens crystallin polypeptides. Modifications result in the
formation of crosslinked material that stays on the top of sodium dodecyl sulfate gels and of
degraded polypeptides with molecular weights below 20,000 daltons. It is hypothesized that the
protein gradient comes about probably because lens crystallins are capable of making extensive
surface contacts to produce a tightly packed matrix. As the lens ages, post-synthetic modifications and hydrolytic breakdown produce a gradual disorganization in these structural
proteins. Local or general disorganization will allow water to fill the gaps and produce hydration, which could predispose the lens to opacity formation. (INVEST OPHTHALMOL VIS Sci
22:606-612, 1982.)
Key words: lens, crystallin polypeptides, SDS-PAGE, freeze-sectioning,
protein gradient, protein modifications, lens hydration
Lens has a higher protein concentration
than any other tissue, and the concentration
in the nucleus is greater than in the cortex. 1
This difference is not discontinuous but is
due to a steady rise in the relative protein
content from periphery toward the center of
the lens in the form of a centripetal gradient. 2 The biochemical and physiological significance of this gradient have thus far not
From The Institute of Biological Sciences, Oakland University, Rochester, Mich.
Supported by NIH grants EY-03319 and EY-00484.
Submitted for publication Feb. 13, 1981.
Reprint requests: Dr. M. K. Mostafapour, Institute of
Biological Sciences, Oakland University, Rochester,
Mich. 48063.
606
been explored. The possibility exists that a
centripetal protein gradient may act as a rising barrier against the influx of water and ions
toward the inner regions of the lens. Breakdown of this barrier may lead to gradual hydration of the lens. Hydration has been implicated as a factor in the process of lens opacification and formation of certain types of
cataract. 3 " 5 One reason for the breakdown of
the lens gradient may be the slow, qualitative, age-related change in lens crystallins
due to enzymatic and random hydrolytic
breakdown of lens crystallins.6"11 Loss of certain crystallins due to leak-out has also been
reported. 12 " 14
In this report we examine the age-related
quantitative changes in the protein concen-
0146-0404/82/050606+07$00.70/0 © 1982 Assoc. for Res. in Vis. and Ophthal., Inc.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933334/ on 06/18/2017
Volume 22
Number 5
Crystallins and lens protein gradient 607
tration gradient in the rabbit lens and the
qualitative changes in the crystallin polypeptides of the rabbit and human lenses.
Materials and methods
Freeze-sectioning of the lens. Freshly excised
albino rabbit lenses were used in these experiments. All lenses used appeared normal and
showed no opacities. Lenses were frozen by covering them with powdered dry ice. The hardened
frozen lenses were then placed on a Petri dish and
allowed to partially thaw at room temperature. If
the lens surface became too soft, it was refrozen by
sprinkling dry ice over it. The objective of these
manipulations was to have a lens frozen all through
to a uniform consistency, which when cut with a
blade, would yield an undeformed, clean-cut surface that was neither too brittle nor too soft and
sticky. Blades and forceps for cutting and handling frozen lenses were kept cold on Dry Ice.
With a single-edge razor blade, the lens was cut at
the poles into two halves by exerting a steady
pressure on the blade. A slice about 1 to 1.5 mm
thick was cut parallel to one of the cut surfaces.
This slice represented a cross section of the lens.
Depending on the need, an equatorial or polar
strip was cut from this disk. The cut strip was
sectioned into several small sections representing
different areas of the lens. These sections were
used for various measurements. The steps involved in freeze-sectioning are depicted in Fig. 1.
Determination of protein and water concentration across lenses. Strips representing cross sections of the lens were obtained as described above.
Two such strips were analyzed for each lens. Each
strip was cut into approximately equal sections.
Each section was placed on a tared piece of Parafilm and weighed. The Parafilm with the section
on it was dropped into a test tube containing 2 ml
of 0. IN NaOH to dissolve the tissue. Typically, an
equatorial strip of a rabbit lens was divided into six
sections and a polar strip into four sections.
After dissolution of tissue, aliquots were analyzed for protein by the Lowry method 15 or by
Bio-Rad protein assay reagent, with bovine serum
albumin used as standard. The protein concentration of each section was determined by using an
aliquot of the NaOH protein solution. To make
sure that determinations were not interfered with
by nonprotein materials originally existing in the
tissue separate aliquots of NaOH-protein solution
were precipitated by the addition of cold 10%
trichloroacetic acid (TCA). The precipitate was collected by centrifugation and redissolved in 0. IN
3
Fig. 1. Diagrammatic representation of lens freezesectioning procedure. Step 1, Lens, frozen on Dry
Ice, is cut along polar axis into two halves. Step 2,
Slice of about 1 to 1.5 mm thick is sliced off from
one of the cut faces. Step 3, Broken lines showing
the strip to be cut along equatorial axes. Step 4,
Strip obtained in step 3 is cut into approximately 2
mm (or smaller) sections to be used for various analyses. For further details see text.
NaOH, brought to its original volume, and subjected to protein assay. Protein concentrations
were calculated as percent wet weight of the tissue
section. To determine the water content, serial
sections of lens were weighed and then dried in a
forced-draft oven at 110° C to constant weight.
The water content is expressed as percent wet
weight of the section. The strips or sections need
not be of exactly the same dimensions because
each section is individually weighed and assayed.
However, after some practice, one can cut sections that are very close to each other in size and
weight.
Sodium dodecyl sufate -polyacrylamide gel electrophoresis (SDS-PAGE). Equatorial strips of
rabbit lenses were cut into several sections of
about 4 to 5 mg each. Sections were weighed and
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933334/ on 06/18/2017
Invest. Ophthalmol. Vis. Sci.
May 1982
608 Mostafapour and Schwartz
1
1
1
o
CN
7 0
6 0
•
100
\
/
G
\
\
-
I
/
#
\ /
1
z
LLI
1
I
\
\
X
0
-
4 0
-
V
/
\
\
o
an
3 0
-
1
"I
T
2 0
N
i
i
1
3
i
i
5
7
S E C T I O N
Fig. 2. Protein and water concentration gradients
in the rabbit lens. Equatorial strips of lens from
8-month-old rabbits were divided into six sections
each. Separate serial sections were used for determining the percent protein and water in the
lens. Values are expressed as percentages of the
wet weight of the sections. Two measurements
were done on each lens, and the points shown are
averages of at least four determinations with a
variation of 6%. N, Lens center.
dissolved in SDS buffer. After complete dissolution, protein concentration was determined by the
Bio-Rad protein assay solution. The presence of
SDS does not interfere with this assay. Aliquots
were subjected to SDS-PAGE under the conditions previously described.16 Quantitation of the
Coomassie blue-stained bands of human lens crystallins was carried out with a soft laser scanning
densitometer (LKB) equipped with an integrator.
Results
Determination of the protein and water concentration gradient. Repeated determination
of protein concentration in polar and equatorial sections of the same lens yielded practically identical protein distribution curves,
Fig. 3. Age-related change in the extent of the lens
protein and water concentration gradients in rabbit lenses. Each point on the graph shows the percent protein and water content at the center of the
lenses of each age group (corresponding to point A7
of Fig. 2). Note the gradual decline in the percent
of protein and the concomitant gain in the water
concentration after the age of 4.
indicating that the changes in protein concentration from periphery toward the center
of the lens were uniform from all directions.
It was thus considered preferable in the present study to use the longer equatorial strips.
When aliquots from the NaOH-protein solution were first treated with excess 10% TCA
and the precipitated protein was assayed,
practically the same values as those for the
non-TCA-treated aliquots were obtained.
This indicated that there was no significant
contribution to the protein value from possible nonprotein materials. Thus values obtained from the assay of the NaOH digest can
be used directly as a measure of protein content. Fig. 2 shows the protein and water distribution in young (8-month-old) rabbit
lenses. The values are given as percentages of
the milligram wet weight of the tissue. Each
point on the graphs is the average of at least
four determinations, with a variation of about
6%. Fig. 2 (solid line) shows that the protein
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933334/ on 06/18/2017
Volume 22
Number 5
concentration rises from about 25% at the
outer cortical region to about 60% at the center of the lens. The rise in protein concentration is accompanied by an equivalent drop in
the water content (broken line).
Age-related changes in the lens protein gradient. In order to study age-related changes
that may occur in the protein gradient, rabbit
lenses ranging in age from 6 months to 9
years were examined. The outermost sections of the lenses examined showed a comparatively similar protein concentration of
about 25%, but the concentration in inner
sections rose with a different slope for each
age group, producing gradient profiles characteristic for that age. The protein concentration rose steadily toward the lens center,
where it attained its highest value. Therefore, as shown in Fig. 3, a plot of the percent
protein concentrations (and the water content) at the center vs. the age of the lens can
be used to demonstrate age-related changes
in the extent of the gradient. Fig. 3 shows
that the protein concentration at the center of
the rabbit lens rose from a value of 60% in the
young rabbits up to 80% in the 4-yearolds—an increase of 20%—but thereafter
began to decline. The decline in the protein
concentration estimated from the sold line in
Fig. 3 was about 5% per year.
Crystallin polypeptide profiles of young
rabbit lenses. It is important to emphasize
that the lens sections obtained by freezesectioning were dissolved directly in the SDS
buffer without any further manipulations.
This prevented possible loss of the polypeptides or distortion in their relative quantities,
which could occur due to procedures such as
homogenization or centrifugation. Fig. 4
shows the SDS-PAGE patterns along the
equatorial axis of a young (6-month-old) rabbit lens. The prominent polypeptides are
numbered arbitrarily 1 to 10 for ease of reference. The identity and molecular weights of
rabbit crystallin polypeptides have been previously reported. 16 Briefly, bands 1 to 7 belong to beta-crystallins, and bands 8 to 10 to
alpha- and gamma-crystallin. Bands 8, 9, and
10 are not well-resolved in this figure because they are a little in excess. However,
Crystallins and lens protein gradient
A
B
C
609
D E F
Fig. 4. Polypeptide profiles in various regions of a
young rabbit lens. A strip representing various
regions of the lens along its equatorial axis was
obtained by the freeze-sectioning technique (see
text). It was serially divided into six sections,
which were analyzed by SDS-PAGE on 5% to 207c
gradient slabs containing 1% SDS (lanes A to F).
The major crystallin polypeptides are arbitrarily
numbered 1 to 10 for ease of reference, but a
number of minor bands are visible below band 10.
Because of the lentoid symmetry of the lens, sections A, B, and C are almost equivalent (mirror
images) of D, E, and F. Note that the relative
amount of some polypeptides decreases in going
from periphery (A) to the center (C) while that of
others increases.
had a smaller quantity of the sample protein
been applied to the gel then some of the
fainter bands in the figure would have disappeared. Bands below 10 are mostly nonspecific degradation products that accumulate by
age (see below). Of the six sections labeled A
to F in Fig. 4, sections A, B, and C represent
subcapsular, cortical, and nuclear areas of the
lens, respectively. The next three sections (D
to F) are mirror images of the first three, and
the polypeptide patterns in them are similar
to those in A to C. A comparison of the polypeptides shown in lanes A, B, and C (or F, E,
and D) reveals that polypeptides 1, 3, and 6
were present in all areas, that 2, 4, and 7
were more abundant in the outer cortical
areas, and that bands 5 and 8, which were
barely detectible in the subcapsular section,
were on the increase towards the center.
Changes in crystallin polpeptides due to
aging. Age-related changes in the lens crystallins become more evident when the polypeptide patterns of young and old rabbit
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933334/ on 06/18/2017
610
Invest. Ophthalmol. Vis. Set.
May 1982
Mostafapour and Schwartz
I 1
2 Yr
4Yr
6Yr
8 Yr.
Fig. 5. Age-related change in crystallin polypeptides of rabbit lenses. Experimental conditions
were the same as described for Fig. 4. Here, only the subcapsular (outermost sections) and the
center (innermost) section from each age group are compared. Age of the lens is indicated
below each set of lanes. Note the decrease in the relative amount of some crystallins by age,
especially in the nuclear sections. C, Cortical section; N, nuclear section; M, marker proteins
with molecular weights (xl(T 3 ) shown on the right.
Table I. Age-dependent changes in the relative percentages of human lens crystallins
22-year-old lenses
5-year-old lenses
68-year-old lenses
Sections
intact
%
x-linked
%
degraded
%
intact
%
x-linked
%
degraded
%
intact
%
x-linked
%
degraded
Subcapsular
Cortical
Outer nucleus
Lens center
100
98
96
95
0
2
0
0
0
0
4
5
91.9
95
93
88
2.7
2.5
2.6
4.8
5.4
2.5
4.4
7.2
85.3
52.8
35.3
25.7
9.8
30.2
41.1
57.1
4.9
17
17.6
17.2
%
X-linked = crosslinked.
Lens sections from 5-, 22-, and 68-year-old lenses were solubilized, electrophoresed, and scanned as described in Materials and Methods
for rabbit lenses. Total densities of each section have been taken as 100%. The integrated density under each region (intact, crosslinked,
and degraded crystallins) is given as the percent of the total. Vertical columns compare the relative quantity of the crystallins in sections
of the same lens. Horizontal rows compare the relative quantities of the crystallins in sections of one lens with those in equivalent sections
of other lenses.
lenses are compared. Fig. 5 shows the SDSPAGE patterns of cortical and nuclear sections of 2-, 4-, 6-, and 8-year-old rabbit
lenses. Electrophoretic conditions were the
same as for Fig. 4, and apparent differences
are due to age-related changes and photographic magnification. For economy of space,
only two sections, one outer cortical and one
nuclear, from each lens are shown. It can be
seen that there was a reduction in the number and quantity of many lens crystallin polypeptides in the nuclear sections of older
lenses, although band 10 (20,000 daltons)
remained relatively high. Reduction in the
quantity of bands 3, 7, and 9 was especially
noticeable. Another difference between the
young and the old lenses was that the quan-
tity of low-molecular-weight polypeptides
(below 20,000 daltons) increased in the nuclear region of older lenses.
In order to corroborate such age-related
changes, SDS-PAGE patterns of human lens
sections of a clear 5-year-old lens, a 22year-old eye bank lens, and a 68-year-old
brunescent brown lens were compared. Results shown in Fig. 6 confirm the observation
in the rabbit lens that there is a relative decline in the quantity of several crystallin
polypeptides together with an increase in the
polypeptides with molecular weights below
20,000 daltons (lanes B, C, and D of the 68year-old lens). Also interesting is the increase, by age, in the amount of material that
deposited on the top of the gel (arrows). This
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933334/ on 06/18/2017
Volume 22
Number 5
Crystallins and lens protein gradient
611
68
43
35
m mi tt*t •
26
A B C D
A B C D
5 Yr.
22 Yr
A B C D M
6 8 Yr.
Fig. 6. Electrophoretic (SDS-PAGE) profiles of sections from different areas of human lenses
of various ages. Experimental conditions were similar to those described for Fig. 4. Polypeptide patterns from four sections of each lens are shown: A, subcapsular (outer cortical), B,
cortical; C, outer nuclear, D, lens center. Note the decrease in the quantity of a number of
crystallin polypeptides, especially in the inner parts of the older lens, and the concomitant
increase in the crosslinked material staying on the top of the gel (arrows) and the increase in
bands below 20,000 daltons. M, Marker proteins. Numbers to the right give the molecular
weight (xlO~3) of the marker polypeptides.
material which was not dissociated by SDSmercaptoethanol-heat treatment must have
been formed through some nondisulfide covalent linkages. Material deposited on the top of
the gel will be referred to as crosslinked crystallins and the bands appearing below the
20,000 dalton mark as degraded crystallins.
The bands in the 35,000 to 20,000 dalton region are considered to be the intact crystallin
polypeptides.
Table I shows the results of soft laser
scanning densitometry of human lens SDSPAGE patterns. Intact polypeptides (those
migrating to the regular position of authentic
crystallins) comprised most of the crystallins
of various sections of the 5-year-old lens and a
very high percentage of that of subcapsular
sections of the 22- and 68-year-old lenses
(91.9% and 85.3%, respectively). The inner
sections of older lenses, however, showed a
progressive age and position-dependent decrease in intact crystallins and a corresponding increase in the crosslinked and degraded
crystallins.
Discussion
The freeze-sectioning method described
allows easy access to various regions of the
lens. This method was utilized to determine
protein concentration as well as crystallin
polypeptide profiles of serial sections of the
rabbit lens. Results demonstrate the existence of a protein gradient in the lens. This
gradient is centripetal and attains its highest
concentration at the center of the lens. It is
further shown that as the lens grows older,
the steepness of the gradient (the difference
in the protein concentration between the
periphery and the center) varies with age.
The maximal protein concentration, which is
about 60% in 6- to 8-month-old lenses, rises
to above 80% by the age of 4 to 5 and then
begins to decline at an average rate of 5% per
year. The decline in the percentage of lens
proteins should not be equated with an absolute decrease in the lens total crystallin content. A gain in lens water alone has a diluting
effect on lens protein concentration without
any actual loss of proteins.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933334/ on 06/18/2017
Invest. Ophthalmol. Vis. Sci.
May 1982
612 Mostafapour and Schwartz
As shown in Fig. 6 and Table I, the loss of
crystallins from their "normal" migrating positions on SDS-PAGE gels can be mostly accounted for by the appearance of crosslinked
and down-sized polypeptides at the top and
bottom of the gel, respectively. Data shown
in Table I indicate that the crystallins in the
outer region of old lenses seem to be mostly
intact but those of inner regions are mostly
transformed into cross-linked and degraded
materials. This could be because crystallins of
the outer cortical fibers have been relatively
newly synthesized and perhaps because the
fiber cells in the outer region are still capable
of providing the conditions that help maintain the crystallins intact.
Qualitative changes in crystallins seem to
precede the gain in lens water. For example,
in Fig. 5, the polypeptide pattern in the nuclear region of the 4-year-old rabbit lens
seems to have changed without any concomitant gain in the lens water content. It is possible that changes in the primary structure
have to ultimately affect the tertiary and
quaternary structures of the lens crystallins
before there is sufficient disorganization to
allow water to gradually come back in.
An important question that arises from the
present study is how lens fibers concentrate
their crystallins to such high levels. One reasonable possibility would be that the crystallins have three-dimensional structures with
surface topographies that enable them to
pack tightly in a crystal-like order. Such a
lattice will leave little space between the protein molecules for the unbound water to occupy. If the assumption that the gradient
comes about because of the tight fit of lens
crystallins is valid, then deteriorations in the
structure and conformation of these proteins
will result in disorganization, the gradient
will be prone to breakdown, and water and
electrolytes will rush in. The gain of water by
the lens (hydration) has long been implicated
in experimental cataracts.7 Perhaps it is in
the later stages of lens senescence and certain
types and degrees of cataract that a frank loss
of lens crystallins may occur. This question
is presently being investigated in our laboratory.
We thank Dr. V. N. Reddy for his helpful advice and
Dr. J. Reddan for donating the older rabbit lenses.
REFERENCES
1. Salit PVV: Biochemical investigation of the nitrogen,
weight and water content of crystallin lenses. Arch
Ophthalmol 5:623, 1931.
2. Philipson B: Distribution of protein within the normal rat lens. INVEST OPHTHALMOL 8:258, 1969.
3. Iwata S and Kinoshita JH: Mechanism of development of hereditary cataract in mice. INVEST OPHTHALMOL 10:504, 1971.
4. Kinoshita JH: Mechanisms initiating cataract formation. INVEST OPHTHALMOL 13:713, 1974.
5. Reddy VN, Schwass B, Chakrapani B, and Lim CP:
Biochemical changes associated with the development and removal of galactose cataracts. Exp Eye
Res 23:483, 1976.
6. Barber GW: Human cataractogenis: a review. Exp
Eye Res 16:85, 1973.
7. Clayton RM and Truman DES: The antigenic structure of chick beta-crystallin subunits. Exp Eye Res
18:495, 1974.
8. Harding JJ and Dilley KJ: Structural proteins of the
mammalian lens: a review with emphasis on changes
in development, aging and cataract. Exp Eye Res
22:1, 1976.
9. Mostafapour MK and Reddy VN: Studies on lens
proteins. II. Soluble and membrane proteins in
normal and cataractous lenses. Doc Ophthalmol
Proc Ser 18:193, 1979.
10. Ostrer H and Piatigorsky J: Relatedness of the polypeptides and their aggregates. Exp Eye Res 30:670,
1980.
11. Nasser S, Bradley R, Alcala J, and Maisel H: Regional differences in the composition of the bovine
lens urea-soluble protein. Exp Eye Res 30:109,
1980.
12. Piatigorsky J, Fukui HN, and Kinoshita JN: Differential metabolism and leakage of protein in an inherited cataract and a normal lens cultured with
ouabain. Nature 274:558, 1978.
13. Russel P, Smith SG, Carper DA, and Kinoshita JH:
Age and cataract related changes in the heavy molecular weight proteins and gamma crystallin composition of the mouse lens. Exp Eye Res 29:245,
1979.
14. Charton JM and Van Heyningen R: An investigation
into the loss of proteins of low molecular size from
the lens in senile cataract. Exp Eye Res 7:47, 1968.
15. Lowry OH, Rosebrough NJ, Farr AL, and Randall
RJ: Protein measurement with the Folin phenol
reagent. J Biol Chem 193:265, 1951.
16. Mostafapour MK and Reddy VN: Studies on lens
proteins. I. Subunit structure of beta crystallins of
rabbit lens cortex. INVEST OPHTHALMOL VIS SCI
17:660, 1978.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933334/ on 06/18/2017