Effects of prior freezing or drying on the swelling behaviour of the bovine cornea.

Abstract

Frozen or dried corneal grafts are commonly used for stromal transplantation such as lamellar keratoplasty (full or partial thickness), keratophakia, epikeratophakia. Structural properties are important for the final optical results of these surgeries but the effects of freezing/thawing and drying/rehydration on the properties of the stroma are known little compared with the corneal endothelium, mainly because of lack of non-invasive technique to evaluate the stromal structure. This study aimed to investigate the swelling and structural properties of the bovine corneal stroma following freezing or drying by X-ray diffraction which was a non-invasive technique and could give ultra-structural information in hydrated tissues.
Bovine corneas were either frozen at -40 degrees C or dried to constant weight in a dessicator over silica gel. Swelling was carried out by placing the corneas into dialysis tubing and equilibrating them against various concentrations of polyethylene glycol (PEG) to obtain a range of tissue hydrations. This method minimises the loss of soluble tissue components during the swelling process. Synchrotron X-ray diffraction was used to measure the average intermolecular spacing, the interfibrillar spacing and the fibril diameter as a function of hydration. Changes in light scattering were detected using a microdensitometer.
Freezing and thawing of the cornea caused an increase in light scattering by 63.9% at tissue hydration (H) = 3.4, and by 50.0% at H = 4.9. Repeated freezing and thawing causes further increased by 38.9% at the second time and another 36.0% at the third time (P < 0.05). There was a tendency for both the frozen and the dried corneas to lose some swelling ability, achieving hydrations respectively of 10% and 18% below those of fresh corneas at 0 PEG. There were no changes in the fibril diameters, interfibrillar or intermolecular spacings as measured by X-ray diffraction in the equilibrated fresh, pre-frozen and pre-dried corneas.
The increase in light scattering and the loss of swelling ability after freezing and thawing probably results from structural changes following the close association of the collagen molecules and fibrils whilst the tissue is in the dry or frozen state. Some unknown changes in the extracellular matrix between the collagen fibrils may also play a role in the light scattering. The equilibration technique may improve the quality of rehydrated corneal graft or lenticules used for corneal surgeries.

Full text

Page 1

Chin Med J 2009;122(2):212-218
212
Original article

Effects of prior freezing or drying on the swelling behaviour of the
bovine cornea

HUANG Yi-fei, Keith M. Meek, WANG Li-qiang and WANG Da-jiang

Keywords: cornea; freezing; stroma; swelling; X-ray diffraction

Background Frozen or dried corneal grafts are commonly used for stromal transplantation such as lamellar
keratoplasty (full or partial thickness), keratophakia, epikeratophakia. Structural properties are important for the final
optical results of these surgeries but the effects of freezing/thawing and drying/rehydration on the properties of the stroma
are known little compared with the corneal endothelium, mainly because of lack of non-invasive technique to evaluate the
stromal structure. This study aimed to investigate the swelling and structural properties of the bovine corneal stroma
following freezing or drying by X-ray diffraction which was a non-invasive technique and could give ultra-structural
information in hydrated tissues.
Methods Bovine corneas were either frozen at –40°C or dried to constant weight in a dessicator over silica gel.
Swelling was carried out by placing the corneas into dialysis tubing and equilibrating them against various concentrations
of polyethylene glycol (PEG) to obtain a range of tissue hydrations. This method minimises the loss of soluble tissue
components during the swelling process. Synchrotron X-ray diffraction was used to measure the average intermolecular
spacing, the interfibrillar spacing and the fibril diameter as a function of hydration. Changes in light scattering were
detected using a microdensitometer.
Results Freezing and thawing of the cornea caused an increase in light scattering by 63.9% at tissue hydration (H)=3.4,
and by 50.0% at H=4.9. Repeated freezing and thawing causes further increased by 38.9% at the second time and
another 36.0% at the third time (P <0.05). There was a tendency for both the frozen and the dried corneas to lose some
swelling ability, achieving hydrations respectively of 10% and 18% below those of fresh corneas at 0 PEG. There were no
changes in the fibril diameters, interfibrillar or intermolecular spacings as measured by X-ray diffraction in the equilibrated
fresh, pre-frozen and pre-dried corneas.
Conclusions The increase in light scattering and the loss of swelling ability after freezing and thawing probably results
from structural changes following the close association of the collagen molecules and fibrils whilst the tissue is in the dry
or frozen state. Some unknown changes in the extracellular matrix between the collagen fibrils may also play a role in the
light scattering. The equilibration technique may improve the quality of rehydrated corneal graft or lenticules used for
corneal surgeries.
Chin Med J 2009;122(2):212-218

T
he cornea is the main refractive component in the eye.
Not only does it have to transmit most of the incident
light in the visible spectrum, it also has to be very strong
to withstand the external and internal forces to which it is
subjected as part of the outer casing of the eye. This
strength comes from the constituent collagen fibrils and
the extrafibrillar matrix in which they are embedded.

The use of both dehydrated corneas and pre-frozen (i.e.
frozen then thawed) lenticules for corneal transplantation
makes it important to understand the structural
consequences of storage procedures that might affect the
performance of the tissue. Such studies are also important
for understanding the effects of storage on corneas used
for research studies. In the past, studies on freezing of the
cornea have concentrated on the effects on the endothelial
cells lining the back surface of the tissue. But little
attention had been paid to the stroma, mainly because of
lack of non-invasive techniques to evaluate the stromal
structure. Cryo-electron microscopy has been used to
study the stroma and has shown that, while frozen, the
extracellular matrix is extremely disrupted.1 Most of the
useful data have, however,
diffraction/scattering studies (XRD). With this
technique, not only the structure of hydrated stroma could
be investigated,2-4 but also it was possible to evaluate the
structural changes from freezing and drying. Fullwood
and Meek5 showed that, while frozen, corneal collagen
fibrils are reduced in diameter and are forced into close
association with each other, with the extrafibrillar
come from X-ray
DOI:10.3760/cma.j.issn.0366-6999.2009.02.019
Department of Ophthalmology, Chinese PLA General Hospital,
Beijing 100853, China (Huang YF, Wang LQ and Wang DJ)
Department of Optometry and Vision Sciences, University of
Cardiff, Cardiff CF1 3AT, United Kingdom (Meek KM)
Correspondence to: Dr. HUANG Yi-fei, Department of
Ophthalmology, Chinese PLA General Hospital, Beijing 100853,
China (Tel: 86-10-66937943. Fax: 86-10-68286682. Email:
301yk@sina.com) and Dr. Keith M. Meek, Department of
Optometry and Vision Sciences, University of Cardiff, Cardiff CF1
3AT, United Kingdom
HUANG Yi-fei and Keith M. Meek contributed equally to this
work.

Page 2

Chinese Medical Journal 2009;122(2):212-218
213
components concentrating on the fibril surface. However,
on thawing, the fibrils gradually separate, regaining their
normal diameters and organization. Scanning electron
microscopy revealed that swollen corneas subjected to
freezing suffered irreversible structural damage.5

The effects of drying on the ultrastructure and swelling
properties of the stroma have also received scant attention.
Goodfellow et al6 used wide angle XRD to study corneal
tissue rehydrated after drying. They found that pre-dried
corneas swell less than fresh corneas. Sayers et al7
showed that, at any given hydration, pre-dried corneas
have a lower average interfibrillar spacing than fresh ones.
They interpreted this to indicate the presence of
collagen-free ‘lakes’ in the dried-then-swollen tissues.
Meek et al8 using wide-angle XRD, showed that, as
corneas dry below physiological hydration, intra-fibrillar
fluid is lost and the fibril diameter reduces significantly.
Fratzl and Daxer9 also used wide angle XRD to examine
corneal drying and suggested that drying is a two-stage
process and that, at a specific hydration, a structural
transformation of the collagen fibrils occurs.

The studies on the swelling properties of the cornea were
usually carried out by placing the corneas directly into the
swelling solution. 8 However, analysis of the content of
the solutions after the experiments revealed that many of
the extrafibrillar matrix constituents (responsible for the
swelling properties as well as for maintaining the stromal
architecture) were removed from the tissue during the
swelling.10,11 The purpose of the present study was to
examine the swelling behaviour of pre-frozen and
pre-dried corneas using an equilibration technique that
inhibits the removal of extrafibrillar matrix during the
experiments. XRD measurements were used to monitor
changes in structural parameters during swelling and
changes in light scattering in the tissue were detected
using a microdensitometer.

METHODS

Samples
Fresh bovine eyeballs were obtained from an abattoir.
Central corneal buttons were excised from the eye within
3 hours of death. By scraping with a scalpel, corneal
epithelium and endothelium were removed. Some of the
corneas were wrapped in cling film and stored at –40°C
until use. Others were placed in a dessicator and were
dried over silica gel at room temperature.6 Constant dry
weight was obtained after at least two weeks. The tissues
were kept in the dessicator until use. In some experiments,
scleral tissue was treated identically to compare the
response with that of the adjacent cornea. The tissue
hydration (H) can be calculated from H=(wet weight–dry
weight)/dry weight.

It is known that structural parameters such as interfibrillar
or intermolecular spacings depend on tissue hydration.
Accordingly, all specimens used in this work were
weighed before being wrapped and stored. This allowed
us to calculate the tissue hydration in the fresh state. For
XRD experiments, samples were weighed again before
and after exposure to obtain a mean wet weight. This
allowed us to calculate the tissue hydration at the time of
the experiment.

Tissue hydration was set by using an equilibration
technique that was developed for studies of cartilage.12
Equilibration was carried out by placing each cornea in a
strip of 14 kD cut-off dialysis tubing and leaving it for
four days at 4°C in a Na2HPO4/NaH2PO4 buffer solution
(PH 7.4) with polyethylene glycol (PEG). The surface of
the cornea remained in contact with the sides of the
dialysis tubing during this equilibration period. Different
corneal sectors were dialysed at a range of PEG
concentrations between 0% and 25%, with sodium
chloride added to give a final ionic strength (µ) of 0.03.
This method prevented the tissue from coming into direct
contact with the solution, so that the PEG could not
diffuse into the tissue and large molecules in the tissue
could not leak out.13 Furthermore, it allowed time for
water concentration gradients within the tissue to even
out gradually.

Light scattering of the cornea
Measurement of light absorbance (optical density) was
carried out using an
microdensitometer (LKB Instruments Inc, Gaithersburg,
USA).14 The wavelength of light produced by the laser
was 633 nm. The shape of the beam/detector was
rectangular (50 µm by 800 µm). The corneal tissue was
placed on a clean glass slide and the tissue was scanned
linearly from the limbus to the centre of the cornea. The
beam was directed through the cornea from posterior to
anterior at 90°. To avoid light scattering caused by
surface irregularities or by wrinkles, a large sector was
cut from each corneal button for the measurement. In this
way, the tissue could be flattened on the glass plate and
the posterior surface of the tissue made contact with the
slide smoothly without any bubbles.

Absorbance was constant across the cornea except at the
edges (Figure 1). The background of the slide was
subtracted from the light absorbance value. As the cornea
loses its transparency, the densitometer records increased
light absorbance because the light scattered by the cornea
cannot be detected by the photocell in the densitometer.
So an increasing value of the light absorbance means
decreasing transparency or increasing light scattering.
Light scattering was therefore expressed in terms of
absorbance units (a.u.). Consequently, the method is
suitable for looking at the change in scattering caused by
a given process such as freezing. For the single
freezing/thawing experiment, the fresh tissue remained on
the glass plate throughout. For the repeated freezing/
thawing/freezing experiment, the fresh whole eyeballs
had been frozen at –40°C for 20 hours before the corneas
were cut and placed on the glass plates. The repeated
Ultrascan XL laser

Page 3

Chin Med J 2009;122(2):212-218
214


Figure 2. The low-angle meridional patterns from rat tail tendon (A). The low-angle first equatorial pattern from normal bovine cornea
(arrow, B). The low-angle subsidiary equatorial pattern from normal bovine cornea (arrow, C) and the high-angle equatorial pattern from
normal bovine cornea (arrow, D).


Figure 1. A plot from a fresh bovine cornea by laser
densitometer scanning. The two peaks of light absorbance are
from the edges of the cornea. The light absorbance between the
two peaks is from the cornea and the glass slide. The absorbance
in the area beyond the cornea is from the glass slide only, which
is the background and needs to be extracted.

freezing lasted for one hour each time at the same
temperature, and the tissue was left on the glass plate
throughout the experiment, wrapped in clingfilm whilst
frozen to prevent evaporation. Then the tissue was thawed
at room temperature for about 20 minutes (longer times
did not further improve the tissue transparency). The
scanning of the tissue took less than one minute so did
not change the tissue hydration much.

X-ray diffraction
XRD was carried out using the synchrotron X-ray source
of CCLRC (the Council for the Central Laboratory of the
Research Councils) at Daresbury, U.K. The tissues were
held in air-tight cells during the exposure. The X-ray beam
was always directed along the optical axis of the cornea.

XRD patterns were collected at two different stations
(Figure 2). Station 7.2b was used for the high-angle X-ray
investigation. The camera length was about 12 cm and the
camera was filled with helium gas during exposure to
reduce air scattering. The wavelength of the radiation was
0.1488 nm and the beam diameter was 0.5 mm. Stations
2.1 were used for the low-angle X-ray investigation. The
camera lengths were typically 6 metres. The camera was
evacuated to reduce air scattering. The wavelength was
0.154 nm and the beam dimensions were 0.25×4 mm.
High-angle X-ray patterns were recorded on Caeverken
AB X-ray film (Caeverken, Strängnäss, Sweden) whereas
the low-angle X-ray station were equipped with a
two-dimensional detector (512×512 image pixels; wire to
wire resolution 1 mm). High-angle patterns were
obtained in 3–7 minutes and low-angle patterns in 0.5–10
minutes.

The exposed X-ray films were scanned using an
Ultrascan XL laser microdensitometer (LKB Instruments
Inc) which produced linear scans across the diffraction
rings. The background scatter was subtracted from the
trace before measurements were made, and the most
frequently occurring Bragg spacing (corresponding to the
maximum intensity in the X-ray reflection) was
calculated from each of these scans. The patterns
collected by the detector were corrected for detector
response. The background due to the empty cell was
subtracted and 1-dimensional scans through the patterns
were taken using the fibre XRD programs "BSL" and
"xotoko".

The X-ray patterns were analyzed as described
previously.5,8,14 The Bragg spacing was calculated from the
innermost low angle equatorial reflection and the average
centre-to-centre spacing between the collagen fibrils (the
interfibrillar spacing) was obtained by multiplying the
Bragg spacing by the factor 1.12 on the assumption that the
fibrils are arranged in a liquid-like manner. 7,14,15 The fibril
diameter was calculated from the first subsidiary equatorial
diffraction ring.15,16 The collagen intermolecular spacing
was obtained by multiplying the high-angle equatorial
Bragg spacing by the factor 1.11.8,14,15 A comprehensive
account of the collection and analysis of X-ray data from
the cornea was reviewed.17

Statistical analysis
All results were expressed as mean ± standard deviation
(SD). The data between the two groups were compared
using the LSD test. All statistical analyses were
performed using SPSS 10.0 for windows (SPSS Inc.,
Chicago, IL, USA). A P value less than 0.05 was
considered statistically significant.

RESULTS

Effects of freezing and thawing on light scattering in
the cornea
Light scattering from the fresh cornea was (0.012±0.005)

Page 4

Chinese Medical Journal 2009;122(2):212-218
215
a.u. at H=3.4 (normal hydration). After being frozen
(–40°C for 24 hours) and then thawed at room
temperature, the light scattering increased by 63.9% on
average. When fresh whole bovine eyeballs were stored at
4°C for 50 hours, the corneal hydration increased to
H=4.9 and the light scattering doubled. When these
corneas were then frozen and thawed, the light scattering
increased by 50.0% on average. The frozen tissue showed
an enormous increase of the light scattering (Table 1).

Table 1. Effect of freezing (–40 °C for 24 hours) and thawing on
the light scattering from bovine cornea
Light scattering (a.u.)
Hydration Fresh
Frozen
3.4±0.1 (n=7)
0.012±0.005 0.992±0.104
4.9±0.2* (n=7)
0.028±0.006 0.959±0.035
*The samples are from eyeballs that had been stored at 4°C for 50 hours in
clingfilm.

The multiple freezing and thawing experiment was
carried out using three corneas, all of which were close to
physiological hydration. The light scattering was
(0.018±0.004) a.u. after the first thawing. This is similar
to the result (0.020±0.005) a.u. showed in Table 1.
Repeated freezing-then-thawing increased the light
scattering by further 38.9% the second time and by
another 36.0% the third time (Table 2). There were
significant increases in the light scattering after either
single or multiple freezing and thawing (P <0.05). We
examined each tissue (including preparation, thawing and
scanning) within half an hour. We have found that if the
same tissue on the glass plate of the densitometer was
scanned repeatedly, and evaporation controlled, the
method provided very reproducible results and, in
practice, there was no significant change in the light
scattering within one or two hours. So any changes after
freezing-then-thawing should be considered as real.

Table 2. Effect of repeated freezing-then-thawing on the light
scattering from bovine cornea
Light scattering (a.u.)
H
1st frozen
&thawed &thawed
&thawed
3.3 0.015±0.002
0.028±0.012 0.043±0.004 0.014 (86.6)
3.4 0.023±0.002
0.025±0.003 0.028±0.005
3.3 0.016±0.001
0.023±0.002 0.032±0.002 0.007 (43.8)
3.3 0.018±0.004
0.025±0.003 0.034±0.007 0.007 (38.9)
H = hydration. The bottom line shows the average values from the three comeras
as above.

Effects of freezing and drying on corneal hydration
Table 3 and Figure 3 show the hydration results from
fresh, pre-frozen, dried bovine corneas and fresh sclera
after equilibration at pH 7.4 (µ=0.03). There was a trend
that, at a lower PEG concentrations, the hydration was
higher in the fresh cornea and lower in the pre-dried
cornea (Hfresh > Hfrozen >Hdried). Statistical analysis showed
that there was a significant difference between the fresh
and pre-dried corneas at 0, 1.5% and 2.5% PEG (P <0.01),
indicating that lower hydrations were achieved in
pre-dried corneas. However, there was no statistical
significance between any of the other corneal groups at
Thawed
0.020±0.005
0.042±0.011
Fresh-thawed
0.008 (63.9%)
0.014 (50.0%)
Difference (a.u.,%)
2nd frozen
3rd frozen
1st–2nd 2nd–3rd
0.015 (53.6)
0.003 (12.0)
0.009 (39.1)
0.009 (36.0)
0.002 (8.7)
Table 3. Hydration of bovine corneal stroma equilibrated at pH 7.4
(µ=0.03)
Hydration (mean ± SE, n = 4 for each mean value) PEG
(%)
Fresh corneaPre-frozen cornea
0 10.92 ± 0.409.79 ± 0.32
0.5 8.10 ± 0.397.96 ± 0.25
1.5 5.53 ± 0.215.25 ± 0.18
2.5 4.05 ± 0.053.70 ± 0.38
5.0 2.45 ± 0.11 2.44 ± 0.20
7.5 1.84 ± 0.181.85 ± 0.38
15.0 0.95 ± 0.101.19 ± 0.02
25.0 0.71 ± 0.010.77 ± 0.00

Pre-dried cornea
8.95 ± 0.22
7.40 ± 0.32
4.37 ± 0.18
2.54 ± 0.30
2.36 ± 0.20
1.83 ± 0.19
0.97 ± 0.03
0.71 ± 0.02
Sclera
2.33±0.21
2.09±0.09
1.94±0.08
1.80±0.12
1.31±0.06
1.11±0.05
0.84±0.01
0.66±0.09


Figure 3. The hydration of bovine corneal stroma equilibrated at
various PEG concentrations at pH 7.4 (µ=0.03). The
corresponding hydrations of bovine sclera are shown for
comparison. The differences among the sets of data are greater
at lower PEG concentrations (i.e. when the tissue is more
swollen).

any PEG concentration. As expected, the hydration of the
sclera was consistently lower than that of the cornea at
any PEG concentration.

Effects of freezing and drying on the interfibrillar
spacings, fibril diameters and intermolecular spacing
Figure 4 shows the results of the interfibrillar spacings of
corneas equilibrated at pH 7.4, as a function of corneal
hydration. Second order polynomials were fitted to the
data to give an idea of the trends. From this figure, it can
be seen that the interfibrillar spacings increased with
hydration in the fresh, pre-frozen and pre-dried bovine
corneas, but when hydration was above 6 or 7, the
interfibrillar spacings increased more slowly with the
hydration, although the changes followed the similar
trends. The apparent drop in the interfibrillar spacing at
H=8 in the dry cornea is probably an artefact due to the
small sample numbers. The interfibrillar spacing at H=3.2
was calculated to be 65 nm for the fresh cornea.

All the samples were at H >1.83 and it has been shown
that there is little change in fibril diameter with hydration
in this range5 so the data within each set were averaged to
obtain fibril diameters. The average values of the fibril
diameter from the fresh, pre-frozen and pre-dried corneas
were (38.9±1.1) nm (n=13), (38.7±1.0) nm (n=9) and
(38.7±0.9) nm (n=9), respectively. There were no
significant differences between any two groups (P >0.05).
When hydration was very low (H <1), no equatorial
reflections for the interfibrillar spacing or fibril diameter

Page 5

Chin Med J 2009;122(2):212-218
216
could be obtained because the collagen fibrils touch each
other at this point.

Figure 5 shows the changes in the mean intermolecular
spacings in fresh, pre-frozen and pre-dried bovine corneas
that had been equilibrated to different hydrations at pH
7.4 (µ=0.03). The intermolecular spacings at H=0 were
obtained from corneas that had been equilibrated to a
very low hydration and then further dried on silica gel,
since H=0 could not be achieved using the PEG
equilibration technique alone. The result showed similar
trends in the fresh, pre-frozen and pre-dried corneas. The
intermolecular spacing changed greatly only at lower
hydration (H <1) and it remained almost unchanged from
H >1 to near 9. The mean intermolecular spacings above
H=1 were (1.731± 0.034) nm (n=6) in the fresh,
(1.713±0.033) nm (n=6) in the pre-frozen and
(1.724±0.020) nm (n=6) in the pre-dried corneas
respectively. There was no difference between these mean
values statistically (P >0.91).


Figure 4. Interfibrillar spacings of bovine corneas equilibrated
at pH 7.4 (µ=0.03). Second order polynomials have been fitted
to show the trends in the data. There is no significant difference
in the relationship between interfibrillar spacing and hydration
between the three sets of data.


Figure 5. Intermolecular spacings in bovine corneas equilibrated
at pH 7.4 (µ=0.03). The intermolecular spacings follow similar
trends in the fresh, pre-frozen and pre-dried corneas. We did not
draw lines for each set of data, because the intermolecular
spacing changes at two stages, When H <1, and H >1. We draw
the line according to the data from all samples (either H <1 or H
>1), showing the two stage changes of the intermolecular
spacing with hydration.

DISCUSSION

Equilibration technique
In previous studies, swelling by direct immersion in the
bathing solution was found to cause a significant loss of
soluble proteins and PGs from the corneal stroma.18 More
recent studies have shown that the equilibration method
essentially prevents the loss of PGs from cartilage19 or
from the corneal stroma,20 and the latter was supported
morphologically by using histochemical techniques.13 So
as not to impede the passage of fluid from the membrane
to the tissue (or vice versa), whenever the tissue was put
in the dialysis tubing and there were no bubbles trapped
between the tubing and the cornea. When this was done
and so long as some pressure existed (such as the osmotic
pressure produced by 0.5% PEG), no water was found
inside the dialysis bag. Goodfellow et al,6 Elliott et al21
and Sayers et al7 immersed the corneal stroma directly in
distilled water or different bathing solutions (without
dialysis tubing) and the stroma could swell much more
than in our experiments. In our case, the corneal stroma
could probably have swelled more but swelling was
constrained by the pressure exerted by the dialysis
membrane, and this may be the reason for the small
amount of surface water on the cornea at 0 PEG.

Effects of freezing and thawing on light scattering of
the cornea
Accurate measurements of light scattering in the cornea
are difficult for a number of reasons, including the need
to prevent scattering from the corneal surface and the
requirement to measure only the light that has passed
directly through the cornea. Most techniques are based on
shining a narrow beam of light normally through the
centre of the cornea and measuring the light that passes
through the cornea. Although the size of the laser beam
and detector used in our experiments was rather large, we
have found that the system provides reproducible data
and is very reliable for detecting changes for comparative
purposes (before and after freezing). Our results showed
that the light scattering from the fresh cornea at normal
hydration (H=3.4) was 0.012 a.u. on average, which
increased by 64% after freezing and thawing (Table 1). At
a higher hydration (H=4.9), the light scattering was about
1.3 times higher than that at normal hydration. After
freezing and thawing, the light scattering of the tissue
increased 50% (Table 1). Repeated freezing and thawing
showed that the light scattering values of the corneas at
H=3.3 were 0.018, 0.025 and 0.034 a.u. after the first,
second and third freezing and thawing, respectively
(Table 2). Although we did not measure the corneas in
their fresh state before the first freezing in this
experiment, the light scattering was increased after the
first freezing and thawing by 67% if compared with the
result from the fresh corneas in Table 1. The second and
the third freezing and thawing made the light scattering
further increase by 38.9% and 36.0%, respectively. The
reasons for the light scattering caused by freezing and
thawing are not known. Freezing and thawing will affect
scattering from the damaged cellular components. It is
also possible that there are some unknown changes in the
extracellular matrix between the collagen caused by
freezing and thawing, but other factors must exist,
because repeated freezing and thawing can further affect