Vitamin A in human eyes: amount,
distribution, and composition
C. D. B. Bridges, Richard A. Alvarez, and Shao-Ling Fong
The amount, distribution, and composition of vitamin A stored in the eyes of 29 postmortem
donors was determined by a combination of techniques, including high-pressure liquid chromatography. The vitamin A concentration in the pigment epithelium-choroid (RPE-Ch) was the
highest observed for human non-liver tissue and amounted to 7.9 ±4.5 nmol/eye (n = 28), or
10.4 ±7.1 fJLg/gm (n = 27). There was no evidence for significant losses during the interval
between death and enucleation or during subsequent storage at 4° C. The vitamin A extracted
from the retina was 15.3% of that in the corresponding RPE-Ch. By measuring rhodopsin
regeneration in retinal homogenates incubated with ii-cis retinal, we estimated that the
amount of vitamin A in the RPE-Ch of fully dark-adapted eyes would represent 2.5 mole
equivalents of the retinal rhodopsin, a value similar to that found in the frog. A preponderance
of the vitamin A in the eye was esterified (98.3% in the RPE-Ch, 79.3% in the retina) and.
consisted principally of stearate and palmitate in the ratio of 1:4.8. A small amount of oleate
was also detected. The ratio o/U-cis isomer over the all-trans averaged 1.52 ± 0.48 (n = 11).
Variable, usually small proportions of 13-cis retinyl esters were also present. Intact RPE-Ch or
isolated RPE cells esterified exogenous all-trans-3H\-retinol to the same fatty acids in roughly
the same proportions as in the endogenous stores. The all-trans configuration was mainly
retained during uptake and esterification, although some isomerization to 13-cis also occurred.
No 11-cis isomer was formed under these conditions. (INVEST OPHTHALMOL VIS SCI 22:706714, 1982.)
Key words: human donor eyes, pigment epithelium, retina, vitamin A,
retinyl esters, 11-cis, a\\-trans, 13-cis, isomerization, enzymatic retinol
esterification, rhodopsin, high-pressure liquid chromatography
V
itamin A* deficiency is a serious consequence of poor nutrition in the developing
countries. It may also be a problem in North
From the Cullen Eye Institute and Program in Neuroscience, Department of Ophthalmology, Baylor College of Medicine, Houston, Texas.
Supported by grants from the Retina Research Foundation, Houston, Texas (C. D. B. Bridges), Fight For
Sight, Inc., New York (Shao-Ling Fong), and by National Institutes of Health grants EY 02489 and EY
02520 (C. D. B. Bridges).
Submitted for publication July 14, 1981.
Reprint requests: C. D. B. Bridges, Cullen Eye Institute, Baylor College of Medicine, Houston, Tex.
77030.
"""Vitamin A" refers to unspecified mixtures of retinol
and retinyl esters.
706
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
America, where it has been estimated that
24% to 41% of the population in some U.S.
states and U.S. and Canadian cities may be
seriously depleted in vitamin A (for review
see ref. 1). Therefore it is important to determine the magnitude of vitamin A reserves
in various human tissues.
The highest concentration of vitamin A is
known to occur in the liver. Non-liver tissues, i.e., the adrenals, fat, heart, kidney,
lung, muscle, pancreas, prostate, spleen,
testis, and thyroid, show little tendency to
accumulate vitamin A.2 The eyes were not
included in these studies, although in many
animals, including frog, rabbit and cattle, 3 " 7
it is known that the ocular tissues contain
stores of vitamin A that have a possible role in
0146-0404/82/060706+09$00.90/0 © 1982 Assoc. for Res. in Vis. and Ophthal., Inc.
Volume 22
Number 6
Vitamin A in human eyes 707
Table I. Vitamin A content of human RPE-Ch
Time before
processing (hr)*
Donor No.
Age and sex
63
64
65
66
67
68
69
70
70
72
82
83
84
86
87
88
90
97
98
99
101
102
105
106
107
108
109
110
111
75F
74F
57M
17F
27M
50F
59M
58F
53F
61F
69M
62M
59F
31M
59F
61M
56M
51M
44M
62F
70M
57F
79M
55M
79M
64F
75M
66M
69M
*i
3
5
8
1
3
2
3
2
2
2
5
6
3
9
1
2
4
6
3
2
5
3
7
2
14
6
5
2
5
Vitamin A content
t2
No. of eyes
1
16
1
3
8
10
45
66
50
16
4
88
76
14
14
21
1
96
2
15
2
34
9
66
20
15
30
4
14
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
1
1
Meant
S.D.
n
nmol/eye
21.7
19.8
10.0
17.6
13.7
(82.5)
25.1
6.2
11.8
11.6
4.2
2.3
10.1
9.0
3.7
7.2
27.6
3.0
11.0
6.0
3.8
4.8
—
7.6
2.3
4.7
17.5
15.5
3.8
10.4
±7.1
27
14.0
11.8
9.9
11.2
7.1
(48.9)
13.8
4.9
6.5
7.4
3.3
1.6
9.2
6.0
3.0
7.1
19.3
2.6
9.4
5.6
3.5
4.3
5.3
10.7
3.0
5.3
14.8
15.7
6.1
7.9
±4.5
28
*t,, Time between death and enucleation; t2) time between enucleation and processing (storage at 4°).
t Excludes data in parentheses (~10 S.D. higher than mean).
the visual cycle and may also buffer the retina
against hypovitaminosis A.
Accordingly, we investigated the amount,
distribution, and composition of the vitamin
A present in the eyes from 29 postmortem
donors. Compared with other non-liver tissues, the pigment epithelium-choroid (RPECh) had the highest concentration of vitamin
A, which consisted mainly of 11-cis and alltrans retinyl palmitate and stearate. On a
molar basis, the quantity averaged 2 to 3
times the amount of rhodopsin in the outer
segments.
Additionally, we established that all-trans
retinyl stearate and palmitate could be synthesized in vitro by isolated RPE-Ch or RPE
cells incubated with all-trans-3H-retinol. An
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
important step in the visual cycle, the uptake
and esterification of all-trans retinol by the
pigment epithelium, can therefore be studied with tissue from donor eyes.
Methods
Source of material. Eyes were provided by the
Lions Eyes of Texas Eye Bank, courtesy of Mr. Bob
Fort and Dr. E. J. Farge, Executive Director of the
Eye Bank Association of America. After enucleation, they were kept protected from light at 4° C in
a moist container. Details concerning the storage
period as well as the delay between death and
enucleation are provided in Table I. Case histories
were available for reference.
Extraction of retinols, retinals, and retinyl esters. Under dim red illumination, the cornea,
lens, and vitreous were removed, the retina was
Invest. Ophthalmol. Vis. Sci.
June 1982
708 Bridges et at.
0.8
0.6
j
ABSOf
5 0.4
0.2
A
B
A
A
/1
J \\
0
330
390
510
570
630
690
750
WAVELENGTH, nm
Fig. 1. A, Absorption spectrum of an extract from
the RPE-Ch of a single eye after chromatography
on alumina (solvent, 10% dioxane in n-hexane;
volume, 2.0 ml; donor 68). B, Absorption spectrum of the color generated when a 200 /JL\ aliquot
of this extract is transferred to 100 /JL\ of chloroform
and mixed with 1 ml of Can-Price reagent (see ref.
10 for details of method).
gently peeled away under mammalian Ringers solution, and the combined RPE-Ch was removed
and weighed. The tissue (RPE-Ch or retina) was
then extracted with 2 x 4 ml volumes of acetone.
The efficacy of this procedure 7 was verified by extracting subsequently with chloroform-methanol
(2:1 v/v).
The acetone extract was filtered through glass
fiber and dried under a stream of purified nitrogen. The residue was dissolved in 1 ml of n-hexane
containing 10% v/v dioxane and was applied to a 1
by 1.5 cm column of n-hexane-washed alumina
(Woelm-Pharma No. 02069, deactivated with 5%
water), and the combined isomers of retinol, retinal, and retinyl esters were eluted with 10% v/v
dioxane/n-hexane. Recovery of these compounds
was 95% to 100%. After its ultraviolet spectrum
had been measured, a portion of this mixture was
transferred to dry chloroform for measurement of
vitamin A and retinal (present in retinas only) by
the Can-Price reaction (see Fig. I). 8 " 10 In these
analyses, all-trans retinol and all-trans retinyl
palmitate purified by high-pressure liquid chromatography (HPLC) were used as standards. The
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Can-Price method measures the total vitamin A;
its isomeric configuration and degree of esterification were determined by HPLC. The remainder
of the extract was therefore transferred to the appropriate running solvent (i.e., ether/n-hexane,
dioxane/n-hexane, or acetonitrile).
HPLC. The equipment has been described
previously.7' l0- " A Waters WISP automatic sample injector (Waters Associates, Inc., Milford,
Mass.) was added to the system to facilitate the
processing of large numbers of samples. For normal phase, two 4.6 mm by 25 cm columns packed
with 5 ju,m adsorbents were used separately or in
series. The first contained Ultrasphere Si 60, the
second Spherisorb CN. This combination gave
highly reproducible results under isocratic conditions. The esters were therefore analyzed with one
injection using 0.4% ether/n-hexane as eluent,
and the retinals and retinols with a second identical injection using 9% dioxane/n-hexane.
The normal-phase system was optimized for
isomer separation. It was somewhat less satisfactory for resolution of the two retinyl esters that
predominate in human eyes, i.e., the stearate and
palmitate. Baseline resolution of these esters was
obtained with a 4.6 mm by 15 cm 5 /u,m Spherisorb
ODS column, with acetonitrile as the mobile
phase (Fig. 2). Isomers could not be separated in
this system.
The system was calibrated by injecting 0.05 to
1.0 nmol quantities of the authentic isomeric retinyl esters, retinals, and retinols (see refs. 7 and
10 for methods of synthesis). Areas of the eluted
peaks were computed with a Columbia Supergrator III. This method is independent of off-column
recoveries, which ranged from 70% to 90%. The
small proportion of free retinol was determined by
adding the amounts of the 13-c/s, 11-cis, and alltrans isomers. No 9-cis retinol was detected.
Esterification of exogenous all-trans-3H-retinol
by RPE tissue. Intact, isolated RPE-Ch in 3 ml of
Ringer's solution (pH 7.4) was incubated in darkness for 3 hr at 38° C with 105 disintegrations per
minute (dpm) of all-trans ll,12- 3 H 2 -retinol (1.6 X
107 dpm per nmol) added to the medium in 20 to
30 fji\ of ethanol. After it had been rinsed gently in
Ringer's solution, the tissue was extracted with
acetone as described above.
The incubation medium used for isolated RPE
cells was as follows: NCTC-135.(Gibco, Grand Island, N.Y.), 50 ml; Eagle's minimum essential
medium with Earle's salts, without L-glutamine
and bicarbonate (Gibco), 50 ml; water, 400 ml;
gentamycin, 50 mg; oxacillin, 250 mg; amphotericin B, 50 mg; ca. 20 ml 0.83M NaHCO 3 to adjust
the pH to 7.4. After the retina had been removed,
Volume 22
Number 6
Vitamin A in human eyes
709
Table II. Rhodopsin content of human retinas
Rhodopsin in retina* (nmol/eye)
Donor No.
Vitamin A in RPE-Ch
(nmol/eye)
Initial
Regenerated
63
64
65
66
67
68
69
70
71
14.0
11.8
9.9
11.2
7.1
(48.9)
13.8
4.9
6.5
2.2
2.1
3.8
2.3
2.3
(1.1)
1.7
1.2
2.9
2.3
3.0
5.2
3.5
2.5
(2-0)
3.4
2.1
3.5
72
7.4
4.0
6.2
9.6
±3.1
2.5
±0.9
3.5
±1.3
Meant
S.D.
Vitamin A /regenerated
rhodopsin
6.09
3.93
1.90
3.20
2.84
(24.45)
4.06
2.33
1.86
1.19
3.04
•Rhodopsin before and after regeneration by incubating the retinal homogenate with 11-cis retinal; see Methods.
t Excludes donor 68 (data in parentheses).
the eyecup was filled with incubation medium,
which was very gently aspirated back and forth
with a Pasteur pipette. Examination of the resulting suspension under phase-contrast showed that
it contained RPE cells with little contamination
from rod outer segments or red blood cells. The
RPE cells were mainly single and did not occur in
clusters or sheets as in similar preparations from
frogs (unpublished observations). The suspension
from one eyecup was allowed to settle in a 20 ml
vial, and the surplus clear supernatant medium
was removed until the volume was reduced to
about 2 ml. All-trans ll,12- 3 H 2 -retinol in 20 yul of
ethanol (10(i cpm) was added to 1 ml portions of
this suspension, which were then incubated in
darkness for 2 to 18 hr under 5% CO 2 , 95% O2 at
38° C. The tissue was then extracted with acetone.
The acetone extracts were treated as described
above and analyzed by HPLC. Fractions from the
HPLC column were collected in 7 ml scintillation
vials at 0.2 to 1.0 min intervals, mixed with 5 ml of
Scintilene (Fisher Scientific Co., Pittsburgh, Pa.),
and counted on a Packard Tricarb liquid scintillation spectrometer.
Both of the above preparations gave essentially
identical results. Labeled retinol was prepared
from the corresponding labeled retinoic acid.10
This was generously donated by Hoffman-La
Roche, courtesy of Dr. W. E. Scott.
Measurement of rhodopsin in the retina. Each
retina was homogenized in 7 ml 0.15M NaCl containing 0.01M Na phosphate buffer (pH 7.4). The
homogenate was divided into two 3 ml aliquots,
one of which was incubated at room temperature
with 10 to 20 nmol of 11-cis retinal added in 25 fx\
of ethanol. After 2Vi hr, 0.3 ml of 0.1M NH 2 OH
was added: The mixture was then centrifuged and
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
the pellet was extracted with 1.5 ml of 2% aqueous
Emulphogene. The amount of rhodopsin was obtained by measuring the spectra before and after
bleaching.
Results
Amount of vitamin A in the RPE-Ch. Total
vitamin A (retinyl esters and retinol) was
measured by the Carr-Price reaction. Fig. 1
illustrates a typical Carr-Price analysis of an
extract from the RPE-Ch of one eye. Fig. 1,
A, is the ultraviolet spectrum (measured in
10% dioxane/n-hexane) after chromatography
on alumina. The Xmax is at 325 nm. Fig. 1, B,
is the spectrum of the blue color, Xmax at ca.
620 nm, generated when this extract is transferred to chloroform and mixed with a saturated solution of antimony trichloride in
chloroform.
The quantity of vitamin A stored in the
RPE-Ch was found to be 7.9 ± 4.5 nmol/eye
(28 donors), or 10.4 ± 7 . 1 /xg/gm wet weight
(27 donors). These findings are summarized
in Table I. The results from one donor (in
parentheses) were omitted from these calculations; the RPE-Ch contained 48.6 nmol in
the left eye and 49.1 nmol in the right eye.
Roth values were about 10 S.D. higher than
the mean.
The average time from death to enucleation was close to 4 hr (Table I). The difference between eyes enucleated at times less
than 4 hr (8.3 ± 3 . 5 nmol/eye, n = 15) and
those greater than 4 hr (7.1 ± 5 . 1 nmol/eye,
n = 13) was not statistically significant.
Invest. Ophthalmol. Vis. Sci.
June 1982
710 Bridges et at.
B
M
325
0.005
STANDARDS
MIN.
Fig. 2. HPLC of retinyl esters from the RPE-Ch
from three pairs of donor eyes. A, Normal-phase
analysis. Peak 2, 11-cis retinyl palmitate and
stearate (shoulder); peak 3, unidentified; peak 4,
all-trans retinyl palmitate and stearate (shoulder);
peak 5, all-trans retinyl oleate. Mobile phase,
0.4% diethyl ether in n-hexane at 1 ml/min; columns, Ultrasphere Si 60 and Spherisorb CN in
series (see Methods); sample, ca. 21 nmol of retinyl esters injected in 40 /x\. B, Reverse-phase
analysis of peaks 2 and 4 collected from five of the
above injections on normal-phase HPLC. Each
peak was transferred to 100 fx\ of acetonitrile.
Samples were injected in 5 /x.1 volumes onto a
Spherisorb ODS column. The standard consisted
of a 10/u.l injection containing 1 nmol each of 11-cis
retinyl palmitate (?) and stearate (S). The 11-cis
and all-trans isomers have the same retention
times in this system. Peak 2.1, 11-cis retinyl palmitate; peak 2.2, 11-cis retinyl stearate; peak 4.1, alltrans retinyl palmitate; peak 4.2, all-trans retinyl
stearate. Mobile phase, acetonitrile at 2 ml/ min.
Loss of vitamin A during subsequent storage in darkness at 4° C was assessed directly
as follows. Seven pairs of eyes were used.
The RPE-Ch from one eye was dissected out
and extracted immediately, and the con-
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
tralateral eye was left in storage for periods of
24 hr (two pairs), 48 hr (two pairs), 72 hr (two
pairs), and 92 hr (one pair). The vitamin A
content of the stored eyes averaged 2% less
than the control eyes on a nmol/eye basis,
and 16% less in terms of/xg/gm wet weight of
RPE-Ch. The amounts of vitamin A obtained
from each eye of a pair processed simultaneously did not differ by more than 20%
Amount of vitamin A in the retina. The
vitamin A in the retinas from 19 eyes (12
donors) was 1.1 ± 0.5 nmol/retina, equivalent to 15.3% of that in the corresponding
RPE-Ch (7.2 ± 4.4 nmol).
Amount of rhodopsin in the retina. In addition to retinyl esters and retinol (see below), the acetone extracts from the retina
contained 13-cis, 11-cis, and all-trans retinal
derived from breakdown of the rhodopsin
(the Carr-Price spectra were corrected for
the presence of retinal, as described in ref.
8). We chose to measure the rhodopsin directly, however, since our technique permitted a simultaneous determination of the
amount of regenerable opsin present. The
amounts of rhodopsin in retinal homogenates
before and after incubation with 11-cis retinal
are presented in Table II. The eyes from nine
donors averaged 2.5 ± 0.9 nmol rhodopsin
per retina initially, and 3.5 ± 1.3 nmol after
regeneration. This suggested that on average
29% had been bleached. For these donors,
the amount of vitamin A in the RPE-Ch averaged 3.0 moles per mole of rhodopsin in the
regenerated retinas (range 1.2 to 6.1).
Donor 68 was exceptional: the vitamin
A/rhodopsin ratio was 24.5. This was because
the vitamin A in the RPE-Ch was abnormally
high (Table I), whereas the rhodopsin content of the retina was within the normal range
(2.0 nmol after regeneration).
Composition of vitamin A in the RPE-Ch
and retina—isomeric configuration and degree of esterification. Small amounts of free
13-cis, 11-cis, and all-trans retinols were
found in the RPE-Ch and retina, but most of
the vitamin A stored in the eye was esterified. The proportion esterified was higher in
the RPE-Ch (98.3% ± 2.6%, n = 15) than in
the retina (79.3% ± 11.9%, n = 15).
A normal-phase HPLC profile for the ret-
Volume 22
Number 6
Vitamin A in human eyes
711
Fig. 3. Retinyl esters from the paired eyes of two donors. A, Donor 63, left eye; B, right eye.
C, Donor 67, left eye; D, right eye. Peak 1, 13-cis retinyl palmitate and stearate (shoulder
visible in C); other peaks identified as in Fig. 2. Mobile phase as in Fig. 2; column, Ultrasphere
Si 60; samples containing approximately 1 nmol retinyl ester were injected in 20 //,!.
inyl esters extracted from the combined
RPE-Ch from three pairs of donor eyes is
illustrated in Fig. 2, A. There are two major
peaks, 2 ' 4 each with a shoulder on the leading
edge. They were identified as follows (cf.
refs. 7 and 10). Co-chromatography with authentic compounds indicated that peak 2 was
mainly 11-cis retinyl palmitate. The shoulder
corresponded to 11-cis retinyl stearate. Similarly, peak 4 appeared to be mainly all-trans
retinyl palmitate, the shoulder corresponding to the stearate. Identification of these
isomers as 11-cis and all-trans was confirmed
by elution, saponification,7' 10 and HPLC.
Peak 2 yielded 11-cis retinol; peak 4 yielded
the all-trans isomer. The isolated peaks were
also subjected to reverse-phase HPLC, as
shown in Fig. 1, B. Peak 4 was resolved into
two peaks that co-chromatographed with alltrans retinyl palmitate (peak 4.1) and stearate
(peak 4.2). Peak 2 was similarly resolved into
its two component 11-cis esters. The all-trans
and 11-cis isomers had identical retention
times in this system.
Peak 3, which was sometimes composite,
was not identified. Peak 5 co-chromatographed with all-trans retinyl oleate.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
The retinyl esters in human RPE-Ch
therefore consist mainly of 11-cis and alltrans retinyl stearate and palmitate. Both
normal- and reverse-phase HPLC showed
that the relative proportions of these esters
was similar for both isomers, the stearate averaging 17% of the palmitate for the 11-cis
isomer and 25% for the all-trans. The retinyl
esters from the retinas had essentially the
same composition as those from the RPE-Ch
and displayed identical HPLC profiles.
Usually, as in Fig. 2, only the 11-cis and
all-trans retinyl palmitate and stearate were
present in significant amount. For 11 donors,
the ratio of 11-cis to all-trans isomers was
found to be 1.52 ± 0.48 (range 0.78 to 2.24).
Variable proportions of the 13-cis retinyl esters were also encountered. They were usually small, as in Fig. 3, A and B, which shows
typical HPLC profiles of the retinyl esters
from the left (A) and right (B) eyes from a
single donor. However, on occasions the
13-cis peak was prominent and even overshadowed the 11-cis, as in Fig. 3, C. The
other eye from the same donor (Fig. 3, D)
displayed a normal HPLC profile. The retinyl esters from eyes kept in the refrigerator
Invest. Ophthalmol. Vis. Sci.
June 1982
712 Bridges et at.
dioactivity was eluted), and analyzed by
HPLC. The lower record represents the
ester region of the elution profile. Esters of
the 13-cis, ll-cis, and all-trans isomers are
present. The main peaks are due to the palmitates. As in Figs. 2 and 3, the shoulders on
the rising segments of the 13-cis and all-trans
bands arise from lesser amounts of their respective stearates.
The upper record shows the measured radioactivity in fractions collected from the column effluent. The major radioactive peak 4
corresponds to all-trans retinyl palmitate.
There is also a shoulder due to the all-trans
retinyl stearate and a small peak 5 corresponding to the putative all-trans retinyl
oleate, demonstrating that the exogenous
all-trans-3H-retinol had been esterified to the
same fatty acids in roughly the proportions
that characterize the endogenous stores.
10,000 r
5,000
'325
0.005
15
MIN.
30
Fig. 4. Esterification of all-trflns-3H-retinol by isolated RPE cells. Lower record, HPLC absorbance
profile of extracted retinyl esters. Upper record,
radioactivity of eluted 0.2 min fractions. Peakidentification as in Figs. 2 and 3. Arrows indicate the positions of the 13-cis and ll-cis isomers.
Mobile phase and columns as in Fig. 2; sample,
0.6 nmol of retinyl esters injected in 20 fx\.
for periods up to 92 hr did not have increased
amounts of 13-cis isomer, showing that it was
not a storage artifact. A second unusual feature in this pair of eyes was that the ratio of
ll-cis to all-trans retinyl esters was 0.8 for
the left eye (Fig. 3, C) and 2.6 for the right
eye (Fig. 3, D). Major differences between
the two eyes from a single donor were rare.
In Fig. 3, A and B, for example, the ratios
were 1.6 and 2.2, respectively.
Esterification of exogenous aH-trans-3Hretinol by isolated human RPE. Fig. 4 illustrates the result of a typical experiment in
which RPE cells aspirated from an eyecup
were incubated with chromatographically purified all-trans ll,12- 3 H 2 -retinol. After incubation, the mixture was extracted, chromatographed on alumina (61% of the applied ra-
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
The small radioactive peak 3 corresponds
to a similar minor peak typically observed in
the HPLC absorbance profile of human ocular retinyl esters (e.g., in Figs. 2 and 3).
Although the all-trans configuration was
mainly retained during uptake and esterification of the exogenous retinol, some
isomerization also occurred, as shown by the
radioactive peak 1 eluting in the position of
13-cis retinyl palmitate. In contrast, no significant radioactivity eluted with the ll-cis
retinyl esters.
Virtually identical results were obtained
with intact, isolated RPE-Ch.
Discussion
The present studies show that the concentration of vitamin A in the RPE-Ch is more
than ten times higher than that reported 2 for
other non-liver tissues. The concentration
must be considerably higher if the RPE is
considered alone, since it is probable that little vitamin A occurs in the choroid. 6 The vitamin A in the retinas was about 15% of that
in the RPE-Ch. Although this is higher than
that reported for the frog (4%, ref. 6), adhesion of RPE cells could have contributed to
the amount extracted.
The eye was found to be variable in its
vitamin A content. Similar variability has
Volume 22
Number 6
been observed for the liver, where autopsy
tissue from 58 traffic fatalities in Iowa averaged 149 ± 132 fJLg/g. Of these, as many as
14 were judged to be vitamin A deficient.
Since we could not compare the ocular vitamin A levels with those in the liver and
plasma in the present study, we were unable
to determine whether the amount of vitamin
A in the eye was less responsive to nutritional
deficiency, as would be the case if it were
buffer against hypovitaminosis A.
Eyes from postmortem donors are not
equivalent to those from healthy individuals.
Infection is believed to lower serum levels of
retinol,12 which in turn could affect the
amount stored in the eye. Further, the eyes
are not handled under ideal conditions: they
are not fresh or dark-adapted. However, we
were able to show that storage at 4° C for the
periods encountered in the present study did
not cause significant losses of vitamin A.
There is less certainty regarding losses that
may have occurred between death and enucleation, although there did not seem to be a
relationship between this delay, which averaged 4 hr, and the amount of vitamin A extracted.
The degree of light adaptation was assessed
by incubating retinal homogenates with 11-cis
retinal. If it is assumed that the opsin in the
rod outer segment membranes had remained
regenerable during storage, the figures in
Table II show that on average 29% of the
rhodopsin had been bleached. If transport
between rod outer segments and RPE had
been normal at that time, the vitamin A generated by bleaching would have augmented
the amount in the RPE by about 1 nmol/eye.
In that case, in the dark-adapted state the
RPE-Ch would have contained 2.5 moles of
vitamin A per mole of rhodopsin, a ratio
comparable with that found in frogs.6 Results
from donor 68 (Tables I and II), where the
content of vitamin A was more than 10 S.D.
above the mean, were not included in the
above calculations. Examination of the medical record did not provide an explanation for
this abnormal value. In contrast, the amount
of visual pigment was not unusual: initially,
the retina appeared to be 45% bleached and
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Vitamin A in human eyes
713
contained 2.0 nmol of rhodopsin after regeneration .
More than 98% of the vitamin A in the
RPE-Ch was esterified and consisted principally of a small proportion of stearate mixed
with palmitate in the ratio of 1:4.8. For comparison, the retinyl esters in the RPE of
dark-adapted frogs (unpublished observations
and ref. 11) and rabbits 7 consist almost exclusively of the palmitate, whereas in lightadapted rats (unpublished observation and
ref. 7) the RPE contains all-trans retinyl
stearate and palmitate in the ratio of 1:1.3. In
the present work, there was no evidence for
any significant difference in fatty acid composition between the 11-cis and all-trans retinyl esters.
Only 79% of the vitamin A in the retina
was esterified. Similarly, in dark-adapted
frogs it has been reported 6 that the proportion
of vitamin A existing as ester is lower in the
retina than in the RPE (51% compared with
99%). In the frog, most of the unesterified retinol occurred in the rod outer segments. 6 ' l3
The retinyl esters in the retina had the
same composition as those in the RPE-Ch.
Futterman and Andrews14 found relatively
more unsaturated fatty acids (oleate and
palmitoleate) in the esters synthesized by
human retinas incubated with exogenous retinol. Their ratio of stearate to palmitate
(1:4.3), however, was similar to that found
for the endogenous esters in the present
work.
The esterification of retinol when it flows
into the RPE from the rod outer segments
during light-adaptation is an integral feature
of the visual cycle. Many tissues can esterify
exogenous retinol. These include liver,15 intestinal mucosa,16 and Sertoli cells.17 Ocular
tissues that have been investigated are cattle
RPE, 4 ' 18 neural retina from several species,19
and retinoblastoma cells.20 The present study
shows that isolated human RPE cells can also
esterify retinol. When the substrate is the
pure all-trans isomer, the product is overwhelmingly in the all-trans conformation,
and the synthesized esters have a chromatographic profile that is similar to that of the
endogenous all-trans retinyl esters. No 11-cis
714 Bridges et al.
retinyl esters were formed. This confirms our
findings with isolated frog RPE (unpublished
observations), although ll-cis retinyl esters
are formed from all-£rans-3H-retinol injected
into the eyes of intact frogs.10' 21 The small
amount of 13-cis retinyl ester observed probably originated from nonspecific isomerization of the all-trans to 13-cis retinol and
its subsequent esterification. Variable proportions of 13-cis retinyl esters were extracted
from our donor eyes. This isomer was not an
artifact of storage, and at present its origin
and possible function are unknown.
REFERENCES
1. Sauberlich HE, Hodges RE, Wallace DL, Kolder
H, Canham JE, Hood J, Raica N, and Lowry LK:
Vitamin A metabolism and requirements in the
human studied with the use of labeled retinol.
Vitam Horm 32:251, 1974.
2. Raica N, Scott J, Lowry L, and Sauberlich HE: Vitamin A concentration in human tissues collected
from five areas in the United States. Am J Clin Nutr
25:291, 1972.
3. Wald G: Carotenoids and the visual cycle. J Gen
Physiol 19:351, 1935.
4. Krinsky NI: The enzymatic esterification of vitamin
A. J Biol Chem 232:881, 1958.
5. Hubbard R and Dowling JE: Formation and utilization of ll-cis vitamin A by the eye tissues during
light- and dark-adaptation. Nature 193:341, 1962.
6. Bridges CDB: Vitamin A and the role of the pigment
epithelium during bleaching and regeneration of
rhodopsin in the frog eye. Exp Eye Res 22:435,
1976.
7. Alvarez RA, Bridges CDB, and Fong SL: Chromatography of fatty acid esters of retinol isomers, Analysis of retinyl esters stored in the eye. INVEST
OPHTHALMOL VIS SCI 20:304, 1981.
8. Hubbard R, Brown PK, and Bownds D: Methodology of vitamin A and visual pigments. Methods
Enzymol 18:615, 1971.
9. Bridges CDB: Storage, distribution and utilization
of vitamin A in the eyes of adult amphibians and
their tadpoles. Vision Res 15:1311, 1975.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Invest. Ophthalmol. Vis. Sci.
June 1982
10. Bridges CDB and Alvarez RA: Measurement of the
vitamin A cycle. Methods Enzymol (vol. 81, in
press).
11. Bridges CDB, Fong SL, and Alvarez RA: Separation by programmed-gradient high-pressure liquid
chromatography of vitamin A isomers, their esters,
aldehydes, oximes and vitamin A 2 —presence of retinyl ester in dark-adapted goldfish pigment epithelium. Vision Res 15:1311, 1980.
12. Arroyave G and Calcano H: Descenso de los niveles
sericos de retinol y su proteina de enlace (RBP)
durante las infecciones. Arch Latinoam Nutr 29:233,
1979.
13. Bridges CDB: 11-cts Vitamin A in dark-adapted rod
outer segments is a probable source of prosthetic
groups in rhodopsin biosynthesis. Nature 259:247,
1976.
14. Futterman S and Andrews JS: The fatty acid composition of human retinal vitamin A ester and the
lipids of human retinal tissue. INVEST OPHTHALMOL
3:441, 1964.
15. Futterman S and Andrews JS: The composition of
liver vitamin A ester and the synthesis of vitamin A
ester by liver microsomes. J Biol Chem 239:4077,
1964.
16. Huang DS and Goodman DS: Vitamin A and carotenoids. I. Intestinal absorption and metabolism of
14
C-labeled vitamin A alcohol and /3-carotene in the
rat. J Biol Chem 240:2839, 1965.
17. Ahluwalia B and Gambhir KK: Retinyl palmitate, a
major retinyl ester in rat testis seminiferous tubules.
II. The Sertoli cell, site of retinol esterification. Int J
Vitam Nutr Res 46:330, 1976.
18. Berman ER, Horowitz J, Segal N, Fisher S, and
Feeney-Burns L: Enzymatic esterification of vitamin A in the pigment epithelium of bovine retina.
Biochim Biophys Acta 630:36, 1980.
19. Futterman S and Andrews JS: Metabolism of the
retina. IV. The composition of vitamin A ester synthesized by the retina. J Biol Chem 239:81, 1964.
20. Saari JC, Bredberg L, and Futterman S: Esterification of retinol by retinoblastoma cell lines WERIRbl and Y79. INVEST OPHTHALMOL VIS SCI 19:1301,
1980.
21. Alvarez RA and Bridges CDB: Isomerization of alltrans to 11-cts retinol in the frog. INVEST OPHTHALMOL Vis SCI 20(ARVO Suppl.):214, 1981.