`
`VITAMIN A DEFICIENCY AND NIGHT BLINDNESS
`
`BY JOHN E. DOWLING AND GEORGE WALD*
`
`BIOLOGICAL LABORATORIES OF HARVARD UNIVERSITY, CAMBRIDGE
`
`Communicated May 16', 1.958
`
`Its
`One of the oldest diseases known to man is nutritional night blindness.
`descriptions go back to the ancient Egyptian medical papyri and are already ac-
`companied by the correct prescription for» its cure, the eating of liver. Toward the
`end of World War I the factor in liver which cures night blindness was identified
`with the then newly discovered vitamin A.’
`Vitamin A is the precursor in the retina of the visual pigments of the rods and
`cones.’
`It seems reasonable to suppose that on a diet deficient in this factor the
`retina eventually synthesizes subnormal amounts of visual pigment, with the
`corresponding decline of visual sensitivity that constitutes night blindness.
`Some of the first studies of experimental human night blindness seemed to
`reveal such a simple and direct relationship.'3
`In two subjects deprived of vitamin
`A, the visual thresholds of both rods and cones began at once to rise, until a mild
`night blindness had been established} On oral administration of vitamin A or
`carotene, the thresholds of both rod and cone vision returned to normal within
`2-3 hours.
`
`It looked for a time, therefore, as though this might be an exemplary instance of
`the origin and cure of a biochemical disease, all elements of which were well under-
`stood. Further studies, however, exposed two major discrepancies:
`(1) Though in
`some subjects placed on a vitamin A—deficient diet the visual threshold began at
`once to rise, in a larger number it remained unchanged for periods ranging from
`several months5 to, in one instance, 2 years.“
`(2) Among the subjects who developed
`night blindness, some were completely cured within a few hours after receiving
`vitamin A, whereas others, though showing some immediate improvement, took
`months of vitamin A supplementation to return to normal.’
`One might take a simple position with regard to the first of these discrepancies.
`The amounts of vitamin A stored in the livers of healthy human subjects are known
`to vary enormously} In Britain, for example, Moore‘ found reserves in adults
`during 1941—44 ranging from about 7 to 750 ug/gm.
`If we take 1,500 gm. as the
`average weight of the adult liver and about 300 Mg. (about 1,000 I.U.) as the daily
`drain upon stored vitamin A, the average Briton stores enough vitamin A in his
`liver—if used economically—to tide him over some 500 days of total deprival. An
`unusually well-supplied Briton—if we can disregard spoilage—might survive seven
`times as long, or almost 10 years! On the other hand, the most poorly supplied
`members of this group might have run through their stored vitamin A within 1-2
`weeks.
`It is not difficult to understand, therefore, why most subjects taken from
`ordinary American or British environmentsfail to respond to vitamin A—deficient
`diets within months or even years.
`It is less clear why a fairly large proportion of
`them responded within a few days, even though in some instances highly supple-
`mented with vitamin A for the preceding period."
`The second discrepancy—the great variability in the times required to cure night
`blindness—raises other issues. The visual pigments are composed of vitamin A
`648
`
`
`
`VOL. 44, 1958
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`649
`
`aldehyde (retinene) joined to specific proteins of the rods and cones called “opsins.”
`The amounts of visual pigment that can be formed in the normal retina are limited,
`not by vitamin A, which is ordinarily present in excess, but by opsin.
`In thinking
`about night blindness, we have tended in the past to be too much preoccupied with
`vitamin A and have paid too little attention to the opsins.2
`When one does consider the opsins, this at once suggests further relationships.
`The outer segment of a rod—and this must be true also of many cones—is com-
`posed in considerable part of visual pigment, that is, of opsin, since the retinene
`chromophore constitutes only about 1 per cent of these molecules. Opsin accounts
`for about 40 per cent of the dry weight of the outer segment of a frog rod and 14
`per cent of that of a cattle rod?
`It is an important structural constituent of the
`rods and probably of the cones; and any loss of this protein might be equivalent to
`the structural deterioration of the visual receptors.
`Tansley” showed some years ago that in vitamin A—deficient rats and dogs,
`somewhat later than the decline in rhodopsin production that should have initiated
`night blindness, the outer segments of the rods deteriorated structurally.
`Johnson 1°
`confirmed and extended these observations in the rat; and recently similar changes
`have been observed in both rods and cones of the monkey.“ According to Johnson,
`after 7-13 weeks of vitamin A deprivation in young rats, many outer segments have
`disappeared, and those that remain stain abnormally. As the deficiency pro-
`gresses, the inner segments of the rods also degenerate, and then successively the
`external limiting membrane, the outer nuclear layer, and the inner nuclear layer.
`These changes occur sooner in central than in peripheral areas of the retina. The
`outer segments of rods which have deteriorated only slightly seem to repair con-
`siderably within 24 hours of feeding vitamin A. Even rods which have degenerated
`completely seem to be replaced Within 10-18 weeks of vitamin A supplementation.
`These observations suggest that the time required to cure night blindness may
`depend on the extent to which vitamin A deficiency has altered the retinal structure.
`Simple lack of vitamin A, through lowering the concentrations of visual pigments,
`might induce a night blindness that is cured as rapidly as vitamin A re-enters the
`retina; but the structural deterioration of the retina, heralded perhaps by the loss
`of opsin, might take much longer to repair.
`For these reasons it seemed Worthwhile to map the entire course of vitamin A
`deficiency and its cure in the rat.
`In single groups of animals we have measured
`simultaneously the vitamin A in the liver and blood, the retinal content of rhodop-
`sin and opsin, the electroretinographic threshold, and the ERG’s obtained over a
`wide range of light intensities.
`In key instances we have also examined the retinal
`histology.
`Not all these things were done for the first time. We have already discussed the
`histological studies of Tansley and Johnson and should mention particularly also
`Tansley’s fine study of rhodopsin synthesis in normal and vitamin A—deficient rats”
`and the measurements of liver, blood, and retinal vitamin A in normal and deficient
`rats by Lewis, Bodansky, Falk, and McGuire.”
`Plan of the E'xperiments.—A number of experiments were performed, all of which
`yielded substantially the same pattern of results. We shall describe primarily the
`last such experiment, because it brings together all the procedures and represents
`most completely and typically all our observations.
`
`
`
`650
`
`BIOCHEMISTRY: DOWLING’ AND WALD
`
`Pnoc. N. A. S.
`
`Male albino rats of the highly inbred Harvard colony, 22-24 days old and
`weighing 36-66 gm., were divided into two groups, one kept on the complete labora-
`tory ration, the other placed on the standard USP vitamin A test diet. The animals
`on the deficient diet continued to gain weight for about 5 weeks, though more
`slowly than normal. At this time they weighed an average of 112 gm. as compared
`with the control weight of 215 gm. but were altogether normal in appearance.
`In
`the fifth to seventh weeks their weights plateaued and thereafter declined rapidly.
`At the same time—in the seventh and eighth weeks—— the classic overt signs of
`vitamin A deficiency appeared, and by the end of the eighth week all the animals not
`sacrificed in the experiments had died.
`For electroretinography, animals that had been dark—adapted overnight were
`anesthetized with nembutal. The eye was held open with threads drawn through
`the lids. Cotton-wick electrodes were used, moistened with Ringer solution, one
`touching the side of the cornea, the other a shaved area on the cheek. The response
`was recorded with a capacity-coupled Grass P4 preamplifier and a Dumont oscillo-
`scope with camera attachment. The stimuli were ‘/5o—second flashes of white
`light, the intensity of which was controlled with neutral filters and photographic
`wedges. The absolute threshold was measured by starting with the light well
`below threshold and flashing it every few seconds at gradually increasing intensities
`until a response could be detected on the oscilloscope. This procedure was repeated
`until constant readings were obtained. Then the ERG was recorded over a wide
`range of intensities.
`After dark-adapting overnight, the same animals were used next morning for the
`biochemical measurements. They were again anesthetized, the body cavity was
`opened, and 5-10 ml. of blood were taken from the heart with an oxalated syringe.
`The entire liver was removed and also both eyes. One eye of each animal was used
`to measure rhodopsin, the other to measure opsin; but, since each of these determi-
`nations requires 2 retinas, animals were paired, usually on the basis of having yielded
`comparable electroretinograms.
`To determine blood vitamin A, the oxalated blood from one animal was centri-
`fuged, and the clear plasma was mixed with an equal volume of ethyl alcohol and
`extracted three times with petroleum ether. This extract was transferred to 0.3
`ml. of chloroform, and a micro»-antimony chloride test was performed by mixing
`0.25 ml. of the extract with 0.50 ml. of antimony chloride reagent, recording the
`absorption spectrum at once in a Cary recording spectrophotometer.
`The livers, weighing 4-12 gm., were ground with anhydrous sodium sulfate to a
`fine powder and extracted by shaking with diethyl ether. An aliquot of this
`extract was transferred to chloroform, and its vitamin A content determined by the
`antimony chloride procedure.
`To measure rhodopsin, two retinas were hardened in 4 per cent alum solution for
`15—20 minutes, then washed with distilled water and buffer, and extracted overnight
`with 0.2 ml. of 2 per cent digitonin solution. After centrifuging, 0.01 ml. of 1 M
`hydroxylamine was added to the extract, and the absorption spectra recorded
`before and after bleaching. The change in extinction at 500 mp measured rhodop-
`sin.
`
`Opsin was determined by measuring the capacity of retinas to regenerate rhodop-
`sin when incubated with neo—b (11—cz's) retinene.“ We found that rat rhodopsin
`
`
`
`VOL. 44, 1958
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`651
`
`in digitonin solution regenerates very little when bleached and incubated with
`neo-b retinene. For this reason whole retinas were exposed to bright light until
`wholly bleached. Then a large excess of neo-b retinene, dissolved in 0.025 ml.
`acetone, was added to the retinas suspended in buffer solution, the mixture was
`stirred periodically during 4-6 hours at room temperature and then left at 5° C.
`overnight. The rhodopsin which had formed was extracted and measured as
`described above. Control measurements showed that 70-80 per cent of rhodopsin
`originally present in a retina was regenerated and recovered by this procedure.
`
`100
`
`8
`
`
`
` 3Percenfaceofnormal
`
`oO
`
`70O
`
`Rhodop s in
`
` 7
`4
`S
`6
`7
`8
`
`Weeks on diet
`
`FIG. 1.—Biochemical changes in a group of white rats on a vitamin A—deficient diet.
`The animals were 22-24 days old when the diet was begun. The liver vitamin A began to
`fall at once and within 3 weeks had reached low values. Then within a week the blood
`level fell from normal to zero. With this, the rhodopsin content of the retina declined,
`marking tie onset of night blindness. Later the opsin also declined, marking the begin-
`ning of the histological deteri)rati )n of the retina.
`
`Vitamin A of Liver and Blood; Rhodopsin and Ops'in.—Figure 1 shows in one
`group of animals the effects of the deficient diet on the vitamin A content of the
`liver, the vitamin A concentration in the blood, and the rhodopsin and opsin of the
`retina. The values are expressed as percentages of normal. For the liver this
`means the percentage remaining of the vitamin A present in control animals at
`the time the diet was begun. The blood vitamin A, rhodopsin, and opsin are
`expressed as percentages of the values found in control animals of the same age.
`The liver vitamin A begins to fall as soon as the diet is begun and within 3 weeks
`has reached a very low value. This depletion proceeded at the average rate of
`2-2.5 pg. daily, the withdrawal rate for animals of this age and weight. Mean-
`While, the control animals on the complete diet increased their liver stores at an
`average rate of 45 ug. daily.
`It is this that makes the age at which the diet is
`
`
`
`652
`
`BIOCHEMISTRY: DOWLING’ AND WALD
`
`Pnoc. N. A. S.
`
`begun decisive for the course of the deficiency. Our control animals when 53 days
`old had livers weighing, on the average, 16 gm. and containing 1,360 ug. vitamin A;
`Withdrawn at a daily rate of even 5 ug., this might have tided them over 9 months
`of a deficient diet.
`
`The blood maintains its normal concentration of vitamin A (10.4 pg. per cent in
`the deprived animals, 11.2 pg. per cent in the controls) until the liver has been
`
`Wee/(5
`
`0
`
`4
`
`Rhodops/n
`/.09 t‘/vres/7a/d
`
`/00
`O
`
`5
`
`74
`I, /5
`
`[.09Lurrmance
`
`4.5 5.0
`
`_
`
`—.
`
`—h‘
`
`g
`
`_
`
`sacnsl
`
`zoofly E E
`
`-
`
`:-
`
`FIG. 2.~Effects of vitamin A deficiency on the electroretinogram (ERG).
`The top three lines show the number of weeks on the deficient diet, the rhodopsin
`content of the retinas as percentages of normal; and the logarithm of the lumi-
`nance of light needed to evoke a just perceptible ERG, the average threshold
`of normal animals being set arbitrarily at 1 (log threshold = 0). ERG’s are
`shown in response to a range of luminances of 5 log units, i.e., l to 100,000.
`Below each ERG, a marker shows the 1/59-second flash. The small rectangles
`at the bottom show a trace of the 60-cycle A.C. to indicate the time scale, and
`weekly calibrations of the oscillographic response to a pulse of 200 microvolts.
`The first two vertical rows of ERG’s show responses from a rat about to begin
`the diet and those from another after 4 weeks, when the vision is still normal.
`Thereafter, records are shown from a pair of animals each week. As the rhodop-
`sin declines, the ERG threshold rises (night blindness), and the ERG displays
`characteristic changes:
`(a) the b-wave at each level of luminance declines;
`(b)
`the a-wave declines still more rapidly; and (c) an inflection on the downward
`sweep of the b-wave is delayed longer and longer until it appears as a separate
`positive wave.
`
`emptied. Then in the space of a few days the blood vitamin A falls precipitately
`to zero.
`
`Up to this time the rhodopsin content of the retina remains normal. The extract
`of two retinas in 0.21—m1. solution possesses an extinction at 500 my of 0.280,
`corresponding to a rhodopsin content of 7.24 X 10*‘ umoles per retina. This is
`equivalent to a vitamin A content of 0.21 pg. per retina.
`(Lewis et al.” found
`only one-fifth to one-third as much vitamin A in the rat retina;
`the description
`of their preparative procedure suggests that it may have involved large losses of
`rod outer segments.)
`
`
`
`VOL. 44, 1958
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`653
`
`Now the rhodopsin also begins to fall and within 3 weeks has reached very low
`values. As we shall see, this marks the beginning of night blindness.
`There is a curious interval of 2-3 weeks in which, though the rhodopsin content
`has declined, the opsin level is still normal. That is, with the liver and blood emp-
`tied of vitamin A, the retina contains opsin which can find no vitamin A with which
`to combine.
`
`Then the opsin level, too, begins to fall, and this marks the beginning of the
`structural deterioration of the retina (see below). At this time also—the seventh
`and eight weeks of the diet—the classic signs of vitamin A deficiency appear.
`(In another of our experiments it took 10-11 weeks to reach this stage.) The
`
`FIG. 3.—Form of the ERG. Both the a- and b-waves are two-
`cusped. This feature appears also in the human ERG, where
`the shorter-latency component in each wave is believed due to
`cone responses,
`the longer-latency component
`to rods. The
`strong inflection evident on the downward sweep of the b-wave
`is delayed longer and longer as night blindness develops. Below
`the ERG, a marker shows the 1/50-second stimulating flash.
`
`animals lose weight rapidly; and by the end of the eighth week all 11ot already
`used in the experiments have died.“
`Physiological Changes.~—For the first 4 weeks on the diet, the animals appear to
`be physiologically normal. During this time, first the liver and then the blood is
`depleted of vitamin A.
`In the fifth week, as the rhodopsin level begins to fall,
`the visual threshold rises, marking the beginning of night blindness.
`Figure 2 shows electroretinograms recorded in pairs of experimental animals.
`At the top of the figure are shown the number of weeks on the diet, and below this
`the average rhodopsin content of the retinas as percentages of the normal Value.
`Below this is the logarithm of the threshold for a just perceptible retinogram;
`the
`normal threshold has been set arbitrarily at 1 (log threshold = 0), so that these
`
`
`
`654
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`‘PROC. N. A. S.
`
`numbers represent the rise in log threshold over the normal value. The rat’s
`eyes were exposed to 1/.,.,—second flashes at the log luminances shown on an arbitrary
`scale at the left. The figure shows a series of ERG’s at various luminances for
`each of two rats each week. Since no changes occur in the first 4 weeks, the first
`pair of records involves one rat just about to begin the diet and another rat that had
`been 4 Weeks on the diet.
`
`In the fifth week, the
`At the end of the fourth week the ERG is entirely normal.
`rhodopsin level falls to 74 per cent, and the visual threshold rises 1.15 log unit, or
`about 14 times.
`In the succeeding weeks, as the rhodopsin level continues to
`
`5.0
`
`E 2.0
`
`3'.’K
`Q 1.5
`‘la
`
`1.0
`
`aou
`
`Weeks on dfef
`
`FIG. 4.—Development of night blindness with time on the vitamin A—deficient diet.
`Each point shows the logarithm of the ERG threshold of a single animal, the normal
`threshold being set arbitrarily at 1 (log threshold = 0). The log threshold rises linearly
`with time on the diet, following a more rapid initial rise.
`
`decline, the threshold rises until, at the end of the eighth week, the rhodopsin is at
`16 per cent, and the visual threshold has risen about 680 times.
`Simultaneously, the ERG undergoes characteristic changes: (1) At all luminances
`but particularly at the lower ones, the height of the positive b-Wave declines as the
`deficiency progresses.
`So, for example, at log luminance 3.0 the b-wave, nearly
`maximal at the end of the fourth week, has sunk to the just perceptible threshold
`level by the end of the eighth week.
`(2) The negative mwave declines in amplitude
`still more rapidly.
`It is initially nearly as large as the b-Wave at log luminance 5;
`but by the end of the eighth week it can hardly be elicited at all, even at this highest
`luminance.
`(3) A small positive inflection, which appears initially only as a
`hump on the downward sweep of the b-wave, is delayed longer and longer as the
`
`
`
`VOL. 44, 1958
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`655
`
`deficiency progresses, until finally it has become a well-separated second positive
`wave, particularly evident at the higher luminances. The source of this delayed
`positive wave has not yet been identified. The possibility that it is an ofi’-efiect
`is not supported by our tests.
`_
`It may be noted in passing that some of the ERG’s show clearly that both the
`a- and the b-waves are two-cusped. This is particularly evident in such records
`as those of the sixth week at log luminance 5. One of this pair of ERG’s has been
`enlarged in Figure 3. Such two-cusped Cr‘ and b-waves have come to be associ-
`ated in the human retinogram with the responses of cones and rods, the shorter
`latency component in each wave presumably representing the cone response.“
`
`100
`
`8
`
`c~0
`
`
`
`Percenz‘rhodopsin NAOO
`
`O
`
`0.5
`
`1.0
`
`1.5
`
`2.0
`
`2.5
`
`3.0
`
`Log 2‘/ire sho/d
`
`FIG. 5.—Rise of the ERG threshold with decline in rhodopsin content in the
`retinas of vitamin A—deficient rats. The normal log threshold is set at 0.
`Rhodopsin is expressed as the percentage of that found in control animals of the
`same age. Over most of its course the log threshold rises linearly as rhodop-
`sin falls, following a disproportionately large rise of threshold with the first
`decline of rhodopsin.
`
`It has frequently been asserted that rats possess only rods; yet Tansleyl‘ and
`Walls” state unequivocally that cones are present, and Sidman“ has recently
`reported finding them in the approximate proportion 1 cone:10 rods. Since in
`the whole human retina the proportion of cones to rods is about 1:20, there may
`be as good anatomical basis in the rat as in man, for the two-cusped ERG’s to
`represent cone and rod responses.
`Figure 4 shows the relation between log threshold, and time on the deficient diet.
`After a somewhat abrupt start in the fifth week, the log threshold rises linearly.
`Similar behavior has been reported in man, the log threshold, rod and cone, rising
`linearly or nearly linearly for long periods on vitamin A—deficient diets.3' 7
`
`
`
`656
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`Paoc. N. A. S.
`
`Figure 5 shows the relation between the rhodopsin level of the retina and the
`logarithm of the visual threshold. Most of this relation again is linear, though
`distorted, as was Figure 4, by a somewhat disproportionate rise of threshold ac-
`companying the first decline of rhodopsin. When the rhodopsin has fallen to
`half its normal content as the result of the deficiency, the threshold has risen about
`100 times; when the rhodopsin is at 10 per cent, the threshold is up about 1,000
`times.
`It would be interesting to determine whether in normal rats the bleaching
`of rhodopsin that accompanies light adaptation causes similar changes of threshold.
`It is now recognized that in the human eye the bleaching of a small fraction of
`rhodopsin raises the threshold enormously ;19 by one estimate, the bleaching of 0.6
`per cent of the rhodopsin raises the threshold about 3,300 times.'~’°
`Histologic Changes.—The deterioration of retinal tissues in vitamin A—deficient
`rats has been described by Tansley9 and by Johnson. 1° Our principal problem is
`
`
`
`FIG. 6.—Effect of vitamin A deficiency on the structure of the rat retina. After 6 Weeks on a
`vitamin A—deficient diet, the retina appears entirely normal. After 8 weeks it has deteriorated
`markedly: only traces remain of the pigment epithelium;
`the layer of rods, particularly the
`outer segments, is disintegrating; and the external limiting membrane has disappeared.
`
`to orient these changes in the pattern of biochemical and physiological events that
`we have described.
`
`After 6 Weeks on the vitamin A—deficient diet, when the rhodopsin had fallen
`to about half its normal value but the opsin was still intact, the retinal histology
`of these animals appeared entirely normal (Fig. 6, left). After 8 weeks on the diet,
`however, when the opsin also had fallen to about half its normal value, the retinal
`. tissues had deteriorated markedly (Fig. 6, right). The outer segments of the rods
`were attenuated, many had a gnawed appearance, and they were irregularly spaced.
`Only vestiges of the pigment epithelium remained. The sharp boundary that
`marked the external limiting membrane was gone; and the blood vessels of the
`choroid layer were frequently occluded.
`Our supposition that, when opsin goes, the outer segments of the rods should
`deteriorate structurally has proved to be correct. By this time, however, the
`
`
`
`VOL. 44, 1958
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`657
`
`animal is deteriorating generally. Not only are other retinal tissues affected as
`just described, but the superficial structures of the eye now begin to display the
`classic signs of vitamin A deficiency: corneal clouding, xerophthalmia, and secre-
`tion of a sticky red exudate about the eyes. The animal is losing weight rapidly,
`the coat is disarranged, some animals have developed an unsteady gait, some
`breathe with difficulty.
`There is no compelling reason, therefore, to single out opsin among the animals’
`disabilities. To the degree that the loss of opsin is responsible for the histological
`decay of the outer segments of the rods, it may be only one of many proteins respon-
`
`Hours
`
`0
`
`18
`
`58
`
`64
`
`85
`
`[.09 threshold 2.05 1.35 0.95
`
`0
`
`0
`
`LogLuminance
`
`2oo,uv E 60C.p.s. H
`
`Fol-
`FIG. 7.—Recovery from night blindness on administration of vitamin A.
`lowing intraperitoneal injection of a large dose of vitamin A, ‘the
`threshold
`returns within 64 hours to normal (log threshold = 0). During this interval the
`ERG retraces in reverse all the changes which had accompanied the development
`of night blindness.
`
`sible for similar manifestations in many tissues. We shall have more to say of this
`below.
`
`Recovery from Night Blind1wss.—A number of animals which have developed
`night blindness were “cured” by administering vitamin A.
`It was found that the
`vitamin could be supplied more effectively by intraperitoneal
`injection than
`orally. Oral administration yielded irregular results, successful in some deficient
`animals, not in others.
`It seemed as though some deficient animals had lost
`temporarily the capacity to take up vitamin A fed by mouth.
`The effect of administering vitamin A to a night-blind animal is shown in Figure 7.
`At the beginning of the experiment this animal exhibited much the same ERG
`responses as did the animals of Figure 2 in the seventh week of the deficiency.
`Its
`ERG threshold was a little more than 100 times normal.
`
`
`
`658
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`Pnoc. N. A. S.
`
`This animal was given a large dose of vitamin A—-340 pg. dissolved in 1 ml.
`cottonseed oil,
`injected intraperitoneally. The ERG threshold slowly fell and
`within 64 hours had reached the normal level. Simultaneously, the ERG re-
`traced in reverse all the changes that had accompanied the development of night
`blindness:
`the a— and b-waves increased to their former sizes, and the delayed
`positive wave was reincorporated into the b-wave.
`Figure 8 shows the return of the ERG threshold to normal in a series of such
`recovery experiments. On the left are shown data from three deficient animals
`exhibiting various degrees of night blindness. Various amounts of vitamin A in
`cottonseed oil, ranging from 320 to 920 ug., were injected in 1-3 doses.
`In every
`
`9’.
`
`Q
`
`
`
`[.09threshold '8i3.
`
`All-tram
`
`O
`
`20
`
`40
`
`60
`
`80
`
`O
`
`20
`
`40
`
`60
`
`80
`
`Hours after Vitamin /1
`
`FIG. 8.—Recoveries from night blindness on administration of vitamin A. Log ERG thresholds
`measured at various times after intraperitoneal injection of large doses of all-tram: vitamin A
`(open circles), or neo-b (11-cis) vitamin A (closed circles). After injection of the all-trans isomer,
`the log threshold falls almost linearly to the normal level; but with neo-b vitamin A there is a lag
`in recovery, caused apparently by the necessity to isomerize this configuration primarily to all-
`trans before it is used by the eye.
`
`case the fall of log threshold to the normal level was approximately linear. The
`time for complete recovery depended primarily on the degree of night blindness,
`not the dosage level, at least within this range of high dosage.
`In one instance two rats, closely matched in their degree of night blindness, were
`injected with equal amounts (ca. 1,000 pg.) of two geometric isomers of vitamin A:
`the all-trdns form, which is most prevalent and which has been shown to be most
`effective in stimulating growth and liver storage in the rat," and the hindered cis
`neo-b isomer (11—cz's), which serves as precursor of the visual pigments. The
`result is shown at the right in Figure 8. The fall of log threshold on injection of all-
`trans vitamin A was, as usual, approximately linear;
`the neo-b isomer, however,
`
`
`
`VoL. 44, 1958
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`659
`
`exhibited a distinct initial lag in its effect, though after 30-40 hours the responses
`to both isomers ran approximately parallel.
`The vitamin A of the liver and blood was measured in these animals and in
`
`another pair which had been treated similarly. The animals injected with all-
`trans vitamin A had stored 71-87 pg. in the liver, in each case about 9 per cent of
`that injected, whereas those that received neo-b vitamin A had stored 32-39 pg., or
`about 4 per cent of that injected.
`Isomerization experiments conducted in both
`instances showed that none of the stored vitamin A was of the neo-b configuration.
`The blood levels after injection of either all-trans or neo-b vitamin A were normal
`(11.9 and 13.8 pg. per cent, respectively); again none of this was neo-b. Rhodopsin
`also was extracted from the retinas of the animals which yielded the data of Figure
`8 (right). The extinctions at 500 mu were 0.259 and 0.252, somewhat low, yet
`within the normal range.
`It appears, therefore, that neo-b vitamin A, when injected intraperitoneally, is
`converted to other isomers, primarily all-trans, before entering the eye.
`It is
`presumably this process of isomerization that causes the observed delay in its
`action and may be responsible also for its relatively low effectiveness in growth and
`
`storage."
`Where this isomerization occurs is not yet known; but it appears to be a general
`feature of vitamin A metabolism, for surveys in our laboratory of representative
`tissues in fishes, rats, and cattle have failed to find the neo-b isomer anywhere but
`in the eye.”
`It seems likely that only other isomers of vitamin A are carried in the
`blood or stored in the liver and that the eye tissues themselves isomerize one or
`several of these to the neo-b configuration for the synthesis of visual pigments.
`The bleaching of the visual pigments by light, on the other hand, yields all-trans
`retinene and vitamin A; so that the visual processes include a continuous cycle of
`geometric isomerization between these two configurations. 14
`At the left of Figure 8 two recoveries are shown from states of night blindness
`involving 2.5-3 log units rise of threshold above normal. Figure 5 shows that this
`degree of night blindness corresponds to the decline of rhodopsin to 10-20 per cent
`of normal and Figure 1 that, by this time, opsin should have fallen to about 60
`per cent of normal, with a corresponding disturbance of retinal histology. We can-
`not be certain that all these lesions occurred in the animals that yielded the data of
`Figure 8, but they were to be expected. The return of the threshold of these
`animals to normal 50 hours after administration of vitamin A implies that, Within
`this interval, not only had neo-b vitamin A again been made available but that the
`retina had regained whatever opsin it had lost and had repaired whatever tissue
`deterioration had occurred.
`It should be said at once that these animals had been
`
`selected because theircorneas were clear, since we wished to avoid the special
`complications of threshold measurement with cloudy corneas. The animals were,
`however, losing weight, and some of them exhibited the rough coats, red eye exudate,
`and postural imbalance already noted.
`In such instances, though a single large
`dose of vitamin A quickly brought the visual threshold back to normal, as shown
`in Figure 8, the external appearance was still poor. Several days elapsed before the
`weight began to increase, and it took several Weeks before they again looked sleek,
`though no more vitamin A had been administered.
`Conclusions.—These experiments reveal a simple and consistent series of changes
`
`
`
`660
`
`BIOCHEMISTRY: DOWLING AND WALD
`
`P300. N. A. S.
`
`directly related to the availability of vitamin A. On being deprived of vitamin A,
`the rat exhausts its store of this substance in the liver at a regular rate, meanwhile
`maintaining the levels in the other tissues. With the liver emptied, the blood
`level falls, and shortly afterward evidences of tissue deprivation appear. The
`first of these is night blindness—the visual pigments lose their prosthetic group,
`vitamin A aldehyde. Later the protein components of the visual pigments, the
`opsins, also decline; but this is a secondary phenomenon accompanied by general
`signs of tissue disintegration in the eye and elsewhere in the organism. On ad-
`ministering vitamin A, all these changes are reversed.
`For some years past we have had to face the embarrassment that the only function
`of vitamin A in the organism that we understand—that of supplying the chromo-
`phores of the visual pigments—p1ays only a trivial part in the whole complex of
`vitamin A deficiency. No animal-dies of nig