0022-202X/ 8 1 / 770 1-005 1 $02.00/ 0 TH E J OU RNAL Of' INVESTIGATIV E DERMATOLOGY, 76:5 1-58, 1981 Copy righ l © 198 1 by The Willia ms & Wilkins Co.
Vol. 77 , No. 1 Prill ted ill U. .A.
The Cutaneous Photosynthesis of Previtamin D3: A Unique Photoendocrine System
MICHAEL F. HOLICK, PH.D., M .D .
Endocrine Unit, M a.ssachusetts General Hospita.l, a.nd Ha.rvard Medica.l School, BostOIl, Massa.chusetts, a.nd Department of Nutrition a.nd Food Sciences, Massa.clwsetts Institute of T ech.nology, Cambridge, Massa.chusetts, U.S.A.
The skin has been recognized as the site for the sunmediated photosynthesis of vitamin D,, ; until recently, however, very little was known about either the sequence of events leading to the formation of vitamin D" in human skin or the factors that regulate the synthesis of this hormone. It is now established that, during exposure to sunlight, the cutaneous reservoir of7-dehydrocholesterol (principally in the stratum Malpighii) converts to previtamin D;I. Once this thermally labile previtamin is formed, it undergoes a temperature-dependent isomerization to vitamin D" over a period of 3 days. The plasma vitamin-D binding protein preferentially translocates vitamin D;j from the skin into the circulation. During prolonged exposure to the sun, the accumulation of previtamin D" is limited to about 10 to 15% of the original 7-dehydrocholesterol content because the previtamin photoisomerizes to 2 biologically inert photoproducts, lumisteroIa and tachysteroh. Increases in either latitude or the melanin concentration in the skin diminish the epidermal synthesis of previtamin D". A single total body exposure to 3 minimal erythemal doses of ultraviolet radiation increased the vitamin-D" levels in the serum almost lO-fold within 2 days and doubled the serum 25-hydroxyvitamin-D levels after 7 days. The unique mechanism for the cutaneous synthesis, storage, and steady release of vitamin D" into the circulation prompted an investigation into the potential therapeutic benefits of using the skin as the site for the synthesis and absorption of vitamin-D3 metabolites.
Since the dawn of man, the sun has been worshipped and revered for its life-sustaining powers. However, the importance of exposure to sunlight for health was not appreciated until the industrialization of northern Emope in the 17th century. Coinciden t with the appearance of the pall of industrial smoke and t h e congregation of people into densely populated towns, where the streets were narrow and poorly exposed to sunlight, was the recognition of a disease, known as rickets, associated with deformities of the skeleton in children [1-3]. As early as 1822, t he Polish physician Sniadecki [4] suggested that the dreaded " English disease" (also known as rickets) was due to the lack of exposm e to sunlight . Further evidence that this disease was promoted by environmental factors was provided in a report, in 1889, by the British Medical Association. A survey of the incidence of rickets in the British Isles showed that rickets abounded in thickly populated industrialized areas, but was
This work was supported in part by NIH grant # AM-25506-01 and NIH grant #AM-27334.
Reprint requests to: Michael F. Holick, Ph.D., M.D., Endocrine Unit, Massachusetts General Hospital, Boston, Massachusetts 02114.
Abbreviations: 7-DHC: 7-dehydrocholesterol 10',25-{OHh -7-DHC: 10',25-dihydroxy-7-dehydrocholesterol 10',25-{OH h-D,,: 10',25-dihydroxyvitamin 0:, 25-0H-D,,: 25-hydroxyvitamin D" MED: minimal erythema dose preD;j: previtamin D"
51
essentially absent in the rural districts [5]. A year later, Palm [6] reported on an epidemiologic survey of rickets in the British Isles and the Orient and observed that in Japan, China, and India, where people were poorly nourished a nd lived in s qualor, rickets was rare, whereas it was endemic in t he industrialized districts of Great Britain, one of the wealthiest nations in t he world. These clinical observations led him to urge t h e "systematic use of sunbaths as a preventive measure in rickets and other diseases." However, despite these ma ny insightful observations, by the t um of the 20th century m ost physicia ns and scient ists could not accept such a simple explanation for the etiology of this dreaded bone-deforming disease. In fact, by 1900 the 3 most popular t heories of the etiology of this disease included (a) an infectious disease, ( b ) a herita ble disorder, and (c) a nutritional deficiency.
In 1919, Huldschinsky [7] provided incontraver t ible evidence that exposure to radiation from a mercury-vapor quartz lamp alone was sufficient to t reat and cure rickets. Furthermore, he showed that this curative effect was not a local effect (because exposw'e of one arm had equal and dramatic curative effects in the non exposed arm as well), prompting th e suggestion that a compound was photosynthesized in the skin and transported into the circulation to act at distant s ites. Two years later, Hess and Unger [8] reported that infants with rickets were cured after simple exposur e to sunligh t. Hence, proof was finally demonstrated 3 decades after the first suggestion by Palm that exposure to sunlight alone had a curative effect on rickets.
CUTANEOUS PHOTOSYNTHESIS OF PREVITAMIN D ;l
At the same t ime that Huldschinsky [7] a nd Hess and Unger [8] were demonstrating that radiation from various a rtificial souJ'ces or the sun could cure r ickets, other investigators were developing rachitic animal models and s howing that rickets could also be cured by administration of cod liver oil [9] or by ultraviolet irradia t ion of th e food that was subsequently eaten [10,11]. The confusion over whether rickets was cured by direct exposure of the body to radiation from the sun (or a mercuryvapor quartz lamp) or by ingestion of a dietary factor was resolved when H ess [12] reported that rats fed ul traviolet (UV)irradiated human cadaver skin did not develop rickets, and Powers et al [13] demonstrated in rachitic }'ats that the healing effects were t he same after treatment with e ither radiation from a mercury-vapor quartz lamp or administration of cod liver oil. Collectively, these observations provided the basis for t he fortification of dairy products and other food stuffs with vitamin D, the consequence of which was responsible for the even tual eradication of this endemic disease.
Once it was established that the ant irachitic factor could be produced in vitro by ul t raviolet irradiation of plant and animal tissues, numerous investigators became involved in the structural characterizat ion of the antirachi t ic factor and its precursor. It was established that the an t irachit ic precursor was a 4-member ring sterol with 2 conjugated double-bonds at carbons 5 and 7. Initial characterization of the antirachitic factor (also known as vitamin D) established that ring B was opened at C9-C lO position (Fig 1) and that 3 double-bonds existed in a 5,6-cis-configuration [14,15]. X-ray crystal analYSis of a heavyatom derivative of vitamin D [16] provided evidence for its
52 HOLICK
24 26
HO
7-0EHYOROCHOLESTEROL
(PROVITAM IN 0 3 )
HO
ERGOSTEROL
(PROVITAMIN O2 )
25
27
VITAM IN 03
VITAMIN 02
FIG 1. Structures for 7-dehydrocholesterol, ergosterol, vitamin Oa and vitamin O2•
spatial orientation as shown in Fig 1. The designations vitamin "D/ ' and vitamin "D/' refer to differences in side-chain structure. Vitamin Da originates from the cutaneous sterol, 7-dehydrocholesterol (7~DHC), whereas vitamin D 2, which has a C24 -
methyl and a double-bond between carbons 22 and 23, originates from the fungal sterol, ergosterol (Fig 1).
With the discovery that exposure of skin to radiation from the sun or artificial-light sources led .to the production of vitamin Da, there was intensive investigation into which part of the sunlight spectrum caused this conversion and what steps led to the foxmation of vitamin Da. It was 'demonstrated in vivo and in vitro that, when human or animal skin was exposed to the ultraviolet-B (UV -B) portion of natural sunlight (spectral range 290-320 nm), the stores of 7-DHC in the epidermis converted ultimately to vitamin D a, but the precise sequence of events was I10t known [17,18). It was well established by 1950 that, when 7-DHC is dissolved in an organic solvent in a quartz vessel and exposed to UV-B radiation, it (a) absorbs a photon of radiation energy, which (b) causes cleavage of the C9- C IO
bond in the precursor, and (e) brings about the formation of the 9,lD-seeo-steroid, pre vitamin Da (preDa) [19]. Vitamin Da is not formed during this photochemical rel;lction, inasmuch as it is derived solely from previtamin D a (20). PreDa is thermally labile and undergoes a temperature-dependent rearrangement of its double-bonds to form the thermally stable 9,l0-seeosteroid, vitamin Da.
A number of investigators have demonstrated that vi.tamin D a can be isolated from animal and human skin after exposure to ultraviolet radiation [21-23]. There appeared to be at least 2 possible mechanisms resulting in cutaneous vitamin-Dol formation from the UV radiation: (a) 7-DHC in the skin converts efficiently and quantitatively directly to vitamin Da, or (b) the process in vivo is identical to that seen in vitro, ttlat is, 7-DHC absorbs a photon of radiation energy and is converted to previtamin Da. Once formed, preDa would then undergo a thermal isomerization to vitamin Da. In 1969, Rauschkolb et al [21] reported the isolation and identification of vitamin Da from human cadaver skin . Okano et al (22) and Esvelt, Schnoes, and DeLuca [23] saponified whole rat skin that had been exposed to ultraviolet radiation and prepared lipid extracts in which, after several chromatographic procedures, they also demonstrated the presence of vitamin D ol . Because saponification was used for the purification in each case, any thermally labile intermediate photoproducts such as preDa were isomerized.
Vol. 77, No.1
During the past 5 yr, my laboratory has been interested in the sequential events OCCUlTing in the skin during exposure to ultraviolet-radiation and culminating in the synthesis of vitamin Da. We chemically synthesized [3cx-:JH]7-DHC [24] and developed chromatographic techniques that would permit the direct analysis of skin lipids without any thermal isomerization. Initially, we topically applied [3cx-3H]7-DHC to previously shaven vitamin-D-deficient rats. Then, 24 hI' after application, groups of animals were exposed to 0.2 J /cm2 of UV radiation (obtained from an F40-T12 Westinghouse sunlamp broad-band UV-radiation 275-350 nm) and the irradiated skin was excised, extracted, and chromatographed. Lipid extracts from the skin of control nonirradiated animals contained 7-DHC and its reduction product, «holesterol. Skin from exposed animals exhibited radiolabeled 7-DHC, cholesterol, and a major photoproduct identified as preDa [24]. To be certain that this phenomenon was not due only to the photoconversion of radiolabeled 7-DHC on the stratum corneum surface but was also an occurrence within the epidermis, 40 vitamin-D-deficient rats were exposed to 0.2 J / cm2 of UV radiation, and the whole skin was immediately excised and extracted in organic solvents. All purification steps were carried out at 4°C to ensure that any thermally labile photoproducts would not be altered during the extraction and chromatography procedures. After numerous chromatographic procedures, preDa was isolated in pure form and identified based upon (a) mass spectrum, m/e 384 (M+), 271 (M-sidechain) , 253 (271- H 20), 136 (ring A + (HC) 6, (HCh, and (H2Ch9, and 118 (136-H20), (6) UV-absorption spectrum Amox260 and A01;,,230 nm, which is characteristic for the 6,7-cis-triene chromophore for pre vitamin Da, and (e) subsequent thermally controlled conversion to vitamin Da as demonstrated by spectral shiJt in the UV absorption from Amux260 to A01;n265 nm (Fig 2). Furthermore, an analysis of the same lipid extract did not demonstrate the presence of any vitamin Da, and thus ruled out the possibility that vitamin Da was being made simultaneously with the preDa (25).
CUT ANEOUS SYNTHESIS OF VITAMIN Da
Once we established that, in vivo, when skin is exposed to UV radiation, cutaneous 7-DHC was photolyzed to preDa, the intriguing question of whether there was any physiologic advantage to this process was explored. We examined some of the physiologic implications of this unique cutaneous photobiologic process, including (a) the temperature dependence of preD3
conversion to vitamin Da, (6) the transport of preD3 and/ or vitamin Da from the skin into the circulation, and (c) the distribution of sites of synthesis of preDa within the various layers of epidermis and dermis. The time course for the thermal isomerization of preDa to vitamin D a in vitro and in vivo at various temperatures was first determined [26]. Incubation of preDa in methanol at O°C for 1 week caused less than 2% formation of vitamin Da, whereas, at 25° C, approximately 50%
PREVITAMIN D3 ,QC. ,---,"'13"6c---0::-:.4:---A-. - -=2-::6-::0---=-----------,
118
40.
w u z « ~02 o en CD «
o WAVELENGTH (nm)
M/E
M' 384
FIG 2. Mass spectrum of previtamin D, isolated fTom rat skin after it was thermally converted to vitamin 0 ". Ultraviolet absorption spectrum of previtamin D" isolated from the rat skin (insert).
July 1981
of the preD:) cO!1verted to vitamin Do in 48 h1" (Fig 3B). When preD3 was incubated in methanol at 37°C, approximately 50% of preD3 converted to vitamin D3 by 28 hr, and equilibrium was reached after 4 days with approximately 80% of preD:J converting to vitamin D 3. In contrast, in vivo (the sill'face temperature of rat skin was determined with a thermal couple to be 36.5 ± 0.5°C, and the dermis, 37.5 ± 0.5°C), 50% of the 1"adiolabeled preD3 that was topically applied 01" generated by exposing [3a-3H]7-DHC to UV-B radiation converted to vitamin Da in 18 hr, and, by 3 days, there was a 95:5 ratio of vitamin D3 to preD3
remaining in the skin (Fig 3B) . We next examined the possibility that there was a transport
system that would preferentially remove vitamin D3 from the skin, leaving behind preD3 to continue its thermal isomerization to vitamin D j . The binding affinity of the plasma vitamin-D binding protein (an a,-globulin important for the transport of vitamin D :) and its metabolites in the circulation) for preD3 and vitamin D 3 was determined [26]. Figure 3A illustrates that the vitamin-D binding protein has essentially no affinity for preD3
compared with the affinity for vitamin D:1• The potential physiologic advantage of this selective binding is summarized in Fig 4. Only after thermal conversion of preDa to vitamin D3 is this seeo-steroid removed from the skin by binding to the vitaminD binding protein in the capillary bed of the dermis.
To investigate the location in the skin where preDa is synthesized during sunlight exposure, we used 2 separate techniques that separate layers of epidermis without destroying the integrity of the cells [26]. Surgically obtained hypopigmented Caucasian skin (from the lower limb) , with subcutaneous fat removed, was cut into samples 6.25 cm2 and separated by either incubation with 20 mg/ml of staphylococcal exfoliatin (a substance that specifically cleaves the epidermis at the stratum granulosum/stratum spinosum interface) or immersion in a 60°C water bath for 30 sec (this technique cleaves epidermis at the stratum-spinosum/stratum-basale interface) (Fig 5). After
0 z 640 CD ...,
0 I 020
"' N
'I' .!2...
* 100
80
0"" 60 z ~
~ 40 5 ;,-'1
20
A 7 - Dehydrocholesterol PreD3
- " .
0'- 0-
1.0
B
I /
l
/
10 100 1000 NANOGRAMS
/ 0 -- '-0 in'l.i.l£ll 36.5-37.5°C
0 /
/ .. ~ / ~ ...... ..... in I'i1ro 25°C
/
'A" " " "" •....... .
/ ~ .... .. ....
.:
!D Vit ro O°C o ___ _ t::. __ _ ___ _ _______ _ _ t::.. ____ _ __ 6
01234567 DAYS
FIG 3. A, Thermal conversion of preDa to Da as a function of time in vitro at O°C (-6 -), 25°C ( . . ·6 · · .), 37°C (-e-), and in vivo at 36.5- 37.5°C (-0 -) (Each time point represents 2 experimen ts that were each determined in triplicate. There was exceUent agreement for each of the data points with less than 2% variation). B, Displacement of ["H]25-hyd.roxyvitamin D" (["H]25-0H-D:,) from ra t D-binding prote in by vitamin D:.I (Da), previtamin D:, (preD:,), and 7-dehyd.rocholesterol (7-DHC) (Each data point represen ts an average of 2 sepa.rate determinations). Reproduced from Holick et a l [26] with permission. (Copyright 1980, American Associat ion for the Advancement of Science.)
PHOTOSYNTHESIS OF PREVITAMIN D3 53
SUN
HO
-~ tt! osb'."o ~ "" -
~ ~ /
7- DEHYOROCHOI..ESTEROI.. Hf PREVITAMIN 03
SKIN JrSK/N TEMP
~ LIVER
Flc 4. Diagrammatic representation of t he formation of previtam in D" in the skin during sun exposure and its subseq uent thermal conversion to vitamin D", which in t urn binds to the plasma vitamin-D binding protein from transport in to t he circulation.
A
8
LOCATION OF PREVITAMIN ~
Staph To)":n
~SC ~q~SG
SM
- HPLC
' .. ' .. , _ HPLC , Js
.
B
.
". DERMIS
'I . " ,- . .,. _.,.- ..........
FIC 5. A , Diagrammatic illustration of t he separa t ion of the stra tum corneum and stratum granulosum from the rest of t he epidermis and dermis by toxin treatment. B, Diagrammatic illustration of the separation of t he skin · into the epidermis minus stratum basale; and the stratum basale and dermis by a heat separation technique.
either treatment, but before physical separation of the skil) layers, the stratum-corneum side of the skin samples was exposed to 0.5 J / cm2 ofUV-radiation with a narrow band centered at 295 nm with a lO-nm half-band width. Control skin samples were kept in an UV -free environment. Immediately after irradiation, the toxin-treated skin was separated into a top layer (stratum corneum + stratum granulosum) a nd the bottom layer (stratum spinosum, stratum basale, and dermis), whereas the heat- treated skin was sepal'ated into a top layer (stratum corneum, stratum granulosum, stratum spinosum) and the bottom layer (stratum basale and dermis). The basal cells of the stra-
54 HOLICK
tum basale were collected by mechanically scraping them off the dermis (Fig 5) . Contents of all skin layers were confIrmed by histologic examination. The separated layers of skin were extracted with 8% ethyl acetate in n-hexane for 24 hr at -20°C, dried under nitrogen and weighed, and a portion of each sample was chromatographed by high-pressure liquid chromatography in duplicate to determine 7-DHC and preD 3 concentrations. As demonstrated in Table I, even though all of the epidermal strata and the dermis contain 7-DHC, the highest concentration of 7-DHC per mg of lipid was in the stratum basale. The largest amount of this sterol per unit surface of skin, however, was present in the stratum spinosum. When human hypopigmented Caucasian skin was exposed to UV radiation, photosynthesis of preDa occurred throughout the entire epidermis and to a small extent in the dermis, with the highest concentration of preD" in the stratum spinosum ano. stratum basale.
Collectively, these data indicate that the photosynthesis of preD" in the epidermis during exposW'e to the sun is advantageous. When hypopigmented Caucasian skin is exposed to sunlight, UV-B radiation penetrates the skin and causes the photochemical conversion of 7-DHC to preD:! throughout the entire epidermis and in the dermis. The regions of the skin that potentially have the highest capacity for preD3 formation per unit area are the stratum spinosum and the stratum basale. However, the a mount of preD" that is formed in the various layers will be dependent on the amount of UV -radiation reaching each layer. Immediately upon its formation, the previtamin begins to isomerize by a nonphotochemical rearrangement of its triene system, the rate of which is dictated by the temperature of the skin at or near the stratum basale, to vitamin 0,1. This process permits the skin to continually synthesize (from preD;I) and release vitamin Da into the circulation for up to 2 to 3 days after a single exposure to sunlight. Once vitamin D;l is formed, the vitamin-D binding protein translocates it preferentially into the circulation and assures the efficient convel'sion of small quantities .of preDa to vitamin D :l by shifting the previtamin D :J ~ vitamin D:l equilibrium to the right at the time when equilibrium is being approached (Fig 4).
FACTORS THAT LIMIT THE CUTANEOUS PHOTOPRODUCTION OF PREVITAMIN Da
In 1922, Hess [27] showed that skin pigmentation in rats influenced tRe antirachitic potency of sunlight. This observation correlated with the clinical observation that the darker the skin, the greater the susceptibility to rickets in childhood, especially during seasons or in environments in which sunlight was reduced [28]. In 1967, Loomis [29] popularized the theory that melanin pigmentation not only limited cutaneous vitamin-0,1 formation but that it was most imporfant for controlling the
TABLE 1. The concentration of 7-dehydrocholesterol (7-DHC) and pre vitamin D:, in the various strata of a. hypopigmented Caucasian
whole shin sample after exposure to 0.5 J/cm2 of UV radiation"
7-DHC 7-DHC previ-Skin Strata (ng/rng (ng/6.25 tamin 0 ;,
lipid) em' ) (ng/6.25 ern' )
Stratum corneum + s tratum 180 360 20 gra nulosum
Stratum spit10sum 630 2459 114 Stratum basale 1720 1892 113 Dermis 88 1670 3
" The concentrations represent the average of 2 determinations from the same thigh skin sample. Analysis of t he 7-DHG concentration in the various 'skin strata from skin samples obtained from different areas of the body or from different individuals suggests that the absolute amount of 7-DHC may vary but that the stratum basale contains the highest concentration of 7-DHC pel' mg of lipid and the stratum spinosum has the high est content of 7-DHC per unit surface area. Reproduced with the kind permission of the publisher [26]. (Copyright 1980, American Association for the Advancement of Science)
Vol. 77, No.1
amount of vitamin D3 synthesized in the skin. He speculated that heavily pigmented individuals living at or near the equator were protected from vitamin-D toxicity principally because the skin pigment absorbed UV-B photons. He further speculated that as mankind moved north and south of the equator, hypopigmentation evolved to permit adequate quantities of UV-B photons to penetrate the epidermis to permit the synthesis of adequate amounts of vitamin D 3 .
For our investigation of the role of pigmentation, latitude, and other factors in the cutaneous synthesis of preD:J, we exposed hypopigmented and hyperpigm~nted human skin for various times to simulated solar UV radiation. We then determined the photoproducts that resulted from 7-DHC in order to study the sequential photochemical events quantitatively, and examined the regulator processes, if any, and the role of melanin and latitude as modifiers of the production of preD3 in the skin [30]. Surgically obtained hypopigmented (type III) and hyperpigmented (types V and VI) human skin specimens obtained
~
u ~ '0 0
~ 0 '6.
~ 0
u I 0
,..:.
""
w u z « CD a:: o Ul CD «
100
\
90 i '. \
BO \
70
60
50
4 0
30
E
I \
\
p'
\
A B
1-QHC 7-DHC
RE TENTION TIME (min)
Tropical Solar Simulation
\
\ \ .0
" . . '1
.0"""
"
Time (J10ur s)
.... 1iL
'X 7 - 0HC
~ PreDo
{; T
FIG 6. High-performance liquid chromatographic profiles of a lipid extr act from the basal cells of sUl'gically obt.ained hypopigmented skin that was previously shielded from (A) or exposed to tropical simulated sola r UV radiation that reaches the Earth at sea leve l at noon for (B) 10 min; (C) ] hI'; and (D) 8 hr. E represents a n analysis of the photolysis of 7 -dehydrocholesterol (7- DHC) (_ ..... -) in the basal-cell laye r and the appearance of the photoproducts previtamin 0 " (preD,,) (-e-), lumisterol:, (L),j--O -- ), and tachysterol:, (T) (-6 -) with increasing exposure time. 6 Represents the SEM from 3 determinations. Reproduced with the kind permission of the publisher [30].
July 1981
from differe nt areas of the body with subcutaneous fat removed were cut into samples (6.25 cm2
) and immersed in a hot-water bath at 60°C for 30 sec. The skin was blotted dry, and the stratum-corneum side was exposed at various times to simulated solar UV-radiation. Conditions were adjusted to approximate UV-radiation that reaches the Earth in June at sea level at either the equator (0° Latitude) or in Boston (42°,20' North) at noontime. Immediately after irradiation, the skin was separated into its component parts (see Fig 5). Control skin samples maintained in an UV-free enviJ"Onment for the same period of time were similarly separated. Figure 6 illustrates the representative chromatographic proflles of lipid extracts of the stratum basale, separated from surgically obtained hypopigmented human skin samples of type III, that were exposed for various times to simulated tropical solar UV-radiation. After 15 min of exposure, the principal photoproduct in the stratum basa le was previtamin D :j • Skin samples subjected to longer exposme time showed the appearance of 2 additional photoproducts (lumisterol and tachysterol), but, as seen in Fig 6-c, -d, -e, tachysterol formation reached the maximum at 5% of the original 7-DHC concentration by 1 hI', whereas lumisteJ"OI formation steadily increased to 50% by 8 hI". Correlated with the increase in photoproduct concentration was the decline in 7-DHC to 30% of its original concentration by 8 hL We next compared t he exposure time necessary to maximize preD:I formation in hypopigmented Type-III and hyperpigmented types-V and VI skin. As pigmentation increased 6'om type III to types V and VI, the exposw'e time necessary to maximize preD:1 formation increased from 30 min to 1 hr and 3 hr, respectively (Table II) . However, regardless of skin type, once preD:1 maximized and plateaued at about 15% of the original 7-DHC concentration, further expOSUl'e to UV radiation produced only increases in
TABLE II. The effect o[ melanin concentration and latitude on the [ormation o[ preuitamin D" and lumisterol3 in hum.an shin a[ter
irradiation with simulated solar UV radiation.
Skin type(s)
Type III" Type V" Type VI" Type III" Type m"
Simulated geographic location
Equator" Equator Equator Equator Boston"
Time (h]') of exposu re to s imulated soIaI' UV radiaLio n to Ilulximize preD;1 for-
mation
0.50-0.75" 1.00-1.50 3.00-3.50 0.25-0.50 0.50-1.00
% Lumist.erob fO rillation after 3-hr exposure of s imulated solar UV ra-
diation
40 20 14 48 30
" Surgically obtained foresk in specimens from healthy males (26-36 yr old).
" Surgically obtained thigh skin from a 38-yr-old male. c A #WG-305-C Schott filter (Schott, Optical Glass Inc., York Ave
nu e, Duryea, PA 18642) was used to simulate noon sunlight at the equator at sea level. Based on the United States Government Standard (USGS) that skin exposed to sunlight on a cloudless day for an 8-h1' period will be irradiated with approximately 175 J /cm2 of UV-A and 4 J/cm2 UV-B.
" A #WG-320-B Schott filter was used to simulate Boston sunlight at noon in June at sea level at a zenith angle of 20° . Based on the USGS report that skin exposed to sunlight on a cloudless day for 8-hr period will be irradiated with approximately 140 J /cm2 of UV-A and 2.4 J /cm2
UV-B. • Skin samples were exposed to simulated solar UV radiation using
a 2.5-kw xenon arc lamp with a dichroic mirror and an appropriate filter. Skin was kept on ice during irradiation to avoid overheating. Because of the high intensity of this light system, exposure times to simulated solar UV radiation were .significantly decreased compared with natural conditions. For example, the amount of time it took to deliver 3-3.5 hI' of simulated equatorial solar UV radiation was approximately 10 min. Each value represents an average of 3 sepal'ate determinations. This was excellent agreement for each value, with less than 10% variation. Reproduced with the kind permission of the publisher [30]. (Copyright 1981, American Association for the Advancement of Science.)
PHOTOSYNTHESIS OF PREVITAMIN D.1 55
lumisterol and tachysterol. Exposure times necessar y to maximize preD:\ formation in hypopigmented type-III skin when it was exposed to simulated solar UV radiation at the equator or in Boston ar e shown in Table II and indicate the effect of latitude on preD3 formation. The decrease in UV radiation from the equator to Boston (due to the increase in the zenith angle of the sun) increased t he exposure t ime required to maximize preD3 formation [3~).
Our data suggest that the photochemical conversion ofpreD3 to lumisterol and tachysterol is the major limit ing factor in preventing vitamin-D:I intoxication due to a single prolonged expOSUl'e to the sun. Our theory is supported by the fact that there aTe no documented cases of vitamin-D toxicity due to prolonged exposure to the sun. For example, if a hypopigmented person and a heavily pigmented person aloe exposed to 3 hr of UV radiation at the equator (Fig 6 and Table II) , the amount of preD:\ formed is essentially the same, i.e., about 15% of the initial 7-DHC content in the epidermis. The major difference is that the photoproduction of the biologically inert photoisomer, lumisterol, in hypopigmented skin is significantly greater than it is in heavily pigmented skin [3D). Melanin competes with 7-DHC for UV -B photons, and therefore an increase in melanin in human skin also increases the t ime of exposure to UV radiation t hat is needed to maximize preD:1 formation (Table II).
Therefore, when we aJ°e exposed to minimum amounts of sunlight, our cutaneous stores of 7-DHC convert to preD". PreD" can either thermally isomerize in the skin to vitamin D :l 01' photochemically isomerize to lumisterol a nd tachysterol (Fig 7). During prolonged exposure to the sun, the synthesis ofpreD.l reaches a plateau at a concentration of about 10-15% of t he original cutaneous 7-DHC concentration, and preD:1 photoisomerizes to 2 biologically iner t products, lumisterol and tachysterol. Tachysterol formation also plateaus at a concentration of abou t 5%, whereas lumisterol formation cont inues to increase with increased exposure to solar UV radiation. An analysis of the binding affini ty of the vitamin-D binding protein for lumisterol and tachysterol has demonstrated that t here is essentially
HO R 7- DEHYDROCHIXESTERIX
SKIN
BLOW
DBP
FIG 7. This is a schematic representation of the formation of previtamin D3 in the skin during exposure to the sun, its thermal isomerization to vitamin D", wh ich is specifically translocated by the vit;a ll'linD-binding protein (DBP) in to the circulation. During continual elCposure to the sun, previtamin Do also photoisomerizes to lumisterob and tachysterola, which are photopl'Oducts that al'e biologically inert (i.e. , they do not stimulate intestinal calcium absorption). Becau e the DBP has no affinity fOl- lumisterob but has minimal affini ty fOJ'tachystet'ol;" the translocation of these photoisomers into the circulation is negligible and these photoproducts are sloughed off during the natural turnover of the skin. Because these photoisomers are in a state of quasiphotoequilibrium as soon as previtamin D3 stores al'e depleted, (due to thertnal isomerization to D,,), exposure of lumisterol and tachysterol to UV radiation will promote these isomers to photoisomerize to preD". Reproduced with the kind permission of the publisher [30].
56 HOLICK
no affinity for lumisterol and minimum affinity for tachysterol, thus making unlikely the translocation of these photoisomers into the circulation (Fig 7) . Therefore, these inert photoproducts would be sloughed off during the natural turnover of the skin . Furthermore, because these photoisomers are in a quasistationary state with preD:J, as soon as preDa stores are depleted due to their thermal isomerization to vitamin D 3 ,
exposure of cutaneous lumisterol and tachysterol to solar UV radiation could promote the photo isomerization of these products to preD~ (Fig 7). Therefore, in order of importance, the significant determinants limiting cutaneous previtamin-Da production appeared to be (a) photochemical regulation, (b) pigmentation, and (c) lati tude [30].
VITAMIN-D METABOLISM
During t he past decade-it has been clearly demonstrated that vitamin Da either ingested in the diet or photosynthesized in the skin is biologically inert. It must undergo 2 successive hydroxylations, first in the liver to 25-hydroxyvitamin D3 (25-OH-Da), and then in the kidney to 1a,25-dihydroxyvitamin Da (la,25-(OHb-D3) before it is able to carry out its 2 main physiologic functions: The stimulation of intestinal calcium absorption and the mobilization of calcium from the bone [31-34]. T herefore, based upon the facts that (a) vitamin D3 is not an absolute nutritional requirement, inasmuch as it can be synthesized in the epidermis when skin is exposed to adequate amounts of sunlight, and, furthermore, that (b) vitamin D3 must be transported to 2 separate organs for metabolism before it can carry out its biological effects at distant organs, vitamin Da is considered a' hormone rather than a vitamin. Very little information has been available on the blood levels of vitamin D3 as a result of exposure to sunlight. However, because the concentration of 25-0H-D 3 (which is the major circulating metaboli te of vitamin D 3 ) in the serum .correlates with dietary intake of vitamin D 3 , many investigators have measured the circulating levels of this metabolite during a val'iety of environmental conditions and demonstrated that serum 25-0H-D levels reflect directly upon the cutaneous syn thesis of vitamin D 3 . Preece et al [35] and Fairney, Fry, and Lipscomb [36] reported a marked lowering of 25-0H-D levels in the serum of submariners and of one Antarctic explorer during sunless periods, wher~as Haddad and Chyu [37] reported that there is almost a doubling of 25-0H-D levels in the serum of lifeguards on the beach. There is now strong evidence in both children and adults that serum 25-0H-D levels vary with the season, with peak levels occurring in the summer when exposure to the sun and the intensity of UV radiation are maximized, whereas the lowest levels occur in the midwinter when exposure to the sun and the intensity of the UV radiation are minimized. The contribu tion of the cutaneous photoproduction of vitamin Da in preventing osteomalacia and rickets relative to the dietary intake of vitamin D2 or vitamin D3 h as been a subject of much examination. In the British Isles, where supplementation of food with vitamin D is not practiced, the cutaneous photosynthesis of preD3 is quite important. In the United States, however, dietary supplementation is widely practiced, and the role of cutaneous preD3 synthesis for the maintenance of calcium homeostasis is less clear. Recently, Neer et al [38] reported on the. concentrations of 25-0H-D in the serum of groups of adult males who either worked primarily indoors or primarily outdoors in 6 cities across the United States, determined at inter vals of 6 weeks after the summer and winter solstic'es and soon after the vernal and autumnal equinoxes. Although all the serum levels of 25-0H-D were within the normal range of between 15 and 80 ng/ml at every season, the blood 25-0H-D levels were 10 to 15 ng/ ml lower in the indoor workers. The conclusion from this study was that the significant determinants of blood 25-0H-D levels in normal adult men are: geography > individual variation> occupational exposure to sun> season of the year.
Morita et al [39] reported the serum concentrations of vi-
Vol. 77, No.1
tamin Da in a group of J apanese who went on vacation to the South Sea Islands for 2 weeks. After 2 weeks of intermittent sunbathing, the vitamin-D values had increased by 92%, and the 25-0H-D values had increased by 45%, whereas the serum 1,25-(OHb-D values did not change significantly. Recently, we exposed both vitamin-D-sufficient and vitamin-D-deficient individuals to a quantitative amount of ultraviolet radiation and determined sequential changes in the serum vitamin-D, 25-0HD, and 1,25-(OHb-D concentrations (J. S . Adams, T. L. Clemens, and M. F. Holick, unpublished results). In healthy young Caucasian adults whose entire bodies we~e exposed to 3 minimal erythema doses (MED) of broad-band UV radiation, the serum vitamin-D values increased by 7- to lO-fold within 24 ill" and then declined to basal values by one week after the initial exposure (Fig 8) . In contrast, the serum 25-0H-D concentrations, after a gradual rise, increased to a maximum of approximately two-fold by 2 to 3-weeks after exposure (Fig 8). The serum 1,25-(OHh -D values also increased by about 20%, although the values remained within normal range. Surprisingly enough, in the patient with vitamin-D deficiency who had received a total-body exposure to 1 MED of UV radiation, the serum 1,25-(OH)2-D concentrations markedly increased from 25 pg/ml before irradiation to 260 pg/ ml 6 days after the exposure, and returned to baseline 3 weeks later.
PHOTOPRODUCTION OF VITAMIN-D3 METABOLITES IN THE SKIN
The detection of specific hereditary or acquired defects in vitamin-D metabolism has provided new insights into the pathophysiology of certain disorders of calcium and phosphorus homeostasis and has focused interest on the clinical use of chemically synthesized vitamin-D metabolites and analogs for the treatment of these disorders [31-34]' Although administration of the active form of vitamin D 3, 1a,25- (OHh-D3, has proved effective in th e treatment of many vitamin-D-resistan t syndromes, such as chronic renal failure, hypoparathyroidism, and vitamin-D-dependent rickets, Type I, occasional hypercalcemia is a problem perhaps related to the high transient local concentrations of the biologically active metabolite in the small intestine. Recent studies on the unique features of the skin for
3 med , 60
E "-co
.0 .5-Cl
20
E 60 "-co .5-Cl '0
:x: 0 0{) 20 N
E "-co 2 Cl
-!:' :x: 2
I 0{) 20
N_
o 2 4 6 8 10 12 14 21
TIME (days)
FIG 8. Serum vitamin D, 25-0H-D, and la,25-(OHh-D levels in a 24-yr-old Caucasian fe male before and after receiving a total body exposure to 3 minimal erythemal doses of UV rad iation.
July 1981
controlling t he synthesis of vitamin D3 suggested that the skin may also serve as a site for the application or synthesis, 01' both, of vitamin-D metaboli tes and would have pharmacologic advantages a nd be preferable to the effects achieved by the oral administration of vitamin-D3 metabolites. To test this hypothesis, we chemically synthesized raruolabeled 10:,25-dihydroxy-7-dehydrocholesterol (l0:,25-(OHh-7-DHC) and topically applied it to vitamin-D-deficient rats to determine whether expOSUl"e to UV radiation would convert t his hydroxylated 7-DHC precursor to the preD:J derivative (Fig 9). Lipid extracts from the skin of rats that received a topical application of r ad iolabeled 10:,25-(OHh-7-DHC either with 01' without UV radiation demonstrated that only the animals that received UV radiation showed t he conversion of 10:,25-(OHh-7-DHC to 10:,25-dihy-
:i ~ SKIN NO UV
A 6
1o<,25-(OH1 2 -7-0HC 5
SMALL INTESTINE
NO UV
C
tf'Q 4 ,
4
:><
:::< a. £:)
'" 'Q
:><
:::< a. £:)
3
2
~:i A B
SKIN +
4 uv 1 ~ .25-(0~ 1 2 - PRE 03
3 lo:.25;(OH12 -03
2 1o<.25-(OHl -7-DHC
t 2
3
2
6
5
4
3
2
SMALL INTESTINE
+ UV
D
o 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
FRACTION NUMBER (2 ml) FRACTION NUMBER (2 ml)
F IG 9. Sephadex LH-20 chl"omatographic profiles of lip id extracts of tissu es from rats t hat received a top ical app lication of [24-"H]-lLf,25-(OHh-7-DHC. (A) Skin and (C ) small-in testine extracts from rats that were exposed to ultraviolet radiation, and (B) skin and (D) sma Uintestine extracts from rats that were exposed to ul traviolet phototherapy. Reproduced wi th t he kind permission of t he publisher [40].
80
o'! ~ 7.0
~ 60 ~ ("j 50
:2 ii! 40 w en
o
A
T 14 21 28 35
DAYS AFTER DOSING
FIG 10. In testina l calcium transport response (A) and serum calcium levels (B) in vitamin-D-deficient rats that received either a topical application of 20 fll of 95% EtOH (-e-), topical app lication of 1 flg of la,25-(OHh-preD" (-6-), topical application of I /lg of l Lf,25- (OHhD a (-.&-), or an ora l dose of 1 flg of lLt,25-( OHh-D" (-0 -). Reproduced with the kind pe rmission of t he publisher [40].
PHOTOSYNTHESIS OF PREVITAMIN Da
UV-8
1 •. 2 5 - (OH)2 -7 - DHC 1 •• 25-(OH)2- PRE - D3
jfSKIN TEMP
"'-' SKIN OH
I p~ 1 • . 25- (OH)2- D3
~-.. DBP Ia. 25- (OH)2D3
/ "'-SMALL INTESTINE BONE
57
F IG 11. D iagrammatic representation of the photochemical conversion of l a,25-(OHh-7-DHC to la,25- (OH)"-preD, in the skin and t he subseq uent thermal conversion of t he previtamin to 1. (t,25-(OHh-D,,; once form ed, lLf,25-(OH)z-D:J is bound to the plasma vitamin-D-binding protein (DBP) for transport through the circulation to its tru'get tissues. Reproduced with t he kind permission of the publisher [40].
droxyprevitamin D3 (10:,25-(OH)2-preD3). We further demonstrated that cutaneous 10:,25-(OHh-preD:J thermally isomerized to 10:,25-(OH)2-D:J, which t hen entered the circulation and local ized in its target t issues: the small intestine and bone [40). We next measUl"ed intestinal calcium transport and serum calcium responses in vitamin-D-deficient anephric rats that received a topical dose of 10:,25-(OHb-7-DHC and were either UV -irradiated or shielded. Rats t hat received either a topical application of l a,25- (OHh-7-DHC without rad iation therapy or a topical application of 50 Ilg of vitamin D:] showed no elevation in intestinal calcium transport or serum calcium when compared with the control group that received only a topica l application of 95% ethanol. However, the anephric rats t hat received a topical dose of l a,25-(OHh-7-DHC and UV-phototherapy demonstrated a marked stimulation in intestinal calcium transport and a subsequent parallel elevation in serum calcium [40].
To study t he long- term biological consequences .of topical application of l a,25- (OHh-7-DHC combined with UV-phototherapy, vitamin-D-deficient rats were given a topical application of 1 Ilg of lcr,25-(OH)2-preD3 (the photoproduct of l a,25-(OHh-7-DHC) to determine serum calcium concentration and intestinal calcium transport a nd to compare the values in similar rats given (a) topical application of a vehicle (95% EtOH), (b) topical application of 1 Ilg of lcr,25-(OH)2-D:], or (c) an oral dose of 1 Ilg of 10:,25-(OHh-D". Animals that had received an oral dose of la,25-(OHh-D:J had a very rapid elevation in intestinal calcium absorption and a parallel rise in serum calcium (Fig 10) that reached a maximum at 12 hr and decreased rapidly to control values by 2 weeks. Rats that received a topical dose of 1 Ilg of 10:,25-(OH)2-D;] or 1lY.,25-(OHh-preD" showed gradual increases in in testinal calcium absorption that were sustained for 14 to 21 days, respectively, before returning to control values. Therefore, the proven historical success of phototherapy for the treatment of rickets along with our experiments that demonstrate the feasibility of using t he skin as the site for the synthesis and absorption of vitamin D metabolites (Fig 11) offers a potentially attract ive a nd easily- ma naged al ternative for t he treatment of vitamin-D-resistant disorders.
CONCLUSION
R ickets evolved into a devastating human affli ction because most of the scientists a nd physicians in the 19th century were
58 HOLICK
unwilling to appreciate the healthful effects of exposure to natural sunlight. Finally, in the 1920's, it was clearly establish e d that vitamin D a was photosynthesized in the skin a nd that this h ormone was responsible for the maintenance of normal skele tal growth a nd developme nt. U ntil recently, h owever, little was known about t h e sequential even ts involved in t h e cutaneous synth esis of vita min D :] and abou t t h e factors t h at r egulated this process. It is now establish ed t h at when skin is initially exposed to solar r a diation , the stores of epidermal 7-DHC convert to preD3. PreDa can e ith er t h ermally isom erize to vitamin D 3 or undergo a photochemical isomerization th at limits the cutaneous synth esis of preD" during prolonged exposure to the sun . The development of (a) n ew chromatographic procedures t h at permit a direct a nalysis of cutaneous 7-DHC and its photoproducts, (b) sensitive assays for vitamin D 3 and its metabolites, a nd (c) new a pproach es for treating v itamin-D-resistan t syndrom es with the cuta neous synth esis a nd absorption of vitamin-D3 metabolites should stimulate new interest in th e basic a nd practical aspects of the cutaneous photobiology of v itamin D a.
REFERENCES
1. Foote JA: Evidence of rickets prior to 1650. Am J Dis Child 34:443-452, 1927
2. Griffenhagen G: A brief history of nu trit ional diseases. II. Rickets. Bull Nat! Inst NutI' 2:9, 1952
3. Guthrie D: In, A History of Medicine. Lippincott, Philadelphia, 1964, p 69
4. Sniadecki J: Cited by W Mozolowski: Jedrzej Sniadecki (1768-1883) on the cure of rickets. Nature 143:121, 1939
5. Owen I: Geogj'aphical distribution of rickets, acute and subacute rheumatism, chorea, cancer, and urinary calculus in the British Islands. Br Med J 1:113-116,1889
6. Palm T A: T he geographic distribution and etiology of rickets. Practitioner 45:270-279,321-342, 1890
7. Huldschinsky K: Heilung von Rachitis durch Kunstliche H ohensonne. Deutsche Med WochenschT xiv:712- 713, 1919
8. Hess AF, Unger LJ: CUTe of infantile rickets by sunlight. JAMA 77: 39, 1921
9. Mellanby E. An experimental investigation on rickets. Lancet 1: 407-412, 1919
10. Hess AF, Weinstock M: Antirachit ic properties imparted to inert fluids and green vegetables by ultraviolet irradiation. J Bioi Chem 62:301-313, 1924
11. Steenbock H: The induction of growth-promoting and calcifying properties in a ration exposed to light . Science 60:224- 225, 1924
12. Hess AF:"I'he antirachitic activation of foods and of cholesterol by ultraviolet irradiation. JAMS 84:1910-1913, 1925
13. Powers GF, Park EA, Shipley PG, McCollum EV , Simmonds N: T he prevention of rickets in the rat by means of radiation with the mercury vapor quartz lamp. Proc Soc Exp Bioi Med 19:120-121, 192 L
14. Fieser LD, F ieser M: Steroids, chap 4, Vitamin D. Reinhold, NY, 1959, pp 90-168 .
15. Havinga E: Vitamin D, example and challenge. Experientia 29: 1181-1193, 1973
16. Crawfoot D, Dunitz JD: Structure of calciferol. Nature 162:608-610, 1948
17. Steenbock H, Black A: The production of growth-promoting and calcifying properties in a ration by exposure to ul traviolet light. J Bioi Chem 61:408-422, 1924 .
18. Hess AF, Weinstock M, Helman DF: The antirachi tic value of . irradiated phytosterol and choleste rol. J Bioi Chem 63:305- 308,
1925 L9. Velluz L, Amiard G, Petit A: Le precalciferol-ses relations
'd'equilibre avec Ie calciferol. Bull Soc Chim Fr 16:501-508, 1949
Vol. 77, No.1
20. Schl\ltmann JLM, Pot J, Havinga E: An investigation into the interconversion of precalciferol and calciferol and analogous compounds. Rec Trav Chim Pays-Bas 83:1173-1184,1964
21. Rauschkolb EW, Winston D, Fenimore DC, Bilack HS, Fabre LF: Identification of vitamin D:J in human skin. J Invest Dermatol 53:289-293, 1969
22. Okano T, Yasumura M, Mizuno K, Kobayashi T: Photochemical conversion of 7-dehydrocholesterol into vitamin D3 in rat skins. J NutI' Sci Vitaminol 23:165-168, 1977
23. Esvelt RP, Schnoes HK, DeLuca HF: Vitamin D;] from rat skins irradiated in vitro with ultraviolet light. Arch Biochem Biophys 188:282-286, 1978
24. Holick MF, Frommer JE, McNeill SC, Richtand NM, Henley JF, Potts JT Jr: Photometabolism of 7-dehydrocholesterol to previtamin D" in skin. Biochem Biophys Res Commun 76:107-114, 1977
25. Holick MF, Richtand NM , McNeill SC, Holick SA, Frommer JE, Henley JW, Potts JT Jr: Isolation and ident ification of previtamin D" from the skin of rats exposed to ultraviolet irradiation. Biochemistry 18: 1003-1008, 1979
26. Holick MF, MacLaughlin JA, Clark MB, Holick SA, Potts JT Jr. , Anderson RR, Blank IH, Parrish JA: Photosynthesis of previtamin D3 in human skin and the physiologic consequences. Science 210:203-205, 1980
27. Hess AF: New aspects of the rickets problem. JAMA 78: 1177-1183, ~~ .
28. Levinsohn S: Rickets in the Negro: Effect of treatment with ultraviolet rays. Am J Dis Child 34:955- 966, 1927
29. Loomis F: Skin-pigment regulation of vitamin D biosynthesis in man. Science 157:501-506, 1967
30. Holick MF, MacLaughlin JA, Doppelt SH : Factors that influence the cutaneous photosynthesis of previta min Do. Science, 211:590-593, 1981
31. Holick MF, Potts JT Jr: Vitamin D, chapt 351, Harrison's Principles of Internal Medicine. KJ Isselbacher, RD Adams, E Braunwald, RG Petersdorf, JD Wilson. New York, McGraw-Hill Book Company, 9th ed, 1980, pp 1843-1849
32. Haussler MR, McCain T A: Basic and clinical concepts related to vitamin D metabolism and action. N E ngl J Med 297:974-983, 1041-1050, 1977
33. DeLuca HF, Schnoes HK: Metabolism and mechanism of action of vitamin D. Annu Rev Biochem 45:631-642, 1976
34. Norman AW, Henry H: 1,25-Dihydroxycholecalciferol-A hormonally active form of vitamin D3. Recent Prog Horm Res 30:431-480, 1974
35. Preece MA, Tomlinson S, Ribot CA, Pietrek J, Korn HT, Davies DM, Ford JA, Dunnigan MG, O'Riordan JLH: Studies of vitamin D deficiency in man. Quart J Med 44:575-589, 1975
36. Fairney A, Fry J, Lipscomb A: The effect of darkness on vitamin D in adults. Postgrad Med J 55:248- 250, 1979
37. Haddad JG, Chyu KJ: Competitive protein binding radioassay for 25-hydroxycholecalciferol. J Clin Endocrinol Metab 33:992-996, 1971
38. NeeI' R, Clark M, Friedman V, Belsey R, Sweeney M, Buonchristiani J , Potts JT Jr: Environmental and nu tri tional influences on plasma 25-hydroxyvitamin D concentration and calcium metabolism in man, Vitamin D: Biochemical, Chemical and Clinical Aspects Related to Calcium Metabolism (Proceedings of the Third Workshop on Vitamin D). Edited by A W Norman, K Schaefer, JW Coburn, HF DeLuca, D Fraser, HG Grigoleit, Dv Herrath. New York, Berlin, Walter de Gruyter, 1977, pp 595-606
39. Morita R, Dokoh S, Fukunaga M, Yamamoto I , Roizuka 1: Effects of sunlight exposure on blood concentration of vitamin D derivatives in healthy Japanese adults, Vitamin D: Basic Research and its Clinical Application (Proceedings of the Fourth Workshop on Vitamin D). Edited by A W Norman, K Schaefer, Dv Herrath, HG Grigoleit , JW Coburn, HF DeLuca, EB Mawer, T Suda . New York, Berlin, Walter de Gruyter, 1979, pp 165-168
40. Holick MF, Uskokovic M, Henley JW, MacLaughlin J, Holick SA, Potts JT JR: The photoproduction in skin of lex,25-hydroxyvitamin Do: An approach to therapy of vitamin-D-resistant syndromes. N E ngl J Med 303:349-354, 1980