hormonal stimulation of melanogenesis in tissue culture · phology of these cells is variable. some...

HORMONAL STIMULATION OF MELANOGENESIS IN TISSUE CULTURE*

FUNAN HU, M.D. AND WALTER CHAVIN, PH.D.

The classic concept of the role of intermedin inthe evocation of pigmentary phenomena was firstdeveloped by Zondek and Krohn in 1932 (1).Recently, interest in the chromatophorotropic ef-fect of ACTH has led to the conclusion that con-tamination of various ACTH preparations withintermedin produces clinical hyperpigmentation(2). However, the inconsistent action of inter-medin in melanogenesis has been demonstratedby many workers (3). This in part may resultfrom the close structural similarities of ACTHand the intermedins which have precluded theircomplete chemical separation. Thus, the so-calledintermedin effects of ACTH preparations andACTH effects of intermedin preparations mustbe evaluated in a completely comparable andparallel series of experiments. In addition, thepresence of endogenous pituitary hormones andother humoral factors in both intact experimentalanimals and humans, has complicated the inter-pretation of biological assays and clinical findingswith ACTH and intermedin. Thus, for the criticalevaluation of the purported role of intermedin inmelanogenesis, it is of paramount importance tomaintain a uniform endocrine environment inwhich various ACTH and intermedin prepara-tions are to be tested. Fulfillment of these require-ments has been the basic concept of the presentstudy.

Recently, intermedin has been found to consistof two polypeptides, a-intermedin and /3-inter-medin (4), both of which have been evaluated inthis report. In addition, as Pickford and Kosto (5)have suggested that prolactin also may affectmelanogenesis, the action of this hormone hasalso been evaluated in tissue culture. The presentstudy demonstrated that a-intermedin, /3-inter-medin, and prolaetin fail to stimulate melano-

* From the Department of Dermatology (Clar-ence S. Livingood, M.D., Director), Henry FordHospital; and the Department of Biology (Contri-bution Number 43), Wayne State University,Detroit, Michigan.

This investigation was supported in part by aresearch grant CY-3345(C1) from the NationalInstitutes of Health, Public Health Service.

Received for publication August 22, 1959.

377

genesis in goldfish integument grown in tissueculture, whereas only ACTH stimulates melano-genesis under similar conditions.

The pigment cell terminology of this report isthat utilized at the Third and Fourth Conferenceon the Biology of Normal and Atypical PigmentCell Growth. Their definitions are as follows:

Melanocyte: a melanin-producing and melanin-containing cell.

Melanophore: a pigment effector cell in loweranimals.

Melanogenesis, melanin formation, and mela-nin synthesis: increase in the amount of mel-anin and refer neither to the darkening of anypre-existing melanin granules nor to the re-distribution of melanin granules in the pig-ment cells.

Melanin dispersion and melanin aggregation:rearrangements of preexisting melanin gran-ules in a pigment-containing cell (e.g. mel-anophores) so that these granules are distrib-uted throughout the cytoplasm (dispersion)or localized in a compact perinuclear mass(aggregation).

To clarify the discussion in this report, theoriginal terminology applied to the hormone origi-nating from the pars intermedia of the pituitary,intermedin, has been retained. Other names havebeen applied to this polypeptide, for example, aterm which is currently widely used, melanocyte-stimulating hormone or MSH.

METHODS AND MATERIALS

The details of roller tube tissue culture technicemployed in this study have been described byRu ci al (6) for the in vitro cultivation of humanskin. The eaudal fins of completely xanthie (non-melanotic) goldfish were excised, washed as de-scribed below, cut into pieces approximately1 x 2 mm in dimensions, and embedded in a plasmaclot on a rectangular eoverslip (12 x 50 mm).These explants were placed in culture tubes con-taining 2 ml of a medium consisting of humanaseitic fluid, diluted chick embryonic extract andGey's balanced salt solution. These cultures weremaintained at a constant temperature of 37° C for72 hours.

378 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

To inhibit bacterial cnntamination andbacterial growth, the pieces of fin were washed intwo changes of Gey's balanced salt solution con-taining 166 meg neomycin sulfate per ml. Thenutrient supernatant fluid also contained thesame concentration of antibiotic.

Freshly prepared solutions of the hormones wereimmediately incorporated in both clot and super-natant fluid. The hormone preparations employedand the hormone concentrations utilized areindicated in Table I, II, III, IV, V and VI. Inaddition, ACTH (146RS3) was subjected toenzymatic, acid and alkali hydrolysis as indicatedin Table VII, and then tested as above.

To obtain reliable results, the experiments wererepeated 3 to 5 times, each utilizing 4 to 6 differentfish. Control cultures were made from each fishin all experiments. Data on control cultures wereomitted from the tables for the reported findingsregarding induced melanogenesis were alwaysmade in comparison to the controls. The caudalfin of each fish was used only once to avoid possiblevariation after fin regeneration. Thus, approxi-mately 150 fish were used to provide the materialfor more than the 7000 explants cultured in thepresent study.

Various methods of microscopic examinationwere employed: (a) phase contrast study and time-lapse cinematography of living cultures, (b)dopa reaction (7), (c) nnstained cultures aftermethyl alcohol fixation, and (d) May-Gruenwald-Giemsa (8) stained preparations.

The presence of melanogenesis was ascertainedas follows: (1) whenever melanin containing cellswere seen in hormone treated cultures, while nomelanin containing cells occurred in control tubesfree of hormone, or (2) when there were definitelymore pigment containing cells in hormone treatedcultures as compared with controls. In addition,the size, degree of ramification, and intensity ofpigment color (as a gauge of their melanin pro-ducing ability) of the melanocytes in treated anduntreated cultures were studied.

RESULTS

Cytological Studies

The growth of goldfish fin explants in tissueculture is rapid and luxuriant. After twenty-fourhours, a sheet-like epithelial outgrowth is obvious,and with time increases in size. Partial retractionand liquefaction of the clot and the formation ofsmall empty spaces in various areas of the out-growth usually occur in from 48 to 72 hours; thecells of the outgrowth appear normal and healthyfor at least another 24 hours.

The outgrowth is predominantly epithelial in

nature, however, scattered fibroeytes occur be-tween the epithelial cells. In hormone treated cul-tures and in those few control cultures in whichoccasional pigment cells are visible, the melano-eytes are usually found immediately adjacent tothe explant and only rarely move into the out-growth itself (Fig. 1). The melanoeytes which arefound in the outgrowth are smaller than thosefound in the explant itself. The details of the cy-tological findings are presented below.

(A) May-Gruenwald-Giemsa Preparations. Thisstain demonstrates the extent of epithelial out-growth well. The four types of cells seen in theoutgrowth are:

1. Epithelial cells in the form of sheets (Fig. 1).2. Fibrocytes scattered in between the epithe-

hal cells or in thin sheets (Figs. 1, 2).3. Presumptive melanocytes which are small

rounded cells, whose morphology and stain-ing reactions resemble those of young pig-ment cells found in cultures of human skin(9). The cells most commonly have short,stubbed processes, a scanty amount of occa-sionally granular cytoplasm and an oval orround, intensely red staining nucleus. Thenuclei of epithelial cells and fibrocytes staina definitely less intense red color (Fig. 2).

4. Large melanocytes were found in large num-bers in ACTH treated cultures and onlyrarely found in control cultures. The mor-phology of these cells is variable. Some arebipolar with long slender processes (Fig. 2);others are stellate and highly branched (Fig.3), while still others are broad and plaque-like. Although their nuclei are not muchlarger than those of the epithelial cells, theentire melanocytes including processes, areseveral times larger than either epithelialcells or fibrocytes. In addition, multinuele-ated melanocytes may often be present. Thedelineation between melanocyte and melan-ophore was not attempted for this was notconsidered pertinent to the present study.

(B) Unstained Cultures. Because of their intra-cellular black melanin granules which contrastwith the unstained background, the melanocytesare readily visible in unstained cultures (Figs. 4,5). As indicated above, these cells are usuallyfound on and immediately adjacent to the ex-plant. These cells represent the population ofmelanocytes which contain pigment granules.Since the propigment cells do not contain melaningranules, they are not visualized in this type of a

MELANOGENESIS IN TISSUE CULTURE 379

Fio. 1. Caudal fin culture. ACTII treated. May-Gruenwald-Giemsa. 100 X. Outgrowth of epitheliacells in sheets with fibroeytes, melanocytes and small round cells scattered in between. Note the largeheavily pigmented polydendritic cells in the dark, solid explant portion.

preparation. These fixed, but unstained prepara-tions, are more satisfactory for the examinationof melanocytes than any method used. Therefore,in the later experiments, only this type of prep-aration was utilized.

(C) Dopa reaction. The dopa reaction has noadvantage over the above technics; on the otherhand, it has several disadvantages. Firstly, inprocessing the cultures during this cytochemicalreaction, the explants frequently loosen from theclots and are lost. In addition, most of the melano-eytes following ACTII treatment contain melaningranules, and it is impossible to determinewhether the intracellular granules are black priorto or after exposure to the dopa reagent. Thus, itis not possible to accurately define a positive dopareaction. In addition, dopa positive cells do notoccur in the control cultures. Consequently, undersuch conditions the dopa reaction does not con-tribute more information regarding new melaninformation. Because of these technical drawbacks,the dopa reaction was omitted from the later ex-

periments.(D) Phase contrast examination of living cul-

tures. All previously described cell types are visi-

ble in the living state. In addition, the followingcells are seen:

1. Lipophores. These cells are visible only inthe living state. Their yellow to red earote-noid pigment is soluble in organic solvents;hence their color is lost following fixationand dehydration. As these cells are not per-tinent to the present study further discus-sion of them is omitted.

2. Motile cells. Careful examination reveals theamoeboid movement of actively motile cells.This motility is obvious with use of time-lapse cinematography. Such cells containeytoplasmic inclusions of either rather large,refraetile globules or small dark granules(Fig. 6). Such cells are not always identifi-able in fixed preparations.

Endocrine Studies

The effects of ACTII, intermedin and prolaetinupon melanogenesis and the appearance of mel-anoeytes are summarized in Table II, III, IV, Vand VI. The ACTII concentrations tested varyfrom 2 X 1O to 21.6 IL' per ml of medium. Theconcentrations of intermedin tested range from

— ..


FIGS. 2 and 3. Caudal fin culture. ACTH treated. May-Gruenwald-Giemsa. 430 XFIG. 2. One long melanocyte with a "foot-plate-like" structure at end of one dendrite. Many small

round cells with distinct deeply staining nuclei and little cytoplasm. A few of these cells contain melaningranules, and appear to be young melanocytes. Epithelial sheet and a few fibrocytes are seen in thebackground.

Fio. 3. Three melanocytes at edge of explant. The largest one is seen in the center with one longprocess extending into the outgrowing epithelial sheet.

t .,4 r1) 4

S * I .• *

44 •t•wa •

I-' irI.

•;S r

I S'I

S. S


Fios. 4 and 5. Caudal fin culture. Fixed and unstained. Culture treated with ACTH, 0.02 lU/mimedium for 72 hours.

FIG. 4. Numerous spindle, stellate, and polydendritic melanocytes on the explant. Note the blackintraeytoplasmie melanin granules in good contrast with the unstained light background. 100X

FIG. 5. High magnification of Fig. 4. 430 X

4.

& _

_'•)

-.-.

r

• a

•

'I'..

4-

1 a


Fio. 6. Living untreated culture of goldfish eaudal fin. Phase contrast microscopy. 430 X. Epithelialsheet forms the lighter background. The darker cells are motile cells with either (1) large round refractilebodies, or (2) fine dark granules.

0.1 Frog skin units (10), FSTJ, to 1.7 x 106 FSUper ml of medium. As indicated in Table II, mela-nin formation occurs at the minimal intermedinconcentration of 1.4 x 10 FSU/ml. On the otherhand, the minimal amount of ACTH necessary toinduce melanogenesis is 2 x l0 IIJ/ml. As theconcentrations of these hormones are obviouslyrepresented in different biological units, it is notpossible to compare them. In addition, as thesehormones do not occur in a constant ratio in eachof the various preparations utilized in this study,it is only possible to compare their actions at vari-ous concentrations depending upon the prepara-tion used.

The data indicate that intermedin concentra-tions are not directly correlated with melanogene-sis. For example, melanoeytes appear in culturesmade with an ACTH preparation (146RS3) con-taining 2 x 10—I IIJ ACTH and 60 FSU inter-medin per ml (intermedin being present as im-purity in the ACTH preparation), andconsistently in concentrations of 300 FSIJ inter-medin and 1 x 102 Ill ACTH. While withanother intermedin preparation (D216-155C)containing 1.4 x 10 FSU intermedin and 2 x10 Ill ACTH per ml (ACTH being present as

impurity in the intermedin preparation) melaninformation does not occur at all times. Therefore,if intermedin is the hormone responsible for mela-nin formation in goldfish, it is difficult to reconcilethe fact that melanogenesis fails to uniformlyoccur after the use of 1.4 x 10FSIJ of a differentpreparation, an increase of 233 times the inter-medin concentration. Further, the results withthe two intermedin preparations of Armour Lab-oratories demonstrate this point rather well(Table IV). Melanogenesis occurs at 1.7 x 10FS1IJ intermedin (AL 1037) containing 2.8 x 10IIJ ACTH, and at 1.4 x 10 FSIJ intermedin(D216-155C) with 2 )< 10 II] ACTH. At theseconcentrations the latter preparation more fre-quently showed a lesser melanogenie action thanthe former, thus again correlating ACTH concen-trations, and not intermedin concentrations, withmelanogenesis.

The melanogenie activities of two intermedinpreparations, LR-I and D216-155C, (Tables IVand V, respectively) are comparable. However,LR-I reveals no detectable ACTH activity asindicated by ascorbic acid depletion method,while D216-155C reveals 2 115/mg of ACTH ac-tivity (Table I). The significance of this finding

MELANOGENESIS IN TISSUE CULTUEE 383

TABLE IAssay data of hypophyseal hormones used in tissue culture studies of goldfish skin

•Eormnne Preparatinn

AnimalOrigin

•Supplier Intermedin (MSII)

FSU/mgiACTHlu/mgi

Luteo-trnpie

Activity.Other Activities

Unit/mg

ACTH:146RS3 Porcine Armour 1 X lOi 3.6 ? trace of thyro-907-100-A42 (Corti- Porcine Armour 2 X 10 100.0 tropin or gonad-

eotropin A) otropinec99H'BiD4 Ovine Gesehwind

& Li1 X 10—1 X 10 78.0

L2821A (ce-Cort.) Ovine Gesehwind 1 1< 10—1 X 10 80.0& Li

Interme din (MSH):D216-155C Porcine Armour 1.4 X 10 2.0AL 1037 Porcine Armour 6.0 X 10 0.1 Histamine activityIntermedin Porcine Lee & 1.0 X 10 No data 0.5 gamma/mg

(MSH)6 LernerLH-I7 (fl-inter- Bovine Gesehwind 1—2 X 10 0

medin, fl-MSH) & LiCCD4 (fl-inter- Porcine Gesehwind 4.0 X 10 0.001

medin, fl-MSH) & Lia-intermedin (a- Porcine Lee & 6.0 )< 106 No data Vasopressin ac-

MSH) Lerner tivity

Prolactin:L2651AP Ovine Geschwind

& Li2.0 X 10 0.02 30—35

JU: International Unit based on ascorbic acid depletion.2 FSU: Frog Skin Unit as defined by Shizume et at, 1954.Dr. IV. F. White of Armour Laboratories informed us thnt tbis sample had not been assayed for

intermedin activity, but several comparable materials had been assayed in Dr. A. B. Lamer's labora-tory and all values were approximately 2 X 10 FSU/gm.

This preparation has not been assayed for intermedin activity, but according to Dr. I. I. Gesehwindthe activity is probably between 10 to 10 FSU/gm.

Alkali treated.Purified by acetic acid extraction, ether purification, oxycellnlose adsorptions and a short counter-

current distribution in 2-BuOH-0.05% triebloroacetic acid.A counter-current purified intermedin preparation with no detectable ACTH activity.

is of particular interest for it is now recognizedthat ACTH has several actions which cannot beevaluated by the ascorbic acid depletion assay.Thus, the melanogenic activity of ACTH may notdepend upon the ACTH potency as indicated bythis assay technic.

An j3-intermedin preparation, CCD4, whichhas ACTH activity of less than 1 x 10 lU/mg,is very different from LR-I in regard to melano-genie activity. The former preparation is active inintermedin and ACTH dilutions far below thosewhich were expected to show melanogenic activity(Table V). Again, these findings are explicable

only if the ascorbic acid depletion potency haslittle correlation to the melanogenic activity ofACTH.

The experimental findings utilizing a porcineintermedin prepared by Lee and Lerner (TableI), indicate that the melanogenic activity is com-parable to that of D216-155C. A preparation ofa-intermedin also prepared by Lee and Lerner,is somewhat more active than both D216-155Cand LR-I (Tables IV and Y). It was purified byearboxyl methyl cellulose column chromatogra-phy, contained a small amount of vasopressin butapparently was free of fl-intermedin.


TABLE IIThe effects of AUTH and intermedin upon melano genesis in tissue culture of goldfish shin

ACTH 145R53 Intennedin D215-155C

Interntedin FSU/ml ACTHI ID/mi Meianogenesis1 Intermedin FSU/mi ACTH ID/mi iMeianogenesis

6X1053 X l03X1046X1033X1036X1023X1026Xl06

21.6 +10.8 +1.08 +

2X10—1 +1X101 I +2X102 +1x102 +2X1082x104 —

1.4X1062.8X1051.4X105l.4X1041.4X103

2X1tr24X1032X1032X1042X101

+++

—

1 See Footnotes 1 and 2, Table I.2 + Indicates the presence of melanin-containing melanocytes in 90—100% of the cultures. The

melanocytes are usually larger, more darkly pigmented and possess more branched dendrites thanthose in the control cultures.

Indicates the presence of melanocytes in about 50% of the cultures.— Like untreated control cultures.

Alkali treatment of intermedin preparationspotentiates the intermedin effect while destroyingat least the contained ascorbic acid and ketogenieactivities of AUTH. An alkali treated intermediri,AL 1037, like the a-intermedin, appears to bemore active than D216-155U, although the latteris also alkali treated. This difference, however, isnot very great and probably is within the rangeof technical linutations. It may be of little or nosignificance (Table IV).

Surprisingly, the melanogenic activities of thethree highly purified AUTH preparations are notas great as that of the less purified preparation(146R53). All three purified hormones producepositive results at intermedin concentration of 6x 10 FSIJ/ml. According to data obtained fromearlier experiments (Table II), intermedin at thisconcentration is not expected to have melano-genie activity. However, on the basis of AUTHpotencies, the ascorbic acid depleting activitiesof the three purified hormones arc at least 200times higher than that of the less purified ma-terial. As these hormones were obtained from twodifferent laboratories, the method of preparationmay be a differential factor. However, this maynot be an adequate explanation, for the two prep-arations from the Armour Laboratories also yielddifferential findings, as the highly purified porcinepreparation (907-109-A4) appears less effective in

stimulating melanogenesis than tbe less purifiedporcine material (146R53). This result appearsinexplicable in terms of either intermedin orAUTH. For example, in comparing these AUTHpreparations, dilutions are made so that the con-centration of intermedin (as impurity) in each is6 X 10 FSTJ/ml. If the melanogenic effect isattributable to AUTH alone, a greater degree ofpigment formation should be expected in thehighly purified preparation (at the above inter-medin level, 4.0 III AUTH per ml). Melanogene-sis is observed in both preparations, but the de-gree of melanosis invariably is more extensive andthe pigment containing cells more numerous inthe less purified preparation (at the above inter-medin level, 0.2 LU AUTH per ml). The findingsare consistent when a series of concentrations ofboth preparations are tested (Table III). Theapparent decrease in melanogenic activity ofhighly purified AUTH preparations also occurs intwo AUTH preparations supplied by Drs. Ge-schwind and Li (Table III). There appears to bea definite loss of melanogenic activity duringpurification of AUTH. This loss is not detectableby the usual bioassay technic which is based uponthe depletion of adrenal ascorbic acid. It is pos-sible that AUTH may lose some of its activities,i.e. mclanogenic factor, during the process ofpurification while retaining its adrenal ascorbic

TA

BL

E I

II

Com

para

tive

actio

n of f

our

AC

TH

pre

para

tions

upo

n m

elan

c gen

esis

in t

issu

e cu

lture

146R

S3

907-

109-

A4

cc99

H'B

eD

L28

21A

Inte

rmed

in'

Mel

ano-

In

term

edin

A

CT

H l

U/m

i M

elan

o-

Inte

rmed

in

FSU

/ml

AC

TII

' lU

/mi

gene

sis2

FS

U/m

l ge

nesi

s FS

U/m

I A

CT

H r

n/m

i M

elan

o-

Tnt

erm

eclin

A

CT

H r

n5i

Mel

ano-

ge

nesi

s FS

TJ/

ml

gene

sis

6X10

5 21.6

+

3X10

5 10.8

+

1—10 )< 1O

7.8

+

1—10

X

10

8.0

+

6X10

3 4.0

+

z 2.

5—25

)( 10

2.0

—

3X10

4 1.

08

+

1—10

X 1

0 7.

8 X

10

z 6X

102

4X10

' 6

X 1

0 2 X 1O

+

3 X 10

2 x 1

0—'

—

2.5—

25 X

10

2 X

10—

1 —

3X10

3 1X10'

+

1—10

X 1

0 7.

8 X

10

_s

6X10

4X102

6 X

102

2

X 1

02

+

3 X

10

2 x

102

2.5-25

2 X 102

—

3X10'

1X10

' +

1—

10

7.8

x i0

6

X 1

0 2

X 1

0 3

2 X

10

0.25

—2.

5 2

X 1

0 1—

10 X

10

7.8

X 1

0'

6 2X

104

—

1—10

X 1

0 7.

8 X

10

1 Se

e fo

otno

tes

1 an

d 2,

Tab

le I

. 2

See

foot

note

2,

Tab

le I

I.

0i


TABLE IVComparative action of three intermedin preparations upon melano genesis in tissue culture

D216-155C j. 1037 Inteemedm (Preparationof Lee & Leener)

Intermedin'FSU/ml ACTH' lU/mi Melano-

genesis2Intermedin

FSU/ml ACTH lU/mi Melano-genesis

IntermedinFSU/ml ACTH lU/mi Melano-

genesis

1.4X1052.8X1011.4X1051.4X1041.4X101

2x1024X10—12X10—'2X1042X105

++++—

1.7XlOt3.4X1051.7X1013.4X1041.7X1011.7X1031.7X1011.7 X 101.7

2.8X10—'5.6X10°2.8X10—15.6X10°2.8X1032.8X1042.8X10—52.8 X 102.8 x 10

+++++++—

—

1.3X103

1.3X101.3X1041.3X10'

Nodata +

+—

1 See footnotes 1 and 2, Table I.2 See footnote 2, Table II.

TABLE VComparative action of a- and 13-intermedin upon melano genesis in tissue culture

Lu-I (fl-intermedin) CCD4 (fi-intermedin) ee-Intermedin (Lee & Lernee)

p-intermedin' FSU/mlACTH'lU/mi

Melann-genesis2

$-intermedinESU/mi ACTH lU/mi Melano-

genesisa-intermedin

FSU/ml ACTH lU/mi Melano-genesis

1—2X101

1—2X104

1—2Xl021—2X1021—2X10

0

0

0 I

00

++

—

—

8X1054X1012X1018X1044X1048X1014X1038Xl02

2X1041x10—45X1052X1011x1012X1051X1052x107

++++++++

6X104

6X101

6X101

6X1066X10'

Nodata ++

—

—

—.

1 See footnote 1 and 2, Table I.2 See footnote 2, Table II.

acid depleting potenty. Of course, other pituitaryhormones may also be involved here. Completeassay data on all the preparations used are notavailable. As indicated in Table I, AL 1037 has ahistamine activity of 0.5 gamma per mg, 146RS3may have trace amounts of thyrotropin or gon-adotropin and ct-intermedin (Lee and Lerner) hassome vasopressin activity. However, Chavin (11)has demonstrated that intact and hypophysecto-mized goldfish do not become melanized afteradministration of TSH, gonadotropins, oxytoeinor vasopressin.

'See footnotes 1 and 2, Table I.2 See footnote 2, Table II.

TABLE VI

Effect of luleotro pin (L26SJAP) upon melano genesisin tissue culture

LuteotenpinUnit/mi

Inter-medin'

FSU/mlACTH' lU/mi Melann-

genesis2

30-353—3.5

202

2 X 10-22X101 —


Ovine prolactin at a concentration of 30 to 35units per ml of medium only occasionally inducesthe appearance of melanin-containing melano-cytes in vitro (Table VI). At this concentrationthe prolactin preparation contains 0.02 ILl ACTHand 20 FSU intermedin/ml. At a concentrationof 3 to 3.5 prolactin units/ml the results wereconsistently negative. Therefore, these data failto confirm Pickford and Kosto's (5) in vivo find-ings dealing with Fundulus even though the con-centrations of the same hormone preparationfused in the present experiment are comparableto those of above authors. Because of the limitedsupply of prolactin, the combined effects of in-termedin and prolactin upon melanogenesiscould not be investigated at this time. TheACTH present in the prolactin experiments ap-pears to be less effective melanogenically thanexpected, for no pigment formation is observedat concentrations of 2 x 10—a ILl ACTH per ml,and only occasional pigment formation at 2 x102 ILl ACTH per ml. These concentrations maybe compared with those of the less purifiedACTH, 146R83, with which 2 X 10_2 ILl ACTHper ml are found to be consistently effective inproducing melanogenesis. A similar loss of mel-

t Drs. Gesehwind and Li who have supplied theprolactin used in tbe present study have indicatedthat this is the same preparation utilized by Pick-ford and Kosto.

anogcnic activity may occur during the isolationof lactogenic hormone as in the purification ofACTH. In any case, under the described experi-mental conditions, prolactin, despite the ACTHcontamination, is not effective in stimulation ofmelanogenesis.

In order to determine whether enzymatic di-gestion, acid or alkali treatment will affect themelanogenic potency of ACTH, lot 146RS3 wassubjected to the treatment as indicated in TableVII. The results indicate that the melanogenicactivity of ACTH is not altered by alkali treat-ment but is lost following trypsin digestion andacid hydrolysis.

Interestingly, alkali treatment is known topotentiate the intermedin effect. Bagnara (12)has shown this to be the case in the preparation(146RS3) used in the present study. However, nodefinite increase in the melanogenic action of thealkali treated hormone occurred when tested intissue culture.

While Bagnara demonstrated little or no loss ofintcrmedin activity following 1 % pepsin diges-tion for 24 hours, the present results indicate analmost total loss of melanogenic activity of thehormone following pepsin digestion as describedby Li (13) or Bagnara (12). After partial acidhydrolysis (0.02% ACTH in 1 N hydrochloricacid at 1000 C for 1 hour), there is a total lossof melanogenic activity, in contrast to the almost

TABLE VIIEffects of various hydrolytic procedures upon melanogenic activity of ACTH (146R,S3) on goldfish skin in

tissue culture, and comparison with re ported findings dealing with other ACTHfactors and intermedin

Treatment Ketngenic Activity Intermedin Activity Maiogenic

0.1 N NaOH, 100 C, 20mm.

Total loss (Engeland Engel, 1957)

Total loss (Eogeland Engel, 1957)

No loss (potentia-tion effect)

No loss

6N HC1, 37.4 C, 24 hrs Total loss (Li, 1948) Total loss

iN HCl, 100 C, 1 hr Little or no loss(Bagoara, 1958)

Total loss

0.05% pepsin, 0.05MHC1, 37.4 C, 4—5 hrs

Little or no loss (Li,1948)

Definiteloss

1% pepsin, 0.05% HCl,37 C, 24 hrs

Little or no loss(Bagnara, 1918)

Total loss

0.1% trypsin, pH 7.0,37 C, 5 hrs

Almost total loss(Bagnara, 1958)

Total loss


complete retention of intermedin potency re-ported by Bagnara (12) after similar treatmentof the same hormone preparation. Thus, it isapparent from these differential reactions thatthe ACTH factor cansing melanogenesis is sep-arate and distinctly different from the melanin-granule-dispersing intermedins.

DIscUsSION

The darkening effect of intermedin upon frogskin is the result of the dispersion of the melaningranules in the melanophores. This observationcan only be made upon cells containing melanin,since intermedin causes the dispersion of the pig-ment granules peripherally in the cytoplasm ofthe melanophores ("expansion" of the cells). Thechanges in light reflectance reading, the basisof the Frog Skin Unit (FSU), indicate only thechanges in the degree of dispersion of the melaningranules in the melanophores. This change israpid and does not indicate the synthesis of newmelanin. The present study deals with the demon-stration of newly formed melanin granules in cellsof ACTH treated tissue cultures of goldfishmelanocytes. Although it is not possible to histo-chemically demonstrate the existence of melaninprecursors in goldfish propigment cells, ACTHactivates pigment synthesis in these cells and thusappears to be responsible for the conversion of acolorless propigment cell into the melanin con-taining cell, the melanocyte. The mechanism bywhich ACTH produces pigment in goldfishmelanocytes in unknown, although the possi-bility exists that ACTH activates an enzymesystem or even releases an inhibitory mechanismpreventing pigment formation. The tissue culturemethod is most suitable for studying the directeffect of ACTH on propigment cells. Utilizingthis technic, the only difference between the ex-perimental and control cultures is the presence orabsence of ACTH. Neural, hormonal, chemicaland physical factors remain constant. This is incontrast to the situation which exists in an intactanimal in which a given stimulus, i.e. ACTH,elicits a series of physiological responses in theorganism. In the latter situation, it is impossibleto distinguish the direct from the indirect effects.It is to be expected, therefore, that agreement ofin vitro and in vivo results may not always occur.

ACTH directly stimulates the formation of pig-ment in the dermal melanoeytes of goldfish invitro when the propigment cells are maintained in

a uniform endocrine environment. Despite thefact that pigment formation occurs in the inter-medin treated cultures, the evidence indicatesthat trace contamination of the intermedinpreparations with ACTH is actually responsiblefor the observed pigmentation. With the use ofconventional assay methods, ACTH contamina-tion is apparently not detected in highly purifiedintermedin preparations, such as lot LR-I. How-ever, melanin-containing melanocytes do appearin cultures treated with this purified preparation.It is evident, therefore, that the presently avail-able ACTH assay methods are not always suffi-ciently sensitive for detection of the minute quan-tities of ACTH required for stimulation ofmelanogenesis in goldfish melanoeytes. It is sug-gested, therefore, that the in vitro melanogenieaction of ACTH may provide an extremely sensi-tive bioassay for this hormone.

ACTH has been found to have a variety ofadrenocortical and extra-adrenal effects. In thisregard, both the weight and ascorbic content ofthe adrenal cortex have long been known to becontrolled by separate adrenocorticotropic "fac-tors," the adrenal weight factor, AWF, and theascorbic acid depleting factor, AAF (14, 15, 16).These factors do not occur in proportionallyequivalent ratios in the ACTH preparationstested by the above authors. A wide variety ofother actions of ACTH upon the cortex andextra-adrenal tissues have been summarized byLi (17) and Engel (18). Interestingly, among theother actions of ACTH, these authors as well asChavin (11), have suggested that ACTH itselfhas a melanin-granule-dispersing activity, on thebasis of biochemical and physiological evidence.In fact, ACTH may not be a uniform hormoneproduct. For example, the recent study of New-man et at (19) suggests that specific varieties ofACTH may be secreted by the hypophysis as afunction of the neural area from which a givenstimulus may originate. If confirmed, this mayoffer a partial explanation of the multifacetedactivity of ACTH. Despite the presently avail-able information concerning the multiple activi-ties of ACTH, the ascorbic acid depletionbioassay procedure is routinely used in the deter-mination of ACTH potency. Thus, in any workdealing with even highly purified ACTH, thenumber of variables is unknown. Perhaps thequestions raised in the present study, as well as inothers, may be resolved to some degree by the


quantitative determination of the known ACTH"factors" in the preparations utilized, for themelanogenie action of ACTH may closely followsome amino acid sequence other than that demon-strated by adrenal ascorbic acid depletion. This isapparent from the tissue culture findings (TableIII) in which a less purified preparation (146RS3)was twenty times more effective in stimulatingmelanogenesis than the more highly purifiedpreparation (907-109-A4) when potency is basedon ascorbic acid depleting activity and intermedinconcentrations are identical. However, on aweight basis, the latter is 1.4 times more effectivein stimulating melanin synthesis than the formerpreparation, thus indicating that not only is thepurified material, indeed, more potent, but alsothat the melanogenic activity is not related tothe ascorbic acid depleting activity of ACTH.

Present evidence indicates that highly purifiedACTH may not be the material of choice forbasic endocrine studies dealing with melanogene-sis. Engel and Engel (20) reported that the treat-ment of purified ACTH with 0.1 N sodium hy-droxide at 1000 C for 20 minutes resulted in theloss of ascorbic acid depleting potency and keto-genie activity. However, such treatment doesnot abolish intermedin activity. The results ofthe present study also indicate the preservationof the melanogenie activity of goldfish melano-cytes following such alkali treatment. Thislends support to the concept that the segment ofthe ACTH molecule possessing melanogenie ac-tivity may not be identical to that portion of themolecular structure responsible for adrenalascorbic acid depletion and ketogenesis (TableVII). In addition, Hungerford et at (21) andJohnsson and Hogberg (22) have found thatalkali-treated ACTH loses its ascorbic acid deplet-ing action but its eosinopenic activity is stilldemonstrable. It appears quite probable that thevarious amino acid sequences of the polypeptideknown as ACTH may each have some specificbiological action which may be dependent uponbut varying in their degree of dependence uponthe integrity of the entire molecule as presentlyknown. It is conceivable, therefore, that themolecular structure of ACTH is actually morecomplex than presently realized, for the pro-cedures involving low pH and heat wbieh are rou-tinely utilized in the extraction and purificationof pituitary material may partially hydrolyzethe "natural" ACTH molecule. Similar hut moreextreme treatment of the "purified" hormone has

been shown by many workers (17) to alter theamino acid sequences. Further consideration ofthis problem in regard to ACTH structure, as wellas that of other pituitary hormones, may clarifythe manifold problems in this area.

The close structural and chemical similarity ofintermedin and ACTH makes their completeseparation difficult. In addition, as indicatedabove, the purification of the hormonal endproduct may lead to partial or complete loss of theprinciple responsible for melanogenesis. From thedata obtained to date, ACTH appears to be theprinciple agent in melanogenesis in goldfish, inter-medin playing a secondary role, if any, in thisprocess. It should be noted, however, an aug-mentation of ACTH action in melanogenesis byintermedin has not been demonstrated. In con-sidering hormonal interaction in melanogenesis,the findings of Pickford and Kosto (5) are ofinterest. Their results indicate that prolactinpotentiates the response of killifisb (Fundutusheteroctitus) melanoeytes to intermedin, permit-ting the formation of new pigment cells in re-sponse to a dose level of intermedin which is in-effective when administered alone. It is possiblethat the prolactin as well as the intermedin usedin their experiments contained small amounts ofACTH melanogenic factor. The parallel bioassayssuggested above might be instrumental in ex-plaining these apparent discrepancies and inbreaking the complex termed ACTH into itscomponent physiological parts.

Evocation of melanogenesis in Fundulus asdescribed by Pickford and Kosto (5) requires avery high dosage of intermedin. For example, 100FSIJ per gram weight of animal barely producedmelanogenesis, while 1,000 FSIJ per gram of ani-mal produced maximal effect. The latter dosagemay well be above physiological limits and if so,the results may fall into the category of pharmaco-logical or paradoxical effects. Earlier, Chavin (3)showed that 1 x l0 FSU of /3-intermedin in-jected into xanthie goldfish did not evokemelanogenesis in this species. The appearance ofmelanoeytes and melanin granules, therefore,may be a more specific response in the goldfishthan in Fundulus. Recently, Chavin (11) demon-strated that the intermedin contained in thepituitaries of xanthie goldfish was approximately20 units (23) per gram of fish. When fish weremaintained over a black background for 115 days,a 66% increase in intermedin occurred. Despitethe gradual increase in endogenous intermedin,


melanogenesis did not occur. It is also of interestin this respect that Chavin (24) has found thatthe intermedin content of the pituitaries ofsaline stressed fish is in the same range as that ofthe control fish; the former are melanized hut thelatter are not. These findings indicate that en-dogenous intermedin is not the hormone whichstimulates melanogenesis.

The melanogenic action of ACTH is direct andwithout adrenocortical mediation. Thus, thehormonal balance between the pituitary secre-tion of ACTH and the adrenal cortex may de-termine the degree of melanoderma. In clinicalhypopituitarism where little ACTH is syn-thesized, a decrease in pigmentation occurs; inhypoadrenalism, increased pigmentation resultsfrom the overproduction of ACTH in response tothe low blood adrenal eortieoid levels. Clinicalhyperpigmentation following the injection ofACTH originally was believed to be a result ofthe intermedin contamination in the ACTHpreparation. No positive evidence is available toindicate that the increase in pigmentation follow-ing intermedin administration is not due to thecontamination of intermedin by a molecularvariety or moiety of ACTH.

The melanin-granule-dispersing effect of in-termedin shown so consistently by variousworkers in frog or fish pigment cells has beenestablished beyond doubt. But it is important tonote that such effects were observed mainly inthe pigment-effector cells, the melanophores.These cells are specialized, of a type found onlyin cold-blooded vertebrates but not in higher ani-mals. In tissue culture, these cells were not seenin mitosis; they were usually degenerate, andwere not replaced on prolonged incubation invitro. Utilizing perfusion technic, Hu and Mon(25) failed to observe pigment dispersion effect ofintermedin in frog melanocytes grown in tissueculture. These pigment-containing cells, on thebasis of their morphological characteristics, werebelieved to be melanocytes instead of melano-phores. It is conceivable that these two types ofcells respond differently to hormonal stimulation.The effect of intermedin on frog melanophoresappears to be rapid, the reaction taking placewithin a matter of minutes. The response of fishpropigment cells and melanocytes to ACTH intissue culture was found to be much slower. Eventhough melanin-containing polydendritie cellswere present in tissue culture at the end of 48hours incubation, invariably many more were

found at the end of 72 hours. It seems apparentthat the response of melanoeytes to the effect ofthe melanogenie factor is relatively slow andrequires several days to reach the maximalresponse. Due to technical difficulties it has notbeen possible to observe the prolonged action ofACTH on cultured fish melanoeytes maintainedin a perfusion chamber. Further study is inprogress. The evidence obtained to date indicatesrather convincingly that the effects of hormonesupon melanophores and melanoeytes are different.Intermedin is effective in stimulating dispersion ofpigment granules in melanophores, while ACTHis effective in inducing melanogenesis in melano-eytes.

SUMMARY

Because of difficulties of complete separationof the ACTH and intermedin activities in eventhe most highly purified hormone preparations,and the relatively scanty experimental evidencedealing with melanogenesis in a completely con-trolled endocrine environment, the actions ofvarious hormones reported to affect melaninsynthesis in the skin were studied. The effects ofACTH, a-intermedin, fl-intermedin and pro-laetin (luteotropin) upon the appearance ofmelanoeytes and stimulation of melanin synthesisin completely xanthie goldfish skin grown intissue culture were compared. The intermedinsand prolaetin were found to have some stimu-latory effect upon melanogenesis in vitro, butonly resulting from the inherent ACTH con-tamination. Only ACTH was found to stimulatemelanin formation. However, the ACTH prepara-tions used varied in melanogenie potency. En-zymatic digestion as well as acid or alkalinehydrolysis studies of ACTH demonstrated thatthe melanogenie activity of ACTH was not corre-lated with the ascorbic acid depleting factor, theketogenie factor or the intermedin activity. Thus,some portion of the ACTH molecule other thanthose responsible for the above activities stimu-lates melanogenesis. The current practice ofindicating ACTH potency in terms of adrenalascorbic acid depleting activity does not indicatethe melanogenie potency of the preparations.

ACKNOWLEDGMENTS

The authors are greatly indebted tn Drs. I.Bunding, E. A. Fullgrabe and W. F. White ofthe Armour Laboratories, Drs. T. H. Lee andA. B. Lerner, Section of Dermatology, Yale


University School of Medicine and Drs. I. I.Gcschwind and C. H. Li, Hormone ResearchLaboratory, University of California for supply-ing the various hormone preparations used in thisstudy.

REFERENCES1. ZONDEK, B. AND KROHN, H.: Hormon des

Zwischenlappens der Hypophyse (Inter-medin). III. Zur Chemie, Darstellung undBiologie des Intermedins. Kim. Wchnschr.,11: 1293—1298, 1932.

2. LEaNER, A. B., SIIIZIJME, K. AND BUNDING, I.:The mechanism of endocrine control ofmelanin pigmentation. J. Clmn. Endocrinol.,14: 1463—1490, 1954.

3. CHAvIN, W.: Pituitary-adrenal control ofmelanization in xanthic goldfish (iera.ssiuseuretus L. J. Exper. Zool., 133: 1—46, 1956.

4. HARRis, J. I. AND LEaNER, A. B.: Amino acidsequence of the a-melanocyte-stimulatinghormone. Nature (Lond.), 179: 1346—1347,1957.

5. PICKFORD, G. E. AND KosTo, B.: Hormonalinduction of melanogenesis in hypophysec-tomized killifish (Fundulus hcteroclitus). En-docrinology, 61: 177—196, 1957.

6. Hu, F., LIvING00D, C. S. AND HILDEBRAND,J. F.: The roller tube tissue culture technicin the evaluation of the primary irritancyproducing capacity of topical medicamentsand chemicals. J. Invest. Dermat., 26: 23—39,1956.

7. LAIDLAW, G. F. AND BLAcKBEEG, S. N.:Melanoma studies: II. A simple technic forthe dopa reaction. Am. J. Path., 8: 491—498,1932.

8. JAcoBsoN, W. AND WEBB, M.: The two typesof nucleoproteins during mitosis. Exper.Cell Res., 3: 163—153, 1952.

9. Hu, F., STABIcCO, R. J., PINKU5, H. ANDFO5NAUGH, H. P.: Human melanocytes intissue culture. J. Invest. Dermat., 28: 15—32,1957.

10. SHIZUiuE, K., LERNER, A. B. AND FITZPATRICK,T. B.: Ju vitro bioassay for the melanocytestimulating hormone. Endocrinology, 54:553—560, 1954.

11. CHAvIN, W.: Pituitary hormones in melano-genesis. Pigment Cell Biology, Chapter IV,pp. 63—83. Academic Press, New York, 1959.

12. BAGNARA, J. T.: Hypophyseal control of

guanophores in Anuran larvae. J. Exper.Zool., 137: 265—283, 1958.

13. Li, C. H.: Biochemistry of adrenocortico-tropic hormone—a review. Conf. Metabol.Aspects of Convalescence, Josiah Macy Jr.Foundation, Trans. 17th Meeting, pp. 114—137, 1948.

14. DixoN, H. B. F., MOORE, S., STAcK-DIJNNE,M. P. AND YOUNG, F. G.: Chromatographyof adrenotropic hormone on ion-exchangecolumns. Nature, 168: 1044—1045, 1951.

15. REINHARDT, W. 0., CE5cHWIND, I. I. AND Li,C. H.: On the evidence suggesting a multi-plicity of adrenocorticotropic hormones: anevaluation of bioassay methods. Acta Endo-ennui., 8:393—424, 1951.

16. MOYER, A. W., VAN DEE SHEER, J., HITTER, R.,TE5AR, W. C., LOGAN, J. B., OLE50N, J. J.AND Cox, H. H.: Comparative assays onadrenocorticotropic hormone preparations.Proc. Soc. Exper. Biol. & Med., 79: 1—3, 1952.

17. Li, C. H.: Hormones of the anterior pituitarygland. Part I. Growth and adrenocortico-tropic hormones. Adv. Protein Chem., 11:101—190, 1956.

18. ENGEL, F. L.: Some unexplained metabolicactions of pituitary hormones with a unify-ing hypothesis concerning their significance.Yale J. Biol. & Med., 30: 201—223, 1957.

19. NEw1uAN, A. E., REDGATE, E. S. AND FARRELL,G.: The effects of diencephalic-mesen-cephalic lesions on aldosterone and cortisonesecretion. Endocrinology, 63: 723—736, 1958.

20. ENGEL, F. L. AND ENGFL, M. C.: The ketogenicactivity of corticotropin, a presumed extra-adrenal action. Endocrinology, 62: 150—158,1958.

21. HUNGEEFOED, C. F., REINHARDT, 0. W. ANDLi, C. H.: Lack of correlation of eosinopenicand adrenal ascorbic acid-depleting activi-ties of various adreuocorticotropic prepara-tions. Proc. Soc. Exper. Biol. & Med., 81:320—323, 1952.

22. JoHNssoN, S. AND H6GBERG, B.: The occurrenceof melanophore hormone in serum fromhuman subjects in different pathologicalconditions and its relation to corticotropin.Acta Endocrinol., 13: 325—342, 1953.

23. WRIGHT, P. A.: A convenient and rapid tech-nique for assay of intermedin. Paper Mich.Acad. Sci., Arts, and Letters, 39: 271—279,1954.

24. CHAVIN, W.: unpublished.25. Hu, F. AND Mom, W.: unpublished.

hormonal stimulation of melanogenesis in tissue culture · phology of these cells is variable. some...

Documents