View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

THE MELANOSOME: A DISTINCTIVE SUBCELLTJLAR PARTICLE OF MAMMALIAN AND THE SITE OF MELANOGENESIS* M. SEIJI, M.D., Pxs.D.,t T. B. FITZPATRICK, M.D., Pss.D.t AND M. S. C. BIRBECK, M.A.t

Tyrosinase from different sources has differentthat the melanized cytoplasmic particles pres- characteristics (1); that obtained from plantent in suspensions of mouse melanoma are tissue can usually be prepared in colloidal solu-modified mitochondria (6,7) because they con- tion, while that obtained from mammalian tissuetain, in addition to , not only cytochrome is held by ultramicroscopic, cytoplasmic particles.oxidase, succinic oxidase in markedly varying Localization of tyrosinase activity in suspensionsdegrees and eytochrome C, but also tyrosinase. of particles was first reported by HerrmannThe need for resolution of these differences of and Boss (2) in preparations of the ciliary bodyopinion has been emphasized by Dalton (8) and of the bovine eye. Later, Lerner et al. (3) alsoBirbeck (9) who make a plea for correlation of reported that in their differential-centrifugationthe findings in electronmicrographs with chemical studies of suspensions of homogenized Harding-data obtained from melanin granules and mito- Passey mouse melanoma, tyrosinase activity wasehondria after centrifugation. Such a study has found to be present in the particulate fraction.been carried out with experimental procedures They conjectured that the structural elementsimproved so as to assure the most effective in this fraction were ", or particlespossible separation of melanin granules and the size of microsomes, which were formerlymitochondria. A preliminary report of these part of larger aggregates that were dispersedresults has already appeared (10) and the follow- during the experiments." ing experiments provide conclusive data in There seem to be two contrasting beliefssupport of the concept that melanin granules about the relationship between mitochondria andand mitoehondria are disparate, subeellular melanin granules in melanin-forming cells: oneparticles in the mammalian . is based on morphological findings in electron- micrographs, the other on histochemical and MATERIALS AND METHODS biochemical observations. Studies by electron- 1. Preparation of the Specific Gravity-Gradient microscopy have led to the belief that melanin Tubes granules are formed from cytoplasmic , Tubes of the Spinco swinging-bucket rotor, SW such as the or "pigment-forming39-L, were prepared by layering 0.5 ml. of eight centers" (4, 5), and the is melanizeddifferent concentrations of sucrose solution in gradually. The proponents of this view concludeserial order, with the most concentrated layer at that melanin granules and mitochondria arethe bottom of the tube. They were then allowed to disparate cytoplasmic particles, since no evidencestand overnight (between 8 and 20 hours) so that has been published to show that there existthe gradient might become smooth; at the end of structures intermediate between mitoehondriathis interval, 1 ml. of freshly prepared "large- granule suspension" was layered carefully over the and melanin granules and since the membranetop of each tube. For low-density studies, the structure of pre-melanin granules differs fromsucrose concentrations used were 2.0 M, 1.8 M, that of mitochondria. Per contra, the interpreta-1.6M, 1.55M, 1.5M, 1.4M, 1.2M, 1.OM (Fig. 1 a) tion of biochemical assays and vital staining isfor high-density studies, they were 2.6 M, 2.4 M, *Fromthe tDepartment of Dermatology of2.2 M, 2.0 M, 1.6 M, 1.55M, and 1.5 M (Fig. 2 a) the Harvard Medical School at the Massachusetts General Hospital, Boston 14, Massachusetts and 2. Preparation of "Large-Granule Suspension" the tChester Beatty Research Institute, In- From Mouse Melanoma stitute of Cancer Research, Royal Cancer Hos- B-lb mouse melanoma was serially transplanted pital, London, England. Presented at the Twenty-first Annual Meetingin C-57 strain mice. The entire, actively growing of the Society for Investigative Dermatology,melanoma was excised wben it reached 1—1.5 cm. Inc., Miami Beach, Florida, June 13, 1960. in diameter, and promptly homogenized in 0.3 M This investigation was supported in part by the Damon Runyon Cancer Fund and U. S. Publicsucrose at about 0°C. All subsequent processing Health Service Research Grant CY-5010. took place in a cold environment (about 3° C.). 243 244 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

PREPARATION OF "LARGE-GRANULE SUSPENSION" Fraction Large number 10 mlgranu• susPens'on. B-16 MOUSE MELANOMA in 10 M 05ml. 2 S vole. isotonic sucrose 12 M 05ml. homoginized a'14 M 05 ml. :3 15 N 05ml.Sucrose SI 0 solution. centrifuged ISSM05 ml. 4 710 ag for 10 mm. 16N 05 ml. iBM OS ml. S Sediment SUPERNATANT cells, nuclei, I 2OM OS ml. 6 celldebrisl 11,000 ag for 10 mm. l.a Lb SEDIMENT Supernatant Fia. 1. a. Density-gradient tube before cen- I (microsomeal trifugation, showing the layering of eight different resuspended concentrations of sucrose solution and the large- 15, 000 xg for 10 mm. granule suspension (in 0.3 M sucrose). The gradi- ent was prepared 18 hours in advance; the SEDIMENT Supe rnatant large-granule suspension was added immediately before centrifugation. resuspended in b. Fractions 1 through 6 as they appear in the isotonic sucrose gradient tube after ultracentrifugation at 103,000 X g for 1 hour in a horizontal rotor. The four "LARGE-GRANULE SUSPENSION" opaque fractions and the places where the tube (mitochondria, melanin granules) was cut are shown diagrammatically. FIG. 3. Preparation of large-granule suspen. shIn.

FRACTION NUMBER 3.Separation of the Various Fractions After LARGE l.OmI GRANULE Centrifuqation SUSPENSION 0.Sml H At the end of ccntrifugation, after the position and thickness of the various strata in each tube _0.5 ml 2 .6 0.5 ml had been recorded (Fig. 1, 2), the strata (fractions) 0.5 ml were separated by means of a specially designed LB 1 2.0 0.5 ml centrifuge-tube cutter (10). Figs. 1 b and 2 b show 2.2 0.5ml where the tube was cut to permit collection of the 2.4 0.5 ml individual fractions and how the fractions were numbered. 2.6 0.5 ml 5 a. 4. Enzyme Assays FIG. 2. a. Tube containing modified density- Determinations of succinoxidase and gluta- gradient with large-granule suspension before centrifugation; the gradient consists of eightmatc-oxidasc content were made immediately different concentrations of sucrose solution, andafter separation of the fractions; determinations was prepared 18 hours previous to centrifugation.of tyrosinase and protein-nitrogen content were The large-granule suspension in 0.3 M sucrosealways completed within 48 hours after separation. was added immediately before centrifugation. b. Fractions 1 through 5 as they appear in the Succinoxidase content of the fractions was gradient tube after ultracentrifugation at 103,000estimated respirometrically by measuring oxygen X g for 1 hour. The two opaque regions and theconsumption in the presence of cytochromc C, as places where the tube was cut are shown diagram-described by Umbreit (11); some estimations were matically. also made with ATP (2.8)< 10 M) and DPN (1.5 X 1)1' M) as co-factors. The homogenate was centrifuged at 700 X g for 10 Glutamate oxidase was estimated respiromctri- minutes. The resulting, low-speed supernatant,cally by measuring oxygen uptake in the reaction when centrifuged at 11,000 X g for 10 minutes,system: potassium phosphate buffer, pH 7.5 (1.2 yielded a sediment which was resuspended in 0.3X 10 M); ATP (2.5 X 102 M); DPN (1.5 X 1O M sucrose and reccntrifugcd at 15,000 X g for 10M); nicotinamidc (1.2 )< 10_2 M); MgC12 (8 X minutes. This sediment was again suspended in 0.3 103); cytochrome C (8.3 X 10—i); and glutamic M sucrose to make the "large-granule" prepa-acid (1 X 102 M) in the presence of KOH in the ration which was used as starting material forcenter well. specific gravity-gradient ccntrifugation (Fig. 3). Tyrosinase content was estimated respiromct- THE MELANOSOME AND THE SITE OF MELANOGENESIS 245 rically by measuring oxygen consumption, using 00 - a 10:1 mixture of L-tyrosine and L-dopa as sub- • S Protein—Nitrogen x strate (1.77 micromols) in the M/10 phosphate 90 - X X Tyrosinase buffer (pH 6.8) (12). v——.-v Succinoxidase 80 - 5. Protein-Nitrogen Determination 70 - V I Protein was precipitated by adding 10 per cent I' I trichioroacetic acid to each sample. The precipi- I' I tate was spun down, resuspended in tricbloroace- 60 - , tic acid and centrifuged once more. The nitrogen- I' I content of this sediment was determined after di- Ca50 - \ gestion in a micro-Kjeldahl flask and steam dis- Lj tillation. 40 -

6. Preparotion of Retinal-Pigment Epithelium of 30 - the Chick Embryo 20 - Retinal pigment-epithelium, dissected from the eyes of Rhode-Island-Red chick embryos, was homogenized in 0.3 M sucrose. The homogenate 10 ci A was centrifuged at 300 X g for 5 minutes and suc- I_.— cessive centrifugations were carried out twice more with the respective supernatants. The last 2 3 4 5 supernatant was recentrifuged at 300 X g for 5 PR4CIO/V minutes and centrifuged once more at 11,000 X g Fin.4. Recovery of protein-nitrogen, tyrosin- ase, and succinic oxidase in Fractions 1 through 6 for 10 minutes. The resulting sediment was thenof large-granule suspension of B-16 mouse mela- resuspended in 0.3 M sucrose and recentrifuged atnoma following density-gradient centrifugation. 11,000 1< g for 10 minutes. Finally, the sedimentPercentages are calculated in terms of the sum of was suspended once more in 0.3 M sucrose to formthe amounts recovered in all fractions. Each point on the curve is the average of six observations. the "large-granule preparation" which was usedThe various fractions are numbered as in Figure 1. as the starting material for density-gradientTotal recoveries were: protein-nitrogen, 97.8% centrifugation in our experiments. 1.26; tyrosinas, 293% 29.3; succinic oxidase, 80.5 2.4. 7. Preparation for Electronmicroscopy gray, turbid layer which contrasted with the For electromicroscopy, suitable aliquots of each fraction were mixed at room temperature with 1relatively clear zone below (Fraction 4). Near per cent osmium tetroxide buffered with veronalthe bottom of the tube, there was a suspension acetate to pH 7.4, centrifuged at 11,000 )< g forof black particles (Fraction 5) and a black pellet 10 minutes, and finally cooled to about 5° C. for(Fraction 6). In experiments with retinal fixation. After fixation for 2 hours, the sedimentpigment-epithelium, centrifugation produced no was dehydrated in a graded series of ethyl alcoholsfraction which corresponded to Fraction 5 of and embedded in Araldite resin. The specimensthe melanoma suspensions. Figures 4 and 5 were sectioned with a modified Cambridge rocker- show in detail the results obtained with B-16 microtome (13) and studied by means of a Siemensmouse melanoma and retinal pigment-cpithclium Elmiskop 1 or a Philips E. M. 100 electron micro-of the 10-day-old chick embryo. In these figures, scope. percentage is calculated in terms of the sum of RESULTS the amounts of enzyme recovered in all fractions. Figure 4 shows the estimates of succinoxidase, Separation by Density-Gradient Centrifugation tyrosinase and protein-nitrogen in B-16 mouse When the specific-gravity gradient shown inmelanoma. Each point on the curve is the average Fig. 1 was centrifuged with "large-granule sus-of six observations. The curve for succinoxidasc pension" at 103,000 X g for 1 hour in a horizontalreached its peak in Fraction 3 which contained rotor, two narrow bands typically formed in the70 per cent of the sum of the activity recovered contents of the tube, one (Fraction 2) at thein all fractions. Some succinoxidasc activity was top of the gradient, the other (Fraction 3) aevident in Fractions 4, 5 and 6. Fraction 6 also 246 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY

100 x e—. Protein— Nitrogen however, sufficiently clear to permit adequate identification of the various elements present. 90 - z XTyrosinase v——--v Succinosidase / The data obtained in experiments with the 80 - retinal pigment-epithelium of 10-day-old chick embryos are shown in Fig. 5. There is some 70 - succinoxidase activity in Fractions 1 and 2 and the curve reaches its peak in Fraction 3; there 80 - is no succinoxidase activity in Fractions 4 and 5. V I' Tyrosinase activity is present only in Fraction 5. 50 - / In order to separate mitochondria satisfactorily I' from melanin granules, the specific-gravity I40 - gradient shown in Fig. 2a was prepared with sucrose and the large-granule suspension. The 30 - typical appearance of the tube after centrifuga- tion at 103,000 xfor 1 hour in a horizontal 20 - g rotor is shown in Fig. 2b; a narrow, brownish- 10 - gray, relatively tightly packed band (Fraction 2) could be seen at the top of the gradient.

C' There was a relatively clear zone between this I 2 3 4 5 layer and the brown or black suspension which Pm4cr/oN filled almost the entire bottom half of the tube. FIG.5. Recovery of protein-nitrogen, tyrosin- Figure 12 shows the typical distribution ob- ase, and succinic oxidase in Fractions 1 throughtained with B-16 mouse melanoma. In this 5 of large-granule suspension of the retinal pig- ment-epithelium of the 10-day-old chick embryoexperiment, glutamate oxidase activity as well following density-gradient centrifugation. Per-as suecinoxidase activity was determined in centages are calculated in terms of the sum of the amounts recovered in all fractions. The variousfractions which had been separated by high fractions are numbered as in Figure 2. Totaldensity-gradient centrifugation. Here, the peaks recoveries were: protein-nitrogen, 100%; tyrosin-of both suecinoxidase and glutamate oxidase ase, 87%; succinic oxidase 64%. activity were found in Fraction 2, which cor- responds to Fractions 2 and 3 in the low-density- contained tyrosinase activity, 94 per cent of the gradient experiments. No succinoxidase activity amount contained in Fractions 1 through 6.was found in Fractions 4 and 5. Tyrosinase Electronmicrographs (Figs. 6—il) show thatactivity appeared only in Fractions 4 and 5. the large-granule suspension contains many cellThe detailed results of this experiment (Table 1) elements, such as cell debris, mitochondria,show that Fraction 5 (melanin granules) con- microsomcs, cndoplasmic reticulum, and melanintains three times as much tyrosinase per milligram granules (Fig. 6, 7). In the melanin granules aof protein-nitrogen as the unfractionated large- single outer membrane and an internal membranegranule suspension and that Fraction 2 (mito- composed of irregular, crumpled, spiral andchondria) contains twice as much succinoxidase parallel strands in various stages of melanizationand three times as much glutamate oxidase per could be recognized. Fraction 2 consisted mainly of cell debris, and Fraction 3 was an almost puremilligram of protein-nitrogen as the uncentrifuged preparation of mitochondria (Fig. 8, 9). Fractionlarge-granule suspension. 5 contained both mitochondria and melanin mscUssION granules, and Fraction 6 was almost exclusively The experiments here described show that the melanin granules (Fig. 10, 11). It is not possible so-called mitochondrial fraction of the large- to prepare good electronmicrographs, particularly of Fractions 5 and 6, because sucrose in highgranule suspension of homogenized mouse concentration tends to distort some structuresmelanoma, obtained by the usual method of (e.g., mitoehondria) osmotically and because thecell fractionation, contains, in addition to micro- hard melanin granules make Fractions 5 and 6somes, at least two structurally distinct elements, difficult to cut. The micrograms obtained were,i.e., mitochondria and melanin granules. Density- THE MELANOSOME AND THE SITE OF MELANOGENESIS 247

I

t2P

FIG.6. Electronmicrograph of large-granule suspension of B-16 mouse melanoma showing cell debris, mitochondria, melanin granules and reticular material. Magnification, X 32,000. The line in the lower part of each elcctronmicrograph denotes the length of one micron. FIG. 7. The large-granule suspension of B-b mouse melanoma (see Figure 6) in more detail, showing the various elements and particularly the different stages of mclanization of melanin granules. Mag- nification, X 80,000. 248 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY

Fm. 8. Electronmicrograph of Fraction 3, isolated from large-granule suspension of B-16 mouse melanoma by density-gradient centrifugation. Particles can be seen to be almost exclusively mitochon- dna. Magnification, X 32,000. Fm. 0. Fraction 3 isolated from large-granule suspension of B-16 mouse melanoma by density-gradi- ent centrifugation (see Fig. 7) in more detail. The arrangement of the internnl membranes can be seen. Magnification, X 80,000. THE MELANOSOME AND THE SITE OF MELANOGENESIS 249

r . —ti }.

"S . .. $ 5ffr 1'4 p.' 0'

1; a' 'SF "9

FIG.10.Electronmicrographof Fraction 6, isolated from large-granule suspension of B-16 mouse melanoma by density-gradient centrifugation. This is a reasonably pure suspension of melanin granules. Magnification, X 32,000. FIG. 11. Fraction 6, isolated from the large-granule suspension of B-16 mouse melanoma by density- gradient centrifugation (see Fig. 10) in more detail. The varied size and degree of melanization of the particles can be seen. Magnification, X 80,000. 250 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY gradient centrifugation made possible the separa-enzymes. Tyrosinase activity was found almost tion of these particles into different zonesentirely in the lowest fractions in the tube. In according to their sedimentation rate. Afterelectronmicrographs it can be seen that these centrifugation, the distribution of tyrosinaselower fractions consist entirely of melanin activity in the gradient tubes was entirely differ-granules. In B-16 mouse-melanoma preparations, ent from the distribution of the mitochondrialthe sum of the tyrosinase activity of the various fractions was three times as great as the tyro- sinase activity of the starting material, i.e., the •—• Protein— Nitrogen unfractionated "large-granule suspension." This X Tyrosinase V——— SuccinajIO will be discussed elsewhere. o—a Glutamate Oxidase The bulk of the succinoxidase and glutamate oxidase activity was recovered in Fractions 3 and2 respectively (the mitochondrial fractious). These are typical mitoehondrial enzymes. No succinoxidasewas found in Fractions 3, 4 and 5 obtained by high density-gradient centrifugation of the large-granule suspension of B-16 mouse melanoma. The experiments show that in suspensions of the mouse melanoma and of the retinal pigment- epithelium of chick embryos there are granules which have the typical appearance of mito- chondria under the electron microscope and contain a high concentration of succinoxidase and glutamate oxidase. These granules are found in a density layer in the gradient tube character- I 2 3 4 5 istic of all mitochondria studied, whatever their FR4crIoIv source. In addition, the tumor and pigment- FIG.12. Recovery of protein-nitrogen, ty- rosinase,succinic oxidase, and glutamate oxidaseepithelium contained denser granules; these had in Fractions 1 through 5 of large-granule suspen-the typical appearance of melanin granules a:nd sion of B-16 mouse melanoma following density- gradient ultracontrifugation. Percentages arecontained the greater part of the tyrosinase calculated in terms of the sum of the amountsactivity of the suspension. Some succinoxidase recovered in all fractions. Total recoveries were: protein-nitrogen, 94.3%; tyrosinase, 417%; sue-activity was found in this denser, granule fraction cinicoxidase,82%; and glutamate oxidase, 124%.as the result of contamination with mitochondria,

TABLE 1 Distribution of protein-nitrogen, tyrosirtase, .succinic oxidase and glutamate oxidase after ultracentrifugation (103,000>< gfor 1 hour) of large-granule suspension of B-16 mouse melanoma over density-gradient of hypertonic sucrose

Protein Nitrogen Tyrosinase Succinoxidase Glutamate oxidase

Fraction Specific Specific Specific TotalRecoveryactivity Total Recoveryactivity TotalRecovery activity TotalRecovery mg/mI mg per centsl 0,/hr/l 0,/hrper centl O,/hr/l 0,/hrper centsi 0,/hr/l 0,/hrper cent mgN ugN mgN

Large 1.58 4.75 100 35.1 167 100 769 3654 100 202 963 100 granules 1 0.02 0.05 1.0 0 0 0 0 0 0 0 0 0 2 0.62 1.67 35.1 0 0 0 1765 2948 80.6 640 1069 111 3 0.15 0.53 11.1 0 + 0 0 0 0 92.4 49 5.1 4 0.14 0.53 11.1 301 160 95.8 64 34 0.9 26.4 14 1.4 5 0.57 1.7 35.7 141 240 143.7 21 36 0.98 37 63 6.6

* Afterincubation for 7hours,the reaction mixture was slightly purple in color. THE MELANOSOME AND THE SITE OF MELANOGENESIS 251

DEVELOPMENT OF MELANIN CRANI/LES IN THE MELANOCYTE

MITOCHONDRIA

• MELANIN GRANULE O.4—2O) (Melonoprotein polymer)

MELANOSOME 004M (Cell orgonelle with internal structure and tyrosinase activity)

PRE-MELANOSOME O.0I—O.05j* (TyrOsinose?) Fm. 13. Diagram showing development of melanin granules in the melanocyte. Melanin granules are distinct cytoplasmic constituents of the melanin-forming cell and have their own characteristic enzyme (tyrosinase). The term "melanosome" is proposed for the distinctive, enzymically active particle which is the site of melanin formation and is located within the of the melanocyte. but by modified density-gradient centrifugationtion within the cytoplasm of mammalian melano- such contamination could be avoided and an es-cytes, and to contain a specialized metabolic sentially pure melanin-granule fraction obtained.pathway that converts tyrosine to melanin, we It is also possible, under the electron micro-would like to suggest that during its enzymically scope, to recognize granules in process of melani-active stages it be called a melanosome (Fig. 13). zation which substantiate the findings of pre-The term "melanin granule" could be reserved viously reported investigations (Fig. 7). Birbeckfor the mature, fully melanized particle which and Barnicot (9), using human hair bulbs, andhas lost its tyrosinase activity and is no longer Dalton (14) and Weffings (15, 16), using mousenecessarily confined to the cytoplasm of the and human melanoma respectively, have pointedmelanocyte. This would facilitate verbal differ- out that in their electronmicrographs there couldentiation between the functionally active particle, be seen a series of steps in the formation of thethe inert particle, and the . melanin granule. They postulate that as the A new concept of melanogenesis emerges from small vesicles of the Golgi zone increase in size,this correlation of morphological and biochemical dense membranous structures become evidentstudies: namely, that melanogenesis is an irrevers- within the bounding membrane of the vesicles.ible process in which an enzymically active These internal membranes appear as concentric,particle (the melanosome) is gradually trans- spiral or parallel strands in their presumed crossformed into a mass of inert melanoprotein (the sections. They gradually become considerablymelanin granule) which is incapable of further thicker and their density increases until it ismelanin formation. approximately the same as that of the outer membrane of the vesicle in its early stage. Later, SUMMARY the spaces between the inner membranes are 1. Suspensions of "large granules" from B-16 filled in, so that eventually the granule becomesmouse melanoma and from the retinal pigment- uniformly dense and structureless. Although theepithelium of the chick embryo have been pre- dense material seen in electronmicrographs maypared in isotonic sucrose and centrifuged at not be precisely the same as melanin, it seemshigh speed over a specific-gravity gradient. quite safe to consider that the changes seen may 2. The different fractions thus obtained have be various stages in the gradual transformationbeen examined for succinoxidase, glutamate of the initial vesieles into structureless, com-oxidase and tyrosinase activity and have also pletely melanized melanin granules. been studied under the electron microscope. Because the melanin granule has been shown 3. The granules which have the typical appear- by electronmicroscopy to be structurally distinctance of mitochondria under the electron micro- from mitoehondria, to be unique in its localiza-scope contain a high concentration of succinic 252 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY oxidase and glutamate oxidase, both of which R. A. Ellis, Pp. 239—253. Academic Press, Inc., New York, 1958. are typical mitochondrial enzymes. These 5. WISSENFELS, N.: Licht-, Phasenkontrast- und granules are found in the gradient tube in a elektronenmikroskopische Untersuchungen uber die Enstehung der propigment granula density layer where mitochondria are char- in Melanoblastenkulturen. Z. Zellforsch., acteristically found, whatever their biological 45: 60, 1956. source. 6. DuBuy, H. G., WOODS, M. W., BUnK, D. AND LACKEY, M. D.: Enzymatic activities of 4.Thegranules which, under the electron isolated amelanotic and melanotic granules microscope, had the typical appearance of of mouse melanomas and a suggested rela- melanin granules have been found to contain tionship to mitochondria. J. Nat. Cancer Inst., 9:325,1949. the bulk of the tyrosinase activity. 7. Woors, IVL. W.: Discussion in Pigment Cell 5. The authors conclude that melanin granules Biology, edited by M. Gordon, P. 560. Academic Press, Inc., New York, 1959. and mitochondria are distinct cytoplasmic con- 8. DALTON, A. J. AND FELIX, M. D.: Phase stituents of the melanin-forming cell and that contrast and electron micrography of the each of these two constituents is equipped with Cloudman S-91 Mouse melanoma. In Pig- ment Cell Growth, edited by M. Gordon. its own characteristic enzymes. p. 267—276. Academic Press, Inc., New York, 6. The term "melanosome" is proposed for 1953. 9. BIRBECK, M. S. C. AND BARNICOT, N. A.: the distinctive, enzyinically active particle which Electron microscope studies on pigment is the site of melanin formation and is located formation in human hair follicles. In Pig- only within the cytoplasm of the melanocyte. ment , edited by M. Gordon, Formation of melanin in a melanocyte is the Pp. 549—561. Academic Press, Inc., New York, 1959. same process as synthesis of melanin granules. 10. BAKER, R. V., BIRBECK, M. S. C.. BLASCHK.O, H., FITZPATRICK, T. B. AND SELJI, M.: ACKNOWLEDGMENT Melanin granules and mitochondria. Nature, 187: 392, 1960. We want to thank Drs. H. Blaschko and Ruth11. UMBREIT, W. W., BuRRIs, R. H. AND STAtJF- FEE, J. F., ET AL.: Manometric techniques, Baker of Oxford University for their help in the P. 173. Burgess Publishing Co., Minneapolis, early phases of this investigation. 1959. 12. MIYAMOTO, M. AND FITZPATRICK, T. B.: Com- REFERENCES petitive inhibition of mammalian tyrosinase by phenylalanine and its relationship to 1. LERNER, A. B. AND FITZPATRICK, T. B.: Bio- hair pigmentation in phenylketonuria. chemistry of melanin formation. Physiol. Nature, 179: 199, 1957. Rev., 30: 91, 1950. 13. BIRBECK, M. S. C., MERCER, E. H. AND BAR- 2. HZRRMANN, H. AND Boss, M. B.: Dopa oxidase NICOT, N. A.: The structure and formation activity in extracts from ciliary body and of pigment granules in human hair. Exp. in isolated pigment granules. J. Cell. Comp. Cell Res., 10: 505, 1956. Physiol., 26:131,1945. 14. DALTON, A. J.: Organization in benign and 3. LERNER, A. B., FITZPATRICK, T. B., CALKINS, malignant cells. Lab. Invest., 8: 510, 1959. E. AND SUMMERSON, W. H.: Mammalian 15. WELLINGS, S. R. AND SIEGEL, B. V.: Electron tyrosinase: preparation and properties. J. microscopy of human malignant melanoma. Biol. Chem., 178: 185, 1949. J. Nat. Cancer Inst., 24: 437, 1960. 4. BARNICOT, N. A. AND BIRBECK, M. S. C.: The 16. WELLING5, S. R. AND SIEGEL, B. V.: Role of electron microscopy of human melanocytes Golgi apparatus in the formation of melanin and melanin granules. In The Biology of granules in human malignant melanoma. Hair Growth, edited by W. Montagna and J. Ultrastruct. Res., 3: 147, 1959.

DISCUSSION Dn. HERMANN PINKUS (Detroit, Michigan): Dn. MAKOTO Saui (in closing): Dr. Pinkus, we I would like to ask Dr. Seiji to clarify once morebelieve that the combination of electron micro- the relationship of the melanosome concept tographs and biochemical assay on particles sepa- the idea, which was prevalent at the Secondrated by density gradient ultracentrifugation is a Pigment Conference, that melanin granules aremore precise method of identifying mitochondra actually mitochondria because they stain withand melanosomes than the use of Janus green B. Janus green. I also would like to ask him toWe have recently demonstrated pre-melanosomes discuss the more recent findings by Wissenfelswith tyrosinase activity in the fraction who described melanogenic centers in culturedin which there were no mitochondria present melanocytes. in the electron micrographs.