Melanosomes Are Specialized Members of the Lysosomal Lineage of Organelles

REVIEW

Seth J. Orlow The Ronald 0. Perelman Department ofDem1atology and the Department of Cell Biology, New York University School of Medicine, N ew York, N ew York

Melanosomes are specialized subcellular organelles nosomes, and has important implications for the in which melanin is synthesized and deposited. Elec chemistry of melanization and the potential phanna tron microscopic, cytochemical, genetic, and biochem cologic manipulation of that process. In addition, the ical evidence all support the contention that melano lysosome-like nature of melanosomes may provide somes are specialized lysosomes. The relationship of insight into the processing and presentation of mela melanosomes and lysosomes provides a framework in nosomal antigens by melanoma cells. Keywords: albi which to understand the pathogenesis of disorders such nismlretitral pigment epitlreliumltyrosinase. ] Invest DeJ'ttra as the Chediak-Higashi syndrome, allows the testing tol 105:3-7, 1995 of hypotheses for the trafficking of proteins to mela-

elanin pigments are tyrosine-based biopolymers Beginning in the 19 50s, the e lectron m icroscope was used to synthesized and deposited within highly special generate much of our knowledge about the biogenesis of the ized subcellular organelles termed melano melanosome. There are four stages in the maturation of the somes. This review addresses the ontogeny of melanosome: stage I, the " premelanosome," a spherical organelle melanosomes. The evidence in support of a with ill-defin ed matrix filaments; stage II, in which the typical cMommon biogen etic pathway shared by melanosomes and the elliptical shape of the melanosome is .filled with a well-defined lysosomal lineage of organelles is reviewed. Melanosomes appear to filamentous or laminar matrix; stage Ill, with deposition of electron represent highly specialized lysosomes. opaque m elanin on tlus matrix; a nd stage IV, with complete The lysosome-like nature of melanosomes allows a large body of opacification of melanosomal con tents by the m elanin deposited knowledge accumulated from the study of lysosomes to be applied therein [3] . While under natural conditions melanin deposition is to pigment cell research. The biogenesis of lysosomes and the limited to stage III-IV melanosom es; cytochemical staining for the subcellular signals goveming the trafficking of proteins to lyso DOPA oxidase activity of tyrosinase reveals the presence of the somes are in large part well understood, and may hold true for enzym e in the Trans-Golgi network (TGN), i11 a population of melanosomes as well. The relationship of melanosomes to lyso coated vesicles, as well as stage I-II m elanosomes [4-6]. T he most somes provides a framework by which the pathogenesis of multi corrunonly accepted explanation for tlus observation involves a system disorders that affect both organelles, such as the Chediak bipartite pathway for melanosomal biogenesis. First, "premelano Higash.i and H ermansky-Pudlak syndromes, can be a ddressed. T he somes" containing i ncompletely o rganized matri.x buds from tl1e acidic nature oflysosome-like organelles predicts that melanization TGN, and the matrix subsequently condenses into its readily normally takes place under acidic conditions, and that pharmaco recognizable form as a stage II melanosome. Concurrently or logic manipulation of this acidification may allow us to modulate subsequently, c oated vesicles that contain tyrosinase arise from the the quantity and quality of melanin synthesized. Finally, the TGN. The fusion of these vesicles with the stage II melanosomes is lysosome-like nature of melanosomes may of!:er clues to the beli eved to initiate the process ofmelanogenesis [3,5]. Purification processing and presentation of melanosomal antigens by melanoma of subsets of coated vesicles from actively melanizing 1uelanoma cells. cells reveals that they have the expected high concentration of THE BIOGENESIS OF MELANOSOMES tyrosinase and m ehnun intem1ediates w itlun tl1em [7 ,8]. More than 40 years ago [1] it was recognized that tyrosinase, the MELANOSOMAL ACID PHOSPHATASE key melanogenic enzyme, existed in a particulate form. Subse By the late 1 960s cytochemical tecluuques allowed pigment cell quently, the subce llula.r organelle that functions as the site of researchers to investigate the distribution of other enzymes and melanin deposition was identified by Seiji and co-workers [2] and compare them with tyrosinase. Acid phosphatase activity is found in termed the melanosome. several acidic organelles, including not only 1ysosom es, but also portions of the Golgi apparatus and various specialized secretory granules. Miyamoto and Fitzpatrick [91 had shown the Reprint requests to: Seth]. Orlow, Department of Dermatology, H-100, presence of NYU Medical Center, 550 First Avenue, New York, NY 10016. acid phophatase within tl1e chick retinal pigment epithelium. Abbreviations: C H S, C hediak-Higashi syndrome; GERL, Golgi-endo Cytochemical staining for tl1 e activity of acid phosphatase revealed plasmic reticulum-lysosome; LAMP, lysosome-associated membrane pro its presence witllin melanosomes in bo tl1 mammalian and avian tein; TGN, T rans-Golgi network; T IL, tumor-infiltrating lymphocyte; melanocytes [10-12]. Furthermore, the identical distribution of TRP, ryrosinasc-related protein. tyrosinase and acid phosphatase within the GERL (Golgi-endoplas-

3 4 ORLOW THE JOURNAL OF INVESTIGATIVE DERMATOLOGY mic reticulum-lysosome) zone of che cell Jed Novikoff and co tions of an acidic intramelanosomal environment. Devi el a/ [31] workers [6] to conclude that melanosomes and lysosomes might had shown previously that pretreatment of murine tyrosinase at an share a common biogenetic pathway. acidic pH causes tl1e enzyme to lose the lag period that it normally demonstrates with tyrosine as a substrate in the absence of added GENETIC EVIDENCE FOR A MELANOLYSOSOMAL DOPA as cofactor. Subsequently, it became evident that this acidic RELATIONSHIP pH might have additional implications for the chemistry of mela Some of the earliest and most compelling evidence in support of a nization. For example, melanins rich in dillydrmryindole-2-carbox shared biogenetic relationship between melanosomes and lyso ylic acid are soluble at neutral pH but become progressively somes was provided by the study of genetic diseases affecting insoluble as the pH is lowered to 5 [32]. pigmentation in humans and mice. Chediak-Higashi syndrome If melanization indeed takes place in a lysosome-like acidic (CHS) is an autosomal recessive disorder in humans c haracterized organelle, then one might hypothesize that che modulation of by oculocutaneous albinism and immunodeficiency [13). Suscepti i.ntramelanosomal pH affords a means of modulating melanogenesis bility co a lymphoma-like phase, a bleeding diathesis, and neurop pharmacologically. Drugs that block the putative melanosomal athy are all well described associated phenomena. The microscopic proton-transporting ATPase, for example, might alter tl1e quantity haU.mark of CHS is the presence of giant granules in m any cell and and/or quality of melanin deposited therein. tissue types, including neutroph.ils, natural killer cells, melanocytes, fibroblasts, liver cells, etc [14,15). Studies at the light and electron MELANOSOMES ARE SPECIALIZED L YSOSOMES microscopic levels s uggested chat these giant granules arise by the aberrant fusion of these " lysosomally derived" organelles with one Although acid phosphatase and proton-translocating ATPase activ another. By contrast, a n additional organelle, the platelet-dense ities are present within melanosomes, the first protein formally granule, is either deficient or markedly reduced in those affected by proved to be shared by lysosomes and melanosomes was the CHS (16). lysosomal membrane glycoprotein Lamp-1 [33,34]. In addition, A second "melanolysosomal" disorder in humans is the Hennan Lamp-1 was identified in vesicles implicated in the trafficking of the sky-Pudlak syndrome [17]. In addition to the presence of oculocu tyrosinase-related protein family to melanosomes [34,35]. We have taneous albinism, persons affected by this autosomal recessive subsequently found that the related protein Lamp-2 is similarly disorder may suffer from a severe bleeding diathesis, interstitial distributed. pubnonary disease, and a granulomatous colitis. ln addition to the If melanosomes are indeed specialized lysosomes, then they absence of platelet-dense granules on electron microscopic exam might be predicted to contain a wide array oflysosomal hydrolases. ination [18), other typical findings include the accumulatiotJ. of a Smit and co-workers [36) studied extracts of human melanocytes ceroid-lipofuscin-like substance in the ly sosomes of macrophages subjected to density gradient fractionation and found that a portion in the lung and the gut [1 9]. ll1creased urinary dolichol excretion of tyrosinase co-migrated with the lysosomal hydrolase a-glucosi also supports a lysosomal involvement [20]. dase but was clearly separable from peroxisomal markers. By A third autosomal recessive disorder associated with both lyso combining cell fractionation techniques with the use of genetic somal and pigmentary abnormalities is in f.1 ntiJe sialic acid storage controls, we have recently demonstrated that melanosomes are the disease. Affected infants succumb to a progressive storage disorder major repository in m elanocytes of such lysosomal hydrolases as typjfied by severe mental retardation and growth retardation. J3-glucuronidase, J3-galactosidase, J3-hexosaminidase, and cathep Hepatosplenomegaly and coarse facies are typicaL Extreme dilution sins B and L [3 7]. of the pigmentation of the skin as well as the associated sparse hair Typical of lysosomes is their accessibility to molecules internal is well described [21,22]. ized via both endocytosis and phagocytosis [38]. Melanosom es In laboratory mice, numerous independent autosomal recessive seem to behave in a similar fashion. Melanocyte-stimulating hor mutations at an equivalent number of loci disrupt "melanolysoso mone internalized 11ia its receptor has been shown to make its way mal" functions [23,24]. In addition to a dilution of coat coloc, each to melanosomes [29]. Even more convincing, however, is the of these mutations also inhibits the usually brisk secretion of certain observation by two independent groups that phagosomes contain lysosomal hydrolases from renal cells into the urine. In addition, ing latex particles internalized by melanocytes or m elanoma cells i11 and reminiscent of the human disorders discussed above, many of culture can fuse with me1anosomes [39,40]. Finally, electron mi these mutations also result in a bleeding diathesis associated with croscopic examination of cultured retinal pigment epithelial cells platelet dysfunction [25]. Detailed analysis of homozygous beige demonstrated the presence of tyrosinase and melanosomaJ matrix mice demonstrate that they are an excellent model for Chediak within phagosomes [41]. Higashi syndrome, with giant granules present in many tissues By 1993, numerous authors could look back upon this growing examined [26,27]. body of data and conclude that m elanosomes were indeed special By 1979, the genetic data combined with the previously m en ized lysosomes [34,40,41]. tioned data regarding acid phosphatase empowered Novak and Swank to state explicitly that "melanosomes and lysosomes share several structural and biosynthetic properties" [23]. TRAFF!CKJNG TO MELANOSOMES The delineation of the relationship between melanosomes a TI-rE INTRAMELANOSOMAL pH IS ACIDIC nd the endolysosomal lineage of organelles has implications not only for A hallmark of endosomes, lysosomes, and related organelles is the bettering our understanding of how the melanosome evolved, but maintenance of an acidic intraorganellar pH [28). Two appr()acbes also for determining the signals responsible for the trafficking of have confirmed the presence of the predicted low intramela.noso proteins to melanosomes. Elucidation of the signals governing mal pH. Moellmann and co-workers employed immunocytocbem protein trafficking are important not merely to the molecular and ica.l methods to show that the imi·amelanosomal pH is acidic [29]. cell biologist, but to the clinician as well; numerous examples now Ramaiah and coll eagues isolated melanosomes ·from B16 melanoma exist of diseases caused by abnormal protein trafficking [42]. cells and assessed the incraorganellar pH with fluorescent probes. The motifs directing membrane proteins to endolysosomal or They estimated the intramelanosomal pH to be as low as 3-5 [30]. ganelles are well studied. A recent review summarizes the role in To date, the presence within m elanosomes of the multisubunit subcellular trafficking of specific tyrosine and basic amino acid vacuolar ATPase responsible for lysosomal acidification has aot yet based motifs found in uumerous endolysosomal proteins [43]. [n been proved formally, although a proton translocating ATPase each case, a stretch of four to five basic amino acids in the protein's activity with a pharmacologic profile similar to that of the vacuolar cytoplasmic tail is followed by a tyrosine residue just proximal to enzyme was demonstrated in melanosomes [30). the protein's carboJI.]' terminus. Like many of these proteins, the Moellmann and coll eagues [29] discussed some of the irnplica- members of the tyrosinase-related protei11 family are type 1 me m- VOL. 105, NO. I JULY 1995 MELANOSOMES AR.E SPE IAL! ZED LYSOSOMES 5 brane proteins, with l o ng intraluminal amino-terminal dotncytoplasm. Given what we know of the potenti process of m elanin deposition, including the endolysosomes, following the "default" pathway for other cellular effects on cell s of the generation of highly reactive oxidative intermediates, it is perhaps m embran e proteins [44,45]. VUayasaradhi and co-workers are this process employing m olecular techniques to investigate this problem and not surprising t hat th e organell e that evolved to contain is a specialized lysosom e. In support of this hypoth esis is . recent preliminary reports suggest that, as for endolysosom al proteins, the amelanotic cytoplasmic tail is critical for correct subceUular trafficking of the evidence w e o btained by exarnining pigm ented versus m elanocytes in culntre. W e have shoWl.l that the levels of five TRP family.* es are We have recently found that the severe ocul ocutaneous hypop typical " lysosomal" hydrolases present w ithin melanosom from black mice when igmentation caused by the pla ti1111111 mutation at the murine albi 11 o greatly elevated u1 m elanocytes cultured e s urprising locus is due to misrouting of an enzymatically active mutant compar ed with those from albino mice [37]. Even mor tyrosinase with a truncated cytoplasmic tail to the m elanocyte are t ~1 e o bservations that the transfection of the gen es encoding into nonmclanocytic cell s such as fibroblasts p lasma m embran e rather than to m elanosom es:r If melanosomes tyrosm ase and TRP-1 are indeed specialized lysosomes, then it might be predicted that (which can be accompanied b y m elru1in deposition within endo of selected m embers of the TRP family expressed in nonmelanocytic cell s lysosomes) is al so associated with elevated levels would traffic to endolysosomes. Numerous studies d emonstrate th e lysosomal m embran e proteins [57] and lysosom al hydrolases.:j: accuracy of this prediction . When tyrosinase and TRP-1 are Take•_, togeth er , these results suggest that the e levation oflysosomal e presence of expressed in fibroblasts, they locali ze to organell es that stain prote_m l ~ve l s IS a gen eral (protective?) response to th positively for the lysosomal hydro.lase ,B-glucuronidase as well as for ongomg mtracellular m elanin synthesis. Lamp-l (46,47]. IMPLICATIONS FOR MELANOMA BIOLOGY Altho ugh the cytoplasmic tail of the Pmel17 I si/ tJt:r locus protein bears som e homology to those of the TRP fa mily [ 48], the mature Cancer biologists_ have long searched for tumo r-associated an tigens, Pmel 17 I si/IJer protein behaves as an intramelanosomal matrix ell:-p ressed exclusively by transformed cell s and tl1erefore potential protein rather than as an integral m embran e protein [ 49]. T he targ~ts fo: im1~1tm e -medi ate d tumor destruction. One approach to predicted transm embran e domain may be lost following proteolysis the Jd entJficatJon of such molecules has entail ed the isolation of within the mclanosome [50,51]. Preceden t for the trafficking of a tumor-infiltrating lymphocytes (TlLs). R ecent studies have e luci m embrane protein to lysosom es directed b y signals in its cytoplas dated_ a number of antigens speci.fica.lly recognized b y melanoma mic tail followed by its s ubsequent cleavage into a soluble protein assoCiated T !Ls [58 - 60]. Altl10ugh a few of these antigens repre is supported b y the study of acid phosphatase. T his enzym e is sent_ h eretofore undescribed m olecules, others are tantalizingly carried to the plasma m embran e via the "default" pathway and is f<~ • mhar . S e~e ~~.l of the recently described T IL targets are major then i11terna Li zed via endocytosis [ 45]. Upon arrival within the h1 stocompatlb1hty complex-associated peptides derived from the lysosom e, the protein is cleaved to soluble form [45]. processing of the m elan osomal proteins tyTosinase and gp100 How do the lysosomal h ydrolases detected w ithin melanosomes [59,60]. traffic to this organ ell e? The classi c pathway by w hich these Tyrosi11ase is a m elanosomal m embrane protein weLl known to hydrolases traffic to lysosom es in volves the modification of their any st u~ e n t of _pigment cell biology. T he gp100 antigen , or its tern1.inal m annose residues to a mannose 6-phosphate (52]. T hese altem atlvely s phced fonn, pmel17 I silver locus protein referred to mannose 6-phospho rylated enzym es bind to the m embrane-bo und above, are, in fa ct, the a ntigens recognized by three different mannose 6-phosphate receptor in the Golgi apparatu s. T hey are n~onocl?n al antibodies employed widely for tl1e histopathologic carried to late endosomes and, in the presence of the low pH in that dwgn oSJS c:>f melanoma: HMBSO, HMB45, and NJGi beteb ( 61]. o rgane!J e, are released from the receptor into the endosome's R ecent ev1dence suggests that gp100 m ay be processed so that its lumen (53,54]. Subsequently, this organel.l e m atures into a lyso final location i s in the m elanosome's interior, as part of the som e . fi lam_e nto us _melanosom al matrL-x; indeed, it is homologous to a D eficiencies of individua.l hydrolases arc well desc•·ibed in hu m atrix protem found in avian retinal pigmented epithelium [62]. m ans, and are not generally associated with hypopigm entation. In How do m elanosom al proteins becom e proteolyzed peptides contrast, in T-cell disease, a wide range oflysosomal hydrolases are bound to major histocompatibility complex molecules on the not phosphorylated on mannose residues due to the a bsence of a m elru10m a cell surface? One possi ble explanation is based on the specific phosphotransferase. A severe lysosomal storage disorder fact that antigen processing i11 m acrophages takes place within eventuates because the unmodified hydrolases are secreted o ut of the endolyso_so mal compartment. It is there that proteins are the ceLl rather than being t ransported to endolysosom es (55]. p~oteo l yzed tnto peptides that may then gain access to major Nonetheless, cel.l s such as lymphocytes are barely affected; recent h1 stocompatibility complex molecules and bu1d to them prior to ~viden ce suggests that, in som e tissues, a m annose 6-phospbate their trafficking to the cell surface. We h ave shown that proteases mdependen t pathway is operative [56]. Even l-cell disease is such as cathepsins Band L are present within the m elan osom e (37], associ:lted with, somes. 1 flll!cst Denuntol cul tured retinal pigment epithelium of cattle. Tiss11 e Cell Res 276:273-279, 52:212- 216, 1969 1994 11. Wolff K, Schreiner E: Mela nosomal acid phosphatase. Arch Dcnll llfol Forsclt 42. Amara JF, C heng SH. Sm.ith AE: Intracellular protein trafticking defects in 241:255-272. 1971 human disease. Tre11ds Ceii/Jio/ 2:14 5-1 49, 1992 12. Boissy RE, MoeUm ann GE, Halaban RO: Tyrosina.s<: and acid phosphatose 43. Sandov~1l IV. Bakke 0: Targeting: of membrane pro teins to endosomes and ac ti vities in melanocytcs from avian albinos.} /rw est Dcmtnto/88:292-300, 1987 lysosomcs. Trertds Cell Bioi 4:292-297, 1994 13. Barak Y, Nir E: Chediak- I-Iigashi syndrome. A111} Pediatr Helllatoi!O IIcol 9:42-55, 44. Will.ia.ms MA, Fukuda M: Accumulation of membrane glycoprotcins in lyso 1987 somes requires a tyrosine residue a t a particular position in the C)'toplasmic tail. 14. Windhorst DB, Zelickson AS, Good R.A: Chediak-Higaslti syndrome: hereditary j Cell /Jio/111 :955-966, 1990 gigantism of cytoplsmic organelles. Sricuce 151:81-83, .1966 45 . Peters C. Braun M, Weber B, W endland M, Schmidt B, Pohlmann R , Waheed 15. Zcli ckson AS , Windhorst DB, White JG, Good R.A : The Chediak-J-Iigashi A, von Fi gura K: Targeting of a lysosomal membran e pro tein: a tyrosin e syndrome: formation of giant rnelanosomcs a.nd the b asis ofhypopigme nt:-~ tio n. containing e ndocytosis signnl in the cytopla sm.ic; tail of acid phosphatase is J l11 11esl Dermat11l 49:575-581, 1967 necessary and sufficient for targeting to lysosomes. EMBOJ 9:3 497-3506, 1990 16. Costa JL, Fauci AS, Wol![ SM: A pl o. ttciCL abn o rm :-~ Jj ty in the Chcdi ak-1-Jig:a shi 46. Winder AJ: Expre ssion of a mo use tyrosinase eDNA in 3TJ Swiss mo use syndrome of man. lJ/oud 48:517-520. 1976 fibroblasts. Biodtclll Biop/rys Res Co 1111111111 178:739-745, 1991 17 . D e Pinho RA, K:-~p ln n KL: The Hcnnansky-Pudlak syn.drome. Report o f three 4 7. Vijay;1 saradh..i S, Doskoch PM, Houghto n A: Biosynthesis and intracellular cases and review of path ophysiology and management considcrntions. Nlcdicim: movement o f th e mclanosomal membrane g lycoprotein gp 75 , the human b 64:192-202, 1985 (Brown) locus product. Exp Cell R es I 96:233-240, 199'1 18. Witkop CJ, Krumwiede M, Sedano 1-1 , WhiteJG: Re liability of abse nt platelet 48. Kwon llS: Pi gmentation genes: the tyrosinase gene family and the pmcl 17 gene dense bodies as a diagnosti c c riterion for HermanskY-Pudlak syndrome. Am 1 fa mily.} III IICS t Derllla/o/100:134S-140S, '1993 J-Jemtlfol 26:305-311 , 1987 49. Zhou B. Kobayashi T, Donatien I'D, Betm ett DC, Hearing VJ, Orlow SJ : 19. Ta.k;1hashi A, Yokoyama T: Herma'nsky-Pudlak syndro •11 e with special reference Identification of a mclan osomal matrix protein encoded by the murine si to lysosomal dysfunction. A case report and review of the literature. Virclwws (sil ver) locus usin g "organ ell e scanning ." Proc Nntl ;lead Sci USA 91:7076- Arc/till A: Patlwl A 11nt J-listopatlw/ 402:247-258, 1984 7080, 1994 20. Witkop C Jr, Wolfe LS, Cal SX, Wh.ite J G, Townsend b, Keenan KM: Elevated 50. Orlow SJ, Zhou BK, .Boissy R.E. Pi.fko-1-lirst S: Id entification of :1 mammalian urinary dolichol excretion in the Hcrmansky-Pudlak:. syndrome. Indica tor of melanosomal rn:1tri..x glycopro tein. 1 lw1cst Demwtol 101 :141.-144, 19 93 lysosomal dysfunction. A 111 j Met/ 82:463-470, 1987 51. Kobayashi T, Urabe K, Orlow SJ, Higashi K, Imok:t wa G, Kwon BS, PotterfB, 21. Stevenson R.E, Lubinsky M, Taylor HA, Wenger D A, Schroer RJ. O lmstead Hearing VJ: The Pmcl 17/siiiiCJ' locus pro tein: characterization and i.nvest.ig:l PM : Sialic acid storage disease w ith sialuria: c lin..i c:1l \l nd biochemical features tion of its melanogenic function. j Bi11f C /i c, 269:29 198-29205, 1994 in the severe infantile type. Pediatrics 72:441-449, 1 983 52. Griffiths G, Ho Aa ck B , Sirnons K, Mellman I, Kornfeld S: T he mannosc 22. Pueschel SM. O 'Shea PA , Alroy J, Ambler MW, bangond F, Daniel PF, 6-phosphatc receptor and the biogenesis of lysosomes. Cell 52:329-341, 1988 53. Geuzc HJ, Stoorvogel W, Strous GJ , Slot JW, nleekemolcn J E, Mellman I: Sorting of mannose 6-pltosphatc receptors and lysosomal membrane proteins § Zhou BK, Orlow SJ, Donatien PD (unpublished results). in endocytic vesicles.} Cell Bioi 107:2491-250'1., 1988 VOL. 105. NO. I JULY 1995 MELANOSOMES ARE SPECIALIZED L YSOSOMES 7

54. Ludwig: T, Griffiths G. Hoflack B: Distribution of newly synthesized lysosomal 61. Adema GJ, de Boer AJ , Vogel AM, Loenen WAM, Figdor CG: Molecular enzymes in the cndocytic pathway of nonnal rat kidney cells. J Cell Bioi characterization of the melanocyte lineage-specific antigen gp 100.] Bio i C l1 em 115:1561-1 572,1991 269:20126-20133, 1994 55. Gorlin RJ. Cohen MM. levin SL: Symlromcs cif tlu:. Head mtd Neck, Jrd ed. Oxford 62. Mochii M, Agata K, Kobayashi H, Yamarnato TS, Eguchi G: E'1Jression of gene University Press, Oxford, 1990, p 124 coding for a mclanosomal matrix protein tra.nscriptionally regulated in the 56. Glickman JN, Komfeld S: Mannosc 6-phosphatc-independent targeting of transdiffercntiation of chick embryo pigmented epitheli al cells. Cell Differcu lysosomal enzymes in 1-cell di sease B lymphoblasts. J Cell Bioi 123:99-108, 24:64-74, 1 988 1993 63. Borovansky J, Mirejovsky P, R.iley PA: Possible relationsllip between abnom1al 57. Luo 0 , Chen H , Jimbow K: Cotransfection of genes encodin g human tyrosinase mclanosomc scructurc and cytotoxic phenomena in malignant melanoma. and tyrosinase-related protcin-1 prevents melanocyte death and enhances Nt"'l''"'"'" 38:393-400, 1991 melanin pigmcnwtio n :wd gene expression of LAMP-l. Exp Cell Res 2 '1 3:231- 64. Erlandson RA: U1 trastructuraJ dhtgnosis of a.mchnotic malignant melanoma: 24 1, 1994 aberrant mclanosomes, myelin figures or lysosomes? Ultrastn/Cf J>atho/11 :191- 58. van dcr Brugg:cn P, Travcrsari C. Chomcz P. Lurquin C, De Placn E, Van den 208, 1987 Eynde B. Knuth A, Boon T: A gene encodin g an antigen recognized by 65. Mazur MT, Katzenstein AL: Metastatic melanoma: the spectrum of ultrastruc cytolytic T lymphocytes on a human melanoma. Scieucc 254:1643-1 647, 1991 tural morphology. Ultrastn/CI J>atlwl I :337-356, 1980 59 . Kawakami Y, Eliyahu S, Delgado CH, Robbins PF, Sakaguchi K, Appclla E. 66. Sloane BF, Dunn JR. Honn KV: Lysosomal cathepsin B: correlation with Yannclli JR. Adema GJ. Miki T, Rosenberg SA : Identification of a human metastatic potential. Scie 11cc 212:11 51-1153, 1981 melanoma antigen recognized by tumor-infiltrating lymphocytes associated 67. Qian F, Bajkowski AS, Steiner DF, Chan SJ, Frankfater A: Expression of five with in vivo tumor rejection. !'roc Nat/ Acad Sci USA 91:6458-6462, 1994 cathepsins in murine melanomas of varying metastatic potential and nom1al 60 . Kawakami Y, Eli yahu S, Delgado C H, Robbins PF, Rivoltini L, Topalian SL, tissues. Cm1ccr R es 49 :4870-4875, 1989 Miki T, Rosenberg SA: Cloning of the gene coding for a s hared human 68. Podhajcer OL, Bover L, Bravo AI, Ledda MF, Kairiyama C, Calb I, Guerra L, melanoma antigen r ecognized by autologous T cell s infiltrating into tumor. Capony F, Mordoh J: Expression of cathepsin D in primarj' and metastatic J>roc Nat / Acad Sci USA 91:3515-3519, 1994 human m elanoma and dysplastic nevi. J llwcst Dcmwo/104:340-344, 1994