Control of melanoblast differentiation in amphibia by alpha-melanocyte stimulating hormone, a serum melanization factor, and a melanization inhibiting factor

A ventrally localized melanization inhibiting factor (MIF) has been suggested to play an important role in the establishment of the dorsal-ventral pigment pattern in Xenopus laevis [Fukuzawa and Ide:Dev. Biol., 129:25-36, 1988]. To examine the possibility that melanoblast expression might be controlled by local putative MIF and melanogenic factors, the effects of alpha-melanocyte stimulating hormone (alpha-MSH), a serum melanization factor (SMF) from X. laevis or Rana pipiens, and MIF on the "outgrowth" and "melanization" of Xenopus neural crest cells were studied. Outgrowth represents the number of neural crest cells emigrating from cultured neural tubes, and melanization concerns the percentage of differentiated melanophores among the emigrated cells. MSH or SMF stimulate both outgrowth and melanization. The melanogenic effect of Xenopus serum in this system is more than twice that of Rana serum. The actions of MSH and Xenopus serum on melanization seem to be different: 1) Stronger melanization is induced by Xenopus serum than by MSH, and the onset of melanization occurs earlier with Xenopus serum; 2) MSH stimulates melanization only in the presence of added tyrosine; and 3) MSH causes young melanophores to assume a prominent state of melanophore dispersion during culture, while Xenopus serum (10%) had only a slight dispersing effect and not until day 3. A fraction of Xenopus serum presumably containing molecules of a smaller molecular weight (MW less than 30 kDa) than that of a pigment promoting factor reported in calf serum [Jerdan et al.: J. Cell Biol., 100:1493-1498, 1985] produces the same remarkable melanogenic effects as does intact serum. While this fraction stimulates outgrowth, another fraction presumably containing larger molecules (MW greater than 100 kDa) does not. MIF contained in Xenopus ventral skin conditioned medium (VCM) inhibits both outgrowth and melanization dose dependently. When VCM is used in combination with MSH, the stimulating effects of MSH on both outgrowth and melanization are completely inhibited. In contrast, the stimulatory effects of Xenopus serum are not completely inhibited when combined with VCM, although melanization is reduced to approximately 40% that of controls. MIF activity was also found to be present in ventral, but not in dorsal, skin conditioned media of R. pipiens when tested in the Xenopus neural crest system.(ABSTRACT TRUNCATED AT 400 WORDS)     

Intrinsic Pigment-Cell Stimulating Activity in the Catfish Integument

In keeping with the concept that local factors in the vertebrate integument affect the expression of pigment cells, the present study was directed toward demonstrating the existence of such factors in the skin of the channel catfish, Ictalurus punctatus. This species has a dark dorsal surface in marked contrast to an almost white midventral surface. Pieces of skin from these two surfaces were used to condition culture media, which were in turn bioassayed using the Xenopus neural tube explant system (Fukuzawa and Ide, 1988, Dev. Biol. 129:25). A certain number of neural crest cells grow out from the explant, and many of these are melanized in a culture medium of Steinberg's basic salt solution (BSS). When the BSS was conditioned with either dorsal or ventral skin, a profound increase in both the number of crest cells emigrated from the neural tubes and the percentage of melanized cells was observed. The effects of dorsal skin were stronger than those of ventral skin and were evident on a dose/response basis. Initial fractionation of conditioned BSS with DEAE ion exchange chromatography produced fractions of particular potency in the stimulation of melanogenesis. A similarly conditioned medium based upon Leibovitz's L-15 was used in the primary culture of mature chromatophores, namely, melanophores, iridophores, and xanthophores from tadpoles of Rana pipiens. Both dorsal and ventral conditioned media stimulated iridophores and xanthophores, but seemed to have little or no effect on tadpole melanophores. A melanization inhibiting factor (MIF) from the ventral surface of adult frogs has been suggested as the basis for the light colored ventrum of amphibians, and although the present experiments were not designed to study catfish MIF, the possible existence of such a factor in this species was supported by the results. The total results of this investigation are discussed in the light of the possible presence of a melanization inhibiting factor (MIF) of greater prevalence in the ventrum and a melanization stimulatory factor (MSF) of greater prevalence in the dorsal integument. It is suggested that the light-colored ventral surface of the catfish and other poikilotherms may result from the presence of higher levels of MIF than MSF. Thus, the expression of melanophores is inhibited while that of iridophores is enhanced. In contrast, higher levels of MSF over MIF in the dark dorsal surface would result in melanophore stimulation and inhibition of iridophore expression.     

The Amphibian Melanization Inhibiting Factor (MIF) Blocks the α-MSH Effect on Mouse Malignant Melanocytes

We have found that a melanization inhibitory factor (MIF) extracted from the ventral skin of Rana forreri has a slight inhibitory effect on the activity levels of tyrosinase and dopachrome tautomerase in B16/F10 and Cloudman S-91 murine melanoma cell lines. Furthermore, this factor appears to block the effects of α-MSH on these enzymatic activities. However, MIF treatment does not affect the melanogenic action of theophylline on the same cells, suggesting that MIF acts proximal to MSH-mediated cAMP formation, possibly by interaction with the MSH receptor. In this way, we show that this amphibian factor has biological activity on mammalian melanocytes. This suggests the existence of mammalian counterparts of amphibian MIF in the mouse integument that might regulate epidermal melanocytes. These peptides might be related to the agouti protein, as they share similar mechanisms of action. The interaction of different peptides with the MSH receptor would be a complex but general mechanism responsible for many mammalian coat color variants.     

A ventrally localized inhibitor of melanization in Xenopus laevis skin

Melanophores normally differentiate in dorsal but not in ventral skin of Xenopus laevis. We have sought factors which might regulate this differentiation pattern, and we have obtained a putative melanization inhibiting factor (MIF) from ventral but not from dorsal skin. Preliminary studies reveal that MIF is destroyed by heat or trypsin treatment, indicating its protein composition, and has a molecular weight in the range of 300 kDa. The effects of MIF on the differentiation of neural crest derivatives to melanophores were examined in vitro in the presence of tyrosine and fetal calf serum (FCS). Tyrosine enhances melanophore differentiation in vitro at concentrations equivalent to those estimated in adult Xenopus blood plasma (20 microM). FCS also stimulates melanization, by way of materials other than the tyrosine contained in FCS. MIF strongly inhibits outgrowth and melanization of neural crest cells from neural tube explants. MIF also inhibits the differentiation of melanoblasts contained in cultured explants of ventral skin. Inhibition of melanization or melanophore differentiation by MIF occurs even in the presence of L-tyrosine and/or FCS. We suggest that MIF plays an important role in the establishment of dorso-ventral pigment patterns in amphibia.     

Melanocyte-stimulating hormone affects melanogenic differentiation of quail neural crest cells in vitro

Quail neural crest cells were treated in vitro with alpha-melanocyte-stimulating hormone (alpha-MSH) or dibutyryl cyclic AMP (dbcAMP) plus theophylline. These treatments increased the proportion of melanocytes to total cells in crest cell outgrowth cultures. Pigmentation of neural crest cell clusters proceeded more rapidly when cultures were treated with alpha-MSH or dbcAMP plus theophylline than when untreated. In clonal cell cultures, the proportion of pigmented colonies to total colonies was increased by MSH treatment. From these results, MSH seems not only to accelerate melanogenic differentiation but also to affect the state of commitment of neural crest cells to melanogenic differentiation in vitro, and this action of MSH appears to be mediated by cAMP.     

Stimulation of Cultured Iridophores by Amphibian Ventral Conditioned Medium

That the ventral integument of adult frogs (Rana pipiens) contains factor(s) that stimulate iridophore expression (adhesion, morphologic appearance, proliferation) was demonstrated on iridophores derived from tadpoles of R. pipiens and Pachymedusa dacnicolor, and maintained in primary culture in a growth medium based upon Leibovitz's L-15. Experimental growth medium (VCM) conditioned by a one-hour exposure to pieces of ventral skin of adult R. pipiens induced iridophores to assume a broad and stellate appearance, to form confluent sheets, and to proliferate over a nine-day period. Iridophores in control medium assumed long thin profiles, detached easily, and exhibited no signs of proliferation. Unknown cells containing reflecting platelets and unusual other organelles appeared uniquely in chromatophore cultures of P. dacnicolor in VCM. The intense stimulation of iridophore expression in VCM is consistent with the known inhibitory effect of this medium on melanization and with its purported role in the determination of dorsal/ventral pigment patterns of amphibians. The results are discussed in terms of a prevailing theory about pigment cell origins and development.     

Unusual Leucophore‐Like Cells Specifically Appear in the Lineage of Melanophores in the Periodic Albino Mutant of Xenopus laevis

In the periodic albino mutant (a^p/a^p) of Xenopus laevis, peculiar leucophore‐like cells appear in the skins of tadpoles and froglets, whereas no such cells are observed in the wild‐type (+/+). These leucophore‐like cells are unusual in (1) appearing white, but not iridescent, under incident light, (2) emitting green fluorescence under blue light, (3) exhibiting pigment dispersion in the presence of α‐melanocyte stimulating hormone (αMSH), and (4) containing an abundance of bizarre‐shaped, reflecting platelet‐like organelles. In this study, the developmental and ultrastructural characteristics of these leucophore‐like cells were compared with melanophores, iridophores and xanthophores, utilizing fluorescence stereomicroscopy, and light and electron microscopy. Staining with methylene blue, exposure to αMSH, and culture of neural crest cells were also performed to clarify the pigment cell type. The results obtained clearly indicate that: (1) the leucophore‐like cells in the mutant are different from melanophores, iridophores and xanthophores, (2) the leucophore‐like cells are essentially similar to melanophores of the wild‐type with respect to their localization in the skin and manner of response to αMSH, (3) the leucophore‐like cells contain many premelanosomes that are observed in developing melanophores, and (4) mosaic pigment cells containing both melanosomes specific to mutant melanophores and peculiar reflecting platelet‐like organelles are observed in the mutant tadpoles. These findings strongly suggest that the leucophore‐like cells in the periodic albino mutant are derived from the melanophore lineage, which provides some insight into the origin of brightly colored pigment cells in lower vertebrates.     

Melanophore sublineage-specific requirement for zebrafish touchtone during neural crest development

The specification, differentiation and maintenance of diverse cell types are of central importance to the development of multicellular organisms. The neural crest of vertebrate animals gives rise to many derivatives, including pigment cells, peripheral neurons, glia and elements of the craniofacial skeleton. The development of neural crest-derived pigment cells has been studied extensively to elucidate mechanisms involved in cell fate specification, differentiation, migration and survival. This analysis has been advanced considerably by the availability of large numbers of mouse and, more recently, zebrafish mutants with defects in pigment cell development. We have identified the zebrafish mutant touchtone (tct), which is characterized by the selective absence of most neural crest-derived melanophores. We find that although wild-type numbers of melanophore precursors are generated in the first day of development and migrate normally in tct mutants, most differentiated melanophores subsequently fail to appear. We demonstrate that the failure in melanophore differentiation in tct mutant embryos is due at least in part to the death of melanoblasts and that tct function is required cell autonomously by melanoblasts. The tct locus is located on chromosome 18 in a genomic region apparently devoid of genes known to be involved in melanophore development. Thus, zebrafish tct may represent a novel as well as selective regulator of melanoblast development within the neural crest lineage. Further, our results suggest that, like other neural crest-derived sublineages, melanogenic precursors constitute a heterogeneous population with respect to genetic requirements for development.     

The Delayed Entry of Thoracic Neural Crest Cells into the Dorsolateral Path Is a Consequence of the Late Emigration of Melanogenic Neural Crest Cells from the Neural Tube

Neural crest cells migrate along two pathways in the trunk: the ventral path, between the neural tube and somite, and the dorsolateral path, between the somite and overlying ectoderm. In avian embryos, ventral migration precedes dorsolateral migration by nearly 24 h, and the onset of dorsolateral migration coincides with the cessation of ventral migration. Neural crest cells in the ventral path differentiate predominantly as neurons and glial cells of the peripheral nervous system, whereas those in the dorsolateral path give rise to the melanocytes of the skin. Thus, early- and late-migrating neural crest cells exhibit unique morphogenetic behaviors and give rise to different subsets of neural crest derivatives. Here we present evidence that these differences reflect the appearance of specified melanocyte precursors, or melanoblasts, from late- but not early-migrating neural crest cells. We demonstrate that serum from Smyth line (SL) chickens specifically immunolabels melanocyte precursors, or melanoblasts. Using SL serum as a marker, we first detect melanoblasts immediately dorsal and lateral to the neural tube beginning at stage 18, which is prior to the onset of dorsolateral migration. At later stages every neural crest cell in the dorsolateral path is SL-positive, demonstrating that only melanoblasts migrate dorsolaterally. Thus, melanoblast specification precedes dorsolateral migration, and only melanoblasts migrate dorsolaterally at the thoracic level. Together with previous work (Erickson, C. A., and Goins, T. L.,Development121, 915–924, 1995), these data argue that specification as a melanoblast is a prerequisite for dorsolateral migration. This conclusion suggested that the delay in dorsolateral migration (relative to ventral migration) may reflect a delay in the emigration of melanogenic neural crest cells from the neural tube. Several experiments support this hypothesis. There are no melanoblasts in the ventral path, as revealed by the absence of SL-positive cells in the ventral path, and neural crest cells isolated from the ventral path do not give rise to melanocytes when explanted in culture, suggesting that early, ventrally migrating neural crest cells are limited in their ability to differentiate as melanocytes. Similarly, neural crest cells that emigrate from the neural tubein vitroduring the first 6 h fail to give rise to any melanocytes or SL-positive melanoblasts, whereas neural crest cells that emigrate at progressively later times show a dramatic increase in melanogenesis under identical culture conditions. Thus, the timing of dorsolateral migration at the thoracic level is ultimately controlled by the late emigration of melanogenic neural crest cells from the neural tube.     

Melanization stimulating factors in the integument of the Mugil cephalus and Dicertranchus labrax

The pigment pattern expression resides in the chromatoblasts of the embryonic skin. The differentiation of these chromatoblasts is influenced by specific local factors such a melanization inhibiting factor (MIF) and a melanization-stimulating factor (MSF). We reveal the presence of these factors by means of a series of experiments on the skin of the marine species of fish Dicertranchus labrax and Mugil cephalus, each with different pigment pattern, the former having a light skin and the latter a darker one. Media conditioned by exposure to dorsal and/or ventral skin, stimulates the melanization of Xenopus laevis neural crest cells throughout a 3 day assay period. Similarly conditioned culture media tested on B16-F10 murine malignant melanocytes, revealed a considerable influence in enzymatic activities: dopachrome tautomerase (DCT), tyrosine hydroxylase and dopa oxidase. The use of media in a dose response basis suggests that the conditioned media may contain both melanophore stimulating and inhibiting factors. The results obtained may actually reflect the resultant activity of the two factors present.     

Extracellular Matrix Constituents and Pigment Cell Expression in Primary Cell Culture

Three types of pigment cells were isolated and cultured from larval Rana pipiens, and their attachment, maintenance, and proliferation were examined in the presence of extra-cellular matrix constituents (ECMs) in primary cell culture. The initial profile of pigment cell types present on day 2 of culture reflects the relative attachment of the cells to the dishes. Changes in the numbers of cells present after day 2 reflects the influence of factors present in the culture media on the maintenance, proliferation, or detachment of each type of pigment cell. Fetal bovine serum (FBS) promoted melanophore expression, but inhibited iridophore expression. FBS had no effect on xanthophores. In contrast, ventral skin conditioned medium (VCM), which contains melanization inhibiting factor, strongly stimulated iridophore expression, while it markedly inhibited melanophore expression. VCM had little effect on xanthophores. Of the ECMs tested, collagen type I had no effect on pigment cells. Fibronectin slightly inhibited melanophore expression, while it moderately stimulated iridophores and xanthophores. The stimulatory effect of fibronectin was not as strong as that of FBS or VCM. Laminin was also tested; however, it did not allow pigment cells to attach to the dishes, at least under the culture conditions utilized. The results of these experiments are discussed in terms of the general mechanisms of pigment pattern formation.     

Regulation of Organelle Movement in Melanophores by Protein Kinase A (PKA), Protein Kinase C (PKC), and Protein Phosphatase 2A (PP2A)

We used melanophores, cells specialized for regulated organelle transport, to study signaling pathways involved in the regulation of transport. We transfected immortalized Xenopus melanophores with plasmids encoding epitope-tagged inhibitors of protein phosphatases and protein kinases or control plasmids encoding inactive analogues of these inhibitors. Expression of a recombinant inhibitor of protein kinase A (PKA) results in spontaneous pigment aggregation. α-Melanocyte-stimulating hormone (MSH), a stimulus which increases intracellular cAMP, cannot disperse pigment in these cells. However, melanosomes in these cells can be partially dispersed by PMA, an activator of protein kinase C (PKC). When a recombinant inhibitor of PKC is expressed in melanophores, PMA-induced pigment dispersion is inhibited, but not dispersion induced by MSH. We conclude that PKA and PKC activate two different pathways for melanosome dispersion. When melanophores express the small t antigen of SV-40 virus, a specific inhibitor of protein phosphatase 2A (PP2A), aggregation is completely prevented. Conversely, overexpression of PP2A inhibits pigment dispersion by MSH. Inhibitors of protein phosphatase 1 and protein phosphatase 2B (PP2B) do not affect pigment movement. Therefore, melanosome aggregation is mediated by PP2A.     

The Protein-Phosphatase Inhibitor Okadaic Acid Mimics MSH-Induced and Melatonin-Reversible Melanosome Dispersion in Xenopus laevis Melanophores

The present study describes the ability of 315 nM okadaic acid to induce melanosome dispersion in cultured Xenopus laevis melanophores. This effect of okadaic acid is similar to that of a-melanocyte stimulating hormone (MSH) and can be reversed by melatonin treatment; it indicates that a member of the protein-phosphatase 1 or 2A families must be active for maintenance of the aggregated state. Higher concentrations of okadaic acid (1 μM) attenuate the response of Xenopus melanophores to melatonin leading to the hypothesis that melatonin action is mediated by the calcium/calmodulin activated phosphatase 2B. This hypothesis seems unlikely, however, since the calcium/calmodulin inhibitors TFP and W7 do not prevent melatonin-induced pigment aggregation, but instead induce aggregation on their own.     

Melanization Stimulating Activity in the Skin of the Gilthead Porgy,Sparus auratus

The presence of a melanization-stimulating factor (MSF) was discovered in dorsal and/or ventral skin of Sparus auratus. Skin from this marine species was used to condition Steinberg's balanced salt solution (BSS), which was subsequently tested with the neural tube assay. BBS conditioned by dorsal and/or ventral skin of S. auratus at 25% and 50% concentrations had a profound stimulatory effect on the percentage of melanization of neural crest cells throughout the 3day assay period. In some cases 90% melanization occurred within the first 24 hr. Such stimulated cells showed a doubling of the number of dendrites per cell. assess the effects of MSF on other indices of melanization, dorsal and/or ventral skin was used to condition MEM used in the culture of B16-F10 murine melanoma cells. During the first 24 hr, B16-F10 murine melanoma cells responded to conditioned media by demonstrating a considerable increase in activities of tyrosine hydroxylase, dopa oxidase, and dopachrome tautomerase, but no effect was observed on melanin content. In contrast, melanin content increased after 48 hr of incubation, whereas the enzymatic activities were inhibited during this period. It seems that MSF activity, expressed in several ways, may be present generally among marine species.     

Xanthophores in chromatophore groups of the premigratory neural crest initiate the pigment pattern of the axolotl larva

Summary The barred pigment pattern (Lehman 1957) of the axolotl larva is best observed from stage 41 onwards, where it already consists of alternating transverse bands of melanophores and xanthophores along the dorsal side of the trunk. The present study investigateswhen the two populations of neural crest derived chromatophores, melanophores and xanthophores become determined andhow they interact to create the barred pigment pattern. The presence of phenol oxidase (tyrosinase) in melanophores (revealed by dopa incubation) and pteridines in xanthophores (visualized by fluorescence) were used as markers for cell differentiation in order to recognize melanophores and xanthophores before they became externally visible. It was found that melanophores and xanthophores were already determined in the premigratory neural crest, at stages 30/31 and 35–36, respectively. Between stages 35–36 and 38 they were arranged in a prepattern of several distinct, mixed chromatophore groups along the dorsal trunk, morphologically correlated in the scanning electron microscope with humps on the original crest cell string. While the occurrence of xanthophores was restricted to the chromatophore groups and around them, melanophores were already uniformly distributed in the dorsolateral flank area, having migrated from trunk neural crest portions including the groups. The bar component of the pigment pattern was subsequently initiated by xanthophores, which caused melanophores in and around the chromatophore groups to fade or become invisible. The barred pattern was established by the formation of alternating clusters of like cells, melanophores and xanthophores.     

Sequential actions of Pax3 and Pax7 drive xanthophore development in zebrafish neural crest

The Pax3/7 gene family has a fundamental and conserved role during neural crest formation. In people, PAX3 mutation causes Waardenburg syndrome, and murine Pax3 is essential for pigment formation. However, it is unclear exactly how Pax3 functions within the neural crest. Here we show that pax3 is expressed before other pax3/7 members, including duplicated pax3b, pax7 and pax7b genes, early in zebrafish neural crest development. Knockdown of Pax3 protein by antisense morpholino oligonucleotides results in defective fate specification of xanthophores, with complete ablation in the trunk. Other pigment lineages are specified and differentiate. As a consequence of xanthophore loss, expression of pax7, a marker of the xanthophore lineage, is reduced in neural crest. Morpholino knockdown of Pax7 protein shows that Pax7 itself is dispensable for xanthophore fate specification, although yellow pigmentation is reduced. Loss of xanthophores after reduction of Pax3 correlates with a delay in melanoblast differentiation followed by significant increase in melanophores, suggestive of a Pax3-driven fate switch within a chromatophore precursor or stem cell. Analysis of other neural crest derivatives reveals that, in the absence of Pax3, the enteric nervous system is ablated from its inception. Therefore, Pax3 in zebrafish is required for specification of two specific lineages of neural crest, xanthophores and enteric neurons.     

Intrinsic pigment cell stimulating activity in the skin of the leopard frog, Rana pipiens.

Consistent with the concept that specific pigment patterns of amphibians might result from the highly localized distribution of stimulators and inhibitors of pigment cell expression in the skin, the spot pattern of the leopard frog, Rana pipiens, was examined through the use of the Xenopus neural tube explant assay system (Fukuzawa and Ide, 1988). Media conditioned with pieces of skin from dorsal black spotted areas promoted melanization of neural crest cells at a significantly higher level than did media conditioned with dorsal interspot skin in the absence of extra tyrosine. All conditioned media contained exceedingly low concentrations of tyrosine. With the addition of supplemental tyrosine, the melanization capacity of conditioned media from the interspot areas was elevated to that of the spotted skin. Control media conditioned with ventral frog skin inhibited melanization, as usual, because of the presumed presence of melanization inhibiting factor (MIF). It is considered that dorsal skin contains a melanization stimulating factor (MSF) which is present in significantly higher levels in spotted skin than in interspot areas and that expression of the particular pigmentary pattern of this leopard frog is regulated by the relative distribution of MIF, MSF, and possibly other intrinsic substances present in the skin.     

Effects of fetal bovine serum and serum-free conditions on white and dark axolotl neural crest explants

Summary Neural crest cells from both white mutant and dark (wildtype) axolotls (Ambystoma mexicanum) were cultured in increasing concentrations of fetal bovine serum (FBS; 2 to 20%). For each explant, the total number of cells that migrated and the percent of differentiated melanophores were recorded. At concentrations of FBS above 2% melanophore differentiation was essentially equivalent (32 to 59%) for both the white and dark neural crest cultures, but subtle differences in cell behavior and differentiation were found between the two phenotypes. By contrast there was a significant difference in the percent melanization of cells in serum-free control cultures, wherein melanophore differentiation in dark neural crest cultures was, on average, 18% compared to 5% in white cultures. Thus, contrary to all previously published work, white and dark neural crest cells are not intrinsically equivalent. Our culture results are discussed with regard to the probable in vivo conditions that cause the white phenotype. This research was supported by grant AR 34478 from the National Institutes of Health, Bethesda, MD, and a University of Kansas Biomedical Science support grant.     

Pigment cell pattern formation in Taricha torosa: The role of the extracellular matrix in controlling pigment cell migration and differentiation

The neural crest is a population of highly migratory mesenchymal cells that ultimately localize in specific sites and differentiate into a variety of cell types. This report describes studies on the factors governing the migratory pathways, differentiation, and ultimate localization of the neural crest-derived pigment cells (black melanophores and yellow xanthophores) in the California newt, Taricha torosa. Melanophores first appear scattered in the dorsal portion of the lateral neural crest migratory pathway (between the somites and the ectoderm). These cells are eventually found in two stripes: a dorsal stripe that runs along the apex of the somites, and a midbody stripe near the somite-lateral plate mesoderm border. Melanophores are not seen in the dorsal fin of prehatching embryos. Xanthophores can be identified with the light microscope using NH₄OH-induced autofluorescence of pteridines and in the transmission electron microscope (TEM) by the presence of pterinosomes. Xanthophores first appear scattered among the melanophores over the surface of the somites; these cells eventually are found between the two melanophore stripes and in the dorsal fin. We were interested in determining the roles of the extracellular matrix (ECM) in controlling the formation of pigment cell patterns in T. torosa. Immunocytochemistry, Alcian blue staining of paraffin sections and ruthenium red staining of thin sections (accompanied by Streptomyces hyaluronidase and chondroitinase ABC digestion) were used to identify the composition and distribution of the ECM surrounding the pigment cells at various stages during development. The adhesive glycoprotein fibronectin is found in the dorsal portion of the lateral neural crest migratory pathway as well as in the dorsal fin matrix. Glycosaminoglycans (GAG) are found primarily in the dorsal fin and in the ECM surrounding the notochord. The dorsal fin ECM contains hyaluronate (HA), which was identified in the TEM as Streptomyces hyaluronidase-sensitive 3–5 nm microfibrils, as well as sulfated proteoglycan aggregates. We then confronted T. torosa neural crest cells in vitro with known ECM molecules. When neural folds are explanted onto tissue culture plastic in half-strength L-15 medium containing 10% fetal calf serum (FCS), cells migrate from the explant and differentiate into melanophores after 6 to 9 days. Xanthophores appear in the cultures 2 to 4 days after the appearance of melanophores. When cultured on three-dimensional collagen gels, xanthophores migrate significantly farther (P < 0.01) onto and into the collagen than melanophores (336 ± 183 vs 196 ± 160 μm from the edge of the explant). When 2.5 mg/ml chondroitin sulfate (CS) is present in the collagen gel, the distance that both pigment cell types migrate from the explant is reduced, with the result being that only xanthophores invade the GAG-rich matrix. When 1 mg/ml HA is present in the collagen gel, the differentiation of pigment cells is inhibited. Melanophores appear 48 hr later than in control gels without HA, and the number of melanophores in the explant after 10 days is significantly reduced (P < 0.01; 26.6 vs 1.1 melanophores/explant). When 1 mg/ml of HA is added to the FCS-enriched medium over neural crest cells spreading on tissue culture plastic, there is a similar delay and inhibition of pigment cell differentiation. With 2 mg/ml of CS there is no effect on pigment cell differentiation in vitro. Melanophores eventually appear in the dorsal fin of T. torosa several weeks after hatching. When fragments of dorsal fin that contain no apparent melanophores are transferred onto tissue culture plastic, melanophores appear in the explants after a few days in culture. These results suggest the following model of ECM-cell interactions during pigment cell pattern formation in T. torosa: Pigment cells differentiate in regions of the embryo that contain relatively little GAG. Xanthophores are able to invade the GAG-rich dorsal fin, but melanophores can not. The melanophores that eventually appear in the dorsal fin are derived from the neural crest cells that invaded the fin during early development, and were delayed in differentiating by the presence of HA.