Cells from a lung metastasis, arising from Cloudman S91 melanoma cells implanted s.c. in the tail of a BALB/c nu/nu mouse, were comprised chiefly of host × tumor hybrids. These lung metastasis cells showed: (a) 30–40% increased DNA content;(b) resistance to 10−4mhypoxanthine, 4 × 10−7maminopterin, and 1.6 × 10−5mthymidine (HAT) + G418; and (c) the presence in genomic DNA of genes for both wt and albino tyrosinase,reflecting the DBA/2J (Cloudman S91) and BALB/c mouse genotypes,respectively. Individual clones of lung metastasis cells expressed enhanced pigmentation, motility, and responsiveness to MSH/IBMX,a behavior similar to that recently reported for artificially generated melanoma × macrophage fusion hybrids. These similarities suggested that the host fusion partner generating the lung metastasis hybrids might have been a macrophage, although formal proof for this was not possible. The results provide the first direct evidence that host × tumor hybridization could serve as an initiating mechanism for melanoma metastasis.

Metastatic transformation is considered to be a process of clonal evolution wherein individual cells of the primary tumor progressively acquire new traits, allowing them to invade surrounding tissue and migrate to lymph nodes and distant organs (1). In addition to the loss of growth control characteristic of primary tumor cells, an imbalance in regulation of motility and proteolysis are key components of invasion and metastasis (2). However, the mechanisms by which these complex traits are acquired are largely unknown. In current models, they arise through cumulative mutations and other genetic changes which, in combination with selective pressures of the tumor environment, lead to the emergence of the metastatic phenotype(1, 2, 3, 4, 5, 6, 7, 8, 9).

An alternative model involves hybridization of tumor-infiltrating leukocytes with primary tumor cells (10, 11, 12, 13, 14, 15, 16). In this concept, the increased metastasis of such hybrids would result from coexpression of traits of motility and homing from the leukocyte parent, coupled with deregulated growth from the neoplastic parent(16). A selective advantage for hybridization is that it produces complex new genotypes rapidly, without the necessity of new mutations. Both mechanisms have strong biological precedent:hybridization in the evolution of the eukaryotic cell (17, 18); and mutation as an essential component of cellular evolution as well as in the onset of deregulated growth in cancer(7, 19, 20, 21). In principle, the mutation and hybrid mechanisms for metastatic transformation are not mutually exclusive,and both mechanisms, or combinations thereof, could be operative.

Munzarova et al.(22, 23) first suggested that macrophage hybridization might play a role in melanoma metastasis,pointing out a number of macrophage-like traits expressed by melanoma cells. To investigate this concept, we created a panel of artificial hybrids between normal DBA/2J peritoneal macrophages and cells of a weakly metastatic Cloudman S91 melanoma subline,6neo, that was also of DBA/2J origin (16, 24). More than half of the 35 hybrids tested were found to be more metastatic in mice than in the parental 6neocells. In addition, hybrids with higher metastatic potential tended to be more pigmented, more motile, and more dendritic than parental 6neo cells, with higher expression of MSH receptors and markedly increased responsiveness to MSH/IBMX treatment(25, 26, 27), a potent stimulator of cyclic AMP levels(26). Initial results indicated that such macrophage × melanoma hybrids expressed enhanced N-glycosylation of a number of cellular proteins, which might account for the multiple phenotypic changes (25, 28;F>).

This report details the spontaneous progression of a poorly differentiated, amelanotic Cloudman S91melanoma from its site of s.c. implantation in a BALB/c mouse to a highly melanotic lung metastasis comprised primarily of host × tumor hybrids. In vitro analyses of individual clones from the lung metastasis revealed that they expressed the same traits of enhanced melanogenesis,motility, and MSH/IBMX responsiveness as did the artificial macrophage × melanoma hybrids described above. This raised the possibility that the BALB/c host fusion partner(s) may also have been of macrophage or other hemopoietic cell origin.

Cells and Culture Conditions.

Cloudman S91 melanoma clone PS-1-HGPRT-1/G418res was cultured as described previously (16). Cells from this line, referred to as 6neo, lacked hypoxanthine guanine phosphoribosyl transferase activity and thus were killed by media containing HAT3(29) and were resistant to G418 (Geneticin; Life Technologies, Inc., Gaithersburg, MD), a neomycin analogue, at concentrations of 400 μg/ml culture medium. To test the sensitivity of cells to HAT/G418, cells (2 × 104) were plated into Corning 25-cm2 flasks in 4 ml of DMEM containing 10%fetal bovine serum. The next day, the medium was replaced with 4 ml of fresh medium with or without HAT and G418 (200 μg/ml). Three days later, and every 2 days thereafter, the cells were fed ± HAT/G418 media with the G418 elevated to 400 μg/ml. At the times designated, cells were harvested in Ca2+/Mg2+-free Tyrodes saline containing EDTA (1 mm) and counted in a Coulter counter.

Flow Cytometric Analyses of DNA Content.

For each sample, about 106 cells were washed,resuspended in ice-cold PBS (2 ml), and then fixed by three consecutive additions (2 ml each) of 95% ethanol on ice. A minimum of 1 h after fixation, the cells were collected by centrifugation, resuspended in RNase (1 mg/ml PBS; Sigma, St. Louis, MO) for 30 min at 37°C, and then stained with propidium iodide (0.05 mg/ml PBS) for 1 h on ice. Flow cytometric analysis was performed with a FACS Vantage flow cytometer (Becton Dickinson, San Jose, CA). The cells were excited at 488 nm, and the emission was collected through a 630/22 nm band pass filter. A minimum of 104 cells were analyzed for each sample. DNA content was calculated relative to that of parental Cloudman S91 6neo cells, which was arbitrarily set at 1.00 (30).

Migration Assay.

Migration was assessed as described previously (26). Cells(5 × 104/0.5 ml culture medium without serum) were seeded into Costar Transwell cell culture chamber inserts (12 μm, pore diameter) and placed into wells containing 1.5 ml of DMEM nutrient medium without serum but with 3T3-fibroblast-conditioned medium (3T3-CM; 33% v/v) as a chemoattractant in a gassed, humidified incubator. After 3 h, the inserts were withdrawn, cells on the upper surface were removed with a cotton swab, and the cells on the underside were fixed with methanol and stained with hematoxylin. The membrane filters were cut with a scalpel, mounted on slides, and counted in 10 microscopic fields with a light microscope at ×430 magnification. Migration was expressed as the percentage of total cells migrating to the underside of the filter/3 h.

Histology.

Tissue specimens were fixed in formalin (10% v/v), embedded in paraffin, sectioned, and stained with H&E by routine histopathological procedures.

RFLPs.

Genomic DNA was prepared using the Stratagene kit. PCR amplification of genomic DNA was carried out in a programmable thermal cycler (MJ Research, Cambridge, MA) using the oligonucleotide primers 5′-TCCGAATTCAAAGGGGTGGATGACCG-3′ (bases 293–312 plus a terminal EcoRI site) and 5′-GACACATAGTAATGCATCC-3′ (bases 633–615)to amplify a region of the tyrosinase gene (bases 293–633; Ref.29) Each reaction mixture contained 200 ng of genomic DNA,40 pmol of primer, and 1.5units of Taq polymerase (Perkin-Elmer) in 100μl of 10 mm Tris-HCl, 50 mm KCl, 1.5 mmMgCl2, and 200 μmdeoxynucleotide triphosphates. The reaction was cycled 30 times through 90 s at 92°C, 90 s at 53°C, and 120 s at 72°C,followed by one cycle of 72°C for 7 min. Restriction endonuclease(DdeI) digestion was performed according to the manufacturer’s instructions (New England Biolabs, Beverle, MA). Electrophoresis was performed in 4% NuSieve GTG-agarose (FMC Bioproducts) gels in 40 mm Tris (pH 7.5)/1 mm EDTA containing ethidium bromide (0.5μg/ml).

DNA Sequencing.

DNA fragments were purified with a Qiaquick gel extraction kit (Qiagen,Inc., Chatsworth, CA). DNA was sequenced using the internal reverse primer (503–520 nucleotides; 5′-GCTGATAGTATGTTTTGC-3′) by the W. M. Keck Biotechnology Resource Laboratory at Yale University (New Haven, CT).

Tyrosinase Activity.

Tyrosinase was assayed by the Pomerantz 3H2O release method from l-[3H]tyrosine (New England Nuclear Corp.) in cell lysates (31). Lysates were prepared after incubation of cells in vitro for 72 h with and without MSH (10−7m) and IBMX (10−4m) to stimulate melanogenesis.

Immunoblotting.

Equal quantities of protein were boiled in the presence of 2% SDS and 2% β-mercaptoethanol and subjected to electrophoresis and immunoblotting as described previously (32). A rat monoclonal antibody to murine LAMP-1, 1D4B, was used as described previously (25, 32).

Tumor Progression in Vivo and Culture of Tumor Cells.

As part of our studies on the potential role of hybrid formation in metastasis, we implanted Cloudman S91 cells homozygous for wt tyrosinase (C/C) into BALB/c nu/nu mice homozygous for albino tyrosinase (c/c). Due to the nu/nu mutation, these mice are immunologically permissive for growth of Cloudman S91 cells. Previous studies had shown that BALB/c albino tyrosinase could be distinguished at the genetic level from wt tyrosinase by DNA restriction fragment analyses (see below). Thus, in this system, metastases comprised of tumor × host hybrids could be identified should individual cells contain copies of both wt and albino mutant tyrosinases. Accordingly,10 BALB/c nu/nu mice (Harlan, Indianapolis, IN)received s.c. implants in the mid-tail region of 3 × 105 Cloudman S91 mouse melanoma 6neo cells. After 1–2 months, small amelanotic tumors were visible at the implantation sites, and, by 4 months, the tumors had reached approximately 0.5 cm in diameter and were oblong in shape, with the long axis parallel to the tail. Amelanotic tumors in vivo are typical for 6neo cells in both BALB/c nu/nu and DBA/2J mice, although 6neo cells have wt tyrosinase. At 4 months, in one mouse, a melanotic tail metastasis was observed approximately 2 mm from the primary tumor and in a more proximal location (Fig. 1,A). Two weeks later, the primary tumor and the tail metastasis had grown together, becoming contiguous (data not shown). At this time, the mouse was anesthetized, the tail was sterilized with ethanol and transected near its base, and the stump was treated with antibiotic ointment. Cells from the most proximal portion of the melanotic tumor and cells from the most distal portion of the amelanotic tumor were aseptically removed with sterile 18-gauge needles; transferred to a 6-well plate (Corning Tissue Culture) with 3 ml/well of DMEM containing 10% fetal bovine serum, penicillin, and streptomycin; and placed in a 37°C gassed, humidified incubator. The cells were adapted to monolayer culture and maintained as such, with early-passage aliquots frozen in liquid N2. A piece of tail that contained the primary tumor and the contiguous metastasis was fixed in formalin. Five weeks after the tail transection, the mouse died. Before the onset of rigor mortis, the animal was aseptically necropsied. A massive melanotic tumor filled the thoracic cavity, obscuring normal tissue (Fig. 1 B). Pieces (<1 mm3) were cut from various regions of this tumor, pooled, and placed in monolayer culture in DMEM/10% FBS with antibiotics, with early-passage aliquots stored frozen in liquid N2, as described above. Thus,three populations of cells were established in monolayer culture, one each from the primary tumor, the tail metastasis, and the lung metastasis. Lung metastasis cells were further cloned in liquid agar(33), and a panel of 12 clones was subsequently established in monolayer from the original lung metastasis.

Effects of HAT + G418 on Growth.

Growth of the cells in HAT + G418-containing medium was assessed (Fig. 2, A–C). Neither cells from the primary tumor (Fig. 2,A) nor tail metastasis cells (Fig. 2,B) were able to grow in HAT + G418; however, lung metastasis cells were completely resistant to these agents, growing at a rate similar to that seen with plain medium, with no lag period (Fig. 2 C). Resistance of lung metastasis cells to HAT + G418 suggested that these cells were hybrids between normal cells(HATres/G418sens) of the host mouse and implanted tumor cells(HATsens/G418res) in which complementation between the fusion partners conferred double drug resistance.

DNA Restriction Fragment Analyses for wt and Albino Tyrosinases.

A comparison of the nucleotide sequences of exons 1 and 2 of the albino tyrosinase gene with that of functional wt cDNA reveals one base change: a G→C transversion at nucleotide residue 387 of the albino gene. This results in a new CTNAG recognition site for DdeI(34, 35, 36) Thus, to test for hybrids at the genetic level,genomic DNA was amplified by PCR and analyzed by RFLP for the presence of wt (C/C) and albino mutant (c/c) tyrosinase by DdeI restriction fragment analyses (Fig. 3). DdeI digestion of wt Cloudman S91 6neo DNA resulted in fragments of 165, 113, and 63 bp. In contrast, digestion of normal liver DNA from a BALB/c nu/nu mouse (c/c) yielded no 165-bp band, but did yield a new band of 130 bp, along with the wt bands of 113 and 63 bp. Such DdeI restriction fragments from wt and albino tyrosinase were reported previously (34, 35, 36). Lung metastasis cells grown in plain culture medium or medium with HAT/G418 and in 12 of 12 individually derived lung metastasis subclones showed both the wt tyrosinase bands of 165, 113, and 63 bp and the albino band of 130 bp (Fig. 3,A). In contrast, primary tumor and tail metastasis cells showed only the wt tyrosinase pattern, with no evidence of the 130-bp albino band. The undigested PCR products are seen in Fig. 3,B. Genomic DNA DdeI restriction fragments of 165 (wt specific) and 130 bp (albino specific) were excised from the gel in Fig. 3 A for sequence analyses (data not shown). Sequences were determined for the 165-bp band from Cloudman S91 6neo cells, the 130-bp band from liver DNA of a BALB/c mouse, and both bands from lung metastasis clone 1. These sequences were then compared with the published sequences for these mouse tyrosinase restriction fragments (34). For both the 165- and 130-bp bands described above, the sequences were identical to those published previously, confirming the simultaneous presence of albino and tyrosinase genomic DNA in lung metastasis cells (data not shown).

DNA Content and Tyrosinase Activity.

In addition to the simultaneous presence of albino and wt tyrosinase DNA, all lung metastasis cell lines showed elevated DNA content,consistent with being hybrids (Table 1). DNA was quantitated by flow cytometry and expressed relative to that of parental 6neo cells, whose value was arbitrarily set at 1.00. The DNA content of uncloned lung metastasis cells (1.41) and all 12 lung metastasis subclones (1.30–1.48) was significantly elevated compared to the 6neo(1.00), primary tumor (1.04), and tail metastasis cells (1.07; P < 0.0001).

The lung metastasis cells also produced much more melanin than the other cells. We thus assayed for tyrosinase, a rate-limiting enzyme in melanogenesis. Tyrosinase activity was measured in lysates of cells with and without exposure to MSH/IBMX (Table 1). Tyrosinase activities were measured after 8–10 passages in culture for all cell lines. In the basal state, i.e., without MSH/IBMX, there was no tyrosinase activity in 6neo parental cells, cells of the primary tumor, or cells of the tail metastasis. In contrast,there was readily measurable basal tyrosinase activity and visible melanization in all but one of the lung metastasis lines. Clone 6 was the only lung metastasis clone that lacked basal tyrosinase activity and melanization; however, all of the lung metastasis clones, including clone 6, were strongly induced by MSH/IBMX for both these functions(Table 1; melanin content not shown). Thus, a strong correlation existed between the increased DNA content of lung metastasis cells and enhanced melanogenesis.

Chemotactic Motility.

Elevated DNA content also correlated with enhanced chemotactic motility in the uncloned lung metastasis population and in most of the lung metastasis clones (Table 2). For example, for cells grown in plain culture medium, 1–6% of parental 6neo cells, cells of the primary tumor,or cells of the tail metastasis migrated to the lower chamber in response to 3T3-conditioned medium in the 3-h migration assay, and treatment with MSH/IBMX had no effect on the migration of these cells. In contrast, e.g., for lung metastasis clone 5, 17% of the cells migrated after growth in plain medium, and 39% migrated after treatment with MSH/IBMX. There were different motility phenotypes in comparison to parental 6neo cells: (a)high basal migration, with little or no induction by MSH/IBMX (uncloned population, clones 6 and 8); (b) high basal migration with further induction by MSH/IBMX (clones 2, 5, and 9); (c) low basal migration with strong induction by MSH/IBMX (clones 1, 7, and 12); and (d) low basal migration and little or no stimulation by MSH/IBMX (similar to parental 6neo cells, e.g., clones 3, 4, 10, and 11).

Histopathology of the Primary Tumor and Lung Metastasis.

H&E-stained sections of the primary tumor were examined by light microscopy. In detailed surveys of serial sections not shown, the primary tumor was found to be devoid of melanin-containing cells,except in the region that was contiguous with the tail metastasis,where melanized cells were plentiful. The latter cells contained large,“coarse” melanin-containing granules that were much larger than individual melanosomes (Fig. 4, A–D). These large granular structures often obscured the nucleus, making it difficult to determine through nuclear atypia whether the cells were melanoma cells or melanophages (benign macrophages that also contain coarse melanin complexes; Refs.37, 38, 39, 40). An example of a melanized cell with its nucleus obscured by coarse melanin is seen in Fig. 4 D (red arrow). However, in sections where nuclei were visible in the melanized cells, the nuclei were large and atypical, similar to those of the nonmelanized primary tumor cells, and indicative of melanoma cells rather than melanophages.

In contrast to the primary tumor, virtually all cells of the lung metastasis were highly melanized, again with large, coarse melanin granules, as seen in a section from the biopsied lung metastasis showing melanoma cells adjacent to normal lung tissue (Fig. 5 A).

Lung Metastasis Cells in Culture.

Lung metastasis cells adapted to culture, and clones derived from them were comprised predominantly of large, coarse melanin-containing cells,as seen in the original culture explant after necropsy (Fig. 5,B). Electron microscopy revealed that these cells contained autophagosome-like structures with heavily melanized melanosomes (Fig. 6). The melanosome complexes varied in size, containing anywhere from less than 10 to hundreds of melanosomes, and appeared to account for the coarse melanin appearance by light and phase microscopy (compare Figs. 4 and 5). Not all lung metastasis melanosomes were in such complexes because melanosomes were also observed free in the cytoplasm(data not shown). In contrast, melanosomes in the parental 6neo cells were never observed within autophagosomal complexes, were few in number, and were amorphous in appearance with poorly defined matrix filaments (25, 28).

Gel Electrophoresis and Immunoblotting of LAMP-1.

We reported recently that the melanosomal proteins tyrosinase and TRP-2, as well as the melanosomal/lysosomal protein LAMP-1, when isolated from artificially constructed macrophage × melanoma fusion hybrids, migrated more slowly on gels than those from the parental Cloudman S91 cells, a phenomenon likely to be due to increased N-glycosylation of these proteins in the hybrids (25, 28). The gel migration of LAMP-1 from the cell lines described herein was thus studied (Fig. 7). In three of three such analyses, LAMP-1 from uncloned lung metastasis cells migrated more slowly on gels than LAMP-1 from the parental Cloudman S91, primary tumor, or tail metastasis cells, consistent with increased N-glycosylation of LAMP-1 in the lung metastasis cells.

We describe the in vivo progression of poorly metastatic amelanotic Cloudman S91 melanoma cells from an amelanotic tumor at their tail implantation site to a highly melanotic metastasis in the lungs, comprised primarily of host × tumor hybrids. By three criteria, the metastatic lung cells were host × tumor hybrids: (a) markedly increased DNA content; (b) growth resistance to HAT/G418; and(c) the simultaneous presence in genomic DNA of genes for both wt (tumor) and albino (host) tyrosinase. Contamination by normal host cells in the cultured lines seems an unlikely explanation for the results. Although such contamination might explain the presence of the 130-bp albino tyrosinase restriction fragment in primary cultures of tumor cells and normal stromal cells, it would not explain growth resistance to G418, to which normal cells are sensitive. Nor would it provide an explanation for the 130-bp albino tyrosinase fragment that was present in 12 of 12 clones isolated from liquid agar, a medium in which normal cells do not grow. Thus, we feel that the host × tumor hybrid nature of the lung metastasis cells was unambiguous.

What conditions might lead to tumor × host hybridization within tumors? We suggested previously that one function that might lead to hybrid formation is the ingestion of apoptotic tumor cells by macrophages (16). Solid tumors are generally rich in apoptotic cells (41). A key feature of apoptotic cells is their recognition by phagocytes and ingestion while still intact,protecting tissues from the potentially harmful consequences of exposure to the contents of dying cells (42, 43). Indeed, in vitro studies have recently demonstrated that genetic information from apoptotic cells can be transferred to both professional and nonprofessional phagocytic hosts, leading to hybrid formation (44, 45, 46). Thus, in principle, phagocytosis and digestion of apoptotic tumor cells could lead to transfer of genetic information and hybrid formation in vivo.

In vitro motility assays showed heterogeneity among individual lung metastasis clones, whose characteristics with regard to basal- and MSH/IBMX-inducible migration could be categorized into at least four phenotypes (Table 2). Interestingly, there was less heterogeneity among the clones with regard to tyrosinase activity,which was elevated in the basal state of 11 of 12 clones and strongly stimulated by MSH/IBMX in all 12 clones (Table 1). Explanations for clonal heterogeneity include considerations as to the number and nature of the hybrids formed. For example, the uncloned lung metastasis population could consist of multiple hybrids with different complements of parental chromosomes. Alternatively, the uncloned lung metastasis population could have arisen from a single hybrid that underwent subsequent chromosomal loss or rearrangement, leading to phenotypic heterogeneity. It is well known that hybrids between tumor cells and normal fibroblasts or epithelial cells undergo chromosome loss(47, 48, 49). Furthermore, karyotypes of artificial hybrids between human macrophages and Cloudman S91 6neocells revealed multiple examples of chromosomes with both human and mouse translocations that would also likely lead to heterogeneity(16). Warner (50) proposed that heterogeneity within tumors could be due to their hybrid nature. Likewise, Munzarova et al.(23) suggested that the heterogeneity of melanoma metastases as described by Clark et al.(51) might be due to hybridization.

Whereas fusion of a variety of cancer cells with normal fibroblasts or epithelial cells generally causes suppression of both malignancy and the expression of differentiated functions (47, 52),normal leukocytes fused with cancer cells have been reported to cause transactivation of differentiated functions between both parental genomes, e.g., in leukocyte × hepatoma hybrids (53, 54), leukocyte × myeloma hybrids (55), the well-studied immunoglobulin-secreting hybridomas (56), and the macrophage × melanoma hybrids described by our laboratory (16, 25, 26, 27, 28). Such suppression by fibroblasts and epithelial cells led to the discovery of tumor suppressor genes (47, 48, 49); however, we know of no explanation for the apparent lack of suppression in leukocyte-derived hybrids.

The lung metastasis cells expressed strikingly similar phenotypes to those of artificial macrophage × melanoma hybrids(16, 25, 26, 27, 28). Similarities included an enhanced basal- and MSH-inducible pigmentary system (16, 25, 28), acquisition of MSH-inducible chemotaxis toward 3T3-conditioned media(26), packaging of melanosomes in autophagosomal complexes, and evidence for macrophage-like N-glycosylation patterns(25, 28). Whereas we favor the notion that the host fusion partner was a macrophage or other professional phagocyte, it was not possible to determine the nature of the host fusion partner because, to our knowledge, no DNA characteristics have been reported that distinguish cells of different developmental lineages.

By the assays used, we were unable to detect hybrids among cells of the tail metastasis or the primary tumor. The observation that the cultured tail metastasis population contained highly melanotic cells, similar to those comprising the lung metastasis population, yet showed no hybrid traits was perplexing. It is possible that tail metastasis cells expressed some hybrid traits, but not the ones we assayed. Another explanation could be that because the number of melanotic cells in the tail metastasis population declined rapidly on passage in culture, they were below the detection limits as hybrids in the population. However,this latter explanation seems unlikely because even in tail metastasis populations where approximately 25% of the cells were melanotic, the cells showed no evidence of HAT/G418 resistance (Fig. 2), nor was there evidence of the 130-bp albino tyrosinase restriction fragment seen in BALB/c liver DNA and lung metastasis cells (Fig. 3,A). Both of these traits should have been within the assay limits. Because tail metastasis cells contained genomic DNA sequences for wt tyrosinase from 6neo cells, apparently one or more progenitor tail metastasis cells were generated in the primary 6neo tumor, migrated 1–2 mm down the tail, and then, for unknown reasons, ceased migrating and formed a tumor. Thus,although progenitor tail metastasis cells must have had the ability to migrate from the primary tumor (Fig. 1,A), at least some of them subsequently lost the ability to migrate further. Therefore,unanswered questions remain regarding potential relationships between melanotic melanoma cells in the primary tumor (Fig. 4) and melanotic cells of the tail and lung metastases.

Because only 10 mice were originally implanted with Cloudman S91 6neo cells, we do not know the true frequency of appearance of such hybrids. In previous studies, Cloudman S91 6neo cells implanted s.c. in the tails of DBA/2J mice developed metastases in 13% (14 of 105) of the mice. These were typically small (1–2 mm in diameter), nonlethal, amelanotic metastases(16). In the single experiment with BALB/c nu/nu mice reported here, metastases developed in only 1 of 10 mice implanted with Cloudman S91 6neo cells. In comparison, 100% of the BALB/c nu/nu mice implanted similarly with B16F10 melanoma cells developed lethal lung metastasis within 4–6 weeks(16).

In other reports, two spontaneous in vivo host × melanoma hybrids were described by others involving B16(57) and Cloudman S91 melanomas (58). Both hybrids were isolated from the primary tumors and not from metastases. We recently determined that the Cloudman S91 hybrid PADA (58) is highly metastatic by the s.c. tail assay used herein and that PADA cells, like the lung metastasis hybrids reported here, exhibit increased sensitivity to MSH/IBMX for both the pigmentary system and motility and show evidence for enhanced N-glycosylation of LAMP-1 (25, 28).

In light of our findings on a relationship between hybridization and enhanced pigmentation, it is interesting that a highly melanotic phenotype has also been noted in a class of human melanomas designated AMM (40). First described in horses (59), a morphological parallel to AMM was later noted in humans(60). Characteristics include abundant “coarse melanin pigment granules,” “often irregularly disposed, manifesting large globular deposits,” frequently “obscuring the nucleus, making melanoma cells difficult to distinguish from melanophages”(40). Because a similar phenotype was observed in both the spontaneous lung metastasis described herein and artificial macrophage × melanoma hybrids,4our results suggest hybridization as a potential explanation for the AMM phenotype.

Thus, several lines of evidence support the notion that host × tumor hybridization might be a mechanism for metastatic transformation. However, it is important to note that the above-mentioned results are correlative and do not necessarily prove a causal relationship between hybridization and metastasis because we could not formally rule out that the 6neo cells underwent metastatic transformation before hybridization. Nonetheless,we believe that our results, coupled with numerous earlier reports of intratumoral hybrid formation in animal models, indicate that this concept deserves serious consideration as a potential initiating event in the metastasis of melanoma and other solid tumors.

Fig. 1.

A, the tail of a BALB/c nu/nu mouse with both an amelanotic primary tumor arising from s.c. implanted cells of the Cloudman S91 melanoma line 6neo and a more proximal melanotic tail metastasis. B, the same mouse 5 weeks after the photo in A. The mouse died and was found to have a large melanotic pulmonary metastasis (arrows). Note that the tail had been transected as described in the text.

Fig. 1.

A, the tail of a BALB/c nu/nu mouse with both an amelanotic primary tumor arising from s.c. implanted cells of the Cloudman S91 melanoma line 6neo and a more proximal melanotic tail metastasis. B, the same mouse 5 weeks after the photo in A. The mouse died and was found to have a large melanotic pulmonary metastasis (arrows). Note that the tail had been transected as described in the text.

Close modal
Fig. 2.

Growth of various cell populations in the presence of culture media with (○) or without (•) HAT/G418. A,primary tumor cells; B, tail metastasis cells; C, lung metastasis cells. Values represent the mean ± SE for triplicate points. Where error bars are absent, the SE fell within the thickness of the circle. The experiments were repeated three times with similar results.

Fig. 2.

Growth of various cell populations in the presence of culture media with (○) or without (•) HAT/G418. A,primary tumor cells; B, tail metastasis cells; C, lung metastasis cells. Values represent the mean ± SE for triplicate points. Where error bars are absent, the SE fell within the thickness of the circle. The experiments were repeated three times with similar results.

Close modal
Fig. 3.

DNA restriction fragment analyses for wt and albino tyrosinases. A, DdeI digestion of genomic DNA prepared by PCR amplification. B, PCR amplification products before DdeI digestion. LM, lung metastasis cells, cloned or uncloned; LM H/N, cells were grown in HAT/G418 medium 1 week before assays.

Fig. 3.

DNA restriction fragment analyses for wt and albino tyrosinases. A, DdeI digestion of genomic DNA prepared by PCR amplification. B, PCR amplification products before DdeI digestion. LM, lung metastasis cells, cloned or uncloned; LM H/N, cells were grown in HAT/G418 medium 1 week before assays.

Close modal
Fig. 4.

A–D, different H&E-stained sections of the primary tumor in Fig. 1 A. Coarse melanin-containing cells with atypical nuclei are delineated by the black arrows. D contains an example of a coarse melanin-containing cell in which the nucleus was obscured (red arrow), preventing us from distinguishing whether this cell was a melanoma cell or a melanophage, as described in the text.

Fig. 4.

A–D, different H&E-stained sections of the primary tumor in Fig. 1 A. Coarse melanin-containing cells with atypical nuclei are delineated by the black arrows. D contains an example of a coarse melanin-containing cell in which the nucleus was obscured (red arrow), preventing us from distinguishing whether this cell was a melanoma cell or a melanophage, as described in the text.

Close modal
Fig. 5.

A, an H&E-stained section of the lung metastasis in Fig. 1 B showing normal lung tissue contiguous to metastatic melanoma tissue (arrowsdelineate melanoma). The melanoma was comprised predominantly of coarse melanin-containing cells. B, a lung metastasis explant after about 1 week in culture as seen by phase microscopy. Heavily melanized, coarse, melanin-containing cells (arrows)grew out of the tumor and formed a monolayer as described in the text.

Fig. 5.

A, an H&E-stained section of the lung metastasis in Fig. 1 B showing normal lung tissue contiguous to metastatic melanoma tissue (arrowsdelineate melanoma). The melanoma was comprised predominantly of coarse melanin-containing cells. B, a lung metastasis explant after about 1 week in culture as seen by phase microscopy. Heavily melanized, coarse, melanin-containing cells (arrows)grew out of the tumor and formed a monolayer as described in the text.

Close modal
Fig. 6.

An electron micrograph of a view within a cultured lung metastasis cell. Melanosomes are complexed in autophagosome-like structures, as described in the text.

Fig. 6.

An electron micrograph of a view within a cultured lung metastasis cell. Melanosomes are complexed in autophagosome-like structures, as described in the text.

Close modal
Fig. 7.

Gel electrophoresis and immunoblotting of LAMP-1 from lysates of Cloudman S91 6neo cells and cultured cells from the primary tumor (Primary), tail metastasis(TM), and lung metastasis (LM) described in the text.

Fig. 7.

Gel electrophoresis and immunoblotting of LAMP-1 from lysates of Cloudman S91 6neo cells and cultured cells from the primary tumor (Primary), tail metastasis(TM), and lung metastasis (LM) described in the text.

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

Supported by a grant from Vion Pharmaceuticals(New Haven, CT). A. C. and S. S. contributed equally to this project.

3

The abbreviations used are: HAT,10−4m hypoxanthine, 4 × 10−7m aminopterin, and 1.6 × 10−5m thymidine; LAMP-1, lysosome-associated membrane protein 1; wt, wild-type; AMM, animal type melanoma; MSH,melanocyte stimulating hormone; IBMX, isobutylmethylxanthine.

4

J. Pawelek and A. Key-Yen, unpublished observations.

Table 1

Tyrosinase activity in various cell lysates with and without exposure of cells to MSH/IBMX in culture (72 h)

DNA content and tyrosinase activity of Cloudman S91 6neo cells was compared to that of uncloned populations of the primary tumor, tail metastasis, and lung metastasis cells (Fig. 1) and a panel of 12 individual clones of lung metastasis cells. DNA content was determined by flow cytometry of a minimum of 15,000 cells/sample and was normalized relative to that of parental Cloudman S91 6neocells, which was set arbitrarily at 1.00. Tyrosinase assays are described in “Materials and Methods.”

Cells of originDNA contentTyrosinase activity (cpm 3H20/106 cells/30 min)
No treatment±MSH/IBMX
6neo (parental cells) 1.00 55 ± 34 
Primary tumor 1.04 180 ± 82 
Tail metastasis 1.07 404 ± 33 
Lung metastasis 1.41 53 ± 11 178 ± 155 
Lung metastasis clone    
1.38 118 ± 78 1388 ± 249 
1.30 223 ± 50 1213 ± 232 
1.41 97 ± 27 775 ± 125 
1.41 386 ± 27 1815 ± 57 
1.45 86 ± 39 1160 ± 144 
1.45 1048 ± 247 
1.45 72 ± 67 1536 ± 113 
1.46 141 ± 31 1586 ± 171 
1.42 423 ± 49 2176 ± 259 
10 1.34 363 ± 32 1461 ± 482 
11 1.40 140 ± 68 1431 ± 181 
12 1.48 97 ± 41 1901 ± 388 
Cells of originDNA contentTyrosinase activity (cpm 3H20/106 cells/30 min)
No treatment±MSH/IBMX
6neo (parental cells) 1.00 55 ± 34 
Primary tumor 1.04 180 ± 82 
Tail metastasis 1.07 404 ± 33 
Lung metastasis 1.41 53 ± 11 178 ± 155 
Lung metastasis clone    
1.38 118 ± 78 1388 ± 249 
1.30 223 ± 50 1213 ± 232 
1.41 97 ± 27 775 ± 125 
1.41 386 ± 27 1815 ± 57 
1.45 86 ± 39 1160 ± 144 
1.45 1048 ± 247 
1.45 72 ± 67 1536 ± 113 
1.46 141 ± 31 1586 ± 171 
1.42 423 ± 49 2176 ± 259 
10 1.34 363 ± 32 1461 ± 482 
11 1.40 140 ± 68 1431 ± 181 
12 1.48 97 ± 41 1901 ± 388 
Table 2

Motility of various cell lines with 3T3-conditioned medium as a chemoattractant

Migration of cells to the underside of a two-chambered Costar Transwell(12 μm, pore size) system was assayed in response to 3T3 fibroblast-conditioned medium (33% v/v DMEM) in the bottom chamber. Migration of Cloudman S91 6neo parental melanoma cells was compared with that of uncloned populations of the primary tumor, tail metastasis, and lung metastasis (Fig. 1) and 12 individual clones of lung metastasis cells. Results are presented as the percentage of total cells migrated in 3 h (mean ± SE) for 3–12 determinations. Cells were grown in plain culture medium or in medium containing MSH/IBMX for 72 h, harvested, and assayed for motility(see “Materials and Methods”).

CellsMotility (% cells migrated)
Control+MSH/IBMX
6neo (parental cells) 4 ± 1 5 ± 2 
Primary tumor 4 ± 2 1 ± 0 
Tail metastasis 5 ± 3 6 ± 1 
Lung metastasis 15 ± 2 12 ± 3 
Lung metastasis clone   
3 ± 1 46 ± 3 
13 ± 3 31 ± 3 
3 ± 1 6 ± 3 
5 ± 2 11 ± 5 
17 ± 2 39 ± 5 
34 ± 8 27 ± 5 
2 ± 1 20 ± 2 
12 ± 6 10 ± 3 
13 ± 5 24 ± 3 
10 2 ± 0 11 ± 3 
11 2 ± 2 8 ± 2 
12 1 ± 0 23 ± 1 
CellsMotility (% cells migrated)
Control+MSH/IBMX
6neo (parental cells) 4 ± 1 5 ± 2 
Primary tumor 4 ± 2 1 ± 0 
Tail metastasis 5 ± 3 6 ± 1 
Lung metastasis 15 ± 2 12 ± 3 
Lung metastasis clone   
3 ± 1 46 ± 3 
13 ± 3 31 ± 3 
3 ± 1 6 ± 3 
5 ± 2 11 ± 5 
17 ± 2 39 ± 5 
34 ± 8 27 ± 5 
2 ± 1 20 ± 2 
12 ± 6 10 ± 3 
13 ± 5 24 ± 3 
10 2 ± 0 11 ± 3 
11 2 ± 2 8 ± 2 
12 1 ± 0 23 ± 1 

Flow cytometry was carried out by Rocco Carbone and supported by the Yale Cancer Center Flow Cytometry Shared Resource. Electron microscopy was performed by Agnes Key-Yen.

1
Van Noorden C. J. F., Meade-Tollin L. C., Bosman F. T. Metastasis.
Am. Sci.
,
86
:
130
-141,  
1998
.
2
Stracke M. L., Liotta L. A. Molecular mechanisms of tumor cell metastasis Mendelsohn J. Howley P. M. Israel M. A. Liotta L. A. eds. .
The Molecular Basis of Cancer
,
:
233
-247, W. B. Saunders Co. Philadelphia  
1995
.
3
Graeber T. G., Osmanian C., Jacks T., Housman D. E., Koch C. J., Lowe S. W., Giaccia A. J. Hypoxia-mediated selection of cells with diminished apoptotic potential in solid tumours.
Nature (Lond.)
,
379
:
88
-91,  
1996
.
4
Nowell P. C. The clonal evolution of tumor cell populations.
Science (Washington DC)
,
194
:
23
-28,  
1976
.
5
Fidler I. J., Kripke M. L. Metastasis results from preexisting variant cells within a malignant tumor.
Science (Washington DC)
,
197
:
893
-895,  
1977
.
6
Collard J. G., Roos E., La Riviere G., Habets G. G. M. Genetic analysis of invasion and metastasis.
Cancer Surv.
,
7
:
691
-710,  
1988
.
7
Cahill D. P., Lengauer C., Yu J., Riggins G. J., Willson J. K. V., Markowitz S. D., Kinzler K. W., Vogelstein B. Mutations of mitotic checkpoint genes in human cancers.
Nature (Lond.)
,
392
:
300
-304,  
1998
.
8
Lengauer C., Kinzler K. W., Vogelstein B. Genetic instabilities in human cancers.
Nature (Lond.).
,
396
:
643
-649,  
1998
.
9
Morita R., Fujimoto A., Hatta N., Takehara K., Takata M. Comparison of genetic profiles between primary melanomas and their metastases reveals genetic alterations and clonal evolution during progression.
J. Investig. Derm.
,
111
:
919
-924,  
1998
.
10
Aichel O. Uber Zellverschmelzung mit qualitative abnormer chromosomenverteilung Roux W. eds. .
Vorträge und Aufsätze über Entvickelungsmechanik Der Organismen
,
:
1
-115, Wilhelm Engelmann Leipzig, Germany  
1911
.
11
Mekler L. B. Hybridization of transformed cells with lymphocytes as one of the probable courses of the progression leading to the development of metastatic malignant cells (in Russian).
Vestn. Akad. Med. Nauk SSSR
,
26
:
80
-89,  
1971
.
12
Kerbel R. S., Lagarde A. E., Dennis J. W., Donaghue T. P. Spontaneous fusion in vivo between normal host and tumor cells: possible contribution to tumor progression and metastasis studied with a lectin-resistant mutant tumor.
Mol. Cell. Biol.
,
3
:
523
-538,  
1983
.
13
Larizza L., Schirrmacher V. Somatic cell fusion as a source of genetic rearrangement leading to metastatic variants.
Cancer Metastasis Rev.
,
3
:
193
-222,  
1984
.
14
DeBaetselier P., Roos E., Brys L., Remels L., Feldman M. Generation of invasive and metastatic variants of a non-metastatic T-cell lymphoma by in vivo fusion with normal host cells.
Int. J. Cancer
,
34
:
731
-738,  
1984
.
15
Lagarde A. E., Kerbel R. S. Somatic cell hybridization in vivo and in vitro in relation to the metastatic phenotype.
Biochim. Biophys. Acta
,
823
:
81
-110,  
1985
.
16
Rachkovsky M., Sodi S., Chakraborty A., Avissar Y., Bolognia J., Madison McNiff J., Platt J., Bermudes D., Pawelek J. Melanoma × macrophage hybrids with enhanced metastatic potential.
Clin. Exp. Metastasis
,
16
:
299
-312,  
1998
.
17
Margulis L., Sagan D. Microcosmos: Four Billion Years of Evolution from Our Microbial Ancestors
281
Simon and Schuster New York  
1991
.
18
Martin W., Muller M. The hydrogen hypothesis for the first eukaryote.
Nature (Lond.)
,
392
:
37
-41,  
1998
.
19
Weinstein I. B., Carothers A. M., Santella R. M., Perera F. P. Molecular mechanisms of mutagenesis and multistage carcinogenesis Mendelsohn J. Howley P. M. Israel M. A. Liotta L. A. eds. .
The Molecular Basis of Cancer
,
:
59
-85, W. B. Saunders Co. Philadelphia  
1995
.
20
Gauwerky C. E., Croce C. M. Molecular genetics and cytogenetics of hematopoietic malignancies Mendelsohn J. Howley P. M. Israel M. A. Liotta L. A. eds. .
The Molecular Basis of Cancer
,
:
18
-37, W. B. Saunders Co. Philadelphia  
1995
.
21
Wikonkal N. M., Brash D. E. Ultraviolet radiation induced signature mutations in photocarcinogenesis J.
Investig. Dermatol.
,
4
:
6
-10,  
1999
.
22
Munzarova M., Kovrik J. Is cancer a macrophage-mediated autoaggressive disease?.
Lancet
,
i
:
952
-954,  
1987
.
23
Munzarova M., Lauerova L., Capkova J. Are advanced malignant melanoma cells hybrids between melanocytes and macrophages?.
Melanoma Res.
,
2
:
127
-129,  
1992
.
24
Pawelek J. M. Melanoma as a macrophage/melanocyte hybrid and the symbiotic nature of eukaryotic cells.
Melanoma Res.
,
3
:
75
-76,  
1993
.
25
Sodi S. A., Chakraborty A. K., Platt J. T., Kolesnikova N., Rosemblat S., Keh-Yen A., Bolognia J. L., Rachkovsky M. L., Orlow S. J., Pawelek J. M. Melanoma × macrophage fusion hybrids acquire increased melanogenesis and metastatic potential: altered N-glycosylation as an underlying mechanism.
Pigm. Cell Res.
,
11
:
299
-309,  
1998
.
26
Rachkovsky M., Pawelek J. Acquired melanocyte stimulating hormone-inducible chemotaxis following macrophage fusion with Cloudman S91 melanoma cells.
Cell Growth Differ.
,
10
:
515
-524,  
1999
.
27
Chakraborty A. K., Funasaka Y., Ichihashi M., Sodi S., Bhattacharya M., Pawelek J. Up-regulation of mRNA for the melanocortin-1 receptor but not for melanogenic proteins in macrophage × melanoma fusion hybrids exhibiting increased melanogenic and metastatic potential.
Pig. Cell Res.
,
45
:
1011
-1027,  
1999
.
28
Pawelek J. M., Chakraborty A. K., Rachkovsky M. L., Orlow S. J., Bolognia J. L., Sodi S. A. Altered N-glycosylation in macrophage × melanoma fusion hybrids.
Cell. Mol. Biol.
,
45
:
1011
-1027,  
1999
.
29
Littlefield J. W. The use of drug-resistant markers to study the hybridization of mouse fibroblasts.
Exp. Cell Res.
,
41
:
190
-196,  
1966
.
30
Crissman H. A., Hirons G. T. Staining of DNA in live and fixed cells.
Methods Cell Biol.
,
41
:
195
-209,  
1994
.
31
Pawelek J. M. Barnes D. W. Sirbasku D. A. Sato G. H. eds. .
Methods for Serum-Free Culture of Neuronal and Lymphoid Cells
,
:
57
-66, Alan R. Liss, Inc. New York  
1984
.
32
Orlow S. J., Boissy R. E., Moran D. J., Pilfko-Hirst S. Subcellular distribution of tyrosinase and tyrosinase-related protein-1: implications for melanosomal biogenesis.
J. Investig. Dermatol.
,
100
:
55
-64,  
1993
.
33
Pawelek J., Sansone M., Morowitz J., Moellmann G., Godawska E. Genetic control of melanization: isolation and analysis of amelanotic variants from cultured melanotic melanoma cells.
Proc. Natl. Acad. Sci. USA
,
71
:
1073
-1077,  
1974
.
34
Muller G., Ruppert S., Schmid E., Schutz G. Functional analysis of alternatively spliced tyrosinase gene transcripts.
EMBO J.
,
7
:
2723
-2730,  
1988
.
35
Shibahara S., Okinaga S., Tomita Y., Takeda A., Yamamoto H., Sato M., Takeuchi T. A point mutation in the tyrosinase gene of BALB/c albino mouse causing the cysteine-serine substitution at position 85.
Eur. J. Biochem.
,
189
:
455
-461,  
1990
.
36
Jackson I. J., Bennett D. C. Identification of the albino mutation of mouse tyrosinase by analysis of an in vitro revertant.
Proc. Natl. Acad. Sci. USA
,
87
:
7010
-7014,  
1990
.
37
Mishima Y. Cellular and subcellular differentiation of melanin phagocytosis and synthesis by lysosomal and melanosomal activity.
J. Investig. Dermatol.
,
46
:
70
-75,  
1966
.
38
Hori Y., Toda K., Pathak M. A., Clark W. H., Jr., Fitzpatrick T. B. A fine-structure study of the human epidermal melanosome complex and its acid phosphatase activity.
J. Ultrastruct. Res.
,
25
:
109
-120,  
1968
.
39
Horikoshi T., Jimbow K., Sugiyama S. Comparison of macromelanosomes and autophagic giant melanosome complexes in nevocellular nevi, lentigo simplex, and malignant melanoma.
J. Cutan. Pathol.
,
9
:
329
-339,  
1982
.
40
Crowson A. N., Magro C. M., Mihm M. C., Jr. Malignant melanoma with prominent pigment synthesis: “animal type” melanoma—a clinical and histological study of six cases with a consideration of other melanocytic neoplasms with prominent pigment synthesis.
Hum. Pathol.
,
30
:
543
-550,  
1999
.
41
Jain R. K. Barriers to drug delivery in solid tumors.
Sci. Am.
,
271
:
58
-65,  
1994
.
42
Savill J., Fadok V., Henson P., Haslett C. Phagocyte recognition of cells undergoing apoptosis.
Immunol. Today
,
14
:
131
-136,  
1993
.
43
Fadok V. A., Henson P. M. Apoptosis.
Getting rid of the bodies. Curr. Biol.
,
8
:
R693
-R695,  
1998
.
44
de la Taille A., Chen M-W., Burchardt M., Chopin D. K., Buttyan R. Apoptotic conversion: evidence for exchange of genetic information between prostate cancer cells mediated by apoptosis.
Cancer Res.
,
59
:
5461
-5463,  
1999
.
45
Holmgren L., Szeles A., Rajnavolgyi E., Folkman J., Klein G., Ernberg I., Falk K. I. Horizontal transfer of DNA by the uptake of apoptotic bodies.
Blood
,
93
:
3956
-3963,  
1999
.
46
Spetz A. L., Patterson B. K., Lore K., Andersson J., Holmgren L. Functional gene transfer of HIV DNA by an HIV receptor-independent mechanism.
J. Immunol.
,
163
:
736
-742,  
1999
.
47
Harris H. The analysis of malignancy by cell fusion: the position in 1988.
Cancer Res.
,
48
:
3302
-3306,  
1988
.
48
Weinberg A. S. Tumor suppressor genes.
Science (Washington DC)
,
254
:
1138
-1146,  
1991
.
49
Levine A. J. Tumor suppressor genes Mendelsohn J. Howley P. M. Israel M. A. Liotta L. A. eds. .
The Molecular Basis of Cancer
,
:
86
-104, W. B. Saunders Co. Philadelphia  
1995
.
50
Warner T. F. C. S. Cell hybridization. An explanation for the phenotypic diversity of certain tumours.
Med. Hypotheses
,
1
:
51
-57,  
1975
.
51
Clark W. H., Jr., Mastrangelo M. J., Ainsworth A. M., Berg D., Bellet R. E., Bernardino E. A. Current concepts of the biology of human cutaneous malignant melanoma.
Adv. Cancer Res.
,
24
:
267
-338,  
1997
.
52
Davidson R. L., Ephrussi R. L. B., Yamamoto K. Regulation of pigment synthesis in mammalian cells, as studied by somatic hybridization.
Proc. Natl. Acad. Sci. USA
,
56
:
1437
-1440,  
1966
.
53
Darlington G. J., Bernhard H. P., Ruddle F. H. Human serum albumin phenotype activation in mouse hepatoma–human leukocyte cell hybrids.
Science (Washington DC)
,
185
:
859
-862,  
1974
.
54
Malawista S. E., Weiss M. C. Expression of differentiated function in hepatoma cell hybrids: high frequency of induction of mouse albumin production in rat hepatoma-mouse lymphoblast hybrids.
Proc. Natl. Acad. Sci. USA
,
71
:
927
-931,  
1974
.
55
Giacomoni D. Tumorigenicity and intracisternal A-particle expression of hybrids between murine myeloma and lymphocytes.
Cancer Res.
,
39
:
4481
-4484,  
1979
.
56
Kohler G., Milstein C. Continuous cultures of fused cells secreting antibody of predefined specificity.
Nature (Lond.)
,
256
:
495
-497,  
1975
.
57
Hu F., Pasztor L. M. In vivo hybridization of cultured melanoma cells and isogenic normal mouse cells.
Differentiation
,
4
:
93
-97,  
1975
.
58
Halaban R., Nordlund J., Francke U., Moellmann G., Eisenstadt J. M. Supermelanotic hybrids derived from mouse melanomas and normal mouse cells.
Somatic Cell Genet.
,
6
:
29
-44,  
1980
.
59
Dick W. Melanosis in men and horses.
Lancet
,
i
:
192
1832
.
60
Darier J. Le melanome malin mesenchymateaux ou melanosarcome.
Bull. Assoc. Fr. Cancer
,
14
:
221
-249,  
1925
.