Matrix Proteins Pathobiology 1991;59:316-323
Matrix Proteins Induce Neuroblastoma Cell Differentiation without Altering Cell Growth Takuro Kidowaki3-1, C.J. Thiele a , H.K. Kleinm anb, Mark A. Israela a Pediatric Branch, National Cancer Institute and the bNational Institute for Dental Research, Bethesda, Md., USA
Key Words. Extracellular matrix • Differentiation • Neuroblastoma • Retinoic acid • Growth regulation
Introduction During development, cells originating in the em bryonal neural crest migrate and proliferate giving rise to highly specialized tissues [1] including the neuronal and glial cells of the peripheral nervous system (PNS) [2, 3], Although the developmental fate of some cells may be established before migration from the neural crest [4, 5], substances in the local environment can affect their migratory behavior [6] as well as their proliferation and differentiation [2,4, 5]. The naturally occurring environ mental signals mediating these processes include hor monal factors [7] and components of the extracellular matrix [8], 1 Dr. Kidowaki was the recipient of a Nakasone Fellow at the National Cancer Institute.
Extracellular matrix proteins influence the differenti ation of several different cell types [8-11]. In vivo they can trigger the diversification of neural crest cells into the various lineages that give rise to PNS tissues [8]. Laminin (LN) and fibronectin (FN) are matrix proteins which, in vitro, promote neuritic outgrowth from cells of both the central and PNS [12-16]. These proteins also promote adrenergic differentiation of neural crest cells in vitro [17-19]. Although neither FN nor LN is known to be synthesized by neural crest cells [13], both are found in varying amounts along neural crest migration pathways where they may act as signals for cellular dif ferentiation and tissue morphogenesis [20-22]. Neuroblastoma (NB), a malignant tumor of very early childhood, is thought to originate in neural crest cells [23]. Although this tumor most commonly arises in the adrenal medulla, it also occurs in sympathetic ganglia
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
Abstract. The maturation of embryonal neural crest cells is thought to be regulated in part by the milieu into which these cells migrate. Neuroblastoma (NB) is a tumor of very early childhood that is thought to arise in asso ciation with the arrested differentiation of embryonal neural crest cells. In culture, neuroblastoma tumor cells differentiate in the presence of retinoic acid, which is also known to influence extracellular matrix protein synthesis. We have cultured neuroblastoma cells on laminin (LN) and fibronectin (FN) substrata to examine the role of extracellular matrix in retinoic acid (RA)-induced differentiation of these tumor cells. These proteins caused mor phologic changes in NB cells indistinguishable from those caused by RA. Antiserum to each of these proteins blocked the effects induced by the corresponding protein, but neither antiserum affected the action of RA. Despite the induction of a neuronal morphologic change, matrix proteins did not alter the proliferation of NB cells. These results indicate that LN and FN modulate the differentiation of NB cells without inducing growth arrest and that RAinduced differentiation does not require these matrix proteins.
and at sites without known PNS tissue [24]. Such sites may be the locations of embryonal neural crest cells that did not successfully complete their migration. The origin of NB in embryonic tissue and its capacity to differen tiate both in vitro and in vivo suggests that arrested dif ferentiation may have a role in its pathogenesis [23, 25, 26]. Retinoic acid (RA) has been the most extensively studied agent to induce differentiation of neuroblastoma cells [27, 28] and is the only known naturally occurring morphogen [29]. We investigated the possible regulatory role of matrix glycoproteins in the differentiation of NB tumor cells [30-32], We found that both FN and LN can initiate the morphologic differentiation of these tumor cells and can enhance the efficacy with which RA induces their in vitro maturation. While RA has been shown to increase the production of matrix proteins as much as 50-fold in embryonal carcinoma cells [33] and chondrocytes [34], we were not able to demonstrate a role for endogenously synthesized matrix proteins in RA-induced neuroblas toma tumor cell differentiation. Materials and Methods Materials Mouse LN, human FN, goat antimouse LN antiserum, and rab bit antihuman FN antiserum were prepared as previously report ed [35-38]. Poly-D-lysine and RA were purchased from Sigma (St. Louis, Mo.). (3H)-thymidine (52.1 Ci/mmol) and (32P)-dCTP (3,000 Ci/mmol) were obtained from New England Nuclear (Bos ton, Mass.). Cells Cells from the human NB cell lines SK-N-BE(2) [39] and SMSK.CNR [40] were maintained in RPMI-1640 supplemented with 10% fetal calf serum (FCS), glutamine, penicillin and streptomycin in 75-cm tissue culture flasks at 37 °C in 5% CO2. Assays o f Morphologic Change and Cell Growth Assays were performed in 96-well tissue culture plates from Cos tar (Philadelphia, Pa.) or immunologic 96-well plates coated with extracellular matrix proteins (10 pg/0.32 cm2) or poty-£>-lysine [16]. To coat plates with matrix proteins, an aqueous solution containing the designated amount of matrix protein was added to each cell, incubated at 37 °C for 1 h, and the supernate was removed. Loga rithmically growing cells in medium containing 10% FCS were har vested with EDTA and dispersed by gentle pipetting. They were then washed, suspended in a solution at 5 X 103 cells/0.2 ml of RPMI-1640 containing 5% FCS, seeded in plates, and cultured at 37 °C in 5% CO2 [16]. Morphologic changes in the cells were observed microscopically and scored after 1, 2, 4, and 6 days. Sim ilar experiments were also performed in the presence of 0.1 \rM alltrans RA (Sigma, St. Louis, Mo.). The effects of antisera to specific matrix proteins on the differentiation of NB cells was assessed by incubating NB cells in matrix protein-coated wells containing media
317
to which antisera against specific matrix proteins [35-37, 40] or a control antisera had been added to a final concentration of 5 %. Cell proliferation was measured by a (3H)-thymidine incorpora tion assay of triplicate cultures: 1 pCi of (3H)-thymidine in 10 pi of phosphate-buffered saline was added to the medium of each well and 20 h later the cells were harvested by a PHD cell harvester (Cambridge Scientific, Cambridge, Mass.). Radioactivity was mea sured in a liquid scintillation counter. Graphical Analysis To display and compare graphically changes in the morphology of NB tumor cells cultured under various conditions, we counted 200 cells in each well and devised a graph to represent the distribu tion of cell types. Typically neuroblastoma tumor cells are small, round, and loosely adherent. Fully differentiated neuroblastoma cells appear indistinguishable from neurons in primary culture with long slender neuritic outgrowths exhibiting varicosities and making contact with distantly located cells. We designated small round tumor cells as - , flat adherent tumor cells as +, cells with processes shorter than the cell body from which they extended as ++, and cells with processes that were longer that the cell bodies as +++. At points corresponding to the time after treatment when we scored these cells, we graphically displayed our results as a vertical line extending from the least differentiated to the most differentiated type seen. This line formed one side of a triangle and indicated the range of the cell shapes. A point corresponding to the cell type most frequently observed was plotted as the third point of the triangle. The kinetics of these changes were indicated by a graph connecting points corre sponding to the most common cell type seen at each evaluation. Adhesion Assay NB cells were labeled by culturing for 24 h in 5 pCi/ml of (3H)thymidine in RPMI-1640 containing 10% FCS. The cells were har vested with EDTA, washed, and plated in 96-well immunologic plates coated with matrix proteins or poly-D-lysine and incubated in RPMI-1640 with 5% FCS for 1 h at 37 °C. Nonadherent cells were removed by washing the cells with media. The remaining adherent cells were harvested with EDTA. Nonadherent and adherent cells were solubilized in Aquasol (New England Nuclear, Boston, Mass.) and their radioactivity measured by liquid scintillation counting. Ribonucleic Acid Isolation and Northern Blots NB cells were cultured in 150-cm2 plates coated with matrix proteins. Total cellular ribonucleic acid (RNA) was isolated from the cells by guanidine thiocyanate extraction and centrifugation over a cesium chloride cushion [41]. Twenty-microgram samples of each RNA preparation were size-fractionated by electrophoresis on a 1% agarose gel containing 2.2 M formaldehyde and were trans ferred to Nytran (Schleicher & Schuell, Keene, N.H.) for Northern blot analysis as previously described [42]. Radiolabeled DNA probes were synthesized by nick translation [43] from the following plasmids: pNF, which contains a 1.3-kb deoxyribonucleic acid (DNA) fragment of pHNFl [intermediate-sized neurofilament, 44]; pHCGA, which contains a cDNA encoding human chromogranin A [45]; pNB-1, which contains a cDNA of N-myc [46]; pLR4-4, which contains a cDNA encoding the LN receptor [40]; and pGAPDH, which contains a cDNA encoding glyceraldehyde-3-phosphate dehy drogenase [47], To evaluate expression of GAP43, we synthesized a 30-nucleotide oligomer corresponding to nucleotide 181-211 of GAP43 (47a) using the Applied Biosystem 381A DNA synthesizer
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
ECM and Neuroblastoma
Kidowaki/Thiele/Kleinman/Israel
318
0 Day
1 Day
2 Days
5 Days
1
A SMS-KCNR SK-N-BEI2) SK-N-BEI2I
B SMS-KCNR SK-N-BEI2) SK-N-BE12)
(Foster City, Calif.). The oligomers were radiolabeled with J2PdATP and terminal transferase. Blots were hybridized for 20 h at 42 °C in 50% formamide and 5X SSPE. The blots were washed twice for 15 min at 25 °C in 6X SSC, twice for 15 min at 37 °C in 2X SSC, and once for 60 min at 65 °C in 0. IX SSC. The mem branes were exposed to Kodak X-Omat AR film with intensifying screens at -7 0 °C.
Fig. 1. Induced differentiation of NB cells. SK-N-BE(2) cells were cul tured in RPMI-1640 supplemented with either 10% FCS and 5 \iM RA 0 day (a), 1 day (b). 2 days (c), and 5 days (d) after the start of culture or 5% FCS and LN (500 pg/150cm2 of well floor) 0 h (e). 6 h (0. 1 day (g). and 2 days (h) after the start of cul ture. The extent of morphologic change was defined as follows: - = round cells, + = flat cells, ++ = cells with processes shorter than cell bod ies, +++ = cells with processes longer than cell bodies. Fig. 2. Effects of extracellular ma trix proteins and RA on human NB cells. SMS-KCNR and SK-N-BE(2) cells were harvested with EDTA and seeded at 5 X I0-1 cells in 0.2 ml of RPMI-1640 containing 5% FCS in 96-well immunologic plates coated with poly-D-lysine (PLYS; 1 pg/0.32 cm2 of well floor), LN. and FN (10 gg/0.32 cm2 of well floor, each). RA was used in a concentration of 0. 1 \iM. A Morphology was observed microscopically 1, 2, 4, and 6 days after the initiation of culture. The ex tent of morphologic changes was shown as four stages defined in figure 1. For each triangle in the figure, height and position represent the range of morphologic types in the cul ture, and the position of the left comer indicates the most common cell type seen. B Cell growth was de termined by a (3H)-thymidine incor poration assay at the indicated time. Values are presented as a mean value of counts per minute. C = Control.
Results
The NB tumor cell line SK-N-BE(2) undergoes se quential morphologic changes in the presence of 5 pAf RA (fig. la-d). The small round cells typical of NB tumor cell lines initially adhere to the cell culture sub
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
Time
strate and then flatten into an epithelioid shape. This change in shape is followed by the sprouting of short processes that can grow to several times the length of the cell body and can associate with other neuritic processes in a manner suggesting the formation of nerve fiber bun dles, after several days in culture [27], Cells from the NB cell line SMS-KCNR undergo similar changes, extending very long neurites [48], and other NB cell lines also change comparably when treated with RA [27, 48]. Ultrastructural and biochemical changes indicative of neuronal maturation also occur following the treatment of NB tumor cell lines with RA [27, 28, 49], When plated in the same media in the absence of RA, however, these cells typically remain small and round growing in loosely adherent clumps, which sometimes associate with rare flatter, more adherent cells on the culture vessel sur face. FN and LN when tested individually induce attach ment, flattening, and neurite outgrowth of SK-N-BE(2) cells indistinguishable from the changes that occurred after RA treatment. The changes observed following FN treatment [data not shown] were indistinguishable from those seen after LN (fig. le-h). In the case of both LN and FN the morphologic changes occurred more rapidly than those following RA treatment. To investigate fur ther the time course of these changes and to compare the effects of RA and matrix proteins quantitatively, SK-NBE(2) and SMS-KCNR cells [48, 50] were evaluated in the presence of either RA or matrix proteins or both (fig. 2). Within 2 days of the time NB cells were plated on either LN or FN, most cells were adherent and had flat tened (fig. 2A). This contrasted with the considerably slower time course over which RA induced these changes. Also of note was the more rapid time course over which morphologic changes suggestive of neuritic maturation could be observed in cultures to which either LN or FN was added. Interestingly, a greater proportion of cells undergo more extensive maturational morpho logic changes when cultures that had been grown on a surface coated with matrix proteins are treated with RA. Treatment with FN or with LN caused changes in SMSKCNR cells indistinguishable from those observed in SK-N-BE(2) cells. Although all cells in these cultures do not undergo morphologic changes synchronously, most have done so after several days of treatment. The in vitro maturation of NB cells is typically asso ciated with diminished growth potential [48] and the continuous proliferation of cells that do not become fully mature may be a key event in the development of embryonal malignancies. We therefore sought to deter
319
mine whether changes in tumor cell proliferation accom panied the maturation of NB cells that occurred in the presence of matrix proteins. In these experiments, we monitored the effects of treatment on the growth of KCNR and SK-N-BE(2) by assaying ,H-thymidine in corporation (fig. 2B) or cell number [data not shown]. In agreement with previous experiments from our lab [48, unpubl. data], we found that RA treatment inhibited the growth of both neuroblastoma cell lines examined, whether in the presence of extracellular matrix proteins or not (fig. 2B). In sharp contrast to this finding was the observation that there was no detectable effect by either LN or FN on proliferation even at late times, e.g. 6 days, when morphologic changes indicative or differentiation had been easily observable for several days (fig. 2B). We examined the effect of poIy-£>-lysine, a highly anionic substance which enhances the attachment of most tissue culture cells to plastic surfaces, to determine whether simply the adherence of neuroblastoma cells to plastic surfaces could induce neurite outgrowth. Poly£>-lysine caused cells to adhere to the growth surface (fig. 3), but did not cause them to change morphologi cally (fig. 2), suggesting that adhesion itself does not bring about subsequent changes in shape. We utilized antisera specifically reactive against either LN or FN to block the effects of these matrix pro teins and to examine whether the effect of RA was likely to be mediated by the production of endogenous extra cellular matrix protein synthesis. Each antiserum blocked the action of its respective matrix protein, but had no effect on the action of the other protein (fig. 4). Neither antiserum inhibited the action of RA. Previous studies have shown that RA- [48. 50, 51] and phorbol esters [52] induced morphologic differenti ation of some neuroblastoma tumor cell lines is accom panied by changes in gene expression. These include increases in mRNA levels of several genes whose expres sion is closely associated with the neuronal phenotype, such as neurofilament [50], and GAP43 [C.J.T. unpubl. data] and decreases in the steady state level of mRNA encoding chromogranin, a highly expressed gene in neu ral crest-derived chromaffin cells [53], In NB cells treated with FN or LN, however, these changes do not occur in association with morphologic differentiation (fig. 5A). Because evidence of neuronal differentiation follow ing the treatment of NB cells with LN or FN is obvious somewhat earlier than it is in cells treated with RA, we also examined the level of mRNA encoded by NMYC, an oncogene commonly amplified in NB tumor cells.
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
ECM and Neuroblastoma
Kidowaki/Thiele/KJcinman/Israel
320
C
The expression of NMYC is known to decrease very early in the course of RA therapy, even before morpho logic changes indicative of a neuronal phenotype can be recognized [48]. Nonetheless, no change in NMYC gene expression was observed by northern blot analysis (fig. 5B). Similarly, laminin receptor expression (fig. 5B), which also has been reported to be reduced following the treatment of tumor cells with another mediator of NB differentiation, sodium butyrate, was not affected [54].
Discussion RA induced differentiation of NB tumor cells is a well-studied in vitro system showing morphologic, genet ic, biochemical, and physiologic changes, indicating that these cells can mature along a ganglionic lineage [27, 28, 48, 50]. Because FN and LN have been implicated in the regulation of neuronal differentiation of cells derived
C
Anti-FN Antiserum
Fig. 4. Effect of antisera against extracellular matrix proteins on NB cells. SK-N-BE(2) cells were cultured in RPMI-1640 supple mented with 2% FCS in wells coated with matrix proteins or in standard tissue culture wells containing 1 \iM RA and treated with antisera (5 % v/v) or normal sera from the same species. Morpho logic effects were observed 1, 2, 4, and 6 days later. C - Control.
from the neural crest, we compared the morphologic changes induced by FN and LN with those observed fol lowing RA treatment of neuroblastoma cells [12-16]. These proteins, which exist in varying concentrations along the migration pathways of these cells throughout their development [8,21, 22], influence the expression of differentiated tissue characteristics in normal neurons [8] and in a number of malignant tissues [8, 48], Matrix proteins including LN and FN also contribute to the growth and maturation of neuronal cells in primary cul ture [12,15, 16], and induce extension of neurites in cells of murine and human nervous system tumors [30, 31], Because both FN and LN induce morphologic changes and enhance RA-induced neuronal differentiation, these proteins and the cellular pathways over which their rec ognition by the cell is mediated may be used to trigger the maturation of NB cells. Antiserum against LN or FN blocked the effect of its corresponding matrix. The spec ificity of these antisera is compatible with other lines of
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
Fig. 3. Adhesion of SK-N-BE(2) cells to matrix proteins. SK-NBE(2) cells, labeled overnight with 5 pCi/ml of (’HJ-thymidine in RPMI-1640 containing 10% FCS, were dispersed, washed and applied to wells of immunologic 96-well plates coated with the indi cated amounts of matrix proteins or poly-D-lysine (PLYS). The cells were incubated in RPMI-1640 with 5% FCS for 1 h at 37 °C. Nonadherent cells were removed by a pipette with medium, and radioac tivity of the adherent cells was counted. Results are shown as a mean ratio of radioactivity of adherent cells to that of total cells in the well. Bars indicate the ranges observed in three independent exper iments. C - Control.
Anti-LN Antiserum
321
ECM and Neuroblastoma
•
II
Ml
G AP 43 C FN LN
CG A
« •
•
NM YC
•
*
LN r
NF
M l
•
GAPDH GAPDH
Fig. 5. A Total RNAs were isolated from SK-N-BE(2) cells cul tured in bactériologie plates with or without matrix proteins. The expression of genes were studied by Northern blot analysis. B Blots were rehybridized to a cDNA probe of the constitutively expressed GADPH gene to provide an estimate of the actual amounts of RNA loaded in each lane.
investigation indicating that FN and LN, despite their indistinguishable effects on cell morphology, have dis tinct receptors through which their activity is mediated. The observation that neither antiserum blocks the RA induced morphological changes suggests that RA effects are not entirely mediated by these matrix proteins, a finding consistent with the obervation of others that RA treatment of neuroblastoma cells does not increase ma trix protein synthesis [48]. LN and FN did not alter the proliferative capacity of the NB cells nor the expression of genes associated with a mature neuronal or neuroendocrine phenotype. In em bryonal tissues, proliferative epithelia, and tumor cells, the association of growth arrest with differentiation sug gests that these two processes are closely linked. In some tissues, growth arrest invariably accompanies matura tion, and the continuous proliferation of some em bryonal malignancies may reflect their immaturity. The linkage in NB tumor cells of differentiation and growth arrest is not, however, invariable. The constitutive ex pression of N-myc can block the ability of RA to induce differentiation of these cells in vitro, but not its ability to
References 1 Le Douarin, N.M.: Migration and differentiation of neural crest cells. Curr. Top. Dev. Biol. 16: 31-85 (1980). 2 Le Douarin, N.M.; Xue, Z.G.; Smith, J.: In vivo and in vitro studies on the segregation of autonomic and sensory cell line ages. J. Physiol., Paris 80: 255-261 (1985). 3 Sanes, J.R.: Roles of extracellular matrix in neural development. A. Rev. Physiol. 45: 581-600 (1983).
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
C FN LN
induce growth arrest [51]. Matrix proteins in our study similarly affected differentiation and cell growth arrest independently by initiating morphologic differentiation without altering growth potential or proliferation. This finding raises the possibility that N-myc induced cellular alterations associated with the arrested differentiation of these cells may be in pathways that are antagonistic to those over which the effects of matrix proteins are me diated. Since neural crest cells are thought to migrate along pathways defined by specific matrix proteins, molecules that are critical for the regulation of cell proliferation might be expected to be located at the target sites to which these cells migrate. In this context, the inability of FN or LN to effect cellular proliferation is not unex pected, although in some migration pathways FN is pref erentially located at sites in which migration ends. These findings would suggest that additional environmental signals are necessary to mediate neural crest cell growth regulation at such target sites. While some migrating neural crest cells may be committed to differentiation along a specific lineage, the maturation of these cells clearly continues after the completion of cell migration and might therefore be expected to require molecular signals not found along the migratory pathway. To understand the regulation of lineage-specific gene expression, it is particularly important to identify extra cellular signals that can trigger the maturation and dif ferentiation of embryonal cells. These signals may be expected to have similar effects on tumor cells. In this regard, the modulation of tumor cell differentiation may ultimately be of therapeutic import for the management of embryonal pediatric tumors. For the study of extracel lular signals and cellular mechanisms that mediate tu mor cell differentiation, knowledge of the milieu in which this differentiation can efficaciously occur is of central importance. The ability of LN and FN to en hance RA-induced neuronal maturation of NB tumor cells provides in vitro evidence suggesting an interaction and relationship between these different environmental signals.
4 Girdlestone, J.; Weston, J.A.: Identification of early neuronal subpopulations in avian neural crest cell cultures. Devi Biol. 109: 274-287 (1985). 5 Le Douarin, N.M.: Cell line segregation during peripheral ner vous system ontogeny. Science 231: 1515-1522 (1986). 6 Thiery, J.P.; Duband, J.L.; Tucker, G.C.: Cell migration in the vertebrate embryo: role of cell adhesion and tissue environment in pattern formation. Annu. Rev. Cell Biol. 1: 91-113 (1985). 7 Le Douarin, N.M.: The neural crest, pp. 22-53 (Cambridge Uni versity Press, Cambridge 1982). 8 Perris, R.; Boxberg, Y. von; Lofberg, J.: Local embryonic ma trices determine region-specific phenotypes in neural crest cells. Science 241: 86-89 (1988). 9 Kleinman, H.K.; Graf, J.; Iwamoto, Y.; Kitten, G.T.; Ogle, R.C.; Sasaki, M.; Yamada, Y.; Martin, G.R.; Luckenbill-Edds, L.: Role of basement membranes in cell differentiation. Ann. N.Y. Acad. Sci. 513: 134-145 (1987). 10 Rutka, J.T.; Giblin, J.R.; Apodoca, G.; DeArmond, S.J.; Stem, R.; Rosenblum, M.L.: Inhibition of growth and induction of dif ferentiation in a malignant human glioma cell line by normal leptomeningeal extracellular matrix proteins. Cancer Res. 47: 3515-3522 (1987). 11 Akers, R.M.; Mosher, D.F.; Lilien, J.E.: Promotion of retinal neurite outgrowth by substratum-bound fibronectin. Devi Biol. 56: 179-188 (1981). 12 Bilozur, M.E.; Hay, E.D.: Neural crest migration in 3D extracel lular matrix utilizes laminin, fibronectin, or collagen. Devi Biol. 125: 19-33 (1988). 13 Gundersen, R.W.: Response of sensory neurites and growth cones to patterned substrata of laminin and fibronectin in vitro. Devi Biol. 121: 423-431 (1987). 14 Manthrope, M.; Engvall, E.; Ruoslahti, E.; Longo, F.M.; Davis, G.E.; Varon, S.: Laminin promotes neuritic regeneration from cultured peripheral and central neurons. J. Cell Biol. 97: 1882— 1890(1983). 15 Rogers, S.L.; Letoumeau, P.C.; Palm, S.L.; McCarthy, J.; Furcht, L.T.: Neurite extension by peripheral and central ner vous system neurons in response to substratum-bound fibronec tin and laminin. Devi Biol. 98: 212-220 (1983). 16 Acheson, A.; Edgar, D.; Timpl, R.; Thoenen, H.: Laminin increases both levels and activity of tyrosine hydroxylase in calf adrenal chromaffin cells. J. Cell Biol. 102: 151-159 (1986). 17 Loring, J.; Glimelius, B.; Weston, J.A.: Extracellular matrix materials influence quail neural crest cell differentiation in vitro. Devi Biol. 90: 165-174 (1982). 18 Sieber-Blum, M.; Sieber, F.; Yamada, K.M.: Cellular fibronectin promotes adrenergic differentiation of quail neural crest cells in vitro. Expl Cell Res. 133: 285-295 (1981). 19 Duband, J.L.; Thiery, J.P.: Distribution of fibronectin in the early phase of avian cephalic neural crest cell migration. Devi Biol. 93: 308-323 (1982). 20 Newgreen, D.; Thiery, J.-P.: Fibronectin in early avian embryos: synthesis and distribution along the migration pathways of neu ral crest cells. Cell Tiss. Res. 211: 269-291 (1980). 21 Rogers, S.L.; Edson, K.J.; Letoumeau, P.C.; McLoon, S.C.: Dis tribution of laminin in the developing peripheral nervous system of the chick. Devi Biol. 113: 429-435 (1986). 22 Bolande, R.P.: Neurofibromatosis - the quintessential neurocristopathy: pathogenetic concepts and relationships. Adv. Neu rol. 29: 67-75 (1981).
Kidowaki/Thiele/Kleinman/Israel
23 Hayes, F.A.; Smith, E.I.: Neuroblastoma; in Pizzo, Poplack, ‘Neuroblastoma’ in principles and practice of pediatric oncolo gy, pp. 607-622 (Lippincott, Philadelphia 1989). 24 Helman, L.J.; Thiele, C.J.; Linehan, W.M.; Israel, M.A.: A role for differentiation arrest in the development of neural crest tumors; in Lapis, Liotta, Rabson, Biochemistry and molecular genetics of cancer métastasés, pp. 73-82 (Nijhoff, Boston 1986). 25 Israel, M.A.: The evolution of clinical molecular genetics. Neu roblastoma as a model tumor. Am. J. Pediatr. Hematol. Oncol. 8: 163-172 (1986). 26 Sidell, N.: Retinoic acid-induced growth inhibition and morpho logic differentiation of human neuroblastoma cells in vitro. J. natn. Cancer Inst. 68: 589-593 (1982). 27 Sidell, N.; Altman, A.; Haussier, M.R.; Seeger, R.C.: Effects of retinoic acid (RA) on the growth and phenotypic expression of several human neuroblastoma cell lines. Expl Cell Res. 148: 2 130(1983). 28 Thaller, C ; Eichele, G.: Identification and spatial distribution of retinoids in the developing chick limb bud. Nature 327:625-628 (1987). 29 Domen, C.; Culp, L.A.: Adhesion sites of neural tumor cells. Morphogenesis of substratum-attached material. Expl Cell Res. 134: 329-338 (1981). 30 Jousimaa, J.; Merenmies, J.; Rauvala, H.: Neurite outgrowth of neuroblastoma cells induced by proteins covalently coupled to glass coverslips. Eur. J. Cell Biol. 35: 55-61 (1984). 31 Rauvala, H.: Neurite outgrowth of neuroblastoma cells: depen dence on adhesion surface - cell surface interactions. J. Cell Biol. 98: 1010-1016 (1984). 32 Cooper, A.R.; Taylor, A.; Hogan, B.L.M.: Changes in the rate of laminin and entactin synthesis in F9 embryonal carcinoma cells treated with retinoic acid and cyclic amp. Devi Biol. 99: 510516 (1983). 33 Hassell, J.R.; Pennypacker, J.P.; Kleinman, H.K.; Pratt, R.M.; Yamada, K.M.: Enhanced cellular fibronectin accumulation in chondrocytes treated with vitamin A. Cell 17: 821-826 (1979). 34 Baron-Van Evercooren, A.; Gansmuller, A.; Gumpel, M.; Bau mann, N.; Kleinman, H.K.: Schwann cell differentiation in vitro: extracellular matrix deposition and interaction. Dev. Neurosci. 8: 182-196 (1986). 35 Hopper, K.E.; Adelmann, B.C.; Gentner, G.; Gay, S.: Recogni tion by guinea-pig peritoneal exudate cells of conformationally different states of the collagen molecule. Immunology 30: 249259 (1976). 36 McGarvey, M.L.; Baron-Van Evercooren, A.; Kleinman, H.K.; Dubois-Dalcq, M.: Synthesis and effects of basement membrane components in cultured rat Schwann cells. Devi Biol. 105:18-28 (1984). 37 Timpl, R.; Rohde, H.; Robey, P.G.: Rennard, S.I.; Foidart, J.M.; Martin, G.P.: Laminin - a glycoprotein from basement membranes. J. biol. Chem. 254: 9933-9937 (1979). 38 Biedler, J.L.; Roffler-Tarlov, S.; Schachner, M.; Freedman, L.S.: Multiple neurotransmitter synthesis by human neuroblastoma cell lines and clones. Cancer Res. 38: 3751-3757 (1978). 39 Reynolds, C.P.; Reynolds, D.A.; Frenkel, E.P.; Smith, R.G.: Selective toxicity of 6-hydroxydopamine and ascorbate for hu man neuroblastoma in vitro: a model for clearing marrow prior to autologous transplant. Cancer Res. 42: 1331-1336 (1982). 40 Chirgwin, J.M.; Przybyla, A.E.; MacDonald, R.J.; Rutter, W.J.:
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
322
ECM and Neuroblastoma
42
43
44
45
46
47
48
49 Tsokos, M.; Scarpa, S.: Ross, R.A.; Triche, T.J.: Differentiation of human neuroblastoma recapitulates neural crest develop ment. Study of morphology, neurotransmitter enzymes, and extracellular matrix proteins. Am. J. Pathol. 128: 484-496 (1987). 50 Thiele, C.J.; Deutsch, L.A.; Israel, M.A.: The expression of mul tiple proto-oncogenes is differentially regulated during retinoic acid induced maturation of human neuroblastoma cell lines. Oncogene 3: 281-288 (1988). 51 Thiele, C.J.; Israel, M.A.: Regulation of N-myc expression is a critical event controlling the ability of human neuroblasts to dif ferentiate. Expl Cell Biol. 56: 321-333 (1988). 52 Bjelfman, C.; Meyerson, G.; Cartwright, C.; Mellstrom, K.; Hammerlin, V.; Pahlman, S.: Early activation of endogenous pp60src kinase activity during neuronal differentiation of cul tured human neuroblastoma cells. Mol. Cell Biol. 10: 361-370 (1990). 53 Helman, L.J.; Ahn, T.G.; Levine, M.A.; Allison, A.; Cohen, P.S.; Cooper, M.J.; Cohn, D.V.; Israel, M.A.: Molecular cloning and primary structure of human chromogranin A (secretory protein I) cDNA. J. biol. Chem. 263: 11559-11563 (1988). 54 Bryant, G.; Habcrem, C.; Rao, C.N.; Liotta, L.A.: Butyrate induced reduction of tumor cell laminin receptors. Cancer Res. 46: 807-811 (1986).
Received: August 22, 1990 Accepted: September 8, 1990 Mark A. Israel, MD Brain Tumor Research Center Department of Neurosurgery University of California Parnassus St. San Francisco, CA 94143-0520 (USA)
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
41
Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18: 5294-5299 (1979). Thiele, C.J.; McKeon, C.; Triche, T.J.; Ross, R.A.; Reynolds, C.P.; Israel, M.A.: Differential protooncogene expression char acterizes histopathologically indistinguishable tumors o f the pe ripheral nervous system. J. clin. Invest. 80: 804-811 (1987). Rigby, P.; Dieckmann, C.; Rhodes, C.; Berg, P.: Labeling deox yribonucleic acid to high specific activity in vitro by nick trans lation with DNA polymerase I. J. molec. Biol. 113: 237-251 (1977). Myers, M.W.; Lazzarini, R.A.; Lee, V.M.; Schlaepfer, W.W.; Nelson, D.L.: The human mid-size neurofilament subunit: a repeated protein sequence and the relationship of its gene to the intermediate filament gene family. EMBO J. 6: 1617-1626 (1987). Helman, L.J.; Ahn, T.G.; Levine, M.A.; Allison, A.; Cohen, P.S.; Cooper, M.J.; Cohn, D.V.; Israel, M.A.: Molecular cloning and primary structure of human chromogranin A (secretory protein I) cDNA. J. biol. Chem. 263: 11559-11563 (1988). Schwab, M.; Alitalo, K.; KJempnauer, K.; Varmus, H.E.; Bishop, J.M.; Gilbert, F.; Brodeur, G.; Goldstein, M.; Trent, J.: Ampli fied DNA with limited homology to myc cellular oncogene is shared by human neuroblastoma cell lines and a neuroblastoma tumour. Nature 305: 245-248 (1983). Fort, P.; Marty, L.; Piechaczyk, M.; El-Sabrouty, S.; Dani, C.; Jeanteur, P.; Blanchard, J.-M.: Various rat adult tissues express only one major mRNA species from the glyceraldehyde-3-phosphate-dehydrogenase multigenic family. Nucl. Acids Res. 13: 1431-1442 (1985). Kants, L.R.; Ng, S.; Freeman, J.A.; Fishman, M.C.: Cloning of complementary DNA for GAP-43, a neuronal growth-related protein. Science 236: 597-600 (1987). Thiele, C.J.; Reynolds, C.P.; Israel, M.A.: Decreased expression of N-myc precedes retinoic acid-induced morphological differ entiation of human neuroblastoma. Nature 313: 404-406 (1985).
323
Matrix Proteins Induce Neuroblastoma Cell Differentiation without Altering Cell Growth Takuro Kidowaki3-1, C.J. Thiele a , H.K. Kleinm anb, Mark A. Israela a Pediatric Branch, National Cancer Institute and the bNational Institute for Dental Research, Bethesda, Md., USA
Key Words. Extracellular matrix • Differentiation • Neuroblastoma • Retinoic acid • Growth regulation
Introduction During development, cells originating in the em bryonal neural crest migrate and proliferate giving rise to highly specialized tissues [1] including the neuronal and glial cells of the peripheral nervous system (PNS) [2, 3], Although the developmental fate of some cells may be established before migration from the neural crest [4, 5], substances in the local environment can affect their migratory behavior [6] as well as their proliferation and differentiation [2,4, 5]. The naturally occurring environ mental signals mediating these processes include hor monal factors [7] and components of the extracellular matrix [8], 1 Dr. Kidowaki was the recipient of a Nakasone Fellow at the National Cancer Institute.
Extracellular matrix proteins influence the differenti ation of several different cell types [8-11]. In vivo they can trigger the diversification of neural crest cells into the various lineages that give rise to PNS tissues [8]. Laminin (LN) and fibronectin (FN) are matrix proteins which, in vitro, promote neuritic outgrowth from cells of both the central and PNS [12-16]. These proteins also promote adrenergic differentiation of neural crest cells in vitro [17-19]. Although neither FN nor LN is known to be synthesized by neural crest cells [13], both are found in varying amounts along neural crest migration pathways where they may act as signals for cellular dif ferentiation and tissue morphogenesis [20-22]. Neuroblastoma (NB), a malignant tumor of very early childhood, is thought to originate in neural crest cells [23]. Although this tumor most commonly arises in the adrenal medulla, it also occurs in sympathetic ganglia
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
Abstract. The maturation of embryonal neural crest cells is thought to be regulated in part by the milieu into which these cells migrate. Neuroblastoma (NB) is a tumor of very early childhood that is thought to arise in asso ciation with the arrested differentiation of embryonal neural crest cells. In culture, neuroblastoma tumor cells differentiate in the presence of retinoic acid, which is also known to influence extracellular matrix protein synthesis. We have cultured neuroblastoma cells on laminin (LN) and fibronectin (FN) substrata to examine the role of extracellular matrix in retinoic acid (RA)-induced differentiation of these tumor cells. These proteins caused mor phologic changes in NB cells indistinguishable from those caused by RA. Antiserum to each of these proteins blocked the effects induced by the corresponding protein, but neither antiserum affected the action of RA. Despite the induction of a neuronal morphologic change, matrix proteins did not alter the proliferation of NB cells. These results indicate that LN and FN modulate the differentiation of NB cells without inducing growth arrest and that RAinduced differentiation does not require these matrix proteins.
and at sites without known PNS tissue [24]. Such sites may be the locations of embryonal neural crest cells that did not successfully complete their migration. The origin of NB in embryonic tissue and its capacity to differen tiate both in vitro and in vivo suggests that arrested dif ferentiation may have a role in its pathogenesis [23, 25, 26]. Retinoic acid (RA) has been the most extensively studied agent to induce differentiation of neuroblastoma cells [27, 28] and is the only known naturally occurring morphogen [29]. We investigated the possible regulatory role of matrix glycoproteins in the differentiation of NB tumor cells [30-32], We found that both FN and LN can initiate the morphologic differentiation of these tumor cells and can enhance the efficacy with which RA induces their in vitro maturation. While RA has been shown to increase the production of matrix proteins as much as 50-fold in embryonal carcinoma cells [33] and chondrocytes [34], we were not able to demonstrate a role for endogenously synthesized matrix proteins in RA-induced neuroblas toma tumor cell differentiation. Materials and Methods Materials Mouse LN, human FN, goat antimouse LN antiserum, and rab bit antihuman FN antiserum were prepared as previously report ed [35-38]. Poly-D-lysine and RA were purchased from Sigma (St. Louis, Mo.). (3H)-thymidine (52.1 Ci/mmol) and (32P)-dCTP (3,000 Ci/mmol) were obtained from New England Nuclear (Bos ton, Mass.). Cells Cells from the human NB cell lines SK-N-BE(2) [39] and SMSK.CNR [40] were maintained in RPMI-1640 supplemented with 10% fetal calf serum (FCS), glutamine, penicillin and streptomycin in 75-cm tissue culture flasks at 37 °C in 5% CO2. Assays o f Morphologic Change and Cell Growth Assays were performed in 96-well tissue culture plates from Cos tar (Philadelphia, Pa.) or immunologic 96-well plates coated with extracellular matrix proteins (10 pg/0.32 cm2) or poty-£>-lysine [16]. To coat plates with matrix proteins, an aqueous solution containing the designated amount of matrix protein was added to each cell, incubated at 37 °C for 1 h, and the supernate was removed. Loga rithmically growing cells in medium containing 10% FCS were har vested with EDTA and dispersed by gentle pipetting. They were then washed, suspended in a solution at 5 X 103 cells/0.2 ml of RPMI-1640 containing 5% FCS, seeded in plates, and cultured at 37 °C in 5% CO2 [16]. Morphologic changes in the cells were observed microscopically and scored after 1, 2, 4, and 6 days. Sim ilar experiments were also performed in the presence of 0.1 \rM alltrans RA (Sigma, St. Louis, Mo.). The effects of antisera to specific matrix proteins on the differentiation of NB cells was assessed by incubating NB cells in matrix protein-coated wells containing media
317
to which antisera against specific matrix proteins [35-37, 40] or a control antisera had been added to a final concentration of 5 %. Cell proliferation was measured by a (3H)-thymidine incorpora tion assay of triplicate cultures: 1 pCi of (3H)-thymidine in 10 pi of phosphate-buffered saline was added to the medium of each well and 20 h later the cells were harvested by a PHD cell harvester (Cambridge Scientific, Cambridge, Mass.). Radioactivity was mea sured in a liquid scintillation counter. Graphical Analysis To display and compare graphically changes in the morphology of NB tumor cells cultured under various conditions, we counted 200 cells in each well and devised a graph to represent the distribu tion of cell types. Typically neuroblastoma tumor cells are small, round, and loosely adherent. Fully differentiated neuroblastoma cells appear indistinguishable from neurons in primary culture with long slender neuritic outgrowths exhibiting varicosities and making contact with distantly located cells. We designated small round tumor cells as - , flat adherent tumor cells as +, cells with processes shorter than the cell body from which they extended as ++, and cells with processes that were longer that the cell bodies as +++. At points corresponding to the time after treatment when we scored these cells, we graphically displayed our results as a vertical line extending from the least differentiated to the most differentiated type seen. This line formed one side of a triangle and indicated the range of the cell shapes. A point corresponding to the cell type most frequently observed was plotted as the third point of the triangle. The kinetics of these changes were indicated by a graph connecting points corre sponding to the most common cell type seen at each evaluation. Adhesion Assay NB cells were labeled by culturing for 24 h in 5 pCi/ml of (3H)thymidine in RPMI-1640 containing 10% FCS. The cells were har vested with EDTA, washed, and plated in 96-well immunologic plates coated with matrix proteins or poly-D-lysine and incubated in RPMI-1640 with 5% FCS for 1 h at 37 °C. Nonadherent cells were removed by washing the cells with media. The remaining adherent cells were harvested with EDTA. Nonadherent and adherent cells were solubilized in Aquasol (New England Nuclear, Boston, Mass.) and their radioactivity measured by liquid scintillation counting. Ribonucleic Acid Isolation and Northern Blots NB cells were cultured in 150-cm2 plates coated with matrix proteins. Total cellular ribonucleic acid (RNA) was isolated from the cells by guanidine thiocyanate extraction and centrifugation over a cesium chloride cushion [41]. Twenty-microgram samples of each RNA preparation were size-fractionated by electrophoresis on a 1% agarose gel containing 2.2 M formaldehyde and were trans ferred to Nytran (Schleicher & Schuell, Keene, N.H.) for Northern blot analysis as previously described [42]. Radiolabeled DNA probes were synthesized by nick translation [43] from the following plasmids: pNF, which contains a 1.3-kb deoxyribonucleic acid (DNA) fragment of pHNFl [intermediate-sized neurofilament, 44]; pHCGA, which contains a cDNA encoding human chromogranin A [45]; pNB-1, which contains a cDNA of N-myc [46]; pLR4-4, which contains a cDNA encoding the LN receptor [40]; and pGAPDH, which contains a cDNA encoding glyceraldehyde-3-phosphate dehy drogenase [47], To evaluate expression of GAP43, we synthesized a 30-nucleotide oligomer corresponding to nucleotide 181-211 of GAP43 (47a) using the Applied Biosystem 381A DNA synthesizer
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
ECM and Neuroblastoma
Kidowaki/Thiele/Kleinman/Israel
318
0 Day
1 Day
2 Days
5 Days
1
A SMS-KCNR SK-N-BEI2) SK-N-BEI2I
B SMS-KCNR SK-N-BEI2) SK-N-BE12)
(Foster City, Calif.). The oligomers were radiolabeled with J2PdATP and terminal transferase. Blots were hybridized for 20 h at 42 °C in 50% formamide and 5X SSPE. The blots were washed twice for 15 min at 25 °C in 6X SSC, twice for 15 min at 37 °C in 2X SSC, and once for 60 min at 65 °C in 0. IX SSC. The mem branes were exposed to Kodak X-Omat AR film with intensifying screens at -7 0 °C.
Fig. 1. Induced differentiation of NB cells. SK-N-BE(2) cells were cul tured in RPMI-1640 supplemented with either 10% FCS and 5 \iM RA 0 day (a), 1 day (b). 2 days (c), and 5 days (d) after the start of culture or 5% FCS and LN (500 pg/150cm2 of well floor) 0 h (e). 6 h (0. 1 day (g). and 2 days (h) after the start of cul ture. The extent of morphologic change was defined as follows: - = round cells, + = flat cells, ++ = cells with processes shorter than cell bod ies, +++ = cells with processes longer than cell bodies. Fig. 2. Effects of extracellular ma trix proteins and RA on human NB cells. SMS-KCNR and SK-N-BE(2) cells were harvested with EDTA and seeded at 5 X I0-1 cells in 0.2 ml of RPMI-1640 containing 5% FCS in 96-well immunologic plates coated with poly-D-lysine (PLYS; 1 pg/0.32 cm2 of well floor), LN. and FN (10 gg/0.32 cm2 of well floor, each). RA was used in a concentration of 0. 1 \iM. A Morphology was observed microscopically 1, 2, 4, and 6 days after the initiation of culture. The ex tent of morphologic changes was shown as four stages defined in figure 1. For each triangle in the figure, height and position represent the range of morphologic types in the cul ture, and the position of the left comer indicates the most common cell type seen. B Cell growth was de termined by a (3H)-thymidine incor poration assay at the indicated time. Values are presented as a mean value of counts per minute. C = Control.
Results
The NB tumor cell line SK-N-BE(2) undergoes se quential morphologic changes in the presence of 5 pAf RA (fig. la-d). The small round cells typical of NB tumor cell lines initially adhere to the cell culture sub
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
Time
strate and then flatten into an epithelioid shape. This change in shape is followed by the sprouting of short processes that can grow to several times the length of the cell body and can associate with other neuritic processes in a manner suggesting the formation of nerve fiber bun dles, after several days in culture [27], Cells from the NB cell line SMS-KCNR undergo similar changes, extending very long neurites [48], and other NB cell lines also change comparably when treated with RA [27, 48]. Ultrastructural and biochemical changes indicative of neuronal maturation also occur following the treatment of NB tumor cell lines with RA [27, 28, 49], When plated in the same media in the absence of RA, however, these cells typically remain small and round growing in loosely adherent clumps, which sometimes associate with rare flatter, more adherent cells on the culture vessel sur face. FN and LN when tested individually induce attach ment, flattening, and neurite outgrowth of SK-N-BE(2) cells indistinguishable from the changes that occurred after RA treatment. The changes observed following FN treatment [data not shown] were indistinguishable from those seen after LN (fig. le-h). In the case of both LN and FN the morphologic changes occurred more rapidly than those following RA treatment. To investigate fur ther the time course of these changes and to compare the effects of RA and matrix proteins quantitatively, SK-NBE(2) and SMS-KCNR cells [48, 50] were evaluated in the presence of either RA or matrix proteins or both (fig. 2). Within 2 days of the time NB cells were plated on either LN or FN, most cells were adherent and had flat tened (fig. 2A). This contrasted with the considerably slower time course over which RA induced these changes. Also of note was the more rapid time course over which morphologic changes suggestive of neuritic maturation could be observed in cultures to which either LN or FN was added. Interestingly, a greater proportion of cells undergo more extensive maturational morpho logic changes when cultures that had been grown on a surface coated with matrix proteins are treated with RA. Treatment with FN or with LN caused changes in SMSKCNR cells indistinguishable from those observed in SK-N-BE(2) cells. Although all cells in these cultures do not undergo morphologic changes synchronously, most have done so after several days of treatment. The in vitro maturation of NB cells is typically asso ciated with diminished growth potential [48] and the continuous proliferation of cells that do not become fully mature may be a key event in the development of embryonal malignancies. We therefore sought to deter
319
mine whether changes in tumor cell proliferation accom panied the maturation of NB cells that occurred in the presence of matrix proteins. In these experiments, we monitored the effects of treatment on the growth of KCNR and SK-N-BE(2) by assaying ,H-thymidine in corporation (fig. 2B) or cell number [data not shown]. In agreement with previous experiments from our lab [48, unpubl. data], we found that RA treatment inhibited the growth of both neuroblastoma cell lines examined, whether in the presence of extracellular matrix proteins or not (fig. 2B). In sharp contrast to this finding was the observation that there was no detectable effect by either LN or FN on proliferation even at late times, e.g. 6 days, when morphologic changes indicative or differentiation had been easily observable for several days (fig. 2B). We examined the effect of poIy-£>-lysine, a highly anionic substance which enhances the attachment of most tissue culture cells to plastic surfaces, to determine whether simply the adherence of neuroblastoma cells to plastic surfaces could induce neurite outgrowth. Poly£>-lysine caused cells to adhere to the growth surface (fig. 3), but did not cause them to change morphologi cally (fig. 2), suggesting that adhesion itself does not bring about subsequent changes in shape. We utilized antisera specifically reactive against either LN or FN to block the effects of these matrix pro teins and to examine whether the effect of RA was likely to be mediated by the production of endogenous extra cellular matrix protein synthesis. Each antiserum blocked the action of its respective matrix protein, but had no effect on the action of the other protein (fig. 4). Neither antiserum inhibited the action of RA. Previous studies have shown that RA- [48. 50, 51] and phorbol esters [52] induced morphologic differenti ation of some neuroblastoma tumor cell lines is accom panied by changes in gene expression. These include increases in mRNA levels of several genes whose expres sion is closely associated with the neuronal phenotype, such as neurofilament [50], and GAP43 [C.J.T. unpubl. data] and decreases in the steady state level of mRNA encoding chromogranin, a highly expressed gene in neu ral crest-derived chromaffin cells [53], In NB cells treated with FN or LN, however, these changes do not occur in association with morphologic differentiation (fig. 5A). Because evidence of neuronal differentiation follow ing the treatment of NB cells with LN or FN is obvious somewhat earlier than it is in cells treated with RA, we also examined the level of mRNA encoded by NMYC, an oncogene commonly amplified in NB tumor cells.
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
ECM and Neuroblastoma
Kidowaki/Thiele/KJcinman/Israel
320
C
The expression of NMYC is known to decrease very early in the course of RA therapy, even before morpho logic changes indicative of a neuronal phenotype can be recognized [48]. Nonetheless, no change in NMYC gene expression was observed by northern blot analysis (fig. 5B). Similarly, laminin receptor expression (fig. 5B), which also has been reported to be reduced following the treatment of tumor cells with another mediator of NB differentiation, sodium butyrate, was not affected [54].
Discussion RA induced differentiation of NB tumor cells is a well-studied in vitro system showing morphologic, genet ic, biochemical, and physiologic changes, indicating that these cells can mature along a ganglionic lineage [27, 28, 48, 50]. Because FN and LN have been implicated in the regulation of neuronal differentiation of cells derived
C
Anti-FN Antiserum
Fig. 4. Effect of antisera against extracellular matrix proteins on NB cells. SK-N-BE(2) cells were cultured in RPMI-1640 supple mented with 2% FCS in wells coated with matrix proteins or in standard tissue culture wells containing 1 \iM RA and treated with antisera (5 % v/v) or normal sera from the same species. Morpho logic effects were observed 1, 2, 4, and 6 days later. C - Control.
from the neural crest, we compared the morphologic changes induced by FN and LN with those observed fol lowing RA treatment of neuroblastoma cells [12-16]. These proteins, which exist in varying concentrations along the migration pathways of these cells throughout their development [8,21, 22], influence the expression of differentiated tissue characteristics in normal neurons [8] and in a number of malignant tissues [8, 48], Matrix proteins including LN and FN also contribute to the growth and maturation of neuronal cells in primary cul ture [12,15, 16], and induce extension of neurites in cells of murine and human nervous system tumors [30, 31], Because both FN and LN induce morphologic changes and enhance RA-induced neuronal differentiation, these proteins and the cellular pathways over which their rec ognition by the cell is mediated may be used to trigger the maturation of NB cells. Antiserum against LN or FN blocked the effect of its corresponding matrix. The spec ificity of these antisera is compatible with other lines of
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
Fig. 3. Adhesion of SK-N-BE(2) cells to matrix proteins. SK-NBE(2) cells, labeled overnight with 5 pCi/ml of (’HJ-thymidine in RPMI-1640 containing 10% FCS, were dispersed, washed and applied to wells of immunologic 96-well plates coated with the indi cated amounts of matrix proteins or poly-D-lysine (PLYS). The cells were incubated in RPMI-1640 with 5% FCS for 1 h at 37 °C. Nonadherent cells were removed by a pipette with medium, and radioac tivity of the adherent cells was counted. Results are shown as a mean ratio of radioactivity of adherent cells to that of total cells in the well. Bars indicate the ranges observed in three independent exper iments. C - Control.
Anti-LN Antiserum
321
ECM and Neuroblastoma
•
II
Ml
G AP 43 C FN LN
CG A
« •
•
NM YC
•
*
LN r
NF
M l
•
GAPDH GAPDH
Fig. 5. A Total RNAs were isolated from SK-N-BE(2) cells cul tured in bactériologie plates with or without matrix proteins. The expression of genes were studied by Northern blot analysis. B Blots were rehybridized to a cDNA probe of the constitutively expressed GADPH gene to provide an estimate of the actual amounts of RNA loaded in each lane.
investigation indicating that FN and LN, despite their indistinguishable effects on cell morphology, have dis tinct receptors through which their activity is mediated. The observation that neither antiserum blocks the RA induced morphological changes suggests that RA effects are not entirely mediated by these matrix proteins, a finding consistent with the obervation of others that RA treatment of neuroblastoma cells does not increase ma trix protein synthesis [48]. LN and FN did not alter the proliferative capacity of the NB cells nor the expression of genes associated with a mature neuronal or neuroendocrine phenotype. In em bryonal tissues, proliferative epithelia, and tumor cells, the association of growth arrest with differentiation sug gests that these two processes are closely linked. In some tissues, growth arrest invariably accompanies matura tion, and the continuous proliferation of some em bryonal malignancies may reflect their immaturity. The linkage in NB tumor cells of differentiation and growth arrest is not, however, invariable. The constitutive ex pression of N-myc can block the ability of RA to induce differentiation of these cells in vitro, but not its ability to
References 1 Le Douarin, N.M.: Migration and differentiation of neural crest cells. Curr. Top. Dev. Biol. 16: 31-85 (1980). 2 Le Douarin, N.M.; Xue, Z.G.; Smith, J.: In vivo and in vitro studies on the segregation of autonomic and sensory cell line ages. J. Physiol., Paris 80: 255-261 (1985). 3 Sanes, J.R.: Roles of extracellular matrix in neural development. A. Rev. Physiol. 45: 581-600 (1983).
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
C FN LN
induce growth arrest [51]. Matrix proteins in our study similarly affected differentiation and cell growth arrest independently by initiating morphologic differentiation without altering growth potential or proliferation. This finding raises the possibility that N-myc induced cellular alterations associated with the arrested differentiation of these cells may be in pathways that are antagonistic to those over which the effects of matrix proteins are me diated. Since neural crest cells are thought to migrate along pathways defined by specific matrix proteins, molecules that are critical for the regulation of cell proliferation might be expected to be located at the target sites to which these cells migrate. In this context, the inability of FN or LN to effect cellular proliferation is not unex pected, although in some migration pathways FN is pref erentially located at sites in which migration ends. These findings would suggest that additional environmental signals are necessary to mediate neural crest cell growth regulation at such target sites. While some migrating neural crest cells may be committed to differentiation along a specific lineage, the maturation of these cells clearly continues after the completion of cell migration and might therefore be expected to require molecular signals not found along the migratory pathway. To understand the regulation of lineage-specific gene expression, it is particularly important to identify extra cellular signals that can trigger the maturation and dif ferentiation of embryonal cells. These signals may be expected to have similar effects on tumor cells. In this regard, the modulation of tumor cell differentiation may ultimately be of therapeutic import for the management of embryonal pediatric tumors. For the study of extracel lular signals and cellular mechanisms that mediate tu mor cell differentiation, knowledge of the milieu in which this differentiation can efficaciously occur is of central importance. The ability of LN and FN to en hance RA-induced neuronal maturation of NB tumor cells provides in vitro evidence suggesting an interaction and relationship between these different environmental signals.
4 Girdlestone, J.; Weston, J.A.: Identification of early neuronal subpopulations in avian neural crest cell cultures. Devi Biol. 109: 274-287 (1985). 5 Le Douarin, N.M.: Cell line segregation during peripheral ner vous system ontogeny. Science 231: 1515-1522 (1986). 6 Thiery, J.P.; Duband, J.L.; Tucker, G.C.: Cell migration in the vertebrate embryo: role of cell adhesion and tissue environment in pattern formation. Annu. Rev. Cell Biol. 1: 91-113 (1985). 7 Le Douarin, N.M.: The neural crest, pp. 22-53 (Cambridge Uni versity Press, Cambridge 1982). 8 Perris, R.; Boxberg, Y. von; Lofberg, J.: Local embryonic ma trices determine region-specific phenotypes in neural crest cells. Science 241: 86-89 (1988). 9 Kleinman, H.K.; Graf, J.; Iwamoto, Y.; Kitten, G.T.; Ogle, R.C.; Sasaki, M.; Yamada, Y.; Martin, G.R.; Luckenbill-Edds, L.: Role of basement membranes in cell differentiation. Ann. N.Y. Acad. Sci. 513: 134-145 (1987). 10 Rutka, J.T.; Giblin, J.R.; Apodoca, G.; DeArmond, S.J.; Stem, R.; Rosenblum, M.L.: Inhibition of growth and induction of dif ferentiation in a malignant human glioma cell line by normal leptomeningeal extracellular matrix proteins. Cancer Res. 47: 3515-3522 (1987). 11 Akers, R.M.; Mosher, D.F.; Lilien, J.E.: Promotion of retinal neurite outgrowth by substratum-bound fibronectin. Devi Biol. 56: 179-188 (1981). 12 Bilozur, M.E.; Hay, E.D.: Neural crest migration in 3D extracel lular matrix utilizes laminin, fibronectin, or collagen. Devi Biol. 125: 19-33 (1988). 13 Gundersen, R.W.: Response of sensory neurites and growth cones to patterned substrata of laminin and fibronectin in vitro. Devi Biol. 121: 423-431 (1987). 14 Manthrope, M.; Engvall, E.; Ruoslahti, E.; Longo, F.M.; Davis, G.E.; Varon, S.: Laminin promotes neuritic regeneration from cultured peripheral and central neurons. J. Cell Biol. 97: 1882— 1890(1983). 15 Rogers, S.L.; Letoumeau, P.C.; Palm, S.L.; McCarthy, J.; Furcht, L.T.: Neurite extension by peripheral and central ner vous system neurons in response to substratum-bound fibronec tin and laminin. Devi Biol. 98: 212-220 (1983). 16 Acheson, A.; Edgar, D.; Timpl, R.; Thoenen, H.: Laminin increases both levels and activity of tyrosine hydroxylase in calf adrenal chromaffin cells. J. Cell Biol. 102: 151-159 (1986). 17 Loring, J.; Glimelius, B.; Weston, J.A.: Extracellular matrix materials influence quail neural crest cell differentiation in vitro. Devi Biol. 90: 165-174 (1982). 18 Sieber-Blum, M.; Sieber, F.; Yamada, K.M.: Cellular fibronectin promotes adrenergic differentiation of quail neural crest cells in vitro. Expl Cell Res. 133: 285-295 (1981). 19 Duband, J.L.; Thiery, J.P.: Distribution of fibronectin in the early phase of avian cephalic neural crest cell migration. Devi Biol. 93: 308-323 (1982). 20 Newgreen, D.; Thiery, J.-P.: Fibronectin in early avian embryos: synthesis and distribution along the migration pathways of neu ral crest cells. Cell Tiss. Res. 211: 269-291 (1980). 21 Rogers, S.L.; Edson, K.J.; Letoumeau, P.C.; McLoon, S.C.: Dis tribution of laminin in the developing peripheral nervous system of the chick. Devi Biol. 113: 429-435 (1986). 22 Bolande, R.P.: Neurofibromatosis - the quintessential neurocristopathy: pathogenetic concepts and relationships. Adv. Neu rol. 29: 67-75 (1981).
Kidowaki/Thiele/Kleinman/Israel
23 Hayes, F.A.; Smith, E.I.: Neuroblastoma; in Pizzo, Poplack, ‘Neuroblastoma’ in principles and practice of pediatric oncolo gy, pp. 607-622 (Lippincott, Philadelphia 1989). 24 Helman, L.J.; Thiele, C.J.; Linehan, W.M.; Israel, M.A.: A role for differentiation arrest in the development of neural crest tumors; in Lapis, Liotta, Rabson, Biochemistry and molecular genetics of cancer métastasés, pp. 73-82 (Nijhoff, Boston 1986). 25 Israel, M.A.: The evolution of clinical molecular genetics. Neu roblastoma as a model tumor. Am. J. Pediatr. Hematol. Oncol. 8: 163-172 (1986). 26 Sidell, N.: Retinoic acid-induced growth inhibition and morpho logic differentiation of human neuroblastoma cells in vitro. J. natn. Cancer Inst. 68: 589-593 (1982). 27 Sidell, N.; Altman, A.; Haussier, M.R.; Seeger, R.C.: Effects of retinoic acid (RA) on the growth and phenotypic expression of several human neuroblastoma cell lines. Expl Cell Res. 148: 2 130(1983). 28 Thaller, C ; Eichele, G.: Identification and spatial distribution of retinoids in the developing chick limb bud. Nature 327:625-628 (1987). 29 Domen, C.; Culp, L.A.: Adhesion sites of neural tumor cells. Morphogenesis of substratum-attached material. Expl Cell Res. 134: 329-338 (1981). 30 Jousimaa, J.; Merenmies, J.; Rauvala, H.: Neurite outgrowth of neuroblastoma cells induced by proteins covalently coupled to glass coverslips. Eur. J. Cell Biol. 35: 55-61 (1984). 31 Rauvala, H.: Neurite outgrowth of neuroblastoma cells: depen dence on adhesion surface - cell surface interactions. J. Cell Biol. 98: 1010-1016 (1984). 32 Cooper, A.R.; Taylor, A.; Hogan, B.L.M.: Changes in the rate of laminin and entactin synthesis in F9 embryonal carcinoma cells treated with retinoic acid and cyclic amp. Devi Biol. 99: 510516 (1983). 33 Hassell, J.R.; Pennypacker, J.P.; Kleinman, H.K.; Pratt, R.M.; Yamada, K.M.: Enhanced cellular fibronectin accumulation in chondrocytes treated with vitamin A. Cell 17: 821-826 (1979). 34 Baron-Van Evercooren, A.; Gansmuller, A.; Gumpel, M.; Bau mann, N.; Kleinman, H.K.: Schwann cell differentiation in vitro: extracellular matrix deposition and interaction. Dev. Neurosci. 8: 182-196 (1986). 35 Hopper, K.E.; Adelmann, B.C.; Gentner, G.; Gay, S.: Recogni tion by guinea-pig peritoneal exudate cells of conformationally different states of the collagen molecule. Immunology 30: 249259 (1976). 36 McGarvey, M.L.; Baron-Van Evercooren, A.; Kleinman, H.K.; Dubois-Dalcq, M.: Synthesis and effects of basement membrane components in cultured rat Schwann cells. Devi Biol. 105:18-28 (1984). 37 Timpl, R.; Rohde, H.; Robey, P.G.: Rennard, S.I.; Foidart, J.M.; Martin, G.P.: Laminin - a glycoprotein from basement membranes. J. biol. Chem. 254: 9933-9937 (1979). 38 Biedler, J.L.; Roffler-Tarlov, S.; Schachner, M.; Freedman, L.S.: Multiple neurotransmitter synthesis by human neuroblastoma cell lines and clones. Cancer Res. 38: 3751-3757 (1978). 39 Reynolds, C.P.; Reynolds, D.A.; Frenkel, E.P.; Smith, R.G.: Selective toxicity of 6-hydroxydopamine and ascorbate for hu man neuroblastoma in vitro: a model for clearing marrow prior to autologous transplant. Cancer Res. 42: 1331-1336 (1982). 40 Chirgwin, J.M.; Przybyla, A.E.; MacDonald, R.J.; Rutter, W.J.:
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
322
ECM and Neuroblastoma
42
43
44
45
46
47
48
49 Tsokos, M.; Scarpa, S.: Ross, R.A.; Triche, T.J.: Differentiation of human neuroblastoma recapitulates neural crest develop ment. Study of morphology, neurotransmitter enzymes, and extracellular matrix proteins. Am. J. Pathol. 128: 484-496 (1987). 50 Thiele, C.J.; Deutsch, L.A.; Israel, M.A.: The expression of mul tiple proto-oncogenes is differentially regulated during retinoic acid induced maturation of human neuroblastoma cell lines. Oncogene 3: 281-288 (1988). 51 Thiele, C.J.; Israel, M.A.: Regulation of N-myc expression is a critical event controlling the ability of human neuroblasts to dif ferentiate. Expl Cell Biol. 56: 321-333 (1988). 52 Bjelfman, C.; Meyerson, G.; Cartwright, C.; Mellstrom, K.; Hammerlin, V.; Pahlman, S.: Early activation of endogenous pp60src kinase activity during neuronal differentiation of cul tured human neuroblastoma cells. Mol. Cell Biol. 10: 361-370 (1990). 53 Helman, L.J.; Ahn, T.G.; Levine, M.A.; Allison, A.; Cohen, P.S.; Cooper, M.J.; Cohn, D.V.; Israel, M.A.: Molecular cloning and primary structure of human chromogranin A (secretory protein I) cDNA. J. biol. Chem. 263: 11559-11563 (1988). 54 Bryant, G.; Habcrem, C.; Rao, C.N.; Liotta, L.A.: Butyrate induced reduction of tumor cell laminin receptors. Cancer Res. 46: 807-811 (1986).
Received: August 22, 1990 Accepted: September 8, 1990 Mark A. Israel, MD Brain Tumor Research Center Department of Neurosurgery University of California Parnassus St. San Francisco, CA 94143-0520 (USA)
Downloaded by: Access provided by the University of Michigan Library 141.216.78.40 - 1/11/2019 10:37:48 AM
41
Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18: 5294-5299 (1979). Thiele, C.J.; McKeon, C.; Triche, T.J.; Ross, R.A.; Reynolds, C.P.; Israel, M.A.: Differential protooncogene expression char acterizes histopathologically indistinguishable tumors o f the pe ripheral nervous system. J. clin. Invest. 80: 804-811 (1987). Rigby, P.; Dieckmann, C.; Rhodes, C.; Berg, P.: Labeling deox yribonucleic acid to high specific activity in vitro by nick trans lation with DNA polymerase I. J. molec. Biol. 113: 237-251 (1977). Myers, M.W.; Lazzarini, R.A.; Lee, V.M.; Schlaepfer, W.W.; Nelson, D.L.: The human mid-size neurofilament subunit: a repeated protein sequence and the relationship of its gene to the intermediate filament gene family. EMBO J. 6: 1617-1626 (1987). Helman, L.J.; Ahn, T.G.; Levine, M.A.; Allison, A.; Cohen, P.S.; Cooper, M.J.; Cohn, D.V.; Israel, M.A.: Molecular cloning and primary structure of human chromogranin A (secretory protein I) cDNA. J. biol. Chem. 263: 11559-11563 (1988). Schwab, M.; Alitalo, K.; KJempnauer, K.; Varmus, H.E.; Bishop, J.M.; Gilbert, F.; Brodeur, G.; Goldstein, M.; Trent, J.: Ampli fied DNA with limited homology to myc cellular oncogene is shared by human neuroblastoma cell lines and a neuroblastoma tumour. Nature 305: 245-248 (1983). Fort, P.; Marty, L.; Piechaczyk, M.; El-Sabrouty, S.; Dani, C.; Jeanteur, P.; Blanchard, J.-M.: Various rat adult tissues express only one major mRNA species from the glyceraldehyde-3-phosphate-dehydrogenase multigenic family. Nucl. Acids Res. 13: 1431-1442 (1985). Kants, L.R.; Ng, S.; Freeman, J.A.; Fishman, M.C.: Cloning of complementary DNA for GAP-43, a neuronal growth-related protein. Science 236: 597-600 (1987). Thiele, C.J.; Reynolds, C.P.; Israel, M.A.: Decreased expression of N-myc precedes retinoic acid-induced morphological differ entiation of human neuroblastoma. Nature 313: 404-406 (1985).
323
