Proc. Nati. Acad. Sci. USAVol. 87, pp. 1922-1926, March 1990Neurobiology

Serotonin uptake, storage, and synthesis in an immortalizedcommitted cell line derived from mouse teratocarcinoma

(adenovirus ElA promoter/simian virus 40 large tumor antigen/differentiation/neuroectoderm)

M. H. BUC-CARONt*, J. M. LAUNAY§, D. LAMBLINt, AND 0. KELLERMANNttLaboratoire de Diffdrenciation Cellulaire de I'Institut Pasteur, Unite Associde Centre National de la Recherche Scientifique 1148, 25, Rue du Dr Roux, 75724Paris Cedex 15, France; and Formation de Recherche Associde Claude Bernard, Neurochimie des Communications Cellulaires, HOpital Saint Louis,1 Avenue C. Vellefaux, 75475 Paris Cedex 10, France

Communicated by Andre Lwoff, December 11, 1989 (received for review August 31, 1989)

ABSTRACT We report the isolation and characterizationof a serotoninergic cell line, iC1l, derived from a mouseteratocarcinoma. The clone iCli was immortalized throughthe expression of the simian virus 40 oncogenes. iCli presentstwo states: an immature epithelial-like state (1Cli precursor)and a more differentiated state (1C01*). After induction bydibutyryl cyclic AMP and cyclohexanecarboxylic acid, almost100% of 1C11* cells continue to divide and have acquired aneural-like phenotype. 1Cl1* cells coexpress several neuralmarkers, such as synaptophysin (the membrane constituent ofsynaptic vesicles), the neuropeptide [Met5Jenkephalin, and theneurotransmitter serotonin. iCl1* cells store endogenous ser-otonin and are able to synthesize serotonin from L-tryptophanand to catabolize it by monoamine oxidase B. Moreover, thecells take up serotonin by a carrier-mediated mechanism verysimilar to that of serotoninergic neurons. The expression of thesimian virus 40 oncogenes, which promoted immortalization,does not therefore prevent further differentiation. This induc-ible cell line constitutes a valuable model for cellular andmolecular studies concerning the physiology and the pharma-cological modulation of the serotoninergic phenotype.

Study of the properties of neuronal or neuroendocrine celltypes is rendered extremely difficult by the failure to isolatethem from complex organs as pure, homogeneous popula-tions of stable cells. An alternative approach is the use ofimmortalized cell lines to obtain a preliminary characteriza-tion and to prepare various kinds of probes. The probes canthen be used to study normal cells in situ. We report here thefirst step in establishing such an experimental system forserotoninergic neurons: the isolation and characterization ofa homogeneous clonal cell line that, upon induction, ex-presses a serotoninergic phenotype.

Adrenergic or cholinergic cell lines (1-4) exhibiting theproperties of differentiated sympathetic neurons (5-9) or ofneuroendocrine adrenal chromaffin cells (e.g., rat pheochro-mocytoma line PC12) (10, 11) have been isolated. Otherneural-like cells, but no clonal cell lines, expressing adren-ergic or cholinergic markers have been obtained from tera-tocarcinomas (12-17).

Several models for serotoninergic neurons have been char-acterized. However, none of them exhibits a complete sero-toninergic phenotype [i.e., active uptake, storage, and me-tabolism of serotonin (5-hydroxytryptamine, 5-HT)]. Bloodplatelets, widely used as a neuronal model to study the uptakeand storage of 5-HT (18, 19), do not synthesize 5-HT, andgenetic studies cannot be performed with such anucleatedcells. Other in vitro models to study 5-HT uptake include aneuroblastoma-glioma hybrid cell line (20) and a simian virus

40 (SV40)-transformed mouse hypothalamic cell line (21)neither of which synthesize or catabolize 5-HT.Other interesting models are tumors of the neuroendocrine

system. Carcinoid tumors originating from enterochromaffincells in the gut maintain a complete serotoninergic pheno-type, but clonal cell lines have not been derived (22). A cellline established from a human medullary thyroid carcinoma,MTC (23), was recently characterized (24). MTC cells, whichoriginate from parafollicular cells of the thyroid, a neural-crest derivative, maintain properties of the original cells,namely, the expression of calcitonin, calcitonin gene-relatedpeptide, and somatostatin. The cells show many features ofserotoninergic cells. However, the presence of active 5-HTuptake has not been formally demonstrated. In addition,some heterogeneity in the cell population persists even afterrepeated subcloning.We have recently developed a clonal selection method that

makes it possible to immortalize teratocarcinoma-derivedcells that have the properties of committed cells capable offurther differentiation along a restricted lineage. The methodis based upon the transformation of embryonal carcinomacells by a recombinant plasmid, pK4, that carries the earlyregion of SV40 under the control of the adenovirus 5 ElApromoter. This promoter is active in early embryonic cells(25). The cell line F9K4b2 was obtained after transfection bypK4 of the embryonal carcinoma F9 (26). After in vitrodifferentiation of F9K4b2 was induced with retinoic acid anddibutyryl cyclicAMP (Bt2cAMP), several clones correspond-ing to immortalized precursors of the neuroectodermal, me-sodermal, and endodermal lineages were selected (27).One of the clones thus obtained, iC11, has features of a

committed precursor cell of the neuronal or neuroendocrinelineage. By addition of Bt2cAMP and cyclohexanecarboxylicacid, a switch between the initial committed cells (iC11) andthe more differentiated cells (1C11*) can be induced in closeto 100% of the population. 1C11* cells have acquired neuron-associated markers and a serotoninergic phenotype-namely, the ability to take up, store, synthesize, and catab-olize 5-HT.

MATERIALS AND METHODSMaterials. Bt2cAMP, dimethyl sulfoxide, and all-

trans-retinoic acid were from Sigma. Cyclohexanecarboxylicacid was from Sanofi (Paris).

[2-14C]- and [1,2-3H(N)]5-HT binoxalate (56.7 mCi/mmoland 26.1 Ci/mmol; 1 Ci = 37 GBq), L-[2,3-3H]tryptophan (55Ci/mmol), and 8-[ethyl-1-14C]phenylethylamine (50 mCi/

Abbreviations: 5-HT, 5-hydroxytryptamine (serotonin); SV40, sim-ian virus 40; Bt2cAMP, dibutyryl cyclic AMP; T antigen, tumorantigen; L-CAM, liver cell adhesion molecule; N-CAM, neural celladhesion molecule; MAO, monoamine oxidase.tTo whom reprint requests should be addressed.

1922

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

Proc. Natl. Acad. Sci. USA 87 (1990) 1923

mmol) were from NEN. [2,5,6-3H]Dopamine (10 Ci/mmol)was from Amersham.Clomipramine, paroxetine, and Ro 4-1284 were generous

gifts from M. Da Prada (Hoffmann-La Roche, Basel).Cell Culture. Cells were grown as described (25). Subclon-

ing of 1C11 was by limiting dilution. To induce 1C11 differ-entiation, cells were seeded in tissue-culture dishes (2 x 103cells per cm2) in the presence of 1 mM Bt2cAMP and 0.05%cyclohexanecarboxylic acid. Experiments were performedon day 4 (2.3 x 104 cells per cm2). When needed retinoic acidwas used at 0.3,.M.Tumor Formation. Injection of cells into syngeneic mice

and processing of the tumors were as described (27).Immunocytochemistry. Indirect immunofluorescence stud-

ies on cells were carried out using the following monoclonalantibodies and antisera. Mouse monoclonal antibodies:Pabl9 (undiluted), detecting viral tumor (T) antigens (28);anti-neurofilament triplet (diluted 1:10) (Immunotech, Lu-miny, France); anti-5-HT transporter system (1:5) (a gift of C.Chopin-Robert, H6pital Saint Louis, Paris); SY38 (1:100)anti-synaptophysin, a gift of B. Wiedenmann (29, 30). Ratmonoclonal antibodies: TROMA 1 (undiluted), recognizingthe cytokeratin Endo A (31); YC5/45 (1:50) anti-5-HT, a giftof C. Milstein (32). Rabbit antisera: anti-tyrosine hydroxylase(1:50), a gift of J. Thibault (33); anti-[Met5]- and anti-[Leu5]-enkephalin (1:50), gifts of F. Cesselin (34); anti-liver celladhesion molecule (L-CAM) (1:100) (35); anti-neural celladhesion molecule (N-CAM) (1:50), a gift of J. R. Sanes (36);anti-type IV collagen (1:20), a gift of J. A. Grimaud (37);affinity-purified anti-a-fetoprotein (38); anti-fibronectin (1:20)(Cappel Laboratories).

Affinity-purified, species-specific secondary antibodiescoupled to rhodamine (1:100; Nordic, Tilburg, The Nether-lands) were used to visualize sites of primary antibodybinding. Negative controls with omission of the primaryantibody were performed in all instances.

Determination of Cellular Content of Bioactive Amines andNeuropeptides. Cells were completely lysed by 250 hemolyticunits of the purified sulfhydryl-dependent toxin alveolysin (agenerous gift of J. E. Alouf, Institut Pasteur, Paris).

Unconjugated 5-HT (39), histamine (40), and catechola-mines (dopamine, norepinephrine, and epinephrine) (41)were measured radioenzymatically. Peptide and protein im-munoreactivities (substance P, [Met5]- and [Leu5]enkepha-lin, neurotensin, vasoactive intestinal peptide, calcitonin,cholecystokinin octapeptide, and bombesin) were analyzedwith commercially available radioimmunoassay kits.Enzymatic Activities. 5-HT biosynthesis [tryptophan hy-

droxylase (EC 1.14.16.4) plus aromatic L-amino acid decar-boxylase (EC 4.1.1.28) activities] was measured (i) by using[3H]tryptophan and a radiochromatographic procedure de-rived from Beevers et al. (42) and (ii) by the 5-HT radioen-zymatic assay (39) after incubation of cell homogenates withtryptophan.Monoamine oxidase (MAO, EC 1.4.3.4) type A and B

activities were determined radioenzymatically according toGuicheney et al. (43), with [14C]5-HT and 3-[14C]phenylethyl-amine as substrates, respectively.Uptake Experiments. Uptake experiments were initiated by

the addition of [3H]5-HT or [3H]dopamine after two washingswith serum-free Dulbecco's modified Eagle's medium(DMEM). As assessed by radioenzymatic measurements,this is sufficient to eliminate the 5-HT initially present in thestandard culture medium. After incubation for various peri-ods at 37°C or 4°C, cells were washed twice and prepared forliquid scintillation counting (44). Cell protein content wasmeasured by the Lowry procedure. Apparent Km and Vm.values were determined by a nonlinear regression performedon an IBM PC AT2 computer using the ENZ-FIT (ElsevierBiosoft) program.

RESULTSImmortalized Clone 1C1l Has Features of a Neuronal (or

Neuroendocrine) Precursor Cell. After retinoic acid- andBt2cAMP-induced differentiation of the murine embryonalcarcinoma F9K4b2, clone iC11 was selected for the expres-sion of the SV40 T antigen, the loss of embryonal carcinomamarkers, and the ability to differentiate in vitro into neural-like derivatives. iC11 cells express the SV40 T antigen in allcell nuclei (Fig. 1 A and B), grow indefinitely in culture,maintaining a stable phenotype, and are tumorigenic insyngeneic mice.The cell population does not express embryonal carcinoma

markers or markers of extraembryonic endoderm (e.g., theextracellular matrix components laminin and type IV colla-gen, or a-fetoprotein) even after addition of retinoic acid, aclassical inducer of embryonal carcinoma cell endodermdifferentiation. Reversion towards the embryonal carcinomastate or endodermal or mesodermal differentiation has neverbeen detected either in vitro or in the >100 tumors analyzed.The tumors (not shown) are composed of immature cellsarranged in nodular structures that resemble immature neu-roblastomas. In some instances, the cells are oriented aroundhost capillaries, with an organization evocative of certainneuroendocrine tumors.When grown in standard medium or after addition of

retinoic acid, 1C11 cells exhibit an epithelial morphology(Fig. 1A) and express markers of epithelial cells such asL-CAM and the cytokeratin Endo A (Fig. 1 C and D). Thisepithelial phenotype holds for up to 99o of the cell popula-tion. However, we consistently find in the cultures a fewpercent of round refringent cells presenting bipolar exten-sions. This heterogeneity is maintained after subcloning ofiC11. Strikingly, after 2-3 days in the presence of Bt2cAMP,a classical inducer of neuroblastoma differentiation (8, 9),95-100% of the cells acquire a neural-like phenotype (Fig.

FIG. 1. iC11 cells. (A) Phase contrast. (B-D) Indirect immuno-fluorescence staining with monoclonal antibody against the SV40 Tantigen, monoclonal antibody against the cytokeratin Endo A, andrabbit antiserum against L-CAM, respectively. (A-C, methanolfixation at -20'C; D, fixation with 3% paraformaldehyde; x400.)

Neurobiology: Buc-Caron et al.

1924 Neurobiology: Buc-Caron et al.

2A) with bipolar extensions that are 5-10 times the length ofthe cell body. Similar morphological changes can be inducedin serum-free medium (12) or when dimethyl sulfoxide or

cyclohexanecarboxylic acid (45) is added to Bt2cAMP. Neu-rite outgrowth is paralleled by the acquisition or increasedexpression of several markers related to neural function. Thecells now express N-CAM (Fig. 2C) in addition to L-CAM.Intense fibronectin immunoreactivity is observed in the cellbodies and is often concentrated at the tips of the neurites(Fig. 2F). Synaptophysin, a membrane constituent of syn-

aptic vesicles of neural and neuroendocrine cells (29, 30), isfound in both the cell bodies and the extensions (Fig. 2D andE), whereas only very weak immunoreactivity is found in theprecursor cells (data not shown).

Interestingly, all Bt2cAMP/cyclohexanecarboxylate-treated iCli cells, now referred to as 1C11*, continue todivide (generation time of 30 hr, vs. 20 hr for iC1i). WhenBt2cAMP is removed from the medium the cells lose theirbipolar extensions within 3-8 days. When such cells are

replated in Bt2cAMP-containing medium, they regenerateprocesses within 24 hr. These properties, cell division andneurite reversibility, unusual for cells harboring many neural-like differentiation markers, may be related to the cells'expression of the "immortalizing" T-antigen gene (Fig. 2 Aand B). Alternatively, 1C11* cells may correspond to an

intermediate differentiation stage, lacking further signal(s)for full differentiation. Finally, iC11 cells are not labeled withanti-neurofilament antibodies, and no specific RNA tran-script is detected by Northern blot analysis with a cDNAprobe specific for mRNA encoding the 68-kDa neurofilamentprotein (J. Deschamps, Hubrecht Laboratory, Utrecht, TheNetherlands, personal communication).We are therefore dealing with an immortalized, immature

cell that is committed to a neuronal or neuroendocrinepathway of differentiation and that can exist in at least twostates: a precursor cell (iC11) and a more differentiated celltype (1C11*) that expresses neuronal (or neuroendocrine)markers. The switch from one state to the other can beinduced in almost 100% of the cell population.

Expression of Serotoninergic Markers. lC11 and 1C11*cells were examined for the presence of neuropeptides,neurotransmitters, and related enzymes. Among the neu-

ropeptides tested (see Materials and Methods), only

D

FIG. 2. 1C11* cells. (A and D) Phase contrast. (B, C, E, and F)Indirect immunofluorescence staining with monoclonal antibodyagainst the SV40 T antigen (same field as A), rabbit antiserum againstN-CAM, monoclonal antibody against synaptophysin (same field as

D), and rabbit antiserum against fibronectin, respectively. (A, B, andF, methanol fixation at -20°C; C, 3% paraformaldehyde fixation; Dand E, fixation with 3% paraformaldehyde and permeabilization with0.2% Triton X-100; x280.)

FIG. 3. 1C11* cells. (A-C) Indirect immunofluorescence stainingwith anti-5-HT, anti-5-HT transporter, and anti-[Met5]enkephalin,respectively. (D) Mepacrine storage in acidic granules. (A, fixationin 3% paraformaldehyde followed by 70%6 ethanol; C, fixation in 3%paraformaldehyde and permeabilization with 0.2% Triton X-100; Band D, unfixed cells; A, x280; B-D, x400.)

[Met5]enkephalin-like immunoreactivity was found in ho-mogenates of 1C11* (19 pg per mg of protein) (see also Fig.3C). Interestingly, [Met5]enkephalin immunoreactivity wasnot present in F9K4b2 and 1C11 homogenates but wasdetected (67 pg per mg of protein) in 1C11 treated withretinoic acid. The key enzyme of the cholinergic pathway,choline acetyltransferase, was not detectable either by itsenzymatic activity in extracts or by immunoreactivity withspecific antibodies. The catecholamines dopamine, norepi-nephrine, and epinephrine were all absent and dopamineuptake was not found. Tyrosine hydroxylase, the limiting

Table 1. Comparison between mature serotoninergic (5-HT)neurons and neural-like cells derived from teratocarcinoma

F9K4b2 lC11 lC11* 5-HTCharacteristic cells cells cells neurons

5-HT biosynthesis(tryptophan hydroxylase) - + + +

MAOTypeA - - - -Type B + + + +

5-HT uptakeEnergy-dependent - - + +

Plasmatic carrier - - + +

Granular carrier - - + +

Subcellular 5-HT storageMepacrine-labeled granules - - + +

Synaptophysin - + + +

Neurofilament triplet - - - +

Mature serotoninergic neurons are characterized by the expres-sion of neurofilaments and by the ability (i) to synthesize 5-HT fromL-tryptophan, (ii) to catabolize excess 5-HT via MAO-B, (iii) to takeup 5-HT through carriers located in the plasma membrane (parox-etine- and clomipramine-sensitive) and storage granules (reserpine-sensitive) and associated with energy-dependent pumps (ouabain-sensitive), and (iv) to store 5-HT in acidic granules.

Proc. Natl. Acad. Sci. USA 87 (1990)

Proc. Natl. Acad. Sci. USA 87 (1990) 1925

Table 2. Serotoninergic properties of neural-like cells derivedfrom teratocarcinoma and of brain raphe nuclei

F9K4b2 iCi iC11* Raphe nucleiProperty cells cells cells (ref. 46)

5-HT, pmol per mgof protein 14 + 5 22 ± 9 208 ± 16 123

5-HT uptakeApparent Km X 107 ND ND 0.6 ± 0.05 1.7Apparent V.,pmol/mg per min <1 <1 36 ± 6.5 1570

InhibitionClomipramine - - + +Paroxetine - - + +Ouabain - - + +Ro 4-1284 - - + +

Enzyme activities,pmol/mg per minTrpOHaset <0.3 5 ± 1 3 ± 1 290MAO-A <0.1 <0.1 <0.1 <0.1MAO-B 24 ± 6 37 ± 8 45 ± 7 1230

Values from this work are means ± SEM (n = 4). ND, notdetermined; -, inhibition not tested in absence of any uptake.tTryptophan hydroxylase.

enzyme in the catecholamine pathway, was not detected byseveral methods (enzymatic activity, immunodetection, spe-cific transcripts). Histamine was also not detected.However, after differentiation by Bt2cAMP and cyclohex-

anecarboxylate, 1C11* cells acquire features of serotoniner-gic cells (Table 1). Specific monoclonal antibodies revealed5-HT both inside the cell bodies (Fig. 3A) and in the exten-sions. The presence of5-HT has been confirmed by biochem-ical measurements: Table 2 shows that 1C11* cells containabout 10 times more 5-HT than either iC11 or F9K4b2 cells.The complete 5-HT synthetic pathway (tryptophan hy-

droxylase and aromatic L-amino acid decarboxylase) wasfound by two independent methods to be functional in bothC1i1 and 1C11* but not in F9K4b2 (Table 2). As calculatedfrom Table 2, tryptophan hydroxylase activity accounts for atleast 10% of the 5-HT present in iC11* cells. As in sero-toninergic neurons (47), MAO-B activity is present (Table 2).1C11* cells also contain the structures required to take up

5-HT. First, specific monoclonal antibodies raised against the5-HT transport system react with cell bodies and extensions,giving a characteristic patchy pattern (Fig. 3B), whereas noimmunoreactivity is found on iC11 cells. Second, in contrastto the precursor iC11, 1C11* cells take up 5-HT. The 5-HT

120

100o 100

h80 <, + ,80

0

060 ~60

0 ~~~~~~0

FI.40 [3HS-H (01,M4pak0yl1*cll ma E

F-4~~~~~~~~~~~~~4

tA20 20

00 10 20 301 2 3 4 5

Time, mmn

FIo. [3H]5-HT (0.1 4f.M) uptake by1C11* cells (mean ± SEM,n = 4). (Left) Kinetics. (Right) Inhibition by various drugs (uptakeincubation time, 5 min). Bar 1, control; bar 2, clomipramine (10 nM);bar 3, culture for 3 days with Ro 4-1284 (10 uM); bar 4, paroxetine(10 nM); bar 5, preincubation for 4 hr with ouabain (0.1 mM).

E160 |

.~1204)4

5 80

0.~~~~~~~~00

s. l/~~~~1(S X 106)4.P 0 0.20 0.40 0.60 0.80

S (= 5-HT, AM)

FIG. 5. Determination of kinetic constants for 5-HT uptake by1C11* cells (5 min, mean ± SEM, n = 4). (Inset) Double reciprocalplot.

uptake process is time-dependent (Fig. 4 Left), saturable(Fig. 5), and blocked at 4TC. The apparent Ki,, 0.6 x 10-7 M(Fig. 5; Table 2), is of the same order of magnitude as that of5-HT neurons of the brain or tumor cells of neuroendocrineorigin (46, 48).Moreover, different elements of the transport machinery

are functional in 1C11* cells. These include (i) the plasmamembrane carrier, inhibited by paroxetine and associatedwith a ouabain-sensitive Na+,K+-ATPase pump; (ii) imipra-mine binding sites downregulating 5-HT uptake, blocked byclomipramine; and (iii) the carrier of the storage granule,sensitive to the reserpine analog Ro 4-1284 (Fig. 4 Right;Table 1). Finally, in contrast to iCi1, 1C11* cells are labeledby mepacrine (Fig. 3D), a molecule known to be concentratedin acidic compartments of the cells, including the acidicstorage granules (49).

DISCUSSIONWe report here the isolation and characterization of a sero-toninergic cell line, 1Cil, derived from a mouse teratocar-cinoma. iCli cells appear to be committed to the neuroec-todermal lineage.1C11 presents two states: an immature phenotype (1Cli

precursor) and a more differentiated state (1C11*). Afterinduction by Bt2cAMP and cyclohexanecarboxylate, 1Ci1*cells extend neurites and express neural markers such asN-CAM, synaptophysin, and [Met5]enkephalin in >95% ofthe clonal population. More importantly, 1C11* cells expressa serotoninergic phenotype, characterized by a high 5-HTcontent and the ability to take up, store, synthesize, andcatabolize 5-HT.The activity of the key enzyme in 5-HT biosynthesis,

tryptophan hydroxylase, is low but significant. MAO-B, thecatabolic enzyme present in the serotoninergic neurons oftheraphe nuclei, is also active. Both activities were also detectedin the noninduced cells, confirming that they are alreadyengaged in the serotoninergic pathway.Four days after induction, the 5-HT content of 1C11* cells

has increased 10-fold and an active uptake system has beengenerated that is inhibited by specific drugs and whosekinetic parameters are comparable to those of serotoninergicneurons. The 1C11* cell population is homogeneously la-beled with anti-synaptophysin and anti-[Met5]enkephalin an-tibodies and exhibits a positive reaction with mepacrine.Such results point to the existence of secretory granules in1C11*, which need to be characterized further by ultrastruc-tural studies. Our data indicate that the 5-HT storage anduptake systems are qualitatively functional in lC11'* cells(Figs. 4 and 5; Table 2). This does not hold true for MTC cells

Neurobioloigy: Buc-Caron et al.

1926 Neurobiology: Buc-Caron et al.

(24), the most complete in vitro serotoninergic model so far,in which active, saturable uptake has not been demonstrated.MTC cells also lack MAO-B and exhibit only 1/10th thetryptophan hydroxylase activity of iC11. Most important,the MTC cell population is heterogeneous with respect toneuropeptides and 5-HT immunoreactivity.Compared to the brain region richest in tryptophan hy-

droxylase activity, the raphe nuclei, 1C11* cells have a lowlevel of this enzyme and show a low rate of 5-HT uptake. Toexplain this, two nonexclusive interpretations can be given.(i) The high 5-HT concentration of the culture medium (0.1-1AM) might downregulate both the rate of5-HT uptake and theactivity of tryptophan hydroxylase. Such an explanation hasbeen put forward for platelets (50) and MTC cells (24). (ii)Based upon the homogeneity of the induced cell population,each 1C11* cell may contain a low number of functionalcarrier molecules compared to adult serotoninergic cells.High 5-HT content associated with low 5-HT uptake andtryptophan hydroxylase activity might be properties of com-mitted, but as yet not fully differentiated, cells (51).

In conclusion, the most remarkable feature of the 1C1lsystem is its inducibility. Almost 100% of the population can

be induced to express [Met5]enkephalin, synaptophysin, andthe serotoninergic machinery. This points to the neuronal or

neuroendocrine origin of the cells (52). Such an inducible cellline has not been available until now. The potential forproliferation and the homogeneity of the immortalized iC11and 1C11* cells make it possible to obtain large amounts ofmaterial. This cell line constitutes a valuable tool not only forcellular and molecular studies concerning the physiology andthe pharmacological modulation of the 5-HT phenotype butalso for genetic studies concerning the onset of differentiationof 5-HT cells.

We thank Prof. F. Jacob, in whose laboratory most of this workwas performed, for constant support and Prof. H. Buc, Prof. M. DaPrada, Prof. M. Hamon, Dr. P. Avner, and Dr. M. Weiss for helpfuldiscussions. We are grateful to J. M. Villette for determination ofmost neuropeptides. This work was aided by grants from theAssociation pour la Recherche contre le Cancer, the Centre Nationalde la Recherche Scientifique, the Fondation pour la RechercheMddicale, the Ligue Nationale Franqaise contre le Cancer, and theFondation Andrd Meyer.

1. Augusti-Tocco, G. & Sato, G. (1969) Proc. Nati. Acad. Sci. USA 64,311-315.

2. Biedler, L., Roffler-Tarlov, S., Schachner, M. & Freedman, L. S.(1978) Cancer Res. 38, 3751-3757.

3. Schubert, D., Humphreys, S., Baroni, C. & Cohn, M. (1969) Proc.Natl. Acad. Sci. USA 64, 316-323.

4. Seeds, N. W., Gilman, A. G., Amano, T. & Nirenberg, M. W.

(1970) Proc. Natl. Acad. Sci. USA 66, 160-167.5. Gurwitz, D. & Cunningham, D. D. (1988) Proc. Natl. Acad. Sci.

USA 85, 3440-3444.6. Krystosek, A. (1985) J. Cell. Physiol. 125, 319-329.7. Amano, T., Richelson, E. & Nirenberg, M. (1972) Proc. Natl. Acad.

Sci. USA 69, 258-263.8. Furmanski, P., Silverman, D. & Lubin, M. (1971) Nature (London)

233, 413-415.9. Ross, J., Olmsted, J. B. & Rosenblaum, J. L. (1975) Tissue Cell 7,

107-136.10. Greene, L. A. & Tischler, A. S. (1976) Proc. Natl. Acad. Sci. USA

73, 2424-2428.11. Yoffe, J. R. & Borchardt, R. T. (1982) Mol. Pharmacol. 21, 368-

373.12. Edde, B. & Darmon, M. (1985) in Cell Culture in the Neurosciences,

eds. Bottenstein, J. E. & Sato, G. (Plenum, New York), pp.273-285.14.

13. Jones-Villeneuve, E. M. V., McBurney, M. W., Rogers, K. E. &Kalnins, V. I. (1982) J. Cell Biol. 94, 253-262.

14. Kuff, E. L. & Fewell, J. W. (1980) Dev. Biol. 77, 103-115.15. Liesi, P., Rechardt, L. & Wartiovaara, J. (1983) Nature (London)

306, 265-267.

16. Pfeiffer, S. E., Jakob, H., Mikoshiba, K., Dubois, P., Gu6net,J. L.,Nicolas, J. F., Gaillard, J., Chevance, G. & Jacob, F. (1981) J. CellBiol. 88, 57-66.

17. Sharma, S. & Notter, M. F. D. (1988) Dev. Biol. 125, 246-254.18. Da Prada, M., Richards, J. G. & Kettler, R. (1981) in Platelets in

Biology and Pathology 2, ed. Gordon, J. L. (Elsevier, Amsterdam),pp. 105-145.

19. Da Prada, M., Cesura, A. M., Launay, J. M. & Richards, J. G.(1988) Experientia 44, 115-126.

20. Hilibrand, A. S., Richardson, U. I., Blusztain, J. K. & Wurtman,R. J. (1987) Neurochem. Int. 10, 185-189.

21. De Vitry, F., Catelon, J., Dubois, M., Thibault, J., Barritault, D.,Courty, J., Bourgoin, S. & Hamon, M. (1986) Neurochem. Int. 9,43-53.

22. Debons-Guillemin, M. C., Launay, J. M., Roseto, A. & Peries, J.(1982) Cancer Res. 42, 1513-1516.

23. Leong, S. S., Horoszewicz, J. S., Shimaoka, K., Friedman, M.,Kawinski, E., Song, M. J., Zeigel, R., Chu, M., Baylin, S. &Mirand, E. A. (1981) in Advances in Thyroid Neoplasia, eds.Andreoli, M., Monaco, F. & Robbins, J. (Field Educational Italia,Rome), pp. 95-108.

24. Tamir, H., Liu, K., Payette, R. F., Hsiung, S., Adlersberg, M.,Nunez, E. A. & Gershon, M. D. (1989) J. Neurosci. 9, 1199-1212.

25. Kellermann, 0. & Kelly, F. (1986) Differentiation 32, 74-81.26. Bernstine, E. G., Hooper, M. L., Grandchamp, S. & Ephrussi, B.

(1973) Proc. Natl. Acad. Sci. USA 70, 3899-3903.27. Kellermann, O., Buc-Caron, M. H. & Gaillard, J. (1987) Differen-

tiation 35, 197-205.28. Montano, X. & Lane, D. P. (1964) J. Virol. 51, 760-767.29. Navone, F., Jahn, R., Di Gioia, G., Stukenbrok, H., Greengard, P.

& De Camilli, P. (1986) J. Cell Biol. 103, 2511-2527.30. Wiedenmann, B., Franke, W. W., Kuhn, C., Moll, R. & Gould,

V. E. (1986) Proc. Natl. Acad. Sci. USA 83, 3500-3504.31. Kemler, R., Brilet, P. & Jacob, F. (1981) Immune Syst. 1, 102-109.32. Milstein, C., Wright, B. & Cuello, A. C. (1983) Mol. Immunol. 20,

113-123.33. Thibault, J., Vidal, D. & Gros, F. (1981) Biochem. Biophys. Res.

Commun. 99, 960-965.34. Cesselin, F., Hamon, M., Bourgoin, S., Buisson, N. & de Vitry, F.

(1982) Neuropeptides 2, 351-369.35. Peyrieras, N., Louvard, D. & Jacob, F. (1985) Proc. Natl. Acad.

Sci. USA 82, 8067-070.36. Covault, J. & Sanes, J. R. (1985) Proc. Natl. Acad. Sci. USA 82,

4544-4548.37. Grimaud, J. A., Druguet, M., Peyrol, S., Chevalier, 0., Herbage,

D. & El Badarawy, N. (1980) J. Histochem. Cytochem. 28, 1145-1156.

38. Buc-Caron, M. H., Darmon, M., Poiret, M., Sellem, C., Sala-Trepat, J. & Erdos, T. (1983) Cold Spring Harbor Conf. CellProliferation 10, 411-419.

39. Walker, R. F., Friedman, D. W. & Jimenez, A. (1983) Life Sci. 33,1915-1924.

40. Haimart, M., Launay, J. M., Zurcher, G., Cauet, N., Dreux, C. &Da Prada, M. (1985) Agents Actions 16, 71-75.

41. Da Prada, M. & Zurcher, G. (1979) in Radioimmunoassay ofDrugsand Hormones in Cardiovascular Medicine, eds. Albertini, A., DaPrada, M. & Peskar, B. (Elsevier, Amsterdam), pp. 175-188.

42. Beevers, S. J., Knowles, R. G. & Pogson, C. I. (1983) J. Neuro-chem. 40, 894-897.

43. Guicheney, P., Soliman, H. R., Launay, J. M., Dreux, C. & Meyer,P. (1988) Clin. Exp. Hypertension Part A 10, 533-544.

44. Costa, J. L., Murphy, D. L. & Kafka, M. S. (1977) Biochem.Pharmacol. 26, 517-521.

45. Portier, M. M., Croizat, B., Berthelot, F., Edde, B., Paulin, D. &Gros, F. (1983) in Gene Expression in Normal and TransformedCells, eds. Celis, J. E. & Bravo, R. (Plenum, New York), pp. 1-43.

46. Osborne, N. N. & Hamon, M., eds. (1988) Neuronal Serotonin(Wiley, London).

47. Westlund, K. N., Denney, R. M., Kochersperger, L. M., Rose,R. M. & Abell, C. W. (1985) Science 230, 181-183.

48. Launay, J. M., Debons-Guillemin, M. C., Roseto, A., Villette,J. M., Dreux, C. & Peries, J. (1983) Diabete Mctab. 9, 16 (abstr.).

49. Lorez, H. P., Da Prada, M., Rendu, F. & Pletscher, A. (1977) J.Lab. Clin. Med. 89, 200-206.

50. Laubscher, A. & Pletscher, A. (1979) J. Pharm. Pharmacol. 31,284-289.

51. Lauder, J. M., Wallace, J. A., Krebs, H., Petrusz, P. & McCarthy,K. (1982) Brain Res. Bull. 9, 605-625.

52. H6kfelt, T., Fuxe, K. & Pernow, B., eds. (1986) Coexistence ofNeuronal Messengers: A New Principle in Chemical Transmission(Elsevier, Amsterdam).

Proc. Natl. Acad. Sci. USA 87 (1990)