Ciliary neurotrophic factor coordinately activates transcription of ...

Proc. Natl. Acad. Sci. USA Vol. 90, pp. 572-576, January 1993 Neurobiology

Ciliary neurotrophic factor coordinately activates transcription of neuropeptide genes in a neuroblastoma cell line AVIVA J. SYMES*t, MAHENDRA S. RAOt, SUSAN E. LEWIS*§, STORY C. LANDISf, STEVEN E. HYMAN*§¶, AND J. STEPHEN FINK*t§II *Molecular Neurobiology Laboratory, tDepartment of Neurology, and IDepartment of Psychiatry, Massachusetts General Hospital and §Program in Neuroscience, Harvard Medical School, Boston, MA 02114; and tDepartment of Neurosciences, Case Western University, Cleveland, OH 44106

Communicated by Gerald D. Fischbach, August 31, 1992

(11-13). After binding to specific receptors, these factors utilize a common signal-transducing transmembrane glycoprotein, gp130, and increase intracellular tyrosine phosphorylation (14-16). Subsequent steps in the intracellular signaltransduction pathway initiated by the interaction of CNTF with its receptor remain to be characterized. To begin to characterize the intracellular signaling mechanisms mediating the diverse cellular effects of CNTF on neuronal cells, we have identified a human neuroblastoma cell line, NBFL, in which CNTF coordinately increases VIP, somatostatin, and calcitonin gene-related peptide (CGRP) mRNAs. Neuroblastoma cells are thought to originate from the neural crest (17), to have properties in common with cells of the sympathoadrenal lineage, and therefore to be cells that might respond to CNTF in a fashion similar to sympathetic neurons. We have investigated the molecular mechanisms by which CNTF activates VIP gene expression in NBFL cells. This study provides new information about the intracellular signaling mechanisms by which CNTF coordinately activates neuropeptide gene transcription to influence neuronal differentiation.

ABSTRACT Differentiation factors have been identified that influence the phenotype of sympathetic neurons by altering expression of classical neurotransmitters and neuropeptides. Investigation of the molecular mechanisms through which such factors act would be facilitated by the availability of a neuronal cell line that responds to these factors in a fashion similar to sympathetic neurons. We have identified a human neuroblastoma cell line, NBFL, that responds to the differentiation factor ciary neurotrophic factor (CNTF) by coordinately inducing multiple neuropeptide genes as do sympathetic neurons. Treatment of NBFL cells with CNTF increases vasoactive intestinal polypeptide (VIP), somatostatin, and calcitonin gene-related peptide (CGRP) mRNAs but does not change other neurotransmitter properties. The induction of VIP mRNA by CNTF in NBFL cells is dose dependent, rapid, sstained, and independent of new protein synthesis. Genomc 5' flanking sequences located within a 1.59-kilobase region of the human VIP gene and distinct from the previously defined cAMP-responsive element subserve trancriptional activation by CNTF. Further examination of NBFL cells should permit the elucidation of the molecular mechanisms by which CNTF and other differentiation factors coordinately activate neuropeptide gene transcription to influence neuronal differentiation. Similar mechanisms may mediate the effect of CNTF on neuronal survival.

MATERIALS AND METHODS Cell Culture. NBFL cells were maintained at 370C under 5% C02/95% air in Dulbecco's modified Eagle's medium (4.5 g of glucose per liter) supplemented with 10%6 (vol/vol) heat-inactivated fetal calf serum, 5% (vol/vol) heatinactivated horse serum, and 1% (wt/vol) penicillin/ streptomycin (GIBCO, BRL). All experiments were performed within a window of 20 passages and were repeated 2-4 times; in later passages the response to CNTF was attenuated. Before all treatments, cells were plated at a density of 106 cells per 100-mm dish. Recombinant rat CNTF (Regeneron Pharmaceuticals, Tarrytown, NY) or nerve growth factor (NGF) (2.5S, 50 ng/ml; Boehringer Mannheim) or both were added to the medium as indicated, 1 day after plating. For long-term experiments, the medium was changed, and the factors were replenished every 2-3 days. Cell Proliferation Assay. Five thousand cells per well were plated in a 96-well plate. One day later the medium was replaced with CNTF-containing medium, and the cells were grown for 48 hr. Cell number was assessed by using cellular conversion of a tetrazolium salt to a blue formazon product (CellTiter 96 assay, Promega), and the absorbance at 560 nm was read with an ELISA reader.

Ciliary neurotrophic factor (CNTF) is a multifunctional cytokine that can regulate the survival and differentiation of many types of developing and adult neurons (1-4). Sympathetic neurons and their precursors have been utilized as a model system to investigate neuronal differentiation (5-9). CNTF has effects on sympathetic neuroblasts and postmitotic sympathetic neurons. When chicken sympathetic neuroblasts in culture are treated with CNTF, they stop dividing and differentiate (5). Primary cultures of postmitotic rat sympathetic neurons undergo differentiation in response to CNTF treatment switching their phenotype from adrenergic to cholinergic (6, 9). In addition to inducing expression of choline acetyltransferase (CHAT) and decreasing expression of tyrosine hydroxylase (TH) in sympathetic neurons, CNTF also alters the expression of several neuropeptides; it increases vasoactive intestinal polypeptide (VIP), somatostatin, and substance P and decreases neuropeptide Y (NPY)

(9).

The molecular mechanisms by which CNTF influences neuronal survival and differentiation are not understood. CNTF is a 24-kDa acidic protein without close sequence homology to any known protein (10). However, based on its proposed tertiary structure and the structure of its receptor, CNTF is considered to be a member of a family of cytokines that includes leukemia-inhibitory factor/cholinergic differentiation factor (LIF/CDF), oncostatin M, and interleukin 6

Abbreviations: CNTF, ciliary neurotrophic factor; LIF/CDF, leukemia inhibitory factor/cholinergic differentiation factor; VIP, vasoactive intestinal polypeptide; CGRP, calcitonin gene-related peptide; NGF, nerve growth factor; TH, tyrosine hydroxylase; CHAT, choline acetyltransferase; NPY, neuropeptide Y; CRE, cAMPresponsive element; CAT, chloramphenicol acetyltransferase. 'To whom reprint requests should be addressed at: Molecular Neurobiology Laboratory, Department of Neurology, Massachusetts General Hospital, Boston, MA 02114.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 572

Neurobiology: Symes et A RNA Isolation and Analysis. Total cytoplasmic RNA was isolated from tissue culture cells after lysis with Nonidet P-40 as described (18). Each RNA sample was electrophoretically separated on formaldehyde/1.4% agarose gels and electrotransferred onto GeneScreen membranes (NEN/DuPont). Northern blots were prehybridized and hybridized at 420C in 50% formamide/10%o dextran sulfate/1% SDS/1 M NaCl/ 0.05 M Tris HCl, pH 7.5/0.1% sodium pyrophosphate/0.2% Ficoll/0.2% bovine serum albumin/0.2% polyvinylpyrrolidone with probes labeled (19) with [a-32P]dCTP to a specific activity of 0.5-1 x 109 cpm/,ug. Membranes were washed with 0.5x SSC (lx SSC = 0.015 M NaCl/0.0015 M sodium citrate, pH 7) containing 0.1% SDS at 500C unless otherwise stated. Each membrane was hybridized with a probe for the unregulated internal reference gene cyclophilin (20). Autoradiographs were digitized with a Microtek scanner and analyzed with the National Institutes of Health IMAGE 1.41 program. The relative densitometric readings were normalized to cyclophilin mRNA to account for loading differences between lanes. Plasmids and Probes. VIP mRNA was detected with a 580-base-pair (bp) HindIII/EcoRI fragment of human VIP cDNA (21). Other probes were a 400-bp EcoRI/HindIII rat somatostatin fragment of pSRl (18) and a 900-bp BamHI fragment of rat cyclophilin in pCD15 (20). Transcription reporter plasmids containing fused genes for VIP and luciferase were constructed by ligating a Kpn I/HindIII polylinker fragment of pGem-3Z (Promega) to Kpn I/HindIIIdigested pA31uc (22) to form pA3Pluc. A 2.1-kilobase (kb) HindIII fragment of VIP20001acz (23) was ligated to HindIIIdigested pA3Pluc to construct VIP1929luc. VIP3401uc was constructed by ligation of a 480-bp Pst I/Nsi I fragment of VIPCAT-1 containing the gene for chloramphenicol acetyltransferase (CAT) (24) to Pst I-digested pA3Pluc. Ligation of Sma I/Pst I-digested pA3Pluc and a 240-bp Nco I (blunt)/Pst I-fragment of VIPCAT-4 (24) generated VIP94luc. Cell Transfection and Analysis. NBFL cells were plated at 1.8 x 106 cells per 10-cm dish 1 day before transfection. Cells were transfected by the calcium phosphate precipitation method exactly as described (25). Luciferase plasmid (30 ,tg) was added to each plate together with 2.5 ,ug of pRSVcat to control for transfection efficiency. Cells were treated beginning 24 hr after transfection; 36 hr later, cells were harvested, and luciferase (26) and CAT activity were assayed (27). Peptide Assays. Neuropeptide levels were determined by radioimmunoassay as described (9). VIP and somatostatin were assayed by general protocols as specified in commercially available RIA kits (Incstar, Stillwater, MI). The primary antibodies have been previously shown to display minimal cross-reactivity with other peptides.

RESULTS Phenotypic Characterization of NBFL Cells. The NB5-B2 cell line, the parent line of NBFL cells, was isolated from a cervical lymph node metastasis of an adrenal neuroblastoma (28). NB5-B2 cells require an adherent substrate such as collagen for attachment. Cells of the NBFL subclone of NB5-B2 were isolated from the parent cell line based on their ability to adhere readily to plastic. NBFL cells have a flat appearance and grow as a monolayer. NBFL cells were examined for a number of transmitter properties characteristic of sympathetic neurons. In untreated NBFL cells, neither mRNAs for TH nor CHAT, synthetic enzymes for catecholamines and acetylcholine, respectively, were detectable by Northern blot. Further, CHAT activity was undetectable. A small percentage of NBFL cells (4-fold and somatostatin immunoreactivity increased 3-fold compared with untreated cells (Fig. 1). To determine whether the elevations in VIP and somatostatin content correlated with increased levels of their mRNAs, VIP and somatostatin mRNA were measured by Northern blot. VIP mRNA was increased 40-fold after 24 hr (Fig. 2 Upper). The increase was maximal (50-fold) after 3 days and was sustained for up to 14 days of continuous treatment with CNTF, the longest time point tested (data not shown). The induction of somatostatin mRNA was also large; somatostatin mRNA from CNTFtreated NBFL cells was readily detectable (Fig. 2 Upper). However, the quantitative induction of somatostatin mRNA by CNTF could not be determined because somatostatin mRNA in untreated NBFL cells was not detectable. The mRNA encoding CGRP increased 3-fold after 7 days of CNTF treatment (Fig. 2 Lower). The levels of NPY mRNA were unchanged after CNTF treatment (5 ng/ml; 1-7 days of treatment; data not shown). CHAT activity and substance P mRNA were not detectable in NBFL cells after CNTF treatment. Treatment of NBFL cells with CNTF produced no morphological changes even after 14 days. CNTF also produced no change in cell proliferation at doses ranging from 0.1 to 50 ng/ml (Table 1). The number of cells expressing TH immunoreactivity or catecholamines and the intensity of TH staining or of catecholamine fluorescence of individual cells were not altered by CNTF (5 days at 5 ng/ml; data not shown). To ascertain the concentration of recombinant CNTF necessary to elicit a detectable change in VIP gene expression in NBFL cells, levels of VIP mRNA were examined after 3 days of treatment with various concentrations of CNTF (0.1-50 ng/ml). CNTF increased VIP mRNA at concentrations as low as 0.5 ng/ml, and a maximal VIP response was produced at 5 ng/ml (Fig. 3). Higher concentrations of CNTF appeared slightly inhibitory. Since both sympathetic neurons and chromaffin cells respond to NGF, we examined the effects of NGF alone and in combination with CNTF on NBFL cells. Treatment of NBFL cells with either NGF or NGF plus CNTF for up to 14 days produced no detectable differences in the morphological

=

300 2 ) 0-

a0control

FIG. 1. CNTF induction of VIP and somatostatin peptides in NBFL cells. NBFL cells were grown in control media or treated with CNTF (5 ng/ml). Five days after treatment began, cells were homogenized and peptide levels were determined in cell lysates by radioimmunoassay. All samples were run in duplicate (n = 3 experiments). Results are expressed as pg per well + SEM.

- ~ ~ mRNA

Neurobiology: Symes et al.

574 0 LU

LU LL C3 Z

L I

Z U

U.

L

z 0

2

L.

+

-

LL

L a

Z

Z

0

Z 0

L

C.)

+

LL U.

U.

UL

U.

z U

z U

LI

+ u

a

Z 0

u.

z

-DAY 3-

-DAY 2-

-DAY 1-

a Z

a Z

U.

z 0

LI

+

U.

Z a U C)Z

a

-DAY 4-

+

L

L.

0 Z O Z 0 Z -DAY 6-

_

Sw * *_ _

Proc. Natl. Acad. Sci. USA 90 (1993)

_

~~~~~VIP

--mRNA SST mRNA

-_ =

=*

* a LU LU

c I-

z

L.

0 z

LI

-

z

-DAY 1

_

LI

U-

z

z

+

U.

LL

0 a z

z -

~~~~~CYCLOPHILIN

*

Table 1. CNTF does not alter cell proliferation Absorbance at 560 nm CNTF, ng/ml 0.281 + 0.046 0 0.258 ± 0.025 0.1 0.227 ± 0.034 0.5 0.247 ± 0.075 1.0 0.267 ± 0.054 2.5 0.254 ± 0.072 5.0 25 0.276 ± 0.061 50 0.242 ± 0.040 NBFL cells were cultured with the stated CNTF concentration. After 48 hr cell proliferation was assayed by spectrophotometric determination. Values are expressed as means ± SEM for six determinations.

LL

0

z

L

z

DAY 7 -

CGRP

EV

mRNA

CYCLOPHILIN mRNA

FIG. 2. CNTF induction of VIP and somatostatin (SST) mRNA (Upper) and CGRP mRNA (Lower) in NBFL cells. CNTF (5 ng/ml) or NGF (50 ng/ml) or both were added to cells as indicated. Thirty micrograms of total cytoplasmic RNA was loaded in each lane.

appearance ofthe cells (data not shown). Treatment of NBFL cells with NGF alone (50 ng/ml) did not alter the basal levels of VIP, somatostatin, or CGRP mRNA levels (Fig. 2). NGF, however, altered the response of the NBFL cells to CNTF. NGF increased the induction of VIP mRNA by CNTF 1.5-fold by 24 hr, but by 7 days the continued presence of NGF attenuated the induction of VIP mRNA by CNTF 20% (Fig. 2 Upper). NGF enhanced the induction of somatostatin mRNA by CNTF 2-fold by day 2 of treatment and attenuated the induction of somatostatin mRNA levels by 50% by day 7 of treatment (Fig. 2 Upper). Thus, while CNTF acts independently, its effects can be differentially modulated by NGF.

A

Induction of VIP mRNA by CNTF Is Rapid and Independent of New Protein Synthesis. To determine the earliest time at which an increase in VIP mRNA occurs, NBFL cells were treated with 5 ng of CNTF per ml and harvested 15 min to 12 hr later. Increased VIP mRNA levels were detectable by 30 min and they continued to increase throughout the 12-hr treatment period. After 12 hr ofCNTF treatment, VIP mRNA was increased 18-fold in NBFL cells (Fig. 4). To determine whether new protein synthesis is required for the induction of VIP mRNA by CNTF, NBFL cells were treated with CNTF (5 ng/ml) for 6 hr in the presence of protein-synthesis inhibitors. Treatment of NBFL cells for 6 hr with cycloheximide (100 jLM) or anisomycin (10 ,uM) alone increased VIP mRNA 3- to 4-fold (Fig. 5). CNTF treatment alone increased VIP mRNA 4-fold. Treatment of NBFL cells with anisomycin produced a slight enhancement of the CNTF effect, increasing VIP mRNA 7-fold. Thus, protein synthesis is not necessary for the increase in VIP mRNA levels in response to CNTF in NBFL cells. Genomic 5' Flanking Sequences Contained Within a 1.59-kb Region of the Human VIP Gene Subserve Transcriptional Activation by CNTF. To determine whether CNTF increases VIP gene transcription, we examined the effect of CNTF on the reporter gene encoding luciferase, fused to 5' flanking sequences from the human VIP gene (21). When NBFL cells were transfected with a VIP-luciferase fusion gene containing 1929 bp of 5' flanking sequences, CNTF treatment of NBFL

0.1 0-1 0.5 095 1.0 1.0 2.5 2.5 5.0 5.0

."

,l~t,.Pw;

25

25

w w .

50

50 w

VIP

CYCLOPHILIN

-1

.s

-0.5

0.5 log

FIG. 3.

1 .5

ug/ani

Dose-response of CNTF induction of VIP mRNA levels. NBFL cells were treated with the doses of CNTF indicated in ng/ml at

the top of each lane 1 day after plating. After 65 hr, plates were harvested, and RNA was isolated. (A) Northern analysis; 20 loaded in each lane. Blots were washed with 0.2 x SSC containing 0.1% SDS at

A.

65°C.

,u&g

of RNA was

(B) Densitometric analysis of Northern blot shown in

Proc. Natl. Acad. Sci. USA 90 (1993)

Neurobiology: Symes et al. z =

0

0

_

N

t

0

0

VYIP

lo_ -sA

z

0 0

o

CYCLOPHILIN

u

U.

>-

T z 4

z

z LL

00 >

_

(f

X

z