Cellular and Molecular Neurobiology, Vol. 12, No. 1, 1992

Retinoids and a Retinoic Acid Receptor Differentially Modulate Thymosin/310 Gene Expression in Transfected Neuroblastoma Cells Alan K. Hall ~ Received January 20, 1991; accepted June 5, 1991 KEY WORDS: retinoid; thymosin; neuroblastoma; gene expression.

SUMMARY

1. Investigations have demonstrated that the gene encoding thymosin ill0 (a 43-amino acid member of a family of related proteins originally described in the rat immune system) is a target for morphogenic retinoids in both human and rat neuroblastoma cells. 2. Structure-activity studies revealed that the stimulatory actions of retinoids upon the thymosin ill0 gene reflect the differing affinities of retinoid analogues for a retinoic acid receptor. 3. To examine further the possibility that the trophic actions of retinoic acid upon expression of the thymosin ill0 gene involved retinoid receptors, neuroblastoma cells were transiently transfected with an expression vector encoding the nuclear retinoic acid receptor (c0 protein. 4. Northern blot and slot-blot analyses revealed that neuronal cells overexpressing RARo:-mRNA exhibited an enhanced sensitivity to exogenous and endogeneous retinoic acid in terms of thymosin filo mRNA. Although the RAR-tr gene was expressed (at low levels) a priori in these neuroblastoma cells, retinoic acid ( 2 x 10-7M for 3 days) slightly stimulated RAR-tr-mRNA accumulation. 5. Collectively, these findings indicate the the retinoic acid receptor (o:) is regulated by retinoid acid and that the developmentally regulated, retinoidresponsive thymosin fllo gene is a target for this nuclear transcription factor in cells derived from the neural crest. 1Department of Surgery, UMDNJ-NJ Medical School, 185 South Orange Avenue, Newark, New Jersey 07103-2757. 45 0272-4340/92/0200-0045506.50/0(~) 1992PlenumPublishingCorporation

46

Hall INTRODUCTION

As a group of biologically active lipids, the retinoids can profoundly influence a diverse array of processes related to cell growth and differentiation. Many of these effects are currently believed to be mediated upon retinoid sensitive genes via specific nulcear receptors. Several receptors for the morphogen retinoic acid have now been identified by molecular cloning experiments including those designated o~, fl, and y (Giguere et al., 1987; Brand et al., 1988; Petrovich et al., 1987; Zelent et al., 1989). These retinoid receptors can differentially transactivate target genes in transfected cells (Astrom et al., 1991). More recently so-called "orphan receptors" (RXRs) have been characterized (Mangelsdorf et al., 1990) and subsequently shown to share homology with the Drosophila ultraspiracle (usp) gene (Oro et al., 1990). Additional molecular cloning experiments have now resulted in the identification of multiple isoforms of the RAR-o~ and R A R - f l genes (Leroy et al., 1991; Zelent et al., 1991). Recently thymosin fllo was shown to be retinoid sensitive in both human and rat neuroblastoma cells (Hall et al., 1990, 1991a). Thymosin ill0 is a small (20 fold over control (mock-transfected) cells. The slot-blot depicted in Fig. 2 was deliberately overexposed to emphasize RAR-o: expression in the mocktransfected neuronal cells. In order to facilitate a more precise examination of the relationship between episomally generated RAR-o: m R N A levels and the amount of expression vector transiently transfected, further analysis was undertaken. Hence, pSG5RAR~xI~

(~g)

)-RNA

r

0

I

5

l

Probe

RAR-,x

t8s

Fig. 3. Slot-blot analysis of episomally generated RAR-y gene expression in transiently transfected rat neuroblastoma cells. Varying amounts of expression vector pSG5RARol0 were transfected. The upper panel represents a filter hybridized with a 3-kb EcoR1 R A R - y fragment and exposed for 3 hr. The lower panel is the same filter rehybridized with an 18S rRNA control cDNA.

51

Retinoids Modulate flloGene Expression in Neuroblastoma Cells

Fig. 3 shows that RAR-tr mRNA is present at relatively low levels in these mock-transfected neuroblastoma cells (upper panel). Nevertheless, transfection with increasing amounts of the pSG5RARa~0 vector led to a corresponding augmentation of endogenous R A R - ~ mRNA expression; ambient levels of RAR-o~ mRNA (in this experiment) were stimulated 3.7-fold and 23.8-fold by 1 and 5/~g of plasmid, respectively (values normalized to 18S rRNA mRNA control). Thymosin fllo Gene Expression in RAR-a-Transfected Cells When RNA was analyzed for thymosin ill0 gene expression, an elevation in thymosin ill0 mRNA levels was observed in RAR-o:-transfected neuronal cells. Figure 4 shows the outcome of a kinetic study in normal (mock-transfected) and transiently transfected cells. Thymosin fllO mRNA was present a priori as a single, constitutively expressed, 0.6-Kb nucleotide species. The addition of retinoic acid (2 x 10-7M) induced a gradual accumulation of j6~0 mRNA, with a maximal induction being observed at 27 hr (fivefold stimulation compared to the zero-time point). A different pattern of thymosin fllo mRNA expression was observed in the RAR-~-transfected cells; levels of fla0 mRNA were induced even before the addition of retinoic acid (Fig. 4, top panel). This HOURS AFTER RETINOIC ACID (2xlO-7M) r"

0

1

6 27 0

1

6

27 mRNA

MOCK

RARd¢

TRANSFECTED

TRANSFECTED

mRNA

Fig. 4. Northern analysis of total RNA (10/tg per lane) isolated from mock-transfected and RAR-7transfected B104 cells treated for different periods of time with 2 × 10-7 M all-trans retinoic acid. The blot was probed with an A v r 2 / E c o R 1 restriction fragment corresponding to the human thymosin ~10 cDNA (Hall et aL, 1990). The upper panel represents thymosin ill0 mRNA levels. The middle panel is the level of fl-actin mRNA. The lower panel depicts the ethidium-stained gel. ETHIDIUM-STAINED GEL

52

Hall

increase was repeatable and not due to unequal R N A loadings as indicated by the beta-actin and 18S rRNA control probes or ethidium staining of gels (Fig. 4, middle and lower panels).

Enhancement of Thymosin ill0 Gene Expression by Retinoic Acid in RAR-a-Transfected Cells To characterize further the impact of overexpression of the R A R gene, further experiments were undertaken to examine in more detail, the possible RETINOIC ACID (M) '

~

~

?

?

~

~

~ ?

'

0 o o o o o o ~ o o

NORMAL

RARo(¢-TRANSFECTED

28S--

18S--

28S 18S

ETHIDIUM-STAINED GEL

Fig. 5. Northern analysis of the differential sensitivity of the thymosin ill0 gene expression to retinoic acid in normal and RAR-er-transfected rat B104 neuroblastoma cells. The upper panel represents thymosin fllo mRNA levels in cells exposed to various concentrations of retinoic acid for 3 days. The middle panel represents fl-actin control The lower panel is an ethidium bromide staining of the ribosomal RNA species indicating some minor discrepancies in gel loading.

Retinoids Modulate fllo Gene Expression in Neuroblastoma Cells

53

differential responsiveness of the thymosin ill0 gene to retinoic acid in mocktransfected and RAR-tr-transfected cells. Accordingly, mock-transfected and RAR-ol-transfected cells were cultured with a range of retinoic acid concentrations (10 -1° t o 10 - 6 M , final) for 3 days. When R N A was analyzed by Northern blots, it was apparent that transfection of the neuroblastoma cells with RAR-o~ increased the responsiveness of thymosin /31o m R N A accumulation to the stimulatory action of retinoic acid (Fig. 5). For example, levels of thymosin fllO mRNA were only slightly stimulated by retinoic acid at 10-8M in the mocktransfected cells; however, the same concentration of retinoic acid provoked a marked increase in thymosin ill0 mRNA in cells harboring the pSG5RARtr0 construct. This effect was exaggerated to some degree by the fact that retinoic acid (in RAR-o~-transfected cells) tended to suppress levels of beta-actin mRNA. Nevertheless, similar results were achieved when the thymosin /31o mRNA was normalized to 18S rRNA m R N A levels. The "trophic effect" of RAR-ol transfection upon thymosin fllO m R N A responsiveness to exogenous retinoic acid declined when transfected cells were passaged. This was presumably due to the loss of episomal expression following proliferation of nontransfected cells. Quantitation of thymosin ill0 m R N A by scanning densitometric analysis of Northern blots revealed that RAR-o~-transfected neuroblastoma cells displayed a responsiveness to retinoic acid which was superior to that exhibited in mocktransfected cells (Fig. 6). The dose-response curve for thymosin fllo mRNA responsiveness to retinoic acid in the RAR-cr-transfected cells was shifted to the left (Fig. 6).

t00

-

?

~--~ NORMAL _& ~RAMTR~

8o 1:2:::

~

60

a: 40 EI 9

~, 2o Fig. 6. Enhanced responsiveness of thymosin ill0 gene expression in response to retinoic acid in mocktransfected and RAR-tr transfected rat B104 neuroblastoma cells. Each value represents the percentage maximal densitometer response.

~'2_~___~

I

I

I

I

I

0 10-I0 10-9 10-8 I0.7 10-6 d/lrons RETINOIC ACID(M)

54

Hall DISCUSSION

High-performance liquid chromatography (HPLC) in combination with molecular cloning experiments was previously used to identify a retinoic acid responsive protein in a rat neuroblastoma cell line (Hall et al., 1991a). This protein was subsequently identified by molecular cloning and gas-phase Edman degradation, as thymosin rio, a small (