ELSEVIER Molecular Brain Research 27 (1994) 315-319

MOLECULAR BRAIN

RESEARCH

Short communication

The 5' flanking region of the serotonin 2 receptor gene directs brain specific expression in transgenic animals

Miklos Toth a,U,,, Daming Ding a,c, Thomas Shenk a a Howard Hughes Medical Institute, Department of Molecular Biology, Princeton University, Princeton, NJ 08544-1014, USA

b Department of Pharmacology, Cornell University Medical College, New York, NY 10021, USA c International Department of Research and Development, Zhuhai S.E.Z. Lizhu Pharmaceutical (Group) Co. Ltd., Zhuhai, Guangdong,

People's Republic of China

Accepted 16 August 1994

A b s t r a c t

The neuron is the predominant cell type expressing the serotonin 2 (5-HT 2) receptor in the central nervous system. Transcriptional control elements involved in the restriction of 5-HT z receptor gene expression to neuronal cells and tissues were studied using both transgenic mice and cultured cells. Sequences extending from a site near the translational initiation codon to - 5.6 kb in the 5' flanking region of the murine receptor gene were found to be sufficient to target gene expression to the brain in transgenic animals. In transient transfection experiments a basal promoter was identified which was functional in both neuronal and nonneuronal cells. Upstream of the basal promoter two repressor domains were found within the 5' flanking sequence of the receptor gene. These sequences repressed gene activity in all cells except cells of neuronal origin, thus the repressor domains are the primary determinants to generate neuronal cell-specific transcription of the 5-HT 2 receptor gene.

Keywords: Serotonin 5-HT 2 receptor; Transcription; Gene regulation; Promoter; Repressor; Transgenic animal; Neuronal cell

Activators and repressors contribute to the cell type-specific regulation of transcription. Cell-specific expression can be achieved through positive regulatory mechanisms in which t rans-act ing factors activate the basal transcription machinery (reviewed in [21,30] or by negative regulation (reviewed in [16]). Gene activity can be repressed in non-expressing tissues through the action of a repressor that acts in some cells types and not in others [15,19,22,23,24,31] or by a repressor that functions in all cell types with subsequent reactivation in the appropriate cell by cell-specific activators [2,5,7,8,11,28]. The combinatorial action of positively and negatively acting transcriptional regulatory pro- teins and mechanisms is essential to provide complex regulation and flexibility in gene expression.

The serotonin 2 (5-HT 2) receptor gene is a useful model for studying transcriptional mechanisms and regulation since it exhibits a broad and complex ex- pression pattern. It is expressed in a variety of cell

* Corresponding author. Fax: (1) (212) 746-8835.

0169-328X/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0169-328X(94)00194-4

types, including neurons, glial cells, platelets, lympho- cytes and smooth muscle cells [13,14,26,27]. However, the 5-HT 2 receptor is not expressed at similar effi- ciency in all neurons and glial ceils. In the brain, the receptor is expressed in cortical layers II to VI, with the highest density of receptor in layer Va [20], where it is most likely associated with interneurons [29]. The 5-HT 2 receptor can also be detected in the cerebellar granula cells [32]. The range of expressing tissues un- doubtedly reflects the broad physiological role of the 5-HT 2 receptor [13,26,27].

We have recently isolated and sequenced the 5' flanking region of the mouse 5-HT z receptor gene, and identified a series of 11 clustered transcriptional initia- tion sites for the promoter that are utilized in brain tissue [5]. We also characterized the transcriptional regulatory functions of the 5' flanking region in the C6 glioma cell line, which expresses the receptor [1,5]. A repressor domain is present that inhibits transcription of the 5-HT 2 receptor gene in C6 glioma cells. A far upstream domain within the 5' flanking region relieves repression and reactivates transcription of the gene

316 M. Toth et al. /Molecular Brain Research 27 (1994) 315-319

A

B

D Lin e -5.6 -0.6 8412

5HT2R-5.6CAT I I I 87/5 -5.1 5HT2R CAT

5HT2R-5.1CAT I I I [ ] 43/7 .~ 103/10

'r.

FP T OB BS Ce Lu He Li Sp in Ki Mu

Cortex

3 -

e~

2 - eJ

e ~

O

Brain

I1_ C OB BS Ce Lu He

Cortex OB BS Ce Lu He Li Sp In Ki

Brain

Fig. 1. Tissue-specific expression of the 5-HT 2 receptor gene in the mouse. A: the 5' flanking region of the 5-HT 2 receptor gene directs brain-specific expression of a CAT reporter gene in transgenic animals. Different amounts of the 5' flanking region of the 5-HT 2 receptor gene were linked to the bacterial CAT gene [18] as described previously [5]. Numbers above the reporter construct diagrams indicate the distance in kb from the translational initiation codon of the 5-HT 2 receptor gene. Harvesting of fertilized eggs and microinjection of purified promoter-CAT fragments (1 /xg/ml) into male pronuclei were performed as described by Hogan et al. [12]. Injected zygotes were cultured overnight and two-cell embryos were implanted into pseudopregnant females. Southern blot analysis identified transgenic animals. Tissue extracts were prepared essentially as described by Wuenschell et al. [31], and CAT activity in tissue extracts was assayed as described [9]. Bars in the graph represent the radioactivity in acetylated chloramphenicol produced in assays containing tissue samples from transgenic animals: FP, frontal-parietal cortex; T, temporal cortex; OB, olfactory bulb; BS, brainstem; Ce, cerebellum; Lu, lung; He, heart; Li, liver; Sp, spleen; In, intestine; Ki, kidney; Mu, muscle. Tissues from non-transgenic animals showed < 100 cpm activities in enzymatic assays, and these background counts were subtracted from the values obtained with samples of transgenic animals. Assays were performed in triplicate on each tissue and the average is presented. B: 5-HT 2 receptor-specific RNA in mouse tissues. A ribonuclease protection experiment was carried out with a probe corresponding to the 5' nontranslated region of the RNA. RNA isolation from different regions of mouse brain and peripheral tissues was carried out using the guanidium thiocyanate-CsCl procedure [4]. An antisense RNA probe for the RNase protection assay was generated by in vitro transcription of SphI-linearized pRSR with T3 polymerase as described [5]. Radioactivity was quantified using a phosphorimage analyzer, and is reported as phosphorimager units (U). The insert shows part of the same gel after 24 h exposure with X-ray film. Tissue designations are as for A, except C represents cortex. The arrow marks the protected fragment and the band above corresponds to undigested probe.

Fig. 2. The 5' flanking region of the 5-HT 2 receptor gene directs preferentially neuronal cell-specific expression. Mouse Neuro-2a (ATCC CCL 131), mouse NB41A3 (ATCC CCL 147), human SK-N-MC (ATCC HTB 10), and human SK-N-SH (ATCC HTB 11) neuroblastoma cells were grown in Dulbecco-modified MEM supplemented with 10% fetal calf serum. Human U-87 MG (ATCC HTB 14) glioblastoma, rat C6 glioma (obtained from Beth Hoffman, NIH, Bethesda, MD), rat CREF fibroblast and human HeLa cells were grown in medium supplemented with 15% horse serum plus 2.5% fetal calf serum. Transfections utilized the calcium phosphate-DNA coprecipitation method [10]. Five /~g plasmid DNA per cell monolayer in 35 mm petri dishes was used for transfection. After overnight incubation cells were washed and fresh medium was added. After 2 days total incubation, cells were collected and CAT activity was determined according to the procedure of Gorman et al. [9] with minor modifications. Reactions received 100/~g of cell protein [determined by the Bradford method [3]] and contained 8 mM coenzyme A. After a 3-4 hr incubation, reaction products were separated on TLC silica plates, and bands corresponding to acetylated and nonacetylated [14C]chlo- ramphenicol were scraped and radioactivity was quantified in a scintillation counter. CAT activities are normalized to that of a control thymidine kinase-CAT construct [17]. The data shown report the mean + standard deviation of four independent experiments. Brain-specific transcription initiation sites [5] are indicated by an arrow.

M. Toth et al. /Molecular Brain Research 27 (1994) 315-319 317

specifically within C6 glioma cells but not within fi- broblasts or epithelial cells where the 5-HT 2 receptor is not expressed [5]. To test whether the 5' flanking sequence of the 5-HT z receptor was able to induce brain-specific expression of a test gene in vivo, trans- genic mice were produced. The animals carried CAT

reporter genes whose expression was directed by 5-HT 2 receptor 5' flanking sequences extending from - 0 . 6 kb (relative to the translational initiation codon at + 1) to either - 5 . 1 kb (5-HTzR-5.1CAT) or - 5 . 6 kb (5- HT2R-5.6CAT). Two independent lines of transgenic mice were generated for each construct, and three out

A

- 5.6 -5.1 I

5HT2R-5.6CAT [ 5HT2R-5.1CAT 5HT2R-4.2CAT 5HT2R-2.TCAT 5HT2R-2.3CAT

; 2 -1.1 -1.0 -0.9 -~ /

--- / - 4 . 2 -2.7-2.3 ATG

I I I I ",, / I I.

I I

5HT2R CAT

/

B

20- -

i

1 5 -

_

• ~ 1 0 -

~ _

~ 5 -

SK-N-MC SK-N-SH Neuro2a

,h& NB41A3 U87

5HT2R-5.6CAT 5HT2R-5.1CAT

C6 CREF HeLa

Neuroblastoma Glia Fibroblast Epithelial

C

a l

b- L)

oJ

Neuro2a NB41A3

N e u r o b l a s t o m a

5HT2R-4.2CAT 5HT2R-2.TCAT

l l l 5HT2R.2.3CAT

U87 C6 CREF

Glia Fibroblast

318 M. Tothet al. / Molecular Brain Research 27 (1994) 315-319

of four lines showed detectable expression of the CAT reporter (Fig. 1). The absolute level of expression var- ied from line to line, most likely reflecting the effects of sequences surrounding the integration sites [25]. The two lines carrying the transgene with flanking sequence extending to - 5 . 6 kb exhibited brain-specific CAT expression (Fig. 1A). However, they differed in the regional distribution of CAT activity. Line 87/5 showed the highest levels of transgene expression in the cere- bellum and in the temporal lobe of the cortex and lower activity in the frontal cortex, brainstem, and olfactory bulb. Line 84 /2 exhibited maximal CAT ac- tivity in the olfactory bulb, lower activity in the cerebel- lum, and no detectable activity in other brain tissues tested. Neither line expressed detectable levels of CAT activity in the non-neuronal tissues tested (lung, heart, liver, spleen, intestine, kidney and muscle). The shorter transgene construct with 5' flanking sequence extend- ing to -5 .1 kb also exhibited brain-specific expression. Line 43 /7 exhibited higher CAT levels in the olfactory bulb and brainstem and lower levels in the frontal cortex and cerebellum. In this particular line, expres- sion of the transgene was also detected in the lung (Fig. 1A, 'Lu'). This may reflect neuron-specific trans- gene expression in the peripheral nervous system. It is more likely, however, that the shorter transgene has a less specific expression pattern or the sequences at the chromosomal integration site fortuitously activated the transgene in the lung.

Because the regional distribution of CAT activity in transgenic mice differed from line to line, it was of interest to compare these patterns with the regional distribution of the 5-HT 2 receptor mRNA in the brain. Fig. 1B shows that the highest receptor RNA level was detected in the cortex. Olfactory bulb and brainstem also contained receptor mRNA but in a lower quantity. The 5-HT 2 receptor RNA level was negligible in the cerebellum and undetectable in the peripheral tissues, including lung. Since the distribution in the brain of the 5-HT 2 receptor mRNA and CAT activity was not identical, it is likely that the 5' flanking region of the receptor gene is not sufficient to direct authentic tis- sue-specific expression within the central nervous sys- tem. Nevertheless, the 5' flanking region extending from - 0 . 6 kb to - 5 . 6 kb clearly contained sequence elements that limited expression to the brain.

In order to correlate the tissue-specific expression observed in transgenic animals with cell type-specific- ity, we employed transient transfection assays of a variety of cell lines, including neuroblastoma, glial, fibroblast and epithelial cell lines. In order to correct for differences in the transfection efficiency in the different cell lines, parallel cultures were transfected with a plasmid containing the CAT gene under the control of the ubiquitously active thymidine kinase promoter of herpes simplex virus. The C6 glioma cell

was the only one of the cell lines studied that ex- pressed its endogenous 5-HT 2 receptor gene [5]. First, the expression of constructs 5-HT2R-5.6CAT and 5- HTzR-5.1CAT, which were previously used for gener- ating transgenic animals, was studied. Fig. 2B shows that both 5-HT2R-5.6CAT and 5-HT2R-5.1CAT were active in all four neuroblastoma cell lines but consider- ably less active in non-neuronal cell lines (U87 glial, fibroblast, and epithelial cell lines). The only exception was the high level of 5-HT2R-5.6CAT expression in C6 glioma cells, and this is due to glial cell-specific activat- ing elements located far upstream ( - 5 . 1 kb to - 5 . 6 kb) in the 5' flanking region [5]. 5-HT2R-5.1CAT, which lacked this far upstream sequence, exhibited a low level of CAT activity in C6 cells, similar to other non-neuronal cells. Since neuronal cells expressing no 5-HT 2 receptor supported the expression of the trans- fected receptor-CAT construct (5-HT2R-5.6CAT) , the expression of the exogenous gene was neuronal specific rather than 5-HT 2 receptor specific. We concluded, that the 5' flanking region of the 5-HT 2 receptor gene contains elements which allow the expression of this gene in neuronal cells. Presumably, elements upstream a n d / o r downstream of the cloned 5' flanking region of the 5-HT 2 receptor gene are responsible for restricting the receptor expression to 5-HT 2 neurons.

To identify the sequences responsible for neuronal cell specificity, different segments of the 5' flanking region of the 5-HT 2 receptor gene between -2 .3 kb and -5 .1 kb were linked to the CAT gene (Fig. 2A), and their activities were tested in two neuroblastoma and three non-neuronal cell lines. 5-HT2R-2.3CAT, containing the upstream sequence between - 0 . 6 kb and - 2 . 3 kb from the 5-HT 2 receptor gene, was ex- pressed equally well in all cell lines regardless of their origin (Fig. 2C). In contrast, the activity of construct 5-HT2R-5.1CAT, containing the flanking region be- tween - 0 . 6 kb and -5 .1 kb, exhibited neuroblastoma cell-specific expression when analyzed in the previous experiment (Fig. 2B). These results indicate that the sequence extending to - 2 . 3 kb in the 5' flanking region of the 5-HT 2 receptor gene contains a weak constitutive promoter, and the basal promoter activity is reduced about 4-5 fold in non-neuronal cells (C6, CREF) by sequences located between - 2 . 3 kb and - 5 . 1 kb. This region was further dissected by analysis of two additional constructs. Inclusion of the DNA region between - 2.3 kb and - 2.7 kb had an inhibitory effect in U87 glial cells only (5-HT2R-2.7CAT), while addition of the region located between - 2 . 7 kb and - 4 . 2 kb reduced CAT activity in all non-neuronal cells including U87 cells (5-HT2R-4.2CAT). Neither region inhibited the constitutive promoter activity in neurob- lastoma cells. The level of CAT activity produced by 5-HTzR-4.2CAT and 5-HTzR-5.1CAT were similar in all cells tested, indicating that the DNA sequence

M. Toth et a l . / Molecular Brain Research 27 (1994) 315-319 319

between -4 .2 kb and -5 .1 kb had no major effect on the cell type-specific expression of the 5-HT 2 receptor gene.

In sum, consistent with the results obtained using transgenic animals, transient transfection experiments indicated that the 5' flanking sequence from the 5-HT 2 receptor gene is capable of directing neuronal-specific gene expression. Cell specificity is achieved by the action of two repressor domains located between - 2.3 kb and -4 .2 kb that inhibited expression in all cells tested except those of neuronal origin. Presumably, non-neuronal cells contain factors that bind within these regions to inhibit transcription. Cells of neuronal origin either lack these repressors or express additional factors that modify their function.

We are grateful to Trish Robinson for helping to generate transgenic animals. M.T. and D.D. were As- sociates of the Howard Hughes Medical Institutes, and T.S. is an American Cancer Society Professor and an Investigator of the Howard Hughes Medical Institute.

[1] Ananth, U.S., Leli, U. and Hauser, G., Stimulation of phospho- inositide hydrolysis by serotonin in C6 glioma cells, J. Neu- rochem., 48 (1987) 253-261.

[2] Baniahmad, A., Muller, M., Steiner, C. and Renkawitz, R., Activity of two different silencer elements of the chicken lysozyme gene can be compensated by enhancer elements, EMBO J., 6 (1987) 2297-2302.

[3] Bradford, M.A., Rapid and sensitive method for the quantita- tion of microgram quantities of protein utilizing the principle of protein-dye binding, AnaL Biochem., 72 (1976) 248-254.

[4] Chirgwin, J.J., Przbyla, A.E., MacDonald, R.J. and Rutter, W.J., Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease, Biochemistry, 18 (1979) 5294-5299.

[5] Ding, D., Toth, M., Zhou, Y., Parks, C., Hoffman, B.J. and Shenk, T., Glial cell-specific expression of the serotonin 2 recep- tor gene: selective reactivation of a silenced promoter, MoL Brain Res., 20 (1993) 181-191.

[6] Dynan, W.S., Promoters for housekeeping genes, Trends Genet., 2 (1986) 196-197.

[7] Emerson, B.M., Nickol, J.M. and Fong, T.C., Erythroid-specific activation and derepression of the chick fl-globin promoter in vitro, Cell, 57 (1989) 1189-1200.

[8] Evans, T., Reitman, M. and Felsenfeld, G., An erythroid-specific DNA-binding factor recognizes a regulatory sequence common to all chicken globin gene, Proc. Natl. Acad. Sci. USA, 85 (1988) 5976-5980.

[9] Gorman, C.M., Moffat, L.F. and Howard, B.H., Recombinant genomes which express chloramphenicol acetyltransferase in mammalian cells, MoL Cell. Biol., 2 (1982) 1044-1051.

[10] Graham, F.L. and van der Eb, A.J., A new technique for the assay of infectivity of human adenovirus 5 DNA, Virology, 52 (1973) 456-467.

[l l] Herbst, R.S., Friedman, N., Darnell, J.E., Jr. and Babiss, L.E., Positive and negative regulatory elements in the mouse albumin enhancer, Proc. Natl. Acad. Sci. USA, 86 (1989) 1553-1557.

[12] Hogan, B., Constantini, F. and Lacy, E., Manipulating the Mouse Embryo, Cold Spring Harbor Press, Cold Spring Harbor, NY, 1986.

[13] Julius, D., Molecular biology of serotonin receptors, Annu. ReL,. Neurosci., 14, (1991) 335-360.

[14] Julius, D., Huang, K.N., Livelli, T.J., Axel, R. and Jessel, T.M., The 5HT2 receptor defines a family of structurally distinct but functionally conserved serotonin receptors, Proc. Natl. Acad. Sci. USA, 87 (1990) 928-932.

[15] Kraner, S.D., Chong, J.A., Tsay, H-J. and Mandel, G., Silencing the type II sodium channel gene: a model for neural-specific gene regulation, Neuron, 9 (1992) 37-44.

[16] Levine, M. and Manley, J.L.,Transcriptional repression of eu- karyotic promoters. Cell, 59 (1989) 405-408.

[17] Luckow, B. and Schiitz, G., CAT constructions with multiple unique restriction sites for the functional analysis of eukaryotic promoters and regulatory elements, Nucl. Acid Res., 15 (1987) 5490-5496.

[18] Maniatis, T., Frisch, E.F. and Sambrook, J., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982.

[19] Maue, R.A., Kraner, S.D., Goodman, R.H. and Mandel, G., Neuron-specific expression of the rat brain type II Na + chan- nels by nerve growth factor, Proc. Natl. Acad. Sci. USA, 85 (1990) 924-928.

[20] Mengod, G., Pompeiano, M., Martinez-Mir, M.I. and Palacios, J.M., Localization of the mRNA for the 5-HT 2 receptor by in situ hybridization histochemistry. Correlation with the distribu- tion of receptor sites, Brain Res., 524 (1990) 139-143.

[21] Mitchell, P.J., Wang, C. and Tjian, R., Positive and negative regulation of transcription in vitro: enhancer binding protein AP2 is inhibited by SV40 T antigen, Cell, 50 (1987) 847-861.

[22] Mori, N., Schoenherr, C., Vandenbergh, D.J. and Anderson, D.J., A common silencer element in the SCG10 and type lI Na + channel genes binds a factor present in non-neuronal but not in neuronal cells, Neuron, 9 (1992) 45-54.

[23] Mori, N., Stein, R., Sigmund, O. and Anderson, D.J., A cell type-preferred silencer element that controls the neural-specific expression of the SCG10 gene, Neuron, 4 (1990) 583-594.

[24] Muglia, L., and Rothman-Denes, L.B. Cell type-specific nega- tive regulatory element in the control region of the rat alpha- fetoprotein gene, Proc. Natl. Acad. Sci. USA 83 (1986) 7653- 7657.

[25] Palmiter, R.D. and Brinster, R.L, Germ-line transformation of mice, Annu. Rev. Genet. 20 (1986) 465-499.

[26] Peroutka, S.J., 5-Hydroxytryptamine receptor subtypes, Annu. Rec. Neurosci. 11 (1988) 45-60.

[27] Peroutka, S.J., 5-hydroxytryptamine receptor subtypes: molecu- lar, biochemical and physiological characterization, TINS, 11 (1988) 496-500.

[28] Seal, S.N., Davis, D.L. and Burch, J.B.E., Mutational studies reveal a complex set of positive and negative control elements within the chicken vitellogenin II promoter, Mol. Cell. Biol., 11 (1991) 2704-2717.

[29] Sheldon, P.W. and Aghajanian, G.K., Serotonin (5-HT) induces IPSPs in pyramidal layer cells of rat piriform cortex: evidence for the involvement of a 5-HT2-activated interneuron, Brain Res., 506 (1990) 62-69.

[30] Struhl, K., Mechanism for diversity in gene expression patterns, Neuron, 7 (1991) 177-181.

[31] Wuenschell, C.W., Mori, N. and Anderson, D.J., Analysis of SCG10 gene expression in transgenie mice reveals that neural specificity is achieved through selective derepression, Neuron, 4 (1990) 595-602.

[32] Xu, J. and Chuang, D.-M., Muscarinic acetylcholine receptor- mediated phosphoinositide turnover in cultured cerebellar gran- ule cells: Desensitization by receptor agonists, J. PharmacoL Exp. Ther., 242 (1987) 238-244.