Expression of the rat aldolase B gene: A liver-specific proximal promoter and an intronic activator

Expression of the rat aldolase B gene: A liver-specific proximal promoter and an intronic activator

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 722-729 Vol. 176, No. 2, 1991 April 30, 1991 EXPRESSION OF THE RAT ALDOLASE B GENE : A LIV...

1MB Sizes 2 Downloads 38 Views

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 722-729

Vol. 176, No. 2, 1991 April 30, 1991

EXPRESSION OF THE RAT ALDOLASE B GENE : A LIVER-SPECIFIC PROXIMAL PROMOTER AND AN INTRONIC ACTIVATOR

Claudine GREGORI, Frederic GINOT, Jean-Fran(~ois DECAUX, Anne WEBER,Tsouria 8ERBAR, Axel KAHN and Anne-Lise PICHARD ICGM, INSERM U. 129 (Laboratoire de Recherches en G~netique et Pathologic Mol~culaires), CHU Cochin, 24 rue du Faubourg Saint-Jacques, 75014 PARIS, France Received

March

15,

1991

The nature and location of the cis-acting DNA sequences regulating expression of the rat aldolase B gene has been investigated. Two liver-specific DNAse I hypersensitive sites were detected, one located just upstream from the cap site, the second in the middle of the first, 4.8-kbp-long, intron. A fragment of 190 bp 5' to the cap site behaved as a tissue-specific but weak core promoter: it directed a detectable reporter gene expression in the Hep G2 cells and hepatocytes, but not in fibroblasts. The tissue-specific expression was stimulated at least 16 fold when constructs contained the entire first intron. The intronic activating sequences could be ascribed to an inner 2 kbp fragment in which the downstream liver-specific DNAse I hypersensitive site was located. ® 1991Academic R . . . . . I n c .

In vertebrates, three genes encode the different aldolase subunits which are associated into active tetramers (1). Aldolase A is ubiquitous and its gene is also strongly expressed from a tissuespecific promoter in adult skeletal muscle (2,3). Aldolase C is weakly expressed in fetal and actively proliferating malignant tissues and strongly expressed in brain (4,5). Aldolase B is specific to liver, kidney and enterocytes where its specific role seems to catalyze a key reaction of fructose metabolism, i.e. cleavage of fructose 1 phosphate. In vivo, the aldolase B gene is positively regulated by carbohydrates and insulin and partially inhibited by fasting and glucagon treatment (6). Aldolase B cDNAs and genes from different species have been isolated and studied by various groups (7-13). As expected from their different tissue specificity and regulatory properties, the promoters of the different aldolase genes share no homology, while the coding sequence is at least 70% conserved. In this paper, we describe the global organization of the aldolase B regulatory regions, namely a proximal tissue-specific promoter and the presence of an activator located in the middle of the large 4.8kbp-long 1st intron. In contrast to the L-pyruvate kinase gene (COGNET et al, personal communication) and the albumin gene (14), the activator is strongly efficient in transient transfection tests, stimulating activity of the promoter 16 to 60 fold. MATERIAL AND METHODS

Cell culture. Monolayers of human hepatoma (Hep G2) cells were maintained in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% (vol/vol) fetal calf serum, 1 p.M Ltriodothyronin,1 #M dexamethasone, 10 p.M insulin. Mouse 3T6 cells were maintained as monolayers in DMEM supplemented with 10% (vol/vol) fetal calf serum. Hepatocytes were isolated from adult rats 0006-291X/91 $1.50 Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

722

Vol. 176, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

fasted for 72 h, they were then maintained in chemically defined medium 199, supplemented with 10% (vol/vol) fetal calf serum, 10 mM lactate, 20 mM glucose, l pM L triodothyronin, I#M dexamethasone and 10-8 M epidermal growth factor (15). Cells were plated at a density of 3.5xl 06 per 78 cm2 dish. Transient transfection assays. Transfection was carried out by the calcium phosphate method (15). Calcium phosphate precipitates were removed 18 h after their addition. Cells were then washed twice with 5 ml physiological saline and fed with 10 ml of fresh culture medium. Hep G2 and 3T6 cells were harvested 40 h and hepatocytes in primary culture 68 h after initial exposure to DNA. Each dish received a total of 20 ~g DNA including 3 ~g of the pRSV luciferase standardization plasmid that was used to monitor variations in transfection efficiency. Pellets of 3.5x106 cells were resuspended in 100 #1 of 100 mM potassium buffer pH 7.8, 1 mM dithiothreitol, broken by three cycles of freezing and thawing, then centrifuged at 9800 g for 15 min at + 4°C. The supernatant was assayed for luciferase activity (17) and for CAT activity according to GORMAN et a1.(18). Transient expression vectors. Different restriction fragments of the aldolase B gene were ligated by conventional procedures into the promoterless CAT vector PeCAT. When possible, the fragments were directly inserted into cohesive sites of the vector. Otherwise, blunt ended fragments were either inserted into the Sma 1 site or ligated to linkers before insertion into cognate restriction sites (19) (Fig. 1). The series of mutants spanning all or parts of the first intron was constructed into the promoterless CAT vector VB 8 CAT. This VB 8 CAT vector was obtained by insertion of a 94 bp fragment, spanning the 86 last bp of the first intron and the 8 first bp of the second exon, i.e. just upstream from the CAT gene (Fig. 2). This fragment, synthesized by polymerase chain reaction, includes the full 3' acceptor splicing site of the first intron and stops before the AUG translation initiator of the second exon.

I

ALu(+33) Xba

I

I -2800

//

Sty (+44)

Eco R I I

AlUl

~J

-677

-192

+1

®

Standardized CAT activity (arbitrary units) Hepatocytes i

//

1.8 + 1.9

2808

Hep

G2

3T6

0.45 + 0.05

=0

3.7±0.7

0.9±0.1

=0

5.8±2.5

1.5±0.4

~0.1

1.3±0.8

0.5±0.1

0.2±0.1

+44

I -677

I U A I ,,,~

~

+44

-192

+33

Fiq 1 .Transient expression analysis of CAT constructs driven by 5' flankinq sequences of the aldolase Bgene A: Scheme of the 5' flanking region up to the 1st exon. The restrictionsites used for constructingthe plasmids are indicated.B: Different CAT constructs and the generated CAT activity in transfected cells. The results are expressed in arbitrary units of CAT/luciferaseactivity (see the method chapter). The means + or - 1 SD of at least three independent experiments, each of them in triplicate, are indicated. 723

BIOCHEMICAL A N D BIOPHYSICAL RESEARCH COMMUNICATIONS

V o l . 176, N o . 2, 1991

Exon 1

Exon 2 Hind l~t

Bgl ]] I

I

-2200

+1

A

[

'

El

+0.67kb

Hind ffl EcoRV II

Bgl 1] C,

~

I

+2.6kb

PCR amplified fragment

+4.Skb I I

Kpn l

/ //

\"/"

',

Sac l

~,bp "[ 8

®

CAT

bp

Standardized CAT activity (arbitrary units)

I',

1

H

El

,

°

H i A '

,

El

2200

Hepatocytes

Hep G2

3T6

24 ,'-7.6

39_+17

=0

Bgl ]3

[ ..............

-2200

I

H

]

EcoRV

[ .............

88±76

52±6

=0

1.5±2.2

<0.2

ND

4.7±1.7

0.6_+0.2

ND

1.3±0.8

0.5__+0.1

0.2±0.1

T ]

H

I -2200

[ ............ H i A ,

, .

f -2200

C

]

Bgl

[ ......... o. cAT ]

Fig 2. Transientexpression analysis of CAT constructs containing intronic sequences derived from the I st intron and driven bv 2.2 kb of 5' flankina seauences of the aldolase B aene A: Scheme of the 5' flanking region of the aidolase B gene up to the 2nd exon. The gene fragment represented by the bar beneath the 1st intron/2nd exon junction was amplified by PCR and ligated between Kpn 1/Sac 1 sites upstream from the CAT gene of the PeCAT plasmid to generate plasmid VB 8-CAT.B: Different CAT constructs and generated CAT activity in transfected cells. Same legend as in Fig. 1.

DNAse I analysis. The animals used were male Wistar rats, fasted for 72 h and refed for 24 h with a carbohydrate-tic h diet. Nuclei were prepared according to BURGOYNE et al. (20), as modified by RAMAIN et al. (21). Nuclei were recovered at a concentration of 1 mg/ml in 0.36 M sucrose, 15 mM Tris-HCI pH 7.5, 15 mM NaCI, 60 mM KCI, 15 mM 6-mercaptoethanol, 0.15 mM spermine, 0.5 mM spermidine. The suspensions were aliquoted in 1 ml fractions and 10 ~1 of 100 mM CaCI 2 were added. Each fraction was then digested with an increasing amount of DNAse I (from 1 to 10 p.g) for 20 min at room temperature. Digestion was stopped by adding 1 ml of the following solution: 50 mM Tris-HCI pH 7.5, 150 mM NaCI, 15 mM EDTA and 0.3% (weight/vol) sodium dodecyl sulfate. DNA was further purified by standard procedures. RESULTS Analysis of the aldolase B promoter region. Promoter activity of aldolase B was examined by transfection of the recombinant plasmids into hepatoma Hep G 2 cells, known to correctly express aldolase B (22), hepatocytes in primary culture that we demonstrated to be well transfectable and that conserved in vitro some important features of the in vivo liver cells (15,23) and into 3T6 fibroblasts, which do not express this gene. 724

Vol. 176, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

The shortest construction (0.2 aid B CAT) contains 192 bp 5' to the cap site, including the TATAA and CAAT consensus sequences, plus 44 bp of the untranslated t st exon, inserted into the promoterless CAT vector. This construct transfected into hepatocytes in primary culture generated a low but detectable CAT activity while, into mouse 3T6 cells, it failed to direct CAT synthesis above the background level for the assay (Fig. 1). This 0.2 aid B CAT plasmid was also weakly transcribed in the Hep G2 cells that have maintained several features of differentiated hepatocytes (22) (Fig. 1). The proximal promoter of the aldolase B gene, located in the 190 bp upstream from the cap site, seems therefore to behave, as judged by a transient transfection assay, as a weak but liver-specific promoter. Longer 5' flanking regions of the gene induced littte or no significant change upon the core promoter activity. At the most, a slight decrease was achieved by sequences 5' to nucleotide 677 (Figs 1 and 3). DNAse I analysis of the first intron. In order to localize potential cis-acting elements, we

performed an analysis of DNAse I hypersensitive sites. We started our analysis with a 15 kbp Sac I-Sac I fragment spanning the aldolase B gene from about 10 kbp upstream from the cap site to the second exon. The probe consisted of the 1.8 kbp-long Hind Ill-Sac I subfragment. Three weak but constant

0.2 ald B CflT

2.2 old B-ABC-CAT

~

~.

3T6

m

GJ

Hop 6 2

Xepatocytes

Fig 3. ExPression of the Aid B - CAT constructs in 3T6 fibroblasts, hepatoma Heo G2 cells and heDatocvtes in primary culture

All the experimentswere performed in triplicate,.butthose concerningexpressionot the controlplasmids are shown in single. Maps of the constructs investigated here are shown in Figs 1 and 2. In all the experiments shown on this picture, luciferase activity generated by the standard pRSV luciferase plasmid was in the same range. 50 ml of cell extracts were used for the CAT assay, undiluted for the aid B-CAT constructsand 16-fold diluted for the pSV2-CAT plasmid used as standard. 725

Vol. 176, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

hypersensitive sites (HS) were detected in liver nuclei, called HS I, II and III. HS I is located around the start site of transcription while HS II and III, 300 bp apart, are located in the middle of the 1st intron. HS II was the only site to be detected in spleen nuclei, which do not transcribe the aldolase B gene (data not shown). Authenticity and location of these sites were confirmed using a 5' probe (a 0.9 kbp Xba I-Hind III fragment) hybridizing at the 5' end of a Xba I-Xba I fragment spanning from position -235 bp with respect to the cap site to the end of the 1st intron. In this system, site I could not be detected. Again, sites II and III were visible using liver nuclei, but only site II when using spleen nuclei (Fig. 4). The very low

DNasel ~g/ml

0

I

233.5

44.$$

6

liver I

I -10

II

III

kb +4.9

I I

kb

kb

I ~1 exon probe

II

DNAasel ~|/ml

,4~,1 I 0

I

Xbal

,

2

33.544.55 liver

B' "



6

?

'ii"

I

3

S

6 7 1015N spleen

Xieal

I

probe

Fig 4. DNAse I hypersensitivity analysis of the aldolase B qene in spleen and liver nuclei Spleen and liver nuclei from a fasted, then refed rat were incubated with DNAse for 20 rain at room temperature, as described in the method chapter. DNA was purified, digested with Sac I (upper panel) or Xba I (lower panel), and analysed by southern blot. l; Liver-specific hypersensitive site located around the start site of transcription. II; Non-specific intronic hypersensitive site, observed in both liver and spleen nuclei. III;.Liver-specific, intronic hypersensitive site. 726

Vol. 176, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

activity of the aldolase B gene 5' promoter, the presence of a conserved large first intron in genes from different species and the existence of a liver-specific HS site in this intron prompted us to look for cisacting element(s)located in this region. Search for cis-acting elements in the first intron. Several aldolase B constructs containing parts or all of the 1st intron were transfected into different cell lines and into hepatocytes in primary culture. A construct containing only fragment A (i.e. the first 600 bp of the intron) displayed similar activity to the construct devoid of intronic sequences. These results indicated that splicing correctly occurred in the transfected cells and that splicing in itself did not interfere with the transient expression of the aldolase B constructs. In hepatocytes, the activity generated by constructs retaining the complete intron was 16-fold higher than that generated by equivalent intronless constructs. Fragment C (i.e the 3' 1.8 kbp-long Hind III-Bgl II fragment of the first intron) did not seem to be required for this activation. Rather, a construct including fragments A and B, but not fragment C, was expressed on average 60-fold more than the construct with only fragment A. The significance of this overactivation compared to the plasmid containing the entire intron is however uncertain because, for reasons not known, the results observed with the plasmid devoid of fragment C were very dispersed (Figs 2 and 3). The same activation by intronic fragment B was observed in Hep G2 cells. In these cells, however, the activity of the construct containing a mini intron consisting of fragment A was so low that it was impossible to quantify (Figs 2 and 3). In 3T6 fibroblasts, the expression of constructs containing the active intronic sequence remained very low and did not significantly differ from the background. We then investigated the effect of the intronic fragment B placed upstream from the 190 bp core promoter. In this non-intronic location, this sequence seemed to be almost inactive (data not shown). DISCUSSION Previous studies have shown that the 200 bp 5' to the cap site of the aldolase B gene were sufficient to direct liver-specific in vitro transcription (24). We now report that 190 bp of 5' flanking sequence are sufficient to ensure cell-specific expression in transient transfection, but at a surprisingly low level. This basal activity was not significantly affected by further upstream flanking sequences. DNAse I hypersensitivity analysis of the gene in aldolase B expressing (liver) and non-expressing (spleen) tissues disclosed three sites, one located around the cap site (I) and two in approximately the middle of the 1st intron (11and III); site III appeared to be tissue-specific. TSUTSUMI et al. (25) were the first to publish an analysis of the DNAse I hypersensitive sites in and around the aldolase B gene. Using Xho I digestion and a cDNA probe, they detected in liver nuclei two sites 2.6 and 0.3 kbp upstream from the cap site, one site in the 8th intron and one site downstream from the polyadenylation site. No site in the 1st intron was reported, but such HS sites would have been difficult to detect with the restriction enzyme and probe they used. In our own DNAse I analysis, site I could correspond to the -0.3 kb site reported by TSUTSUMI et al. (25) but we were unable to confirm the -2.6 kbp site which, in our hands, does not seem to be associated with cis-acting sequences detectable in transient expression. Consistently with the presence of a liver-specific DNAse I hypersensitive site in the first intron, we found that the inner 2 kbp-long Hind III-Eco RV fragment (fragment B) of the 1st intron contains a major activator element. The presence of the entire 1st intron in a CAT construct driven by 2.2 kbp of the 5' flanking region of the aldolase B gene resulted in a 16 fold increase in its transcription rate. The activation region could be localized to the inner fragment B because a construct devoid of 727

Vol. 176, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

fragment C was strongly expressed while a construct devoid of both fragments B and C was as weakly expressed as the intronless plasmids. The requirement for both a tissue-specific proximal promoter and a distal activating element to achieve a full stimulation of gene expression is not unique to the aldolase B gene. In fact, such organization has already been described for numerous hepatic genes (14, 26-29). Sometimes, the distal enhancer seems to be indispensable only in the context of the chromosomal structure, probably to open chromatin during development (30); in other cases, the enhancers are also required for gene stimulation in transient transfection tests, as demonstrated here for the aldolase B constructs. If the presence of a distal activating element is common in hepatic genes, the aldolase B gene seems to be the 1st example in which such an element proves to be located in an intron. However, intronic enhancers or activators have already been described in numerous nonhepatic genes, e.g. genes for immunoglobulin, adenosine deaminase, (31,32) and a number of muscle-specific genes, e.g. mouse a-2 type collagen, quail troponin I, muscle creatine kinase and drosophila 63 tubulin (33-36). In conclusion, the aldolase B gene needs a proximal, tissue-specific promoter and an intronic activator for full expression. We have currently investigating molecular dissection of these two entities as well as their interaction, in vitro by cell-free transcription, ex vivo by transfer in cultured cells and in vivo by creation of transgenic mice. REFERENCES 1.

2. 3. 4.

5. 6. 7. 8. 9.

10. 11. 12. 13. 14. 15. 16. 17 18. 19. ,20.

Penho~t, E., Kochman, M., Valentine, R. & Rutter, W.J. (1967) Biochemistry 6, 2940-2949. Mennecier, F., Daegelen, D., Schweighoffer, F., Levin, M. & Kahn, A.(1986) Biochem. Biophys Res. Com. 134, 1093-1100. Maire, P., Gautron, S., Hakim, V., Gregori, C., Mennecier, F. & Kahn, A. (1987) J. Mol. Biol. 197, 425-438. Schapira, F., Hatzfeld, A. & Weber, A. (1975)in Isozymes (Market, C.L., ed.), vol. 3, pp 9871003, Academic Press, New-York. Vibert, M., Henry~ J., Kahn, A. & Skala, H. (1989) Eur. J. Biochem.181,33-39. Munnich, A., Besmond, C., Darquy, S., Reach, G., Vaulont, S., Dreyfus, J.C. & Kahn, A. (1985) J. Clin. Invest. 75, 1045-1052. Paolella, G., Santamaria, R., Izzo, P., Costanzo, P. & Salvatore, F. (1984)Nucleic Acids Res. 12, 7401-7410. Rottman, W.H., Tolan, D.R. & Penho~t, E.E. (1984) Proc. Natl. Acad. Sci. USA 81,2738-2742. Simon, M.P., Besmond, C., Cottreau, D., Chaumet-Riffaud, P., Weber, A., Dreyfus, J.C., Marie, J. & Kahn, A. (1983)J. BM. Chem. 258, 14576-14584. Tsutsumi, K.I., Mukai, T., Tsutsumi, R., Mori, M., Daimon, M., Tanaka, T., Yatsuki, H., Hori, K. & Ishikawa, K. (1984) J. Biol. Chem. 259, 14572-14575. Tsutsumi, K.I., Mukai, T., Tsutsumi, R., Hidaka, S., Arai, Y., Hori, K. & Ishikawa, K. (1985) J. Mol. Biol. 181,153-160. Burgess, D. & Penho~t, E. (1985) J. Biol. Chem. 260, 4604-4614. Tolan, D. & Penho~t, E. (1986) Mol. Biol. Med. 3,245-264. Herbst, R.S., Boczko, E.M., Damell Jr., J.E. & Babiss, LE. (1990) Mol. Cell. Biol.10, 3896-3905. Ginot, R, Decaux, J.F., Cognet, M, Berbar, T, Levrat, F., Kahn, A. & Weber, A. (1989) Eur. J. Biochem. 180, 289-294. Graham, F.L. & Van tier Eb, A.J. (1973) Virology 52, 456-467. De Wet, J.R., Wood, K.V., De Luca, M, Helinski, D.R. & Subramani, S. (1987) Mot. Cell. Biol. 7, 725-737. Gorman, C., Moffat, L. & Howard, B. (1982) Mol. Cell. Biol. 2, 1044-t 051. Maniatis, T., Fdtsch, F. & Sambrook, J. (1982) Cold Spring Harbor Laboratory Press, Cold Spdng Harbor, N.Y. Burgoyne, L.A., Wagar, M.A. & Atkinson, MR. (1970)Biochem. Biophys. Res. Com. 39,254. 728

Vol. 176, No. 2, 1991 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Ramain, P., Bouronis, M., Dretzen, G., Richards, G., Sobkowiack, H. & Bellard, M (1986) Cell 45,545-553. Kelly, J.H. & Darlington, G.J. (1989) In Vitro Cell Dev. Biol. 25 ; 217-222. Decaux, J.F., Marcillat, O., Pichard, A.L., Henry, J. & Kahn, A. (1991) J. Biol. Chem. in press. Tsutsumi, K.I., Ito, K. & Ishikawa, K. (1989) Mol. Cell. Biol. 9, 4923-4931. Tsutsumi, K.I., Tsutsumi, R. & Ishikawa, K. (1987) J. Biochem. 102, 1013-1021. Herbst, R, Friedman, N, Darnell, E & Babiss, L (1989) Proc. Natl. Acad. Sci. USA 86, 1553-1557. Pinkert, C.A., Ornitz, D.M., Brinster, R.L. & Palmiter, R.D. (1987) Genes Dev. 1,268-276. Godbout, R., Ingram, R.S. & Tilghman, S.M. (1988) Mol. Cell. Biol. 8, 1169-1178. Watanabe, K., Saito, A. & Tamoki, T. (1987) J. Biol. Chem. 262, 4812-4818. Voss, S.D., Schlokat, V. & Gruss, P. (1986) Trends Biochem. Sci. 11,287-289. Gillies, S.D., Morrison, S.L., Oi, V.T. & Tonegawa, S. (1983) Cell 33, 717-728. Aronow, B., Lattier, D., Silbiger, R., Dusing, M, Hutton, J., Stock, J., McNeish, J., Potter, S., Witte, D. & Wiginton, D. (1989) Genes Dev. 3, 1384-1400. Bornstein, P.J., McKay, D.J., Liska, S., Apone, S. & Devarayalu, S.(1988) Mol. Cell. Biol. 8, 4851-4857. Konieczny, S.F. & Emerson Jr, C.P. (1987) Mol. Cell. Biol. 7, 3065-3075. Sternberg, E.A., Spizz, G., Perry, W.M., Vizard, D., Well, T. & Olson, E.N. (1988) Mol. Cell. Biol. 8, 2896-2909. Gasch, A., Hinz, U. & Renkawitz-Pohl, R. (1989) Proc. Natl. Acad. Sci. USA 86, 3215-3218.

729