Vol.
174,
January
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS Pages
31, 1991
TRANSACTIVATION
OF THE HUMAN INSULIN CAAT/ENHANCER BINDING Catherine
McKeon
RECEPTOR PROTEIN
721-728
GENE BY THE
and Thang Pham
Diabetes Branch, National Institute of Diabetes and Kidney and Digestive Diseases, National Institutes of Health, Bethesda, MD 20892 Received
November
5, 1990
Within human insulin receptor gene there are three consensus binding sites for the CAAT/enhancer binding protein (C/EBP). Two sites are located in the 5 flanking region and the other is in the first intron. We have studied the ability of these sequences to be regulated by C/EBP. A eukaryotic expression vector containing these sequences can be transactivated in a dose-dependent manner by a C/EBP expression vector when co-transfected into NIH-3T3 cells. In addition, double stranded oligonucleotides corresponding to two of these sequences can bind C/EBP in a gel retardation assay. These two oligonucleotides can compete with each other to bind C/EBP. These findings suggest that this transcription factor may play a role in the regulation of insulin receptor gene expression in viva. cc1991Academic Press,1°C.
The insulin receptor mediates the action of insulin in its target cells presumably through the activation of its tyrosine kinase. Although most cells contain insulin receptors, there are increased numbers of insulin receptors in major target tissues for insulin action such as liver and adipocytes.
In order to elucidate the
mechanism of tissue-specific regulation of insulin receptor gene expression, we and others have cloned and analysed the human insulin receptor promoter (l-5). The region immediately upstream of the translation initiation site contains a promoter characteristic of “housekeeping” genes in that the sequence is very GC-rich and there are multiple transcriptional start sites (3-5). The flanking sequences have not previously been shown to contain binding sites for any known tissue-specific regulators. The gene is divided into 22 exons and spans over 150 kb of DNA (2). The first exon which encodes the signal peptide and the first seven amino acids of the alpha-subunit, is located in a 3 kb region of unique DNA which is embedded in Alu repeat sequences. The first intron is approximately 25 kb in length containing multiple Alu repeats (4,5). We now report that the 5’ flanking region and the unique domain at the 5’ end of the first intron of the insulin receptor gene contain consensus binding sites for the transcription regulatory factor, C/EBP. These sequences can regulate insulin receptor gene expression in vitro. A plasmid containing these sequences can be transactivated by an expression vector for the DNA binding protein C/EBP. This DNA
Vol.
174,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
binding protein has been shown to have a tissue-specific distribution and has been implicated in the regulation of both adipocyte-specific (6-8) and liver-specfic genes (9,101. MATERIALS
AND METHODS
Expression Vectors. The expression vectors HIR575, HIR535 and HIR1454 contain varying lengths of 5’ flanking region of the human insulin receptor gene (HIR) cloned in front of the chloramphenicol acetyltransferase gene (CAT) in the vector, pSVo-CAT. These vectors have been previously designated as HIR94, HIR81, and HIR170, respectively (5). In HIRINl and HIRIN2 portions of the first intron containing the C/EBP binding site shown in Fig. 1 have been inserted into the Barn HI site of HIR535. HIRINl contains a region from Nae I (+15) to Hind III (+3900) and HIRIN2 contains the region from Nae I (+15) to Sac I (+1703). The fragments were isolated from a genomic recombinant phage and Barn HI linkers were added (5). In the two intron containing vectors, the intron segments are oriented in opposite directions as shown in Figure 2A. Transfection. Cells were transfected using the calcium- phosphate technique. Cotransfection experiments were done using the CfEBP expression vector, pMSV-C/EBP-wt, generously provided by Steve McKnight and Alan Friedman, Carnegie Institute, Baltimore, MD (9). Each plate was seeded with 5 X 105 mouse NIH-3T3 cells. A total of 20 pg of DNA was co-precipated; this consisted of 2 ug of the appropriate expression vector, varying amounts of C/EBP vector as indicated and the balance was made up of pSV2-C, a pSV2-CAT vector with a deletion of the CAT gene provided by Ira Pastan, NCI, NIH (11). CAT assays were carried out according to Gorman (12). Conversion to acetylated 14Cchloramphenicol was quantitated using a escanner (Ambis Systems, Inc., San Diego, CA) Gel Retardation Studies. A set of complementary oligonucleotides 20 nucleotides long were synthesized to two of the C/EBP binding site containing sequences shown in Fig. 1 on a Gene Assembler (Du Pont, Wilmington, DE). The oligonucleotides for the HIR C/EBP sites were labeled by phoshorylation with (r-32P) ATP and incubated with nuclear extracts. The nuclear extracts were prepared from 3T3-Ll cells and 3T3-Ll cells following transfection with the C/EBP expression vector using a modification of the procedure of Dignam et al. (13,14). 10,000 cpms of labeled fragment was incubated with 20 pg of protein in a buffer containing 12mM HEPES pH 7.9, 12 mM KCI, 0.6 mM MgCI2, 1.2 mM DTT, 50 mM NaCI, 2.5 ug poly dl/dC and 10% glycerol at room temperature for 30 min. Competition was achieved by pre-incubating for 10 minutes with a 200-fold excess of an unlabeled competitor. The oligonucleotide containing the Ap3 binding site was purchased from Stratagene (La Jolla, CA). The protein-DNA complexes were separated on a 5% polyacrylamide gel (6O:i) in 0.5 XTBE (45mM Tris-Borate pH 7.8, 45 mM boric acid, and 1 mM EDTA) at 120 volts for 3 hours. The gels were then fixed, dried and autoradiographed. RESULTS Recent studies have demonstrated that transcription of the DNA binding protein C/EBP increases during adipocyte differentiation(6,7). This protein has been shown
to regulate several
adipocyte-specific
722
genes including
the insulin
Vol.
174,
C/EBP
No.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
2, 1991
CONSENSUS
SEQUENCE
r-lTGNN
GTAAG C
I I T
+343 HIR
INTRON
SEQUENCE
T +324
CGCACTGGG
GCAAGGTTCCT I
I
-1316 HIR
PROMOTER
SEQUENCE
-1297 TCTCCTGCA
GTAAGGTAGGT
-188 ROUS
SARCOMA
VIRUS
SEQUENCE
-207
ACCATGTT
GCAAGACTACAA
-267 GLUCOSE
TRANSPORTER
SEQUENCE
-240 ATTCTTTCA
u
FIGURE 1. ClEBP consensus sequences are found in the insulin receptor gene. The sequence of one strand of the oligonucleotides used in the gel retardation experiments is shown above 5’ to S’.The antisense sequence is shown for the HIR intron C/EBP site and the RSV C/EBP site and the sense sequence is shown for one of the HIR promoter sites and the GLUT4 C/EBP site. The consensus C/EBP site as determined by Ryden and Beeman (18) is boxed. The GLUT4 sequence has a consensus C/EBP binding site on both strands. The HIR promoter contains an additional site, -1429 TTGTGTAAT -1437. responsive
glucose
transporter
(GLUT4)(15),
stearoyl-CoA
desaturase
1
(SCDl)(7) expression
and aP2, a fatty acid binding protein (7,8). Since insulin receptor gene increases during adipocyte differentiation, we looked for a possible role
for C/EBP in insulin receptor gene regulation (16,17). Inspection of the DNA sequence surrounding the insulin receptor promoter revealed the presence of three consensus
UEBP
region between
binding sites (Fig. 1)(18). nucleotides
1429 in the opposite
One site is found in the 5’ flanking
-1311 and - 1303, another
orientation,
opposite orientation between translation initiation site). 4 expression
nucleotides
vectors
the marker
gene, chloramphenicol
diagrammed
in Fig. 2A. HIM75
enhancer sequences containing
element
+338 to +330 (nucleotide
Fragments expression
regulatory
that fused portions acetyltransferase contains
+l is the
sequences,
of the insulin receptor gene to (CAT).
These plasmids
the the insulin receptor
have been previously
we
defined
(5).
promoter
HIRl454
are and an
contains
further upstream including two putative C/EBP sites and a region 4 putative Spl binding sites. This construction exhibits an approximately
50% increase truncated
which
-1437 to -
and the other is found in the first intron in the
In order to study these and other potential constructed
is found between
in activity when compared
promoter
that expresses
to HIM75
(5).
HIR 535 contains
CAT at a level that is equivalent
a
to HIR575 (5).
of intron A containing the putative C/EBP binding site were inserted into vector HIR535 to yield HlRlNl and HIRIN2. HIRINl contains sequences
from +15 to approximately +1703. However
+3900 and HIRIN2
since the intron fragments
contains
sequences
are located in opposite
these two vectors, the C/EBP site is located at different distances as indicated in Fig. 2A.
723
from +15 to orientations
in
from the promoter
Vol.
174,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
B
COMMUNICATIONS
Transactivation
HIR575
HIR1454
by C/EBP
HlRlNl
HIR Epressbn
HIRINZ Vector
FIGURE 2. Insulin receptor sequences 8re tramactivated by C/EBP. A) Four vectors are diagrammed. Each vector contains the insulin receptor promoter fused to the chloramphenicol acetyltransferase gene. The nucleotide numbers indicate the 5’ end of the promoter. HlRlNl and HIRIN2 also contain portions of exon 1 indicated by the dark box and portions of intron A as indicated by the nucleotide numbers. Nucleotides are numbered consecutively from the translation initiation site. The location of the C/EBP consensus sequence is indicated by the triangle. In HIP1454 the triangle marks the location of two C/EBP sites. Since the intron fragment is in opposite orientation in these two constructions, the distances of the C/EBP site from the promoter is indicated for both directions. B) Each of these insulin receptor promoter vectors was co-transfected into NIH-3T3 cells with either a defective expression vector or the C/EBP expresion vector. CAT activity was assayed after 48 hrs. and quantitated by P-scanning. The bar graph shows the ratio between the CAT activity after co-transfection with the ClEBP vector to the CAT activity after co-transfection with the defective expression vector. These values are the mean of 3 separate experiments.
To investigate if CYEBP can regulate the expression of the insulin receptor gene, we co-transfected the C/EBP expression vector and each of the insulin receptor vectors into NIH-3T3 cells. After 48 hours, CAT activity was determined and the results are graphed in figure 2B. Co-transfection with C/EBP had no effect on the minimal promoter contained in HIR575. HIR1454 which extends the promoter to include the C/EBP sites which are upstream, showed a e-fold increase with C/EBP. Although both HlRlNl and HIRIN2 contain the portion of the intron with the C/EBP site, they are transactivated to different extents by C/EBP. HIRINl increases its CAT expression
2.3-fold
whereas
HIRIN2
increases
its CAT expression
6- to 8-fold
when
co-transfected with C/EBP. In HIRIN2 the intron fragment is truncated and inverted which moves the C/EBP site closer to the promoter than in HIRINl (Fig. 2A). In both of these constructions, the C/EBP site is located further away from the promoter than its natural position 750 bp downstream. The induction of CAT expression from 724
Vol.
174,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Transacfivation
by UEBP
0 0 0
4 P9
8
1-19
18
12Pcl
w
10
20
ug DNA
ClEBP FIGURE 3. ClEBP transactivates in a dose-dependent manner. 2 pg ot HIRIN2 was co-precipitated with increasing amounts of the C/EBP expression vector as indicated. A defective expression vector was added so that the DNA per transfection always totalled 20 pg. CAT activity was determined after 48 hours. The CAT assay is shown on the left and the percent conversion as determined by pscanning is graphed on the right.
HIRIN2 by ClEBP demonstrates dose-dependence (FIG. 3). The effect was first observed with 8 pg of C/EBP expression vector DNA and increased progressively with increased amounts of the C/EBP expression vector To investigate whether these potential C/EBP sites actually can bind C/EBP, we used a gel retardation assay. Synthetic oligonucleotides were synthesized corresponding to the sequences of two of the C/EBP sites (Fig. 1). These were labeled and incubated with 20 pg of protein from two nuclear extracts; a 3T3-Ll fibroblast extract and an extract from 3T3-Ll cells transfected with the C/EBP expression vector (Fig. 4). The undifferentiated 3T3-Ll cells have been shown to express very low endogenous levels of C/EBP (6,7). Both oligonucleotides showed increased binding with the nuclear extracts from the cells which had been transfected with C/EBP (FIG. 4, Lanes C and K). The binding of C/EBP to the oligonucleotides corresponding to the intron sequence can be competed by the oligonucleotides to the binding site in the insulin receptor promoter (Fig. 4, Lane E). Oligonucleotides corresponding to two previously characterized C/EBP binding sites, the GLUT4 binding site and one of the binding sites in the Rous sarcoma virus long terminal repeats (RSV LTR) were both able to compete for the protein as well (Fig. 4, Lanes F and G). An oligonucleotide to the DNA binding site for Ap3 competes 725
Vol.
174,
No.
BIOCHEMICAL
2, 1991
INTRON
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
QEBP SEQUENCE I
ABCDEFGH
JK
-
DNAICIEBP complex
FIGURE 4. The ClEBP sequences located in the insulin receptor gene can bind CYEBP. The oligonucleotide for the intron C/EBP sequence and one of the promoter C/EBP sequences were labeled with 32P and incubated with the following nuclear extracts. Lanes A & I) no extract, Lanes B & J) undifferentiated 3T3-Ll extract, Lanes C-H & K) 3T3-Ll cells transfected with the C/EBP expression vector extract. In lanes D-H, the labeled oligonucleotide was competed with cold oligonucleotides to D) C/EBP sequence from the intron, E) C/EBP sequence from the promoter, F) C/EBP sequence from GLUT4, G) C/EBP site from RSV-LTR , H) Ap3 site oligonucleotide.
slightly
for the binding
protein
(Fig. 4, Lane
H).
These
studies
demonstrate
that these
DNA sequences can bind authentic C/EBP. DISCUSSION In the 5’-end of the insulin receptor gene, there are 3 consensus binding sites for C/EBP. Expression vectors containing these sites can be transactivated to varying degrees by a C/EBP expression vector. The vector HIRIN2 showed the largest increase with an 8-fold stimulation by C/EBP. Other genes which are induced during adipocyte differentiation including GLUT4 and aP2, have been shown to be transactivated by C/EBP (15,8). In these genes the C/EBP site is located in the 5 flanking region near the start site of transcription and inductions up to 80-fold have been reported (15). The insulin receptor gene is the first gene to be reported where a WEBP site is located in an intron. It is interesting to speculate that the induction of the endogenous insulin receptor gene is only lo-fold because each of the C/EBP sites are located at a distance from the start site of transcription. In our expression vector constructions, the intron fragment and therefore the C/EBP site are not in their natural location. This probably affects their ability to respond to C/EBP. A positional 726
Vol.
174,
No.
2, 1991
BIOCHEMICAL
AND
effect could explain why the same sequence is deleted and inverted as seen in HIRIN2. the induction
should be drawn
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
has a greater effect when the fragment No conclusions
about the magnitude
until the intron can be tested in its physiologic
of
location
within the gene. C/EBP has also been implicated in the induction of several liver-specific genes including albumin (9) and Factor IX (10). It is possible that this protein plays a role in the elevated studies
expression
of insulin receptor
show a similar transactivation
hepatoma
cell line, HepG2.
endogenous
of HIR1454
mRNA found in liver. and HIRIN2
This cell line has also been shown
levels of the CiEBP transcription
expression
Preliminary in a human
to contain
low
factor.
The fact that C/EBP has been shown to be involved in the induction of both liver-specific and adipocyte-specific genes and the finding that it can transactivate the insulin receptor promoter in vitro make it a good candidate for an in vivo regulator of insulin gene expression. Further studies will be required to identify in which tissues
and at what points in development
this form of regulation
may be
operating. ACKNOWLEDGMENTS: We are grateful to Charles Roberts, Jorge Alemany, Gillian Walker and Hui Chen for their helpful suggestions, Steven McKnight for giving us the C/EBP expression vector, Ira Pastan for the vector pSV2-C, and Simeon I. Taylor for his guidance and continued support for this work. T. Pham was supported by Grant #188739 from JDF International. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Araki, E., Shimada, F., Uzawa, H., Mori, M. and Ebina, Y. (1987) J. Biol. Chem. 262, 16186-16191. Seino, S., Seino, M., Nishi, S. and Bell, G. (1989) Proc. Natl. Acad. Sci. USA 86, 114-118. Mamula, P.W., Wong, K., Maddux, B.A., McDonald, A.R. and Goldfine, I.D. (1988) Diabetes 37, 1241-l 246. Tewari, D.S., Cook, D.M. and Taub, R. (1989) J. Biol. Chem. 264, 1623816245. McKeon, C., Moncada, V., Pham , T., Salvatore , P., Kadowaki, T., Accili, D. and Taylor, S.I. (1990) Mol. Endocrinol. 4, 647-656. Birkenmeier, E.H., Gwynn, B., Howard, S., Jerry, J., Gordon, J.L., Landschulz, W.H. and McKnight, S.L. (1989) Genes Dev. 3, 1146-l 156. Christy, R.J., Yang, V.W., Ntambi, J., M., Geiman, D.E., Landschulz, W.H., Friedman, A.D., Nakabeppu, Y., Kelly, T.J. and Lane, M.D. (1989) Genes Dev. 3, 1323-l 335. Herrera, R., Ro, H.S., Robinson, G.S., Xanthopoulos, K.G. and Spiegelman, B.M. (1989) Mol. Cell. Biol. 9, 5331-5339. Friedman, A.D., Landschulz, W.H. and McKnight, S.L. (1989) Genes Dev. 3, 1314-1322. Crossfey, M. and Brownlee, G.G. (1990) Nature 345, 444-446. Kageyama, R. and Pastan, I. (1989) Cell 59, 815-825. Gorman, C.M., Moffat, L.F. and Howard, B.H. (1982) Mol. Cell. Biol. 2, 10441051. Dignam, J.D., Lebovitz, R.M. and Roeder, R.G. (1983) Nucl. Acids Res. 11, 1475- 1489. 127
Vol.
174,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
14. Wildeman, A.G., Sassone-Corsi, P., Grundstrom, T., Zenke, M. and Chambon, P. (1984) EMBO J. 3, 3129-3133. 15. Kaestner, K.H., Christy, R.J. and Lane, M.D. (1990) Proc. Natl. Acad. Sci. USA 87, 251-255. 16. Ullrich, A., Bell, J.R., Chen, E.Y., Herrera, Ft., Petruuelli, L.M., Dull, T.J., Gray, A., Coussens, L., Liao, Y.-C., Tsubokawa, M., Mason, A., Seeburg, P.H., Grunfeld, C., Rosen, O.M. and Ramachandran, J. (1985) Nature 313, 756 -761. 17. Sibley, E., Kastelic, T., Kelly, T.J. and Lane, M.D. (1989) Proc. Natl. Acad. Sci. USA 86, 9732-9736. 18. Ryden, T.A. and Beemon, K. (1989) Mol. Cell. Biol. 9, 1155 -1164.
728