DNA AND CELL BIOLOGY Volume 9, Number 10, 1990 Mary Ann Liebert, Inc., Publishers Pp. 777-781
LABORATORY METHODS
Simplified Method for the Preparation of Transcriptionally Active Liver Nuclear Extracts A
MASAHIRA HATTORI,* ANTONIO TUGORES,* LINDA VELOZ,* MICHAEL and DAVID A. BRENNER*
KARIN.t
ABSTRACT We have developed a simplified method for the preparation of liver nuclear extracts to study gene regulation and protein-DNA interactions. This protocol uses conventional laboratory equipment and standard reagents. The liver tissue is homogenized in a low-salt solution at physiological molarity with subsequent adjustment of the molarity and purification of nuclei by density sedimentation. The nuclear extracts are transcriptionally active in a validated cell-free transcription assay and contain functional DNA-binding proteins. This protocol results in the rapid preparation of highly reproducible and active liver nuclear extracts.
require the construction of new equipment. Here we describe a simplified method for the preparation of nuclear extracts from rat liver. The quality of these extracts was tested by in vitro transcription and mobility-shift assays. They are highly active in transcription and show appropriate tissue specificity.
INTRODUCTION Of genes ÍS the interaction of irons-acting DNA-binding proteins with specific as-acting elements in the DNA (Gorski et ai, 1986; Scheidereit et ai, 1987; Karin et ai, 1990). The first step in the biochemical analysis of this complex system requires the preparation of nuclear extracts from the tissue or cell type of interest. These extracts are then used for in vitro transcription and the analysis of DNA-protein binding. Most studies in this area utilize immortal tissue culture cells as a source of nuclear extracts that are relatively simple to prepare (Ohlsson and Edlund, 1986; Bodner et ai, 1987; Scheidereit et ai, 1987). However, transformed cell lines may not reflect accurately the state of differentiation seen in vivo. In addition, expression of certain tissue specific genes is either highly reduced or completely extinguished in cell lines compared to normal tissue. Preparation of nuclear extracts from the primary tissue, although advantageous, is far more difficult because of the presence of organ capsules, vascular structures, interstitial connective tissue, extracellular proteases, and cellular heterogeneity. To circumvent these problems, elaborate protocols have been designed for the preparation of transcriptionally active nuclear extracts, some of which
mediated by TISSUE
AND CELL-TYPE-SPECIFIC EXPRESSION
Departments of 'Medicine and 92093.
tpharmacology,
METHODS
Preparation of nuclear
extracts
All solutions, tubes, and centrifuges were maintained at 0-4°C. For each nuclear extract preparation, two male Fisher rats (150-250 grams) were anesthetized and killed by exsanguination. The preparation of rat liver nuclear extracts was based on previous methods that produced functional DNA-binding proteins (Graves et ai, 1986; Johnson et ai, 1987) or transcriptionally active extracts (Gorski et ai, 1986; Lichtsteiner and Schibier, 1989). Hepatectomies were performed, and the livers were cut with scissors into approximately 3-mm cubes in 10 ml of homogenization buffer (0.3 M sucrose, 10 mM HEPES pH 7.6, 0.74 mM spermidine, 0.15 mM spermine, 0.1 mM EDTA, 0.1 mM EGTA, 10 mM KC1, 1 mM DTT, 0.5 mM PMSF, and 2 fig/m\ each of aprotinin, leupeptin, and bestatin). Excess
Center for Molecular Genetics,
777
University of California, San Diego,
La
Jolla, CA
778 tissue and blood were removed. Half of the minced tissue was added to 25 ml of homogenization buffer in a 55-ml Potter homogenizer and was homogenized with two or three strokes by a motor-driven Teflon pestle. The homogenate was transferred to a precooled 250-ml cylinder by filtering through a cheese cloth to remove debris. The remaining half of the tissue was homogenized following the same procedure. The homogenized tissue (50 ml in solution) was then mixed with 100 ml of cushion buffer (identical to homogenization buffer except that the concentration of sucrose was 2.2 M), resulting in a final sucrose concentration of 1.57 M. The homogenate was laid over 10 ml of cushion buffer in 38-ml polyallomer ultracentrifuge tubes. The tubes were spun at 24,000 rpm for 50 min at FC in an SW28 ultracentrifuge rotor. The supernatant, including an upper layer of lipids and intact cells, was removed and the tube was inverted for 10 min in ice to drain remaining buffer from the nuclear pellet. At this point, the resulting hepatic nuclei may be flashfrozen in liquid nitrogen after resuspending the six nuclear pellets in 20 ml of nuclear storage buffer (25% glycerol, 50 mM HEPES pH 7.6, 3 mM MgCl2, 0.1 mA/ EDTA, 1.0 mM DTT, 0.1 mM PMSF). Subsequently, a volume of 5 ml of frozen nuclei was reconstituted by the addition of 25 ml of storage reconstitution buffer (12.5% glycerol, 2 mM HEPES pH 7.6, 116 mM KC1, 3 mM MgCl2, 0.1 mM EDTA, 1 mM DTT, 0.1 mM PMSF, 1.0 mM NaMo04, 2 /¿g/ml aprotinin, 2 /¿g/ml leupeptin, 2 /¿g/ml bestatin). The nuclei were resuspended with one stroke of a B pestle in a 40-ml Dounce homogenizer (Wheaton) and lysed as described below. Alternatively, each fresh nuclear pellet was resuspended in 5 ml of nuclear lysis buffer (10% glycerol, 10 mM HEPES pH 7.6, 100 mM KC1, 3 mM MgCL, 0.1 mM EDTA, 1.0 mM DTT, 0.1 ¡iM PMSF, and 2 /tg/ml each of aprotinin, leupeptin, and bestatin). The nuclei were resuspended by one stroke of a B pestle of a 40-ml Dounce homogenizer and brought to a final volume of 40 ml with nuclear lysis buffer; 20 ml of this nuclear suspension was transferred to each 30-ml Oakridge tube. Next 2 ml of 4 M ammonium sulfate pH 7.9 was added to each tube, mixed well, and incubated for 30-60 min with gentle agitation on a rocker platform at 4°C. The tubes were spun at 35,000 rpm for 60 min at FC in a 55.2 Ti ultracentrifuge rotor. The supernatant was measured and transferred to clean Oakridge tubes avoiding the chromatin pellets. Then, 0.33 mg of finely powdered ammonium sulfate per milliliter of supernatant was added to each tube. After the ammonium sulfate was dissolved, the tubes continued incubating without shaking at 0°C for 15 min. The tubes were spun at 35,000 rpm for 20 min at FC in a Ti 55.2 ultracentrifuge rotor. The supernatant was discarded and the tubes were inverted for 10 min on ice. The resulting nuclear protein pellet was resuspended in 500 ¡A of nuclear dialysis buffer (20% glycerol, 20 mM HEPES pH 7.6, 0.1 M KC1, 0.2 mM EDTA, 2 mM DTT, 0.1 mM PMSF, 1 mM NaMoO,, and 2 /ig/ml each of aprotinin, leupeptin, and bestatin). This pellet can be kept on ice overnight. Each tube was washed with an additional 100 ¡A of nuclear dialysis buffer and transferred to a sterile, washed dialysis tubing. The
HATTORI ET AL.
nuclear protein extract was dialyzed against 250 ml of nuclear dialysis buffer with one change of dialysis solution. After 4 hr of dialysis, the nuclear extract was clarified by spinning in a microfuge at 4°C for 10 min. The supernatant was aliquoted, flash-frozen in liquid nitrogen, and stored in liquid nitrogen.
Other methods Plasmid DNA was prepared by the alkaline-NaDodSO., method followed by RNase A treatment and PEG precipitation, as previously described (Hattori and Sakaki, 1986). Cell-free transcription assays were performed with up to 100 /ig of protein per reaction using the "G-free cassette" plasmids AdML 390 and Alb 400 according to the method of Gorski et ai (1986). Mobility-shift assays were performed using 5-10 ¡ig of protein per reaction according to the method of Fried and Crothers (1981) with some modifications (Hattori et ai, 1990). The double-stranded oligonucleotides used in the mobility-shift assays were the HNF-1 binding site in rat albumin (TGTGGTTAATGATCTACAGTTA) (Cereghini et ai, 1988), the consensus sequence of the NF-1 binding site (GATCATTTTGGCTTGAAGCCAATATGA) (Chodosh et ai, 1988), the consensus sequence of the AP-1 binding site (CTAGTGATGAGTCAGCCGGATC) (Angel et ai, 1987), the acutephase response element (APRE) in rat c*2-macroglobulin
(CTAGAGTGAGCAGTTTCTGGGAATTCTTTAATCCTTCTGGGAATTCTGGCT) (Hattori et ai, 1990), and the NFxB binding site in the mouse Ig x enhancer (GGATCCTCAACAGAGGGGACTTTCCGAGGCCA) (Sen and Baltimore, 1986). RESULTS
protocol described above yields 7-15 mg of nuclear protein per liver. The most stringent evaluation of these nuclear extracts was their use in cell-free transcription. Three liver nuclear extracts were independently prepared and tested in cell-free transcription assays using "Gfree cassette" plasmids containing the adenovirus-2 major late promoter (AdML 360) (Fig. 1, lanes 3-5) and the mouse albumin promoter (Alb 400) (Fig. 1, lanes 8-10) as templates. Alb400 contains 650 bp of DNA upstream from the start site of transcription of the rat albumin promoter. These sequences were previously shown to be sufficient for conferring cell type specificity in vitro (Gorski et ai, 1986; Lichtsteiner et ai, 1987). AdmL390 contains 404 bp upstream from the start of transcription of the adenovirus-2 major late promoter, which is expressed efficiently in a wide variety of tissues. The transcriptional activity of these extracts was highly reproducible, as demonstrated in the three extracts shown in Fig. 1, and in 27 other nuclear extracts prepared in this laboratory. Two liver nuclear extracts prepared by the method of Lichtsteiner and Schibler (1989) were also utilized in cell-free transcription for comparison (Fig. 1, lanes 1, 2, 6, and 7). The basic parameters for the cell-free transcription assay were optimized using the well-characterized Alb 400 plasThe
extract
SIMPLIFIED METHOD FOR LIVER NUCLEAR EXTRACTS
AdML390
779
B.
Alb400
MWI 603-
Alb400 20 40 80
310-
120 160
Alb400
15' 30' 45' 60' 90'
Alb400
a°° *0°
%°v
-
-¿>0-oV
Alb400
AdML390
281271-
234-
FIG. 2.
Effect of varying conditions on the cell-free tranA. Effect of nuclear extract concentration on the cell-free transcription of Alb 400. A total of 800 ng of template plasmid was incubated for 45 min with different concentrations of nuclear extract (20, 40, 80, 120, and 160 fig). B. Effect of incubation time on cell-free transcription of Alb 400. A total of 800 ng of template plasmid and 70 fig of nuclear extract protein were incubated for different lengths of time (15, 30, 45, 60, and 90 min). C. Effect of template plasmid concentration on the cell-free transcription of Alb 400. A total of 70 fig of nuclear extract protein was incubated for 45 min with different concentrations of Alb 400 plasmid (200, 400, 800, 1,200, 1,600, and 2,000 ng). D. Effect of a-amanitin on cell-free transcription. A total of 70 fig of nuclear extract and 800 ng of template plasmid (Alb 400 or AdML 390) were incubated for 45 min in the absence (-) or presence (+) of 1 /ig/ml of a-amanitin.
scription products.
2
3
4
5
6
7
8
9 10
FIG. 1. Cell-free transcription products. Three independent liver nuclear extracts prepared using this protocol (lanes 3-5 and 6-10) and two nuclear extracts prepared by the method of Lichtsteiner and Schibler (1989) (lanes 1-2 and 6-7) were used to transcribe the G-free cassette plasmids AdML 390 and Alb 400. The assay included 70 fig of nuclear extract protein and 800 ng of template plasmid incubated for 45 min.
mid as template. Using 800 ng of template DNA, the transcription assay had maximal activity with 80 fig of nuclear extract per reaction (Fig. 2A). The transcriptional activity
increased with time of incubation until 30 min and then subsequently reached a plateau (Fig. 2B). Using 70 fig of nuclear extract, the transcriptional activity was maximal with 800 ng of plasmid template per reaction (Fig. 2C). Transcription from AdML 390 and Alb 400 was eliminated in the presence of 1 ¡ig/ml of a-amanitin (Fig. 2D), demonstrating that the observed transcripts were synthesized by RNA polymerase II. The nuclear extracts were also used to characterize DNA binding proteins by mobility shift assays (Fig. 3) and DNAse I protection assays (Hattori et ai, manuscript in preparation). The extracts prepared by this protocol yielded highly reproducible binding to the binding site of the general transcriptional factor NF-1 (Fig. 3A) and to the liver-specific transcriptional factor HNF-1 (Fig. 3B). Com-
petition experiments with unlabeled oligonucleotides in 500-fold excess demonstrated the specificity of the DNA protein binding (Fig. 3). DISCUSSION The biochemical mechanism involved in the regulation cell-type and tissue-specific gene expression can be effectively analyzed by using cell-free in vitro transcription systems (Bodner and Karin, 1987; Scheidereit et ai, 1987; Lichtsteiner and Schibler, 1989). This approach resulted in the characterization of transcription factors mediating cell-
of
HATTORI ET AL.
780
homogenize the tissue in highly viscous solutions using high-torque engines for driving the pestle or a modisary to
///V/
fied Waring Blendor. Previous methods have prepared DNA-binding proteins by homogenizing liver cells in a low-sucrose buffer prior to
purifying nuclei by antrifugation through a high-sucrose cushion (Graves et ai, 1986; Johnson et ai, 1987). However, the protocol used in the present study consistently produced a higher yield of nuclear extract with a higher transcriptional activity. The higher final concentration of sucrose in the overlaying homogenization solution in this study might explain in part the higher yield of nuclei (i.e., less nuclei were lost at the interface between the overlaying solution and the cushion). Also, the extensive use of protease inhibitors resulted in greater transcriptional activity per milligram of nuclear extract protein. Finally, the NaMo04 increased the yield of phosphorylated DNAbinding proteins, such as the interleukin-6 responsive factor in the a2-macroglobulin gene (Hattori et ai, unpublished observations). The extracts prepared by this protocol give faithful transcription initiation and are highly active in both transcription and binding to specific DNA sequences. The use of this procedure for preparations of nuclear extracts should facilitate the biochemical analysis of tissue-specific gene FIG. 3. Sequence specificity of protein DNA complexes. expression. Mobility-shift assays were performed with 0.1 ng of radiolabeled probe [the binding site of NF-1 (A) or HNF-1 (B)] after incubating with no protein (0), 5 /¿g of liver nuclear REFERENCES extract alone (+ne), or 5 /ig of liver nuclear extract in the presence of 50 ng of cold competing oligonucleotides [the binding sites for NF-1, HNF-1, the acute-phase responsive ANGEL, P., IMAGAWA, M., CHIU, R., STEIN, B., IMBRA, factor of a2-macroglobulin (APRE), AP-1 and NFxB]. R.J., RAHMSDORF, H.F., JONAT, C, HERRLICH, P., _
LABORATORY METHODS
Simplified Method for the Preparation of Transcriptionally Active Liver Nuclear Extracts A
MASAHIRA HATTORI,* ANTONIO TUGORES,* LINDA VELOZ,* MICHAEL and DAVID A. BRENNER*
KARIN.t
ABSTRACT We have developed a simplified method for the preparation of liver nuclear extracts to study gene regulation and protein-DNA interactions. This protocol uses conventional laboratory equipment and standard reagents. The liver tissue is homogenized in a low-salt solution at physiological molarity with subsequent adjustment of the molarity and purification of nuclei by density sedimentation. The nuclear extracts are transcriptionally active in a validated cell-free transcription assay and contain functional DNA-binding proteins. This protocol results in the rapid preparation of highly reproducible and active liver nuclear extracts.
require the construction of new equipment. Here we describe a simplified method for the preparation of nuclear extracts from rat liver. The quality of these extracts was tested by in vitro transcription and mobility-shift assays. They are highly active in transcription and show appropriate tissue specificity.
INTRODUCTION Of genes ÍS the interaction of irons-acting DNA-binding proteins with specific as-acting elements in the DNA (Gorski et ai, 1986; Scheidereit et ai, 1987; Karin et ai, 1990). The first step in the biochemical analysis of this complex system requires the preparation of nuclear extracts from the tissue or cell type of interest. These extracts are then used for in vitro transcription and the analysis of DNA-protein binding. Most studies in this area utilize immortal tissue culture cells as a source of nuclear extracts that are relatively simple to prepare (Ohlsson and Edlund, 1986; Bodner et ai, 1987; Scheidereit et ai, 1987). However, transformed cell lines may not reflect accurately the state of differentiation seen in vivo. In addition, expression of certain tissue specific genes is either highly reduced or completely extinguished in cell lines compared to normal tissue. Preparation of nuclear extracts from the primary tissue, although advantageous, is far more difficult because of the presence of organ capsules, vascular structures, interstitial connective tissue, extracellular proteases, and cellular heterogeneity. To circumvent these problems, elaborate protocols have been designed for the preparation of transcriptionally active nuclear extracts, some of which
mediated by TISSUE
AND CELL-TYPE-SPECIFIC EXPRESSION
Departments of 'Medicine and 92093.
tpharmacology,
METHODS
Preparation of nuclear
extracts
All solutions, tubes, and centrifuges were maintained at 0-4°C. For each nuclear extract preparation, two male Fisher rats (150-250 grams) were anesthetized and killed by exsanguination. The preparation of rat liver nuclear extracts was based on previous methods that produced functional DNA-binding proteins (Graves et ai, 1986; Johnson et ai, 1987) or transcriptionally active extracts (Gorski et ai, 1986; Lichtsteiner and Schibier, 1989). Hepatectomies were performed, and the livers were cut with scissors into approximately 3-mm cubes in 10 ml of homogenization buffer (0.3 M sucrose, 10 mM HEPES pH 7.6, 0.74 mM spermidine, 0.15 mM spermine, 0.1 mM EDTA, 0.1 mM EGTA, 10 mM KC1, 1 mM DTT, 0.5 mM PMSF, and 2 fig/m\ each of aprotinin, leupeptin, and bestatin). Excess
Center for Molecular Genetics,
777
University of California, San Diego,
La
Jolla, CA
778 tissue and blood were removed. Half of the minced tissue was added to 25 ml of homogenization buffer in a 55-ml Potter homogenizer and was homogenized with two or three strokes by a motor-driven Teflon pestle. The homogenate was transferred to a precooled 250-ml cylinder by filtering through a cheese cloth to remove debris. The remaining half of the tissue was homogenized following the same procedure. The homogenized tissue (50 ml in solution) was then mixed with 100 ml of cushion buffer (identical to homogenization buffer except that the concentration of sucrose was 2.2 M), resulting in a final sucrose concentration of 1.57 M. The homogenate was laid over 10 ml of cushion buffer in 38-ml polyallomer ultracentrifuge tubes. The tubes were spun at 24,000 rpm for 50 min at FC in an SW28 ultracentrifuge rotor. The supernatant, including an upper layer of lipids and intact cells, was removed and the tube was inverted for 10 min in ice to drain remaining buffer from the nuclear pellet. At this point, the resulting hepatic nuclei may be flashfrozen in liquid nitrogen after resuspending the six nuclear pellets in 20 ml of nuclear storage buffer (25% glycerol, 50 mM HEPES pH 7.6, 3 mM MgCl2, 0.1 mA/ EDTA, 1.0 mM DTT, 0.1 mM PMSF). Subsequently, a volume of 5 ml of frozen nuclei was reconstituted by the addition of 25 ml of storage reconstitution buffer (12.5% glycerol, 2 mM HEPES pH 7.6, 116 mM KC1, 3 mM MgCl2, 0.1 mM EDTA, 1 mM DTT, 0.1 mM PMSF, 1.0 mM NaMo04, 2 /¿g/ml aprotinin, 2 /¿g/ml leupeptin, 2 /¿g/ml bestatin). The nuclei were resuspended with one stroke of a B pestle in a 40-ml Dounce homogenizer (Wheaton) and lysed as described below. Alternatively, each fresh nuclear pellet was resuspended in 5 ml of nuclear lysis buffer (10% glycerol, 10 mM HEPES pH 7.6, 100 mM KC1, 3 mM MgCL, 0.1 mM EDTA, 1.0 mM DTT, 0.1 ¡iM PMSF, and 2 /tg/ml each of aprotinin, leupeptin, and bestatin). The nuclei were resuspended by one stroke of a B pestle of a 40-ml Dounce homogenizer and brought to a final volume of 40 ml with nuclear lysis buffer; 20 ml of this nuclear suspension was transferred to each 30-ml Oakridge tube. Next 2 ml of 4 M ammonium sulfate pH 7.9 was added to each tube, mixed well, and incubated for 30-60 min with gentle agitation on a rocker platform at 4°C. The tubes were spun at 35,000 rpm for 60 min at FC in a 55.2 Ti ultracentrifuge rotor. The supernatant was measured and transferred to clean Oakridge tubes avoiding the chromatin pellets. Then, 0.33 mg of finely powdered ammonium sulfate per milliliter of supernatant was added to each tube. After the ammonium sulfate was dissolved, the tubes continued incubating without shaking at 0°C for 15 min. The tubes were spun at 35,000 rpm for 20 min at FC in a Ti 55.2 ultracentrifuge rotor. The supernatant was discarded and the tubes were inverted for 10 min on ice. The resulting nuclear protein pellet was resuspended in 500 ¡A of nuclear dialysis buffer (20% glycerol, 20 mM HEPES pH 7.6, 0.1 M KC1, 0.2 mM EDTA, 2 mM DTT, 0.1 mM PMSF, 1 mM NaMoO,, and 2 /ig/ml each of aprotinin, leupeptin, and bestatin). This pellet can be kept on ice overnight. Each tube was washed with an additional 100 ¡A of nuclear dialysis buffer and transferred to a sterile, washed dialysis tubing. The
HATTORI ET AL.
nuclear protein extract was dialyzed against 250 ml of nuclear dialysis buffer with one change of dialysis solution. After 4 hr of dialysis, the nuclear extract was clarified by spinning in a microfuge at 4°C for 10 min. The supernatant was aliquoted, flash-frozen in liquid nitrogen, and stored in liquid nitrogen.
Other methods Plasmid DNA was prepared by the alkaline-NaDodSO., method followed by RNase A treatment and PEG precipitation, as previously described (Hattori and Sakaki, 1986). Cell-free transcription assays were performed with up to 100 /ig of protein per reaction using the "G-free cassette" plasmids AdML 390 and Alb 400 according to the method of Gorski et ai (1986). Mobility-shift assays were performed using 5-10 ¡ig of protein per reaction according to the method of Fried and Crothers (1981) with some modifications (Hattori et ai, 1990). The double-stranded oligonucleotides used in the mobility-shift assays were the HNF-1 binding site in rat albumin (TGTGGTTAATGATCTACAGTTA) (Cereghini et ai, 1988), the consensus sequence of the NF-1 binding site (GATCATTTTGGCTTGAAGCCAATATGA) (Chodosh et ai, 1988), the consensus sequence of the AP-1 binding site (CTAGTGATGAGTCAGCCGGATC) (Angel et ai, 1987), the acutephase response element (APRE) in rat c*2-macroglobulin
(CTAGAGTGAGCAGTTTCTGGGAATTCTTTAATCCTTCTGGGAATTCTGGCT) (Hattori et ai, 1990), and the NFxB binding site in the mouse Ig x enhancer (GGATCCTCAACAGAGGGGACTTTCCGAGGCCA) (Sen and Baltimore, 1986). RESULTS
protocol described above yields 7-15 mg of nuclear protein per liver. The most stringent evaluation of these nuclear extracts was their use in cell-free transcription. Three liver nuclear extracts were independently prepared and tested in cell-free transcription assays using "Gfree cassette" plasmids containing the adenovirus-2 major late promoter (AdML 360) (Fig. 1, lanes 3-5) and the mouse albumin promoter (Alb 400) (Fig. 1, lanes 8-10) as templates. Alb400 contains 650 bp of DNA upstream from the start site of transcription of the rat albumin promoter. These sequences were previously shown to be sufficient for conferring cell type specificity in vitro (Gorski et ai, 1986; Lichtsteiner et ai, 1987). AdmL390 contains 404 bp upstream from the start of transcription of the adenovirus-2 major late promoter, which is expressed efficiently in a wide variety of tissues. The transcriptional activity of these extracts was highly reproducible, as demonstrated in the three extracts shown in Fig. 1, and in 27 other nuclear extracts prepared in this laboratory. Two liver nuclear extracts prepared by the method of Lichtsteiner and Schibler (1989) were also utilized in cell-free transcription for comparison (Fig. 1, lanes 1, 2, 6, and 7). The basic parameters for the cell-free transcription assay were optimized using the well-characterized Alb 400 plasThe
extract
SIMPLIFIED METHOD FOR LIVER NUCLEAR EXTRACTS
AdML390
779
B.
Alb400
MWI 603-
Alb400 20 40 80
310-
120 160
Alb400
15' 30' 45' 60' 90'
Alb400
a°° *0°
%°v
-
-¿>0-oV
Alb400
AdML390
281271-
234-
FIG. 2.
Effect of varying conditions on the cell-free tranA. Effect of nuclear extract concentration on the cell-free transcription of Alb 400. A total of 800 ng of template plasmid was incubated for 45 min with different concentrations of nuclear extract (20, 40, 80, 120, and 160 fig). B. Effect of incubation time on cell-free transcription of Alb 400. A total of 800 ng of template plasmid and 70 fig of nuclear extract protein were incubated for different lengths of time (15, 30, 45, 60, and 90 min). C. Effect of template plasmid concentration on the cell-free transcription of Alb 400. A total of 70 fig of nuclear extract protein was incubated for 45 min with different concentrations of Alb 400 plasmid (200, 400, 800, 1,200, 1,600, and 2,000 ng). D. Effect of a-amanitin on cell-free transcription. A total of 70 fig of nuclear extract and 800 ng of template plasmid (Alb 400 or AdML 390) were incubated for 45 min in the absence (-) or presence (+) of 1 /ig/ml of a-amanitin.
scription products.
2
3
4
5
6
7
8
9 10
FIG. 1. Cell-free transcription products. Three independent liver nuclear extracts prepared using this protocol (lanes 3-5 and 6-10) and two nuclear extracts prepared by the method of Lichtsteiner and Schibler (1989) (lanes 1-2 and 6-7) were used to transcribe the G-free cassette plasmids AdML 390 and Alb 400. The assay included 70 fig of nuclear extract protein and 800 ng of template plasmid incubated for 45 min.
mid as template. Using 800 ng of template DNA, the transcription assay had maximal activity with 80 fig of nuclear extract per reaction (Fig. 2A). The transcriptional activity
increased with time of incubation until 30 min and then subsequently reached a plateau (Fig. 2B). Using 70 fig of nuclear extract, the transcriptional activity was maximal with 800 ng of plasmid template per reaction (Fig. 2C). Transcription from AdML 390 and Alb 400 was eliminated in the presence of 1 ¡ig/ml of a-amanitin (Fig. 2D), demonstrating that the observed transcripts were synthesized by RNA polymerase II. The nuclear extracts were also used to characterize DNA binding proteins by mobility shift assays (Fig. 3) and DNAse I protection assays (Hattori et ai, manuscript in preparation). The extracts prepared by this protocol yielded highly reproducible binding to the binding site of the general transcriptional factor NF-1 (Fig. 3A) and to the liver-specific transcriptional factor HNF-1 (Fig. 3B). Com-
petition experiments with unlabeled oligonucleotides in 500-fold excess demonstrated the specificity of the DNA protein binding (Fig. 3). DISCUSSION The biochemical mechanism involved in the regulation cell-type and tissue-specific gene expression can be effectively analyzed by using cell-free in vitro transcription systems (Bodner and Karin, 1987; Scheidereit et ai, 1987; Lichtsteiner and Schibler, 1989). This approach resulted in the characterization of transcription factors mediating cell-
of
HATTORI ET AL.
780
homogenize the tissue in highly viscous solutions using high-torque engines for driving the pestle or a modisary to
///V/
fied Waring Blendor. Previous methods have prepared DNA-binding proteins by homogenizing liver cells in a low-sucrose buffer prior to
purifying nuclei by antrifugation through a high-sucrose cushion (Graves et ai, 1986; Johnson et ai, 1987). However, the protocol used in the present study consistently produced a higher yield of nuclear extract with a higher transcriptional activity. The higher final concentration of sucrose in the overlaying homogenization solution in this study might explain in part the higher yield of nuclei (i.e., less nuclei were lost at the interface between the overlaying solution and the cushion). Also, the extensive use of protease inhibitors resulted in greater transcriptional activity per milligram of nuclear extract protein. Finally, the NaMo04 increased the yield of phosphorylated DNAbinding proteins, such as the interleukin-6 responsive factor in the a2-macroglobulin gene (Hattori et ai, unpublished observations). The extracts prepared by this protocol give faithful transcription initiation and are highly active in both transcription and binding to specific DNA sequences. The use of this procedure for preparations of nuclear extracts should facilitate the biochemical analysis of tissue-specific gene FIG. 3. Sequence specificity of protein DNA complexes. expression. Mobility-shift assays were performed with 0.1 ng of radiolabeled probe [the binding site of NF-1 (A) or HNF-1 (B)] after incubating with no protein (0), 5 /¿g of liver nuclear REFERENCES extract alone (+ne), or 5 /ig of liver nuclear extract in the presence of 50 ng of cold competing oligonucleotides [the binding sites for NF-1, HNF-1, the acute-phase responsive ANGEL, P., IMAGAWA, M., CHIU, R., STEIN, B., IMBRA, factor of a2-macroglobulin (APRE), AP-1 and NFxB]. R.J., RAHMSDORF, H.F., JONAT, C, HERRLICH, P., _
