Oncotic pressure regulates gene transcriptions of albumin and apolipoprotein B in cultured rat hepatoma cells ATSUSHI YAMAUCHI, YOSHIFUMI FUKUHARA, SHIGEO YAMAMOTO, FUMIO YANO, MASARU TAKENAKA, ENYU IMAI, TAM10 NOGUCHI, TAKEHIKO TANAKA, TAKENOBU KAMADA, AND NAOHIKO UEDA First Department of Medicine and Department of Nutrition and Physiological Chemistry, Osaka University Medical School, Osaka 553, Japan Yamauchi, Atsushi, Yoshifumi Fukuhara, Shigeo Yamamoto, Fumio Yano, Masaru Takenaka, Enyu Imai, Tamio Noguchi, Takehiko Tanaka, Takenobu Kamada, and Naohiko Ueda. Oncotic pressure regulates gene transcriptions of albumin and apolipoprotein B in cultured rat hepatoma cells. Am. J. PhysioL. 263 (Cell PhysioL. 32): C397C404, 1992.-The mechanism of the accelerated syntheses of albumin and apolipoprotein B (apo B) in response to decreased oncotic pressure was investigated in cultured rat hepatoma H4II-E cells. Addition of dextran (mol wt 6-9 X lo*) to the culture medium decreased the levels of albumin and apo B mRNAs in an oncotic pressure-dependent manner. The reductions of both mRNAs were attenuated with increase in the molecular weight of dextran, which resulted in a decrease in oncotic pressure. Addition of macromolecule increased the viscosity in medium; however, alteration of viscosity appeared not to correlate with albumin and apo B mRNA levels. Transcriptional run-on assays with isolated nuclei from dextran-treated vs. untreated hepatoma cells indicated that the changes in steady-state mRNA levels were mainly controlled at the transcriptional step. Treatment with cycloheximide increased albumin mRNA to the basal level, which was effectively suppressed by dextran, and resulted in superinduction of apo B mRNA. These changes occurred primarily at the transcriptional step. These results suggest that regulations of the expressions of the albumin and apo B genes for adaptive increases in the mRNAs may require the continued synthesis of a labile protein(s) or a limiting transcription factor(s). We conclude that oncotic pressure plays an important role in regulation of expression of the albumin and apo B genes at the transcriptional step. messenger
ribonucleic
acid; H4-II-E
cells; nephrotic
syndrome
NEPHROTIC SYNDROME is characterized by increased urinary excretions of albumin and other proteins hypercholesterassociated with hypoalbuminemia, olemia, and edema. Recently, we demonstrated that the steady-state level of albumin mRNA in the liver of nephrotic rats was increased at the transcriptional level (28). Kaysen et al. (12) also reported that a reduced serum oncotic pressure and increased level of dietary protein in combination stimulated albumin synthesis in nephrotic rats at the level of gene transcription. The molecular mechanism by which protein synthesis is increased in nephrotic animals is still unknown. Several factors have been found to contribute to regulation of albumin synthesis in vivo: hormones such as insulin, thyroid hormones, growth hormone, and glucagon regulate albumin synthesis (2, 27). Moreover, cytokines, which are secreted from various cells in inflammations, suppress albumin synthesis at the transcriptional level (20). In experimental models of the nephrotic syndrome, these factors that regulate albumin gene transcription must be taken into consideration. The mechanisms responsible for hyperlipidemia in
THE
0363-6143/92
$2.00
Copyright
this syndrome are more complex and may include both reduced clearance of lipoproteins from the circulation (5,8,17) and enhanced hepatic syntheses of lipoproteins (17,18,23). From studies on a rat model of the nephrotic syndrome, Marsh and Drabkin (17) proposed that hyperlipidemia was due to overproduction of hepatic lipoproteins and that hypoalbuminemia acted as a signal for increase in hepatic production of plasma proteins in general. Increases in hepatic secretions of lipoproteins of all density classes and of several apolipoproteins were demonstrated in perfused rat liver (18) and rat liver slices (23). The levels of very-low-density lipoprotein (VLDL) and low-density lipoprotein (LDL), in particular, increase in the nephrotic syndrome, and hepatic synthesis of VLDL is markedly increased in clinical and experimental nephrotic subjects (3, 15, 18). Apolipoprotein B (apo B) is a major component of VLDL and LDL, and its abundance is increased in nephrotic patients cm There are several reports of a negative correlation between the serum albumin and cholesterol levels (3, 5, 13). A closer correlation has, however, been found between the plasma oncotic pressure and the plasma cholesterol level (3). Baxter et al. (4) demonstrated that daily albumin infusion led to decreases in the plasma lipid and lipoprotein levels in adult nephrotic subjects. The infusion of dextran into patients and animals with nephrotic syndrome, however, decreased hepatic albumin synthesis (26) and the serum lipid level (l), suggesting that decrease in plasma oncotic pressure, rather than in albumin, is a signal for accelerated hepatic syntheses of albumin and lipoproteins. Despite these findings supporting the original hypothesis of Marsh and Drabkin (17), many investigators have been unable to demonstrate a correlation between either plasma albumin or the plasma oncotic pressure and t,he rates of syntheses of albumin and lipoproteins (10, 25). Yedgar et al. (29, 30), however, reported that change in plasma viscosity may be the principal signal sensed by hepatocyte for control of protein production. To avoid the involvement of hormonal and paracrine factors in this study, we used cultured differentiated hepatoma H4-II-E cells for direct examination of the hypothesis that oncotic pressure regulates the syntheses of albumin and apo B by activating gene transcription. We investigated the effect of change in oncotic pressure of the culture medium on the steady-state levels of albumin and apo B mRNAs in the cells. We also measured the transcription rates of the albumin and apo B genes in isolated nuclei from dextran-treated and untreated rat hepatoma cells by nuclear transcription assay.
0 1992 the American
Physiological
Society
c397
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 17, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
C398
ONCOTIC
MATERIALS
AND
PRESSURE
AND
METHODS
Materials. Bovine serum albumin (essentially fatty acid free), dextran (mol wt 6-9 x 104, 15 x 104, and 48 X lo*; produced by Leuconostoc mesenteroides, strain no. B-512), human y-globulin (Cohn fraction II), actinomycin D, and cycloheximide were all obtained from Sigma (St. Louis, MO). Fetal calf serum (FCS) was obtained from MA Bioproducts (Walkersville, MD) and was heat inactivated at 56°C for 30 min before use. Minimal essential medium (MEM) and Dulbecco’s phosphate-buffered saline were obtained from Nissui Pharmaceutical (Tokyo, Japan). A multiprime labeling system was from Amersham International (Buckinghamshire, UK). a-:j2P-labeled deoxycytidine triphosphate (dCTP; 400-800 Ci/mmol) was from Du Pont-New England Nuclear (Boston, MA). “H-labeled uridine 5’-triphosphate (UTP; 35-50 Ci/mmol) and [a-““P]UTP (>600 Ci/mmol) were purchased from ICN Radiochemicals (Irvine, CA). Nitrocellulose filters (BA85) and nylon filters (Hybond-N) were obtained from Schleicher & Schuell (Dassel, Germany) and Amersham International (Buckinghamshire, UK), respectively. DNA polymerase I was from Bethesda Research Laboratories (Rockville, MD), deoxyribonuclease I from Takara Shuzo (Kyoto, Japan), ribonuclease A from Boehringer Mannheim (Mannheim, Germany), RNA polymerase from Pharmacia (Piscataway, NJ), and proteinase K from Merck (Darmstadt, Germany). P-Actin cDNA was obt,ained from Wako Pure Chemical Industries (Osaka, Japan). Other chemicals were of the highest purity available from commercial sources. Cell line and culture conditions. The H4-II-E rat hepatoma cell line (22), producing and secreting albumin and other plasma proteins, was obtained from the American Type Culture Collection (Rockville, MD). The H4-II-E cells (seeding density 3 X lo6 cells/dish) were grown in MEM containing 10% FCS in plastic petri dishes (10 cm diam) for 3 days to confluent monolayers. In studies on the effect of oncotic pressure, the culture medium was then changed to MEM containing 10% FCS with or without various concentrations of macromolecule, and incubation was continued for 6 h at 37°C. The cell monolayers were then washed twice with Dulbecco’s phosphate-buffered saline and harvested. Other procedures are described in RESULTS.
The oncotic pressure of the culture medium was measured with a colloid osmometer (model 4400, Wescor, Logan, UT). The viscosity of the culture medium was measured with a coneplate microviscometer (Biorheolyzer model E, Tohki Sangyoh, Tokyo, Japan), using normal pooled plasma as a standard. Isolation of total cellular RNAs and RNA blot hybridization assay. Total cellular RNAs were extracted from the cultured cells by the guanidinium thiocyanate-lithium chloride method, as described previously (28). Individual mRNA levels were usuTable 1. Oncotic of culture media
pressures
and
20 rpm
(rpm),
respectively.
ally quantitated by dot-blot hybridization in at least three independent experiments as described previously (28). Northern blot analysis was carried out and analyzed by densitometry as described elsewhere (28), using ““P-labeled rat albumin cDNA (1,200 bp) (18) and rat apo B cDNA (230 bp) (16). After being stripped of these probes, human /3-actin cDNA (440 bp) (21) was rehybridized and p-actin mRNA level was measured for correcting albumin and apo B mRNA levels. The human P-actin gene is highly homologous to the corresponding sequence in the rat P-actin gene. These cDNAs were labeled with [a-“2P]dCTP to a specific activity of 2.0-4.0 X lOa counts. min-l . pg-’ in a multiprime labeling system. Nuclear transcription assay. Isolation of nuclei from H4-II-E cells and subsequent in vitro transcription assay were performed essentially as described previously (28). Each cDNA insert was ligated into the pUC vector, which was used as control DNA to monitor for nonspecific binding of RNA. Nuclei were incubated with 90 &i of [o(-:‘“P]UTP for 30 min at 25°C. After treatment of deoxyribonuclease I and proteinase K, the ‘s2P-labeled products (3.5 X 10” counts/min) from each reaction were hybridized to nitrocellulose filters containing 3 pg of denatured plasmid, which had been linearized by digestion with an appropriate restriction enzyme. Filter hybridization and washing were performed as described previously (28). The hybridization efficiency was assessed using 3,000 counts/min of [“H]cRNA synthesized from each cDNA insert as a control. The relative rates of mRNA synthesis were determined by densitometric scanning of autoradiogram, and values were corrected for hybridization
dext ran concent ration 0 2 4 8(wdU Albumin 2.2 ItbB-Act i n 2.2 kb-
Concentration, g/d1
0
2
4
1.7
9.3
18.4
1.13 1.68 2:/o
1.37 1.61 1.89
1 .a9 2.39 2.94
8
37.2 2.77 3.12 3.78
10 rpm Values are means for duplicate determinations. Oncotic pressure of normal pooled serum was -25 mmHg. Viscosity of normal pooled serum was 1.146-1.962, 1.135-1.965, and 1.156-1.934 CP at 10, 20, and 50 revolutions/min
TRANSCRIPTION
viscosities
Dextran
Osmotic pressure, mmHg Viscosity, cP 50 rpm
GENE
Fig. 1. Northern
blot analysis of RNAs isolated from H4-II-E
cells
incubated with various concentration of dextran. Samples of 4.5 bg of total RNA were isolated by electrophoresis on 0.8% agarose gel containing 2.2 M formaldehyde. RNAs were then transferred to nylon filters and hybridized with ““P-labeled albumin and apolipoprotein (apo) B cDNAs together. After being exposed to X-ray films, probes were removed and rehybridized with fl-actin cDNA for correction.
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 17, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
ONCOTIC
PRESSURE
AND
efficiency,.which varied from 10 to 16%. Statistical analysis. Values are given as means + SD. Student’s t test was used for statistical evaluation of the data, and P < 0.05 was considered to be significant. Linear correlations between various parameters were calculated by the least-squares method.
on mRNA levels in H4-II-E understanding of the regula-
tion of hepatic protein synthesis through alteration of oncotic pressure, we investigated the effect of macromolecule that produces a colloidosmotic pressure on the levels of albumin, apo B, and P-actin mRNAs in H4-II-E cells in culture. Three days after cell seeding, the growth medium was changed to MEM containing 10% FCS plus 0, 2, 4, or 8 g/d1 of dextran (mol wt 6-9 x 104), which conferred 1.7,9.3,18.4, or 37.2 mmHg of oncotic pressure, respectively (Table 1). As the concentration of macromolecule increased, the viscosity also increased (Table 1). After 6 h incubation, total cellular RNAs were extracted from the cells and the level of each mRNA was determined. Northern analysis revealed that the levels of albumin and apo B mRNAs decreased inversely with increase in the concentration of dextran added to the culture medium (Fig. 1). The decreases in the levels of albumin and apo B mRNAs caused by dextran were evident within 4 h after the medium change and maximal after 6 h and persisted until at least 24 h after the medium change (data not shown). On the other hand, the level of @-actin mRNA appeared to remain constant. The Dextran Concentration
4
(gidl)
TRANSCRIPTION
c399
changes in the steady-state levels of albumin, apo B, and P-actin mRNAs were quantitated by dot-blot hybridization (Fig. 2). Albumin mRNA levels of hepatoma cells treated with 2,4, and 8 g/d1 of dextran were decreased to 85, 82, and 71% of dextran-free control, respectively (Fig. 2). Apo B mRNA levels also suppressed to 84, 75, and 68% of control in the presence of 2, 4, and 8 g/d1 of dextran, respectively (Fig. 2). The ratios of albumin and apo B mRNAs to P-actin mRNA in medium with 8 g/d1 of dextran were -40% less than those of cells in medium without added dextran (P < 0.01). Addition of macromolecule also increases the viscosity in the medium. Yedger et al. (29, 30) argued that increment of the plasma viscosity may be the principal signal sensed by hepatocyte for protein production. We further studied the effect of oncotic pressure and viscosity using various molecular weights of dextran (6-9 x 104, 12 x 104, and 48 X 104) or y-globulin (mol wt 15 X 104), which conferred various oncotic pressure and viscosity values (Fig. 3). When 8 g/d1 of different macromolecules was added to the medium, oncotic pressure increased in reciprocal proportion to molecular weight of dextran, whereas viscosity increased in proportion to increase in molecular weight of dextran. The levels of albumin and apo B mRNAs decreased with increase in the oncotic pressure (Fig. 3). In contrast to the effect of oncotic pressure, the steady-state level of albumin mRNA or apo B mRNA was not suppressed in the presence of high viscosity. To distinguish which factor, the oncotic pressure or the
RESULTS
Effects of macromolecule cells. As a basis for further
GENE
8
Albumin Apo B O-Actin
Albumin mRNA level
(0,0)
100+6
8521
Apo B mRNA level
w
100+7
841~6’
O-Actin mFiNA level
(“A)
100+5
96i3
Albumin/O-Actin mRNA ratio (%)
100+4
88i3’*
80210
6OklO”
Apo B/O-Actin mRNA ratio (%)
10027
87i-7 **
73+9**
61+5**
l
82k2”
71+3**
75il4’
68ilO”
102+9
llO-tll
Fig. 2. Levels of albumin and apo B mRNAs in H4-II-E cells incubated with various concentrations of dextran. Three days after seeding, confluent monolayers were incubated in medium containing dextran (mol wt 6-9 x 104) at indicated concentrations for 6 h at 37°C. Total RNA then was extracted, and level of each mRNA was measured. Representative autoradiograms of dot-blot hybridization are shown. Data were analyzed by densitometry from 6 independent experiments, and values are shown as percentage of control without dextran. * P < 0.05. ** P < 0.01.
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 17, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
c400
ONCOTIC
PRESSURE
AND
25
Control (0 WV
oncotic pressure (mmHg) viscosity
Dex(6) (8 gidl)
1.7
37.2
Dex(12) (8 gidl)
Y-Glob(l5) (8 gidl)
Dex(48) (8 gidl)
31.0
nd
27.2
(Cp)
50 rpm
1.13
2.26
4.01
nd
so
20 rpm
1.68
3.03
4.44
nd
7.73
10 rpm
2.22
3.42
5.10
8.31
Fig. 3. Effects of y-globulin and dextrans with different molecular weights on levels of albumin and apo B mRNAs. Three days after seeding, confluent monolayers were incubated in medium containing 8 g/d1 of y-globulin or dextrans with indicated molecular weights for 6 h, and then levels of mRNAs were measured. Average values of ratio of albumin mRNA to p-actin mRNA (open bars) or apo B mRNA to /3-actin mRNA (hatched bars) are shown as percentages of value without additions (n = 2). Viscosities of respective culture media are also shown. Dex(G), Dex(lZ), and Dex(48), dextrans of mol wt 6-9 X 104, 12 x lo”, and 48 x lo”, respectively; r-Glob(l5), y-globulin of mol wt 15 X 103; rpm, revolutions/min; so, scaled over; nd, not determined.
viscosity, was more important for regulating the levels of albumin and apo B mRNAs, we plotted the albumin and apo B mRNA levels against the oncotic pressure or viscosity (Fig. 4). Albumin mRNA levels, relative to dextran-free control as lOO%, were negatively correlated with the oncotic pressure (r = -0.806, P < 0.001). We also observed a negative correlation between apo B mRNA and the oncotic pressure (r = -0.737, P < 0.001). On the other hand, we could not observe any significant correlation between the viscosity and albumin or apo B mRNA level. From these results we conclude that the oncotic pressure is more important than the viscosity in regulating the steady-state levels of albumin and apo B mRNAs. Restoration of levels of albumin and apo B mRNAs by decrease in dextran concentration. Next we tested the effect of decrease in oncotic pressure on the suppressed levels of albumin and apo B mRNA. For this, confluent monolayers were preincubated in medium containing 8 g/d1 of dextran (mol wt 6-9 x 104) for 15 h at 37°C and then incubated in medium containing various concentrations (8, 4, 2, or 0 g/dl) of dextran for 6 h. The levels of albumin and apo B mRNAs were then measured. Figure 5 shows that the levels of albumin and apo B mRNAs increased inversely with the concentration (or oncotic pressure) of dextran in the incubation medium. Addition of actinomycin D (2 pg/ml) to the medium without dextran (0 g/dl) inhibited the increases in the levels of albumin and apo B mRNAs. Thus the steady-state levels of albumin and ano B mRNAs seem to be regulated mainlv
GENE
TRANSCRIPTION
at the transcriptional step. Effect of oncotic pressure produced by dextran on transcription rates of albumin and apo B genes. For direct evaluation of whether the oncotic pressure produced by dextran affects the transcriptional step of expression of the albumin and apo B genes, we measured the transcription rates of these genes by a nuclear run-on transcription assay. Confluent monolayers were incubated in medium containing 0,4, or 8 g/d1 of dextran (mol wt 6-9 x 104) for 4 h at 37°C and then the nuclei were isolated and examined by in vitro run-on transcription assay. The basal transcription rate of the albumin gene was lower than that of apo B gene in H4-II-E cells (Fig. 6). The addition of dextran decreased the transcription rates of the albumin and apo B genes concentration dependently, with a concentration of 8 g/d1 dextran decreasing the transcription rates of both genes m 50%. These results directly demonstrate that the change in oncotic pressure produced by dextran regulates the transcriptions of both genes. Effect of cycloheximide on transcription rates of albumin and apo B genes. The expressions of several genes are known to require ongoing protein synthesis (7, 24). Therefore we investigated whether protein synthesis was required for the regulation of transcription of the albumin, apo B, and P-actin genes. First, we measured the levels of albumin, apo B, and P-actin mRNAs of cells incubated for 6 h in medium with 0 or 8 g/d1 of dextran in the presence or absence of cycloheximide (5 pg/ml). In the absence of dextran, the presence of cycloheximide did not affect the level of albumin or apo B mRNA but increased the level of p-actin mRNA. In medium containing 8 g/d1 of dextran, the presence of cycloheximide increased the level of apo B mRNA markedly compared with the level in its absence and increased the albumin mRNA level slightly (Table 2). There are two possible mechanisms for these increases in mRNA levels: increase in the rate of synthesis of mRNA and decrease in the rate of degradation of mRNA. To distinguish between these possibilities, we measured the transcription rates of these genes directly by nuclear transcription assay (Fig. 6). H4-II-E cells were incubated for 4 h in medium containing 8 g/d1 of dextran in the presence or absence of cycloheximide. The transcription rate of the apo B gene was 3.0 times higher in the presence than in the absence of cycloheximide. The transcription rate of the albumin gene was 1.8 times higher in the presence than in the absence of cycloheximide and was similar to the basal transcription rate. The increase in P-actin gene transcription in the presence of cycloheximide was only -30%. DISCUSSION
The object of our study was to clarify the mechanism regulating the syntheses of albumin and apo B protein in response to alteration of oncotic pressure. To investigate this problem directly, we employed cultured H4-II-E cells and evaluated the effects of the oncotic pressure of the culture medium on the levels of albumin and apo B mRNAs. We used the H4-II-E cell line, derived from Reuber H-35 hepatoma, because this line has been found to be useful for studying the regulations of syntheses of albumin and other nroteins (22). We found that addition of
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 17, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
ONCOTIC
y = 98.058 p
ribonucleic
acid; H4-II-E
cells; nephrotic
syndrome
NEPHROTIC SYNDROME is characterized by increased urinary excretions of albumin and other proteins hypercholesterassociated with hypoalbuminemia, olemia, and edema. Recently, we demonstrated that the steady-state level of albumin mRNA in the liver of nephrotic rats was increased at the transcriptional level (28). Kaysen et al. (12) also reported that a reduced serum oncotic pressure and increased level of dietary protein in combination stimulated albumin synthesis in nephrotic rats at the level of gene transcription. The molecular mechanism by which protein synthesis is increased in nephrotic animals is still unknown. Several factors have been found to contribute to regulation of albumin synthesis in vivo: hormones such as insulin, thyroid hormones, growth hormone, and glucagon regulate albumin synthesis (2, 27). Moreover, cytokines, which are secreted from various cells in inflammations, suppress albumin synthesis at the transcriptional level (20). In experimental models of the nephrotic syndrome, these factors that regulate albumin gene transcription must be taken into consideration. The mechanisms responsible for hyperlipidemia in
THE
0363-6143/92
$2.00
Copyright
this syndrome are more complex and may include both reduced clearance of lipoproteins from the circulation (5,8,17) and enhanced hepatic syntheses of lipoproteins (17,18,23). From studies on a rat model of the nephrotic syndrome, Marsh and Drabkin (17) proposed that hyperlipidemia was due to overproduction of hepatic lipoproteins and that hypoalbuminemia acted as a signal for increase in hepatic production of plasma proteins in general. Increases in hepatic secretions of lipoproteins of all density classes and of several apolipoproteins were demonstrated in perfused rat liver (18) and rat liver slices (23). The levels of very-low-density lipoprotein (VLDL) and low-density lipoprotein (LDL), in particular, increase in the nephrotic syndrome, and hepatic synthesis of VLDL is markedly increased in clinical and experimental nephrotic subjects (3, 15, 18). Apolipoprotein B (apo B) is a major component of VLDL and LDL, and its abundance is increased in nephrotic patients cm There are several reports of a negative correlation between the serum albumin and cholesterol levels (3, 5, 13). A closer correlation has, however, been found between the plasma oncotic pressure and the plasma cholesterol level (3). Baxter et al. (4) demonstrated that daily albumin infusion led to decreases in the plasma lipid and lipoprotein levels in adult nephrotic subjects. The infusion of dextran into patients and animals with nephrotic syndrome, however, decreased hepatic albumin synthesis (26) and the serum lipid level (l), suggesting that decrease in plasma oncotic pressure, rather than in albumin, is a signal for accelerated hepatic syntheses of albumin and lipoproteins. Despite these findings supporting the original hypothesis of Marsh and Drabkin (17), many investigators have been unable to demonstrate a correlation between either plasma albumin or the plasma oncotic pressure and t,he rates of syntheses of albumin and lipoproteins (10, 25). Yedgar et al. (29, 30), however, reported that change in plasma viscosity may be the principal signal sensed by hepatocyte for control of protein production. To avoid the involvement of hormonal and paracrine factors in this study, we used cultured differentiated hepatoma H4-II-E cells for direct examination of the hypothesis that oncotic pressure regulates the syntheses of albumin and apo B by activating gene transcription. We investigated the effect of change in oncotic pressure of the culture medium on the steady-state levels of albumin and apo B mRNAs in the cells. We also measured the transcription rates of the albumin and apo B genes in isolated nuclei from dextran-treated and untreated rat hepatoma cells by nuclear transcription assay.
0 1992 the American
Physiological
Society
c397
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 17, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
C398
ONCOTIC
MATERIALS
AND
PRESSURE
AND
METHODS
Materials. Bovine serum albumin (essentially fatty acid free), dextran (mol wt 6-9 x 104, 15 x 104, and 48 X lo*; produced by Leuconostoc mesenteroides, strain no. B-512), human y-globulin (Cohn fraction II), actinomycin D, and cycloheximide were all obtained from Sigma (St. Louis, MO). Fetal calf serum (FCS) was obtained from MA Bioproducts (Walkersville, MD) and was heat inactivated at 56°C for 30 min before use. Minimal essential medium (MEM) and Dulbecco’s phosphate-buffered saline were obtained from Nissui Pharmaceutical (Tokyo, Japan). A multiprime labeling system was from Amersham International (Buckinghamshire, UK). a-:j2P-labeled deoxycytidine triphosphate (dCTP; 400-800 Ci/mmol) was from Du Pont-New England Nuclear (Boston, MA). “H-labeled uridine 5’-triphosphate (UTP; 35-50 Ci/mmol) and [a-““P]UTP (>600 Ci/mmol) were purchased from ICN Radiochemicals (Irvine, CA). Nitrocellulose filters (BA85) and nylon filters (Hybond-N) were obtained from Schleicher & Schuell (Dassel, Germany) and Amersham International (Buckinghamshire, UK), respectively. DNA polymerase I was from Bethesda Research Laboratories (Rockville, MD), deoxyribonuclease I from Takara Shuzo (Kyoto, Japan), ribonuclease A from Boehringer Mannheim (Mannheim, Germany), RNA polymerase from Pharmacia (Piscataway, NJ), and proteinase K from Merck (Darmstadt, Germany). P-Actin cDNA was obt,ained from Wako Pure Chemical Industries (Osaka, Japan). Other chemicals were of the highest purity available from commercial sources. Cell line and culture conditions. The H4-II-E rat hepatoma cell line (22), producing and secreting albumin and other plasma proteins, was obtained from the American Type Culture Collection (Rockville, MD). The H4-II-E cells (seeding density 3 X lo6 cells/dish) were grown in MEM containing 10% FCS in plastic petri dishes (10 cm diam) for 3 days to confluent monolayers. In studies on the effect of oncotic pressure, the culture medium was then changed to MEM containing 10% FCS with or without various concentrations of macromolecule, and incubation was continued for 6 h at 37°C. The cell monolayers were then washed twice with Dulbecco’s phosphate-buffered saline and harvested. Other procedures are described in RESULTS.
The oncotic pressure of the culture medium was measured with a colloid osmometer (model 4400, Wescor, Logan, UT). The viscosity of the culture medium was measured with a coneplate microviscometer (Biorheolyzer model E, Tohki Sangyoh, Tokyo, Japan), using normal pooled plasma as a standard. Isolation of total cellular RNAs and RNA blot hybridization assay. Total cellular RNAs were extracted from the cultured cells by the guanidinium thiocyanate-lithium chloride method, as described previously (28). Individual mRNA levels were usuTable 1. Oncotic of culture media
pressures
and
20 rpm
(rpm),
respectively.
ally quantitated by dot-blot hybridization in at least three independent experiments as described previously (28). Northern blot analysis was carried out and analyzed by densitometry as described elsewhere (28), using ““P-labeled rat albumin cDNA (1,200 bp) (18) and rat apo B cDNA (230 bp) (16). After being stripped of these probes, human /3-actin cDNA (440 bp) (21) was rehybridized and p-actin mRNA level was measured for correcting albumin and apo B mRNA levels. The human P-actin gene is highly homologous to the corresponding sequence in the rat P-actin gene. These cDNAs were labeled with [a-“2P]dCTP to a specific activity of 2.0-4.0 X lOa counts. min-l . pg-’ in a multiprime labeling system. Nuclear transcription assay. Isolation of nuclei from H4-II-E cells and subsequent in vitro transcription assay were performed essentially as described previously (28). Each cDNA insert was ligated into the pUC vector, which was used as control DNA to monitor for nonspecific binding of RNA. Nuclei were incubated with 90 &i of [o(-:‘“P]UTP for 30 min at 25°C. After treatment of deoxyribonuclease I and proteinase K, the ‘s2P-labeled products (3.5 X 10” counts/min) from each reaction were hybridized to nitrocellulose filters containing 3 pg of denatured plasmid, which had been linearized by digestion with an appropriate restriction enzyme. Filter hybridization and washing were performed as described previously (28). The hybridization efficiency was assessed using 3,000 counts/min of [“H]cRNA synthesized from each cDNA insert as a control. The relative rates of mRNA synthesis were determined by densitometric scanning of autoradiogram, and values were corrected for hybridization
dext ran concent ration 0 2 4 8(wdU Albumin 2.2 ItbB-Act i n 2.2 kb-
Concentration, g/d1
0
2
4
1.7
9.3
18.4
1.13 1.68 2:/o
1.37 1.61 1.89
1 .a9 2.39 2.94
8
37.2 2.77 3.12 3.78
10 rpm Values are means for duplicate determinations. Oncotic pressure of normal pooled serum was -25 mmHg. Viscosity of normal pooled serum was 1.146-1.962, 1.135-1.965, and 1.156-1.934 CP at 10, 20, and 50 revolutions/min
TRANSCRIPTION
viscosities
Dextran
Osmotic pressure, mmHg Viscosity, cP 50 rpm
GENE
Fig. 1. Northern
blot analysis of RNAs isolated from H4-II-E
cells
incubated with various concentration of dextran. Samples of 4.5 bg of total RNA were isolated by electrophoresis on 0.8% agarose gel containing 2.2 M formaldehyde. RNAs were then transferred to nylon filters and hybridized with ““P-labeled albumin and apolipoprotein (apo) B cDNAs together. After being exposed to X-ray films, probes were removed and rehybridized with fl-actin cDNA for correction.
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 17, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
ONCOTIC
PRESSURE
AND
efficiency,.which varied from 10 to 16%. Statistical analysis. Values are given as means + SD. Student’s t test was used for statistical evaluation of the data, and P < 0.05 was considered to be significant. Linear correlations between various parameters were calculated by the least-squares method.
on mRNA levels in H4-II-E understanding of the regula-
tion of hepatic protein synthesis through alteration of oncotic pressure, we investigated the effect of macromolecule that produces a colloidosmotic pressure on the levels of albumin, apo B, and P-actin mRNAs in H4-II-E cells in culture. Three days after cell seeding, the growth medium was changed to MEM containing 10% FCS plus 0, 2, 4, or 8 g/d1 of dextran (mol wt 6-9 x 104), which conferred 1.7,9.3,18.4, or 37.2 mmHg of oncotic pressure, respectively (Table 1). As the concentration of macromolecule increased, the viscosity also increased (Table 1). After 6 h incubation, total cellular RNAs were extracted from the cells and the level of each mRNA was determined. Northern analysis revealed that the levels of albumin and apo B mRNAs decreased inversely with increase in the concentration of dextran added to the culture medium (Fig. 1). The decreases in the levels of albumin and apo B mRNAs caused by dextran were evident within 4 h after the medium change and maximal after 6 h and persisted until at least 24 h after the medium change (data not shown). On the other hand, the level of @-actin mRNA appeared to remain constant. The Dextran Concentration
4
(gidl)
TRANSCRIPTION
c399
changes in the steady-state levels of albumin, apo B, and P-actin mRNAs were quantitated by dot-blot hybridization (Fig. 2). Albumin mRNA levels of hepatoma cells treated with 2,4, and 8 g/d1 of dextran were decreased to 85, 82, and 71% of dextran-free control, respectively (Fig. 2). Apo B mRNA levels also suppressed to 84, 75, and 68% of control in the presence of 2, 4, and 8 g/d1 of dextran, respectively (Fig. 2). The ratios of albumin and apo B mRNAs to P-actin mRNA in medium with 8 g/d1 of dextran were -40% less than those of cells in medium without added dextran (P < 0.01). Addition of macromolecule also increases the viscosity in the medium. Yedger et al. (29, 30) argued that increment of the plasma viscosity may be the principal signal sensed by hepatocyte for protein production. We further studied the effect of oncotic pressure and viscosity using various molecular weights of dextran (6-9 x 104, 12 x 104, and 48 X 104) or y-globulin (mol wt 15 X 104), which conferred various oncotic pressure and viscosity values (Fig. 3). When 8 g/d1 of different macromolecules was added to the medium, oncotic pressure increased in reciprocal proportion to molecular weight of dextran, whereas viscosity increased in proportion to increase in molecular weight of dextran. The levels of albumin and apo B mRNAs decreased with increase in the oncotic pressure (Fig. 3). In contrast to the effect of oncotic pressure, the steady-state level of albumin mRNA or apo B mRNA was not suppressed in the presence of high viscosity. To distinguish which factor, the oncotic pressure or the
RESULTS
Effects of macromolecule cells. As a basis for further
GENE
8
Albumin Apo B O-Actin
Albumin mRNA level
(0,0)
100+6
8521
Apo B mRNA level
w
100+7
841~6’
O-Actin mFiNA level
(“A)
100+5
96i3
Albumin/O-Actin mRNA ratio (%)
100+4
88i3’*
80210
6OklO”
Apo B/O-Actin mRNA ratio (%)
10027
87i-7 **
73+9**
61+5**
l
82k2”
71+3**
75il4’
68ilO”
102+9
llO-tll
Fig. 2. Levels of albumin and apo B mRNAs in H4-II-E cells incubated with various concentrations of dextran. Three days after seeding, confluent monolayers were incubated in medium containing dextran (mol wt 6-9 x 104) at indicated concentrations for 6 h at 37°C. Total RNA then was extracted, and level of each mRNA was measured. Representative autoradiograms of dot-blot hybridization are shown. Data were analyzed by densitometry from 6 independent experiments, and values are shown as percentage of control without dextran. * P < 0.05. ** P < 0.01.
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 17, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
c400
ONCOTIC
PRESSURE
AND
25
Control (0 WV
oncotic pressure (mmHg) viscosity
Dex(6) (8 gidl)
1.7
37.2
Dex(12) (8 gidl)
Y-Glob(l5) (8 gidl)
Dex(48) (8 gidl)
31.0
nd
27.2
(Cp)
50 rpm
1.13
2.26
4.01
nd
so
20 rpm
1.68
3.03
4.44
nd
7.73
10 rpm
2.22
3.42
5.10
8.31
Fig. 3. Effects of y-globulin and dextrans with different molecular weights on levels of albumin and apo B mRNAs. Three days after seeding, confluent monolayers were incubated in medium containing 8 g/d1 of y-globulin or dextrans with indicated molecular weights for 6 h, and then levels of mRNAs were measured. Average values of ratio of albumin mRNA to p-actin mRNA (open bars) or apo B mRNA to /3-actin mRNA (hatched bars) are shown as percentages of value without additions (n = 2). Viscosities of respective culture media are also shown. Dex(G), Dex(lZ), and Dex(48), dextrans of mol wt 6-9 X 104, 12 x lo”, and 48 x lo”, respectively; r-Glob(l5), y-globulin of mol wt 15 X 103; rpm, revolutions/min; so, scaled over; nd, not determined.
viscosity, was more important for regulating the levels of albumin and apo B mRNAs, we plotted the albumin and apo B mRNA levels against the oncotic pressure or viscosity (Fig. 4). Albumin mRNA levels, relative to dextran-free control as lOO%, were negatively correlated with the oncotic pressure (r = -0.806, P < 0.001). We also observed a negative correlation between apo B mRNA and the oncotic pressure (r = -0.737, P < 0.001). On the other hand, we could not observe any significant correlation between the viscosity and albumin or apo B mRNA level. From these results we conclude that the oncotic pressure is more important than the viscosity in regulating the steady-state levels of albumin and apo B mRNAs. Restoration of levels of albumin and apo B mRNAs by decrease in dextran concentration. Next we tested the effect of decrease in oncotic pressure on the suppressed levels of albumin and apo B mRNA. For this, confluent monolayers were preincubated in medium containing 8 g/d1 of dextran (mol wt 6-9 x 104) for 15 h at 37°C and then incubated in medium containing various concentrations (8, 4, 2, or 0 g/dl) of dextran for 6 h. The levels of albumin and apo B mRNAs were then measured. Figure 5 shows that the levels of albumin and apo B mRNAs increased inversely with the concentration (or oncotic pressure) of dextran in the incubation medium. Addition of actinomycin D (2 pg/ml) to the medium without dextran (0 g/dl) inhibited the increases in the levels of albumin and apo B mRNAs. Thus the steady-state levels of albumin and ano B mRNAs seem to be regulated mainlv
GENE
TRANSCRIPTION
at the transcriptional step. Effect of oncotic pressure produced by dextran on transcription rates of albumin and apo B genes. For direct evaluation of whether the oncotic pressure produced by dextran affects the transcriptional step of expression of the albumin and apo B genes, we measured the transcription rates of these genes by a nuclear run-on transcription assay. Confluent monolayers were incubated in medium containing 0,4, or 8 g/d1 of dextran (mol wt 6-9 x 104) for 4 h at 37°C and then the nuclei were isolated and examined by in vitro run-on transcription assay. The basal transcription rate of the albumin gene was lower than that of apo B gene in H4-II-E cells (Fig. 6). The addition of dextran decreased the transcription rates of the albumin and apo B genes concentration dependently, with a concentration of 8 g/d1 dextran decreasing the transcription rates of both genes m 50%. These results directly demonstrate that the change in oncotic pressure produced by dextran regulates the transcriptions of both genes. Effect of cycloheximide on transcription rates of albumin and apo B genes. The expressions of several genes are known to require ongoing protein synthesis (7, 24). Therefore we investigated whether protein synthesis was required for the regulation of transcription of the albumin, apo B, and P-actin genes. First, we measured the levels of albumin, apo B, and P-actin mRNAs of cells incubated for 6 h in medium with 0 or 8 g/d1 of dextran in the presence or absence of cycloheximide (5 pg/ml). In the absence of dextran, the presence of cycloheximide did not affect the level of albumin or apo B mRNA but increased the level of p-actin mRNA. In medium containing 8 g/d1 of dextran, the presence of cycloheximide increased the level of apo B mRNA markedly compared with the level in its absence and increased the albumin mRNA level slightly (Table 2). There are two possible mechanisms for these increases in mRNA levels: increase in the rate of synthesis of mRNA and decrease in the rate of degradation of mRNA. To distinguish between these possibilities, we measured the transcription rates of these genes directly by nuclear transcription assay (Fig. 6). H4-II-E cells were incubated for 4 h in medium containing 8 g/d1 of dextran in the presence or absence of cycloheximide. The transcription rate of the apo B gene was 3.0 times higher in the presence than in the absence of cycloheximide. The transcription rate of the albumin gene was 1.8 times higher in the presence than in the absence of cycloheximide and was similar to the basal transcription rate. The increase in P-actin gene transcription in the presence of cycloheximide was only -30%. DISCUSSION
The object of our study was to clarify the mechanism regulating the syntheses of albumin and apo B protein in response to alteration of oncotic pressure. To investigate this problem directly, we employed cultured H4-II-E cells and evaluated the effects of the oncotic pressure of the culture medium on the levels of albumin and apo B mRNAs. We used the H4-II-E cell line, derived from Reuber H-35 hepatoma, because this line has been found to be useful for studying the regulations of syntheses of albumin and other nroteins (22). We found that addition of
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 17, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
ONCOTIC
y = 98.058 p
