JOURNAL OF CELLULAR PHYSIOLOGY 153:288-296 I 1992)

Differential Regulation of Glucose Transporter 1 and 2 mRNA Expression by Epidermal Growth Factor and Transforming Growth Factor-Beta in Rat Hepatocytes D A V I D M I S C H O U L O N , BASABI RANA, NATAL10 KOTLIAR, PAUL F. PILCH, N A N C Y L.R. BUCHER, AND STEPHEN R. FARMER* Departments of Biochemistry (D.M., B.R., N.K., P.F.P., S.R.F.1 and Pathology (N.L.R.B.), Boston Univer5ity School of Medicine, Boston, Massachusetts 02 1 18 We have examined by Northern blot analysis the expression of two members of the glucose transporter family of genes (GLUT-1 and GLUT-2) in regenerating liver and in hepatocytes cultured under various conditions. GLUT-1, although thought to be a growth-associated gene, is not expressed in normal or regenerating liver, whereas GLUT-2, a liver-specific gene, is abundant in normal liver and gradually up-regulated during liver regeneration. Conversely, in hepatocytes cultured conventionally on dried rat tail collagen (RTC) in the premice of CGr and insulin, which potentiate proliferation, GLUT-1 mRNA is rapidly and abundantly expressed, whereas GLUT-2 i s depressed. To investigate the causes of this "switch" in glucose transporter expression seen when hepatocytes are removed from the liver and cultured under the conventional proliferative conditions, we examined the effects of specific growth factors and extracellular matrices on cultured hepatocytes. ECF, a potent liver mitogen, although causing a threefold induction of GLUT-1, was found to have no effect on GLUT-2 expression, suggesting that the increase in GLUT-? seen in regenerating liver is not due to EGF. Inhibition of protein synthesis by cycloheximide in cultured hepatocytes does not prevent the induction of GLUT-1 mRNA. In addition, treatment of cells with cyclohexiniide appears to stabilize the GLUT-2 mRNA, preventing the usual down-regulation of this gene in cultured hepatocytes. The expression of the two glucose transporter mRNAs also differed when the hcpatocytcs were adherent to particular cell matrices. Culture of hepatocytes on a reconstituted basement membrane gel matrix (EHS) is known to restrain their growth and mediate high levels of differentiated hepatocytic functions that are lost under conventional culture conditions. Unlike cells on RTC, hepatocytes on EHS expressed low levels of GLUT-1 mRNA, and decreased GLUT-2 mRNA. TGF-P, an attenuator of DNA synthesis, when added to cultures on RTC, substantially down-regulated GLUT-2 but had no effect on GLUT-1. We propose that the effectors, EGF, TGF-P and basement membrane components, play a significant role in the regulation of expression of GLUT-1 and GLUT-2 in hepatocytes. IO"/pg DNA by random priming using Klenow fragment (Bethesda Research Laboratories) and "P-dCTP (Amersham). Autoradiography was performed with Kodak X-Omat film and Cronex Lightening-Plus intensifying screens. Autoradiographic signals were quantified using a scanning laser densitometer (Pharmacia LKB Biotechnology). Plasmids used were: human GLUT-2 (Fukumoto et a]., 1988); rat GLUT-2 (Thorens et al., 1988);human GLUT-1 (Mueekler et al., 1985); albumin (a gift from Douglas Cooper and Harvey Lodish); c-jun (Ryder et al., 1988); mouse histone genomic clone (pRAH3.2) (Delisle et al., 1983).

MISCHOULON ET AL.

290

A

Regeneration Time (Hrs) 4

0

16

96

Time in Culture ( H r s ) 0

5

24

B

GLUT 2

GLUT 1 Albumin

B

a Sh

24

Hx

Sh

Hx

GLUT 2

Histone

Albumin

Experiments were carried out at least three times and data shown represent typical results.

RESULTS Expression of GLUT-1 and GLUT-2 in regenerating liver and in cultured hepatocytes To explore how glucose transporters respond to normal endogenous growth factors in the physiological environment within the animal, we performed Northern blot analysis on regenerating livers a t intervals after partial hepatectomy. We found that GLUT-1 expression was not induced during the first 96 hr, a period during which proliferative activity rises manyfold from essentially nil to reach a sharp peak by 24 hr, followed by a steep, then gradual decline. GLUT-2 slowly increases during this period, attaining several times its normal level by 96 hr (Fig. 1A). During the course of our studies, Yamada et al. (1990) published a similar result demonstrating no change in GLUT-1 expression and a 2-3-fold induction in GLUT-2 mRNA levels between 4-12 h r posthepatectomy. To ascertain that the increase in GLUT-2 expression is due to liver growth and not to the stress of the surgery, we subjected a series of control rats to sham operations. Figure 1B shows that by 8 h r posthepatectomy, the amount of GLUT-2 mRNA expressed is greater than in the corresponding sham timepoint, and this effect is even greater by 24 hr. Note the extensive in-

Fig. 1. Expression of glucose transporter mRNA in regenerating liver and in hepatocytes cultured under conventional conditions. A. Total RNA was isolated from livers at the indicated times following a 213 partial hepatectomy, or in hepatocytes cultured on dried rat tail collagen in the presence of EGF and insulin. Total RNA (25 pg) was electrophoresed through a 1% agarose gel containing 610 formaldehyde and 0.5 pg EtBr. Northern blot transfer and filter hybridization were carried out according to Bond and Farmer (1983).B. Northern blot of total RNA isolated from regenerating livers (Hx) at the indicated times and from the corresponding sham controls (Sh). The mRNAs shown correspond to cDNAs used: rat GLUT-2, human GLUT-1, histone, albumin. Note: B = rat brain mRNA, used as a positive control for GLUT-1, and as ii negative control for GLUT-2.

duction of histone 3.2 mRNA a t this time indicating the transition of hepatocytes through S phase of the cell cycle. To examine effects of specific growth factors, we turned to freshly isolated normal hepatocytes cultured on dry collagen in the presence of EGF and insulin, conventional conditions that potently stimulate DNA synthesis. Under these conditions, expression of the two glucose transporters was completely reversed when compared to regenerating livers, with GLUT-1 rising dramatically within a few hours and GLUT-2 declining significantly during the initial 24-hr period (Fig. 1A). Albumin mRNA levels remained essentially unchanged both during liver regeneration and during these short-term hepatocyte cultures, as previously documented by ourselves (Bucher et al., 1990) and others (Clayton and Darnell, 19831, thereby demonstrating equal lane loading. Note that all Northern blots shown here and elsewhere are representative of three or more experiments.

EGF stimulates expression of GLUT-1 mRNA in cultured hepatocy tes For further assessment of glucose transporter activity in hepatocyte proliferation, we focused on EGF because it has been shown to induce GLUT-1 mRNA in mouse fibroblast cultures, a response that is characteristic of the immediate early genes (Hiraki et al., 1988).

29 1

GLUCOSE TRANSPOKTER 1 AND 2 mRNA EXPRESSION

A Time in Culture (Hours) EGF-+

0

-I-

-

2

+

-

3

-

4

-

+ 24

-

L

B

GLUT 2 GLUT 1 Albumin

B 7r

I

0

2

3

4

6

8

N

L

TIME (HRS)

Fig. 2. The effect of EGF on the expression of glucose transporter mRNA in cultured hepatocytes. A. Liver was perfused for -10 min, removed, and hepatocytes were isolated, filtered through a nylon membrane, and suspended in Williams E plating medium (WEiIEPj. EGF was either present ( + I at a concentration of 10 ng/ml or absent (pi.Hepatocytes were plated a t a concentration of 350 x 10:’ cellsil.5 ml of WE/IEP onto Lux 100 mm tissue culture dishes coated with dried rat tail collagen. Cells were maintained at 37”C,in air + 5%,CO,, and harvested at the indicated times. Total RNA (25 Fg) isolated from hepatocytes was subjected to Northern blot analysis as in Figure 1. Note: L = normal liver mRNA; B = rat brain mRNA. B. The intensity of the GLUT-1 autoradiographic signal was quantified by laser densitometry and normalized to the albumin signal. HistobTam includes timepoints not shown in A.

-EGF

In addition, EGF may have a growth promoting role in liver regeneration, although this is not incontrovertible (Fausto, 1989; Marti et al., 1989). Figure 2 shows that EGF potently stimulated expression of GLUT-1 mRNA in the hepatocyte cultures, resulting in a threefold enhancement within 2 h r compared to control conditions lacking EGF, and the difference persisted for the duration of the 24-hr culture period. Effects on GLUT-2 were negligible. It is important to note that the expression of GLUT 1 mRNA in culture cannot be accounted for by the presence of contaminating nonparenchymal cells. In fact, Northern blot analysis of total RNA isolated from hepatocytes cultured for 72 h r reveals negligible amounts of vimentin mRNA (a marker for nonhepatocytes) when compared to a n equivalent quantity of total fibroblast RNA (unpublished data).

Interestingly, inhibition of hepatic protein synthesis did appear to slow down the normal rate of decrease of GLUT 2 mRNA during this short-term culture period. As a control for adequacy of drug dosage, we analyzed the expression of c-jun, a gene that is known to be superinduced in hepatocytes by cycloheximide (Mohn et al., 1990). As shown in Figure 3, the drug dosage was highly effective in superinducing c-jun mRNA, while having a limited effect on the induction of GLUT 1 mRNA.

Adhesion of hepatocytes to a reconstituted basement membrane (EHS gel) attenuates the induction of GLUT-1 mRNA in culture Hepatocytes in the artificial environment of a culture are influenced not only by the growth factors and other soluble components of the culture medium but Induction of GLUT 1 mRNA is not dependent on also to a large extent by the substratum on which they prior protein synthesis are plated. On dried collagen they soon lose the ability To determine whether the induction of GLUT 1 t o transcribe liver-specific genes (i.e., genes coding for mRNA was a n immediate response to EGF activity, we other than their own maintenance of “housekeeping” blocked protein synthesis in freshly isolated hepato- functions) and also the capacity to manifest the normal cytes with cycloheximide. Addition of 10 Fgiml of this phenotype (Clayton and Darnell, 1983).On EHS, howdrug t o cells during the first 10 h r of culture did not ever, these properties are maintained to a far greater affect the ability of EGF to induce the expression of extent (Ben-Ze’ev et al., 1988). GLUT 1 mRNA; however, it did slightly dampen the We examined expression of the glucose transporters extent of activation at the 10-hr time point (Fig. 3). in hepatocytes cultured for up to 44 hr, a time by which

292

MlbCHUULUN KI AL.

Cycloheximide+

NL

0

-

-

+

2 -

+

6

-

+

10

-

GLUT 1

GLUT 2

c-jun

Albumin

Fig. 3. Effect of cycloheximide on expression of GLUT-1 and GLUT-2 mRNA in hepatocytes cultured in growth medium. Northern blot analysis of total RNA (25 pg per lane) isolated from hepatocytes cultured under conventional conditions as described in Figure 1 in the presence ( + ) or absence ( - ) of 10 pgiml cycloheximide. Blot was probed with GLUT-1, GLUT-2, c-jun, and albumin cDNAs. NL = RNA fi-om normal rat liver; 0 = RNA from freshly isolated hepatocytes prior to plating.

Oh

lh

5h

44h

GLUT 2

GLUT 1

Albumin Fig. 4. Expression of glucose transporter mRNA in hepatocytes cultured on a reconstituted basement membrane gel matrix (EHS).Hepatocytes were isolated as in Figure 2A and plated on either dried rat tail collagen (C) or EHS gel (EX Cells were harvested at indicated times after plating. Total KNA (2.5 +g) was subjected to Northern blot analysis as in Figure 1.

drastic changes in cell behavior have occurred, as our earlier studies have shown (Ben-Ze’ev et al., 1988). For the cells cultured on collagen for 5 h r in the presence of EGF and insulin, the levels of GLUT-1 mRNA were increased manyfold, whereas those of GLUT-2 were decreased (Fig. 4).By 44 h r of culture, GLUT 1mRNA has decreased significantly from the peak levels reached at 5 hr, but the levels are still elevated above those detected in the normal liver. GLUT 2 mRNA gradually decreases during the entire culture period to levels that are manyfold lower than those in liver. On a reconstituted basement membrane gel matrix (EHSI, the expression of GLUT-1 mRNA was significantly lower

than that observed on RTC at all times analyzed. GLUT-2 mRNA levels were also lower than those seen on RTC (Fig. 4). The inhibition of GLUT-1 expression by the EHS matrix is consistent with the notion that culture of hepatocytes under these conditions will maintain a nonproliferative, highly differentiated state (Ben-Ze’ev et al., 1988), but the decreased expression of GLUT-2 was unanticipated and not consistent with this concept.

TGF-P inhibits expression of GLUT-2 mRNA in cultured hepatocytes EHS is known to contain TGF-P, which, even if initially present in a n inactive state, is probably activated by proteases released from cultured hepatocytes and could conceivablv influence glucose transDorter exDression. TGF-P is also known tz inhibit hepitocyte pGoliferation both in liver regenerating within the animal (Braun et al., 1988) and in cultures (Nakamura et al., 1985; Carr et al., 1986; McMahon et al., 1986). To determine possible effects of TGF-@on glucose transporter expression in hepatocyte cultures, we examined the effect of a dose a t least 10-fold greater than needed to inhibit DNA synthesis (Nakamura et al., 1985; Carr et al., 1986; McMahon et al., 1986). TGF-P was added to hepatocytes cultured on RTC (Fig. 5 ) and the results quantified by laser densitometry (Fig. 6). Figure 5A,B shows that TGF-P inhibits GLUT-2 mRNA expression without affecting GLUT-1 or albumin mRNA levels. The low level of histone 3.2 mRNA expression a t 4 2 4 8 h r (Fig. 5A) in the presence of TGF-P clearly shows the inhibition of entry into S-phase during this time period. The GLUT-2 signals visualized in these and additional Northern blots were subjected to quantitation by laser densitometry. Different autoradiographic exposures were obtained to ensure that the intensity of the signal was within the linear range of detection by the

GLUCOSE TRANSPORTER 1 AND 2 mRNA EXPRESSION

A. TGF-6

-

0

42

5

48

- + - + - +

B

GLUT 1 GLUT 2

293

Fig. 5. Effect of TGF-P on the expression of' glucose transporter mRNAs in hepatocytes cultured in growth medium. A. Hepatocytes were isolated as in Figure 2A, and plated on dried rat tail collagen in the presence of EGF and insulin. TGF-6 was either present ( + ) a t a concentration of 2 ngiml or absent (-). Cells were harvested at indicated times. Total RNA (25 pg) from hepatocytes representing two separate experiments (A and B) was subjected to Northern blot analysis as in Figure 1. In A the ethidium bromide stained blot demonstrates equal loading of RNA per lane; B = brain RNA. In B the blot was probed with an albumin cDNA in wrder to assess for equal loading of RNA.

Histone 1

2

3

4

5

6

7

8

1 2 3 4 5 6 7 8

28S-

18s-

24

NL

TGF-B

+

48

-

+

72

-

+

-

GLUT 2

Albumin

densitometer. The absorbance data are presented as a series of histograms shown in Figure 6. I t is apparent that GLUT-2 expression decreases by -75%-90% during the culture of hepatocytes in the presence of EGF and insulin without TGF-P during the 48-72-hr period. Addition of 2 ngiml of TGF-P down-regulates the expression of GLUT-2 mRNA by a n additional 50%-75% during this time.

DISCUSSION Activation of quiescent hepatocytes into the cell cycle during compensatory hyperplasia resulting from a 213 hepatectomy is accompanied by the induction of several members of the family of immediate early growth responsive genes (Thompson et al., 1986; Mohn et al., 1990). This observation is consistent with the notion that the transit of hepatocytes from Go into GI phase may depend on induction of a specific set of genes. The transition can be brought about in cultures of several different cell types by various growth factors, including EGF. Moreover, studies in the animal have suggested that activation of the EGF receptor, either by EGF itself or by TGF-alpha, is a n early requisite event in liver regeneration (Mead and Fausto, 1989). The recent demonstration by Hiraki et al. (1988) that EGF as well a s

other growth factors can activate GLUT-1 expression in quiescent mouse fibroblasts, thus raising the possibility of its inclusion in the early growth-response family, encouraged us to find whether expression of this gene is activated during liver regeneration, along with other immediate early growth genes such a s c-fos and c-jun. To our surprise, GLUT-1 mRNA expression did not respond to the burst of hepatocyte growth following partial hepatectomy (Fig. l ) , whereas the bidirectional liver glucose transporter, GLUT-2, was augmented during the regeneration process. The kinetics of increase in GLUT-2 mRNA, seen t o occur between 16-96 hr, suggest that GLUT-2 mRNA is not regulated along with the immediate early growth genes, since the induction of these mRNAs occurs within the first 1-3 hr of regeneration (Thompson et al., 1986; Mohn et al., 1990). EGF was found to have a negligible effect on GLUT-2 mRNA expression in cultured hepatocytes, but did, however, cause a significant increase in GLUT-1 expression (Fig. 2 ) . These observations are not consonant with a n earlier report (Rhoads et al., 1988) that EGF at a concentration of 10 pg/ml(l,OOO-foldgreater than the concentration we used) had no effect on GLUT-1 mRNA expression in cultured hepatocytes, but the data and

MISCHOULON ET AL.

294

A.

B.

6 1

slw

x2

W

u z 4

m

a

a 0

$

v)

9

m 4

0

=

5

(-)TGF-B

17 29 TIME (HRS)

42

48

(+)TGF-B

24

=

48

72

TIME (HRS) (-ITGF-B

(+)TGF- B

Fig. 6. Quantitation of GLUT-2 mRNA levels in hepatocytes cultured in the presence or absence of TGF-p. The intensities of the GLUT-2 autoradiographic signals shown in Figure 5 A and B (and additional tirnepoints) were quantified by laser densitometry; the intensity of the GLUT-2 signal in (B) was normalized to the albumin signal.

the specific experimental conditions are not presented in sufficient detail to permit comparative evaluation. We can only speculate that differences in experimental conditions are the likely source of the divergence. We have noted, e.g. (unpublished data), that the widely used collagenase perfusion procedure used to isolate hepatocytes itself potently induces the abundant expression of certain immediate early growth genes, including c-fos, c-myc, jun B, and c-jun, but not GLUT-1. Its mRNA, however, increases very early during hepatocyte culture, even in the absence of EGF. If GLUT-1 is not activated during hepatocyte proliferation in the animal, why do primary cultures of hepatocytes express such high levels of GLUT-1 mRNA in response to EGF? A recent study by Wertheimer et al. (1991) suggests that GLUT-1 belongs to the glucose regulated protein family of stress-inducible genes (GRPs). The initial induction of GLUT-1 during culture could conceivably result from the activation of these GRPs by the stress of the culture conditions. The major surgical stress imposed by partial hepatectomy, which does not induce GLUT-1, seems to preclude a t least this form of stress as a GLUT-1 inducer, but some form of metabolic stress is not excluded. Culture of hepatocytes under conditions that facilitate extensive spreading and proliferation, i.e., on dried rat tail collagen with EGF and insulin, normally results in a n extensive decrease in liver-specific gene expression (Clayton and Darnell, 1983; Ben-Ze’ev et al., 1988; Bucher et al., 1990). It is not surprising, therefore, that we observe a decrease in GLUT-2 mRNA expression during a 2-%day culture period (Fig. 5) under these conventional conditions. One possible reason for this down-regulation of hepatic function is the disruption of the appropriate hepatocyte morphology and cell-cell and cell-extracellular matrix interactions during cell isolation and culture. These physical associations most likely influence the expression of a series of liver-specific transcription factors that regulate ex-

pression of GLUT-2. Culture of hepatocytes on the EHS gel, however, while maintaining expression of most liver-specific functions (Ben-Ze’ev et al., 19881,failed to maintain GLUT-2 expression. The studies outlined in Figures 4-6 strongly suggest that this is due to the presence of biologically active TGF-P in the EHS gel preparation. The gel is known to contain very high concentrations (in the range of 10 ng/ml) of immunoreactive TGF-f3that is considered likely to exist in a latent, inactive form (Hynda Kleinman and Anita Roberts, personal communication). It is possible that significant amounts of the active peptide are also present or that release of proteases from the cultured hepatocytes may activate the latent form. Adhesion of hepatocytes to the basement membrane gel also inhibited the expression of GLUT 1 mRNA. This response does not appear to result from the presence of TGF-p in the EHS gel environment since addition of this factor to hepatocytes cultured on collagen has no effect on GLUT 1 mRNA expression. Other studies by us (Rana, Xie, Bucher, and Farmer, unpublished data) indicate that adhesion to EHS arrests hepatocyte growth in Go phase of the cell cycle and in doing so inhibits the expression of the immediate early genes, including GLUT 1. TGF-f3, however, appears to inhibit hepatocyte proliferation in late G, and has no effect on the immediate early genes. The mechanisms that control the differential expression of these two glucose transporter mRNAs in response to these effectors may likely include both transcriptional and posttranscriptional processes. The growth factor induction of GLUT-1 found in fibroblasts is reported to result primarily from a n increase in transcriptional activity of the gene that is not dependent on prior protein synthesis (Hiraki et al., 19881, and it seems likely that a similar mechanism exists for its EGF dependent activation in hepatocytes (Fig. 3). With regard to the inhibition of GLUT-2 in the hepatocyte cultures by TGF-P, however, it must be borne in mind

GLUCOSE TRANSPORTER 1 AND 2 mRNA EXPRESSION

that TGF-P can alter the expression of many genes by affecting different levels of regulation. For example, TGF-P has been shown to down-regulate the metalloproteinase gene, stromelysin, via activation of a fosijun complex that inhibits transcription of the gene (Kerr et al., 1990). The two isoforms of glucose transporter differ in various ways, as noted, and appear to subserve different metabolic functions, although both are facilitative glucose carriers that accelerate glucose diffusion across cell membranes by a n energy independent process (Bell e t al., 1990). GLUT-1, which is a high affinity type of transporter, is saturated at physiological blood glucose concentrations (5-10 mM) (Wheeler and Hinkle, 1985) and is primarily involved in cellular uptake of glucose from the plasma (Thorens et al., 1990a). GLUT-1 is restricted to a single cell layer around the hepatic venule, where nutrients may be in reduced supply so that emergency import of glucose is needed, a s evidenced by the GLUT-l induction in a few additional perivenous cells during prolonged fasting and its reversal on glucose refeeding (Tal et al., 1990; Thorens et al., 1990b). The perivenous hepatocytes may not be in the same state of differentiation as the others, as they differ metabolically in other respects as well (Jungermann et al., 1982; Gebhardt and Mecke, 1983). GLUT-2 is expressed in all hepatocytes. Unlike GLUT-1, GLUT-2 is a low affinity transporter and although bidirectional (Craik and Elliott, 19791, it is most abundant in cells that are capable of gluconeogenesis and that primarily export glucose into the plasma. The primary function of GLUT-2 seems to be to maintain plasma glucose homeostasis. This is probably the most vital of the multitudinous functions of the liver, as the brain metabolizes only glucose, except during prolonged starvation, when i t can metabolize ketone bodies. The hepatic stores of glycogen and high gluconeogenic potential, in conjunction with the increase in GLUT-2, appear sufficient to maintain blood glucose levels during liver regeneration despite the excision of two-thirds of the normal complement of hepatocytes. In cultures, in contrast, where glucose levels are constant, GLUT-2 is slightly down-regulated, probably due to a general loss of liver-specific control of gene expression. The metabolic roles of GLUT-1 and GLUT-2 will be more sharply defined when the molecular mechanisms through which their effectors operate become clearer.

ACKNOWLEDGMENTS We are grateful to Dr. Graeme Bell for providing the human GLUT-1 and GLUT-2 cDNAs, and to Drs. Bernard Thorens and Harvey Lodish for providing the rat GLUT-2 cDNA. We also thank our research associates Yu-Hong Xie, Fei Lin, and Babette Radner for valuable assistance and suggestions. D.M. acknowledges the support of the Charles E. Culpeper Foundation and the Karin Grunebaum Cancer Research Foundation. This work was supported by grants #CA39049 and #DK30425 from the National Institutes of Health and #2237 from the Council for Tobacco Research, and by a grant from the American Diabetes Association.

295

LITERATURE CITED Bell, G., Kayano, T., Buse, J.B., Burant C.F., Takeda, J., Lin, D., Pukumoto, H., and Seino, S. (1990) Molecular biology of mammalian glucose transporters. Diabetes Care, 13:198-208. Ben-Ze'ev, A,, Robinson, G.S., Bucher, N.L.R., and Farmer, S.R. (1988) Cell-cell and cell-matrix interactions differentially regulate the expression of hepatic and cytoskeletal genes in primary cultures of rat hepalocytes. Proc. Natl. Acad. Sci. USA, 85t2161-2165. Birnbaum, M.J., Haspel, H.C., and Rosen, O.M. (1986) Cloning and characterization of a cDNA encoding the rat brain glucose-transporter protein. Proc. Natl. Acad. Sci. USA, 83:5784-5788. Bond, J.F., and Farmer, S.R. (1983) Regulation of tubulin and actin mRNA production in rat brain: expression of a new (J-tubulin mRNA with development. Molec. Cell Biol., 3:1333-1342. Bucher, N.L.R., Robinson, G.S., and Farmer, S.R. (1990) Effects of extracellular matrix on hepatocyte growth and gene expression: lmplications for hepatic regeneration and the repair of liver injury. Seminars in Liver Disease, 10t11-19. Carr, B.I., Hayashi, I.?Branum, E.L., and Moses, H. (1986) Inhibition of DNA synthesis in rat hepatocytes by platelet-derived type p transforming growth factor. Cancer Res., 46.2330-2334. Chomczynski, P., and Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal. Biochem., 162:15&159. Clayton, D.F., and Darnell, J.E. (19831 Changes in liver-specific compared to common gene transcription during primary culture of mouse hcpatocytcs. Mol. Cell. Biol., 3:1552-1561. Craik, J.D., and Elliott, K.R.F. (1979) Kinetics of 3-0-methyl-D-glucose transport in isolated rat hepatocytes. Biochem. J., 182.503508. Delisle, A.J., Graves, R.A., Marzluff, W.F., and Johnson, L.F. (1983) Regulation of histone mRNA production and stability in serumstimulated mouse 3T6 fibroblasts. Mol. Cell. Riol., 3:1920-7929. Fausto, N. (1989) Hepatic Regeneration. In: Hepatology, 2nd ed. D. Zakim, T.D. Boyer, eds. W.B. Saunders, London, pp. 4 9 4 5 . Flier, J.S., Mueckler, M., McCall, A.L., and Lodish, H.F. (1987) Distribution of glucose transporter messenger RNA transcripts in tissues of rat and man. J. Clin. Invest., 7Yt657-661. Fukumoto, H., Seino, S., Imura, H., Seino, Y . ,Eddy, R.L., Fukushima, Y., Byers, M., Shows, T.B., and Bell, G.I. (1988)Sequence, tissue distribution, and chromosomal localization of rnRNA encoding a human glucose transporter-like protein. Proc. Natl. Acad. Sci. USA, 86.54346438.

Gebhardt, R., and Mecke, D. (19831 Heterogenous distribution of glutamine synthetase among rat liver parenchymal cells in situ and in primary culture. EMBO J.,2567-570. Gould, G.W., and Bell, G.I. (1990) Facilitative glucose transporters: An expanding family. Trends Bioch. Sci., 15:1%23. Higgins, G.M., and Anderson, R.M. (1931) Experimental pathology of the liver. I. restoration of the liver of white rat following partial surgical rcmoval. Arch. Pathol., 12r186202. Hiraki, Y., Rosen, O.M., and Birnbaum, M.J. (1988) Growth factors rapidly induce expression of the glucose transporter gene. J. Biol. Chem., 2(;,'~,19655-13662. Jungermann, K., Heilbronn, R., Katz, N., and Yasse, D. (1982) The glucose/glucosc-6-phosphate cycle in the periportal and perivenous zone of rat liver. European J. Biochem., 123:429436. Kerr, L.D., Miller, D.B., and Matrisian, L.M. (1990)TGF-p1 inhibition of transinistromelysin gene expression is mediated through a fos binding sequence. Cell, 61267-278. Marti, U., Burwen, S.J., and Jones, A.L. (1989) Biological effects of epidermal growth factor, with emphasis on the gastrointestinal tract and liver: An update. Hepatology, 9:126-138. McMahon, J.B., Richards, W.L., del Campo, A.A., Song, M.K.H., and Thorgeirsson. S.S.119861Differential effects of transforming growth factor-p on proliferation of normal and malignant rat liver epithelial cells in culture. Cancer Rcs., 46:46654671. Mead, J.E., and Fausto, N. (1989)Transforming growth factor 01 may be a physiological regulator of liver regeneration by means of an autocrine mechanism. Proc. Natl. Acad. Sci. USA, 86:155%1562. Mohn, K.L., Laz, T., Melby, A.E., and Taub, R. (1990)Immediate-early gene expression differs between regenerating liver, insulin-stimulated H-35 cells and mitogen-stimulated Balb/c3T3 cells. J. Biol. Chem., 265t21916-21921. Mueckler, M., Caruso, C., Baldwin, S.A., Panico, M., Blench, M., Morris, H.R., Allard: W.J., Lienhard, G.E., and Lodish, H.F. (1985) Sequcncc and structure of a human glucose transporter. Science, 229:941-945.

296

MISCIIOULON ET AL

Nakamura, T., Tomita, Y., Hirai, R., Yamaoka, K., Kaji, K., and Ichinara, A. (1985)Inhibitory effect of transforming growth factor p on DNA synthesis of adult rat hepatocytes in primary culture. Bioch. Biophys. Res. Comm., I33:104%1050. Rhoads, D.B., Takano, M., Gattoni-Celli, S., Chen, C-C, and Isselbacher, K.J. (1988)Evidence for expression of the facilitated glucose transporter in rat hepatocytes. Proc. Natl. Acad. Sci. USA, 85:90429046. Russell, W.E., Coffey, R.J., Oullette, A.J., and Moses, H.L. (1988). Type p transforming growth factor reversibly inhibits the early proliferative response to partial hepatectomy in the rat. Proc. Natl. Acad. Sci. USA, 855126-5130. Ryder, K., Lau, L.F., and Nathans, D. (1988) A gene activated by growth factors is related to the oncogene v-jun. Proc. Natl. Acad. Sci. USA, 85:1487-1491. Seglen, P.O. (1976) Preparation of isolated rat liver cells. Meth. Cell. Biol., 13.2943. Tal, M., Schneider. D.L., Thorens, B., and Lodish, H.F. (1990) Restricted expression of the erythroidibrain glucose transporter isoform to perivenous hepatocytes in rats. J.Clin. Invest., 86986-992. Thompson, N.L., Mead, J.E., Braun, L., Goyette, M., Shank, P.R., and Fausto, N. (1986) Sequential protooncogene expression during rat liver regeneration. Cancer Res., 46t3111-3117. Thorens, B., Sarkar, H.K., Kaback, H.R., and Lodish, H.F. (1988)

Cloning and functional expression in bacteria of a novel glucose transporter present in liver, intestine, kidney, and P-pancreatic islet cells. Cell, 55:281-290. Thorens, B., Cheng, Z.Q., Brown, D., and Lodish, H.F. (1990a) Liver glucose transporter: a basolateral protein in hepatocytes and intestine and kidney cells. Amer. J. Physiol., 259:C279-C285. Thorens, B., Flier, J.S., Lodish, H.F., and Kahn, B.B. (1990b)Differential regulation of two glucose transporters in rat liver by fasting and refeeding and by diabetes and insulin treatment. Diabetes, 39:712719. Umek, R.M., Friedman, A.D., and McKnight, S.L. 11991) CCAATEnhancer Binding Protein: A component of a differentiation switch. Science,251288-292. Wertheimer, E., Sasson, S., Cerasi, E., Ben-Neriah, Y. (1991) The ubiquitous glucose transporter GLUT-1 belongs to the glucose-related protein family of' stress-inducible proteins. Proc. Nat'l. Acad. Sci. USA, 88r2525-2529. Wheeler, T..J., and Hinkle, P.C. (1985) The glucose transporter of mammalian cells. Annu. Rev. Physiol., 47.503-517. Yamada, Y., Seino, Y., Takeda, J., Fukumoto, H., Yano, H., Inagaki, N., Fukuda, Y., Seino, S.,and Imura, H. (1990) Increase in liver glucose transporter mRNA levels during rat liver regeneration. Bioch. Biophys. Res. Comm., 168:127&1279.