- Email: [email protected]
Developmental and hormonal regulation of carbamoyl-phosphate synthase gene expression in rat liver: Evidence for control mechanisms at different levels in the perinatal period
Biochimica et Biophysica Acta 866 (1986) 61-67 Elsevier
61
BBA91562
Developmental and hormonal regulation of carbamoyl-phosphate synthase gene expression in rat liver: evidence for control mechanisms at different levels in the perinatal period C.J. d e G r o o t , D. Z o n n e v e l d , R . T , M . de Laaf, M.A. D i n g e m a n s e , P.G. M o o r e n , A . F . M . M o o r m a n , W . H . L a m e r s * a n d R. C h a r l e s Department of Anatomy and Embryology, University of Amsterdam, AMC, Meibergdreef15, 1105 A Z Amsterdam (The Netherlands) (Received October 15th, 1985)
Key words: Carbamoyl-phosphate synthase; Gene expression; Hormonal regulation; (Rat liver)
Carbamoyl-phosphate synthase gene expression is found to be primarily regulated by conditions that enhance hepatic glucocorticosteroid levels (hormone injections) and cyclic AMP levels (induction of diabetes). After birth, changes in the level of carhamoyl-phosphate synthase protein follow changes in the level of carbamoylphosphate synthase mRNA, suggesting a pretranslational control mechanism. In fetal rats, carbamoyl-phosphate synthase gene expression is regulated by the same factors as in adults. However, both the level to which carbamoyl-phosphate synthase mRNA can accumulate and the extent to which mRNA can be translated appear to be limited, indicating control mechanisms at the pretranslational and translational level. Finally, in the immediate postnatal period, a transient but pronounced decrease in the rate of degradation of carbamoyl-pbosphate synthase protein may play a role in the accumulation of the enzyme.
Introduction Carbamoyl-phosphate synthase (EC 6.3.4.16) is an abundant hepatocyte-specific protein [1,2]. It offers an attractive model system for studying the developmental regulation of gene expression, as both the timing and the site of expression within the liver are under developmental control. Rat hepatocytes acquire the capacity to synthesize carbamoyl-phosphate synthase as soon as they differentiate from the embryonic foregut [3], but start to accumulate the enzyme only shortly before birth [4]. After birth, the adult capacity to synthesize carbamoyl-phosphate synthase is acquired [5], as is
* To whom correspondence should be addressed.
the typica! periportal localization [6]. The factors that regulate carbamoyl-phosphate synthase gene expression at the protein level have been extensively studied, showing glucocorticosteroids and cyclic AMP to be the primary regulatory factors [3,4,7,8]. A preliminary exploration showed that the same factors affected carbamoylphosphate synthase mRNA levels [9]. In the present study we have analyzed further the role of hormones in regulating the levels of carbamoylphosphate synthase mRNA. Furthermore, we have established a detailed developmental profile of carbamoyl-phosphate synthase mRNA and have compared it to that of carbamoyl-phosphate synthase protein [10]. The results suggest that in the perinatal period, both pretranslational and translational control mechanisms may be operative.
0167-4781/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)
62 Materials and Methods
Animals. Wistar rats, obtained from the TNO animal farm in Zeist, The Netherlands, were used. Fetal age was estimated from the time of conception and was accurate to within 12 h; birth normally occurred between 21.5 and 22.5 days after conception. Neonatal age was calculated from the time of birth. Rats were weaned at 4 weeks post partum. Treatment of animals, For hormonal treatment, either 3-4-month-old adult or 19-day fetal rats were used. Diabetes was induced in adult rats by intraperitoneal injection of streptozotocin (70 m g / k g body weight) after an overnight fast. 24 h later, the animals were checked for glucosuria. Bilateral adrenalectomy and gonadectomy were performed under ether anesthesia and checked for completeness at death. Glucose (5 g / k g body weight) was administered to 24-h-starved rats by gastric gavage. Hormone-treated adult rats received triamcinolone acetonide (4 m g / k g body weight) by intraperitoneal injection. Hormonetreated fetuses received a cocktail of dexamethasone (2.5 m g / k g body weight), triiodothyronine (3.3 m g / k g body weight), dibutyryl cyclic AMP (50 m g / k g body weight) and theophylline (50 m g / k g body weight) by intraperitoneal injection, after laparotomy of the dam. Although triiodothyronine does not have an effect on carbamoyl-phosphate synthase mRNA (see Results) or protein [7] in adult animals, it was included in this cocktail as the hormone does have a stimulatory effect on carbamoyl-phosphate synthase protein before birth [7]. Determination of RNA content. The total RNA content of liver was determined by the orcinol method [11]. RNA was isolated from liver homogenates in guanidinium thiocyanate either by differential precipitation (Ref. 12, as modified in Ref. 9) or by centrifugation through a CsCI cushion [12]. Concentrations were calculated from the absorbance at 260 nm. The RNA samples were diluted in a high-salt buffer (1.8 M NaC1/100 mM N a H 2 P O / 1 0 mM EDTA (pH 7.4)) and aliquots were passed through nitrocellulose filters using a commercial manifold. After heating at 80°C for 2 h, filters were prehybridized and hybridized at 42°C, essentially as
described [13]. The cDNA probes for carbamoylphosphate synthase mRNA and albumin mRNA [9] and phosphoenolpyruvate carboxykinase mRNA [14] were labeled by nick translation to a specific activity of approx. 10 ~ cpm//~g. In more recent experiments a nearly full-length carbamoyl-phosphate synthase cDNA clone was used (designated pBR-CPS5; insert length, 5.5 kb). After washing, autoradiographs of the filters were made. For quantitative analysis, spots were cut out and radioactivity was determined by liquid scintillation counting. In the graphs and tables, means _+ S.E. of 3-7 independent estimates are given. In Fig. 1, the standard error of the ratio of carbamoyl-phosphate synthase mRNA and carbamoyl-phosphate synthase protein was calculated with the delta method. Results
Table I describes the effects of long-lasting changes in the levels of glucocorticosteroids and cyclic AMP on the level of carbamoyl-phosphate synthase mRNA. Durable changes in hepatic cyclic AMP concentration were induced by diabetes, starvation, and starvation followed by glucose refeeding [15,16]. As a control, albumin mRNA levels were estimated as well. It is clear that administration of glucocorticosteroids (triamcinolone) enhances carbamoyl-phosphate synthase m R N A levels more than 2-fold after 24 h, whereas a decrease in the levels of circulating glucocorticosteroids as caused by adrenalectomy decreases the mRNA levels to 60% after 96 h. In intact animals, additional glucocorticosteroids have no effect on albumin mRNA levels; however, albumin mRNA levels decrease 2-fold upon removal of the adrenals. Glucocorticosteroids may therefore be necessary for the maintenance of 'normal' levels of albumin mRNA (cf. Ref. 17). Elevated hepatic cyclic AMP levels, as seen in diabetic animals, are associated with a more than 2-fold increase in carbamoyl-phosphate synthase mRNA levels after 24 h. Furthermore, the effects of glucocorticosteroids (triamcinolone treatment) and cyclic AMP (diabetes) are additive, causing a 3.5fold increase after 24 h. Despite the continued presence of the inducing factors, their effects on
63
carbamoyl-phosphate synthase mRNA levels are transient, showing a 30-40% decrease when effects at 96 h are compared with those at 24 h. No effects of thyroid hormones on carbamoyl-phosphate synthase mRNA levels were observed (data not shown). The diabetic condition causes a decrease in albumin mRNA levels (cf. Ref. 18). Apparently, glucocorticosteroids are able to counteract some of the deleterious effects of diabetes on albumin synthesis. Starvation slightly decreases carbamoyl-phosphate synthase mRNA levels, while glucose refeeding restores the mRNA levels to normal. Starvation for 24 h has little effect on albumin mRNA levels. Fig. la shows the developmental profile of the ratio of the levels of hepatic carbamoyl-phosphate synthase mRNA and total RNA. Between 5 days before birth, when only a very small amount of m R N A can be detected, and 1 day before birth, carbamoyl-phosphate synthase mRNA levels increase 4-fold. Between this time and birth, mRNA levels remain constant, whereas in the first few postnatal days the mRNA levels drop to 50% of those at birth. Thereafter, mRNA levels incrase to attain maximum levels during the weaning period (3rd and 4th postnatal weeks). Adult mRNA levels are approx. 40% of those at weaning. In order to find the relationship between carbamoyl-phosphate synthase mRNA levels and carbamoyl-phosphate synthase protein levels it was necessary to
determine the total RNA content per gram liver (Fig. lb). After birth, the total RNA level was found to be 8-9 m g / g liver. However, before birth much higher RNA contents were found (12-13 m g / g liver at 4-5 days before birth), that rapidly decreased thereafter so that adult values were reached at 1 day before birth. The carbamoyl-phosphate synthase protein content per g liver (Fig. lc) was recalculated from previously published data [10], taking into account the perinatal differences in immunoreactivity of the enzyme protein. The average of the ratios of the liver content of carbamoyl-phosphate synthase m R N A and carbamoyl-phosphate synthase protein calculated from the data between 5 and 32 days after birth was set arbitrarily at 1 (Fig. ld). It is clear that large changes in this ratio occur in the perinatal period, decreasing from a high value of 10 at 4 days before birth to a low value of 0.5 at 1-3 days after birth, i.e., a 15-20-fold change. These data may point to an enhanced efficiency of carbamoyl-phosphate synthase protein synthesis from its mRNA a n d / o r a decreased rate of protein degradation in the perinatal period (see Discussion). For reference, the changes in carbamoyl-phosphate synthase m R N A levels were compared to those of phosphoenolpyruvate carboxykinase mRNA (Fig. 2a) and albumin mRNA (Fig. 2b). Phosphoenolpyruvate carboxykinase mRNA levels
TABLE I EFFECT OF HORMONAL TREATMENT AND STARVATION ON THE LEVELS OF CARBAMOYL-PHOSPHATE T H A S E A N D A L B U M I N m R N A IN A D U L T R A T L I V E R
SYN-
m R N A levels found in control a n i m a l s (n = 9) are regarded as 100%, m R N A levels of e x p e r i m e n t a l a n i m a l s ( n = 3 - 6 ) are expressed as p e r c e n t a g e s + S . E , of control values, m R N A levels were d e t e r m i n e d by h y b r i d i z a t i o n of radioactively labeled c D N A to serial d i l u t i o n s of purified total R N A that was fixed on nitrocellulose. Diabetes was induced with streptozotocin (70 m g / k g ) , 48 h before the start of the experiment. T r i a m c i n o l o n e acetonide (4 m g / k g ) t r e a t m e n t was started at t = 0 h and repeated daily. A d r e n a l e c t o m y was p e r f o r m e d 48 h before the start of the experiment. Starvation was started at t = 0 h. Glucose (5 g / k g ) was a d m i n i s t e r e d by gavage 2 h before death. - = no d e t e r m i n a t i o n . Period of t r e a t m e n t
Triamcinolone Diabetes Diabetes + triamcinolone Adrenalectomy Starvation Starvation + glucose
C a r b a m o y l - p h o s p h a t e synthase m R N A
Albumin mRNA
24 h
96 h
24 h
96 h
250 + 30 220 _.9_10 350 ___30 75 + 15 115 4-10
170 _+ 30 125 4- 15 230 + 50 60 + 15
95 + 5 50 4- 10 80 _+ 10 9 0 + 10 95 + 5
110 60 90 45 -
+ 25 4- 20 4- 10 + 10
64
are extremely low up to parturition. Immediately after birth the levels rapidly accumulate to their developmental maximum (cf. Refs. 14,19 and 20). Thereafter, m R N A levels hardly change. Albumin m R N A levels are relatively low before birth, and increase in the first postnatal week to reach a
< Z rr E v
1.00
0.75
0.50
D. >m 0 2 5 o~
a c
J
I
I
I
+
,~,,
I
J
,
~
1.00
g
,< Z 0.75 n"
•
÷
E 0.50 -
÷
0
.E
+
025.Q
¢> 0.25 -
, • + , 4..1~, #
a
o
-1
b
i
I
I
I
0
1
2
3
4
13-
Z
age
't
-
b
> •6 4
~
2-
O
10
-
+
2-
'
* -
T A B L E I1 E F F E C T OF H O R M O N A L T R E A T M E N T ON T H E LEVELS OF C A R B A M O Y L - P H O S P H A T E S Y N T H A S E A N D A L B U M I N m R N A IN FETAL R A T LIVER
T
TI
0
C
5-
0 o1. Z ~ E
(weeks)
Fig. 2. Developmental profiles of: (a) phosphoenolpyruvate carboxykinase (PEPCK) m R N A and (b) albumin m R N A . m R N A levels were determined by dot-blot analysis and expressed in arbitrary units per mg total RNA. Estimates were derived from the same R N A preparations as used for the determination of carbamoyl-phoshate synthase m R N A content (Fig. ]a).
4 y
~
A
developmental maximum during the 2rid, 3rd and 4th postnatal weeks. Thereafter, levels decline to adult values (cf. Refs. 21-23).
$ 5
~
5
I
i
0
1
i 2
* i 3
] 4
d r .~e 5
I A
age (weeks)
Fig. 1. Developmental profiles of: (a) carbamoyl-phosphate synthase (CPS) m R N A ; (b) total liver RNA; (c) carbamoylphosphale synthase protein and (d) the ratio of carbamoylphosphate synthase m R N A and carbamoyl-phopshate synthase protein content (a × b : c). Carbamoyl-phosphate synthase m R N A levels were determined by dot-blot analysis and expressed in arbitrary units per mg total RNA. Total liver R N A content was estimated by the orcinol method [11]. Carbamoylphosphate synthase protein content was recalculated from previous results [10]. The average ratio of carbamoyl-phosphate synthase m R N A and protein between 5 and 32 days after birth was arbitrarily set at 1. Means__. S.E. of 3 7 determinations are depicted.
m R N A levels found in control animals (n = 3-9) of the same age are regarded as 100%. m R N A levels of experimental animals (2 pools of 6 fetuses) are expressed as percentage of control values and, between brackets, as a percentage of maximal (inducible) adult values, m R N A levels were determined by hybridization of radioactively labeled c D N A to serial dilutions of purified total R N A that was fixed on nitrocellulose. The fetuses were injected 3 days before birth with dexamethasone (2.5 m g / k g ) , triiodothyronine (3.3 m g / k g ) , dibutyryl cyclic A M P (50 m g / k g ) and theophylline (50 m g / k g ) (t = 0 h). This treatment was repeated after 24 h for the group examined at 48 h. Period of treatment
Carbamoyl-phosphate synthase m R N A
Albumin mRNA
6 12 24 48
420 (16) 210 (12) 145 (11) 120(11)
80 80 80 50
h h h h
65 To find out whether the limited capacity to accumulate carbamoyl-phosphate synthase protein prior to birth [4,5,7,8] is caused by a limited capacity to accumulate carbamoyl-phosphate synthase mRNA, fetal rats were treated with a hormonal cocktail that had previously been found to be optimal for prenatal enzyme induction [8]. Table II shows that, as in adult rats, the effects of hormonal treatment on carbamoyl-phosphate synthase mRNA levels are transient. In fetal rats, the maximal effect (a 4-fold increase over basal levels) is observed after 6 h. Moreover, the maximal levels of carbamoyl-phosphate synthase mRNA that are observed in these fetuses represent only 16% of those observed in adults, showing that the capacity to accumulate carbamoyl-phosphate synthase mRNA is limited prenatally. For comparison, albumin mRNA levels were estimated as well. Table II shows that prolonged hormonal treatment did affect albumin mRNA levels. This effect is presumably due to the cyclic AMP in the hormonal cocktail (cf. Table I). Discussion
If one assumes that the rate of synthesis of carbamoyl-phosphate synthase protein is a linear function of the concentration of carbamoyl-phosphate synthase mRNA ( ~ = k~[mRNA]) then the change in carbamoyl-phosphate synthase protein level (d E/d t) should equal the difference between the rate of protein synthesis ( ~ ) and protein degradation (Vd = ka[E]), i.e. d E / d t = k~[mRNA]ko[E ] (Ref. 24). The first assumption is allowed when the translational machinery is saturated with mRNAs of different types, while changes in the level of specific mRNAs occur. In steady-state conditions (dE/dt = 0), the ratio [mRNA]/[E]= kd/k ~. After calculation of d E / d t from the data of Fig. lc, and assuming that either k s o r k d is constant, a curve is obtained which is virtually identical to that Fig. ld. This demonstrates that d E / d t is negligible within the time scale of our observations. After birth, changes in the level of carbamoylphosphate synthase protein usually follow changes in the level of carbamoyl-phosphate synthase mRNA (Fig. ld; cf., for example, Refs. 25 and 26). The regulation of carbamoyl-phosphate syn-
thase gene expression by glucocorticosteroids and cyclic AMP in rat liver in this period of development is therefore predominantly regulated at the pretranslational level (k~ and k d constant). Similarly, the effects of different protein contents in the diet on carbamoyl-phosphate synthase gene expression have been localized at this pretranslational level [27]. It is most likely that the effects of protein content in the diet are mediated by cyclic AMP via changes in the rate of glucagon secretion [15,28]. Furthermore, glucocorticosteroids and insulin have been found to affect carbamoyl-phosphate synthase mRNA levels in Reuber hepatoma cells [25]. It was very interesting to find that the carbamoyl-phosphate synthase mRNA/protein ratio is not constant in the perinatal period (Fig. ld). The differential increase of carbamoyl-phosphate synthase mRNA and protein can in principle be explained by a 10-15-fold difference in the half-life of the mRNA and that of the protein molecule [24]. However, a similar increase in mRNA levels in the postnatal period is followed by a proportional increase in protein levels, as is evident from the constant ratio between both parameters (Fig. ld). We have already argued that steady-state conditions for carbamoyl-phosphate synthase protein levels can be assumed to exist within the time scale of our observations. This then leaves the possibility that the carbamoylphosphate synthase mRNA/protein ratio changes as a result of changes in the half-life of carbamoyl-phosphate synthase protein (ku) or as a result of changes in the efficiency of protein synthesis from the carbamoyl-phosphate synthase mRNA (k~). In view of the fact that in cultures of 14-day embryonic hepatocytes the half-life of carbamoyl-phosphate synthase protein is the same as that in adult liver (Van Roon et al., unpublished data), it seems most likely that the changes in the carbamoyl-phosphate synthase m R N A / c a r b a moyl-phosphate synthase protein ratio in the prenatal period are the result of changes in the efficiency of protein synthesis from the cellular mRNA (translational control). However, in the immediate postnatal period a transient decrease in the rate constant of degradation of carbamoylphosphate synthase protein cannot be excluded, as such a phenomenon has been shown to exist for
66
phosphoenolpyruvate carboxykinase in this period [29]. Therefore, carbamoyl-phosphate synthase gene expression in the prenatal period appears to be characterized by control at the pretranslational (regulation of mRNA levels (Fig. la, Table II)) as well as at the translational (regulation of k~) level, while in the immediate postnatal period, posttranslational (regulation of k d) control may exist. In view of the long half-life of carbamoyl-phosphate synthase protein under 'normal' conditions [23] and the transient nature of the observed phenomena, it is probably best to verify the nature of the control mechanisms by measuring in vivo the rate of carbamoyl-phosphate synthase protein synthesis as a function of carbamoyl-phosphate synthase mRNA concentration (k~). The discrepancy between the rates of accumulation of carbamoylphosphate synthase mRNA and protein in the perinatal period was not previously observed [30], probably because too few observations were made. Many examples of translational (posttranscriptional) control mechanisms exist. In fact, in some cases, factors such as cyclic AMP can affect both the transcriptional and translational efficiency of genes [31]. The data in Table II clearly show that in order to explain the regulation of carbamoyi-phosphate synthase gene expression in the perinatal period, additional factors have to be identified. In particular, the question arises as to why glucocorticosteroids and cyclic AMP are not able to stimulate gene expression to adult levels. Two still poorly defined factors have been identified by us as responsible for the limited capacity of the fetal hepatocytes to respond to hormones: the intrauterine environment [5] and the degree of developmental maturation of the hepatocytes (Ref. 3, and Van Roon et al., unpublished data). The availability of an in vitro culture system in which maturation towards the adult capacity of gene expression occurs (Van Roon et al., unpublished data) should allow further characterization of these factors and of their effects on carbamoyl-phosphate synthase gene expression.
Acknowledgements These investigations were supported in part by the Foundation for Medical Research, F U N G O ,
Grant No. 900-528-038. We want to acknowledge the stimulating discussions with Drs. H. Westerhoff and H. Oosting, the help of Mrs. J.H. van Horssen-Medema and Mr. C. Hersbach in preparing the drawings and photographs and the help of Ms. J. Husslage in preparing the manuscript.
References 1 Charles, R., De GraM, A. and Moorman, A.F.M. (1980) Biochim. Biophys. Acla 629, 36-49 2 Gaasbeek Janzen, J.W., Lamers, W.H., Moorman, A.F.M., De Graaf, A., Los, J.A. and Charles, R. (1984) J. Histochem. Cytochem. 32, 557-564 3 Lamers, W.H., Zonneveld, D. and Charles, R. (1984) Dev. Biol. 105, 500 508 4 Lamers, W.H. and Mooren, P.G. (1980) Biol. Neonat. 37. 113-137 5 Lamers, W.H., Mooren, P.G., De Graaf, A. and Charles, R. (1985) Eur. J. Biochem. 146, 475-480 6 Gaasbeek Janzen, J.W., Moorman, A.F.M., Lamers, W.H. and Charles, R. (1985) J. Histochem. Cytochem. 33, 1205-1211 7 Earners, W.H. and Mooren, P.G. (1980) Biol. Neonat. 37, 264 284 8 kamers, W.H. and Mooren, P.G. (1981) Mech. Ageing Dev. 15, 93 118 9 De Groot, C.J., Van Zonneveld, A.J., Mooren, P.G., Zonneveld, D., Van den Dool, A., Van den Bogaert, A.J.W., Lamers, W.H., Moorman, A.F.M. and Charles, R. (1984) Biochem. Biophys. Res. C o m m u n . 124, 882-888 10 Charles, R., De Graaf, A., Lamers, W.H. and Moorman, A.F.M. (1983) Mech. Ageing Dev. 22, 193-203 11 Munro, H.N. and Fleck, A. (1966) Methods Biochem. Anal. 14, 113 176 12 Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter, W.J. (1979) Biochemistry 18, 5294 5299 13 Meinkoth, J. and Wahl, G. (1984) Anal. Biochem. 138, 267-284 14 Cimbala, M.A., Lamers, W.H., Nelson, K., Monahan, J.E., Yoo-Warren, H. and Hanson, R.W. (1982) J. Biol. Chem. 257, 7629-7636 15 Seitz, H.J., Krone, W. and Tarnowski, W. (1977) Acta Endocrinol. 85, 389-397 16 Seitz, H.J., Li~th, W. and Tarnowski, W. (1979) Arch. Biochem. Biophys. 195, 385-391 17 Moshage, H.J., De Haard, H.J.W., Princen, H.M.G. and Yap, S.H. (1985) Biochim. Biophys. Acta 824, 27-33 18 Jefferson, L.S., Liao, W.S., Peavy, D.E., Miller, T.B., Appel, M.C. and Taylor, J.M. (1983) J. Biol. Chem. 258, 1369 1375 19 Mencher, D., Cohen, H., Benvenisty, N., Meyuhas, O. and Reshef, L (1984) Eur. J. Biochem. 141, 199-203 20 Ruiz, J.P.G., Ingrain, R. and Hanson, R.W. (1978) Proc, Natl. Acad. Sci. USA 75, 4189-4193 21 Liao, W.S.L., Conn, A.R. and Taylor, J.M. (1980) J. Biol. Chem. 255. 10036-10039
67 22 Sala-Trepat, J.M., Dever, J., Sargent, T.D., Thomas, K., Sell, S. and Bonner, J. (1979) Biochemistry 18, 2167-2178 23 Selten, G.C.M., Princen, H.M.G., Selten-Versteegen, A.M.E., MoI-Backx, G.P.B.M. and Yap, S.H. (1982) Biochim. Biophys. Acta 699, 131-137 24 Berlin, C.M. and Schimke, R.T. (1965) Mol. Pharmacol. 1, 149-156 25 Kitagawa, Y., Ryall, J., Nguyen, M. and Shore, G.C. (1985) Biochim. Biophys. Acta 825, 148-153 26 Kitagawa, y. and Sugimoto, E. (1985) Eur. J. Biochem. 150, 249-254
27 Ryall, J., Rachubinsky, R.A., Nguyen, M.. Rozen, R., Broglie, K.E. and Shore, G.C. (1984) J. Biol. Chem. 259, 9172-9176 28 Eisenstein, A.B., Strack, I., Gallo-Torres, H., Georgiadis, A. and Miller, O.N. (1979) Am. J. Physiol. 236, E20-E27 29 Philippidis, H., Hanson, R.W., Reshef, L., Hopgood, M.F. and Ballard, F.J. (1972) Biochem. J. 126, 1127-1134 30 Adcock, M.W. and O'Brien, W.E. (1984) J. Biol. Chem. 259, 13471-13476 31 Noguchi, T., Diesterhaft, M. and Granner, D. (1982) J. Biol. Chem. 257, 2386-2390
61
BBA91562
Developmental and hormonal regulation of carbamoyl-phosphate synthase gene expression in rat liver: evidence for control mechanisms at different levels in the perinatal period C.J. d e G r o o t , D. Z o n n e v e l d , R . T , M . de Laaf, M.A. D i n g e m a n s e , P.G. M o o r e n , A . F . M . M o o r m a n , W . H . L a m e r s * a n d R. C h a r l e s Department of Anatomy and Embryology, University of Amsterdam, AMC, Meibergdreef15, 1105 A Z Amsterdam (The Netherlands) (Received October 15th, 1985)
Key words: Carbamoyl-phosphate synthase; Gene expression; Hormonal regulation; (Rat liver)
Carbamoyl-phosphate synthase gene expression is found to be primarily regulated by conditions that enhance hepatic glucocorticosteroid levels (hormone injections) and cyclic AMP levels (induction of diabetes). After birth, changes in the level of carhamoyl-phosphate synthase protein follow changes in the level of carbamoylphosphate synthase mRNA, suggesting a pretranslational control mechanism. In fetal rats, carbamoyl-phosphate synthase gene expression is regulated by the same factors as in adults. However, both the level to which carbamoyl-phosphate synthase mRNA can accumulate and the extent to which mRNA can be translated appear to be limited, indicating control mechanisms at the pretranslational and translational level. Finally, in the immediate postnatal period, a transient but pronounced decrease in the rate of degradation of carbamoyl-pbosphate synthase protein may play a role in the accumulation of the enzyme.
Introduction Carbamoyl-phosphate synthase (EC 6.3.4.16) is an abundant hepatocyte-specific protein [1,2]. It offers an attractive model system for studying the developmental regulation of gene expression, as both the timing and the site of expression within the liver are under developmental control. Rat hepatocytes acquire the capacity to synthesize carbamoyl-phosphate synthase as soon as they differentiate from the embryonic foregut [3], but start to accumulate the enzyme only shortly before birth [4]. After birth, the adult capacity to synthesize carbamoyl-phosphate synthase is acquired [5], as is
* To whom correspondence should be addressed.
the typica! periportal localization [6]. The factors that regulate carbamoyl-phosphate synthase gene expression at the protein level have been extensively studied, showing glucocorticosteroids and cyclic AMP to be the primary regulatory factors [3,4,7,8]. A preliminary exploration showed that the same factors affected carbamoylphosphate synthase mRNA levels [9]. In the present study we have analyzed further the role of hormones in regulating the levels of carbamoylphosphate synthase mRNA. Furthermore, we have established a detailed developmental profile of carbamoyl-phosphate synthase mRNA and have compared it to that of carbamoyl-phosphate synthase protein [10]. The results suggest that in the perinatal period, both pretranslational and translational control mechanisms may be operative.
0167-4781/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)
62 Materials and Methods
Animals. Wistar rats, obtained from the TNO animal farm in Zeist, The Netherlands, were used. Fetal age was estimated from the time of conception and was accurate to within 12 h; birth normally occurred between 21.5 and 22.5 days after conception. Neonatal age was calculated from the time of birth. Rats were weaned at 4 weeks post partum. Treatment of animals, For hormonal treatment, either 3-4-month-old adult or 19-day fetal rats were used. Diabetes was induced in adult rats by intraperitoneal injection of streptozotocin (70 m g / k g body weight) after an overnight fast. 24 h later, the animals were checked for glucosuria. Bilateral adrenalectomy and gonadectomy were performed under ether anesthesia and checked for completeness at death. Glucose (5 g / k g body weight) was administered to 24-h-starved rats by gastric gavage. Hormone-treated adult rats received triamcinolone acetonide (4 m g / k g body weight) by intraperitoneal injection. Hormonetreated fetuses received a cocktail of dexamethasone (2.5 m g / k g body weight), triiodothyronine (3.3 m g / k g body weight), dibutyryl cyclic AMP (50 m g / k g body weight) and theophylline (50 m g / k g body weight) by intraperitoneal injection, after laparotomy of the dam. Although triiodothyronine does not have an effect on carbamoyl-phosphate synthase mRNA (see Results) or protein [7] in adult animals, it was included in this cocktail as the hormone does have a stimulatory effect on carbamoyl-phosphate synthase protein before birth [7]. Determination of RNA content. The total RNA content of liver was determined by the orcinol method [11]. RNA was isolated from liver homogenates in guanidinium thiocyanate either by differential precipitation (Ref. 12, as modified in Ref. 9) or by centrifugation through a CsCI cushion [12]. Concentrations were calculated from the absorbance at 260 nm. The RNA samples were diluted in a high-salt buffer (1.8 M NaC1/100 mM N a H 2 P O / 1 0 mM EDTA (pH 7.4)) and aliquots were passed through nitrocellulose filters using a commercial manifold. After heating at 80°C for 2 h, filters were prehybridized and hybridized at 42°C, essentially as
described [13]. The cDNA probes for carbamoylphosphate synthase mRNA and albumin mRNA [9] and phosphoenolpyruvate carboxykinase mRNA [14] were labeled by nick translation to a specific activity of approx. 10 ~ cpm//~g. In more recent experiments a nearly full-length carbamoyl-phosphate synthase cDNA clone was used (designated pBR-CPS5; insert length, 5.5 kb). After washing, autoradiographs of the filters were made. For quantitative analysis, spots were cut out and radioactivity was determined by liquid scintillation counting. In the graphs and tables, means _+ S.E. of 3-7 independent estimates are given. In Fig. 1, the standard error of the ratio of carbamoyl-phosphate synthase mRNA and carbamoyl-phosphate synthase protein was calculated with the delta method. Results
Table I describes the effects of long-lasting changes in the levels of glucocorticosteroids and cyclic AMP on the level of carbamoyl-phosphate synthase mRNA. Durable changes in hepatic cyclic AMP concentration were induced by diabetes, starvation, and starvation followed by glucose refeeding [15,16]. As a control, albumin mRNA levels were estimated as well. It is clear that administration of glucocorticosteroids (triamcinolone) enhances carbamoyl-phosphate synthase m R N A levels more than 2-fold after 24 h, whereas a decrease in the levels of circulating glucocorticosteroids as caused by adrenalectomy decreases the mRNA levels to 60% after 96 h. In intact animals, additional glucocorticosteroids have no effect on albumin mRNA levels; however, albumin mRNA levels decrease 2-fold upon removal of the adrenals. Glucocorticosteroids may therefore be necessary for the maintenance of 'normal' levels of albumin mRNA (cf. Ref. 17). Elevated hepatic cyclic AMP levels, as seen in diabetic animals, are associated with a more than 2-fold increase in carbamoyl-phosphate synthase mRNA levels after 24 h. Furthermore, the effects of glucocorticosteroids (triamcinolone treatment) and cyclic AMP (diabetes) are additive, causing a 3.5fold increase after 24 h. Despite the continued presence of the inducing factors, their effects on
63
carbamoyl-phosphate synthase mRNA levels are transient, showing a 30-40% decrease when effects at 96 h are compared with those at 24 h. No effects of thyroid hormones on carbamoyl-phosphate synthase mRNA levels were observed (data not shown). The diabetic condition causes a decrease in albumin mRNA levels (cf. Ref. 18). Apparently, glucocorticosteroids are able to counteract some of the deleterious effects of diabetes on albumin synthesis. Starvation slightly decreases carbamoyl-phosphate synthase mRNA levels, while glucose refeeding restores the mRNA levels to normal. Starvation for 24 h has little effect on albumin mRNA levels. Fig. la shows the developmental profile of the ratio of the levels of hepatic carbamoyl-phosphate synthase mRNA and total RNA. Between 5 days before birth, when only a very small amount of m R N A can be detected, and 1 day before birth, carbamoyl-phosphate synthase mRNA levels increase 4-fold. Between this time and birth, mRNA levels remain constant, whereas in the first few postnatal days the mRNA levels drop to 50% of those at birth. Thereafter, mRNA levels incrase to attain maximum levels during the weaning period (3rd and 4th postnatal weeks). Adult mRNA levels are approx. 40% of those at weaning. In order to find the relationship between carbamoyl-phosphate synthase mRNA levels and carbamoyl-phosphate synthase protein levels it was necessary to
determine the total RNA content per gram liver (Fig. lb). After birth, the total RNA level was found to be 8-9 m g / g liver. However, before birth much higher RNA contents were found (12-13 m g / g liver at 4-5 days before birth), that rapidly decreased thereafter so that adult values were reached at 1 day before birth. The carbamoyl-phosphate synthase protein content per g liver (Fig. lc) was recalculated from previously published data [10], taking into account the perinatal differences in immunoreactivity of the enzyme protein. The average of the ratios of the liver content of carbamoyl-phosphate synthase m R N A and carbamoyl-phosphate synthase protein calculated from the data between 5 and 32 days after birth was set arbitrarily at 1 (Fig. ld). It is clear that large changes in this ratio occur in the perinatal period, decreasing from a high value of 10 at 4 days before birth to a low value of 0.5 at 1-3 days after birth, i.e., a 15-20-fold change. These data may point to an enhanced efficiency of carbamoyl-phosphate synthase protein synthesis from its mRNA a n d / o r a decreased rate of protein degradation in the perinatal period (see Discussion). For reference, the changes in carbamoyl-phosphate synthase m R N A levels were compared to those of phosphoenolpyruvate carboxykinase mRNA (Fig. 2a) and albumin mRNA (Fig. 2b). Phosphoenolpyruvate carboxykinase mRNA levels
TABLE I EFFECT OF HORMONAL TREATMENT AND STARVATION ON THE LEVELS OF CARBAMOYL-PHOSPHATE T H A S E A N D A L B U M I N m R N A IN A D U L T R A T L I V E R
SYN-
m R N A levels found in control a n i m a l s (n = 9) are regarded as 100%, m R N A levels of e x p e r i m e n t a l a n i m a l s ( n = 3 - 6 ) are expressed as p e r c e n t a g e s + S . E , of control values, m R N A levels were d e t e r m i n e d by h y b r i d i z a t i o n of radioactively labeled c D N A to serial d i l u t i o n s of purified total R N A that was fixed on nitrocellulose. Diabetes was induced with streptozotocin (70 m g / k g ) , 48 h before the start of the experiment. T r i a m c i n o l o n e acetonide (4 m g / k g ) t r e a t m e n t was started at t = 0 h and repeated daily. A d r e n a l e c t o m y was p e r f o r m e d 48 h before the start of the experiment. Starvation was started at t = 0 h. Glucose (5 g / k g ) was a d m i n i s t e r e d by gavage 2 h before death. - = no d e t e r m i n a t i o n . Period of t r e a t m e n t
Triamcinolone Diabetes Diabetes + triamcinolone Adrenalectomy Starvation Starvation + glucose
C a r b a m o y l - p h o s p h a t e synthase m R N A
Albumin mRNA
24 h
96 h
24 h
96 h
250 + 30 220 _.9_10 350 ___30 75 + 15 115 4-10
170 _+ 30 125 4- 15 230 + 50 60 + 15
95 + 5 50 4- 10 80 _+ 10 9 0 + 10 95 + 5
110 60 90 45 -
+ 25 4- 20 4- 10 + 10
64
are extremely low up to parturition. Immediately after birth the levels rapidly accumulate to their developmental maximum (cf. Refs. 14,19 and 20). Thereafter, m R N A levels hardly change. Albumin m R N A levels are relatively low before birth, and increase in the first postnatal week to reach a
< Z rr E v
1.00
0.75
0.50
D. >m 0 2 5 o~
a c
J
I
I
I
+
,~,,
I
J
,
~
1.00
g
,< Z 0.75 n"
•
÷
E 0.50 -
÷
0
.E
+
025.Q
¢> 0.25 -
, • + , 4..1~, #
a
o
-1
b
i
I
I
I
0
1
2
3
4
13-
Z
age
't
-
b
> •6 4
~
2-
O
10
-
+
2-
'
* -
T A B L E I1 E F F E C T OF H O R M O N A L T R E A T M E N T ON T H E LEVELS OF C A R B A M O Y L - P H O S P H A T E S Y N T H A S E A N D A L B U M I N m R N A IN FETAL R A T LIVER
T
TI
0
C
5-
0 o1. Z ~ E
(weeks)
Fig. 2. Developmental profiles of: (a) phosphoenolpyruvate carboxykinase (PEPCK) m R N A and (b) albumin m R N A . m R N A levels were determined by dot-blot analysis and expressed in arbitrary units per mg total RNA. Estimates were derived from the same R N A preparations as used for the determination of carbamoyl-phoshate synthase m R N A content (Fig. ]a).
4 y
~
A
developmental maximum during the 2rid, 3rd and 4th postnatal weeks. Thereafter, levels decline to adult values (cf. Refs. 21-23).
$ 5
~
5
I
i
0
1
i 2
* i 3
] 4
d r .~e 5
I A
age (weeks)
Fig. 1. Developmental profiles of: (a) carbamoyl-phosphate synthase (CPS) m R N A ; (b) total liver RNA; (c) carbamoylphosphale synthase protein and (d) the ratio of carbamoylphosphate synthase m R N A and carbamoyl-phopshate synthase protein content (a × b : c). Carbamoyl-phosphate synthase m R N A levels were determined by dot-blot analysis and expressed in arbitrary units per mg total RNA. Total liver R N A content was estimated by the orcinol method [11]. Carbamoylphosphate synthase protein content was recalculated from previous results [10]. The average ratio of carbamoyl-phosphate synthase m R N A and protein between 5 and 32 days after birth was arbitrarily set at 1. Means__. S.E. of 3 7 determinations are depicted.
m R N A levels found in control animals (n = 3-9) of the same age are regarded as 100%. m R N A levels of experimental animals (2 pools of 6 fetuses) are expressed as percentage of control values and, between brackets, as a percentage of maximal (inducible) adult values, m R N A levels were determined by hybridization of radioactively labeled c D N A to serial dilutions of purified total R N A that was fixed on nitrocellulose. The fetuses were injected 3 days before birth with dexamethasone (2.5 m g / k g ) , triiodothyronine (3.3 m g / k g ) , dibutyryl cyclic A M P (50 m g / k g ) and theophylline (50 m g / k g ) (t = 0 h). This treatment was repeated after 24 h for the group examined at 48 h. Period of treatment
Carbamoyl-phosphate synthase m R N A
Albumin mRNA
6 12 24 48
420 (16) 210 (12) 145 (11) 120(11)
80 80 80 50
h h h h
65 To find out whether the limited capacity to accumulate carbamoyl-phosphate synthase protein prior to birth [4,5,7,8] is caused by a limited capacity to accumulate carbamoyl-phosphate synthase mRNA, fetal rats were treated with a hormonal cocktail that had previously been found to be optimal for prenatal enzyme induction [8]. Table II shows that, as in adult rats, the effects of hormonal treatment on carbamoyl-phosphate synthase mRNA levels are transient. In fetal rats, the maximal effect (a 4-fold increase over basal levels) is observed after 6 h. Moreover, the maximal levels of carbamoyl-phosphate synthase mRNA that are observed in these fetuses represent only 16% of those observed in adults, showing that the capacity to accumulate carbamoyl-phosphate synthase mRNA is limited prenatally. For comparison, albumin mRNA levels were estimated as well. Table II shows that prolonged hormonal treatment did affect albumin mRNA levels. This effect is presumably due to the cyclic AMP in the hormonal cocktail (cf. Table I). Discussion
If one assumes that the rate of synthesis of carbamoyl-phosphate synthase protein is a linear function of the concentration of carbamoyl-phosphate synthase mRNA ( ~ = k~[mRNA]) then the change in carbamoyl-phosphate synthase protein level (d E/d t) should equal the difference between the rate of protein synthesis ( ~ ) and protein degradation (Vd = ka[E]), i.e. d E / d t = k~[mRNA]ko[E ] (Ref. 24). The first assumption is allowed when the translational machinery is saturated with mRNAs of different types, while changes in the level of specific mRNAs occur. In steady-state conditions (dE/dt = 0), the ratio [mRNA]/[E]= kd/k ~. After calculation of d E / d t from the data of Fig. lc, and assuming that either k s o r k d is constant, a curve is obtained which is virtually identical to that Fig. ld. This demonstrates that d E / d t is negligible within the time scale of our observations. After birth, changes in the level of carbamoylphosphate synthase protein usually follow changes in the level of carbamoyl-phosphate synthase mRNA (Fig. ld; cf., for example, Refs. 25 and 26). The regulation of carbamoyl-phosphate syn-
thase gene expression by glucocorticosteroids and cyclic AMP in rat liver in this period of development is therefore predominantly regulated at the pretranslational level (k~ and k d constant). Similarly, the effects of different protein contents in the diet on carbamoyl-phosphate synthase gene expression have been localized at this pretranslational level [27]. It is most likely that the effects of protein content in the diet are mediated by cyclic AMP via changes in the rate of glucagon secretion [15,28]. Furthermore, glucocorticosteroids and insulin have been found to affect carbamoyl-phosphate synthase mRNA levels in Reuber hepatoma cells [25]. It was very interesting to find that the carbamoyl-phosphate synthase mRNA/protein ratio is not constant in the perinatal period (Fig. ld). The differential increase of carbamoyl-phosphate synthase mRNA and protein can in principle be explained by a 10-15-fold difference in the half-life of the mRNA and that of the protein molecule [24]. However, a similar increase in mRNA levels in the postnatal period is followed by a proportional increase in protein levels, as is evident from the constant ratio between both parameters (Fig. ld). We have already argued that steady-state conditions for carbamoyl-phosphate synthase protein levels can be assumed to exist within the time scale of our observations. This then leaves the possibility that the carbamoylphosphate synthase mRNA/protein ratio changes as a result of changes in the half-life of carbamoyl-phosphate synthase protein (ku) or as a result of changes in the efficiency of protein synthesis from the carbamoyl-phosphate synthase mRNA (k~). In view of the fact that in cultures of 14-day embryonic hepatocytes the half-life of carbamoyl-phosphate synthase protein is the same as that in adult liver (Van Roon et al., unpublished data), it seems most likely that the changes in the carbamoyl-phosphate synthase m R N A / c a r b a moyl-phosphate synthase protein ratio in the prenatal period are the result of changes in the efficiency of protein synthesis from the cellular mRNA (translational control). However, in the immediate postnatal period a transient decrease in the rate constant of degradation of carbamoylphosphate synthase protein cannot be excluded, as such a phenomenon has been shown to exist for
66
phosphoenolpyruvate carboxykinase in this period [29]. Therefore, carbamoyl-phosphate synthase gene expression in the prenatal period appears to be characterized by control at the pretranslational (regulation of mRNA levels (Fig. la, Table II)) as well as at the translational (regulation of k~) level, while in the immediate postnatal period, posttranslational (regulation of k d) control may exist. In view of the long half-life of carbamoyl-phosphate synthase protein under 'normal' conditions [23] and the transient nature of the observed phenomena, it is probably best to verify the nature of the control mechanisms by measuring in vivo the rate of carbamoyl-phosphate synthase protein synthesis as a function of carbamoyl-phosphate synthase mRNA concentration (k~). The discrepancy between the rates of accumulation of carbamoylphosphate synthase mRNA and protein in the perinatal period was not previously observed [30], probably because too few observations were made. Many examples of translational (posttranscriptional) control mechanisms exist. In fact, in some cases, factors such as cyclic AMP can affect both the transcriptional and translational efficiency of genes [31]. The data in Table II clearly show that in order to explain the regulation of carbamoyi-phosphate synthase gene expression in the perinatal period, additional factors have to be identified. In particular, the question arises as to why glucocorticosteroids and cyclic AMP are not able to stimulate gene expression to adult levels. Two still poorly defined factors have been identified by us as responsible for the limited capacity of the fetal hepatocytes to respond to hormones: the intrauterine environment [5] and the degree of developmental maturation of the hepatocytes (Ref. 3, and Van Roon et al., unpublished data). The availability of an in vitro culture system in which maturation towards the adult capacity of gene expression occurs (Van Roon et al., unpublished data) should allow further characterization of these factors and of their effects on carbamoyl-phosphate synthase gene expression.
Acknowledgements These investigations were supported in part by the Foundation for Medical Research, F U N G O ,
Grant No. 900-528-038. We want to acknowledge the stimulating discussions with Drs. H. Westerhoff and H. Oosting, the help of Mrs. J.H. van Horssen-Medema and Mr. C. Hersbach in preparing the drawings and photographs and the help of Ms. J. Husslage in preparing the manuscript.
References 1 Charles, R., De GraM, A. and Moorman, A.F.M. (1980) Biochim. Biophys. Acla 629, 36-49 2 Gaasbeek Janzen, J.W., Lamers, W.H., Moorman, A.F.M., De Graaf, A., Los, J.A. and Charles, R. (1984) J. Histochem. Cytochem. 32, 557-564 3 Lamers, W.H., Zonneveld, D. and Charles, R. (1984) Dev. Biol. 105, 500 508 4 Lamers, W.H. and Mooren, P.G. (1980) Biol. Neonat. 37. 113-137 5 Lamers, W.H., Mooren, P.G., De Graaf, A. and Charles, R. (1985) Eur. J. Biochem. 146, 475-480 6 Gaasbeek Janzen, J.W., Moorman, A.F.M., Lamers, W.H. and Charles, R. (1985) J. Histochem. Cytochem. 33, 1205-1211 7 Earners, W.H. and Mooren, P.G. (1980) Biol. Neonat. 37, 264 284 8 kamers, W.H. and Mooren, P.G. (1981) Mech. Ageing Dev. 15, 93 118 9 De Groot, C.J., Van Zonneveld, A.J., Mooren, P.G., Zonneveld, D., Van den Dool, A., Van den Bogaert, A.J.W., Lamers, W.H., Moorman, A.F.M. and Charles, R. (1984) Biochem. Biophys. Res. C o m m u n . 124, 882-888 10 Charles, R., De Graaf, A., Lamers, W.H. and Moorman, A.F.M. (1983) Mech. Ageing Dev. 22, 193-203 11 Munro, H.N. and Fleck, A. (1966) Methods Biochem. Anal. 14, 113 176 12 Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter, W.J. (1979) Biochemistry 18, 5294 5299 13 Meinkoth, J. and Wahl, G. (1984) Anal. Biochem. 138, 267-284 14 Cimbala, M.A., Lamers, W.H., Nelson, K., Monahan, J.E., Yoo-Warren, H. and Hanson, R.W. (1982) J. Biol. Chem. 257, 7629-7636 15 Seitz, H.J., Krone, W. and Tarnowski, W. (1977) Acta Endocrinol. 85, 389-397 16 Seitz, H.J., Li~th, W. and Tarnowski, W. (1979) Arch. Biochem. Biophys. 195, 385-391 17 Moshage, H.J., De Haard, H.J.W., Princen, H.M.G. and Yap, S.H. (1985) Biochim. Biophys. Acta 824, 27-33 18 Jefferson, L.S., Liao, W.S., Peavy, D.E., Miller, T.B., Appel, M.C. and Taylor, J.M. (1983) J. Biol. Chem. 258, 1369 1375 19 Mencher, D., Cohen, H., Benvenisty, N., Meyuhas, O. and Reshef, L (1984) Eur. J. Biochem. 141, 199-203 20 Ruiz, J.P.G., Ingrain, R. and Hanson, R.W. (1978) Proc, Natl. Acad. Sci. USA 75, 4189-4193 21 Liao, W.S.L., Conn, A.R. and Taylor, J.M. (1980) J. Biol. Chem. 255. 10036-10039
67 22 Sala-Trepat, J.M., Dever, J., Sargent, T.D., Thomas, K., Sell, S. and Bonner, J. (1979) Biochemistry 18, 2167-2178 23 Selten, G.C.M., Princen, H.M.G., Selten-Versteegen, A.M.E., MoI-Backx, G.P.B.M. and Yap, S.H. (1982) Biochim. Biophys. Acta 699, 131-137 24 Berlin, C.M. and Schimke, R.T. (1965) Mol. Pharmacol. 1, 149-156 25 Kitagawa, Y., Ryall, J., Nguyen, M. and Shore, G.C. (1985) Biochim. Biophys. Acta 825, 148-153 26 Kitagawa, y. and Sugimoto, E. (1985) Eur. J. Biochem. 150, 249-254
27 Ryall, J., Rachubinsky, R.A., Nguyen, M.. Rozen, R., Broglie, K.E. and Shore, G.C. (1984) J. Biol. Chem. 259, 9172-9176 28 Eisenstein, A.B., Strack, I., Gallo-Torres, H., Georgiadis, A. and Miller, O.N. (1979) Am. J. Physiol. 236, E20-E27 29 Philippidis, H., Hanson, R.W., Reshef, L., Hopgood, M.F. and Ballard, F.J. (1972) Biochem. J. 126, 1127-1134 30 Adcock, M.W. and O'Brien, W.E. (1984) J. Biol. Chem. 259, 13471-13476 31 Noguchi, T., Diesterhaft, M. and Granner, D. (1982) J. Biol. Chem. 257, 2386-2390