Plasma clearance and liver metabolism of native and asialo-human transcortin in the rat

Plasma clearance and liver metabolism of native and asialo-human transcortin in the rat

543 Biochim~cd et Bi~physica Acta, 585 (1979) 543--553 © Elsevier/North-Holland Biomedical Press BBA 28972 PLASMA CLEARANCE AND LIVER METABOLISM OF...

727KB Sizes 1 Downloads 39 Views

543

Biochim~cd et Bi~physica Acta, 585 (1979) 543--553 © Elsevier/North-Holland Biomedical Press

BBA 28972

PLASMA CLEARANCE AND LIVER METABOLISM OF NATIVE AND ASIALO-HUMAN TRANSCORTIN IN THE RAT

KIM L. HOSSNER and REINHART B. BILLIAR Department of Reproductive Biology, Case Western Reserve University, Cleveland, OH 44106 (U.S.A.) (Received November 21st, 1978)

Key words: Plasma clearance; Asialoglycoprotein; Transcortin; Steroid-binding protein; (Liver metabolism)

Summary Plasma kinetics and liver metabolism of iodinated human corticosteroid. binding protein have been studied in ovariectomized female rats. 12SI-labeled human corticosteroid-binding globulin prepared by a modified chloramine T reaction was shown to be physically intact and biologically active. Intravenously injected t2SI-labeled human corticosteroid-binding globulin was shown to give a complex clearance pattern from the plasma, with half-lives of 7.5 and 51 min. Estrogen injections had no effect on plasma clearance rate. Direct involvement of liver plasma membrane receptors for asialoglycoproteins in t2sIlabeled human corticosteroid-binding globulin metabolism was demonstrated in vivo and in vitro using asialofetuin as a competitive inhibitor. ~2SI-labeled human asialo~orticosteroid-binding globulin was cleared from the plasma with a half-life of less than 1 min, while the simultaneous injection of 5 mg asialofetuin maintained the circulating plasma levels. Asialofetuin also slowed the clearance of intact ~2SI-labeled human corticosteroid4~inding globulin from the plasma (tl/2 = 9 0 m i n ) . Binding of i2SI-labeled human asialo~orticosteroidbinding globulin to rat liver plasma membranes in vitro was inhibited in a dosedependent manner by asialofetuin, but not by intact human corticosteroidbinding globulin or fetuin, t2SI-labeled human corticosteroid-binding globulin did not bind significantly to the membranes. It is concluded that human corticosteroid-binding globulin clearance from rat plasma is rapid and that the carbohydrate moiety of human corticosteroid-binding globulin is involved in its clearance and catabolism by the liver.

544 Introduction Transcortin or corticosteroid.binding globulin, a serum a-globulin glycoprotein [1] is found in all mammalian species studied [2] and has a high binding affinity for corticosteroids and progesterone [1,3]. Iodinated human corticosteroid-binding globulin has been shown to have a long half-life in human subjects, in which it was estimatoJ to be about five days [4]. These authors observed an initial rapid loss of about 50% of the radioactivity during the first hour of circulation which was assumed to be due to equilibrium with extravascular spaces, as there was no significant accumulation of iodine in the thyroid. They also observed that estrogen injections did not affect the corticosteroid.binding globulin clearance rate. Ashwell and MoreU [5] have shown that the mammalian liver has specific hepatocyte-binding sites for the galactose terminal of asialoglycoproteins which leads to the subsequent degradation of such glycoproteins. The binding of asialoglycoproteins can also be demonstrated in vitro with isolated rat liver plasma membrane [6]. Exposure of the binding sites, in vivo or in vitro, to an excess of asialoglycoproteins prevents binding of a trace label of asialoglycoprotein, which results in a prolongation of the labeled glycoprotein's half-life in the intact animal. Since corticosteroid-binding globulin is a glycoprotein, we have performed competition experiments in vivo and in vitro with asialofetuin to study if the hepatic receptor is involved in corticosteroid-binding globulin metabolism. Recent studies have shown that the liver is the only source of asialofetuin metabolism [ 7 ]. Methods and Materials

Adult Sprague-Dawley female rats (250--350 g) were ovariectomized via flank incisions and allowed to recover at least 10 days before being used. Estradiol-17~ (0.5 or 2.0 ~ ) was injected subcutaneously in 0.10 ml of sesame oil at 2 4 ~ intervals for two days before injection of radiotmcers. Affinity column purified human corticosteroid-binding globulin [8] was generously provided by Dr. W. Rosner, Columbia University, and was free of any contaminating protein as indicated by polyacrylamide electrophoresls. Radioactivity tracers (107 cpm in 0.5 ml 5% bovine serum albumin) were injected into the saphenous vein under ether anesthesia. Heparinized blood samples (0.2-0.3 ml) were taken by jugular puncture of ether-immobilized rots and 50-~1 duplicates of plasma were assessed for the total radioactivity and for acid-precipitated 125I. Urine was collected in metabolic cages and the urinary protein was precipitated with perchloric acid after the addition of 100 bovine serum albumin. Organs from rats were collected on ice after aortic exsangnination. The tissues were homogenized in 5 vols. of glass~tistilled water (4°C) with two 154 bursts of a Brinkman Polytron tissue homogenizer. Perchloric acid was used to precipitate the protein in the plasma, urine and in 0.4 ml duplicates of tissue homogenates. Proteins were iodinated by a modification of the original chloramine T reaction [9]. Iodinations were carried out at room temperature for 15--30 s with 5--10 ~g protein and 2 : 1 or 7 : 1 (w/w) chloramine T : protein ratio. Un-

545

reacted iodine was separated from protein-bound iodine by passage through a bovine serum albumin-precoated Sephadex G-100 column (0.7 × 18 cm). Attempts at using DEAE-cellulose ion-exchange chromatography to purify ~2SI-labeled h u m a n corticosteroid-binding globulin were unsuccessful, as degradation products eluted with the 12SI-labeled human corticosteroid-binding globulin peak. The specific activity was estimated to be 75--130 Ci/g protein. Iodination products were stored in 5.0% albumin in 0.05 M sodium phosphate buffer, pH 7.4, at --20 ° C and were discarded after two weeks. Asialofetuin was prepared from fetuin (Gibco) according to Tuppy and Gottschalk [10]. Free N-acetyl neuraminic acid was analyzed by the thiobarbituric acid assay [11] as modified [12]. All of the N-acetyl neuraminic acid was removed from acid-treated fetuin. 100 ~g of human corticosteroid-binding globulin was desialylated using a 1 ml column of agarose~oupled neumminidase (N-acetylneuraminate glycohydrolase, EC 3.2.1.18, Clostridium perfringens, Sigma Chemical Co.) at pH 5.0 (0.1 M sodium acetate buffer). After extensive dialysis, 45/~g of protein was recovered. 5/zg of the human asialo~orticosteroid-binding globulin was iodinated with a 2 : 1 ratio of chlorarnine T : protein (w/w). 86% of the iodinated human asialo-corticosteroid-binding globulin comigrated with radioinert human asialo~orticosteroid-binding globulin on polyacrylamide gel electrophoresis. Affinity chromatography with cortisolSepharose indicated that at least 60% of the 12SI-labeled human asialocorticosteroid-binding globulin bound cortisol. Rat liver plasma membranes were prepared by the m e t h o d of Emmelot et al. [13]. Binding of 12SI-labeled human corticosteroid-binding globulin to the plasma membranes was performed in duplicate at 37°C according to the assay m e t h o d of van Lenten and AshweU [6] except that the membranes were collected on prewet 25-mm Gelman glass fiber filters (Type AE). To evaluate the biological activity of 12SI-labeled corticosteroid-binding globulin, cortisol hemisuccinate-Sepharose was prepared according to the method of Rosner and Bradlow [8]. The/~mol of cortisol hemisuccinate bound/ml Sepharose was 1.76 and 0~70, respectively, for two preparations. Small affinity columns (bed volume = 100/~1) were prepared in 1.0 ml disposable glass pipets, precoated with 250/lg gelatin, and washed with 1.0 ml 0.05 M sodium phosphate buffer (pH 7.4) before use. 12SI-labeled human corticosteroid-binding globulin was applied to the column at room temperature. After 10 rain, the columns were washed with 0.9 ml of 0.050 M sodium phosphate buffer, pH 7.4, to remove non-bound radioactivity and then with 0.9 rnl of 0.1 M NaC1 in 0.050 M sodium phosphate buffer, (pH 6.5), which contained 200/zg/ml of cortisol to competitively remove intact ~2SI-labeled human corticosteroid-binding globulin bound to the column. Results 71--83% of the radioactivity for various preparations of ~2SI-labeled human corticosteroid-binding globulin (n ffi 4) comigrated in polyacrylamide gel electrophoresis with radioinert h u m a n corticosteroid-binding globulin (results not shown). Bovine serum albumin was required to stabilize the ~2SI-labeled human corticosteroid-binding globulin during storage and lowering the chloremine T:

546

protein ratio from 7 : I to 2 : 1 (w/w) resulted in a slightly higher proportion of the counts comigrating with human corticosteroid-binding globulin. Binding of i2SI-labeled human corticosteroid-binding globulin to the cortisol affinity columns was specific, since there was an absence of binding of ~2sIlabeled human corticosteroid-binding globulin in the presence of cortisol and heat inactivated ~2SI-labeled human corticosteroid-binding globulin bound nonspecifically and irreversibly. The level of chloramine T used in iodination had no effect on the a m o u n t of 12SI-labeled human corticosteroid-binding globulin specifically bound to the column. Three different preparations of ~2SI-labeled human corticosteroid-binding globulin bound specifically to the affinity columns at 77, 76 and 73% of the total counts added. Thus, the a m o u n t of biologically active (specifically bound) 12SI-labeled human corticosteroid-binding globulin was the same as the a m o u n t of 12SI-labeled human corticosteroidbinding globulin comigrating with human corticosteroid-binding globulin in polyacrylamide gel electrophoresis. Fig. 1 shows the plasma clearance of both total and acid-precipitable i2sIlabeled human corticosteroid-binding globulin at various times after intravenous injection. The non-linear curve of acid-precipitable radioactivity indicates complex clearance kinetics of 12SI-labeled human corticosteroid-binding globulin. This is especially evident when compared to the clearance of a non-

10o

~

5o

1oo 50

"6 100.

50

~o

0

I

1

I

I

3 6 Time efter mjectton (h)

I

12

F i g . 1. P l a s m a c l e a r a n c e c u r v e o f 1 2 5 1 4 a b e l e d h u m a n c o r t i c o s t e r o i d - b i n d i n g g l o b u l i n a n d 1 2 $ 1 . l a b e l e d bovine serum albumin in ovariectomized rats. Ovarieetomized rats were intravenously injected with a p p r o x . 1 0 7 c p m o f 1 2 5 l - l a b e l e d h u m a n c o r t i c o s t e r o i d - b i n d / n g g l o b u l i n o r 125 I - l a b e l e d b o v i n e s e r u m a l b u m i n a t t i m e 0 . 0 . 2 - - 0 . 3 m l b l o o d waS c o l l e c t e d b y j u g u l a r v e i n p u n c t u r e a t i n d i c a t e d t / r a e s . @ i, t o t a l 7 : 1 1 2 5 I - l a b e l e d ; p l a s m a p a r c h l o r l c a c i d - p r e c i p i t e b l e 1 2 5 I w i t h 7 : 1 (o o) a n d 2 : 1 (A-----.--.-A) e h l o r a m i n e T : p r o t e i n ( w / w ) r a t i o d u r i n g i o d i n a t i o n (-+S.D., n ffi 4). I n s e t . C o m p a r i s o n o f t h e plasma clearance of acid-precipitable 12$I-labeled human cortieosteroid-binding globulin and 125Ilabeled bovine serum albumin at early times in ovarleetomized rats. • ~, 1 2 5 I - l a b e l e d b o v i n e s e r u m a l b u m i n ; o - - - - - - - - ~ , 1 2 5 I . l a b e l e d h u m a n c o r t i e o s t e r o i d - b i n d i n g g l o b u l i n ( ~ S . D . , n ffi 4).

547 glycoprotein such as iodinated bovine serum albumin. As shown in Fig. 1 (inset), i2sI-labeled bovine serum albumin has a slow, linear exponential clearance during the first 30 min after intravenous injection. The initial half-lives (tl/2) calculated for 12SI-labeled bovine serum albumin and 12SI-labeled human corticosteroid-binding globulin are 61.5 and 7.5 min, respectively. If the initial rapid clearance during the first 15 min is ignored, the ill 2 for 12SI-labeled human corticosteroid.binding globulin is 51 min. A similar clearance pattern of 12SI-labeled human corticosteroid-binding globulin was observed in rats which were injected with either 0.5 or 2/~g estradiol. Sephadex G-100 column chromatography of rat plasma during the first hour after injection indicates that the proportion of high molecular weight radioactivity is virtually unchanged during the first 15 min for both 12SI-labeled human corticosteroid-binding globulin (Fig. 2) and 12SI-labeled bovine serum albumin. The small portion of low molecular weight radioactivity present in the injected radiolabel is rapidly removed during this interval. In rats receiving 12Sllabeled human corticosteroid-binding globulin this fraction begins to increase again at 30 min and is prominent at 60 min. There is a concomitant decrease in the high molecular weight fraction during this time in rats receiving 12SI-labeled human corticosteroid-binding globulin. In contrast, only the high molecular weight 12sI component was present in the plasma of rats receiving 12SI-labeled bovine serum albumin during the 30 rain after injection (results not shown). From rats studied in metabolic cages, the amount of radioiodine accumu-

30-

~o

zero time

t

20-

8-

CL

?

u

o Fracbon number Fig. 2. S e p h a d e x G - 1 0 0 gel filtration of rat plzum~a a f t e r injection o f 1 2 S I - l a b e l e d h u m a n cort/eosteroidbinding g l o b u l i n . See Fig. 1 f o r e x p e r i m e n t a l d e t a i l s . S a m p l e s consisted o f 1 0 0 ~d p l a m n a f r o m a d n 8 1 e rat a t t h e i n d i c a t e d t i m e s a f t e r i n j e c t i o n . S e p h a d e x c o l u m n w a s t~m a t r o o m t e m p e r a t u r e a f t e r e q u i l i b r a t i o n w i t h 0 . 0 5 0 M s o d i u m p h o s p h a t e b u f f e r , p H 7 . 4 . C o l u m n d i m e n z d o n s w e r e 0 . 7 X 1 8 e m , 0 . 5 rnl f r a c t i o n s . V 0 ffi b l u e D e x t r a n 2 0 0 0 v o i d v o l u m e ; V t = N a l 2 $ 1 b e d v o l u m e .

548 lated in the urine over a 48 h period represents approximately 50% of the total tracers injected (n = 12). The radioactivity present in the urine was greater than 99% acid soluble. No change in these patterns was seen when 0.5 or 2.0 ~g estradiol-17/3 was injected (data n o t shown). Thus, the urine represents a major, but n o t the sole, route of long-term elimination. Injection of estradiol-17~ had no effect on either the plasma kinetics or the urinary accumulation of radiotracers, but did induce a cornified vaginal smear, evidence that the rats were, indeed, estrogenized. Tissues sampled directly for the total a m o u n t of radioactivity accumulated at various time intervals over a 48 h period after injection of l~SI-labeled human corticosteroid-binding globulin are shown in Table I. The thyroid compartment does not significantly contribute to the clearance of the radioactivity, as it accumulated less than 1% of the injected dose after I h and only approx. 7% of the injected dose at 48 h. The kidney compartment contains a small proportion of the injected radioactivity at all times examined, remaining relatively constant for the first hour and gradually tapering off as the plasma radioactivity levels decline. Only the liver accumulated significant levels of radioactivity during the course of the experiment, especially in the early time intervals: 36% of the injected dose at 30 rain, 50--60% of which is acid precipitable. This is in contrast to the liver accumulation of 12SI-labeled bovine serum albumin which is only 2.7% of the injected dose at 30 rain. From these data it is apparent that the liver is a major source of the rapid metabolism of human corticosteroid-binding globulin in the rat. That the liver metabolism is directly related to the glycoprotein nature of human cortico: steroid-binding globulin is suggested by the study shown in Fig. 3. When 12SI-labeled human corticosteroid-binding globulin is injected into rats in the presence of 5 mg asialofetuin, the half-life of the 12SI-labeled human cortico-

TABLE I DI STR IBUT ION OF 125I-LABELED HUMAN CORTICOSTEROID-BINDING IN PLASMA, KIDNEY, L I V E R AND T H Y R O I D AT V A R I O U S TIMES A F T E R I N T R A V E N O U S INJECTIONS Overleetomized female rats were injected with approx. 107 epm 125I-labeled h u m a n eort/costeroidbinding at t = 0. A t i n d i c a t e d times, k / d n e y s and liver were re move d and h o m o g e n i z e d as desc,4bed i n the text. Thyro ids were c o u n t e d i n t a c t and plasma values are from 50-pl duplicates. All values represent the t o t a l c o u n t s p r e s e n t in t h e organs d/vided b y the injected dose. Plasma is assumed t o be 5 ~ of the b o d y weight for calculations. Numbers in parentheses are + S.D. Time

n

Plamna

Thyroid

Kidne y

Liver

Tot a l

2 rain

4

93% (18)

0.0"/2% (0.026)

1.22% (0.166)

12.18% (1.90)

106%

30 rain

4

28.60 (2.88)

0.243 (0.056)

1.06 (0.05)

35.68 (6.33)

66%

60 m ~

4

22.78 (1.22)

0.788 (0.319)

1.70 (0.12)

7.36 (2.50)

32.6

24 h

4

1.83 (0.80)

6,62

(0.35)

0.251 (0.1M)

(0.202)

0,570 (0.270)

6.90 (2.60)

0.088 (0.028)

0.440 (0,141)

48 h

4

0.864

9.6

8.0%

549

50"

"~ " ' ~

I,,

"\

"\

,~

\

\

~ 10o

\\\

"---.. N

\\\ ,I

~

10

0

I 05

, 10 0 Ttme after injection (h)

t 30

Fig. 3. Plasma clearance curves of 12SI-labeled h u m a n corti e oi t e roi dJ oi ndi ng globulin and 12SI*labeled h u m a n asialo-cortieosteroid-binding globulin in the presence of aJdalofetuin. Rats were injected and p l a s m a samp led as in legend to Fig. 1. 5 m g of -.dmlofetuin or native fe t ui n were injected w i t h t he radiotracer, o . . . . . . o, 12$I-labeled h u m a n eorticosteroid-binding gl obul i n + na!-lofetutn; • . . . . . .e, 12SIlabeled h u m a n c o r t i e o i t e r o i d - b i n d i n g globulin + f e t u i n ; Q n 12Si.labeled h u m a n ~elm]o-corticosterold-binding g/obtttin + n _ ~ o f e t u i n ; • ~--, 1251-1abeled h u m a n n~llo-cortlcosteroid-bin0~n_i globulin + fetu ln (+S.D., n ffi 4).

steroid-binding globulin is extended to about 90 rain as opposed to the short half-life of about 12 min observed for 12SI-labeled h u m a n corticosteroid-binding globulin in the presence of 5 mg native fetuin. When human corticosteroidbinding globulin is desialylated with neuraminidase, iodinated and injected simultaneously with 5 mg fetuin, plasma levels of the acid-precipitable radioactivity drop rapidly to 4% of the injected dose in 5 rain and remain stable at 2--3% for 2 h. The half-life of 12SI-labeled human asialo~orticosteroid-binding globulin under these conditions is less than 1 rain. In the presence of 5 mg of the competitive inhibitor, asialofetuin, ~2sI-labeled human aslalo-corticosteroidbinding globulin plasma levels are maintained at 90--100% of the injected dose of 20 rain and then decline in a biphasic exponential curve with a half-life of 40 rain to less than 1% of the injected dose at 2 h after injection. The data indicate that the carbohydrate moiety of human corticosteroidbinding globulin is intimately involved in this glycoprotein's rapid clearance from the plasma and subsequent binding to the liver. The results do n o t indicate whether this rapid clearance is a result of desialylation in vivo, damage from the iodination procedure, carbohydrate heterogeneity in the purified h u m a n corticosteroid-binding globulin, or simply a species effect. Therefore, a series of binding experiments was performed with isolated rat liver plasma membranes. As shown in Fig. 4, rat liver plasma membranes bind ~2SI-labeled h u m a n asialo~orticosteroid-binding globulin on sites which are competed for in a dose~iependent manner by asialofetuin. In three experiments, 27, 26 and 21% of the t2SI-labeled human asialo~:orticosteroid-binding globulin bound to

550 loo E"

80-

5

60-

c

40-

o

20-

~

o oool

I I 001 0 0 2

I I Q05 01

I 10

I 10

A s t a l o f e t w n ( IJg )

Fig. 4. I n h i b i t i o n of 12 S I-labeled h u m a n asialo.corticosteroid-binding globulin bi ndi ng to rat liver plasma m e m b r a n e s by asialofetuin. Liver plasma m e m b r a n e b i n d i n g was assessed at 37°C using 60 pg pl a s ma m e m b r a n e p r o t e i n in the assay described by Van L e n t e n and Ashwell [6]. Each p o i n t represents t he average of d u p l i c a t e tu bes in two separate experiments.

the plasma membranes in the absence of any inhibitor and binding was inhibited 92, 93 and 89% by 10/~g of asialofetuin.10/~g of intact fetuin or human corticosteroid-binding globulin had an inhibitory effect on the binding of ~2Sllabeled human asialo¢orticostexoid~bindingglobulin to the plasma membranes of only 27.3 and 38.5%, respectively.This level of inhibition corresponds with the inhibition seen with 0.010 and 0.022 #g of asialofetuin,respectively,and indicates that these preparations of fetuin and human corticoste~oid-binding globulin contain a m a x i m u m of 0.10 and 0.22% by weight of desialylated protein, assuming an equal affinityof human asialo~orticosteroid-bindingglobulin and asialofetuin for the plasma membranes. Studies of 125I-labeled human corticosteroid-bindingglobulin binding to the plasma membranes indicate a low affinity for the native iodinated human corticosteroid,bindingglobulin,as only 2.35% of the 12SI-labeled human corticosteroid.binding globulin preparation bound to the plasma membranes, and 0.001--1/~g of asialofetuln did not reduce this binding nor did 0.01--10/~g human corticosteroid-bindingglobulin. Discussion

The iodinated human corticosteroid-binding globulin used in this study is both physically intact and biologically active. Gel electrophoresis of the 12sIlabeled human corticosteroid-binding globulin indicated that 70--80% of the radioactivity applied to the columns comigrates with human corticosteroidbinding globulin. Approximately the same proportion (76%) of the ~2SI-labeled human corticosteroid-binding globulin binds to and is subsequently eluted with cortisol from a cortisol-Sepharose affinity column at room temperature. Binding studies of rat liver plasma membrane preparations also indicate that iodination of the human corticosteroid-binding globulin did n o t produce asialo, or asialo-like corticosteroid.binding globulin. Plasma kinetics of the trace a m o u n t of 12SI-labeled human corticosteroidbinding globulin used in these studies were complex, with the slopes of the clearance curves indicating a three or four c o m p o n e n t clearance. The majority (approx. 75%) of the 12SI-labeled human corticosteroid-binding globulin is

551 removed from the plasma during the first 15 min of circulation with a half-life of 7.5 rain. Direct sampling of the thyroid showed that it was not significantly involved in the initial or subsequent clearance of radioactivity. The Sephadex G-100 profiles of plasma taken at various intervals up to 1 h of circulation show that the radioactivity removed from the plasma is associated with the high molecular weight fraction and that it is replaced by radioactivity associated with the low molecular weight eluate. The observation that 12SI-labeled bovine serum albumin, also a foreign protein, is slowly cleared from the circulation suggests that the rapid clearance seen with ~2SI-labeled human corticosteroid-binding globulin is not an immune response and that the iodination procedure did not have a deleterious effect (e.g. oxidation or agglutination) on this protein. The half-fife of 12SI-labeled human corticosteroid-binding globulin in the rat is much more rapid than the value reported for humans by Sandberg et al. [4]. The early sampling times used in this report versus the relatively late sampling used by Sandberg and colleagues (at 1 h post-injection and then at daily intervals) may account for some of the discrepancy. Their estimate of corticosteroid-binding globulin's half-life (4.7--6.0 days} was derived from the latter portion of a multicomponent curve after 2--4 days when only 10--15% of the total injected radioactivity was remaining in the plasma. Also, the rat is a much smaller animal than the human, with a higher metabolic rate, and this probably contributes to a faster turnover of plasma proteins [14]. In a perfusion study using male rat liver, human asialo-corticosteroid-binding globulin was removed from the perfusate in 30 min while 80% of the native human corticosteroid-binding globulin was still present at 60 min of perfusion [15]. The tl/2 values estimated from the graph in this publication are approximately 12 min for human asialo-corticosteroid.binding globulin and 167 min for human corticosteroid-binding globulin, both of which are much greater than our values (less than 1 min and 7.5--51 min, respectively} obtained in vivo. The discrepancy might be explained in part by the high concentration of human asialo-corticosteroid-binding globulin and human corticosteroid-binding globulin used in their study (1 mg), which may be saturating the liver receptors [7] and/or sialidases, thereby slowing human asialo-corticosteroid-binding globulin and human corticosteroid-binding globulin metabolism. The absence of any other organs or blood in the perfusion system also eliminates other tissue sources of sialidase, which may also decrease the liver metabolism of human corticosteroid.binding globulin. It is not known if rat corticosteroid-binding globulin is cleared from the circulation as quickly as human corticosteroid.binding globulin in the rat. Measurement of the rat corticosteroid-binding globulin concentrations in rat plasma at different times after injection of donor rat serum to acceptor rats indirectly suggests that rat corticosteroid-binding globulin is also rapidly cleared from the circulation [16]. Efforts to purify rat corticosteroid-binding globulin are underway in our laboratory. Corticosteroid-binding globulin clearance in our study is similar to that seen with other native and asialoglycoproteins [5,14], such as human glucocerebrosidase, which has a half-life of 21 min before and 1.2 rain after desialylation [17]. Human thyroxine binding globulin (TBG) has been shown to exist in the

552 serum as both a fully sialylated species with a half-life in the rabbit of 0.8-3.4 days and as an electrophoretically slow species (STBG), which has a halflife (tl/2) of approx. 3 min in the rabbit [18]. STBG has been shown to have about one-fourth the sialic acid content of TBG [19] and to be rapidly removed by selective liver uptake [18]. The relatively rapid corticosteroidbinding globulin disappearance in the rat indicates that this protein is not the stable, unchanging entity previously proposed and thus leaves open the question of what effect this rapid turnover has on steroid metabolism and activity. Both direct sampling of the liver and binding in vitro to liver plasma membranes indicates that the liver is a primary organ responsible for the uptake of corticosteroid-binding globulin. The presence of approximately 35% of the injected trace in the liver at 30 rain after injection demonstrates a high affinity of this organ for the labeled corticosteroid-binding globulin, while the simultaneous injection of asialofetuin inhibits this uptake and maintains plasma levels of ~2SI-labeled human corticosteroid-binding globulin (Fig. 3). Asialofetuin blocks the liver uptake of asialoglycoproteins, thereby lengthening their plasma half-lives [14,20]. Asialofetuin has been shown to be catabolized exclusively by the liver and is taken up preferentially by the hepatocytes and n o t the non-parenchymal cells of the liver in vitro [21]. That intact ~2SI-labeled h u m a n corticosteroid-binding globulin is maintained longer in the circulation in the presence of asialofetuin, but does not bind to isolated rat liver plasma membranes indicates that its sialic acid is being rapidly removed during its circulation throughout the body. Sialidase (neuramidase) is a ubiquitous enzyme occurring throughout the body in almost a dozen different organs (Ref. 23) and is believed to be the first step in the catabolism of glycoproteins and also of the blood cells including erythrocytes, platelets, thymocytes and lymphocytes [23]. The present studies indicate that the desialylation process is the rate-limiting step for corticosteroid-binding globulin's metabolism, as asialocorticosteroid.binding globulin is almost instantaneously removed from the circulation (tl/2 ~ 1 rain) while intact human corticosteroid-binding globulin has a much slower clearance as indicated above. This is in line with the concept that the liver receptor for asialoglycoproteins has a high capacity [7], while endogenous sialidases, working at suboptimal pH [23] would be expected to be the first and perhaps the rate-limiting event leading to corticosteroid-binding globulin metabolism. In summary, evidence has been reported here to show that human corticosteroid-binding globulin in the rat has a rapid, complex clearance from the plasma which is primarily due to liver catabolism, although extrahepatic routes, such as the gastrointestinal tract and urinary excretion may play a role, especially in long-term metabolism. The liver uptake of both native and aisalo~ortocosteroid-binding globulin is blocked by asialofetuin in vivo, but binding of native human corticosteroid-binding globulin to liver plasma membranes in vitro does n o t occur, suggesting that this glycoprotein is desialylated by endogenous sialidases in vivo, this being the first step in its catabolism.

Acknowledgements We thank D.K. Mahajan, Ph.D., for assistance in preparation of the affinity columns. The support and advice of Doctor B. Little is appreciated. The assis-

553 tance o f Tim Billiar in some of the studies is gratefully acknowledged. K.H. was suppo~ed by, NIH T32 HD-07120 and this work was supported by NIH HD02378 (Dr. B. Little) and NIH T32 HD-07120. References 1 Seal, U.S. and Doe, R.P. (1966) in Steroid Dynamics (Pineus, G., Nakao, I. and Talt, J.F., eds.), p. 69, Academic Press, New York 2 Seal, U.S. and Doe, R.P. (1965) Steroids 5, 827---841 3 Westphal, U. (1971) Steroid-Protein Interactions, Springer-Verlag, New York, Heidelberg, Berlin 4 Sandberg, A.A., Woodruff, M., Rosenthal, H., Nienhouse, S. and Slaunwhite, W.R., Jr. (1964) J. Clin. Invest. 4 3 , 4 6 1 - - 4 6 6 5 Ashwell, G. and Morell, A.G. (1974) Adv. Enzymol. 41, 99--128 6 Van Lenten, L. and Ashwell, G. (1972) J. Biol. Chem. 247, 4633--4640 7 Regoeczi, E., Debanne, M.T., Hatton, M.W.C. and Koj, A. (1976) Biochim. Biophys. Acta 5 4 1 , 3 7 2 ~ 364 6 Rosner, W. and Bradlow, H.L. (1971) J. Clin. Endocrinol. 32, 193--198 9 Greenwood, F.C., Hunter, W.M. and Glover, J.S. (1963) Biochem. J. 89,114--123 10 Tuppy, H. and Gottschalk, A. (19"/2) in Glycoproteins (Gottschalk, E., ed.), 2nd edn., p. 440, Elsev/er, Amsterdam 11 Warren, L. (1959) J. Biol. Chem. 234, 1971--19#5 12 Keene, E.L. (1972) Methods Enzymol. 28, 413--421 13 Emmelot, P., Bos, C.J., van Hoeven, R.P. and van BlitterswiJk, W.J. (1974) Methods Enzymol. 31, 75---9O 14 Gregoriadis, G. (1975) Front. Biol. 43, 273--274 15 Van Baelen, H. and Mannaerts, G. (1974) Arch. Biochem. Biophys. 163, 53---56 16 Yamamoto, S. and Ohsawa, H. (1976) Biochem. Biophys. Res. C o m m u n . 72, 489---498 17 Furbish, F.S., Steer, C.J., Barranger, J.A., Jones, E.A. and Brady, R.O. (1978) Biochem. Biophys. Res. C o m m u n . 81, 1047--1053 18 Refetoff, S., Fang, V.S. and Marshall, J.S. (1975) J. Clin. Invest. 56, 177--187 19 Marshall, J.S., Pensky, J. and Green, A.M. (1972) J. Clin. Invest. 51, 3173--3181 20 Morell, A.G., Gregoriadis, G., Scheinberg, I.H., Hickraan, J. and Ashwell, G. (1971) J. BioL Chem. 246, 1461--1467 21 ToLleshaug, H., Berg, T., NiIsson, M. and Norum, K.R. (1977) Biochira. Biophys. Aeta 499, 73---84 22 LaBadie, J.H., Chapman, K.P. and Aronson, N.N., Jr. (1975) Biochem. J. 152, 271--279 23 Bocci, V. (1976) Experlentia 32, 135--140