Unique tissue distribution of two distinct cellular retinoic acid binding proteins in neonatal and adult rat

Unique tissue distribution of two distinct cellular retinoic acid binding proteins in neonatal and adult rat

Biochimica et Biophysica Acta, 1033 (1990) 267-272 267 Elsevier BBAGEN 23265 Unique tissue distribution of two distinct cellular retinoic acid bind...

548KB Sizes 0 Downloads 4 Views

Biochimica et Biophysica Acta, 1033 (1990) 267-272

267

Elsevier BBAGEN 23265

Unique tissue distribution of two distinct cellular retinoic acid binding proteins in neonatal and adult rat J. S t u a r t B a i l e y a n d C h i - H u n g

Siu

Banting and Best Department of Medical Research and Department of Biochemistry, Charles H. Best Institute, Unioersity of Toronto, Toronto, (Canada)

(Received 30 August 1989)

Key words: Retinoic acid binding protein; HPLC; Tissuedistribution; (Rat) Two different species of cellular retinoic acid binding proteins, CRABP-I and CRABP-II, have been found in neonatal rat pups (Bailey, J. S. and Siu, C.-H. (1988) J. Biol. Chem. 263, 9326-9332). In this report, we describe a sensitive radio-ligand binding assay for CRABP in crude tissue extracts. The assay makes use of anion-exchange high-pressure liquid chromatography which effectively separates CRABP-II from CRABP-I, thus permitting the simultaneous quantitation of these two proteins. The distribution CRABP-I and CRABP-II in various neonatal and adult tissues of the rat has been examined. CRABP-I is the predominant species of CRABP and is present in high levels in the brain, skin and testis. CRABP-II is apparently unique to the skin of neonatal animals and it becomes undetectable in adult skin. Interestingly, CRABP-II is detected at a significant level only in the adrenals of adult animals, while neonatal adrenals express only CRABP-I and not CRABP-II. CRABP-I is present in higher levels in most organs at the neonatal stage than in the adult.

Introduction Retinoic acid, the oxidized form of vitamin A, is known to play a vital role in a number of tissues during embryonic growth and development [1-3]. It is especially important in the regulation cell proliferation and the maintenance of proper differentiation of epithelial tissues [4,5]. Recently, retinoic acid has been shown to be a natural morphogen in chick and amphibian limb buds, relaying positional information to cells in the developing limb [6-8]. The mechanisms by which retinoic acid elicit its effects are still largely unknown. A family of cellular retinoic acid binding proteins (CRABP) has been found in many tissues [9]. The observation that neonatal tissue is a rich source of CRABP and cellular retinol binding proteins is consistent with the notion that these binding proteins may be intimately involved in the action of retinoic acid in the developing animal [10,11]. It is possible that they may function as a carrier protein targetting retinoic acid specifically to its nuclear receptor [12,13] a n d / o r as a storage protein regulating the intracellular concentration of retinoic acid [14].

Correspondence: C.-H. Siu, Charles H. Best Institute, University of Toronto, 112 College Street, Toronto, Ontario, Canada, M5G 1L6.

Recently, we have discovered a novel species of CRABP (CRABP-II) in neonatal rodent tissues [15]. CRABP-II represented approx. 5% of the predominant CRABP species, CRABP-I. Although CRABP-I and CRABP-II share many similar properties, these two proteins are immunologically distinct and have different affinities for retinoic acid. This suggests that they may have unique roles in mediating the biological activities of retinoic acid. As an initial step to investigate the role of CRABP in retinoic acid action, we sought to determine the tissue distribution of CRABP-II in both neonatal and adult rat tissues. The fact that CRABP-I and CRABP-II have similar molecular weights has necessitated the development of an assay which can distinguish between these two proteins. The ligand binding assay described here makes use of the inherent advantange of anion-exchange HPLC, which allows the rapid separation of CRABP-I and CRABP-II in crude tissue extracts. Materials and Methods Materials

Adult tissues were obtained from actively growing, sexually mature Wistar-Firth rats (275-300 g) supplied by Charles River Laboratories (Montreal, Quebec). For neonatal tissue, pregnant females were taken to term

0304-4165/90/$03.50 © 1990 ElsevierScience Publishers B.V. (BiomedicalDivision)

268 and the newborn pups were sacrificed within 1 h postpartum. All-trans-[11,12- 3H] retinoic acid (48 C i / m m o l ) was purchased from New England Nuclear. Unlabelled, all-trans-retinoic acid was purchased from Sigma (St. Louis, MO.). The DEAE-5PW H P L C column, manufactured by Toya Soda Kaisha (TSK), was purchased from Mandel Scientific (Toronto, Canada). The bicinchoninic acid protein determination kit was obtained from Pierce (Rockfod Ill).

Tissue preparation Tissue from freshly killed animals was homogenized immediately in 10 mM Tris-HC1 (pH 7.2). The initial homogenate was centrifuged for 15 min at 13 000 x g to remove particulate material. The supernatant was collected and then titrated to pH 5.1 with 1 M acetic acid and the resultant precipitate was removed by centrifugation at 13000 x g for 15 rain. The supernatant was titrated to pH 7.5 with 1 M N a O H and used immediately for binding assays. All procedures were performed on ice.

(0.25% activated charcoal, 0.025% Dextran 50, 10 mM Tris-acetate (pH 8.0)) was added to each sample and allowed to mix. After an incubation period of 15 rain at 4°C, the Dextran-charcoal was removed by centrifugation at 13000 rpm for 5 rain. 200/~1 of the supernatant was then injected onto the DEAE-5PW HPLC column and run under the conditions described above. The column eluate was collected in 0.25 ml fractions at 15 s intervals between 11 and 15 min of the column gradient. Under these conditions, two specific [3H]retinoic acid binding peaks were resolved, which were eluted at 12.75 and 13.75 min, respectively.

Protein purification CRABP-I and CRABP-II were purified from neonatal rat pups as described previously [15].

Protein determination The concentration of protein in different samples was determined using the bicinchoninic acid protein assay kit, with bovine serum albumin as the standard [17].

High-performance fiquid chromatography Protein samples were injected onto a DEAE-5PW column (7.5 x 75mm, equipped with a 4 x 20 mm guard column), equilibrated in 10 mM Tris-acetate (pH 8.0) (buffer A). The column was eluted at 1 m l / m i n with 0.5 M Tris-acetate (pH 8.0) (buffer B), utilizing the following gradient profile: 0-5 rain, 0% buffer B; 5-20 min, linear to 100% buffer B; 20-25 min, 100% buffer B; 25-30 min, linear to 0% buffer B. The column eluate was routinely monitored for absorbance at 350 nm, which is specific for the bound retinoid in the protein/ligand complex. Under these conditions, purified CRABP-I and II eluted with retention times of 13.82 and 12.60 min, respectively. To quantitate the amount of bound [3H]retinoic acid, 0.25 ml fractions of the column eluate were collected at the appropriate times and added directly to 5 ml of Redi-Solv liquid scintillation cocktail (Beckman) for determination of radioactivity. Under these conditions, the counting efficiency of tritium was approx. 35%.

[3H]Retinoic acid binding assay Tissue samples were routinely exposed to long-wavelength ultraviolet light for a period of 4 h at 4°C prior to the assay procedure to destroy the endogenous ligand [16]. After this initial step, aliquots of tissue extracts ranging in amounts from 0.1 mg total protein were added to 10 mM Tris-acetate buffer (pH 8.0) to a final volume of 200/~1. 1 #Ci of [3H]retinoic acid (21 pmol) in 1 ffl of ethanol (containing 2.5 mM (+)-a-tocopherol was added to the tissue extract. To the control samples, 1 /tl of 5 mM unlabelled retinoic acid was added prior to the addition of the labelled ligand. After an overnight incubation at 4°C, 50 #1 of Dextran-charcoal solution

Results

Separation of CRABP-I and CRABP-II by HPLC We have previously shown that the two closely related retinoic acid binding proteins CRABP-I and CRABP-II can be distinguished based on their chro-

lOO C2J ('¢3

100

50

,t --]

o 0

I

%B

Z I

I

I

loo

B

50

_

d~-

- -

i

_

J

I

I

I

I

o

10

20

30

Thne (rain) Fig. 1. Separation of CRABP-I and CRABP-II by anion-exchange HPLC. 40 #g of partially purified CRABP-I (A), and an equivalent amount CRABP-II (B) were injected separately onto a TSK DEAE 5PW column equilibrated with 10 mM Tris-acetate buffer (pH 8.0). The protein bound to the column was eluted at a flow rate of 1 ml/min to a final buffer concentration of 0.5 M Tris-acetate (pH 8.0), utilizing the gradient profile shown in panel A. The eluate was monitored for absorbance at 350 nm (0.02 AUFS).

269 matographic behavior on a D E A E ion-exchange column [15]. This property was further exploited to develop a rapid separation m e t h o d for these two binding proteins using a D E A E H P L C column. Fig. 1 shows the elution profiles of partially purified samples of C R A B P - I and C R A B P - I I separated under optimal conditions. The protein samples were loaded onto the column in 10 m M Tris-acetate buffer ( p H 8.0) and eluted using a linear gradient of 10 m M to 500 m M Tris-acetate ( p H 8.0). Since both proteins retained their ability to bind retinoic acid under these conditions, the column eluate was

100

A

o

50 o

0

!

I

I

I

B

4000

2000

i

I

0

l0

i

I

20 Time (rain)

30

2000

C

m

1000

i't

l'z

1~

! 14

i 15

Time (min) Fig. 2. Resolution of two [3H]retinoic acid binding activities in the cytosol of neonatal rat by anion exchange HPLC. Approx. 1 mg of a soluble neonatal rat extract was injected onto a TSK DEAE-5PW column and eluted under conditions as described in the legend of Fig. 1. Prior to chromatography, the extract was incubated with 1 ttCi of [3H]retinoic acid in the presence or absence of a 250-fold excess of unlabelled RA. The column eluate was monitored for absorbance at 350 nm (A) and 1 ml fractions were collected for the determination of bound [3H]retinoic acid (B). Panel C, shows the radioactivity recovered in 0.25 ml fractions collected between 11 and 15 min of an identical injection. Open bars, recovery of [3H]retinoic acid; closed bars, recovery of [3H]retinoic acid in the presence of an excess of unlabelled retinoic acid.

monitored for absorbance at 350 n m which is specific for the retinoid when complexed with the binding protein. U n d e r these conditions, C R A B P - I I eluted with a retention time of 12.60 rain while C R A B P - I eluted slightly later with a retention time of 13.84 min.

[SH]Retinoic acid binding assay Based on the above separation procedure, a radioligand binding assay was developed for the simultaneous quantitation of C R A B P - I and C R A B P - I I in tissue extracts. T o demonstrate that C R A B P - I and C R A B P - I I in crude extracts would be separated under identical conditions, a neonatal rat extract was incubated with [3H]retinoic acid either in the presence or in the absence of excess unlabelled retinoic acid and then subjected to H P L C separation. The sample was resolved into several 350 n m absorbance peaks, two of which contained specific [3 H]retinoic acid binding activity (Fig. 2). A l t h o u g h the two retinoic acid binding activities were not well resolved by collecting 1-ml fractions (Fig. 2B), they could be effectively resolved into two distinct radioactive peaks by collecting 0.25 ml fractions of the column eluate. The retention times of these two specific retinoic acid binding activities were centered at 12.75 min and 13.75 min (Fig. 2), corresponding closely to those of the purified proteins (Fig. 1). Free retinoic acid remained b o u n d to the column under these conditions and could be removed by washing the column with 20% methanol. The linearity of the assay was determined over a wide range of protein concentrations. Varying a m o u n t s of soluble rat testis protein were incubated with [3H]retinoic acid under standard assay conditions and the a m o u n t of radioactivity recovered from the C R A B P - I peak was quantitated for each sample. The results showed that the a m o u n t of [ 3H]retinoic acid associated with the C R A B P - I peak was directly proportional to the a m o u n t of protein added to the assay (Fig. 3). The assay was linear up to 1 mg of testis protein, which yielded 1.7 pmole of C R A B P - I , an a m o u n t c o m p a r e d favorably with the reported value. Also, the maximal a m o u n t of cytosolic protein which could be loaded onto the colu m n was 1 mg. The C R A B P - I and C R A B P - I I concentrations tested in all subsequent studies were adjusted to fall within this linear range. The lower limit of detection of C R A B P - I or C R A B P - I I was estimated to be about 2.0 pmol per g of tissue or 1 ng per mg protein.

Expression of CRABP-I and CRABP-H in neonatal and adult rat tissues The levels of C R A B P - I and C R A B P - I I were determined in a variety of neonatal and adult rat tissues using the above assay procedure. The only neonatal tissue that contained a substantial a m o u n t of C R A B P - I I was skin (Table I). Fig. 4 shows a typical [3H]retinoic

270 20

_

A

,~,~ (..).~ 60 30 15

0.25

0.5

0.75

1.0

'C, r..-,

10

Protein, mg Fig. 3. Linearity of the binding assay. Varying amounts of rat testis extract were added to the standard assay mixture containing [3H]retinoic acid with or without a 250-fold excess of unlabelled retinoic acid. The samples were subjected to HPLC separation. The amount of radioactivity recovered in the CRABP-I peak was determined after subtracting the amount recovered in the presence of an excess of unlabelled ligand. Values represent the average of two independent determinations.

acid b i n d i n g assay performed on n e o n a t a l a n d adult rat skin cytosol. The a m o u n t of C R A B P - I I represented approx. 5% of the a m o u n t of C R A B P - I detected in n e o n a t a l skin. In both adult a n d n e o n a t a l skin, high levels of C R A B P - I were f o u n d in the cytosol (Table I). A l t h o u g h the level of C R A B P - I did n o t change significantly between the n e o n a t a l tissue a n d the adult tissue, C R A B P - I I was n o longer detectable in the adult skin. The patterns of C R A B P - I a n d C R A B P - I I expression in the adrenal gland is also distinctly different from other tissues. Fig. 5 shows the results of a typical

L)

0

25

20 o X v

15

£) 10

\x~\x•,xx£xxx "~ I

11 TABLE I Level of cellular retinoic acid-binding proteins in adult and neonatal tissues

Soluble protein (mg/g tissue)

CRABP-I (pmol/g tissue)

CRABP-II (pmol/g tissue)

Neonatal Brain Liver Adrenal Kidney Small Intestine Skin

43 82 52 63 25 17

38.7 + 4.5 n.d. 6.5 + 0.8 13.5 _+ 1.1 2.9+_ 0.3 82.5+-12.2

n.d. n.d. n.d. n.d. n.d. 3.8+-0.3

Adult Brain Liver Adrenal Kidney Small Intstine Skin Testis

35 45 59 79 34 22 40

11.7+_ 0.8 n.d. n.d. 4.0 +- 0.5 n.d. 84.7 + 10.1 64.6 +- 8.8

n.d. b n.d. 5.5 + 0.6 n.d. n.d. n.d. n.d.

Tissue a

a Tissue samples were tested for retinoic acid binding activity as described in Materials and Methods. Values represent the mean of three determinations. b n.d.; not detected; less than 2.0 pmol/g wet weight.

i

12

L !

13

14

15

Time (min) Fig. 4. Separation retinoic acid binding activities in the cytosol of neonatal and adult skin. 275 ~g of neonatal (A) or adult (B) skin cytosolic protein, equilibrated with [3H]retinoic acid, was injected onto a TSK DEAE anion-exchange column and eluted under conditions as described in the legend of Fig. 1. Beginning at time 11 min, 0.25 ml fractions were collected for the determination of radioactivity. Open bars, recovery of [3H]retinoic acid; closed bars, recovery of [3H]retinoic acid in the presence of an excess of unlabelled retinoic acid.

[3H]retinoic acid b i n d i n g assay for n e o n a t a l a n d adult a d r e n a l gland. Quite unexpectedly, C R A B P - I I was f o u n d to be expressed exclusively in the adrenal g l a n d of the adult animal. A m o n g the tissues tested, the a d r e n a l was the only adult tissue in which C R A B P - I I was detected (Table I). Interestingly, C R A B P - I was detected exclusive of C R A B P - I I in the cytosol of the n e o n a t a l a d r e n a l . T a b l e I summarizes the results obtained for the q u a n t i t a t i o n of C R A B P - I a n d C R A B P - I I in various n e o n a t a l a n d adult tissues. The values were normalized to represent the a m o u n t of p r o t e i n recovered in terms of p m o l of b i n d i n g p r o t e i n recovered per g of tissue wet weight. I n n e o n a t a l rat, C R A B P - I was present in relatively high levels in the cytosol of

271 20

A

10

0

20

B

o. U

10

0

1'2

;3

1'5

Time (min) Fig. 5. Retinoic acid binding activities in the cytosol of neonatal and adult adrenal gland. 425/Lg of neonatal adrenal (A), or adult adrenal (B) cytosolic protein, equilibrated with [3H]retinoic acid, was injected directly onto a TSK DEAE anion-exchange column and eluted under conditions as described in the legend of Fig. 1. Beginning at 11 min, 0.25 ml fractions were collected for the determination of radioactivity. Open bars, recovery of [3H]retinoic acid; closed bars, recovery of [3H]retinoic acid in the presence of an excess of unlabelled retinoic acid.

skin, brain and kidney. Low levels of CRABP-I were found in the cytosol of small intestine and adrenal. The liver was the only tissue tested in which CRABP-I was not detected. In adult tissues, the highest levels of CRABP-I were found in the skin, testis and brain cytosol. CRABP-I was not detected in the cytosol of spleen, liver, small intestine or adrenal. Only a small amount of CRABP-I was recoverable from the cytosol of adult kidney. As indicated above, CRABP-II rather than CRABP-I was present in moderate levels in the adult adrenal gland. With the exception of skin, the level of CRABP-I in a given neonatal tissue was generally at least 3 fold higher than that in the corresponding adult tissue. Discussion

In this report, we have described a sensitive radioligand binding assay for the simultaneous quantification of CRABP-I and CRABP-II in crude tissue extracts. Since CRABP-I and CRABP-II have similar molecular weights [15], classical ligand binding assays, which rely on sucrose gradient separation [18,19], size exclusion chromatography [20], or polyacrylamide gel electrophoresis [21] to separate the unbound ligand from the protein/ligand complex, were not able to distinguish between CRABP-I and CRABP-II. The assay we have developed relies on anion-exchange H P L C to

separate the bound from the unbound ligand, achieving at the same time effective separation of CRABP-I and CRABP-II. The sensitivity of this ligand binding assay was on the order of 2.0 pmol retinoic acid bound per g of tissue. This degree of sensitivity compares favorably with the sensitivity of other quantitative methods for CRABP-I, including specific radioimmunoassays [22,23], and the levels of CRABP-I in various tissues tested here are in broad agreement with results reported previously for some of these tissues [20-23]. Since C R A B P - I I has a lower affinity for retinoic acid and its association with retinoic acid is more labile [15], it is possible that the CRABP-II content in different tissues may be underestimated by our method. This aspect should be evaluated when alternative methods for C R A B P - I I determination become available. Retinoic acid has been implicated as being an important regulator of fetal growth and development [1-3,24]. Our results show that the level of expression of CRABP-I is generally about 3-fold higher in neonatal rat tissue than in the corresponding tissue of the adult animal. This observation is true for all tissues tested with the exception of skin, where the level of CRABP-I does not appear to change between the neonatal animal and the adult animal. Skin is an important target organ for retinoids. It has been suggested that retinoic acid controls both the proliferation of mitotic basal cells and their differentiation into mucus-secreting colunmar cells or keratin-producing squamous cells [4,25]. The high levels of CRABP-I in both neonatal and adult skin is likely a reflection of the special requirement of skin for retinoic acid. It is also notable that the neonatal skin is the only neonatal organ that contains a significant level of CRABP-II, suggesting that this binding protein m a y have a unique role in skin development. Further studies on the relative distribution of CRABP-I and C R A B P - I I in the various cell types of embryonic skin should provide some clues on the role of these binding proteins in skin development. The adult testis has the second highest level of CRABP-I. The testis has long been recognized as an important target organ for vitamin A and retinoic acid [4,26,27]. Retinoids are obligatory for normal sperm development. Cellular binding proteins for retinol and retinoic acid are present in distinct cell types of this complex organ [28,29]. Although detailed mechanisms are still unknown, it is becoming evident that retinoids have multiple functions in several cell types of the testis. It is of interest to note that adult adrenal gland contains primarily CRABP-II. CRABP-I is not detectable in rat adrenals, consistent with the observation of Ong et al. [22]. On the other hand, bovine adrenals have been found to be a rich source of C R A B P [30]. The discrepancy between these two species may reflect the different roles of retinoic acid in adrenals, which are still unknown.

272 A current model concerning the role of CRABP-I in mediating the action of retinoic acid suggests that the protein functions to shuttle its ligand to specific binding sites within the nucleus [12,13]. Recently, Daly and Redfern [31] have demonstrated the presence of nuclear retinoic acid receptors (RAR) in F9 embryonal carcinoma cells, cDNA clones for three specific nuclear retinoic acid receptors, RARa, RARfl and RARa, have also been isolated and characterized and all three forms of RAR apparently belong to the superfamily of steroid hormone receptors [32-36]. These nuclear receptors are thought to regulate gene transcription in a retinoic acid-dependent manner. It is, therefore, conceivable that CRABP delivers retinoic acid specifically to these nuclear acceptor proteins, which in turn regulate the transcription of retinoic acid responsive genes. RARct and RARfl have an approximate 10-fold difference in their affinity for retinoic acid, the estimated K d being 10 -8 and 10 - 9 for RARa and RARfl, respectively [34]. These values parallel the respective affinity of CRABP-II and CRABP-I for retinoic acid [15,19]. It is possible that target tissues may make use of the different CRABP and RAR species to fine tune their respective responses to retinoic acid. In this regard, it is of interest to note that RARfl mRNA has been detected in adult tissues that express a high level of CRABP-I [35,37]. On the other hand, the expression of R A R a mRNA is more restricted and RARe, is apparently unique to skin [36]. Adult adrenal glands which express CRABP-II is among the organs that express RARa. The significance of such correlations is not clear. However, one possibility is that each species of CRABP may target retinoic acid to a specific species of RAR in the nucleus. Further studies to evaluate the relationship of CRABP and RAR in target tissues should help us better understand their respective roles in retinoic acid action.

Acknowledgments We thank Dr. Robert Murray for discussion and advice. This work was supported by grants from the Medical Research Council of Canada and the National Cancer Institute of Canada. J.S.B. was supported by a Medical Research Council studentship.

References 1 Summerbell, D. (1983) J. Embryol. Exp. Morphol. 78, 269-289. 2 Tickle, C., Alberts, B., Wolpert, L. and Lee, J. (1982) Nature 296, 564-566.

3 Edwards, M.K., Harris, J.F. and McBurney, M.W. (1983) Mol. Cell. Biol. 3, 2280-2286. 4 Wolbach, S.B. and Howe, P.R. (1925) J. Exp. Med. 42, 753-777. 5 Shapiro, S.S. (1986) In Retinoids and cell differentiation. (Sherman, M.I., ed.), pp.29-60, CRC Press, Boca Raton. 6 Maden, M. (1984) Dev. Biol. 98, 409-416. 7 Thorns, S.D. and Stocum, D.L. (1984) Dev. Biol. 104, 319-328. 8 Thaller, C. and Eichele, G. (1987) Nature 327, 625-628. 9 Chytil, F. and Ong, D.E. (1984) In The retinoids (Sporn, M.B., Roberts, A.B. and Goodman, D.S., eds.), Vol. 2, pp.90-119, Academic Press, New York. 10 Ong, D.E. (1984) J. Biol. Chem. 259, 1476-1482. 11 Bailey, J.S. and Siu, C.-H. (1988) Biochem. Cell Biol. 66, 750 757. 12 Takase, S., Ong, D.E. and Chytil, F. (1986) Arch. Biochem. Biophys. 247, 328-334. 13 Barkai, U. and Sherman, M.I. (1987) J. Cell Biol. 104, 671-678. 14 Maden, M., Ong, D.E., Summberbell, D. and Chytil, F. (1988) Nature 335, 733-735. 15 Bailey, J.S. and Siu, C.-H. (1988) J. Biol. Chem. 263, 9326-9332. 16 Ong, D.E. and Chytil, F. (1980) Methods Enzymol. 67, 288-296. 17 Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner, H.F., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J. and Klenk, D.C. (1985) Anal. Biochem. 150, 76-85. 18 Ong, D.E. and Chytil, F. (1975) J. Biol. Chem. 250, 6113-6117. 19 Ong, D.E. and Chytil, F. (1978) J. Biol. Chem. 253, 4551-4554. 20 Rainer, S., Herrera, J.M. and McCormick, A.M. (1983) Arch. Biochem. Biophys. 225, 818-825. 21 Siegenthaler, G. and Saurat, J.-H. (1987) FEBS Lett. 166, 209-214. 22 Ong, D.E., Crow, J.A. and Chytil, F. (1982) J. Biol. Chem. 257, 13385-13389. 23 Kato, M., Blaner, W.S., Mertz, J.R., Das, K., Kato, K. and Goodman, D.S. (1985) J. Biol. Chem. 260, 4832-4838. 24 Roberts, A.B. and Sporn, M.B. (1983) Cancer Res. 43, 3034-3042. 25 Wolf, G. (1980) In Human Nutrition. (Alfin-Slater, R.B. and Kritchevsky, D., eds.), pp.97-203, Plenum Press, New York. 26 Howell, J., Thompson, J.N. and Pitt, G.N. (1963) J. Reprod. Fertil. 5, 159-167. 27 Thompson, N., Howell, J.M. and Pitt, G.A. (1964) Proc. R. Soc. Biol. 159; 510-535. 28 Kato, M., Sung, W.K., Kato, K. and Goodman, D.S. (1985) Biol. Reprod. 32, 173-189. 29 Porter, S.B., Ong, D.E., Chytil, F. and Orgebin-Crist, M.C. (1985) J. Androl. 6, 197-212. 30 Sundelin, J., Anundi, H., Tragardh, L., Eriksson, U., Lind, P., Ronne, H., Peterson, P.A. and Rask, L. (1985) J. Biol. Chem. 260, 6488-6493. 31 Daly, A.K. and Redfern, C.P.F. (1987) Eur. J. Biochem. 168, 133-139. 32 Petkovitch, M., Brand, N.J., Krust, A. and Chambon, P. (1987) Nature 330, 444-450. 33 Giguere, V., Ong, S., Segui, P. and Evans, R.M. (1987) Nature 330, 624-629. 34 Brand, N., Petkovitch, M., Krust, A., Chambon, P., De The, H. Marchio, A., Tiollais, P. and Dejean, A. (1988) Nature 332, 850-853. 35 Benbrook, D., Lernhardt, E. and Pfahl, M. (1988) Nature 669 672. 36 Zelent, A., Krust, A., Petkovich, M., Kastner, P. and Chambon, P. (1989) Nature 339, 714-717. 37 de The, H., Marchio, A., Tiollais, P. and Dejean, A. (1989) EMBO J.B, 429-433.