Ontogenesis lodothyronine-5'- Deiodinasedm5migu4zj3pb.cloudfront.net/manuscripts/111000/111652/JCI8411165… · Ontogenesis of lodothyronine-5'-Deiodinase ... purification. AG-IX-10

Ontogenesis of lodothyronine-5'-DeiodinaseInduction of 5'-Deiodinating Activity by Insulin,

Glucocorticoid, and Thyroxine in CulturedFetal Mouse Liver

Kanji Sato, Hisoko Mimura, Doo Chol Han, ToshioTsushima, and Kazuo ShizumeDepartment of Medicine, Institute of Clinical Endocrinology,Tokyo Women's Medical College, Ichigaya-Kawada-cho 10,Shinjuku-ku, Tokyo, Japan #162; Foundation for Growth Sciencein Japan, Shinjuku-ku, Tokyo, Japan

Abstract. To elucidate the regulatory mecha-nism of ontogenetic development of iodothyronine-5'-deiodinase in the fetal and neonatal period, fetal mouseliver of the 19th day of gestation, in which no iodothy-ronine-5'-deiodinating activity was detectable, was cul-tured in Dulbecco-Vogt medium supplemented with10% thyroid hormone-depleted fetal calf serum, insulin,hydrocortisone, and thyroid hormones. Iodothyronine-5'-deiodinating activity of the homogenate was assessedby the amount of iodide released from outer-ring-labeledreverse T3 and expressed as picomoles of 1271- permilligram of protein per minute. The enzyme activitywas induced in a dose-dependent manner; optimal con-centrations for insulin, hydrocortisone, and thyroxinewere 1 ;ig/ml, 0.4 jg/ml, and 10-6 M, respectively.Without supplementation of either hydrocortisone orthyroxine, no 5'-deiodination was detected. The enzymeactivity was observed after 3 d of culture, peaked at days14-20, and then gradually decreased. Lineweaver-Burkanalysis revealed that the increase in activity was pri-marily due to an increase in Vmax (day 3, 0.2 pmol/mgprotein per min; day 20, 2.5 pmol/mg protein per min).

A part of this work was presented at the International Symposium for"Peripheral Metabolism of Thyroxine: Biochemical Background andClinical Meaning," 15-18 May 1983, at the Reisensburg Castle, Ulm,Germany.

Address reprint requests to Dr. Sato, Department of Medicine,Institute of Clinical Endocrinology, Tokyo Women's Medical College,Ichigaya-Kawadacho 10, Shinjuku-ku, Tokyo #162, Japan.

Received for publication 10 February 1984 and in revised form 23August 1984.

Half maximal thyroxine (T4) and triiodothyronine (T3)concentrations were 1 X l0-7 M(free T4: 4 X 10-10 M),and 2 X 10-9 M(free T3: 5.0 X 10-" M), respectively,whereas reverse T3 did not elicit any activity at 10-8_10-6 M.

These results suggest that ontogenetic developmentof iodothyronine-5'-deiodinase in the liver of the fetaland neonatal mouse is induced by physiological concen-trations of glucocorticoid and thyroid hormones, andthat insulin plays a permissive role in enhancing T3formation from T4 in the liver.

IntroductionRecently, the ontogenesis of thyroid hormone metabolism inperipheral tissues was extensively studied in man, sheep, chick,and especially rodents (1-5). In contrast to newborn of humansand sheep, in which hypothalamic-pituitary-thyroid functionis already mature and the sera contain adult thyroxine (T4)'levels, the rat (and probably mouse) is delivered in the hypo-thyroid state. The serum T4 concentration of newborn of ratsis very low (1.6 ,ug/dl) and triiodothyronine (T3) is hardlydetectable (-5 ng/dl) (1-3). Active thyroid hormone (T3) thenincreases in parallel with serum T4. In fetal liver of mice (6)and rats (3), iodothyronine-5'-deiodinating activity is hardlydetectable. However, in the neonatal period, there occurs aprogressive increase in the enzyme activity, which peaks at 2-3 wk after birth in the case of mice (6). In parallel with theincrease in enzyme activity, serum T3 increases progressivelyduring the first month to a peak concentration of 60-90 ng/dl (2, 3). The increase in T3 production in the peripheraltissues, especially in the liver, seems to be advantageous for

1. Abbreviations used in this paper: DTT, dithiothreitol; DV-T4(-)FCS,Dulbecco-Vogt medium supplemented with nonessential amino acids,penicillin, streptomycin sulphate, and 10% resin-treated fetal calfserum; GSSG, glutathione disulfide; rT3, 3,3',5'-triiodo-L-thyronine;3,3'-T2, 3,3'-diiodo-L-thyronine; T3, 3,3',5-triiodo-L-thyronine; T4, thy-roxine; T-GSH, total glutathione.

2254 Sato, Mimura, Han, Tsushima, and Shizume

J. Clin. Invest.© The American Society for Clinical Investigation, Inc.0021-9738/84/12/2254/09 $ 1.00Volume 74, December 1984, 2254-2262

the survival of newborn mammals. However, the factorscontrolling this maturation process are not yet known.

On the assumption that iodothyronine-5'-deiodinase isregulated by certain hormones, we developed an organ culturesystem of fetal mouse liver. We found that the ontogeneticdevelopment of iodothyronine-5'-diodinase, which converts T4to T3 and reverse T3 (rT3) to 3,3'-T2 (7, 8), is induced byglucocorticoid and thyroid hormones at their physiologicalconcentrations and that insulin plays a permissive role ininducing the enzyme activity.

Methods

[3',5'-'251]triiodo-L-thyronine (1251I-rT3) with a specific activity of about1 mCi/Mg was obtained from New England Nuclear Corp. (Boston,MA). [3'-'25Iltriiodo-L-thyronine (125II-T3) and [3',5'-'23I]thyroxine ('2I-T4) with a specific activity of 1.5 mCi/Mg was prepared (9) and purified(10) as described elsewhere. Nonradioactive thyroxine (T4), 3,3',5-triiodo-L-thyronine (T3), and hydrocortisone hemisuccinate were pur-chased from Sigma Chemical Co. (St. Louis, MO) and rT3 fromCalbiochem-Behring Corp. (San Diego, CA). Our T4 was contaminatedby 0.8% T3 (determined by radioimmunoassay), and used withoutpurification. AG-IX-10 anion exchange resin (chloride form, minus400 mesh) and Dowex 50W-X4 cation exchange resin (200-400 mesh)were obtained from Bio-Rad Laboratories (Richmond, CA). Fetal calfserum and nonessential amino acids were obtained from Gibco Lab-oratories (Grand Island, NY) and other amino acids and vitamins fortissue culture from Wako Pure Chemicals, Ltd. (Tokyo, Japan).Dithiothreitol (DTT) was purchased from Nakarai Chemicals, Ltd.(Kyoto, Japan). Silicon rubber tubing (inner diam, 2 mm; outer diam,3 mm) was obtained from Imamura Co., Ltd. (Bunkyo-ku, Tokyo,Japan), and sterile membrane filters (0.45 Am, 47 mm, HAWG047)were from Millipore Corp. (Bedford, MA).

Resin treatment offetal calf serum. To remove iodothyronines infetal calf serum, it was treated with AG-iX-10 resin according to themethod of Samuels et al. (11). T4, T3, and cortisol concentrations inthe original fetal calf serum were 8 Ag/dl, 130 ng/dl, and 1.0 gg/dl,respectively. In one experiment, '251-T4 or '25I-T3 was added to thefetal calf serum and the serum was treated with resin. The serum wasprecipitated by TCA and radioactivity in the precipitate was counted.Approximately 93% of T4 and 96% of T3 were depleted by the resintreatment.

Culture conditions offetal mouse liver. Pregnant mice on the 19thday of gestation were sacrificed by dislocation of the neck, and thefetuses were removed from the placenta through an abdominal incision.They were immediately decapitated and placed in a culture dish (60mm) kept on ice. Then, a traverse abdominal incision was made andthe fetal liver was removed with forceps. Combined livers from 20-24fetuses from two pregnant mice were then cut into explants of <1mm3, which were placed on a filter membrane that had been immersedin Dulbecco-Vogt medium. Usually, 10-20 pieces of liver were placedon a membrane. The membrane with liver pieces was transferred to aculture flask (60 X 20 mm)containing 12 ml of Dulbecco-Vogt mediumsupplemented with 1% nonessential amino acids, penicillin (100 U/ml), streptomycin sulphate (100 jg/ml), 10% resin-treated fetal calfserum (DV-T4(-)FCS), and hormones (insulin, hydrocortisone hemi-succinate, and thyroid hormones) as specified in each experiment. Toculture the liver pieces floating on the surface on the medium, thefilter membrane was placed on two floating loops with diameters of

30 and 40 mm. These floating loops were made of silicon tubing. Apolyethylene shaft (2.5 X 8 mm) cut from a disposable micropipet tip(Exel tips for 20 Al, Shoei Co., Tokyo) was inserted into the tube tomake a loop, which had been sterilized by autoclaving. Pieces of fetalliver were cultured at 370C under 5% CO2and 95% air for 9 d unlessotherwise specified. Every 3 d, two thirds of the medium was replacedwith fresh medium.

At the end of incubation, the membrane was washed with serum-free Dulbecco-Vogt medium; then, the membrane surface was scrapedwith a sharp-edged spatula and hepatocytes that proliferated on themembrane were transferred to a Potter-Elvehjem homogenizer. Theywere homogenized in 0.4 ml of ice-cold 0.25 M sucrose solutioncontaining 10 mMTris-HCl buffer (pH 7.0) with a motor-driventeflon pestle at 3,000 rpm for 30 s. A 50-Ml aliquot of homogenate wastaken for measurement of protein by the method of Lowry et al. (12).The homogenate was kept at 0-40C when deiodinating activity wasmeasured on the same day. Otherwise, it was frozen at -80'C untilassayed. The enzyme activity was stable for at least 1 wk under thiscondition.

Assay for iodothyronine-'-deiodinating activity. The enzyme activitywas assessed as described elsewhere with slight modifications (13). Inbrief, hepatocyte homogenates (10-100 Mg of protein) were incubatedin 200 Ml of sodium phosphate buffer (100 mM, pH 7.0) containing10 mMDTT, 1.3 mMEDTA, and l0-7 M [3,5'-'251I]rT3. After 15-min incubation under air, the reaction was stopped by adding 0.1 mlhuman serum followed by 1 ml of cold 10% TCA. After centrifugationat 3,000 rpm for 20 min, the supernatant was applied on Dowex 50W-X4 cation exchange resin columns (200-400 mesh, 0.7 X 1.0 cm)equilibrated with 8 N acetic acid. Columns were washed twice with 1ml of 8 N acetic acid. '25I released from outer-ring deiodination of rT3was collected in 3.2 ml eluate and counted in a gamma-spectrometer.In each assay, three or four tubes without hepatocyte homogenatewere processed exactly as the test samples, and the radioactivityobtained, representing nonspecific deiodination, was subtracted fromthat of test samples. Usually, nonspecific deiodination was 2-3% ofthe total. Since monodeiodination from the 3'- and 5'- positions of rT3occurs equally, the quantity of "2l- released was multiplied by twoand iodothyronine-5'-deiodinating activity was expressed as picomolesof 127j- per milligram of protein per minute (13). Generally, thesensitivity of the assay was <0.08 pmol of '27I-/mg of protein per min.

In another experiment, '251I-rT3 metabolites were analyzed by thinlayer chromatography. In brief, the reaction was stopped after 30-minincubation by adding 10 Mg of rT3. The reaction mixture was imme-diately acidified with 0.1 ml 4 N HC1, and the acidified sample wasextracted twice with 2 ml of n-butanol. Then, 4 ml of chloroform wasadded to the combined n-butanol and mixed well. The turbid mixturewas extracted with 2 and 1 ml of 2 N NH4OH, successively. Thecombined extract was lyophilized and rT3 metabolites were analyzedby thin layer chromatography with the solvent system of n-butanolsaturated with 2 MNH4OH(13).

To investigate whether T3 is produced from T4 in cultured fetalliver, hepatocyte homogenate (-0.5 mg) was incubated in 200 Ml ofsodium phosphate buffer (100 mM, pH 7.0) containing 10 mMDTT,1.3 mMEDTA, and 1 Mg/ml T4. After 60-min incubation at 370Cunder air, the reaction was stopped by adding 0.8 ml ice-cold ethanol,and stored at -20°C for 1 h. After centrifugation at 3,000 rpm for 30min, the supernatant (40 Ml) was analyzed using the radioimmunoassaykit for T3 (Eiken Co., Tokyo). Each assay included tubes (zeroincubation tubes) that were handled exactly as described except thatthe homogenate was added immediately before extraction with ethanol.

2255 Ontogenesis of T4-Deiodinating Activity

The amounts of T3 in the zero incubation tubes were subtracted fromthe T3 in test samples. The results were expressed as nanograms of T3generated per 200 nanograms T3 per milligram protein per hour.

Measurement of total glutathione (T-GSH) and glutathione disulfide(GSSG) levels in cultured fetal liver. In some experiments, T-GSH andGSSGwere measured enzymatically (14) as described elsewhere (15,16). In brief, hepatocytes cultured as described above were washed,scraped from the membrane, and homogenized in 400 Ml of ice-cold40 mMHCl solution in a Potter-Elvehjem homogenizer. Homogenateswere transferred to another tube in which 40 ul of 50% TCA wasadded. After 30-min centrifugation at 3,000 rpm, the supernatant waskept frozen at -80'C until assayed for T-GSH. For GSSGassay,hepatocytes were homogenized in 400 ul of 100 mMN-ethylmaleimide.The homogenates were transferred to another tube, in which 40 ul of50% TCA was added. The supernatant obtained as described abovewas also kept at -80'C until assayed. T-GSH and GSSGlevels wereexpressed as micrograms per milligrams of protein.

Iodothyronine-5'-deiodinating activity of the liver offetal, neonatal,and adult mouse. Liver of fetal mice of the 19th day of gestation,neonatal mice 3 d after birth, and adult mice were weighed andhomogenized (1:20, wt/vol) in 0.25 Msucrose solution containing 10mMTris-HCl buffer (pH 7.0). The enzyme activity in the homogenatewas determined as described above.

Measurement of free T4 and T3 in the medium. Free T3 and T4concentrations in DV-T4(-)FCS medium supplemented with insulin(I M&g/ml), hydrocortisone hemisuccinate (0.1 Mg/ml), and variousconcentrations of thyroid hormones were assessed according to themethod of Uchimura et al. (17).

Results

Effects of insulin, hydrocortisone, and T4 on iodothyronine-5'-deiodinating activity. In preliminary experiments, pieces offetal mouse liver were cultured in DV-T4(-)FCS mediumwithout addition of hormones. No 5'-deiodinating activity wasdetected even if the liver was cultured for up to 4 wk.Therefore, insulin (1 ,ug/ml), hydrocortisone hemisuccinate (1jAg/ml), and/or T4 (10-7 M) were added to the DV-T4(-)FCSmedium in various combinations. As shown in Table I, neitherinsulin, hydrocortisone, or T4 alone could induce the enzymeactivity. However, when fetal mouse livers were cultured for9 d in the DV-T4(-)FCS medium supplemented with T4together with hydrocortisone, iodothyronine-5'-deiodinatingactivity was clearly demonstrated. Increased enzyme activitywas demonstrated when the DV-T4(-)FCS medium was furthersupplemented with insulin. However, supplementation of theDV-T4(-)FCS medium either with insulin plus T4 or insulinplus hydrocortisone could not induce the enzyme activity,suggesting that the fetal and neonatal development of iodothy-ronine-5'-deiodinase is regulated mainly by steroid and thyroidhormones.

Effects of T4 concentration on iodothyronine-S'-deiodinatingactivity. The dose dependence of T4-stimulated iodothyronine-5'-deiodinating activity in fetal liver cells cultured for 9 d inthe DV-T4(-)FCS medium supplemented with insulin (1 ;ig/ml) and hydrocortisone hemisuccinate (0.1 and 1.0 ttg/ml) isshown in Fig. 1. At the higher hydrocortisone hemisuccinate

Table I. Effects of Insulin (), Hydrocortisone (HC),and T4 on Iodothyronine-S'-Deiodinating Activity

Hormones added to the medium n Enzyme activity

I 4 NDHC 3 NDT4 3 NDI + HC 10 NDI + T4 3 NDHC+ T4 3 0.35±0.17I + HC+ T4 11 0.89±0.32

HC, hydrocortisone; I, insulin.Fetal liver explants were cultured in DV-T4(-)FCS medium supple-mented with insulin (I ug/ml), hydrocortisone hemisuccinate (I ;&g/ml), and thyroxine (10-' M) at various combinations for 9 d. Then,liver explants were homogenized and iodothyronine-5'-deiodinatingactivity in the homogenate was determined. The enzyme activity wasexpressed as picomoles of 2'I- per milligram of protein per minutereleased from outer ring deiodination of rT3. Data are the mean±SDof 3 to 11 experiments. ND, not detectable (<0.08 pmol of '27I-/mgprotein per min).

concentration (1.0 jig/ml), T4 at 10-8 Melicited a questionableenzyme activity; in three of five experiments, little enzymeactivity was detected. The maximum iodothyronine-5'-deiodin-ating activity was observed at 10-7_10-6 MT4. At 10-' MT14,however, the 5'-deiodinating activity was distinctly lower.

c2.0u2

2.0*0~

10 ND N

0 to9 10 8olo106 10yThyrox ine concentration (M )

Figure 1. Effects of thyroxine on iodothyronine-5'-deiodinating activ-ity. Pieces of fetal mouse liver of the 19th day of gestation in whichno thyroxine-5'-deiodinating activity was demonstrated were culturedin DV-T4(-)FCS medium supplemented with insulin (1 Mg/ml), hy-drocortisone hemisuccinate (0.1 Mg/ml or 1 Mg/ml), and variousconcentration of T4. After 9-d culture, the liver was homogenizedand the enzyme activity was determined as described in Methods.Iodothyronine-5'-deiodinating activity was expressed by picomoles ofI- released from the outer ring of rT3 per milligram of protein perminute. Dotted and white columns indicate that the livers werecultured at 0.1 and 1 Mg/ml hydrocortisone hemisuccinate, respec-tively. The data are the mean±SEMof five experiments. Note thatno enzyme activity was detected (ND) when-the livers were culturedin the medium containing <IO-8 MT4.


At 0.1 tg/ml hydrocortisone hemisuccinate, T4 elicited aslight but significant iodothyronine-5'-deiodinating activity at10-8 M T4. The activity increased steeply between 108 and10-7 Mand peaked at 10' M, but also distinctly decreased atI0-' M. At all T4 concentrations, the enzyme activity wasgreater in liver cells cultured in the DV-T4(-)FCS mediumsupplemented with physiological concentration of hydrocorti-sone hemisuccinate (0.1 gg/ml).

Effects of hydrocortisone concentrations on iodothyronine-S'-diodinating activity. As expected from the above data, therewas a dose-dependent induction of iodothyronine-5'-deiodinat-ing activity by glucocorticoid (Fig. 2). When hydrocortisonewas not added to the culture medium, no enzyme activity wasdemonstrated even if the medium was supplemented withinsulin (1 gg/ml) and T4 (10-7 M). The addition of hydrocor-tisone hemisuccinate (0.01 jig/ml) induced a minimal enzymeactivity, which increased to 1.6 pmol of 1271-/mg protein permin at 0.1 sg/ml, and peaked at 0.4 gg/ml (2.2 pmol of 1251-/mg protein per min). At a supraphysiological hydrocortisonehemisuccinate concentration (1-10 ,ug/ml), however, the en-zyme activity decreased.

Effects of insulin concentrations on iodothyronine-5'-deio-dinating activity. Iodothyronine-5'-deiodinating activity wasdemonstrated in fetal liver cells cultured for 9 d in the DV-T4(-)FCS medium supplemented with hydrocortisone hemi-succinate (0.1 ug/ml) and T4 (10-7 M), even though insulinwas not added (Table I, Fig. 3). The enzyme activity furtherincreased when insulin was added to the medium; its optimalconcentration was observed at about 1 gg/ml (Fig. 3).

Effects of T4, T3, and rT3 on the induction of iodothyronine-5'-deiodinating activity. As shown in Fig. 4, T3 was much moreeffective than T4 in inducing iodothyronine-5'-deiodinating

Cc> E

4

z._

,-

oEr E

._ 0

oEIo

Lo

2

oo 0.01 0.1 1 10 (jig/ml)

0 168 17 16 10-5 10-4 (M)Hydrocort isone hemisuccinate

Figure 2. Effects of hydrocortisone concentration on iodothyronine-5'-deiodinating activity. Fetal mouse liver on the 19th day of gesta-tion was cultured in DV-T4(-)FCS medium supplemented with insu-lin (1 Ag/ml), T4 (10-' M), and various concentration of hydrocorti-sone hemisuccinate. After 9-d culture, liver was homogenized andiodothyronine-5'-deiodinating activity was determined. Three experi-ments were done and data from one representative experiment areshown. Note that no enzyme activity was determined in the absenceof glucocorticoid.

C~~~~~~~>, t.0Et

3

n E

to

_E

0 0.01 0.1 1.0 10 (jig/ml)0 0.1 1 10 100 (mU/ml)

Insulin concentration

Figure 3. Effects of insulin on iodothyronine-5'-deiodinating activity.Fetal mouse liver on the 19th day of gestation was cultured in DV-T4(-)FCS medium supplemented with T4 (10-' M), hydrocortisonehemisuccinate (0.1 jIg/ml), and various concentration of insulin.After 9-d culture, liver was homogenized and iodothyronine-5'-deio-dinating activity was determined. The data are the mean±SEMofthree experiments. *P < 0.05.

activity in fetal liver explants cultured for 9 d in the DV-T4(-)FCS medium supplemented with insulin (1 tg/ml) andhydrocortisone hemisuccinate (0.1 gJml). While the minimumT4 concentration for induction was 0O-8 M, T3 was effectiveat 10-0 M. Maximal enzyme activity (2.0 pmol of 127T/mgof protein per min) was attained at l0-7 MT3 compared with10-6 MT4. At 10-6 MT3, the enzyme activity decreased to60% of the maximum level in two other experiments (datanot shown). In contrast to T3 and T4, rT3 was completelyinactive at concentrations from 10-8 to 10-6 M(Fig. 4).

The approximate T4 and T3 concentrations for inducing

w~~~v

0 1610 169 1618 17 166 1i5Thyroid hormone concentrations( M)

Figure 4. Effects of T4, T3, and rT3 on iodothyronine-5'-deiodinatingactivity. Fetal mouse liver on the 19th day of gestation was culturedin DV-T4(-)FCS medium supplemented with insulin (1 gg/ml), hy-drocortisone hemisuccinate (0.1 jIg/ml), and various concentrationsof thyroid hormones. After 9-d culture, liver was homogenized andiodothyronine-5'-deiodinating activity was determined. Total thyroidhormone concentrations in the medium are indicated in the abscissa.Representative data from three experiments are shown. -, T3; *, T4;v, rT3.


2-?:l E-';: 2.z

to ZCL1

c (I:r- Eto

.c 1'a.2 160) 0

'D 4,010 E0v.& 0

- I

1

half-maximal 5'-deiodinating activity were 7 X 10-8 and 2X 10-9 M, respectively. The corresponding free T4 and T3concentration determined by equilibrium dialysis were 2.8X 10`0 and 5.0 X 10-11 M, respectively, suggesting that T3 isfive to six times more active than T4.

Time course of induction of iodothyronine-S'-deiodinatingactivity by T4. To investigate the time course of induction ofiodothyronine-5'-deiodinating activity by T4 in fetal liver cells,they were cultured in the DV-T4(-)FCS medium supplementedwith insulin (1 gg/ml), hydrocortisone hemisuccinate (0.1 ftg/ml), and with or without T4 (10-7 M). As shown in Fig. 5, theenzyme activity was distinctly detected on day 3, graduallyincreased thereafter, and peaked at day 15-20. The enzymeactivity was barely detected on culture day 1 and decreasedby 4 wk in other experiments (data not shown). Lineweaver-Burk analysis revealed that the increase in enzyme activitywas primarily due to an increase in maximum velocity (Vma,,)rather than alteration of the Michaelis constant (Km) (Fig. 6).Vmax increased from 0.20 pmol of '271/mg protein per min(day 3) to 2.5 pmol 1271-/mg protein per min (day 20), whereasKm for the outer-ring deiodination of rT3 was not alteredthroughout the culture period (7.7 X 10-8 M). These Kmvaluesare comparable with those obtained from adult mouse liverhomogenate, suggesting that the iodothyronine-5'-deiodinaseinduced in vitro under the present experimental conditionreflects the 5'-deiodinase of adult mouse liver.

1.5

C.-E

*' 0 1.0

ir, a,

a.

E,c

c

& o

_o0

0 5 10 15 20

Days in culture

Figure 5. Time course of iodothyronine-5'-deiodinating activity in-duced by hormones in cultured fetal liver. Fetal mouse liver on the19th day of gestation was cultured in DV-T4(-)FCS medium supple-mented with insulin (I sg/ml), hydrocortisone hemisuccinate (0.1 g/ml), and in the presence (-0-) and absence (-o-) of T4 (10-' M).After 3-20 d in culture, liver was homogenized and iodothyronine-5'-deiodinating activity was determined. Note that no enzyme wasdetectable in the fetal liver on the 19th day of gestation (n) and thatno enzyme activity was induced when the fetal liver was cultured inthe absence of T4 for up to 15 d. In other experiments, no enzymeactivity was detected for up to 25 d in the liver cultured in theabsence of T4 in the medium. Representative data from three experi-ments are shown.

V 6I

._

cn4

E

5,

_ 5

so

a.

-1\

0

6 days

9 days12 days20 days

-2 -1 0 1 2 3 4L (10-7M)-1

Figure 6. Lineweaver-Burk analysis of iodothyronine-5'-deiodinase.Fetal liver on the 19th day of gestation was cultured in DV-T4(-)FCS medium supplemented with insulin (1 ;Lg/ml), hydrocorti-sone hemisuccinate (0.1 g/ml), and T4 (10-' M). After 3-20 d inculture, liver was homogenized and iodothyronine-5'-deiodinating ac-tivity was determined. Note the progressive increase in Vmax with nochange of Km (7.7 X 10-8 M). Vm..x and Km values of liver homoge-nate cultured for 3 d are 0.20 pmol of 1/mg of protein per min and8.0 X 10-8 M, respectively (data not shown in the figure). Data aremeans of duplicate determinations.

Characterization of iodothyronine-S'-deiodinating activityinduced in cultured fetal liver of mouse. Iodothyronine-5'-deiodinating activity in homogenates of fetal liver cultured for9 d at nearly optimal hormone concentrations (insulin, 1.0 sg/ml; hydrocortisone hemisuccinate, 0.4 ug/ml; and T4, l0-7 M)were characterized. First, the deiodinating activity was linearwith protein concentrations over the range in the assay (20-200 jg/tube). When 90 jig of protein was used, the reactionwas linear at least for 30 min. Propylthiouracil (10-6 M)inhibited 5'-deiodinating activity uncompetitively in the assaycondition. When '25I-rT3 metabolites were extracted and ana-lyzed by thin layer chromatography, only three bands corre-sponding to iodide (- 10%), 3,3'-T2 (- 11%), and rT3 (79%)were demonstrated.

Furthermore, when the homogenate was incubated withT4 for 1 h, a distinct amount of T3 was produced from T4(0.46±0.16 ng T3/mg protein per h, mean±SD, n = 4), whereasthe homogenates of fetal liver cultured for 9 d with insulin (1;g/ml) and hydrocortisone hemisuccinate (0.4 ug/ml), butwithout T4 synthesized no detectable amount of T3. Thesefindings suggest that iodothyronine-5'-deiodinase induced incultured fetal liver by hormones accepts not only rT3 but alsoT4 and converts them to 3,3'-T2 and T3, respectively.


T-GSH and GSSGconcentrations in the cultured fetal liver.Glutathione levels were investigated in fetal liver cells culturedfor 9 d in the DV-T4(-)FCS medium supplemented withinsulin (1 ,ug/ml), hydrocortisone hemisuccinate (I gg/ml),and various T4 concentrations (0-10-5 M). Irrespective of T4concentration, T-GSH concentration was - 6 ,ug/mg of protein.GSSG comprised -2% of total glutathione and was notinfluenced by thyroid hormone levels in the culture medium(Table IIA). T-GSH concentration also was not dependent onT4 concentration even though fetal liver was cultured atoptimal hormone concentrations (insulin, 1.0 ,ug/ml; hydro-cortisone hemisuccinate, 0.1 tg/ml) (Table IIB). There was nocorrelation between GSSGconcentration and iodothyronine-5'-deiodinating activity (Table II).

Iodothyronine-5'-deiodinating activity in the liver of fetal,neonatal, and adult mice. As reported previously (6), little orno iodothyronine-5'-deiodinating activity was detected in thefetal mouse liver on the 19th day of gestation. The enzymeactivity gradually increased in the neonatal period (6). Line-weaver-Burk analysis revealed that Vm. of the enzyme activityincreased from 2.45±1.87 (SD, n = 6, determined in fourexperiments) in the liver of neonatal mice (3 d after delivery)to 32.2±5.4 (SD, n = 5, determined in four experiments) pmolP-/mg protein per min in the adult. However, no significantdifference in Km was observed between the neonatal liver(6.6±2.4 X 10-8 M, mean±SD, n = 6) and the adult liver(7.2±2.9 X Io-0 M, mean±SD, n = 5) (P > 0.1), indicatingthat the increase in enzyme activity in the extrauterine life isprimarily due to an increase in the enzyme capacity.

Table H. T-GSH and GSSGLevelsin Cultured Fetal Liver of Mouse

A BT4 con- T-GSH GSSG GSSG/T-GSH T-GSHcentration (n = 4) (n = 2) (n = 2) (n = 2)

M ag/mg protein Ag/mg protein %

0 5.7±1.7 0.11 1.8 5.10-9 7.2±1.5 0.075 1.7

10-8 5.8±1.5 0.13 1.9 5.1l0-7 6.0±0.8 0.11 1.5 4.710-6 7.2±1.1 0.21 2.4 4.61-0 7.3±1.0 0.15 1.8 5.3

Fetal liver of mouse on the 19th day of gestation was cultured inDV-T4(-)FCS medium supplemented with insulin (1 Ag/ml), hydro-cortisone hemisuccinate (1.0 Ag/ml [A] or 0.1 ,ug/ml [B]), and T4 (0-10' M). After culturing for 9 d, liver was homogenized and T-GSHand GSSGwere measured. Results are the mean±SEMof four (T-GSH[A]) and the mean of two (GSSG and T-GSH[B]) experiments.Note that, in contrast to iodothyronine-5'-deiodinating activity, gluta-thione concentration is not dependent on T4 concentration in themedium.

Discussion

A number of recent studies revealed that thyroid hormonemetabolism in the peripheral tissues is regulated multifactorially,e.g., by hormones such as glucocorticoids, insulin, glucagon,and by putative cytosol cofactors ( 18-24). Therefore, it is quitereasonable to speculate that ontogenetic development of io-dothyronine-5'-deiodinase, which converts T4 to T3 and rT3 to3,3'-T2 (7, 8), is also under multihormonal regulation. Toelucidate the regulatory mechanism of this ontogenetic devel-opment, we have developed an organ culture system usingfetal mouse liver of the 19th day of gestation, in which noiodothyronine-5'-deiodinating activity was detected (6). Wefound that iodothyronine-5'-deiodinase is induced in the fetalliver by the synergistic action of glucocorticoid and thyroidhormones at their physiological concentrations attained in thesera of the fetal or neonatal period.

Although at pharmacological doses, glucocorticoid decreasesserum T3 levels in the adult human (18) and animals bydecreasing iodothyronine-5'-deiodinating activity in the periph-eral tissues such as the liver (7, 20), a physiological dose ofglucocorticoids does the opposite particularly in the fetus (1,5). Thomas et al. (25) showed in prematurely delivered sheepthat infusion of cortisol simulating the prepartum plasmacorticosteroid surge was accompanied by an increase in fetalplasma T3 concentration, which was ascribed to an increase inthe T3-generating capacity of the fetal liver (26). In this respect,the present findings that minimum and optimum hydrocorti-sone concentrations to induce iodothyronine-5'-deiodinatingactivity are on the order of 0.01 and 0.4 gg/ml, respectively,are physiologically very feasible, since these corticosteroidlevels are those attainable in the perinatal period of humansand rodents (27, 28). A supraphysiological concentration ofglucocorticoid, however, inhibits the enzyme activity in accor-dance with clinical observations in human subjects (18) (Fig.2). Our present report directly demonstrates that iodothyronine-5'-deiodinase should be added to the list of corticosteroid-inducible enzymes in the fetal liver (29). Glucocorticoids,however, cannot induce 5'-deiodinase activity in the absenceof thyroid hormone in the culture medium. A similar phenom-enon was observed in cultured GH, cells in which glucocorticoidand thyroxine synergistically increase GHsecretion by stimu-lating the synthesis of mRNAfor GH(30).

In the presence of glucocorticoid, thyroxine caused a dose-dependent increase in iodothyronine-5'-deiodinating activityin cultured fetal liver of the mouse. Stimulation by T4 ofconversion of T4 to T3 was previously reported by Kaplan andUtiger (31) and Balsam et al. (32) in the liver of normal,thyroidectomized, and hypophysectomized adult rats. Ourpresent data, however, shows directly that a physiologicalconcentration of T4 is also capable of inducing iodothyronine-5'-deiodinating activity in the fetal liver. The minimum T4and T3 concentrations in medium containing 10% of fetal calfserum to induce the enzyme activity were 10-8 M(free T4,3 X 10" M) and 10-'° M(free T3, 3 X 10-12 M), respectively.


These free T3 and T4 concentrations are attained in the serumof neonatal rodents; free T4 concentration in rat serum increasesfrom 0.79 ng/dl (- I0- " M) at birth to 2.5 ng/dl (3.2 X 10-M) at 14-22 d (1).

On the basis of free thyroid hormone concentrations, half-maximal enzyme activity was induced by 5.0 X 10-1 MT3and 2.8 X 10-' T4. These values are comparable with thoseobserved in GH1 cells (33), rat hepatocytes (34), and especiallyin chick liver embryo cells in which free T3 concentrations of4 X 10-" Mexerted 50% of maximum effect in inducing thesynthesis of malic enzyme (35). Since little or no T3 could beformed from T4 at the beginning of the organ culture, hormonalactivity should be attributed to T4 itself. It is the free thyroidhormones that are biologically active (36). Therefore, T4 mayhave one-sixth the intrinsic biological activity of T3. However,as the T4 used was contaminated with 0.8% T3, half maximalconcentration of T4 (7 X 10-8 M) contains 5.6 X 10-10 MT3,which is able to induce a significant enzyme activity (0.3 pmolof 1271-/mg. protein per min). Therefore, the dose responsecurve of T4 shifts to the right, giving rise to the real halfmaximal T4 concentration of 1 X 10-7 M. Since the corre-sponding free T4 concentration is - 4 X 10-10 M, T3 is eighttimes more active than T4 in inducing the enzyme activity.The apparent equilibrium dissociation constants (Kd) of thyroidhormones for nuclear receptors of rat liver varies widely,dependent on the study procedures. Oppenheimer et al. (37)reported Kd for T3 of 2.1 X 10-12 Mand for T4 of 4.5 X 101"M from in vivo studies. High Kd values for T3 were reportedin isolated rat liver nuclei (2.1 X 1010 M) (38). The estimatedKd values for T3 and T4 in intact GH, cells are 2.9 X 1011and 2.6 X 1010 M, respectively (39), which are virtuallyidentical with the free T3 and T4 concentrations in the mediumrequired to elicit 50% of the maximal stimulation of iodothy-ronine-5'-deiodinating activity in fetal liver. These findingssuggest that T3 induction of iodothyronine-5'-deiodinase ismediated by control of regulatory events at the level of thehepatocyte nucleus.

Since fetal calf serum contains about l0-7 MT4 and 2-3X 10-1 Mfree T4, the total T4 level in the medium containing10% fetal calf serum is 10-8 M. However, the free T4 concen-tration is largely independent of serum dilution over this range(40). Therefore, without having used the thyroid hormone-depleted serum, we could not have developed such a sensitiveculture system for thyroid hormone. These findings were

emphasized by Samuels et al. (38) who first reported a cellculture system responsive to physiological concentrations ofthyroid hormones (33).

In addition to thyroid hormone and glucocorticoid, insulinfurther stimulated iodothyronine-5'-deiodinating activity in thepresence of both hormones. However, when either thyroidhormone or hydrocortisone is not present in the culturemedium, insulin could not induce the enzyme activity. Thesefindings suggest that insulin plays a permissive role in enhancingthe enzyme activity induced by thyroid and glucocorticoidhormones. Stimulation by insulin of T3 production from T4

or thyroxine-5'-deiodination was reported in diabetic rat liver(20) and rat hepatocytes in monolayer culture (23). Further-more, a bolus injection of insulin to normal subjects (22) andpatients with diabetic ketoacidosis (41) caused a gradual increasein serum T3 concentration. Whether this stimulatory action ofinsulin resides in the stimulation of synthesis of iodothyronine-5'-deiodinase, putative cytosol cofactor, or other effects remainto be elucidated.

T-GSH concentrations in the cultured fetal liver (6 /Ag/mgof protein) were lower than our previous data found incultured hepatocytes of adult rats (15, 16). This is partly dueto an intrinsic low glutathione level in the fetal liver (42), buta more likely explanation may be that types of cells (i.e.,fibroblasts) that contain lower intracellular glutathione levelsthan hepatocytes (43) might have replicated more rapidly inthe liver explants. It should be noted that the medium used inthe present experiment was the usual Dulbecco-Vogt medium,and not the arginine-free, ornithine-supplemented mediumthat selectively favors the growth of hepatocytes since theycontain enzymes for the urea cycle (44). A lack of correlationbetween T-GSH concentration and the enzyme activity in thepresent experiment confirms our previous observation that aputative sulfhydryl reductant in the cytosol, mainly glutathione,does not modulate the enzyme activity (15). Furthermore, theGSSG level, which may modulate thyroxine-5'-deiodinationunder certain pharmacological condition ( 16), also is unrelatedto the enzyme activity in the organ culture.

The time course of the in vitro induction of iodothyronine-5'-deiodinating activity in fetal liver reflects that found in thefetal and neonatal mouse (6). The increase in enzyme activityin vitro as well as in vivo is primarily due to an increase inVmax, suggesting an increase in the enzyme capacity, whereasan alteration of Kmwas not observed. It should also be pointedout that the Km values for outer ring deiodination of rT3 inthe cultured fetal liver is identical to that of adult liver ofmouse (7 X 10-8 M) and rat (6.5 X 10-8 M) (7, 45). Takingthese observations together, we assume that fetal liver of mousecultured under the present experimental conditions in thepresence of thyroid and glucocorticoid hormones have differ-entiated enough to synthesize iodothyronine-5'-deiodinase asnewborn mice do in their neonatal period. Although the Vm..obtained in the present experiments (-3 pmol of3271jmg ofprotein per min) is less than that of adult mouse liver (- 30pmol/mg of protein per min), this may be due to the growthof nonparenchymal cells that contain less enzyme activity. Wepresume that a similar mechanism is involved in the ontogeneticdevelopment of iodothyronine-5'-deiodinase in the prenataland postnatal period of newborn sheep and humans.

AcknowledgmentsThe authors are greatly indebted to Dr. Jacob Robbins of the NationalInstitutes of Health, Bethesda, MD, for advice in the preparation ofthe manuscript, and Mrs. K. Matsumoto for typing the manuscript.

This work was supported by a research grant from the Foundationfor Growth Science, Shinjuku-ku, Tokyo, Japan.


References

1. Fisher, D. A., J. H. Dussault, J. Sack, and I. J. Chopra. 1977.Ontogenesis of hypothalamic-pituitary-thyroid function and metabolismin man, sheep and rat. Recent Prog. Horm. Res. 33:59-116.

2. Dubois, J. D., and J. H. Dussault. 1977. Ontogenesis of thyroidfunction in the neonatal rat. Thyroxine (T4) and triiodothyronine (T3)production rates. Endocrinology. 101:435-441.

3. Harris, A. R. C., S. L. Fang, J. Prosky, L. E. Braverman, andA. G. Vagenakis. 1978. Decreased outer ring monodeiodination ofthyroxine and reverse triiodothyronine in the fetal and neonatal rat.Endocrinology. 103:2216-2222.

4. Borges, M., J. LaBourene, and S. H. Ingbar. 1980. Changes inhepatic iodothyronine metabolism during ontogeny of the chick embryo.Endocrinology. 107:1751-1761.

5. Thomas, A. L., and P. W. Nathanielsz. 1983. The fetal thyroid.In Fetal Endocrinology and Metabolism. Martin, L., and V. H. T.James, editors. Academic Press, New York. 5:97-116.

6. Sato, K., and J. Robbins. 1984. Does glutathione regulatethyroxine deiodinase activity in cells? Horm. Metab. Res. In press.

7. Kaplan, M. M., and R. D. Utiger. 1978. lodothyronine metab-olism in rat liver homogenate. J. Clin. Invest. 61:459-471.

8. Eisenstein, Z., A. Balsam, J. R. Garber, and S. H. Ingbar. 1980.A study of the properties of the enzyme in rat liver that deiodinates3,3',5'-triiodothyronine to 3,3'-diiodothyronine. Endocrinology. 107:530-537.

9. Kochupillai, N., and R. S. Yalow. 1978. Preparation, purification,and stability of high specific activity '25l-labeled thyronines. Endocri-nology. 102:128-135.

10. Sato, K., and H. Cahnmann. 1980. Synthesis of [3,5-'25ljtriiodo-L-thyronine of high specific activity. Anal. Biochem. 102:237-242.

11. Samuels, H. H., F. Stanley, and J. Casanova. 1979. Depletionof L-3,5,3'-triiodothyronine and L-thyroxine in euthyroid calf serum

for use in cell culture studies of the action of thyroid hormone.Endocrinology. 105:80-85.

12. Lowry, 0. H., N. J. Rosebough, A. L. Farr, and R. J. Randall.1951. Protein measurement with the Folin phenol reagent. J. Biol.Chem. 193:265-275.

13. Sato, K., and J. Robbins. 1980. Thyroid hormone metabolismin cultured monkey hepatocarcinoma cells. Monodeiodination activityin relation to cell growth. J. Biol. Chem. 255:7347-7352.

14. Tietze, F. 1969. Enzymatic method for glutathione determinationof nanogram amounts of total and oxidized glutathione. Applicationto mammalian blood and other tissues. Anal. Biochem. 27:502-522.

15. Sato, K., and J. Robbins. 1981. Gutathione deficiency inducedby cystine and/or methionine deprivation does not affect thyroidhormone deiodination in cultured rat hepatocytes and monkey hepa-tocarcinoma cells. Endocrinology. 109:844-852.

16. Sato, K., H. Mimura, K. Wakai, T. Tsushima, and K. Shizume.1983. Modulating effect of glutathione disulfide on thyroxine-5'-deio-dination by rat hepatocytes in primary culture: effect of glucose.Endocrinology. 113:878-886.

17. Uchimura, H., S. Nagataki, T. Tabuchi, M. Mizuno, and S. H.Ingbar. 1976. Measurements of free thyroxine in diluted and undilutedsera. J. Clin. Endocrinol. Metab. 42:561-566.

18. Chopra, I. J., D. E. Williams, J. Orgiazzi, and D. H. Solomon.1975. Opposite effects of dexamethasone on serum concentrations of3,3',5'-triiodothyronine (reverse T3) and 3,3',5-triiodothyronine (T3). J.Clin. Endocrinol. Metab. 41:911-920.

19. Chopra, I. J. 1978. Sulfhydryl groups and the monodeiodinationof thyroxine to triiodothyronine. Science (Wash. DC). 199:904-906.

20. Balsam, A., and S. H. Ingbar. 1978. The influence of fasting,diabetes, and several pharmacological agents on the pathways ofthyroxine metabolism in rat liver. J. Clin. Invest. 62:415-424.

21. Harris, A. R. C., S. L. Fang, L. Hinerfeld, L. Braverman, andA. G. Vagenakis. 1979. The role of sulfhydryl groups on the impairedhepatic 3,3',5-triiodothyronine generation from thyroxine in the hy-pothyroid, starved, fetal, and neonatal rodent. J. Clin. Invest. 63:516-524.

22. Tevaawerk, G. J. M., C. J. Hurst, P. Ursik, and L. Reese.1979. Effect of insulin-induced hypoglycemia on the serum concentra-tions of thyroxine, triiodothyronine, and reverse triiodothyronine. Can.Med. Assoc. J. 121:1090-1093.

23. Sato, K., and J. Robbins. 1981. Thyroid hormone metabolismin primary cultured rat hepatocytes. Effects of glucose, glucagon, andinsulin. J. Clin. Invest. 68:475-483.

24. Kabadi, U., and B. N. Premachandra. 1983. Effect of glucagonon thyroid hormone levels in euthyroid subjects. Program of the 65thAnnual Meeting of the Endocrine Society, San Antonio, TX. 211(Abstract #521).

25. Thomas, A. L., E. J. Krane, and P. W. Nathanielsz. 1978.Changes in the fetal thyroid axis after induction of premature parturitionby low dose continuous intravascular cortisol infusion to the fetalsheep at 130 days of gestation. Endocrinology. 103:17-23.

26. Wu, S. Y., A. H. Klein, I. J. Chopra, and D. A. Fisher. 1978.Alterations in tissue thyroxine-5'-monodeiodinating activity in perinatalperiod. Endocrinology. 103:235-239.

27. Murphy, B. E. P., and C. T. L. Branchaud. 1983. Fetalmetabolism of cortisol. In Fetal Endocrinology and Metabolism.Martin, L., and V. H. T. James, editors. Academic Press, New York.5:197-229.

28. Martin, C. E., M. H. Clark, P. E. Hartman, and I. F. Cook.1977. Relationship between fetal corticosteroids, maternal progesteroneand parturition in the rat. Acta Endocrinol. 84:167-176.

29. Liggins, C. C. 1976. Adrenocortical-related maturational eventsin the fetus. Amer. J. Obstet. Gynecol. 126:931-941.

30. Shapiro, L. E., H. H. Samuels, and B. M. Yaffe. 1978. Thyroidand glucocorticoid hormones synergistically control growth hormonemRNAin cultured GH, cells. Proc. Nat. Acad. Sci. USA. 75:45-49.

31. Kaplan, M. M., and R. D. Utiger. 1978. lodothyroninemetabolism in liver and kidney homogenates from hyperthyroid andhypothyroid rats. Endocrinology. 103:156-161.

32. Balsam, A., F. Sexton, and S. H. Ingbar. 1978. The effect ofthyroidectomy, hypophysectomy, and hormone replacement on theformation of triiodothyronine from thyroxine in rat liver and kidney.Endocrinology. 103:1759-1767.

33. Samuels, H. H., J. S. Tsai, and R. Cintron. 1973. Thyroidhormone action: a cell-culture system responsive to physiologicalconcentrations of thyroid hormones. Science (Wash. DC). 181:1253-1256.

34. Mariash, C. N., C. R. McSwigan, H. C. Towle, H. L. Schwartz,and J. H. Oppenheimer. 1981. Glucose and triiodothyronine bothinduce malic enzyme in the rat hepatocyte culture: evidence thattriiodothyronine multiplies a primary glucose-generated signal. J. Clin.Invest. 68:1485-1490.

35. Goodridge, A. G., and T. G. Adelman. 1976. Regulation ofmalic enzyme synthesis by insulin, triiodothyronine, and glucagon inliver cells in culture. J. Biol. Chem. 251:3027-3032.


36. Robbins, J., and J. E. Rall. 1957. The interaction of thyroidhormone and protein in biological fluids. Recent Prog. Horm. Res.13:161-208.

37. Oppenheimer, J. H., H. L. Schwartz, D. Koerner, and M. I.Surks. 1974. Limited binding capacity sites for L-triiodothyronine inrat liver nuclei. Nuclear cytoplasmic interaction, binding constants,and cross-reactivity with L-thyroxine. J. Clin. Invest. 53:768-777.

38. Samuels, H. H. 1978. In vitro studies on thyroid hormonereceptors. In Receptors and Hormone Action III. L. Birnbauer andB. W. O'Malley, editors. Academic Press, New York. 3:35-74.

39. Samuels, H. H., and J. S. Tsai. 1973. Thyroid hormone actionin cell culture: demonstration of nuclear receptors in intact cells andisolated nuclei. Proc. Nati. Acad. Sci. USA. 70:3488-3492.

40. Oppenheimer, J. H., and M. I. Surks. 1964. Determination offree thyroxine in human serum: a theoretical and experimental analysis.J. Clin. Endocrinol. 24:785-793.

41. Madsbad, S., P. Lauerberg, J. Weeke, H. 0rskov, 0. K. Faber,C. Binder, T. Krarup, and L. Regeur. 1981. Very early change incirculating T3 and rT3 in diabetic patients. Acta Med. Scand. 209:385-387.

42. Lambert, G. H., and S. S. Thorgeirsson. 1976. Glutathione inthe developing mouse liver I: developmental curve and depletion afteracetaminophen treatment. Biochem. Pharmacol. 25:1777-1781.

43. Schulman, J. D., J. A. Schneider, K. H. Bardley, and J. E.Seegmiller. 1972. Cystine, cysteine, and glutathione metabolism innormal and cystinoic fibroblasts in vitro, and in cultured normalamniotic fluid cells. Clin. Chim. Acta. 37:53-58.

44. Leffert, H. L., and D. Paul. 1972. Studies on primary culturesof differentiated fetal liver cells. J. Cell. Biol. 52:559-568.

45. Chopra, I. J., S. Y. Wu, Y. Nakamura, and D. H. Solomon.1978. Monodeiodination of 3,5,3'-triiodothyronine and 3,3',5'-triio-dothyronine to 3,3'-diiodothyronine in vitro. Endocrinology. 102:1099-1106.


Documents

Ontogenesis lodothyronine-5'- Deiodinasedm5migu4zj3pb.cloudfront.net/manuscripts/111000/111652/JCI8411165… · Ontogenesis of lodothyronine-5'-Deiodinase ... purification. AG-IX-10