Mammary Gland Type I Iodothyronine Deiodinase Is Encoded by a ...

77 downloads 131 Views 2MB Size Report
Maia AL, Berry ML, Sabbag R, Harney JH, Larsen PR 1995 Structural and functional differences in the dio1 gene in mice with inherited type 1 deiodinase.
0013-7227/97/$03.00/0 Endocrinology Copyright © 1997 by The Endocrine Society

Vol. 138, No. 10 Printed in U.S.A.

Mammary Gland Type I Iodothyronine Deiodinase Is Encoded by a Short Messenger Ribonucleic Acid* LUZ NAVARRO, ABRAHAM LANDA, CARLOS VALVERDE-R,

AND

CARMEN ACEVES

Departamento de Neuroendocrinologı´a, Centro de Neurobiologı´a, Campus Juriquilla Queretaro, y Departmento de Microbiologı´a y Parasitologı´a, Facultad de Medicina, Mexico, D. F., Universidad Nacional Autonoma de Mexico, Mexico ABSTRACT Lactating rat mammary gland expresses a deiodinating activity that, on the basis of kinetic characteristics, corresponds to the socalled 59-deiodinase type I (D1). In the present study we amplified and sequenced several D1 complementary DNA (cDNA) fragments from rat lactating mammary gland. The mammary cDNA was found to be identical to the previously reported rat liver cDNA in the coding region, but 465 nucleotides shorter on its 39-untranslated region, suggesting that the D1 is the same in both tissues. D1 messenger RNA

(mRNA) was also detected by reverse transcriptase-PCR in mammary glands from puberal and late pregnant rats, but not in virgin animals. Densitometric analysis showed a close and direct correlation between mRNA content and enzyme specific activity in mammary gland. Our results also show that rat liver contains both D1 mRNA forms and that the large form may respond to the thyroid status. These data suggest a differential and organ-specific expression of these mRNA forms, which could play a role in the functional regulation of D1 activity. (Endocrinology 138: 4248 – 4254, 1997)

T

Northern analysis and a reverse transcriptase-PCR assay (RT-PCR). The researchers suggested that the D1 in lactating mammary gland could be encoded by a different mRNA (12). In this context, sequence analysis of several D1 complementary DNAs (cDNAs) have revealed that the enzyme is highly homologous in all species studied (2, 15–17). However, a detailed analysis of these data suggest that D1 seems to be encoded by at least two forms of mRNA. Thus, Berry et al. (2) described a mRNA of 2094 bp in the rat liver with two potential polyadenylation signal sites (the first at 1612 bp and the second at 2069 bp); this clone was designated G21. Recently, hepatic D1 mRNAs for the dog (15), mouse (16), and human (17) have been described. Dog and mouse messengers exhibit 76 – 86% homology compared with G21, both contain only the first polyadenylation signal and end at approximately 1600 bp, thus being about 450 bp shorter than G21 on the 39-untranslated region (39UTR). In the case of the human, only one sequence has been described and is incomplete because the polyadenylated [poly(A)] tail was not found (17). However, as in the case of rat and mouse, the reported human sequence also contains a polyadenylation signal close to and downstream of the selenocysteine insertion sequence (SECIS) region (1825 bp). The physiological importance of these different transcripts has not been elucidated. However, several studies on other transcripts have demonstrated that the 39UTR is involved in the control of stability of mRNA and represents a site of functional regulation (18 –21). The present study was undertaken with the aim to reevaluate the presence of D1 mRNA in lactating rat mammary gland. To this end, a set of specific primers to G21 were used in the RT reaction of the RT-PCR procedure, thus increasing the probability of identifying D1 mRNA(s). The results strongly suggest that 1) lactating mammary gland D1 is identical to that expressed in the rat liver, but is encoded by a shorter mRNA form; 2) during the breeding cycle, there is a close and direct correlation between mRNA content and

HE PERIPHERAL conversion of T4 to T3 or rT3 is catalyzed by a group of enzymes known as iodothyronine deiodinases (1). All members of this family contain the uncommon amino acid selenocysteine (Se-Cys) at their active sites (2– 4). Type I deiodinase (D1) is a microsomal protein of approximately 29 kDa that, depending on the sulfation state of its substrate, is able to deiodinate either the phenolic or the tyrosyl ring of T4 and other iodothyronines. D1 activity is sensitive to inhibition by propylthiouracil (PTU) or gold thioglucose, a property that is useful in distinguishing this activity from the rest of deiodinases. (5). This enzyme is expressed predominantly in the liver, kidney, and thyroid gland, but it is also detected in other organs, such as heart, anterior pituitary gland, and lactating mammary gland (5, 6). Within the past few years several studies have provided convincing evidence showing that lactation is accompanied by a characteristic euthyroid sick-like syndrome and opposite changes in mammary gland and liver deiodinase activity. As lactation progresses there is a clear-cut increase in this enzyme activity in the mammary gland (6, 7) and a concomitant decrease in the liver (7–9). Furthermore, the kinetic characterization of this enzymatic activity in rat lactating mammary gland has revealed that it corresponds to D1 (6, 10, 11). Rat mammary D1 activity during lactation is only about 0.5% of that observed in the liver (11, 12) and exhibits a significant increase in the presence of specific stimuli, such as lactation intensity (7, 13) or overfeeding (14). However, recently, Jack et al. (12) were unable to detect D1 messenger RNA (mRNA) in the rat lactating mammary gland using Received February 5, 1997. Address all correspondence and requests for reprints to: Dr. Luz Navarro, Centro de Neurobiologia, Universidad Nacional Autonoma de Mexico, Apartado Postal 1–1141, Queretaro 76001, Qro. Mexico. E-mail: [email protected]. * This work was supported in part by Grants DGAPA-UNAM IN203492 and IN206496 as well as NIH-Fogarty Award TWO5215– 01.

4248

MAMMARY D1 IS ENCODED BY A SHORT mRNA

4249

TABLE 1. Schematic representation of G21 ATG 7 M1s 3

TGA 382 M2s 3

TAG 778 M3s 3

M4s 3 4 M6as

SECIS 1516-1600 J2s 3

J3s 3 4 M7as

Primer

Size

Location

M1s M2s M3s M4s M6as M7as J3s J2s J2as J1as As Bas AP UAP

17-mer 20-mer 20-mer 16-mer 18-mer 18-mer 20-mer 17-mer 17-mer 17-mer 20-mer 20-mer 36-mer 20-mer

66-82 377-396 483-502 523-538 610-627 660-677 1138-1157 1477-1493 1477-1493 1582-1598 1693-1712 1809-1829

4 J2as

As 3 AAAAA 4 Bas

4 J1as Sequence

CTT GCA CAC CAG CTG ATC GTT TTG GCT CAT CTG CCC GTC ATA

GGA CCT AGA CAC GCT CTG TGC GAC ACC CTT CCG TTC CAC CTA

GGT GAC TGG CGA GCT CCT CAC TGA CGT CCC TCT CAC GCA GTC

GGC CTT ATG AGC CTG TCC TCA CGG CAG GAC TTT AAC TCG GAT

TAC CAT GGC CTC GTT TGT TAG GTA TCC ATT CCG ACT ACT GCG

Purpose

GG TTC TTT C CTG ATC AAT GC AA TT ATA TCA AGT TGG

TT TA

CA

GC CA AT(17) AC

PCR PCR PCR Blotting PCR RT Blotting PCR PCR RT PCR PCR RT-RACE PCR-RACE

The upper panel shows a schematic representation of G21. Arrows represent the position and direction of the different oligonucleotides used. Exact position, length, and sequence are illustrated in the lower panel.

mRNA form that corresponds to G21 increases during hyperthyroidism and decreases during hypothyroidism and lactation. Together, these data show a differential and organspecific expression of D1 mRNA forms, which suggests that they may play a role in the functional regulation of this enzyme. Materials and Methods Animals Sprague-Dawley rats were obtained from Charles River Breeding Laboratories (Burlington, MA) and kept under conditions of controlled lighting (12 h of light, 12 h of darkness) and temperature. Food and water were available ad libitum. The animals used in this study were adult male euthyroid (7 weeks old), hypothyroid (thyroidectomized 3 weeks before killing), hypothyroid and T4 replaced (daily sc injections of 1.5 mg of T4/100 BW for 8 days before killing), hyperthyroid (daily sc injections of 20 mg T4/100 BW for 8 days before killing), female puberal (4 weeks old), virgin (7 weeks old), 14-day pregnant, and lactating (1- and 10-day postpartum) with litter adjusted to 10 pups each. Animals were killed by decapitation, and tissues were immediately dissected, homogenized, and processed for RNA extraction or deiodinase activity determination.

59-Deiodinase assay

FIG. 1. Northern blot analysis of D1 mRNA. Poly(A) RNA (5 mg/lane) was extracted from liver of hyperthyroid and euthyroid male rats and from mammary glands of 12-day pregnant female (G-12 M.G.) and 10-day lactating female (L-10 M.G.). A, At low stringency; B, at high stringency.

enzyme specific activity in mammary gland, suggesting that this enzyme is regulated predominantly by pretranslational mechanisms; 3) rat liver D1 is encoded by two messengers that differ in the length of their 39UTR; and 4) the large

D1 activity was determined using a modification of the methods previously described (22) and characterized for mammary gland (7). In brief, tissues were homogenized in ice-cold 0.32 m sucrose, 10 mm HEPES (pH 7), 5 mm dithiothreitol, and 1 mm EDTA and centrifuged at 1000 3 g at 4 C. The reaction mixture containing 5 mg (liver) or 200 mg (mammary gland) supernatant protein, and 0.5 mm of an isotopic solution of [125I]rT3 (New England Nuclear, Boston, MA; SA, 1250 mCi/mg) and 5 mm dithiothreitol. The assay was carried out with and without 1 mm PTU. Incubations were performed at 37 C for 1 h (liver) or 3 h (mammary gland). The released acid-soluble radioiodide was isolated by chromatography on Dowex 50W-X2 columns. Proteins were measured by the Bradford method (Bio-Rad protein assay, Bio-Rad, Richmond, CA). Results are expressed as picomoles of radioiodide released per mg protein/h.

4250

MAMMARY D1 IS ENCODED BY A SHORT mRNA

FIG. 2. Amplification of the mammary gland mRNA 59-region. We used M7as primer for RT reaction. PCR was carried out on a 5-ml (mammary glands) or 1-ml (liver) aliquot of the reverse transcription with M1s-M6as oligonucleotides. Lanes 1 and 2 correspond to euthyroid liver, with (RT1) and without (RT2) RT, respectively. Lanes 3 and 4, Twelve-day pregnant mammary gland (RT1 and RT2, respectively). Lanes 5 and 6, Ten-day lactating mammary gland (RT1 and RT2, respectively); lane 7, control (H20 with all the reagents). A, Gel stained with ethidium bromide. B, Southern blotting of the RTPCR products. Hybridization was carried with labeled M4as oligonucleotide.

Endo • 1997 Vol 138 • No 10

FIG. 3. PCR-RACE amplification of 39UTR of D1 mRNAs. Oligo(dT) with a specific sequence denoted AP (see Table I and Materials and Methods) was used for the RT reactions. In A, PCR was carried out with the complementary sequence oligo(dT), UAP, and J2s. Southern blotting was performed using the nested J1AS oligonucleotide as a probe. In B, PCR was carried out with oligonucleotides As and Bas, and radiolabeled As was used for Southern blotting. PCR amplifications were carried out on 1-ml (livers) or 5-ml (mammary glands) aliquots of the RT reaction. Lanes are as follows: eu, euthyroid male liver; hyper, hyperthyroid male liver; Tx, hypothyroid male liver; Tx1T4: replaced T4 hypothyroid male liver; eu, euthyroid female liver; lactating, lactating female liver; mammary gland; RT-, euthyroid male liver without RT and H2O with all the reagents.

Northern blot analysis Total RNA was extracted using a modification of the Chirgwin method as previously described (23). Poly(A) RNA was isolated from total RNA using an oligo(deoxythymidine)-cellulose [oligo(dT)-cellulose] column. Northern blot analysis was performed using 5 mg poly(A) RNA from liver or mammary glands. Two different stringency conditions were used in the hybridization; low [6 3 SSC (standard saline citrate), 5 3 Denhardt’s, 0.5% SDS, and 100 mg/ml salmon sperm DNA; overnight at 55 C; washes in 0.1 3 SSC and 0.1% SDS at room temperature and at 55 C] and high [5 3 SSC, 5 3 Denhardt’s, 0.1 m phosphate buffer (pH 6.5), 100 mg/ml salmon sperm DNA, and 50% formamide; washes in 0.1 3 SSC and 0.1% SDS at room temperature and at 42 C]. The G21 D1 cDNA, provided by Drs. M. Berry and P. R. Larsen (Brigham and Women’s Hospital, Boston, MA), was used as a probe in these studies.

labeled oligonucleotide probe (M4s). Similar methods were used to detect and amplify the 39-region of G21-associated transcripts, except that J1as oligonucleotide was used in the RT reaction, and oligonucleotides M2s and J2as were used as primers in the PCR amplification. Southern blotting of these amplicons was performed using the nested J3s oligonucleotide as a probe. Mixture containing a RNA sample and the appropriate oligonucleotide primers, but without the RT, was used as control. RT-PCR assay was also used to analyze the presence of D1 mRNA in mammary gland during the different stages of the breeding cycle. M7as oligonucleotide was used for the RT reaction. PCR was carried out on a 5-ml (mammary glands) or a 1-ml (liver) aliquot of the reverse transcription with M2s-M6as oligonucleotides. Radiolabeled M4s oligonucleotide was used for Southern blotting. Densitometric analysis was performed in a PhosphorImager analyzer (Molecular Dynamics, Sunnyvale, CA).

RT-PCR assays Oligonucleotide primers used in the RT-PCR assays are shown in Table 1 and were derived from the sequence of G21. Primers were synthesized by Life Technologies (Grand Island, NY). To detect and amplify the coding region from G21-associated transcripts, 2 mg lactating mammary gland or liver total RNA were reverse transcribed using a Superscript RT (Life Technologies) and a specific D1 antisense oligonucleotide primer (M7as). PCR was then carried out on a 5-ml aliquot of the RT reaction mixture, using oligonucleotide primers M1s and M6as. Other components of the PCR mixture included 2.5 pmol of each oligonucleotide primer, 200 mm deoxy-NTPs, and 2.5 U Taq polymerase (Promega) in a 50-ml total volume reaction. Amplification was carried out for 32 cycles of 94 C for 45 sec, 54 C for 45 sec, and 72 C for 1 min. Amplified fragments were electrophoresed on 1% agarose gel, transferred to a nylon membrane, and then hybridized with a nested radio-

Amplification of the 39UTR The length of the 39UTR D1 mRNA was assessed using 2 mg total RNA from liver or 10-day lactating mammary gland and a rapid amplification of 39-DNA ends (39-RACE) procedure (Life Technologies) as previously described (4). Oligo(dT) with a terminal-specific sequence (AP oligonucleotide; see Table 1) was used for the RT reactions. PCR amplification was carried out with the complementary terminal sequence of oligo(dT) denoted universal amplification sequence or UAP and J2s for 32 cycles. To verify the presence of the long form of D1 mRNA in both tissues, a PCR amplification was carried out with As and Bas oligonucleotides, the sequences of which are present only in the long mRNA form. PCR reactions contain the standard mixture and a 1-ml (liver) or a 5-ml (mammary gland) aliquot from the previous RT reaction.

MAMMARY D1 IS ENCODED BY A SHORT mRNA

4251

FIG. 4. RT-PCR amplification strategy for mammary gland D1 transcript. Schematic representation of the rat liver D1 cDNA derived from the published sequence (2). The solid bar corresponds to the coding region. The open bars correspond to the 59- and 39UTRs. The amplified fragments are shown as discontinuous lines with the oligonucleotides used for the PCR. The oligonucleotide used for the RT reaction are shown within parentheses. Arrows below each fragment indicate the direction and length of the region sequenced.

Sequence analysis The amplified fragments (see Fig. 4) were sequenced by an automated sequence system that uses the PRISM Ready Reaction Dye Deoxy Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA).

Statistical analysis All values are expressed as the mean 6 sem. Differences between groups were analyzed by one-way ANOVA and Tukey’s test or KruskalWallis one-way ANOVA. P , 0.05 was considered statistically significant.

Results

Figure 1 shows that at lower stringency (A), lactating mammary gland poly(A) RNA yields a hybridizing band smaller than that from the liver (1.6 vs. 2.1 kilobases, respectively). However, this signal disappears under high stringency conditions (Fig. 1B). Similar results were obtained when using mammary gland total RNA and Northern or slot blot (data not shown). To confirm the presence of D1 mRNAs in the lactating mammary tissue, in vitro amplifications by RT-PCR were performed. As expected, using mammary total RNA and primers M1s and M6as, a 561-bp DNA fragment was amplified (Fig. 2). A longer DNA fragment (1116 bp) was obtained by using primers M2s and J2as (data not shown). To verify that the D1 gene expression is not a ubiquitous phenomenon in the mammary gland, similar experiments were carried out using total RNA from male rat liver and mammary gland from 12-day pregnant rats. DNA fragments of the predicted size were amplified only from liver and lactating mammary gland, but not from 12-day pregnant mammary gland. The PCR-RACE method was employed to analyze the

39-terminal region D1 mRNA from liver and mammary gland, using oligonucleotide AP for the RT reaction, and J2s and UAP for the PCR amplification. Figure 3A shows that two fragments were amplified from liver cDNA, one of the expected size (617 bp) and another of 150 bp, indicating the presence of two D1-mRNA forms. The large fragment is more abundant in hyperthyroid rat liver and practically disappears in liver from hypothyroid and lactating animals. Furthermore, the large fragment is restored in T4replaced hypothyroid animals. In contrast, the short fragment remains unchanged. On the other hand, just one short fragment of 150 bp was amplified using cDNA from lactating mammary gland, thus indicating that this tissue only presents the short D1 mRNA form. To corroborate these results, we carried out a PCR amplification using oligonucleotides As and Bas, which only amplify cDNAs corresponding to the long D1 mRNA form. Fragments of the expected size (136 pb) were amplified from livers, and no amplification was obtained from mammary gland (Fig. 3B). These data confirm that the long D1 mRNA form present in liver is modified by the thyroid status and is absent in mammary gland. Figure 3A also shows a more intense signal for the short fragment than for the larger one, suggesting that the short mRNA is more abundant. Nevertheless, it is difficult to interpret the relative quantities of the two isoforms considering that a coamplification took place, and the replication efficiency is dependent on the fragment size (24). The nucleotide sequences of amplified fragments from lactating mammary gland were determined (Fig. 4) and compared with the sequence of G21 (Fig. 5). All fragments showed sequences identical with that of G21 except for 15

4252

MAMMARY D1 IS ENCODED BY A SHORT mRNA

Endo • 1997 Vol 138 • No 10

FIG. 5. Nucleotide sequence comparison of D1 cDNAs from liver (G21) and lactating mammary gland (MG) in the rat. The alignment of nucleotide sequences refers to the liver cDNA sequence (2). Dashes indicate nucleotides identical to those of the reference sequence. The initiator ATG and stop TAG codons and the two polyadenylation signals are in bold fonts; the TGA selenocysteine codon and SECIS element are underlined.

residues on the 39UTR. These data strongly suggest that both tissues express identical D1. However, the mammary gland mRNA is ;450 bp shorter in the 39UTR, suggesting that it ends at the first polyadenylation signal. We also amplified and sequenced fragments obtained from the hepatic short

mRNA form, and the sequence was identical to the sequence from G21 (data not shown). To explore the physiological regulation of the D1 gene in the mammary gland, D1 mRNA was examined in female rats during the breeding cycle and compared with the specific

MAMMARY D1 IS ENCODED BY A SHORT mRNA

enzymatic activity (Fig. 6). D1 transcript levels were determined using a RT-PCR assay and mammary gland total RNA extracted from puberal, virgin, pregnant, and lactating rats. PCR products generated using primers M2s and M6as showed that D1 mRNA is present in puberal, 14-day pregnant, and 1- and 10-day lactating animals. The data also show that there was a direct correlation between the amount of amplified fragments and the specific enzyme activity. Discussion

A short form of D1 mRNA in lactating rat mammary gland was identified by using a modified RT-PCR analysis that allows the identification of scarce or not abundant transcripts. Indeed, densitometry approximations revealed that the lactating mammary gland contains about 200 times less D1 mRNA than the liver. These findings are consonant with our failure to detect this mRNA by direct methods such as Northern or slot blot analyses. There is an apparent controversy between our present results and the unsuccessful efforts by Jack et al. (12) to identify this messenger using ran-

FIG. 6. Comparison between D1 enzymatic activity and mRNA content in mammary glands during the breeding cycle. A, Enzymatic activity. B, Densitometric units (PhosphorImager analyzer) of the Southern analysis for D1 RT-PCR fragments shown in C. M7as oligonucleotide was used for the RT reaction. PCR was carried out on a 5-ml (mammary glands) or a 1-ml (liver) aliquot of the reverse transcription with M2s-M6as oligonucleotides. Radiolabeled M4s oligonucleotide was used for Southern blotting. ML, Male liver; P, puber; V, virgin; G-14, 14 days of pregnancy; L-1, 1 day of lactation; L-10, 10 days of lactation. Values are presented as the mean 6 SEM (n 5 3). Means with different letters are significantly different [by one-way ANOVA and Tukey’s HSD test (capital letters) or Kruskall-Wallis test (small letters)].

4253

dom hexanucleotides as primers for the RT reaction and a competitive PCR. However, a plausible explanation could be that these researchers used a shorter internal control for their competitive PCR than the expected fragment. It is known that when two DNA fragments are coamplified, the longer fragment experiences a decrease in its replication efficiency before the shorter one does (24). Mammary D1 cDNA sequencing revealed that despite being 465 bp shorter than G21, its coding region is identical to the corresponding region of the hepatic clone. Furthermore, in both cDNAs, the SECIS region is present in the same position. This finding indicates that the D1 expressed in the rat lactating mammary gland and liver are identical and is consistent with previous studies showing that in operational terms (Km, preferential substrate, PTU sensitivity, etc.), mammary deiodinase activity corresponds to a D1 (6, 10 –11). Similarly, the demonstration of a close and direct correlation between mammary D1 mRNA content and specific deiodinase activity in the breeding cycle allows us to propose the involvement of pretranslational mechanisms in their regulation. This agrees with previous studies indicating that mammary D1 is confined to the alveolar epithelium (11) and that it is suckling stimulus dependent (25). The present findings also demonstrate that rat liver contains two different D1 mRNA forms that differ in their lengths at the 39UTR. The size of the short D1 mRNA as well as that of the mammary D1 mRNA (1630 bp) closely correspond to the use of the first polyadenylation signal site contained in G21. Although the length of the long form corresponds to the expected size throughout the use of the second polyadenylation signal site also present in G21. Several studies have demonstrated that the 39UTR represents a site of functional regulation in mRNA stability (18 –21). In particular, the presence of an AUUUA motif between two polyadenylation signals, as in the case of G21, has been associated with the translation-dependent instability of mRNA encoding for other transcripts (20, 21). The mechanisms involved in the alternative election between polyadenylation signal sites are practically unknown and constitute an exciting new field in understanding the control of mRNA translation (19). Data obtained from this work not only confirm that thyroid hormones regulate liver D1 mRNA expression (26), but also suggest that they could form some of the factors that regulate the differential expression of D1 mRNA forms. Thus, although thyroid status has no evident effect on the hepatic short messenger, we here show that the large form increases in hyperthyroidism and decreases in hypothyroidism. These findings and the well documented euthyroid sick-like syndrome that characterizes lactation (7–9) could also explain the decrease in the long D1 mRNA form observed in lactating rat liver. Similarly, the data presented here are in agreement with the differential D1 response that occurs in the liver, kidney, and thyroid gland during hyperthyroidism (27–29), TSH administration (30), and/or selenium depletion (31). Thus, despite the fact that there is no available information concerning which mRNA forms are expressed in kidney or thyroid gland, the present results suggest that not only thyroid hormones, but TSH and selenium as well, may exert a differential and organ-specific regulation of the expression of these D1 mRNA forms. Furthermore, this differential re-

4254

MAMMARY D1 IS ENCODED BY A SHORT mRNA

sponse is consonant with the fact that G21 corresponds to the large mRNA form and was isolated from hyperthyroid male rats (2), whereas the short D1 mRNA forms described in the dog and mouse arose from euthyroid animals (15, 16). Moreover, the presence of two thyroid hormone response elements in the human Dio1 gene that have not been identified in the mouse (32) suggests a possible species-specific regulation or the existence of more than one D1 gene. However, further studies will be required to analyze the possible differences in translational efficiencies and/or stability of these messenger forms as well as the influence that factors such as TSH, selenium, or iodine may exert on their regulation. Acknowledgments We thank Dr. Galton’s laboratory at Dartmouth College, especially Jennifer Davey for their helpful advice with the molecular biology techniques. We also thank Aurea Orozco for her critical review of this paper.

References 1. Larsen PR, Silva JE, Kaplan MM 1981 Relations between circulating and intracellular thyroid hormones: physiological and clinical implications. Endocr Rev 2:87–102 2. Berry MJ, Banu L, Larsen PR 1991 Type I iodothyronine deiodinase is a selenocysteine-containing enzyme. Nature 349:438 – 440 3. St Germain DL, Schwartzman RA, Croteau W, Kanamori A, Wang Z, Brown DD, Galton VA 1994 A thyroid hormone-regulated gene in Xenopus laevis encodes a type III iodothyronine 5-deiodinase. Proc Natl Acad Sci USA 91:7767–7771 4. Davey CJ, Becker KB, Schneider MJ, St Germain DL, Galton VA 1995 Cloning of a cDNA for the type II iodothyronine deiodinase. J Biol Chem 270:26786 –26789 5. Kohrle J 1994 Thyroid hormone deiodination in target tissues–a regulatory role for the trace element selenium? Exp Clin Endocrinol 102:63– 89 6. Aceves C, Valverde-R C 1989 Type I, 59-monodeiodinase activity in the lactating mammary gland. Endocrinology 124:2818 –2820 7. Valverde-R C, Aceves C 1989 Circulating thyronines and peripheral monodeiodination in lactating rats. Endocrinology 124:1340 –1344 8. Kahl S, Capuco AV, Bitman J 1987 Serum concentrations of thyroid hormones and extrathyroideal thyroxine-59-monodeiodinase activity during lactation in the rat. Proc Soc Exp Biol Med 184:144 –150 9. Giralt M, Iglesias R, Villarroya F, Mampel T 1987 Changes in liver iodothyronine 59deiodinase activity during pregnancy and lactation in the rat. Horm Metab Res 19:510 –511 10. Kahl S, Jack LWJ, Capuco AV 1993 Characterization of thyroxine-59-deiodinase in mammary tissues of the cow, sow and rat. Livestock Prod Sci 35:177–178 11. Aceves C, Rodon C, Ramirez-C I, Wilson S. Pineda O, Lopez L. Mancilla R, Valverde-R C 1995 Mammary 59deiodinase (59D) during the breeding cycle of the rat: indirect evidence that the 59D type is specific to the alveolar epithelium. Endocr J 3:95–99 12. Jack LJ, Kahl S, St Germain DL, Capuco AV 1994 Tissue distribution and regulation of 59D processes in lactating rats. J Endocrinol 142:205–215

Endo • 1997 Vol 138 • No 10

13. Kahl S, Bitman J, Capuco AV, Keys JE 1991 Effect of lactational intensity on extrathyroidal 59deiodinase activity in rats. J Dairy Sci 74:811– 818 14. Aceves C, Navarro L, Ramirez del Angel A, Luna M, Valverde-R C 1994 Lactation selectively modifies 59monodeiodinative responses to glucose overfeeding. Endocr J 2:547–551 15. Toyoda N, Harney JW, Berry MJ, Larsen PR 1994 Identification of critical amino acids for 3,5,39-triiodothyronine deiodination by human type 1 deiodinase based on comparative functional-structural analyses of the human, dog and rat enzymes. J Biol Chem 269:20329 –20334 16. Maia AL, Berry ML, Sabbag R, Harney JH, Larsen PR 1995 Structural and functional differences in the dio1 gene in mice with inherited type 1 deiodinase deficiency. Mol Endocrinol 9:969 –980 17. Mandel SJ, Berry MJ, Kieffer JD, Harney JW, Warne RL, Larsen PR 1992 Cloning and in vitro expression of the human selenoprotein, type I iodothyronine deiodinase. J Clin Endocrinol Metab 75:1133–1139 18. Carter BZ, Malter JS 1991 Biology of disease. Regulation of mRNA stability and its relevance to disease. Lab Invest 65:610 – 621 19. Wahle E, Keller W 1992 The biochemistry of 39end cleavage and polyadenylation of messenger RNA precursors. Annu Rev Biochem 61:419 – 440 20. de Savage F, Kruys V, Marnix O, Huez G, Octave JN 1992 Alternative polyadenylation of the amyloid protein precursor mRNA regulates translation. EMBO J 11:3099 –3103 21. Kilk A, Laan M, Torp A 1995 Human CuZn superoxide dismutase enzymatic activity in cells is regulated by length of the mRNA. FEBS Lett 362:323–327 22. Leonard JL, Rosenberg IN 1980 Iodothyronine 59-deiodinase from rat kidney: substrate specificity and the 59-deiodination of reverse triiodothyronine. Endocrinology 107:1376 –1383 23. Galton VA, Morganelli CM, Schneider MJ, Yee K 1991 The role of thyroid hormone in the regulation of hepatic carbamyl phosphate synthetase activity in Rana catesbeiana. Endocrinology 129:2298 –2304 24. Stolovitzky G, Cacchi G 1996 Efficiency of cDNA replication in the polimerase chain reaction. Proc Natl Acad Sci USA 93:12947–12952 25. Aceves C, Morales T, Pineda O, Rodo´n-Fonte C, Navarro L, Valverde-R C, Mena F, T3 and iodine in milk. Mammary 59deiodinase is neurally regulated. 5th International Symposium on Hormones and Bioactive Substances in Milk, Eslovenia, 1996 26. Berry MJ, Kates AL, Larsen PR 1990 Thyroid hormone regulates type I deiodinase messenger RNA in rat liver. Mol Endocrinol 4:743–748 27. Smallridge RC, Wartofsky L, Burman KD 1982 The effect of experimental hyperthyroidism and hypothyroidism on 59-monodeiodination of 3,39,59-triiodothyronine and 39,59-diiodothyronine by rat liver and kidney. Endocrinology 11:2066 –2069 28. Kaplan MM, Utiger RD 1978 Iodothyronine metabolism in liver and kidney homogenates from hyperthyroid and hypothyroid rats. Endocrinology 103:156 –159 29. Toyoda N, Nishikawa M, Horimoto M, Yoshikawa N, Mori Y, Yoshimura M, Masaki H, Tanaka K, Inada M 1990 Synergistic effect of thyroid hormone and thyrotropin on iodothyronine 59-deiodinase in FRTL-5 rat thyroid cells. Endocrinology 127:1199 –1205 30. Erickson VJ, Cavalieri RR, Rosemberg L 1982 Thyroxine 59-deiodinase of rat thyroid, but not that of liver, is dependent on thyrotropin. Endocrinology 111:434 – 440 31. Bermano G, Nicol F, Dyer JA, Sunde RA, Beckett GJ, Arthur JR, Hesketh JE 1995 Tissue-specific regulation of selenoenzyme gene expression during selenium deficiency in rats. Biochem J 311:425– 430 32. Toyoda N, Zavacki AM, Maia AL, Harney JW, Larsen PR 1995 A novel retinoid X receptor-independent thyroid hormone response elements is present in the human type 1 deiodinase gene. Mol Cell Biol 15:5100 –5112