Thyroid Hormone Receptor Expression in Rat Placenta

Placenta (2001), 22, 353–359 doi:10.1053/plac.2000.0617, available online at http://www.idealibrary.com on

Thyroid Hormone Receptor Expression in Rat Placenta A. J. Leonard, I. M. Evans, M. R. Pickard, R. Bandopadhyay, A. K. Sinhaa and R. P. Ekins Division of Molecular Endocrinology, UCL Medical School, Mortimer Street, London W1N 8AA, UK Paper accepted 8 November 2000

The expression of c-erbA
and - encoded thyroid hormone receptors (TR) was investigated in rat placenta between 16 and 21 days of gestation (dg), and in fetal liver and brain at 16 dg, using semi-quantitative RT-PCR and nuclear 3,5,3 -triiodothyronine (T3) binding. TR
1, TR1, c-erbA
2 and c-erbA
3 mRNA abundance was unchanged in placenta between 16 and 21 dg, as was the dissociation constant (Kd) of T3 binding. The maximal T3 binding capacity (Bmax) in placenta doubled over this period, suggesting placental TR binding activity is post-transcriptionally regulated. Transcript abundance in tissues at 16 dg can be summarized: TR
1, placenta=fetal liverfetal brain; c-erbA
2 and
3, placenta=fetal liver
INTRODUCTION Maternal hypothyroidism during pregnancy disturbs rat and human fetal brain development, resulting in neurological deficits in offspring (Man, Brown and Serunian, 1991; Porterfield and Hendrich, 1993; Pickard et al., 1997; Sinha et al. 1997; Evans et al., 1999; Haddow et al., 1999; Pop et al., 1999). The underlying mechanisms are unclear, however, in the rat thyroid hormones (TH)—predominantly thyroxine (T4)—cross the placenta from mother to fetus throughout much of gestation (Obregon et al., 1984; Woods, Sinha and Ekins, 1984) and T4 selectively enters the fetal brain, where it is converted to 3,5,3 -triiodothyronine (T3) by type II 5 -monodeiodinase (Calvo et al., 1990). This is also thought to occur in humans, since TH are present from well before the onset of fetal thyroid function, in coelomic and amniotic fluid (Contempre et al., 1993), and in fetal brain and other tissues (Bernal and Pekonen, 1984; Sinha et al., 1997). TH are also present in blood from neonates suffering thyroid agenesis or total organification defects (Vulsma, Gons and Vijlder, 1989). The primary mechanism of TH action is via nuclear TH receptors (TR), which preferentially bind T3 and function as ligand-regulated transcription factors (Brent, Moore and Larsen, 1991; Lazar, 1993). Multiple TR isoforms exist—TR
1 being derived from the c-erbA
gene and TR1 and TR2 from differentially spliced c-erbA gene transcripts (Brent, Moore and Larsen, 1991; Lazar, 1993). Related c-erbA
-derived proteins, produced by differential splicing a

To whom correspondence should be addressed.

0143–4004/01/040353+07 $35.00/0

(c-erbA
2 and c-erbA
3; Koenig et al., 1989) and N-terminal truncation (c-erbA
1 and c-erbA
2; Chassande et al., 1997), do not bind T3 or act as classical TRs, but may modulate transcription by interfering with TR function (Koenig et al., 1989; Lazar, 1993; Chassande et al., 1997; Sinha et al., 1997). TR mRNA and protein, predominantly TR
1, are found in human and rat fetal brain from early gestation, as are non-T3 binding isoforms (Bernal and Pekonen, 1984; Perez-Castillo et al., 1985; Bradley, Towle and Young, 1992; Falcone et al., 1994; Iskaros et al., 2000). Maternal hypothyroidism may therefore modify TH levels and TR activity within fetal brain, disturbing its development. Maternal hypothyroidism may also however induce placental dysfunction, perturbing the flux of nutrients, oxygen and/or trophic factors from mother to fetus, disrupting brain development as a secondary consequence. This possibility seemed unlikely, since gross placental development is unaffected by maternal hypothyroidism in some studies in rat (Morreale de Escobar et al., 1985; Pickard et al., 1993), though not in others (Bonet and Herrera, 1988; Hendrich and Porterfield, 1992). A preliminary study also suggested that rat placenta expressed low levels of TR transcripts and T3 binding (Bandopadhyay et al., 1996). TRs are however expressed in human placenta (Banovac, Ryan and O’Sullivan, 1986; Ashitaka et al., 1988; Kilby et al., 1998) and aspects of human placental function are TH-sensitive (Matsuo et al., 1993; Maruo et al., 1995). Furthermore, maternal hypothyroidism induces subtle defects in rat placental glycogen (Pickard et al., 2000) and glucose transporter (Pickard et al., 1999) levels, and c-fos and c-jun expression (Leonard  2001 Harcourt Publishers Ltd

Placenta (2001), Vol. 22

354

Table 1. Primer sequences, product sizes and reaction conditions for RT-PCR of fetal tissue cDNA Target

Primer sequences; annealing sites in targeta

Product size

Ta (C)b

Cycles

18S rRNA

U: GTCCCCCAACTTCTTAGAG; 1436–1454 L: CACCTACGGAAACCTTGTTAG; 1834–1854 U: CACCCCGGCCATCACC; 990–1005 L: TGGGGCACTCGACTTTCATGT; 1484–1504 U: CACCCCGGCCATCACC; 629–644 L: ACTTCCCGCTTCACCAAACTG; 1218–1238 U: GGCGTGGTGCATTGAAGAAT; 383–402 L: ATCCGTGGGCTCTGGCTTAT; 878–897 U: AAGCCTTTTCCTCAAGTGCG; 453–472 L: TTCCCCATTCAAGGTTAGAGT; 1228–1248

419 bp

53

24

515 bp

60

34

610 bp (
2) 493 bp (
3) 515 bp

60

34

60

35

762 bp

60

60

TR
1 c-erbA
2/
3 TR1 TR2c

a Sequences used for primer design and nucleotide numbering for annealing site designation: 18S rRNA (Chan et al., 1984); TR
1 cDNA (Thompson et al., 1987); c-erbA
2/
3 cDNA (Mitsuhashi, Tennyson and Nikodem, 1988); TR1 cDNA (Koenig et al., 1988); TR2 cDNA (Hodin et al., 1989). b Annealing temperature during PCR. c TR2 mRNA consists of a TR2-specific 5 end and a TR1/TR2-common 3 end. The sequence of the former was derived from nucleotides 1–494 of a 536 base TR2 cDNA (Hodin et al., 1989), the latter from nucleotides 529–4535 of a TR1 cDNA (Koenig et al., 1988). The upper primer annealing site is described relative to the TR2 sequence, the lower primer site relative to the TR1 sequence.

et al., 1999). Thus a mechanism for TH action appears to exist in rat placenta. To clarify whether TR comprise part of such a mechanism, the expression of c-erbA
and - transcripts and nuclear T3 binding activity were evaluated in normal rat placenta.

MATERIALS AND METHODS Sprague–Dawley rat dams were mated, then stunned and killed by cervical dislocation after 16, 19 or 21 days of gestation (dg). Placentae were dissected on ice at all ages, as were fetal liver and brain at 16 dg as positive control tissues. Tissues were weighed, pooled (from at least four fetuses per pregnancy) and stored at
20C for RNA isolation, or used immediately to assay nuclear T3 binding. A maternal brain was dissected and stored at
20C for RNA isolation, to provide a positive control for TR2 mRNA estimation. Post mortem maternal blood samples were taken for RNA isolation, as discussed below. The expression of c-erbA
and c-erbA transcripts was measured by semi-quantitative RT-PCR (Dieffenbach and Dveksler, 1995; Iskaros et al., 2000), with primer pairs specific for c-erbA
and c-erbA isoforms or for 18S rRNA as a control (Table 1). Primers were designed by target sequence analysis using the program Oligo 5 (National Biosciences Inc., Plymouth MN, USA) and synthesized by Genosys Biotechnologies (Cambridge, UK). The c-erbA
-common upper primer annealed upstream of the start site of c-erbA
1 and c-erbA
2 transcription (Chassande et al., 1997), therefore only full length c-erbA
transcripts were amplified. The c-erbA
2/3 primers annealed each side of the 117 bp region in c-erbA
2 transcripts, i.e. nucleotides 1075–1191 (Mitsuhashi, Tennyson and Nikodem, 1988), which is differentially spliced

from c-erbA
3 mRNA. Thus the c-erbA
2 and
3 amplicons differed in length. Total RNA was isolated from placenta and fetal tissues by acid–phenol–chloroform extraction and its integrity verified by gel electrophoresis (Sambrook, Fritsch and Maniatis, 1989). Intact RNA was treated with DNase I (Life Technologies, Paisley, UK) and 1 g aliquots reverse-transcribed to cDNA with Superscript II (Life Technologies, Paisley, UK) and random hexamer primers, using the manufacturer’s protocols. Superscript II was replaced with water in parallel reactions (RT
) to control for any residual DNA contamination during PCR. Serial cDNA dilutions were subjected to PCR in 20 l reactions containing 2 l cDNA, 1 U Amplitaq Gold (Applied Biosystems, Warrington, UK), 1 PCR buffer, 0.5 m dNTP, 3 m MgCl2 and primers at 0.3 . The thermal profile consisted of enzyme activation (94C/15 min) followed by a target-dependent (Table 1) number of cycles (94C/ 30 sec; annealing temperature/30 sec; 72C/1 min), then final elongation at 72C/5 min to ensure products were fully extended. Products were electrophoresed on 2 per cent agarose gels, stained with 0.5 g/ml ethidium bromide and subjected to UV-transillumination. Product band intensities were integrated from digitized gel images using the program NIH Image 6.1 (US National Institutes of Health; http:// rsb.info.nih.gov/nih-image/) with optical density step tablet calibration. Each cDNA sample was assayed in duplicate at four dilutions, the PCR conditions being optimized to give reproducible, highly specific amplification proportional to cDNA input. The PCR product intensity/l cDNA was calculated and expressed relative to 18S rRNA amplicon intensity/l cDNA. The 18S rRNA and TR
1 PCR products were analyzed by Southern blotting (Sambrook, Fritsch and Maniatis, 1989) to verify their identity, with oligonucleotides complementary to

Leonard et al.: T3 Receptors in Rat Placenta

(A) ×8

M

×4

×2

×1 18S; 419 bp TRα1; 515 bp TRβ1; 515 bp c-erbAα2; 610 bp c-erbAα3; 492 bp

(B) 4

15

3 2 1 0

(ii) TRβ1

* TRβ1/18S

TRα1/18S

(i) TRα1

5

*

P16 P19 P21 FL16 FB16

*

2 1

P16 P19 P21 FL16 FB16

c-erbAα3/18S

0.6 (iii) c-erbAα2

0

10

0

P16 P19 P21 FL16 FB16

3 c-erbAα2/18S

their expected sequence: TAAGTGCGGGTCATAAGCTT GCGTTGATTA i.e. nucleotides 1654–1683 of rat 18S rRNA (Chan et al., 1984); CCTGCCACGCCAGCCGCTTCC i.e. nucleotides 1460–1480 of rat TR
1 cDNA (Thompson et al., 1987). TR
1 and TR1 PCR products were also purified and investigated by restriction mapping with StuI, PstI, PvuII, BamHI and XbaI (all from Promega UK, Southampton, UK). The placental c-erbA
2 and c-erbA
3 PCR products were cloned using the TA cloning kit (Invitrogen BV, Groningen, Netherlands) and sequenced (Cambridge Biosciences; Cambridge, UK). TR transcript abundance was assessed in maternal blood from two dams at 16 dg and two at 21 dg. Total RNA was isolated from 1 ml blood samples using the SV total RNA isolation kit (Promega UK, Southampton, UK), then half of each RNA yield was reverse transcribed, in parallel with 0.025–1 g total RNA from a 16 dg placenta, as described above. The remaining blood RNA was used in the RT
control. Dilutions of cDNA were subjected to PCR for 18S rRNA (26 cycles) and for TR
1 (36 cycles), TR1 (37 cycles), c-erbA
2 and c-erbA
3 transcripts (37 cycles), the number of cycles being increased relative to the main study because less RNA was reverse transcribed; the same annealing temperature was used however. PCR products were quantified as described above. Whole nuclear T3 binding was estimated in tissues from another set of pregnancies as previously described (Gullo et al., 1987; Hubank et al., 1990). Tissues were homogenized and nuclei prepared by density centrifugation in 2  sucrose/ 10 m MgCl2. Nuclear pellets were suspended in binding assay buffer (BAB; 0.25  sucrose; 3 m MgCl2; 1 m dithiothreitol; 20 m Tris-HCl, pH 7.4), centrifuged at 1000 g/ 10 min, then re-suspended in BAB. DNA content was assayed using bis-benzimide (Labarca and Paigen, 1980). Binding assays were performed in quadruplicate in 0.2 ml BAB containing nuclei (60–200 g DNA), 0.2 n [125I]-T3 (>44 MBq/ g; BM Browne, Calcot, UK), and unlabelled T3 (SigmaAldrich, Gillingham, UK) to 0.2–20 n total T3. After 30 min at 37C, tubes were put on ice and 0.2 ml ice-cold BAB containing 2 per cent (v/v) Triton X-100 (Sigma-Aldrich, Gillingham, UK) was added. After 15 min, nuclei were pelleted (3000 g/10 min) and washed in 0.4 ml ice-cold BAB containing 1 per cent (v/v) Triton X-100. After re-centrifugation, bound T3 in the nuclear pellet was measured by -spectrometry. Non-specific binding was measured at 10  total T3 and was subtracted to adjust total binding to specific binding. Specific binding data were analyzed using Prism 2.0 software (Graphpad Software Inc, San Diego, USA), which fitted data by non-linear regression to a single binding site equation, providing estimates for the dissociation constant (Kd) and maximal binding capacity (Bmax) of the T3 binding. Statistical evaluation of data was by two-way ANOVA, using prior loge transformation where appropriate to satisfy homogeneity of variance requirements; Fisher PLSD test was used for post hoc analysis. Values are mean.

355

(iv) c-erbAα3

*

0.4 0.2 0.0

P16 P19 P21 FL16 FB16

Figure 1. RT-PCR analysis of rat placental c-erbA
and c-erbA isoform mRNA abundance. Serial dilutions of cDNA from placenta at 16, 19 and 21 dg (P16, P19, P21) and from fetal liver and fetal brain at 16 dg (FL16, FB16) were subjected to PCR using specific primers. Only products of the expected size were amplified using each primer set, as indicated on representative ethidium bromide stained gels (A). Product intensity was proportional to cDNA input; bars adjacent to the 100 bp marker lanes (M) indicate the more intensely stained 600 bp marker. Product intensities were expressed relative to 18S rRNA product intensity, to control for variation in reverse transcription efficiency, providing ontogenic profiles of TR
1, TR1, c-erbA
2 and c-erbA
3 mRNA abundance (B). *P<0.01, FB16 versus P16 and FL16; n=4 pregnancies.

RESULTS For RT-PCR of c-erbA
and c-erbA transcripts, primer pairs were designed to provide specific, linear amplification—as shown by the production of only amplicons of the expected size, at intensities proportional to the cDNA input across at least three serial dilutions [Figure 1(A)]. The number of PCR cycles differed for primer pairs (Table 1) because each target varied in abundance. The same number of cycles was used, however, for a primer pair for all tissue samples, to ensure valid comparison. The range of cDNA dilutions investigated was optimized for each tissue to ensure linear amplification. No PCR products were seen in RT- controls or in reactions containing water instead of cDNA. There was no change in the expression of TR
1, TR1, c-erbA
2 and c-erbA
3 transcripts in placenta between 16 and 21 dg, though differences were noted between placenta and fetal tissues at 16 dg [Figure 1(B)]. TR
1 mRNA was more abundant in fetal brain than in placenta and fetal liver, by 296 per cent and 183 per cent, respectively. In contrast, TR1

Placenta (2001), Vol. 22

356

×2

×1

×2

×1

TRβ1 ×2

(a)

×1

16 dg 21 dg P16 RNA 0.2 µg

×2

×1 ×2

×1

×2

×1

0.1 µg

(b) Bound T3 (fmol/ml)

MB RNA

TRα1

Bound T3 (fmol/ml)

18S

40 30 20 10 0

5

0.05 µg

mRNA abundance in placenta was not significantly different from that in fetal liver, but was 120 per cent greater than in fetal brain. Equivalent levels of c-erbA
2 and c-erbA
3 transcripts were expressed in placenta and fetal liver, but their abundance in fetal brain was, respectively, 764 per cent and 1180 per cent greater than in placenta. TR2 transcripts were not detected after 70 cycles of amplification in placenta at 16 and 21 dg, or in fetal liver and fetal brain at 16 dg, though the expected 762 bp product was detected in maternal brain after 60 cycles (data not shown). Placental TR2 expression was therefore negligible and not investigated further. The identities of the 18S rRNA and TR
1 PCR products from placenta, fetal brain and fetal liver were confirmed by Southern blotting using complementary oligonucleotide probes. High stringency binding to products of the expected size was observed in all cases. Furthermore, when purified placental TR
1 and TR1 PCR products were restriction mapped, the fragment sizes concurred with those predicted from analysis of the published sequences—StuI, PstI and PvuII cutting TR
1 but not TR1, and vice versa for BamHI and XbaI. Sequence analysis of the placental c-erbA
2 and
3 PCR products confirmed their identity, as they exhibited >99.8 per cent homology to the published c-erbA
2 and
3 sequences (Thompson et al., 1987; Mitsuhashi, Tennyson and Nikodem, 1988). PCR products may have been amplified from transcripts from blood within the placenta, therefore TR transcript abundance was assessed in total RNA isolated from 1 ml maternal blood samples. Too little blood RNA was isolated to quantify spectrophotometrically, instead, half of each yield was reverse transcribed in parallel with 0.025–1 g total RNA from 16 dg placenta. After PCR for 18S rRNA, product intensity was proportional to the amount of placental RNA reverse transcribed and maternal blood RNA exhibited equivalent product to 0.05–0.1 g of placental RNA (Figure 2). Maternal

(c)

20

Bmax

30 20 10

(d)

300

5

0

20

10 15 T3 (nM)

Kd 10

†

8 200

† nM

fmol T3 /mg DNA

Figure 2. RT-PCR analysis of c-erbA
and c-erbA isoform mRNA abundance in maternal blood. RNA from 1 ml maternal blood (MB) at 16 and 21 dg was reverse transcribed in parallel with 0.025–1 g RNA from 16 dg placenta (P16). cDNA was then subjected to PCR undiluted (1) and at two-fold dilution (2). 18S rRNA product was proportional to the amount of P16 RNA reverse transcribed (18S, lower panel). MB contained similar levels of 18S product to 0.05–0.1 g P16 RNA (18S, upper versus lower panel), suggesting it contained approx. 0.1–0.2 g RNA/ml. TR
1 mRNA abundance in MB was approx. fivefold that in 0.1 g P16 RNA (TR
1; upper versus lower panel) whereas TR1 transcripts were similarly expressed (TR1; upper versus lower panel). The abundance of c-erbA
2 and
3 transcripts was also similar (data not shown).

10 15 T3 (nM)

40

100

6 4 2

0

P16 P19 P21 FL16 FB16

0

*

*

P16 P19 P21 FL16 FB16

Figure 3. Nuclear T3 binding in placenta at 16, 19, and 21 dg (P16, P19, P21) and fetal liver and fetal brain at 16 dg (FL16, FB16). Representative specific T3 binding data are shown for P16 compared with P19 and P21 (a) and FL16 and FB16 (b). Statistical analysis of binding parameters indicated that the Bmax of nuclear T3 binding in P19 and P21 was more than double that P16 (c; †P<0.05, P19 and P21 versus P16; n=4 pregnancies), whereas the Kd showed no ontogenic change (d). T3 binding in FL16 and FB16 exhibited equivalent Bmax and Kd (c and d), the Kd being less than half of that in P16 (d; *P<0.005, FB16 and FL16 versus P16; n=4 pregnancies), though the Bmax was unchanged (c). : P16; 400 g DNA/ml; : P19; 400 g DNA/ml; : P21; 400 g DNA/ml; : FL16; 400 g DNA/ml; : FB16; 300 g DNA/ml.

blood thus contained approx. 0.1–0.2 g RNA/ml, assuming equal 18S rRNA abundance in placenta and blood. The RNA yield from placenta at 16, 19 and 21 dg was 3203628, 3219435 and 3210561 g RNA/g tissue, respectively. Maternal blood RNA therefore comprised <0.01 per cent of a placental RNA sample and to confound placental TR transcript estimation, TR mRNA abundance in maternal blood would have to be at least 10 000-fold greater than in placenta. Maternal blood TR
1 abundance was however only approx. fivefold greater than in 0.1 g placental RNA, which exhibited equivalent 18S rRNA product, and TR1 and c-erbA
2 and
3 transcripts were equivalently expressed (Figure 2). Nuclear T3 binding was investigated in placenta at 16, 19 and 21 dg, and in fetal liver and brain at 16 dg, as a measure of the relative abundance of TR
1 and TR1 protein (Figure 3). All binding data were consistent with a single binding site model (data not shown). T3 binding activity was detected in placenta at all ages, indicating that placental TR
1 and/or TR1 transcripts are translated. The Bmax of nuclear T3 binding in placenta increased from 69 to 150 fmol T3/mg DNA between 16 and 19 dg, then remained constant to 21 dg [Figure 3(C)]. The Kd of T3 binding in placenta showed no ontogenic change, it being 4.4 n at 16 dg [Figure 3(D)]. At 16 dg, the Kd in placenta was more than double that in fetal liver and brain [Figure 3(D)], whereas the Bmax was equivalent in all three tissues [Figure 3(C)].

Leonard et al.: T3 Receptors in Rat Placenta

DISCUSSION Rat placenta abundantly expressed TR
1, TR1, c-erbA
2 and c-erbA
3 mRNA between 16 and 21 dg, but no ontogenic changes were seen. At 16 dg, levels were similar to those in fetal liver, which is known to express higher levels of TR1 mRNA than TR
1 (Strait et al., 1990; Mellstrom et al., 1991). The patterns of c-erbA
and - transcript expression in fetal liver and brain were in agreement with previous observations (Strait et al., 1990; Mellstrom et al., 1991; Bradley, Towle and Young, 1992). TR2 mRNA was not detected in placenta or the fetal tissues but was apparent in adult brain. TR2 transcripts are detectable by in situ hybridization in discrete regions of fetal brain from 15 dg (Bradley, Towle and Young, 1992), but were probably not detected in the present study due to dilution in the whole fetal brain RNA preparations. It is possible that highly localized placental TR2 mRNA expression may also have gone undetected; in situ hybridization analysis will be required to investigate this possibility. Transcript abundance was conventionally expressed relative to 18S rRNA in this study, rather than per mg DNA which provides a measure of cellular abundance. When transcript abundances per mg DNA were approximated, from the mean abundances relative to 18S rRNA and the tissue RNA/DNA ratio, similar expression patterns were apparent. Rat placenta at 21 dg contains approx. 275 l of blood/g tissue (unpublished data) and blood RNA contamination may have perturbed estimation of placental c-erbA
and - mRNA abundance. Indeed, relative to 18S rRNA, TR
1 mRNA was fivefold more abundant in maternal blood than in placenta, and the other c-erbA
and - transcripts were at equivalent abundance. This would not however confound estimation of placental transcript abundance, since the RNA concentration in maternal blood was considerably less than in placenta. Fetal blood is also present in placenta, but again is unlikely to exhibit the high TR expression required to be problematic. These results were not unexpected since the RNA concentration in blood is known to be low. Furthermore, leukocytes, which are the only TR expressing blood cell type, contain, in human, 300 T3 binding sites/nucleus (Samuels et al., 1980)—similar to the content in rat placenta (a Bmax of 100 fmol T3/mg DNA corresponds to approx. 400 T3 binding sites/nucleus)—but comprise only a minor blood cell fraction. TR
1 and TR1 transcripts are translated, as nuclear T3 binding activity was detected in rat placenta, as well as in fetal liver and brain. The Bmax estimates in fetal liver and brain were comparable with previous values (Perez-Castillo et al., 1985). In placenta, the Bmax increased more than twofold between 16 and 21 dg with no change in TR
1 and TR1 transcript abundance, suggesting that T3 binding activity is posttranscriptionally regulated. Indeed, a lack of correspondence between TR transcript and protein levels is not uncommon and in some tissues has been ascribed to variable translation rates (Hodin, Lazar and Chin, 1990; Strait et al., 1990; Schwartz et al., 1992). This may pertain to the placenta although other mechanisms are possible, influencing either TR

357

abundance (e.g. changes in protein degradation or sequestration) or T3 binding activity (Eberhardt et al., 1979; Samuels et al., 1980; De Nayer and Dozin, 1985). The Kd of nuclear T3 binding in rat fetal liver and brain was 2.0 and 1.6 n respectively, while in placenta it was 4.4 n. The elevated Kd in placental nuclei compared with the control tissues suggests a tissue-specific difference in TR binding affinity. These Kd estimates are higher than those observed for in vitro translated TR (Weinberger et al., 1986; Murray et al., 1988; Hodin et al., 1989) and are also, in general, slightly higher than for detergent-washed nuclei and salt-extracted receptors from fetal and postnatal rat brain and liver (PerezCastillo et al., 1985; De Nayer and Dozin, 1985). The decreased affinity in the present study may occur because chromatin-associated factors, which decrease T3 binding affinity in postnatal rat brain and liver preparations (De Nayer and Dozin, 1985), are active in the whole nuclei. Indeed, estimates in the present study are similar to those in whole nuclei from postnatal brain (Hubank et al., 1990). The TR
1/TR1 protein ratios for fetal liver and brain are approx. 0.25 and 10, respectively (Schwartz et al., 1992; Falcone et al., 1994), which broadly reflects transcript abundance in these tissues (Strait et al., 1990; Mellstrom et al., 1991; Bradley, Towle and Young, 1992). If the same is true in placenta, the TR1 isoform predominates, as in fetal liver. Indeed, c-erbA
and - transcript expression was generally similar in placenta and fetal liver, suggesting that TH may act within them in a similar manner. These possibilities are tentative however, since the abundance of different mRNAs should not be compared if estimated by RT-PCR, unless quantitative standards are used. Furthermore, TR transcript levels do not correspond with protein levels in many tissues (Hodin, Lazar and Chin, 1990; Strait et al., 1990; Schwartz et al., 1992). An immunological approach will be needed to assess the relative levels of c-erbA
and - encoded proteins in placenta. In a preliminary report (Bandopadhyay et al., 1996), northern hybridization indicated that rat placenta contained low levels of TR
1 and c-erbA
2 mRNA relative to fetal brain. Both tissues contained few TR1 transcripts and placental nuclei exhibited ‘limited’ T3 binding. The results of the present study are in some agreement—e.g. TR
1 and, to a greater extent, c-erbA
2 mRNA were more abundant in fetal brain than placenta. TR1 transcripts were however more abundant in placenta than in fetal brain in the present study, perhaps because RT-PCR is more sensitive than northern hybridization, and better able to quantify TR1 transcripts. Placental nuclear T3 binding was easily detected in this study, because more nuclei and 125I-T3 were present in the assay, allowing better determination of the difference (i.e. specific T3 binding) between total T3 binding and the relatively high non-specific binding of placental nuclei. The expression of c-erbA
and - transcripts in rat placenta is qualitatively similar to that in human placenta (Kilby et al., 1998), although transcript abundance exhibits an ontogenic increase in the latter. This may be related to the period of

358

gestation investigated—first, second and third trimester in human but only late gestation in rat. Species differences in placental development and structure are probably also influential and, in general, comparisons between rat and human pregnancy should be cautiously made. Indeed, placental TR2 transcripts were not detected in this study but are seen in human (Kilby et al., 1998). The Bmax of T3 binding in 21 dg rat placental nuclei was similar to that in human trophoblast nuclei at term (Banovac, Ryan and O’Sullivan, 1986; Ashitaka et al., 1988). Kd values in these human studies are lower than in rat placenta, however neither investigated whole nuclei. When measured immunohistochemically, TR protein expression in human placenta increases with gestation (Kilby et al., 1998), as it did in rat placenta in the present study. When estimated by T3 binding however, human placental TR abundance decreases with gestation (Nishii et al., 1989). This contradiction is perhaps due to methodological differences since in the latter study, T3 binding was assessed in saltextracted nuclear preparations from enzyme-digested tissue (Nishii et al., 1989). T4 and T3 accumulate in rat placenta as gestation progresses and type II 5 -deiodinase, expressed in the maternal side, may convert T4 to T3 for local action within placental cells (Calvo et al., 1992). In hypothyroid rat dams, fetal tissue TH levels normalize after the onset of fetal thyroid function at 17.5 dg, but placental TH levels remain depressed up to 20 dg (Morreale de Escobar et al., 1985). The presence of c-erbA
and - transcripts and T3 binding activity in rat placenta from 16 dg onwards, provides a mechanism via which the prolonged changes in placental TH levels in hypothyroid dams may disturb placental development and function, producing deficits in fetal development. Indeed, recent studies in the rat indicate that maternal hypothyroidism perturbs placental c-fos and c-jun expression (Leonard et al., 1999), glycogen metabolism (Pickard et al., 2000) and glucose transporter expression (Pickard et al., 1999). TH also influences placental function in humans (Matsuo et al., 1993; Maruo et al., 1995). In pregnancies complicated by pre-eclampsia and intrauterine growth retardation (IUGR), conditions for which maternal hypothyroidism is a risk factor (Millar et al., 1994), placental dysfunction occurs and c-fos and c-jun expression (Faxen et al., 1997) and glycogen metabolism (Arkwright et al., 1993) are also disturbed. Indeed, fetal free T4 and T3 levels are depressed and placental TR expression elevated in IUGR pregnancies (Kilby et al., 1998). Comparisons between rat and human pregnancy must be tentatively made, however, it is possible that maternal hypothyroidism in rat may act, via the placental TRs demonstrated in this study, to disturb placental function in a manner similar to that seen in human pregnancies complicated by IUGR and pre-eclampsia.

REFERENCES Ashitaka Y, Maruo M, Takeuchi Y, Nakayama H & Mochizucki M (1988) 3,5,3 Triiodothyronine binding sites in nuclei of human trophoblastic cells. Endocrinol Jpn, 35, 197–206.

Placenta (2001), Vol. 22 Arkwright PD, Rademacher TW, Dwek RA & Redman CW (1993) Pre-eclampsia is associated with an increase in trophoblast glycogen content and glycogen synthase activity, similar to that found in hydatidiform moles. J Clin Invest, 91, 2744–2753. Bandopadhyay R, Leonard AJ, Sinha AK, Pickard MR, Evans IM & Ekins RP (1996) 3,5,3 Triiodothyronine receptor expression in rat placenta. J Endocrinol Invest, 19 (Suppl.), 53. Banovac K, Ryan EA & O’Sullivan MJ (1986) Tri-iodothyronine nuclear binding sites in human placenta and decidua. Placenta, 7, 543–549. Bernal J & Pekonen F (1984) Ontogenesis of the nuclear 353 triiodothyronine receptor in the human fetal brain. Endocrinology, 114, 677–679. Bonet B & Herrera E (1988) Different response to maternal hypothyroidism during the first and second half of gestation in the rat. Endocrinology, 122, 450–455. Bradley DJ, Towle H & Young W (1992) Spatial and temporal expression of
and -thyroid hormone receptor mRNAs, including the 2 subtype, in the developing mammalian nervous system. J Neurosci, 12, 2288–2302. Brent GA, Moore D & Larsen P (1991) Thyroid hormone regulation of gene expression. Ann Rev Physiol, 53, 17–35. Calvo R, Obregon J, Ruiz de Ona C, Escobar del Rey F & Morreale de Escobar G (1990) Congenital hypothyroidism, as studied in rats. J Clin Invest, 86, 889–899. Calvo R, Obregon J, Escobar del Rey F & Morreale de Escobar G (1992) The rat placenta and transfer of thyroid hormones from the mother to the fetus. Effects of maternal thyroid status. Endocrinology, 131, 357–365. Chan Y-L, Gutell R, Noller HF & Wool I (1984) The nucleotide sequence of a rat 18S ribosomal ribonucleic acid gene and a proposal for the secondary structure of an 18S ribosomal ribonucleic acid. J Biol Chem, 259, 224–230. Chassande O, Fraichard A, Gauthier K, Flamant F, Legrand C, Savatier P, Laudet V & Samarut J (1997) Identification of transcripts initiated from an internal promoter in the c-erbA
locus that encode inhibitors of RAR
and T3R activities. Mol Endocrinol, 11, 1278–1290. Contempre B, Jauniaux E, Calvo R, Jurkovic D, Campbell S & de Escobar GM (1993) Detection of thyroid hormones in human embryonic cavities during the first trimester of pregnancy. J Clin Endocrinol Metab, 77, 1719–1722. De Nayer P & Dozin B (1985) Thyroid hormones and brain development: modulation of the binding activity of the T3 nuclear receptor by chromatinassociated factors. Mol Physiol, 7, 303–310. Dieffenbach CW & Dveksler GS (1995) PCR Primer—a Laboratory Manual. Plainview, New York: CSHL Press. Eberhardt NL, Ring JC, Johnson LK, Latham KR, Apriletti JW, Kitsis RN & Baxter JD (1979) Regulation of activity of chromatin receptors for thyroid hormone: possible involvement of histone-like proteins. Proc Natl Acad Sci USA, 76, 5005–5009. Evans IM, Sinha AK, Pickard MR, Edwards PR, Leonard AJ & Ekins RP (1999) Maternal hypothyroxinemia disrupts neurotransmitter metabolic enzymes in developing brain. J Endocrinol, 161, 273–279. Falcone M, Miyamoto T, Fierro-Renoy F, Macchia E & DeGroot LJ (1994) Evaluation of the ontogeny of thyroid hormone receptor isotypes in rat brain and liver using an immunohistochemical technique. Eur J Endocrinol, 130, 97–106. Faxen M, Nasiell J, Lunell NO & Blanck A (1997) Differences in mRNA expression of endothelin-1, c-fos and c-jun in placentas from normal pregnancies and pregnancies complicated with preeclampsia and/or intrauterine growth retardation. Gynecol Obstet Invest, 44, 93–96. Gullo D, Sinha AK, Woods R, Pervin K & Ekins RP (1987) Triiodothyronine binding in adult rat brain: Compartmentation of receptor populations in purified neuronal and glial nuclei. Endocrinology, 120, 325–331. Haddow J, Palomaki G, Allan W, Williams J, Knight G, Gagnon J, O’Heir C, Mitchell M, Hermos R, Waisbren S, Faix J & Klein R (1999) Maternal thyroid deficiency during pregnancy and subsequent neuropsychological development of the child. N Engl J Med, 341, 549–555. Hendrich CE & Porterfield SP (1992) Serum growth hormone levels in hypothyroid and GH-treated thyroidectomized rats and their progenies. Proc Soc Exp Biol Med, 201, 296–302. Hodin RA, Lazar MA & Chin W (1990) Differential and tissue specific regulation of the multiple rat c-erbA messenger RNA species by thyroid hormone. J Clin Invest, 85, 101–105. Hodin RA, Lazar MA, Wintman BI, Darling DS, Koenig RJ, Larsen PR, Moore DD & Chin W (1989) Identification of a thyroid hormone receptor that is pituitary specific. Science, 244, 76–79. GenBank Accession M25071.

Leonard et al.: T3 Receptors in Rat Placenta Hubank M, Sinha AK, Gullo D & Ekins RP (1990) Nuclear triiodothyronine (T3) binding in neonatal rat brain suggests a direct glial requirement for T3 during development. J Endocrinol, 26, 409–415. Iskaros J, Pickard M, Evans I, Sinha A, Hardiman P & Ekins R (2000) Thyroid hormone receptor gene expression in first trimester human fetal brain. J Clin Endocrinol Metab, 85, 2620–2623. Kilby MD, Verhaeg J, Gittoes N, Somerset DA, Clark PMS & Franklyn JA (1998) Circulating thyroid hormone concentrations and placental thyroid hormone receptor expression in normal human pregnancy and pregnancy complicated by intrauterine growth restriction (IUGR). J Clin Endocrinol Metab, 83, 2964–2971. Koenig RJ, Warne RL, Brent GA, Harney JW, Larsen PR & Moore DD (1988) Isolation of a cDNA clone encoding a biologically active thyroid hormone receptor. Proc Natl Acad Sci USA, 85, 5031–5035. GenBank Accession J03819. Koenig RJ, Lazar MA, Hodin RA, Brent GA, Larsen PR, Chin W & Moore DD (1989) Inhibition of thyroid hormone action by a non-hormone binding c-erbA protein generated by differential mRNA splicing. Nature, 337, 659–661. Labarca C & Paigen K (1980) A simple, rapid and sensitive DNA assay procedure. Anal Biochem, 102, 344–352. Lazar MA (1993) Thyroid hormone receptors: multiple forms, multiple possibilities. Endocr Rev, 14, 184–193. Leonard AJ, Pickard MR, Sinha IK, Edwards PR, Evans IM & Ekins RP (1999) Maternal thyroid dysfunction and c-fos and c-jun expression in rat placenta. Placenta, 20, 727–731. Man EB, Brown JF & Serunian SA (1991) Maternal Hypothyroxinemia: Psychoneurological deficits in progeny. Ann Clin Lab Sci, 21, 227–239. Maruo T, Matsuo H, Otani T & Mochizuki M (1995) Role of epidermal growth factor and its receptor in the development of the human placenta. Reprod Fert Dev, 7, 1465–1470. Matsuo H, Maruo T, Murata K & Mochizuki M (1993) Human early placental trophoblasts produce an epidermal growth factor-like substance in synergy with thyroid hormone. Acta Endocrinol, 128, 225–229. Mellstrom B, Naranjo JR, Santos A, Gonzalez M & Bernal J (1991) Independent expression of the
and  c-erbA genes in developing rat brain. Mol Endocrinol, 5, 1339–1350. Millar LK, Wing DA, Leung AS, Koonings PP, Montoro MN & Mestman JH (1994) Low birth weight and preeclampsia in pregnancies complicated by hyperthyroidism. Obstet Gynecol, 84, 946–949. Mitsuhashi T, Tennyson G & Nikodem V (1988) Nucleotide sequence of novel cDNAs generated by alternative splicing of a rat thyroid hormone receptor gene transcript. Nucl Acid Res, 16, 5697–5703. GenBank Accession X07409. Morreale de Escobar G, Pastor R, Obregon MJ & Escobar del Rey F (1985) Effects of maternal hypothyroidism on the weight and thyroid hormone content of rat embryonic tissues, before and after onset of fetal thyroid function. Endocrinology, 117, 1890–1900. Murray MB, Zilz ND, McCreary NL, MacDonald MJ & Towle HC (1988) Isolation and characterization of rat cDNA clones for two distinct thyroid hormone receptors. J Biol Chem, 263, 12770–12777. Nishii H, Ashitaka Y, Maruo M & Mochizuki M (1989) Studies on the nuclear 3,5,3 triiodothyronine binding sites in cytotrophoblast. Endocrinol Jpn, 36, 891–898. Obregon MJ, Mallol J, Pastor R, Morreale de Escobar G & Escobar del Rey F (1984) L-thyroxine and 3,5,3 -triiodo-L-thyronine in rat embryos before onset of fetal thyroid function. Endocrinology, 114, 305–307.

359 Perez-Castillo A, Bernal J, Ferreiro B & Pans T (1985) The early ontogenesis of thyroid hormone receptor in the rat fetus. Endocrinology, 117, 2457–2461. Pickard MR, Sinha AK, Ogilvie L & Ekins RP (1993) The influence of the maternal thyroid hormone environment during pregnancy on the ontogenesis of brain and placental ornithine decarboxylase activity in the rat. J Endocrinol, 139, 205–212. Pickard MR, Evans IM, Bandopadhyay R, Leonard AJ, Sinha AK & Ekins RP (1997) Thyroid hormone action in rat brain from fetal to adult life. In Recent Research Developments in Neuroendocrinology—Thyroid Hormone and Brain Maturation (Ed.) Hendrich CE, pp. 15–30. Trivandrum, India: Research Signpost. Pickard MR, Sinha AK, Ogilvie LM, Leonard AJ, Edwards PR & Ekins RP (1999) Maternal hypothyroxinemia influences glucose transporter expression in fetal brain and placenta. J Endocrinol, 163, 385–394. Pickard MR, Sinha AK, Ogilvie LM, Leonard AJ & Ekins RP (2000) Maternal thyroid status regulates placenta and liver glycogen level. J Endocrinol, 164 (Suppl), OC2. Pop V, Kuijpens J, van Baar A, Verkerk G, van Son M, de Vijlder J, Vulsma T, Wiersinga W, Drexhage H & Vader H (1999) Low maternal free thyroxine concentrations during early pregnancy are associated with impaired psychomotor development in infancy. Clin Endocrinol, 50, 149–155. Porterfield SP & Hendrich CE (1993) The role of thyroid hormones in prenatal and neonatal neurological development—current perspectives. Endocr Rev, 14, 94–106. Sambrook J, Fritsch EF & Maniatis T (1989) Molecular Cloning. Plainview, New York: CSHL Press. Samuels H, Stanley F, Casanova J & Shao T (1980) Thyroid hormone nuclear receptor levels are influenced by the acetylation of chromatinassociated proteins. J Biol Chem, 255, 2499–2508. Schwartz HL, Strait K, Ling N & Oppenheimer J (1992) Quantitation of rat tissue thyroid hormone binding receptor isoforms by immunoprecipitation of nuclear triiodothyronine binding capacity. J Biol Chem, 267, 11794–11799. Sinha AK, Prabakaran D, Godbole MM, Chattopadhyay N, Karmarkar M, Pickard MR, Leonard AJ & Ekins RP (1997) Thyroid hormones and the human fetal brain. In Recent Research Developments in Neuroendocrinology—Thyroid Hormone and Brain Maturation (Ed.) Hendrich CE, pp. 1–14. Trivandrum, India: Research Signpost. Strait K, Schwartz H, Perez-Castillo A & Oppenheimer J (1990) Relationship of c-erbA mRNA content to tissue triiodothyronine nuclear binding capacity and function in developing and adult rats. J Biol Chem, 265, 10514–10521. Thompson CC, Weinberger C, Lebo R & Evans RM (1987) Identification of a novel thyroid hormone receptor expressed in the mammalian central nervous system. Science, 237, 1610–1614. GenBank Accession M18028. Vulsma T, Gons M & Vijlder J (1989) Materno-fetal transfer of thyroxine in congenital hypothyroidism due to a total organification defect or thyroid agenesis. New Engl J Med, 321, 13–16. Weinberger C, Thompson CC, Ong ES, Lebo R, Gruol DJ & Evans RM (1986) The c-erb-A gene encodes a thyroid hormone receptor. Nature, 324, 641–647. Woods RJ, Sinha AK & Ekins RP (1984) Uptake and metabolism of thyroid hormones by the rat fetus in early gestation. Clin Sci, 67, 359–363.