Relationship of c-erbA mRNA content to tissue triiodothyronine nuclear binding capacity and function in developing and adult rats.

We have quantitated in adult and developing rat tissues the molar concentrations of c-erbA alpha 1- and beta 1-mRNAs, which code for nuclear T3-binding proteins, and c-erbA alpha 2-mRNA, which is generated by alternate splicing of the alpha gene transcript and codes for a receptor variant that does not bind T3. Comparison of the concentrations of c-erbA alpha 1-mRNA, beta 1-mRNA, or their sum to the T3 nuclear binding capacity per mg of DNA in adult liver, kidney, heart, cerebrum, and cerebellum and during the ontogeny of liver and brain shows that the T3 binding capacity/c-erbA mRNA ratio is tissue-specific and related to developmental state. Administration of T3 resulted in a 40-50% fall in the alpha 1 signal of adult liver, kidney, and heart without changing either the beta 1 signal or T3 binding capacity. A 40-fold increase in rat brain beta 1-mRNA occurred in the transition between the 19-day gestational fetus and the 10-day-old neonate. This corresponds to the period during which the T3 content rises in brain and during which T3 is known to influence central nervous system development. Our findings indicate that important translational or post-translational factors influence nuclear binding capacity and raise the possibility that c-erbA beta 1 may play a primary role in mediating T3 effects in developing and adult animals.

We have quantitated in adult and developing rat tissues the molar concentrations of c-erbA a,-and B1-mRNAs, which code for nuclear Ta-binding proteins, and c-erbA a2-mRNA, which is generated by alternate splicing of the a gene transcript and codes for a receptor variant that does not bind TS. Comparison of the concentrations of c-erbA aI-mRNA, fil-mRNA, or their sum to the TI nuclear binding capacity per mg of DNA in adult liver, kidney, heart, cerebrum, and cerebellum and during the ontogeny of liver and brain shows that the T3 binding capacity/c-erbA mRNA ratio is tissuespecific and related to developmental state. Administration of TS resulted in a 40-50% fall in the a, signal of adult liver, kidney, and heart without changing either the & signal or T3 binding capacity. A 40-fold increase in rat brain PI-mRNA occurred in the transition between the 19-day gestational fetus and the loday-old neonate. This corresponds to the period during which the Ts content rises in brain and during which T3 is known to influence central nervous system development.
Our findings indicate that important translational or post-translational factors influence nuclear binding capacity and raise the possibility that c-erhA ,f31 may play a primary role in mediating T3 effects in developing and adult animals.
Recent studies have shown an unexpected diversity in the genes and mRNAs coding for putative triiodothyronine (T# nuclear receptors (l-5). Two members of the c-erbA gene family, LY and /3, are located on human chromosomes 17 and 3, respectively (6). Alternate splicing of the 01 gene transcript results in the formation of three distinct mRNAs, one coding for a 5.0-kb mRNA (LYJ and two similarly sized 2.6-kb mRNAs (Q-I and ~~-11) (2). Whereas CY~ codes for a nuclear protein which binds T3 with high affinity and the expected analogue specificity, both a*-mRNAs code for proteins which do not bind TB and are thus designated as "receptor variants." Alternate splicing of the p gene is responsible for the generation of two 6.0-kb mRNAs (& and p2), both of which bind T3 (7). '  The &mRNA is localized to the anterior pituitary of the rat (7), whereas the ,&-mRNA is widely distributed in several tissues (8). A high degree of homology exists between all the c-erbA proteins in the purported DNA binding region. In addition, the three c-erbAs that code for T, binding proteins ((Ye, &, and fiZ) show a high degree of homology in the TB binding region, particularly the last 40 amino acids which are completely absent from the nonbinding CQ variants. The functional significance of the diverse proteins coded by the c-erbA genes has received considerable attention.
In transient transfection experiments with cDNAs for the TX binding receptors, both a, and pi c-erbA products show a capacity for facilitating the expression of specific target genes either in a T3-dependent (9-11) or Ts-independent (12) manner. Recently, transient transfection of the cDNA coding for the nonbinding a2 variants has been shown to block the effect of a co-transfected (Ye-or &-cDNA in facilitating TO regulation of a third co-transfected reporter gene (13). This blockade has been ascribed to an ability of the CYZ protein to compete with 0~~ and PI for the thyroid hormone response element situated in the 5'-flanking region of the reporter gene. These findings clearly demonstrate the functional potential of the c-erbA products in transient transfection experiments. However, the role of these receptor forms in the intact animal remains less certain. A logical first step in deducing such function is to determine the relationship between mRNA levels and the Ts nuclear binding capacity. Previous studies in this area had relied on the use of Northern blot analysis to estimate the levels of specific mRNAs. Whereas Northern blots allow comparison of the relative concentrations of the same mRNA in different tissues, this technique does not permit a quantitative comparison of signals generated in the same tissue by different probes hybridizing with distinctive mRNAs. Variation in the degrees of labeling of such probes and in the efficiency of mRNA transfer from agarose pose major obstacles.
Accordingly, we applied techniques which would allow us to measure in normal euthyroid rat tissues the mass of the specific c-erbA mRNA/mg of DNA and to compare the values with each other and the nuclear T3 binding capacities. We further attempted to assess these relationships under circumstances associated with changing levels of receptors. Since Ta lowers both the binding capacity (14) and c-erbA mRNA (15) content of cultured GH cells, we considered the possibility that the Ts receptor content and c-erbA mRNAs in animal tissues might also be sensitive to hormonal state. Lastly, we examined these parameters in the neonatal development of liver and brain, processes which have been documented to be associated with changes in nuclear binding capacity (16).
The results of our studies highlight the importance of translational or post-translational determinants of tissue receptor levels. They show a striking rise in the brain &-c-erbA mRNA Levels in Neonatal and Adult Rat Tissues mRNA between fetal day 19 and neonatal day 10, a period during which TB is known to exert important effects on central nervous system development.
Our results suggest that the p-mRNA plays a predominant role in mediating thyroid effects in adult rat tissue and in the tissues of the developing neonate. cross-linked to the paper using a Stratalinker (120 pJ; Stratagene).
The blots were prehybridized at 60 "C in 0.1% SDS and 0.1 2 SSC for 60 min, foilowed by prehybridization in 50% formamide (deionized with AG50 l-X8 mixed bed resin. Bio-Rad). 50 mM NaH2P0,, pH 6.5,0.8 mM NaCl, 1 mM EDTA, 2.5 k Denhardt's, 250 rg/ml salmon testis DNA, and 500 pg/ml yeast tRNA for 4 h. Hybridizations were done overnight at 42 "C with 2 X lo6 cpm/ml of nick-translated "P-labeled c-erbA cDNA probes (2.3-kb EcoRI p1 probe (8) or a common PstI region of the No-and N~-cDNA (4)) in the prehybridization solution with the addition of 10% dextran sul-fate. Blots were washed twice at room temperature in 2 X SSC, 0.1% SDS, then once in the same solution at 60 'C followed by two more washes in 0.2 x SSC, 0.1% SDS at 60 "C. All washes were for 45 min each. The washed blots were autoradiographed on Kodak XAR-5 film and quantitated by computer-assisted video densitometry (22) in whole nuclei as previously described (27).

RESULTS
Tissue Distribution of c-erbA mRNAs in Euthyroid rat: Correlation with T3 Binding Capacity-A representative Northern blot of 20 pg of total brain cerebrum RNA illustrates the characteristic sizes of the c-erbA mRNAs hybridized with the probes used in these studies (Fig. 1). The first lane was hybridized to a common (Y probe that anneals to both the 2.6kb oc2-mRNAs ((~~-1 and ~~~-11; collectively refered to as (Ye in the present studies) and the 5.0-kb oil-mRNA. After washing, the blot was rehybridized with the p1 probe to yield the characteristic 6.0-kb c-erbA fil signal. The &mRNA has previously been shown to be pituitary-specific and thus not present in brain (7).
Northern blot analyses have already demonstrated a heterogeneous distribution of c-erbA mRNAs in various rat tissues (8). The p1 mRNA signal has been shown to be abundant in liver, whereas the levels of the a2 signal are high in brain. We have overcome the quantitative limitations of Northern blot analysis by placing in each gel a reference sample of cerebrum RNA, the c-erbA mRNA content of which had been predetermined by means of cRNA solution hybridization analysis. Measurement of the RNA/DNA ratios of individual tissues allowed us to express the mass of each mRNA species per mg of DNA. Six rat tissues were analyzed: liver, heart, kidney, spleen, testis, and brain (Table I). The markedly lower T, nuclear binding capacity per mg of DNA previously observed in cerebellum as compared to cerebrum prompted us to divide the brain to allow analysis of these areas separately (28).
FIG. 1. Northern blot hybridization of rat brain RNA using c-erbA cDNA probes.
Total RNA was extracted from euthyroid adult rat brain, electrophoresed in an agarose formaldehyde gel, electrotransferred to Zeta-bind paper, and hybridized to a nicktranslated cDNA probe for c-erbA cul and (~2 ((u) or a cDNA probe for c-erbA /?& (/3), as described in detail under "Experimental Procedures." cu,-mRNA was virtually absent in spleen and was not detectable in testis. Cerebrum also had the highest level of the nonbinding (Ye variant mRNA (Column B), followed closely by the level in cerebellum. The levels of a*-mRNA in cerebrum and cerebellum are higher by almost an order of magnitude than any other c-erbA mRNA measured in this series. Testis, which is devoid of both mRNAs that code for Ta binding proteins ((pi and pi), does contain appreciable amounts of the (~2 variant. Cerebrum also contained the highest tissue content of Pi-mRNA (Column C), exceeding the liver content by some 3-fold. Whereas (pi-and a*-mRNA levels in cerebrum and cerebellum are comparable, the Pi-mRNA levels in cerebellum are 25fold lower than in the cerebrum. The level of the /3,-mRNA in spleen was also exceedingly low, 120-fold less than that in cerebrum, and comparable in magnitude to that of the aI-and az-mRNAs in spleen. Thus, the expression of the (Y and fi c-erbA genes differs markedly in various tissues and in different areas of the brain. The sum of the masses of the mRNAs coding for T3+binding proteins are listed in Column D. Cerebrum had the highest combined signal of all tissues. The fractional contribution of the Pi-mRNA to the sum of both binding forms is listed in Column E. The conventionally Ts-responsive tissues, liver, kidney, and heart show the highest proportion of /3-mRNA: 82, 85, and 73'S, respectively. In contrast, for spleen and cerebrum, the pi mRNAs represent only about 50% of the binding forms. In the cerebellum, the Pi-mRNA constitutes only 7% of the binding forms of mRNA. The relative tissue content of (Ye and a2 is also determined by tissue-specific factors (Column F). In liver, the LY~ constitutes 48% of the sum of q + a2 mRNA whereas the corresponding value in testis is 0.
The binding capacities listed in Column G are consonant with previously published results (28,29), with liver exhibiting the highest value and in declining order, cerebrum, heart, kidney, cerebellum, spleen, and testis. The last is devoid of measurable Ta binding capacity. As in previous studies, the Ts binding capacity/mg of DNA in cerebellum was 3-4-fold lower than that in cerebrum. The lack of any clear-cut correlation between TO binding capacity and the mass of c-erbA mRNA in a tissue is apparent from the ratio of binding capacity to the sum of the q and @, mRNA (Column H = Column G/Column D). There is a lo-fold variation in this ratio among the tissues studied. A similar variation also occurs if the ratio of the binding capacity to the mass of either /3i- The amounts of the individual c-erbA mRNAs in several rat tissues were determined by Northern blot analysis in conjunction with solution hybridization and normalized per mg of DNA as described under "Experimental Procedures." The amounts of c-erbA N, in the tissues is shown in Column A, 01~ levels in Column B, and fll levels in Column C, expressed in fmol/mg of DNA. Column D is the sum of the c-erbA mRNAs (Y, and /3r, which code T%-binding proteins. The fraction of the sum in Column D contributed by c-erbA 6, is shown in Column E. The fraction of c-erbA u gene expression resulting in nr-mRNA is shown in Column F. Ta binding capacity (B,,,, ng of TJmg of DNA) was determined in isolated whole nuclei as described under "Experimental Procedures" and is shown in Column G. Finally, the relationship between T, binding capacity and the levels of the c-erbAs which code for Ts binding proteins ((Ye + &) is shown in Column H. All measurements of c-erbA mRNA levels were from six animals per group (mean + SD. In an attempt to further define the physiologic roles of the various c-erbA mRNAs, we compared changes in the levels of the c-erbA mRNAs and T3 binding capacity under circumstances that previously have been reported to alter the binding capacity: 1) thyroidal status and 2) neonatal brain and liver development.
Effect of Thyroidal Status on the Levels of c-e&A mRNAs and T3 Binding Capacity-Previous studies designed to assess the effect of thyroidal state on the T3 nuclear receptor content of tissues and cell lines have yielded conflicting results. Samuels et al. (14) have shown that addition of TS to the medium of GH cells lowered nuclear Ts binding capacity to approximately one-half of the initial value. The fall in T3 binding sites in GH cells is accompanied by a reduction in al, a2, and ,& mRNA, while /I1 mRNA content remained constant (7,15). In contrast, earlier studies in our laboratory showed thyroidal state had no influence on the hepatic nuclear TB binding capacity (30). The following series of experiments were designed to determine whether thyroidal state influenced the expression of the (Y and p erbA genes or the T3 binding capacity in liver, kidney, heart, and brain.
The effects of thyroidal state on c-erbA mRNA levels are shown in Fig. 2. TB had no effect on the levels of c-erbh &-mRNA in any of the four tissues studied. Thyroid hormone reduced the a1 and 01~ signal in liver (A) by approximately 50% and 60%, respectively. The levels of c-erbA CQ-and a*-mRNAs in the kidney (B) and heart (C) also declined by approximately 50% in response to TB. In sharp contrast to the results in other tissues, Ts had no effect on either al-or a2-mRNAs in the brain (D). In none of these tissues was there a significant difference between the nuclear TB binding capac-ity between the hypothyroid and euthyroid states (data not shown). No efforts were made to assess the nuclear Ts binding capacity in hyperthyroid rats given the technical difficulty involved in making these measurements in the face of the large doses of TB required to establish the hyperthyroid state. The reduction in the al-mRNA by TB in liver, kidney, and heart without a corresponding decrease in nuclear binding capacity confirms our data in Table I  well defined changes in the T3 binding capacity of rat liver and brain during the first 15 days of neonatal life (16,32,43). These developmental changes in T3 binding presented us with the opportunity to correlate changes in T3 binding capacity with the levels of individual c-erbA mRNAs. Fig. 3A illustrates c-erbA mRNA levels in the livers of normal rats from fetal day 19 to the 2-month-old adult animal. The relatively low levels of both the cyl-and a*-mRNAs remained constant during this period. The p1 mRNA on the other hand fell by 60% from 19-day fetus to neonatal animals 4 days after birth. From day 4 to adult, the levels of the PI-mRNA remained constant. During this same developmental period, the TB nuclear binding capacity began to increase, culminating in adult levels &fold the 19-day fetal level (Fig.  3B). The development of the liver is therefore associated with a 5-fold increase in Ts binding without an increase in the levels of the c-erbA mRNA. These findings support the fact that no simple relationship exists between the c-erbA mRNAs and nuclear Ts binding capacity of a tissue. Fig. 4A  Nuclear T3 binding capacity was measured in isolated whole nuclei of liver by Scatchard analysis as described under "Experimental Procedures." Nuclei were isolated from portions of the same pools of liver (4-6 rats per pool) used for c-erbA mRNA studies. mRNAs in cerebrum during the same period. In the Is-day fetus, as in the adult, the predominant c-erbA mRNA is (Ye. The (Ye mRNA is at all times approximately IO-fold higher than the cq mRNA. Both LY c-erbA mRNAs rise coordinately to a plateau value on day 4, a 4-fold rise from fetal levels. After day 10, both a mRNAs fall to an adult value 1.5-fold higher than that of the 19-day fetus.
The developmental patterns of the /I1 mRNA contrast sharply with the pattern of the Q and (Ye mRNAs. The /I1 mRNA in the 19-day fetus is barely detectable at levels less than 10% that of the cyl. From these low fetal levels there is a sharp increase in the p1 mRNA with neonatal development. Ta binding capacity in the developing rat brain.
Nuclear TS binding capacity was measured in isolated whole nuclei of brain by Scatchard analysis as described under "Experimental Procedures." Nuclei were prepared from portions of the same pools of tissue used for the c-erbA mRNA studies.
in adult brain, some 3-4 times greater than the (Ye mRNA. Changes in the nuclear Ta binding capacity in cerebrum during this developmental period also demonstrates a dissociation between mRNA and binding capacity. The binding capacity rises 3-4-fold from the fetal value to reach a maximum on the 4th day after birth. Thereafter, the binding capacity falls to an adult value which is approximately 1.5fold that of the fetus, in accordance with previously published data (16,32). Although the developmental time course of the TB binding capacity in rat cerebrum resembles that of the 01i-mRNA, it differs sharply from the Pi-mRNA as well as the sum of the al-and P1-mRNA. Thus, unless one is prepared to assume that in the brain the P1-mRNA is not expressed as a functional nuclear TB-binding protein, these data like those from the developing liver do not support a simple proportional c-erbA mRNA Levels in Neonatal and Adult Rat Tissues The c-erbA P1-mRNA levels were measured as described under "Experimental Procedures" in euthyroid and hypothyroid rat brain at various times during development by Northern blot analysis and solution hybridization. TB levels in the cytosol were measured by immunoassay as described under "Experimental Procedures." Animals were rendered hypothyroid by addition of methimazole to the drinking water of the pregnant mothers from 12 days of gestation throughout the period of study. Cytosol T, levels in hypothyroid brains were all at or below the level of detectability, 15 pg of Ta/ml. For c-erbA PI determinations, n = 4 animals per group for neonatal days 10 and 15, and n = 4 pools of 4-6 brains for the 19-day fetus and neonatal day 1, 4, and 6 animals (mean + S.D.). For cytosol TB determinations, n = 4 animals per group (mean + SD. between the level of the c-erbA mRNAs and TB binding. The 40-fold rise in the Pi-mRNA during the first 10 days of neonatal life takes on special significance in view of the fact that most of the developmental changes in the central nervous system induced by TS take place in this interval (31). During this period of development, there is a well documented rise in the level of plasma TS (16). The present studies indicate that this is reflected by a proportional increase in T3 in the brain cytosol (Fig. 5). To ascertain whether the increase in T3 initiated the rise in the PI-mRNA, neonatal animals were rendered hypothyroid by administering methimazole to the mothers at 12 days of gestation and continued throughout the neonatal development.
Despite the fact that this procedure completely depleted plasma and cellular TB in the neonates, the rise in the Pi-mRNA was indistinguishable from that observed in untreated animals (Fig. 5). In a related experiment, administration of receptor-saturating doses of TS to newborn rat pups did not accelerate the rise in P1-mRNA levels in the brain above that seen in the hypothyroid or euthyroid animal (data not shown). Therefore, the rise in pl-mRNA apparently represents a developmental phenomenon causally unrelated to the T3 content of the brain.

DISCUSSION
The methods described in this manuscript have permitted for the first time a quantitative comparison of the levels of LYE-, LYE-, and Pi-mRNAs in multiple tissues in adult and developing rat. These measurements have facilitated a comparison of the content of individual c-erbA mRNAs with nuclear Ts binding capacity as determined by isotopic displacement analysis in the same tissues.
Quantitation of the molar content of c-erbA mRNA has also made possible an estimate of the average number of molecules per cell. Such calculations, based on Avogadro's number and the assumption of 8 pg of DNA per cell, indicate that in the average hepatic cell the number of c-erb mRNA molecules is 1.1, 1.2, and 4.8 for (pi, (Ye, and &, respectively. These mRNAs are thus clearly in the low abundancy class. In contrast, we have estimated the number of receptor molecules per hepatic cell to be 4800 (33). Since the tJh of the receptor has been estimated to be about 4 h (34-36), we calculate that each mRNA molecule generates one receptor every 20 s. This is in general accord with published estimates of protein synthesis rates in eukaryotic cells (one molecule every 20-60 s) (37).
One of the major conclusions of this study is that there is no simple relationship between the content of (pi-and pi-mRNA and the total Ts binding capacity. This is apparent from the lo-fold variation in the ratio of the binding capacity to the sum of (pi-and &-mRNAs in various tissues of the euthyroid rat and from the change in this ratio in developing rat liver and brain. Moreover, a selective 50-60% reduction of the al-mRNA by thyroid hormone in liver, kidney, and heart results in no detectable changes in binding capacity. These findings supersede conclusions reached in our earlier studies based on heterologous cDNA probes which suggested a direct relationship between binding capacity and expression of the p gene (20).
The dissociation of binding capacity from mRNA suggests the operation of translational or post-translational factors in determining the expression of the c-erbA mRNAs as protein. Such factors could include different degrees of sequestration of the mRNA in cytosol, variation in the efficiency of translation at the ribosomal level, post-translational modification of the translational product, and alterations in the stability of the receptor protein. Although unlikely, the possibility has not been entirely excluded that the observed discrepancies are due to as yet unidentified receptor mRNA species. Our studies emphasize that tissue and developmental factors determine both the expression of the c-erbA genes as mRNA and the various mRNAs as protein. The tissue-specific regulation of the cy gene transcript is particularly instructive in this regard. At one extreme lies the testis which contains substantial quantities of a2-mRNA but no measurable amounts of al-mRNA.
At the other extreme, the gene products in liver are equally represented by the LYE-and as-mRNA species. The changes in total LY gene expression observed in response to TS administration in heart, liver, and kidney as well as the developmental changes noted in brain and liver development are characterized by a constant ratio of cyl-to cut-mRNA. These findings indicate that, although TR and developmental factors influence a gene expression, these factors do not influence the selective processing of the LY gene transcript to the oyl and (Ye forms. Despite our inability to establish a direct relationship between mRNA content and binding capacity, our results may be helpful in suggesting functional roles for the mRNA species. The striking 40-fold rise in the level of the Pi-mRNA in the transition from the 19-day-old fetal to the lo-day-old neonatal brain is a case in point. It is well established that the absence of thyroid hormone in early development causes profound alterations in central nervous system development (38), including deficiency in myelination (39), and arborization of neuronal dendrites (40), as well as the proliferation and migration of cells (41).
In normal animals, these effects of thyroid hormone on rat brain development appear to occur largely in the first lo-15 days of life (31). During this period, there is a steady rise in the level of plasma T3 from essentially undetectable levels in c-e&A mRNA Levels in Neonatal and Adult Rat Tissues fetal serum. In the present studies, we have also demonstrated that there is a parallel rise in the Ts tissue content of Ts. In the euthyreotic neonatal rat, the effects of thyroid hormone deficiency can be reversed if Ts treatment is started prior to day 15. Delays in starting Ts replacement past 15 days produce increasingly abnormal brain development.
The clinical counterpart of this phenomenon is observed in human cretinism. Our studies revealed that accompanying the increase in serum and brain T3 during the first 10 days of neonatal life was an unanticipated 40-fold increase in the level of PI-mRNA. The rise in &-mRNA is not due to the rising level of TB, since hypothyroid animals with undetectable cytosolic brain TB showed an identical increase in &-mRNA during this period. The coordinate increase in TJ concentrations, pi-mRNA levels, and the well established functional and structural changes in the central nervous system during this time period suggest that the c-erbA fil receptor is involved in the transduction of TB effects on normal brain development. The target genes of TS during this period may well be responsible for initiating the structural changes which occur during the early neonatal period. These genes may function only during a restricted time frame. This may account for the failure of TB to reverse the structural damage of neonatal hypothyroidism in the rat if T3 is administered after neonatal day 15. The finding that PI-mRNA is the predominant mRNA in the conventionally thyroid-responsive tissues liver, kidney, and heart provides further suggestive evidence favoring a functional role of the & translational product. Furthermore, fetal rat brain, which is generally not believed to be responsive to TS, is almost completely devoid of PI-mRNA. Thus, the correlation of P,-mRNA with thyroid response (rather than binding capacity) in certain tissues supports the suggestion that the /3-mRNA product may play a central role in thyroid hormone action.
The potential role of the cuP-mRNA, which codes for a non-T3-binding protein, has also received attention in the literature. Most recently, Koenig et at. (13) have shown by transfection experiments that the a*-cDNA product can block the effects of co-transfected LYE and & in facilitating the regulation of reporter genes by T3. These observations have prompted speculation that the high levels of (Ye in brain account for the nonresponsive nature of this tissue to Ta. Our findings provide the first quantitative assessment of the relative content of c+ (pi, and p1 in cerebrum. The levels of ox in adult cerebrum are 7-fold higher than those of /Ii and g-fold higher than those of al-mRNA, thus lending credence to the possibility that the o(~ product could block the accessibility of target genes to Ts-binding receptors. Such a view would necessarily be based on the unproved assumption that in cerebrum the LY? protein is greatly in excess of that coding for the p1 or (Ye product.
The possibility that (Y* protein under some circumstances stimulates constitutive gene expression rather than opposing the effects of & or LY* deserves consideration.
Although the gene for malic enzyme is regulated by TS in many tissues it is not regulat,ed by TB in brain. The malic enzyme gene in brain is, nevertheless, highly expressed. Experiments in our laboratory have suggested that the level of malic enzyme mRNA in adult brain approaches that in the hyperthyroid liver (data not shown). It is possible that the level of expression of genes such as malic enzyme are rendered maximal by the presence of the cu2-mRNA product. The possibility should be seriously considered in light of previous findings by Samuels and coworkers (12) that transfection of (Y-and @-cDNAs can result in both TZ-dependent and Ta-independent stimulation of gene expression. Further, our findings in testis demonstrate that this tissue contains substantial quantities of cuP-mRNA but is totally lacking Q-and PI-mRNA and receptors demonstrable by displacement methods. Therefore, if (Ye has any function in this tissue it is not that of blocking the binding of T3 receptor to putative thyroid hormone response elements of TS target genes.
The possibility that the high concentrations of LYE in brain are responsible for preventing the regulation of many genes by TS, either by blocking access to the TB receptor complex or by maximal constitutive stimulation, presents a potential problem in understanding any direct effects of thyroid hormone effects on brain development.
However, the present studies have clearly demonstrated that (Y and /3 gene products are differentially distributed in the brain. Whereas the relative (Ye-and cun-mRNA content in cerebrum and cerebellum are roughly similar, the relative levels of & are much lower in the cerebellum.
Examples of similar segregation of (Y and /3 gene products has also recently been demonstrated by the in situ hybridization studies of Bradley et al. (42). Such segregation would make it possible for certain cell groups rich in @I and poor in (Ye to respond to Ta. A definitive resolution of many of the issues raised in this study will depend upon the development of specific antibodies for identifying the translational products of the LY~-, ti2-, and &-mRNA and the identification and quantitation of specific brain genes which transduce the developmental effects of thyroid hormone.