| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Developmental Genetics Programme, The Babraham Institute, Cambridge CB2 4AT, UK; and
* Department of Physiology, University of Cambridge, Cambridge CB2 3EG, UK
1Correspondence: The Babraham Institute, Bldg. 540, Babraham Hall, Cambridge CB2 4AT, UK. E-mail: joy.dauncey{at}bbsrc.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUDING REMARKSREFERENCES |
|---|
1, TRß 1, TRß 2 (TH
binding), and TR 2 (non-TH binding) in functionally distinct porcine
muscles during prenatal and postnatal development. Use of a novel and
highly sensitive RNase protection assay revealed striking
muscle-specific developmental profiles of all four TR isoform mRNAs in
cardiac, longissimus, soleus, rhomboideus, and diaphragm. Distribution
of TR isoforms varied markedly between muscles; TR expression was
considerably greater than TRß and there were significant differences
in the ratios TR 1:TR 2, and TRß 1:TRß 2. Together with
immunohistochemistry of myosin heavy chain isoforms and data on
myogenesis and maturation of the TH axis, these findings provide new
evidence that highlights central roles for 1) TR
isoforms in fetal myogenesis, 2) the ratio TR 1:TR
2 in determining cardiac and skeletal muscle phenotype and function;
3) TRß in maintaining a basal level of cellular
response to TH throughout development and a specific maturational
function around birth. These findings suggest that events disrupting
normal developmental profiles of TR isoforms may impair optimal
function of cardiac and skeletal muscles.—White, P., Burton, K. A., Fowden, A. L., Dauncey, M. J. Developmental expression
analysis of thyroid hormone receptor isoforms reveals new insights into
their essential functions in cardiac and skeletal muscles.
Key Words: cardiac and skeletal muscle • myosin • prenatal and postnatal development • TR
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUDING REMARKSREFERENCES |
|---|
2
3
4)
. Their major action is via nuclear TH receptors
(TR); members of a superfamily of hormone receptor transcription
factors that regulate expression of numerous target genes
(5
6
7)
. TR are encoded by two proto-oncogenes—c-erbA-
and c-erbA-ß—located on human chromosomes 17 and 3, respectively,
and alternative splicing leads to formation of functionally distinct
isoforms: TR 1, TR 2, TRß 1, and TRß 2. Whereas TR 1,
TRß 1, and TRß 2 can bind TH and trans-activate TH response
elements within the promoter region of target genes, the TR 2
variant cannot bind TH because of structural changes in the
carboxyl-terminal domain, and hence is not strictly a receptor for TH.
The precise functions of TR 2 are unknown, but it is thought to
compete with the TH binding isoforms for specific DNA response elements
and thus inhibit transcription.
TH have an especially important influence on muscle development and
function. Prenatally, they induce myoblasts to exit the cell cycle,
differentiate, and express a muscle-specific phenotype
(8)
. Coincident with the perinatal increase in circulating
TH levels (9
, 10)
, embryonic and neonatal myosin heavy
chain (MyHC) isoforms are progressively repressed and adult MyHC
isoforms are accumulated (11)
. Postnatally, TH continue to
influence muscle phenotype, changing transcription levels of myogenic
regulatory factors, altering metabolic properties, and inducing
switching from type I slow to type II fast MyHC, with the extent of
these changes being dependent on muscle type (12
, 13)
.
Whereas the developmental and tissue-specific functions of TH have been
relatively well documented, much less is known about the ontogeny and
function of their receptors. Ligand binding studies have been used to
investigate TR ontogeny (14
15
16)
, but these provide no
information about the specific TR isoforms and cannot detect the
presence of the non-TH binding variant, TR 2. Cloning of the TR
cDNAs has enabled assessment of TR isoform expression, but most studies
have been undertaken in altricial species such as the rat in which
maturation of the thyroid system occurs postnatally rather than
prenatally, as in precocial species such as the human and pig. Recent
gene inactivation studies of mice have suggested specific functions for
TR and TRß isoforms in muscle (17
18
19)
, but there
have been no studies of TR isoform expression in functionally distinct
muscles. The aim of the present study was therefore to undertake a
detailed analysis of TR isoform expression in diverse muscles during
prenatal and postnatal development, and to relate this to changes in
structure and function. A novel and highly sensitive method for
quantification of the four TR mRNAs revealed unique developmental
profiles of TR isoform expression. Together with MyHC
immunocytochemistry and previous data on myogenesis and maturation of
the TH axis, this study provides new evidence for the essential roles
of TR isoforms in myogenesis and acquisition of optimal cardiac and
skeletal muscle function.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUDING REMARKSREFERENCES |
|---|
, 20
, 21)
. Our previous findings demonstrated that both energy
intake and energy expenditure profoundly affect total TR numbers
(22)
, and food intake also influences expression levels of
the individual TR isoforms (23)
. Therefore, all tissues
used in this study were obtained from animals kept under carefully
controlled conditions of food intake and thermal environment. Animals
before term were delivered by Caesarian section under general
anesthesia (sodium pentobarbitone 20 mg/kg) and tissues were obtained
after administration of a lethal dose of anesthetic. Postnatally,
animals were sedated by an intramuscular injection of ketamine
hydrochloride (1.0 ml Vetalar, 100 mg/ml; Parke Davis Veterinary,
Pontypool, UK) and killed with a 0.7 ml/kg body weight intracardiac
injection of sodium pentobarbitone (Lethobarb, 200 g/l; Duphar
Veterinary, Southampton, Hampshire, UK). All procedures were approved
by the UK government under the Animals (Scientific Procedures) Act
1986.
Tissue-specific pattern of TR isoform expression
The pattern of TR isoform expression in a range of functionally
distinct muscles and liver was assessed in neonatal and postnatal
animals. The following tissues were obtained from 10 pigs, each taken
from a separate litter—5 at birth and 5 at 7 wk postnatally: cardiac,
soleus, diaphragm, rhomboideus, longissimus dorsi, and liver. Tissue
samples were obtained from neonates within 3 h of birth for tissue
sampling. Pigs were weaned at 3 wk and housed at thermal neutrality
(26°C) in separate pens to allow careful control of food intake. For
the next 4 wk, the animals were provided with a controlled amount of
food (UltraWeen, Dalgety, Bristol, UK) that enabled optimal growth
(24)
. The food contained 14 kJ gross energy/g wet weight
and comprised 32% carbohydrate, 22.5% protein, 5.5% fat, 3.5%
fiber, and 6% ash with added vitamins and minerals; the remaining 30%
was water. Water was freely available and lighting was on for a 12 h period, from 9:00 AM to 9:00 PM. Tissue
samples were obtained at 7 wk, 20 h after the last meal. Samples
were always taken from the same relative anatomical position of the
different animals investigated. For all animals, tissues were dissected
rapidly, divided into 1–5 g portions, frozen in liquid nitrogen, and
stored at -70°C. These tissues were used subsequently for RNA
extraction and TR isoform mRNA expression using RNase protection
analysis. In addition, samples of the four skeletal muscles were taken
from the 7 wk animals for morphological assessment. Samples of 1
cm3 were mounted on cork blocks, surrounded by
Cryo-m-bed (Bright Instrument Company Ltd., Cambridge, UK), frozen
immediately in isopentane cooled in liquid nitrogen, and stored at
-70°C.
Developmental profiles of TR isoform expression
A detailed assessment of the developmental profiles of TR and
TRß isoform expression was undertaken in cardiac and longissimus
dorsi muscles. Samples were obtained throughout development, with the
focus on fetal, neonatal, and postnatal life. For fetal analysis,
longissimus was obtained from between one and three piglets at each of
the following ages: 72, 81, 88, 94, 99, 104, and 111 days gestation,
i.e., at -42, -33, -26, -20, -15, -10, and -3 days before birth
(term=114 days); cardiac ventricular muscle was obtained at -42, -26,
-20, -15, and -3 days before birth. For neonatal and early postnatal
analysis, tissues were obtained from four litters each of four pigs
aged 0, 2, 5, or 14 days. All procedures were standardized to ensure
that differences in nutritional or thermal environment did not
influence the results. Within each litter, one animal was removed from
the sow at birth, placed in a thermally neutral environment (34°C),
and killed within 4–6 h. For the three other time periods, animals
were removed from the sow at 8:00 AM and kept at thermal
neutrality for 2 h before tissue sampling. Samples were also
obtained from five 7-wk-old pigs (described in detail in the previous
section) and one mature adult pig aged 2 years. Samples were always
taken from the same relative anatomical position of the different
animals investigated. For all animals, tissues were dissected rapidly,
divided into 1–5 g portions, frozen in liquid nitrogen, and stored at
-70°C. Tissues were used subsequently for RNA extraction and TR
isoform mRNA expression using RNase protection analysis.
Skeletal muscle morphology, myofiber types, and myosin
immunocytochemistry
Functionally distinct muscles were investigated: longissimus
dorsi, predominantly fast oxidative-glycolytic dorsal-lumbar, used in
rapid movement acts to stabilize and flex the vertebral column;
rhomboideus, mixed slow-fast oxidative-glycolytic dorsal interscapular
is important for postural maintenance and may play a key role in
thermoregulation; soleus, slow oxidative postural muscle of the hind
leg; diaphragm, mixed slow-fast oxidative-glycolytic; and cardiac, slow
oxidative (25
, 26)
. We had performed a rigorous assessment
of the relative merits of myosin immunocytochemistry and ATPase
histochemistry with preincubation at different pH values for
determining myofiber type (26)
. In young animals,
immunocytochemistry for type I slow/type II fast MyHC expression gives
clear-cut repeatable results. By contrast, although ATPase staining
at acidic and alkaline preincubation pHs can resolve most fiber types
in mature muscle, this is not always the case during prenatal and
postnatal development. In the present study, therefore, differences in
morphology between the four skeletal muscles were assessed using myosin
immunocytochemistry (26)
. Indirect immunoperoxidase
staining was used with myofiber type-specific monoclonal antibodies:
anti-slow (Clone No. MHCs; Biogenesis Ltd., Poole, Dorset, UK), which
reacts with type I slow MyHC; and anti-fast (Clone No. MY-32;
Sigma-Aldrich Company Ltd., Poole, Dorset, UK), which reacts with type
II fast MyHC. Determination of type II subtypes at the protein level
was not carried out because antibodies suitable for mature muscle
cannot always distinguish between fiber types postnatally (20
, 26)
. Frozen sections were fixed in 4% paraformaldehyde, washed
in Tris-buffered saline (TBS), blocked with normal horse serum and
incubated with the primary MyHC antiserum for 1 h at room
temperature. After washing in TBS-Tween, sections were incubated with a
biotinylated anti-mouse IgG secondary antibody (Vector Laboratories
Ltd., Peterborough, UK), washed, incubated with an
avidin-biotin-peroxidase complex (ABC reagent; Vector), rinsed, and
incubated in 0.03% H2O2 in
TBS + 1 mg/ml diaminobenzidine. The sections were finally dehydrated
and mounted in a xylene-based mounting medium. Relative proportions of
type I slow and type II fast fibers in each muscle sample were
determined in four random fields of view within a standard field of
119,000 µm2 using a Seescan A010 research-grade
image analysis system (Cambridge, UK).
Assessment of TR and TRß isoform expression
Extraction and measurement of total RNA
Total RNA was extracted from 0.5 g samples of tissue using
a method based on the guanidine thiocyanate method of Chomczynski and
Sacchi (27)
. The final RNA pellet was dissolved in 0.3 M
sodium acetate, pH 5.2, and quantified by duplicate absorbance readings
at 260 nm. A constant relation between A260 and
poly(A)+ content was found for RNA from all tissues. RNA samples were
stored in ethanol at -70°C.
Construction of TR and TRß riboprobes
Antisense riboprobes were designed according to the porcine
TR 1 and TRß 1 cDNA sequences (EMBL Nucleotide Sequence Database
accession numbers AJ005797 and AJ238614, respectively) using in vitro
transcription in the presence of [32P]UTP
(Amersham, UK) as described previously (28)
. The TR
riboprobe had a full-length transcript of 295 bp, of which 218 bp
hybridized fully to TR 1, corresponding to nucleotides 958 to 1175.
Since this probe spanned the carboxyl-terminal region where TR 1 and
TR 2 have differences in the ligand binding domain, a second
protection product of 153 bp was formed corresponding to nucleotides
958 to 1110 of TR 2. Similarly, the TRß riboprobe had a
full-length transcript of 293 bp, of which 230 bp hybridized fully to
TRß 1, corresponding to nucleotides 189 to 418. Since this probe
spanned the amino-terminal region where TRß 1 and TRß 2 sequences
diverge, a second protection product of 154 bp was formed corresponding
to nucleotides 265 to 418 of TRß 2. Thus, these single probes allowed
accurate measurement of the expression levels of both TR isoforms or
both TRß isoforms within a single assay.
RNase protection analysis
RNase protection analysis was used for developmental and
muscle-specific expression studies of TR isoform abundance using
methods described earlier (28
29
30)
. In brief, samples of
total RNA (20 µg) were hybridized with a small molar excess of the
radiolabeled TR isoform riboprobe to ensure linearity of the assay with
respect to RNA. After 16 h hybridization at 45°C, excess
nonprotected RNA was digested with RNase A (50 µg/ml, 
1 U/sample;
Ambion, Austin, TX) and RNase T1 (300 U/ml, 80 U/sample; Ambion).
The protected hybridization products were purified by extraction in
phenol:chloroform:isoamyl alcohol (25:24:1) and ethanol precipitation.
After separation by denaturing polyacrylamide gel electrophoresis,
followed by autoradiography, expression levels of the four TR isoforms
were quantified by image analysis (Seescan). The system was linear over
the range of OD values measured. Expression levels of TR 1, TR 2,
TRß 1, and TRß 2 are presented as relative abundance of each mRNA.
TR isoform expression at the mRNA and protein levels
The present investigation focuses on TR isoform expression at
the mRNA level. In general, it is thought that changes in mRNA level of
TR isoforms are closely matched by changes in number of the specific TR
isoforms. However, most current investigations focus on TR isoform
mRNAs because these can be detected accurately and, using the RNase
protection assay described here, to a very high level of sensitivity.
By contrast, there are major difficulties in reliably measuring TR
isoforms at the protein level. Our experience and that of others in the
field show there are limitations in the TR isoform antibodies currently
available for such investigations and problems associated with optimal
isolation of TR protein from the nucleus. Hence, reliable data on the
relation between TR mRNA and TR protein are limited. Although analysis
of overall TR numbers is possible with ligand binding assays, these
fail to give any information on isoform composition and do not detect
the non-TH binding isoform, TR 2.
Statistical analysis
Significance of differences were tested using the Student’s
paired t test. Probabilities were considered significant at
the 5%, 1%, and 0.1% levels. The association between TR isoforms
and myofiber type was investigated by calculating the product-moment
correlation coefficient in order to derive the corresponding
P value. All results are expressed as mean values ±
SE.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUDING REMARKSREFERENCES |
|---|
illustrates differences in morphology between the four skeletal muscles
investigated as assessed using myosin immunocytochemistry and
demonstrates the predominance of myofibers containing type II fast MyHC
compared with type I slow MyHC in longissimus. This contrasts with the
fiber type proportions in the three other skeletal muscles and with
cardiac muscle, which contains 100% slow-twitch fibers. Proportions of
type I slow and type II fast fibers for longissimus, rhomboideus,
soleus, and diaphragm are presented in Table 1
.
|
|
Tissue-specific expression of TR isoforms
A detailed assessment of TR and TRß isoform expression was
made in a range of functionally distinct muscles and liver in newborn
and postnatal animals. This revealed major differences with respect to
the relative abundance of all four TR isoforms both between and within
tissues in the neonate and at 7 wk postnatally.
TR isoforms
Autoradiographs from RNase protection analyses and mean
values ± SE obtained by image analysis are
presented for TR isoforms in Fig. 2
and Fig. 3
, respectively. The tissue-specific expression patterns of
TR 1 and TR 2 at birth were similar to those at 7 wk. Expression
of TR 1 tended to be constant between muscles, especially
postnatally, whereas TR 2 expression showed a much more variable
expression pattern between muscles. Regression analysis revealed a
striking positive relation between myofiber type and TR 2
(Fig. 4
). Thus, at 7 wk, the correlation between proportion of fibers
expressing type I slow MyHC and TR 2 expression was significant
(correlation coefficient=0.97; P<0.002). By contrast, there
was no significant relation between type I slow MyHC fiber
proportion and TR 1 expression (correlation coefficient=0.05;
P>0.9).
|
|
|
The major difference between birth and 7 wk with respect to both TR
isoforms was in the diaphragm: TR expression was low at birth, but
by 7 wk had increased to levels similar to those in the other muscles
investigated. Strikingly, TR 2 expression was extremely high in the
heart. At birth, TR 2 was 3-fold greater in the heart than in the
soleus, rhomboideus, and longissimus, 7-fold greater than in diaphragm,
and 40-fold greater than in the liver. At 7 wk, TR 2 expression in
the heart was between 2.5-fold and 3-fold greater than in the soleus,
diaphragm, and rhomboideus, 5-fold greater than in longissimus, and
10-fold greater than in the liver. At 7 wk in longissimus, by contrast,
TR 2 expression was 46% lower than in the other skeletal muscles
and 80% lower than in the heart. In the neonate and at 7 wk,
expression of both TR isoforms was lowest in the liver.
Particularly marked were differences in the relative levels of TR
isoforms within a tissue. In the heart, the relative level of TR 2
was more than twofold greater than that of TR 1 at birth
(P<0.001) and 7 wk (P<0.01). At both ages,
expression of TR 2 in the soleus, diaphragm, and rhomboideus was
slightly lower than TR 1; only in the diaphragm was the difference
statistically significant. In longissimus, however, TR 1 was
expressed at a markedly greater level than TR 2, and the difference
increased with age: TR 1 mRNA was 30% greater than TR 2 at birth
(P<0.05) and 105% greater at 7 wk (P<0.001).
In the liver, expression of TR 1 was threefold greater than TR
2 at birth (P<0.01) and twofold greater at 7 wk
(P<0.001).
TRß isoforms
Autoradiographs from the RNase protection analysis and mean
values ± SE obtained by image analysis are
presented for TRß in Fig. 5
and Fig. 6
, respectively. Although TRß isoforms were expressed at a
considerably lower level than TR, a tissue-specific pattern of
distribution was again seen at birth and at 7 wk. However, by contrast
with TR, there was a striking difference in the tissue-specific
pattern of TRß expression in newborns compared with postnatal
animals. At birth, the highest levels of TRß 1 mRNA occurred in
rhomboideus, values were intermediate in the soleus, diaphragm, and
liver and lowest in the heart and longissimus. Postnatally, this
pattern of expression changed: levels of TRß 1 mRNA were greatest in
the heart, intermediate in the soleus, diaphragm, rhomboideus, and
liver, and lowest in longissimus.
|
|
Expression of TRß 2 was extremely low in all tissues examined and, at birth, it could not be quantified because levels were below the detection limit of the assay. TRß 2 then increased between birth and 7 wk, at which age its tissue-specific expression pattern was similar to that of TRß 1. However, for all tissues examined TRß 2 was significantly lower than TRß 1.
Developmental profiles of TR and TRß isoform expression
A detailed assessment of the ontogenic profiles of TR and TRß
isoform expression was performed in the two muscles that showed the
most striking differences in relative abundance of TR isoforms: cardiac
and longissimus. This analysis revealed major changes in expression of
TR and TRß throughout the time period analyzed. Developmental
profiles of these nuclear receptors were markedly different in the two
tissues. Moreover, the developmental expression pattern within each
tissue of the two TR isoforms was quite different from that of the
two TRß isoforms.
Cardiac
The developmental profiles of TR isoforms in cardiac muscle are
presented in Fig. 7
. At every age investigated, levels of the non-TH binding TR 2
isoform were two- to fourfold greater than those of TR 1. For both
TR isoforms, expression was greatest in the youngest animal
investigated at 72 days gestation (i.e., 42 days before birth). As
development progressed, expression fell gradually to reach low levels
at 7 wk postnatally and declined even further in adulthood.
|
By contrast with TR, cardiac TRß was at its lowest level at 72
days gestation and increased gradually as development progressed,
peaking sharply at 111 days gestation (i.e., 3 days before birth) and
falling at birth to levels similar to those seen earlier in
development. After birth, levels increased gradually and at 7 wk were
similar to those in the adult. Expression patterns of the two TRß
isoforms were similar, although TRß 1 was consistently expressed at a
level two- to fourfold greater than TRß 2. However, levels of both
TRß isoforms were considerably lower than TR 1 and TR 2.
Longissimus
Developmental profiles of TR and TRß isoforms in longissimus
muscle are shown in Fig. 8
. Expression patterns of TR 1 and TR 2 followed similar patterns
and were strikingly different from those in the heart. Levels increased
between 72 and 88 days gestation (i.e., 42 and 26 days before birth,
respectively), then declined slightly during the period leading up to
birth. By the second day postnatally, both TR isoforms fell sharply
to a level 60% below that at birth. Subsequent changes were small:
expression increased slightly by 14 days, then declined over the next
few weeks to reach similar levels at 7 wk and at 2 years. Whereas there
was a preponderance of TR 2 in the heart throughout development,
levels of TR 1 and TR 2 in longissimus were similar before birth,
and TR 1 was greater than TR 2 postnatally. The difference was
significant by 14 days (P<0.01), and this preponderance of
TR 1 over TR 2 persisted throughout the rest of development.
|
In contrast to the results for TR isoforms, TRß 1 had an
expression pattern similar to that in the heart, with lowest levels at
72 and 81 days gestation. It then increased gradually as development
progressed to reach its highest levels in the adult. In addition, there
was a distinct peak in TRß 1 at birth, which fell by 2 days
postnatally. Expression of TRß 2 was extremely low in longissimus and
below the detection limit of the assay. As in the heart, levels of
TRß in longissimus were considerably lower than those of TR.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUDING REMARKSREFERENCES |
|---|
and TRß isoform expression in functionally
distinct muscles during fetal, neonatal, and postnatal development.
Relative abundance and developmental expression patterns of TR were
markedly different from those for TRß and there were significant
differences in the ratios TR 1:TR 2 and TRß 1:TRß 2.
Moreover, TR 2 isoforms in different muscles were closely related to
MyHC abundance and myofiber type. The significance of these findings
will be discussed in the context of our previous observations on
developmental changes in circulating TH levels, the postulated
functions of TR isoforms, and the specific roles of different muscles
in infancy. These findings are relevant not only to a fundamental
understanding of the functions of TR isoforms, but also to the optimal
development of cardiac and skeletal muscles in the human infant.
Developmental regulation of TR isoform expression
Detailed assessment of the developmental changes in TR isoforms
revealed striking tissue-specific differences in the patterns of TR
and TRß isoform expression throughout fetal and postnatal life. TR
expression was much greater than that of TRß at all stages of
development. Moreover, TR mRNAs were already relatively abundant at
the earliest age investigated (72 days gestation) whereas TRß mRNAs
were always expressed at relatively low levels.
At 72 days gestation, expression of cardiac TR isoforms was
extremely high; throughout development TR 2 expression was between
two- and fourfold greater than TR 1. As development progressed,
cardiac TR isoform expression decreased whereas TRß gradually
increased, showing a small but pronounced peak around birth. In
longissimus, the two TR isoforms were expressed at similar levels
and had similar ontogenic profiles. Expression of both TR isoforms
increased between 72 days gestation and 90 days gestation, then
declined slightly toward birth. There was a precipitous fall in
longissimus TR mRNAs between 0 and 2 days postnatally. This profile
is in accord with our previous findings on nuclear TH binding in
porcine longissimus muscle: nuclear TR numbers are already high at 80
days gestation, increase slightly at birth, and decrease postnatally
(16)
. Profiles for longissimus TRß mRNA were similar to
those observed in the heart in that there was a gradual increase in
expression with age and a small but pronounced peak in the period
around birth.
Changes in ontogeny of the thyroid axis may explain in part these
developmental changes in TR isoform expression. In precocial species
such as the human, pig, and sheep, the thyroid system matures before
birth (9)
. The porcine thyroid gland begins to function at
46–47 days gestation, but fetal plasma TH levels remain extremely low
until 80 days gestation. As in humans, levels of TH increase between
mid-gestation and birth, followed by a dramatic postnatal surge between
birth and 6 h (10)
. After this surge, plasma TH
decrease gradually with age; in the human, they plateau during
adolescence (4)
. TH have been implicated in negative
feedback regulation of their own receptors, although results in vitro
sometimes differ from those in vivo (22)
. In cultured
cardiomyocytes, TR 1, TRß 1, and TRß 2 mRNAs are all
down-regulated by TH (31)
, whereas administration of TH to
adult rats reduces both TR isoforms but does not affect TRß
(32)
. The overall decrease in TR isoform expression
levels toward birth may therefore result from the increase in fetal TH
levels, and the immediate postnatal surge in TH levels could explain
the dramatic fall in longissimus TR isoforms between birth and 2
days. Clearly, however, not all changes in TR isoforms can be explained
by developmental changes in TH levels. Glucocorticoids may be important
in this respect. They are potent regulators of many hormone systems,
and whether the prenatal cortisol rise that occurs in precocial species
(3)
is responsible for the perinatal induction of TRß
isoforms should be investigated. In addition, transient changes in
nuclear associations and structure play a key role in developmental
regulation of transcription (7
, 33)
. In view of the
different chromosomal locations of the TR and TRß genes, the
possibility is that specific changes in chromatin remodeling may be
involved in differential regulation of TR isoforms during development.
Functional significance of muscle-specific and developmental
differences in TR isoform expression
The distribution of TR isoforms varied markedly between muscles,
and TR isoforms were expressed at a considerably higher level than
TRß isoforms. Particularly striking was the finding that cardiac
muscle expressed an extremely high level of the non-TH binding 2
variant compared with the TH binding 1. However, there was no such
dominance of TR 2 in skeletal muscles, and in some instances TR 1
was the predominant isoform. Muscle phenotype and function are probably
regulated not simply by differences in total TR numbers, but also by
differences in relative proportions of TR isoforms (28)
.
The present results support this hypothesis and indicate that relative
expression levels of TR 1 and TR 2 within a muscle may be
critical in determining its phenotype. TR 2 probably inhibits
transcription of target genes by competing with the TH binding TR
isoforms for specific DNA response elements. High TR 2 levels would
thus protect the tissue from TH action, whereas muscles where TR 1
expression predominates will be the most sensitive to TH. Many
muscle-specific genes are regulated by TH, and our results suggest that
the pattern of TR isoform expression will directly affect the
functional characteristics of muscles by modulating the action of TH in
switching type I slow to type II fast MyHC (34)
. In
cardiac muscle, slow sustained contractility will be enhanced by high
TR 2 and hence an increase in slow MyHC, whereas in longissimus, the
ability for rapid movement will by emphasized by its relatively high
levels of TR 1 and increase in fast MyHC. Similar levels of TR 1
and TR 2 in the soleus, diaphragm, and rhomboideus further accord
with this hypothesis because these muscles contain similar proportions
of fast and slow MyHC.
The developmental changes in TR isoform expression in skeletal
muscle appear to be closely associated with changes that take place
concurrently in muscle development. As in most mammals, myogenesis in
the pig is biphasic. A primary generation of myofibers forms between 35
and 55 days gestation, followed by the formation of a secondary
generation between 55 and 95 days gestation (20)
. Apart
from postnatal recruitment of tertiary fibers, total myofiber numbers
are largely established at this age (12)
. The present
study shows that the peak in expression of TR 1 and TR 2 in
longissimus occurs at the time that secondary myofibers are formed. In
view of the role of TH in stimulating fast MyHC expression, the
increase in sensitivity to TH conferred by a high TR 1 level is
consistent with the fact that secondary fibers express only fast MyHC.
When total fiber numbers have been established, sensitivity to TH will
fall with the reduction in TR 1 expression during late gestation.
In the diaphragm, the low level of TR 1 expression at birth may have
particular functional significance. One of the proposed functions of
the immediate postnatal rise in TH is in regulatory thermogenesis
(10)
. The fall in ambient temperature at birth probably
induces this increase in TH via the hypothalamic-pituitary-thyroid
axis; in lambs, this increase can be prevented by delivering the
newborn into an environment at 37–39°C (9)
. Low levels
of TR 1 in diaphragm would protect it from high TH levels, whereas
its greater expression in the other skeletal muscles would enable an
increase metabolic rate.
In all muscles, expression of TRß isoforms was considerably lower
than TR, and at birth levels of TRß 2 mRNA were below the
detection limit of the assay. Nevertheless, at 7 wk the high
sensitivity of the assay allowed detection and quantification of TRß
2 in all tissues investigated. TRß 2 was originally thought to be
localized to the pituitary gland (35)
. It has since been
detected in other tissues, albeit at much lower concentrations than
TRß 1; the average cellular content of TRß 2 mRNA in rat pituitary
is 0.6 molecules, but elsewhere it is less than 0.007 molecules per
cell (36)
. The marked changes in TRß isoforms around
birth may be important in relation to specific maturational events and
the otherwise general increase in TRß with age may be important in
maintaining a basal level of cellular response to TH.
|
CONCLUDING REMARKS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUDING REMARKSREFERENCES |
|---|
1 and TR 2 isoforms in regulating muscle phenotype is suggested. The
role of the TRß isoforms is less clear because they are expressed at
extremely low levels, but in addition to a specific perinatal function,
they may maintain a basal level of responsiveness to TH. Studies of
mice deficient in one or several TR isoforms support these hypotheses.
Inactivation of the TR gene leading to abrogation of both TR
isoforms results in death within 5 wk postnatally (37)
.
Inactivation of the TRß gene has a much less dramatic effect on
development, and mice exhibit a phenotype similar to that in TH
resistance in humans (38
, 39)
. Recent findings in mouse
skeletal muscles indicate that abrogation of TR 1 or TR 1 and
TRß genes together markedly up-regulates type I slow MyHC and
down-regulates type IIA fast MyHC in the soleus, whereas the lack of
TRß does not alter MyHC composition (19)
. Moreover,
there are concomitant changes in soleus contraction and relaxation
times, and the phenotype is similar to that associated with
hypothyroidism (40)
.
The present study also highlights a potentially critical role for TR
expression in the development and optimal function of the heart. This
accords with a recent analysis of cardiac function in TR 1- and
TRß-deficient mice that suggests that TR 1 has a major role in
determining basal heart rate and body temperature and that TRß
mediates a hormone-induced increase in heart rate (18)
.
There is, however, the possibility that recent gene inactivation
studies have overlooked the potentially critical role of TR 2 in
muscle development; there are currently no available data that focus
specifically on this non-TH binding isoform. Finally, our studies
indicate that events disrupting normal developmental profiles of muscle
TR isoforms may impair numerous contractile, metabolic, respiratory,
postural, and thermogenic functions. In view of the major effect of
energy status on the TH axis and nuclear TH binding (22
, 41
, 42)
, the extent to which TR and TRß isoforms are modulated
by both prenatal and postnatal undernutrition needs to be
investigated.
|
ACKNOWLEDGMENTS |
|---|
Received for publication November 15, 2000.
Revision received February 5, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUDING REMARKSREFERENCES |
|---|
and TRß in the control of thyroid hormone production and post-natal development. EMBO J 18,623-631[Medline]
1 and ß. Am. J. Physiol. 276,H2006-H2012 1, 2, and ß 1 T3-receptor mRNAs in the fasted rat measured using competitive PCR. Biochem. Biophys. Res. Commun. 242,492-496[Medline]
gene encoding a thyroid hormone receptor is essential for post-natal development and thyroid hormone production. EMBO J 16,4412-4420[Medline]
1- or -ß-deficient mice. Am. J. Physiol. 278,R598-R603This article has been cited by other articles:
|
|
 |
|
  C. M Villicev, F. R S Freitas, M. S Aoki, C. Taffarel, T. S Scanlan, A. S Moriscot, M. O Ribeiro, A. C Bianco, and C. H A Gouveia Thyroid hormone receptor {beta}-specific agonist GC-1 increases energy expenditure and prevents fat-mass accumulation in rats J. Endocrinol., April 1, 2007; 193(1): 21 - 29. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Pantos, I. Mourouzis, C. Xinaris, A. D Kokkinos, K. Markakis, A. Dimopoulos, M. Panagiotou, T. Saranteas, G. Kostopanagiotou, and D. V Cokkinos Time-dependent changes in the expression of thyroid hormone receptor {alpha}1 in the myocardium after acute myocardial infarction: possible implications in cardiac remodelling Eur. J. Endocrinol., April 1, 2007; 156(4): 415 - 424. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Bee, M. Calderini, C. Biolley, G. Guex, W. Herzog, and M. D. Lindemann Changes in the histochemical properties and meat quality traits of porcine muscles during the growing-finishing period as affected by feed restriction, slaughter age, or slaughter weight J Anim Sci, April 1, 2007; 85(4): 1030 - 1045. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Irrcher, P. J. Adhihetty, T. Sheehan, A.-M. Joseph, and D. A. Hood PPARgamma coactivator-1alpha expression during thyroid hormone- and contractile activity-induced mitochondrial adaptations Am J Physiol Cell Physiol, June 1, 2003; 284(6): C1669 - C1677. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||
