| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Endocrinology (E.A.N.), Department of Medicine, Brown University School of Medicine, Rhode Island Hospital, Providence, Rhode Island 02903; and Yale University School of Medicine and Connecticut Veterans Affairs Healthcare System (K.A.S.), Department of Psychiatry, Research 151, West Haven, Connecticut 06516
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
TRH, produced in the paraventricular nucleus of the hypothalamus (PVN),2 stimulates the biosynthesis and secretion of TSH from the anterior pituitary (8 9 ). TSH, in turn, stimulates thyroid hormone biosynthesis and release (10 ). TRH is central in regulating the hypothalamic-pituitary-thyroid (HPT) axis. TRH influences the release of other hormones, including PRL, GH, vasopressin, and insulin (11 12 13 ), and the classic neurotransmitters noradrenaline and adrenaline (14 ). Further, TRH is present in many brain loci outside of the hypothalamus, supporting a potential role as a neuromodulator or neurotransmitter outside of traditional HPT axis function (15 16 ). For example, TRH is implicated as a modulator of seizure activity (17 ) and gastrointestinal function (18 ). TRH has been also found outside the central nervous system (CNS) in the gastrointestinal tract, pancreas, reproductive tissues including placenta, ovary, testis, seminal vesicles, and prostate, retina, and blood elements (19 ). The widespread distribution of TRH within and outside the CNS supports a diverse range of roles for this molecule, roles likely to involve many functions outside of the traditional HPT axis.
Of the many peptide products derived from the TRH precursor (pro-TRH
described below), until recently only TRH itself was studied
extensively. In the last few years, a new wave of research has
identified many other products derived from pro-TRH and suggests
potential biological functions for these non-TRH peptides. The
immediate precursor to TRH, TRH-Gly, independent of conversion to TRH,
stimulates gastric acid secretion in a dose-dependent manner, although
it is 100-fold less potent than TRH (18 ). Other peptides derived from
pro-TRH, originally referred to as cryptic peptides because their roles
are incompletely understood, are now being studied. In the
hypothalamus, prepro-TRH160–169 (pST10, also
known as Ps4) and prepro-TRH178–199 (pFE22),
peptides that lie between the third and fourth and the fourth and fifth
progenitor sequences for TRH in the TRH precursor, respectively, are
released from perifused rat hypothalamic slices and the median eminence
(20 ). [The reader should note that in Fig. 1
, pro-TRH-derived peptides are named by
"p" for peptide followed by the single letter amino acid
designation for the first and last amino acid of the peptide, along
with the peptide length in subscript. Where these peptides are first
mentioned, they are followed by the longer prepro-TRH name that
describes their amino acid residue positions within the precursor.]
prepro-TRH160–169 potentiates TSH release from anterior
pituitary and stimulates TSHß gene promoter activity (21 ). This
peptide also potentiates TRH-induced gastric acid secretion when
microinjected into the dorsal motor nucleus of the vagus (22 ). Thus,
prepro-TRH160–169 acts in concert with TRH both within and
outside traditional HPT roles. prepro-TRH178–199 is
proposed to be a CRH-inhibiting factor (Refs. 23 24 but see Ref.
25 ).
|
Substantial progress has been made in the last few years in
understanding the biosynthesis and processing of prohormone and
neuropeptide precursors. For pro-TRH (Fig. 1
), we have developed a
model of its processing to mature peptides (29 30 31 ). Like many other
secreted peptides, processing of the primary translation product,
prepro-TRH, begins with removal of the signal peptide during its
passage into the lumen of the rough endoplasmic reticulum (RER). From
our current knowledge, processing of pro-TRH takes place within the
regulated secretory pathway (RSP) (for full definition see
Section IIA). Two recently discovered serine proteases,
which are members of the family of prohormone convertases (PCs), PC1
(SPC3) and PC2 (SPC2), related to subtilisin and the yeast-processing
enzyme Kex 2 (32 33 34 ), are the primary PCs involved in
posttranslational processing of pro-TRH (31 35 36 37 ).
Recent contributions to the understanding of pro-TRH biosynthesis and processing provide a useful framework for uncovering the diversity of function displayed by pro-TRH-derived peptides and the way that this diversity is generated. We will review the current knowledge of pro-TRH processing to the most studied biological end product, TRH, and importantly, other pro-TRH-derived peptides. The importance of understanding pro-TRH processing is then underscored by a comprehensive review of the wide range of potential biological roles subsumed by non-TRH products derived from pro-TRH. Although data are limited, where possible we summarize where these peptides are produced and how their levels might be regulated. We then review the HPT axis, with emphasis of its effects on pro-TRH processing. This is followed by a comprehensive review of the function of extrahypophysiotropic TRH. A review of TRH and other pro-TRH-derived peptide receptors then explains mechanisms of pro-TRH peptide signal transduction. Finally, TRH degradation provides an additional way by which peptide levels are tightly controlled, and this area is reviewed. The degradation of other pro-TRH-derived peptides is not yet understood.
|
II. Biosynthesis of TRH and Other pro-TRH-Derived Peptides |
|---|
|
TopAbstractI. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
Even though the amino acid structures of vasopressin and insulin had been elucidated in the 1950s (41 ), it was not until the early 1960s that the mechanisms of protein biosynthesis began to be understood and the genetic code was fully defined. In 1964, Sachs and Takabatake (42 ) provided the first evidence that the biosynthesis of vasopressin can be inhibited with puromycin, a protein synthesis inhibitor, and that newly synthesized vasopressin could not be detected in tissues until more than 1 h after pulse labeling. In these studies, Sachs and Takabatake demonstrated that before vasopressin becomes a biologically active peptide, it exists in a modified, or proform, state. Posttranslational modification was then required to convert the proform into an active peptide.
While the structure of insulin peptides was described early on (41 ), it
was difficult to envision how the combination of A and B chains was
attained in ß-cells of the pancreas. Experiments initiated by Steiner
in 1965 using tritiated leucine and phenylalanine to label
proinsulin and insulin represented a landmark in prohormone theory.
Using a pancreatic insulinoma derived from a patient, it was possible
to determine that insulin could be derived in vitro from a
single molecule that was converted to the A and B chains by trypsin
treatment (43 ). Studies done in rat islets subsequently demonstrated
conversion of proinsulin to insulin, in a relatively slow process
taking approximately 40 min (44 ). During the same period, work done by
Howell and Taylor (45 ) on insulin biosynthesis showed that newly
synthesized insulin was released several hours after its biosynthesis.
The emerging view of these findings was that some sort of orderly
vectorial transport occurred involving the RER, the Golgi complex (GC),
and secretory granules (SG). These data (46 ), along with the major
contributions of other investigators who established that the
biosynthesis of serum albumin, PTH, and glucagon also originate from
larger precursors, formed the basis of the prohormone theory. This
theory states that synthesis of peptide hormones and neuropeptides
begins with mRNA translation into a large, inactive precursor peptide,
followed by limited posttranslational proteolysis to release bioactive
end products. Chretien and Li (47 ) also made an important contribution
to the prohormone theory when they determined the amino acid sequences
of ß-lipotropin (ß-LPH), 
-LPH, and ß-melanotropin (ß-MSH).
They observed that ß-MSH was part of the ß-LPH sequence, providing
evidence that ß-MSH was a conversion product of ß-LPH. They also
observed that cleavage occurred at the C-terminal side of paired basic
lysine or arginine residues. More definitive evidence for a
precursor/product was provided with the cloning of POMC (29 kDa), which
revealed that the ACTH and ß-LPH sequences were present within the N
terminus of POMC (38 ).
In summary, the biochemical processing of neuropeptides, peptide hormones, and other secreted proteins begins with limited posttranslational proteolysis of larger inactive precursors. Prohormones generally have their hormone sequences flanked by a single, a pair or four (tetra) basic amino acids where subtilisin-like processing enzymes produce their initial endoproteolytic cleavage (48 49 ). This endopreoteolytic cleavage is produced at the C-terminal side of the single or paired basic amino acid residue(s) which is followed by removal of the basic residue(s) by carboxypeptidase enzymes (CP) (50 51 ). Further modifications can occur in the form of N-terminal acetylation, pyroglutamate formation, or C-terminal amidation, which confers bioactivity to many peptides (52 ).
At the intracellular level, hormone precursors are synthesized on membrane-bound ribosomes, by which they are translocated into the lumen of the RER via a signal recognition peptide. During vectorial transport through the GC and beyond, the newly synthesized proteins are subjected to posttranslational modifications including glycosylation, phosphorylation, amidation, acetylation, and proteolytic conversion (48 ). Ultimately, partially processed proteins reach the last compartment of the GC, the trans-Golgi Network (TGN). At the TGN, unprocessed or partially processed products are sorted to the RSP (29 53 54 ) to be stored in immature secretory granules (ISG). Upon maturation, electron-dense SGs containing sorted products can fuse with the plasma membrane in response to an extracellular stimulation in a calcium-dependent manner, thereby releasing their contents into the external milieu (55 ). Two pathways of unstimulated release are proposed for AtT20 cells: constitutive (nongranular) secretion and basal release from compartments that form after sorting into the RSP (56 ). The mechanism whereby constitutive and regulated proteins are differentially sorted into separate vesicles after budding from the TGN is still under intensive investigation (57 58 ).
Two hypotheses have been proposed to explain how proteins are selectively targeted from the TGN to the RSP. In the first hypothesis, proteins are sorted by passive aggregation, in which the proteins condense within forming ISG, thereby excluding other proteins from entering in the granule. This process occurs under acidic pH and high calcium concentrations (59 60 ). Support for the aggregation hypothesis comes from studies done with chromogranin A (61 62 ), chromogranin B (62 ), carboxypeptidase E (CPE) (63 ), and prohormone convertase 2 (PC2) (64 ). However, there are data suggesting that aggregation alone is not sufficient for sorting. Modifications of the chromogranin B sequence (65 ) can prevent the correct sorting of these peptides to the RSP, while their in vitro aggregation properties appeared unaltered. The insulin-like growth factor-1 (66 ) does not aggregate in the TGN, but is still sorted in the RSP.
The second hypothesis, originally proposed in 1985 by Kelly (67 ),
involves cis-acting sorting signals within a protein
destined for the RSP that interact with membrane-bound sorting
receptors. Sorting receptors, possibly located in the forming SG,
direct segregation of the protein for further packaging into SGs.
Protein aggregation within the ISG can occur in this model, but is more
critical for product concentration than sorting per se.
Evidence supporting this second hypothesis has come from experiments
involving chimeric proteins (68 69 70 71 72 73 ), where the fusion of
constitutively secreted protein to a protein destined for RSP caused a
rerouting of this protein to the RSP. Conversely, proteins that have
their sorting signal domains modified may be misrouted from the RSP
into the constitutive pathway, as demonstrated for POMC (74 ),
chromogranin A and B (75 ), PC2 (69 ), and glycine 
-amidating
monooxygenase (PAM) (76 ). A related hypothesis is sorting by retention,
in which all proteins are initially targeted to ISGs, after which
proteins that do not belong in the RSP are removed to their final
destination, e.g., lysosomal enzymes (57 77 ).
It has been proposed recently that the membrane form of CPE, localized to the TGN, is a sorting signal receptor (78 ). CPE is proposed to direct POMC, proinsulin, proenkephalin, but not chromogranin A, into the RSP (79 ). Thus, CPE is a common sorting receptor for some, but not all, prohormones, and there must be other sorting receptors to direct trafficking of other proteins to the RSP (79 ). However, Irminger et al. (80 ) have provided evidence refuting the claim that CPE is a sorting receptor for proinsulin. In those studies they used pancreatic islets isolated from CPE-deficient (Cpefat/Cpefat) and control (Cpefat/+) mice to examine whether the trafficking of proinsulin and insulin was affected. They found that CPE was not essential for the sorting of proinsulin to the RSP (80 ). However, similar experiments with procholecystokinin in Cpefat/Cpefat mice indicate that CPE does function as a sorting receptor (81 ).
With this as a background, we turn specifically to the processing of
pro-TRH. The elucidation of the rat prepro-TRH sequence in 1986 was of
key importance to understanding the processing of pro-TRH to TRH and
non-TRH peptides. Rat prepro-TRH is a 29-kDa polypeptide composed of
255 amino acids. The rat precursor contains an N-terminal 25-amino acid
leader sequence, 5 copies of the TRH progenitor sequence
Gln-His-Pro-Gly flanked by paired basic amino acids (Lys-Arg or
Arg-Arg), 4 non-TRH peptides lying between the TRH progenitors, an
N-terminal flanking peptide, and a C-terminal flanking peptide (82 83 ). The N-terminal flanking peptide
(prepro-TRH25–50-R-R-prepro-TRH53–74) is
further cleaved at the C-terminal side of the arginine pair site to
render prepro-TRH25–50 and prepro-TRH53–74,
thus yielding a total of 7 pro-TRH-derived peptides (Fig. 1).
The biosynthesis of TRH and other pro-TRH-derived peptides follows the same prohormone-processing mechanisms described above, beginning with mRNA-directed ribosomal translation, followed by posttranslational limited proteolysis of the larger precursor, proTRH. This process occurs while pro-TRH is transported from the TGN to newly formed ISGs (84 ). These granules then mature and are targeted to sites of secretion at the plasma membrane of the cell. Rat, mouse, and human pro-TRH, similar to other peptide hormone precursors such as pro-enkephalin, contains multiple copies of one of its peptide products, in this case, the progenitor for TRH, Gln-His-Pro-Gly. Most of the products derived from pro-TRH are targeted into the RSP. Cleavage of the precursor to generate biologically active TRH occurs at paired basic residues by the action of PC1 and PC2 (31 35 37 ) followed by the action of CPE to remove the basic residue(s) (85 ). Gln-His-Pro-Gly is then amidated by the action of PAM, which uses the C-terminal Gly as the amide donor, and the Gln residue undergoes cyclization to a pGlu residue to yield TRH.
In the last few years, this laboratory has elucidated the processing steps involved in the synthesis of TRH and pro-TRH-derived peptides. An understanding of the biosynthesis and processing of pro-TRH is critical to appreciating how, when, and where modulation of this central regulator of the HPT axis takes place during physiologically appropriate modulation of thyroid function. It is also critical to understanding the function of TRH and other pro-TRH-derived peptides in extrahypothalamic regions of the brain, or outside of the nervous system, as discussed in Section VI of this review. Initial studies of TRH biosynthesis were difficult because of the low levels of TRH and other pro-TRH-derived peptides produced in hypothalamic tissue in vivo. In the search for a better system to perform these studies, initial work was done by stably transfecting prepro-TRH cDNA into transformed cell lines (86 87 ) to attain higher levels of pro-TRH expression. Among the cell lines investigated, only AtT20 (corticotroph, mouse) and RIN 5F (insulinoma, rat) cells were able to efficiently cleave the TRH precursor at paired basic amino acid residues to generate mature TRH, as well as pro-TRH-derived peptides, which were identical to those previously identified in vivo (88 89 ). GH4C1 (somatommamotroph, rat) and 3T3 (fibroblast, rat) cells were unable to process the pro-TRH precursor. When these studies were performed, little information was available about PCs, their role in prohormone processing, or their expression in different tissues and cell lines. Retrospectively, it was determined that the cell lines capable of processing proTRH, AtT20 and RIN 5F cells, also express PC1 and PC2 (90 ).
A model of pro-TRH posttranslational processing was developed in
experiments with transfected AtT20 cells expressing rat
prepro-TRH (30 ). By means of Western blot analysis, immunoprecipitation
followed by SDS-PAGE, and RIA, it was determined that pro-TRH is
present in transfected AtT20 cells and in primary cultures
of hypothalamic neurons, an endogenous source of pro-TRH, as a 26-kDa
protein (37 ). Pulse-chase studies indicated that the 26-kDa precursor
is cleaved at two mutually exclusive sites to generate the first
intermediate forms (Fig. 1). One cleavage generates a 15-kDa N-terminal
peptide (prepro-TRH25–151 or 157) and a 10-kDa C-terminal
peptide (prepro-TRH154 or 160–255). An alternate cleavage
generates a 9.5-kDa N-terminal peptide (prepro-TRH25–106 or
112) and a 16.5-kDa C-terminal peptide (prepro-TRH109 or
115–255). These cleavage steps occur in the TGN (Fig. 2),
before packaging into ISGs (29 84 ), in agreement with similar studies
of the cellular location of early processing for POMC and somatostatin
(SRIF) (53 ).
|
). Processing of the 9.5-kDa N-terminal fragment
arising from the alternative cleavage of the 26-kDa prohormone at
residues 107–108 is postulated to result in the production of the
N-terminal peptides, prepro-TRH25–50,
prepro-TRH53–75, and prepro-TRH83–106, while
the 16.5-kDa fragment is processed to produce the 5.6-kDa
prepro-TRH160–199 and the 5.4-kDa
prepro-TRH208–255 (Fig. 1).
Recent experiments confirm the proposed model that the C-terminal
10-kDa peptide derived from an initial cleavage at residues
Lys152-Arg153 is the precursor to two peptides
of 5.6- (prepro-TRH160–199) and 5.4-kDa
(prepro-TRH208–255) (Fig. 1). After the cleavage of basic
residues at positions 199–200 or 207–208, the 5.6-kDa peptide is
further processed to generate the prepro-TRH160–169 and
prepro-TRH178–184. prepro-TRH178–184 is
further cleaved to two novel peptides, prepro-TRH178–184
and prepro-TRH186–199 (91 ), Fig. 1).
B. Intracellular sites of pro-TRH processing
Initial studies in 1993 showed that processing of the 26-kDa
pro-TRH precursor to smaller intermediates occurred before packaging
into ISGs (29 ). The experimental strategy used to more precisely define
the intracellular sites of pro-TRH processing was to block peptide
transport from one cellular compartment to another and to characterize
processing that occurred before the point of blockade (84 ). To study
processing that occurs in the RER, peptides were blocked in their exit
from this compartment with brefeldin A (BFA) treatment of
AtT20 cells expressing the prepro-TRH cDNA (30 87 ). BFA is
a fungal metabolite that blocks ER-to-Golgi transport of proteins by
reversibly inhibiting the exchange of GDP for GTP in the GTP-binding
protein that is a key component in the vesicular transport, the ARF
protein. This prevents protein recruitment to intracellular membranes
and inhibits subsequent vesicle formation. To study the processing
steps that occur in the GC, cells were incubated at reduced
temperatures. Incubation of cells at 20 C prevents packaging of
proteins into ISGs at the TGN and has been used previously to study the
processing of other prohormones (53 92 ). Figure 2 summarizes the results of these
blockade experiments, described more fully below.
When transfected AtT20 cells expressing prepro-TRH were treated with BFA to block ER-to-GC transport, the 26-kDa pro-TRH precursor accumulated 4-fold over control levels, indicating a significant degree of post-ER processing. However, some processing of 26-kDa pro-TRH to the 15-kDa and the 9.5-kDa N-terminal intermediates, and the 16.5-kDa and 10-kDa C-terminal intermediates, was seen. The accumulation of the 16.5-kDa intermediate was not as great as that of the 15-kDa intermediate when compared with control levels (4-fold for the 15 kDa and about 2-fold for the 16.5 kDa) (84 ). To further clarify where the initial site of pro-TRH cleavage occurred, a combination of temperature blockade and BFA treatment was performed. With this strategy it was demonstrated that the 26-kDa precursor protein (accumulated 4-fold) is processed in the GC (possibly in the TGN) to generate the 15-kDa/10-kDa and the 9.5-kDa/16.5-kDa intermediate pairs. When the fate of the 15- and 16-kDa intermediates was analyzed, while they were retained within the TGN at 20 C, the 16-kDa C-terminal intermediate was further processed at basic residues 206–207 to the 5.4-kDa C-terminal peptide. In contrast, the 15-kDa N-terminal intermediate appeared to undergo processing in a post-GC compartment, i.e., the SGs. This observation strongly suggests that these two intermediates follow different paths of processing (84 ).
Evidence supporting differential distribution for the N- and C-terminal
peptides comes from recent immunocytochemical (ICC) studies using
transfected AtT20 cells, which indicate that pro-TRH, as
well as the 15- and the 6-kDa N-terminal intermediates, are located in
the GC and TGN (Fig. 3A). In
contrast, end products, including
prepro-TRH25–50, prepro-TRH160–169,
and TRH, are only present in SGs along the plasma membrane and in cell
processes (Fig. 3F). C-terminally directed antisera that recognize
pro-TRH and the 16.5- and 5.4-kDa C-terminal peptides result in
positive immunostaining in the GC, along the plasma membrane, and in
cell processes (Fig. 3H). Thus, C-terminal intermediates appear to
reach further along the RSP before processing than their N-terminal
counterparts. This differential processing might serve as a mechanism
to regulate the timing of production of peptides such as
prepro-TRH160–169, prepro-TRH178–199, and
prepro-TRH53–75, and possibly TRH. For example, the
16.5-kDa intermediate, which is processed in the TGN, contains
prepro-TRH178–199 and preproTRH160–169. A
portion of such peptides, formed before their entry into SGs, might
exit the cell via the constitutive pathway to maintain a basal level of
release independent of TRH secretion.
|
, B and
E). Immunoelectron microscopy (IEM) confirms positive staining in the
GC (Fig. 3E). Some ISGs budding from the TGN-like were also
stained. Since positive staining was detected in boutons, it was
proposed that these intermediates are processed to mature
pro-TRH-derived peptides near the axon terminal, before secretion. ICC
using anti-TRH, anti-prepro-TRH53–75, and
anti-prepro-TRH160–169 showed positive staining only in
the neuronal processes and axon terminals while the soma remained
unstained (Fig. 3G). The positive staining observed in neuronal
processes and axon terminals with these antibodies was confirmed by IEM
(Fig. 3G, inset).
Note that in transfected AtT20 cells using the same
antiserum, immunostaining was seen only in GC and ISGs, but not in
cellular processes (Fig. 3A). ICC on transfected AtT20
cells using anti-TGN38, a TGN marker, indicated that the
TGN can be projected away from the GC toward budding ISGs (Fig. 3C).
Thus, in AtT20 cells processing of the 15-kDa N-terminal
peptide takes place somewhere between the TGN and ISGs, suggesting that
prepro-TRH25–50, prepro-TRH53–75, and
prepro-TRH83–106 are already formed by the time SGs
mature.
Using the C-terminal antiserum that recognizes pro-TRH and the 16.5-,
10-, and 5.4-kDa peptides, positive staining was visualized in a patchy
cytoplasmic distribution (Fig. 3I), often closely associated with the
nucleus. Immunoreactivity was also observed in neuronal boutons, axon
terminals, and unbranched growth cones. IEM confirmed that the patchy
cytoplasmic areas were within the ER and GC (36 ) (Fig. 3I, inset); all layers of the GC (cis, medial, and
trans) were immunostained with this antibody (36 ). The
contrasting staining patterns for the two antisera (N- and C-terminal)
(Fig. 3, B and I) suggest the existence of a different peptide
distribution for N-terminal vs. C-terminal peptides,
possibly due to different intracellular routing of intermediates to
SGs. However, no conclusive data are yet available to confirm this
hypothesis.
C. Tissue-specific processing of pro-TRH
Increased knowledge of differential processing has led to a better
understanding of how multiple biological peptides, with different
functions, are generated from the same prohormone. This concept is
further reinforced by the observation that certain regions in the brain
can give rise to several different pro-TRH-derived peptides in addition
to, or instead of, TRH. This will become clear in a review of the
neuroanatomical distribution of pro-TRH and pro-TRH-derived peptides.
TRH-positive axons in the ME originate from neuronal perikarya in the PVN. The PVN is composed of two major components, the magnocellular and the parvocellular divisions. The parvocellular division contains most of the TRH neurons that project to the ME. A large population of immunoreactive neurons is located in medial and periventricular parvocellular subdivisions, organized in a triangular configuration, symmetric to the dorsal aspect of the third ventricle, whereas the anterior parvocellular subdivision neurons are more disperse (93 ). However, not all TRH-containing neurons in the PVN project to the ME (94 ). In addition to the PVN, TRH neurons are present in many other regions of the hypothalamus (93 ). As these populations of neurons have no known projections to the ME, and are not regulated in conjunction with the thyrotropic neurons of the PVN, it is presumed that they do not subserve a direct hypophysiotropic function.
The largest concentration of hypothalamic TRH neurons outside of the
PVN are found in the dorsomedial nucleus, lateral hypothalamus, and
preoptic area, including medial, periventricular, suprachiasmatic, and
the sexual dimorphic nucleus of the preoptic area (93 ). In addition to
the PVN, TRH neurons are present in many other regions of the CNS
including regions in the diencephalon, telencephalon, mesencephalon,
myelencephalon, and spinal cord (Table 1). An extensive anatomic description of
TRH neurons and TRH fibers in these tissues has been reported
previously (93 ). While the presence of TRH in these regions is clear,
relative levels are largely unknown because most mapping has depended
on ICC detection in animals pretreated with colchicine. Detailed
microdissection studies, combined with RIA, have been done for some
areas, including the brainstem and hippocampus. These studies do not
require colchicine pretreatment, are more quantitative than ICC
detection, and reveal an even broader distribution for TRH than
appreciated by ICC. However, their neuroanatomical resolution falls
short of that of ICC, and these latter studies are the primary source
of TRH data presented in Table 1.
|
). In the case
of pro-TRH, the reticular nucleus of the thalamus contains abundant
prepro-TRH mRNA and several pro-TRH-derived peptides in their extended
forms, but does not contain mature TRH (95 ). Moreover, the N-terminal
extended forms of TRH, TRH-prepro-TRH160–169 and
TRH-prepro-TRH178–199 are major end products of pro-TRH
processing in the olfactory lobe (OB) (96 97 ), but not in the
hypothalamus where pro-TRH is completely processed to non-TRH peptides
and TRH (96 ). In the ME, PVN, and preoptic area (POA), pro-TRH is fully
processed to its mature forms, while in the OB less than 60% of
N-terminal prepro-TRH25–50 is formed. Similarly, while in
the OB the 10-kDa C-terminal intermediate (see Fig. 1) is the main end
product, in the ME and PVN this intermediate is fully processed to
prepro-TRH160–169 and prepro-TRH178–199 (Fig. 1). Finally, in the POA partial processing of the 10-kDa peptide is
observed, and the lateral hypothalamus contains lower levels of both N-
and C-terminal pro-TRH-derived peptides as compared with the ME and POA
(E. A. Nillni, unpublished data). Table 1 summarizes the current
information regarding neuroanatomical distribution of intermediate
forms of pro-TRH processing, mature pro-TRH-derived peptides, and TRH
peptides in perikarya and fibers.
D. The role of PCs and CPs
The PCs are a family of seven subtilisin/kexin-like endoproteases
including furin, PC1 (also known as PC3), PC2, PC4, PACE4, PC5-A (also
known as PC6-A), its isoform PC5-B (also known as PC6-B), PC7 (also
known as LPC), and PC8 (also known as SPC7) (32 33 34 98 99 100 101 )). The
structure of these serine proteinases resembles both the bacterial
subtilisins and yeast kexin (90 102 103 ). These enzymes cleave at the
C-terminal side of single, paired, or tetra basic amino acid residue
motifs (104 ), followed by removal of remaining basic residue(s) by CPs
(50 51 ). The selective expression of PC1 and PC2 in endocrine and
neuroendocrine cells suggests they are significant in prohormone
processing (32 90 98 105 ). PC1 and PC2 have been shown to process
pro-TRH (31 35 37 106 ), proinsulin (104 107 108 ), proenkephalin
(109 ), prosomatostatin (110 111 ), and POMC (112 113 ) to various
intermediates and end products in coexpression experiments.
Like their substrates, PC1 and PC2 undergo maturation from larger precursor proteins. Maturation of pro-PC1 begins in the ER and continues in the TGN (90 ). In contrast, pro-PC2 maturation begins in the TGN and continues in the SGs; active PC2 and PC1 accumulate in SGs (114 115 ). PC2 has been proposed to produce smaller peptides from intermediates resulting from cleavage by PC1 (109 112 113 ). Consistent with this, PC1 is implicated in the early cleavage of POMC (112 113 116 ) and proinsulin (117 ). Thus, the compartments where these enzymes are active also are determined by differential timing of their respective maturation.
The distribution of neurons containing mRNAs encoding prepro-TRH, PC1,
and PC2 in the PVN and other areas of the brain has been determined
using in situ hybridization (106 ) (Table 1). The glomerular
layer of the OB displays coexpression of prepro-TRH with PC2 mRNA, but
not PC1, whereas in the tenia tecta coexpression of mRNA for prepro-TRH
with PC1, but not PC2, is evident. The PVN displays prepro-TRH mRNA
coexpression with both enzymes, whereas the lateral hypothalamus shows
coexpression of prepro-TRH mRNA with PC2 mRNA, but not PC1 mRNA. Double
in situ hybridization indicates that in the PVN, PC2 mRNA is
present in 60–70% of TRH neurons, and PC1 is present in 37–46% of
TRH neurons (118 ). Even though these investigators found a trend for
more coexpression of mRNA for prepro-TRH with PC2 than PC1,
coexpression alone does not define which enzyme is more important in
the processing of pro-TRH in vivo.
During the last few years, Nillni’s laboratory has provided unequivocal evidence for the role of PC1 and PC2 in the processing of pro-TRH (31 35 36 37 ). In coinfection experiments, where recombinant vaccinia viruses are used to coexpress PC1, PC2, PACE4, PC5-B, and furin, together with rat prepro-TRH in constitutively-secreting LoVo cells or in the regulated-secreting endocrine GH4C1 cell line, RIA of LoVo-derived secreted products demonstrates that furin cleaves the precursor to generate both N- and C-terminal intermediates, while PC5-B does not produce any peptide. PC1, PC2, and PACE4 only produce N-terminal intermediates, and less efficiently than furin. Interestingly, in LoVo cells, furin cotransfection produces TRH-Gly at much greater levels than any of the other PC enzymes. Recent data indicate that furin, which is ubiquitously expressed in all tissues, may serve a role in processing of prosomatostatin within the constitutive pathway (110 111 ). Since LoVo cells only contain the constitutive secretory pathway, these results suggest that pro-TRH can be processed to a certain extent without entry into the RSP. However, caution must be taken with this interpretation because under conditions of viral expression, the unusually high level of virus in coinfected cells can produce disruption of cellular compartments. The products resulting from coexpression of prepro-TRH with either furin or PC1 are similar, in agreement with their similar specificity observed in a number of cell coexpression experiments (109 111 ) and in vitro data (119 ).
In GH4C1 cells, PC1, PC2, furin, PC5-B, and PACE4 produce
both N-terminal and C-terminal peptides derived from pro-TRH.
Significantly, TRH-Gly and TRH are produced in highest amounts by PC1,
PC2, and furin. Further analysis of the cleavage specificity of PC1 and
PC2 reveals that PC1 is primarily responsible for cleavage of the
entire TRH precursor to mature TRH, as it can generate all products at
significantly higher levels than PC2 (Fig. 4). While 7B2 is known to be involved in
the maturation of PC2 (Fig. 4) (90 120 ) it does not augment the
ability of PC2 to cleave pro-TRH to either N- or C- terminal forms.
Subsequently, we have examined the role of PC1 and PC2 in the formation
of prepro-TRH178–199 by coexpressing rat prepro-TRH cDNA
with PC1, PC2, and 7B2 in GH4C1 cells (91 ). PC1
effectively cleaved pro-TRH to immunoreactive forms recognized by
anti-prepro-TRH178–199, while PC2 played a minor role,
even in the presence of 7B2 (Fig. 1).
|
) (37 ).
|
-Ser186-Trp187-Glu188-Glu188-Lys189...
) generating two novel peptides, pFQ7 and pSE14
(Fig. 1).
These two novel peptides, prepro-TRH178–184 and
prepro-TRH186–199 are present in rat PVN, lateral
hypothalamus, and ME (91 ). Thus, the antibody generated against the
prepro-TRH178–199 sequence recognizes the 10-kDa peptide,
a 5.6-kDa form that probably is prepro-TRH160–199 (30 ), a
2.6-kDa peptide that is prepro-TRH178–199, and two smaller
moieties of 1.6 and 0.84 kDa that are proposed to be
prepro-TRH186–199 and prepro-TRH178–184,
respectively (Fig. 1). Figure 6 shows a
diagrammatic representation of rat pro-TRH and its cleavage by PC1 and
PC2 as proposed from the most recent studies and compared with previous
in vitro studies (31 35 91 ).
|
combines current neuroanatomical distribution data for
prepro-TRH mRNA, pro-TRH, and pro-TRH-derived products with the
distribution of PC mRNAs and enzymes. Also included is a summary of
pro-TRH processing in transfected cells and primary cultures of
hypothalamic neurons and pituitary cells. Several important conclusions
can be drawn. pro-TRH is widely distributed in the hypophysiotropic and
extrahypophysiotropic areas of the brain. The widespread expression of
pro-TRH, PC1, and PC2 mRNAs, with their overlapping distribution in
many areas of the rat CNS, indicates the striking versatility provided
by tissue-specific processing in generating quantitative and
qualitative differences in non-TRH peptide products as well as TRH.
Examples of these differences for several tissues are presented in the
first column of the table and described above. A most striking example
is the reticular nucleus of the thalamus, where PC1 and PC2 are not
coexpressed with pro-TRH. TRH is not produced in this nucleus,
indicating a central role for PC1 and PC2 in maturation to TRH.
However, other pro-TRH intermediates are present in the reticular
nucleus, suggesting PCs other than PC1 and PC2 might be involved in
processing of pro-TRH for this particular region of the CNS (106 ).
Differential processing has been reported for other prohormones, and
these differences relate to alterations in the expression of various
PCs within different cell types. POMC is processed primarily to ACTH,
ß-endorphin, and N-POMC1–77 in the anterior pituitary
(melanotrophs). In turn, these products are further processed to
-MSH, ß-endorphin1–31, N-POMC1–49,
and -MSH in the intermediate lobe and brain (38 ). Differential
processing of a common polypeptide precursor is dependent upon the
processing enzymes expressed in each specific cell type. Proenkephalin,
which contains seven identical copies of met-enkephalin, is processed
to large intermediate forms in the adrenal medulla, whereas this
precursor is cleaved primarily to the pentapeptide met-enkephalin in
the brain (39 ). The biological actions of substance P (SP) depend on
the enzymatic processing of its precursor by the processing enzymes
prolylendopepetidase to yield SP5–11, and endopepetidase
3.4.24.11 to yield SP1–7. While SP1–7 acts as
an analgesic, inhibits aggression, and enhances learning and memory,
the SP5–11 enhances pain transmission, stimulates
aggression, and blocks learning and memory (40 ). In the brain,
procholecystokinin (pro-CCK) is processed to produce only CCK8 amide,
while in the gut the precursor is cleaved to larger molecules, such as
CCK12, 22, 33, 38, 58, and 83 amide (121 ). Transfection experiments
have shown that proneuropeptide Y (pro-NPY) can be cleaved by cell
lines expressing either PC1 or PC2, but pro-NPY is primarily processed
by PC2 in superior cervical ganglia (122 ). Thus, differential
processing of neuropeptitdes including pro-TRH, pro-NPY, POMC, pro-CCK,
SP, and proenkephalin provides a critical mechanism through which cells
regulate the levels of specific peptides to fulfill different
physiological requirements, a mechanism potentially more versatile than
the alternative splicing of mRNA.
As mentioned in Section IIA, CPs remove remaining C-terminal basic residues from prohormone intermediates that are initially cleaved by PCs. Experiments with the fat/fat mouse model of CPE deficiency (123 ) support a role for CPE in the processing of pro-TRH (85 ). Mice homozygous for the fat/fat mutation are obese, diabetic, and infertile. These mice have a missense (Ser to Pro) mutation at CPE residue 202 that abolishes enzymatic activity (123 ). Hypothalamic TRH levels are depressed 65% in fat/fat mice relative to heterozygous controls. SDS-PAGE demonstrates hypothalami from both wt/fat and fat/fat mice contain moieties different from those of the wt/wt mice. Specifically, 6.5-, 3.0-, 2.6-, and 1.6-kDa forms of the pro-TRH sequence are detected, and their levels differ significantly between the two groups. Compared with wt/fat mice, fat/fat mice hypothalami contain 20-, 3-, and 2-fold elevations in the 6.5-, 3.0-, and 2.6-kDa species. These data indicate that the fat/fat mutation produces qualitative changes in pro-TRH processing, and that CPE is involved in the later stages of pro-TRH processing. However, cell transfection experiments would help to rule out secondary phenotypic changes caused by the CPE mutation in these mice. Further, since hypothalami from fat/fat mice contain immunoreactive TRH, additional CPs must also be able to process pro-TRH to TRH, assuming that the TRH detected is not a cross-reactive non-TRH species. CPs such as carboxypeptidase D, with similar enzymatic properties to CPE, are also present in compartments of the secretory pathway and are distributed in many tissues, including the brain (50 ).
Interestingly, in fat/fat mice, levels of TSH,
T3, and T4 were normal, suggesting that 34% of
normal TRH levels is sufficient to maintain the thyroid function. This
last observation is important because it is hypothesized that the five
identical progenitor sequences of TRH contained in the prohormone may
not be processed to mature TRH at all times, and that only a few
of them may be needed to maintain the thyroid function. In cultured
hypothalamic cells at steady state, previous studies had shown that the
ratio of mature TRH to prepro-TRH25–50 and the 5.4-kDa
C-terminal peptide (Fig. 1) is 3:1 instead of the theoretical 5:1 for
complete processing (36 ). A similar ratio is seen in transfected
AtT20 cells (30 ). In the rat brain, the ratio of TRH to
other pro-TRH-derived peptides sequences is almost 1:1 in the
hypothalamus, and 1.5:1 in the olfactory lobe (89 ). The above ratios
are indicative of incomplete yields of TRH from pro-TRH, although only
if degradation rates do not contribute significantly to these various
ratios. Further, the relative immunoreactivites of various antisera
used to their iodinated tracers has not been defined, leaving exact
molar ratios difficult to calculate. Still the range of TRH to
pro-TRH-derived peptides found in various tissues makes it likely that
full TRH yields are not achieved in all, or even most, tissues.
If not all TRH progenitor sequences are cleaved from pro-TRH, the resulting TRH progenitor sequences, linked to amino- or carboxyl-terminal extensions, would not be detected in usual TRH RIAs. The TRH assay is specific for mature TRH, needing both the amino-terminal pyro-Glu and carboxy-terminal Pro-amide for detection. Nonimmunoreactive TRH progenitors might also retain untrimmed basic amino acid residues, carboxyl-terminal glycine residues, etc. Another explanation for the lower than 5:1 ratio of TRH to other products could be that TRH is extracted less efficiently from cells than these other peptides, although this seems unlikely because TRH is the smallest of the peptides, and doping experiments indicate TRH is extracted with high efficiency under our conditions. As described above, it is proposed that the processing of peptides derived from the N-terminal portion of pro-TRH is substantially different from the processing of those of the C-terminal end. This suggests the possibility of differential maturation of TRH molecules depending on location within the pro-TRH sequence (84 ). If this is the case, either the three TRH molecules derived from the C-terminal side, or the two from the N-terminal side, may play a more primary role in hormonogenesis. It is also possible that certain TRH molecules become biologically active at different times than others, or only some of them reach maturity while the rest are degraded by the TRH-degrading enzymes, depending upon physiological needs (see Section VII). Finally, cells may simply produce excess TRH that may or may not be used.
In summary, we have presented an overview of the current knowledge of pro-TRH biosynthesis, its processing, its tissue distribution, and the role of known processing enzymes in pro-TRH maturation. Evidence is presented suggesting differential processing for pro-TRH at the intracellular level is physiologically relevant. The data indicate that PC1 is primarily responsible for most pro-TRH cleavage events. PC2 is involved in specific processing events that occur later in the secretory pathway, specifically in the formation of the second TRH molecule from the N-terminal side of prepro-TRH83–106, and the proteolytic cleavage of prepro-TRH178–199 to generate the novel prepro-TRH178–184 and prepro-TRH186–199 peptides.
E. Neuropeptide and catecholamine regulation of pro-TRH
biosynthesis and processing
Immunoreactive TRH (iTRH) axon terminals are present in high
density in the external layer of the rat ME, in close apposition to
capillaries of the hypophysial-portal system (124 ). These axons
originate from neuronal perikarya located in the PVN, the
"thyrotrophic area" of the rat hypothalamus. Destruction of this
region results in disappearance of up to 94% of TRH in the external
layer of the ME and reduction of TSH secretion from the anterior
pituitary gland (125 ). As described in Section IIC (Table 1), in addition to the PVN, iTRH neurons are present in other regions
of the hypothalamus, including the POA, anterior hypothalamus, and
supraoptic, arcuate, dorsomedial, and premmamilary nuclei, as well as
basolateral and prefornical hypothalmus (93 ).
Although many neurons in the PVN contain more than one peptide, TRH neurons are unique in being almost always unassociated with other known peptides (126 ). This makes the regulation of pro-TRH-derived peptide biosynthesis very specific. As described in more detail below, TRH neurons in the PVN are located in a region where they can be regulated by a number of neuroendocrine inputs. TRH neurons are densely innervated by norepinephrine (NE)-containing axons that stimulate TRH secretion (124 ). TRH neurons are also densely innervated by neuropeptide Y (NPY) neurons. In smaller numbers, SRIF and endogenous opioid peptide (EOP) terminals are also in contact with TRH neurons (124 ). In vivo, these various neuroendocrine inputs may affect the levels of prepro-TRH mRNA and the posttranslational processing of pro-TRH by influencing the biosynthesis and maturation of PC1 and PC2. At the present time, the effect of this input on pro-TRH processing is unknown. Evidence for coordinated regulation of mRNAs for processing enzymes and their substrates has been documented in several cases (127 ). In contrast, outside the hypothalamus TRH is colocalized with other substances. For example, in the descending bulbospinal pathway, TRH is colocalized with SP and serotonin (5-HT) (128 ). Regulation of these TRH neuronal systems is much less well characterized than the thyrotropic neurons of the hypohyseal-portal system.
TRH-synthesizing neurons in the rat PVN receive a large number of
afferent neuroendocrine inputs. Axon collaterals of parvocellular
neurons ramify within the medial parvocellular PVN and establish
numerous synaptic contacts with perikarya and dendrites of other
parvocellular as well as magnocellular neurons, e.g., SRIF
afferents are derived from the PVN itself (129 ). The majority of inputs
to TRH neurons are derived from the diencephalon, telencephalon, and
brainstem (130 ). The paraventricular and medial parvocellular divisions
of the PVN are densely innervated by NE-containing and epinephrine
(E)-containing inputs from the medulla and pons (131 ). Further,
NE-containing neurons densely innervate the midregion of the external
layer of the ME. These inputs activate tuberoinfundibular neurons.
Intracerebroventricular (icv) injections of NE, E, and 2-adrenergic
agonists stimulate basal TSH secretion (132 133 ), and NE/E treatment
of hypothalamic preparations stimulates TRH release (134 ). Inhibitors
of catecholamine (CA) biosynthesis or 2-adrenergic antagonists lead
to a fall in basal TSH secretion. Thus, NE and E exert a tonic,
stimulatory regulation on TSH secretion principally through
2-adrenergic receptors. Stimulated release from the ME appears to be
postsynaptically mediated via 1-adrenergic receptors (135 ). In
contrast, locus coeruleus (LC) afferents are inhibitory, being
activated during stress (136 ). NE/E excitation of PVN TRH neurons
mediates the rise in TSH in response to acute cold exposure or
hypovolemia (135 137 138 ). However, it has also been proposed that
1-adrenergic receptors mediate a phasic inhibitory regulation of TSH
release. The data on NE/E modulation of TRH biosynthesis may be
reconciled by an examination of how these inputs affect the
posttranslational processing of pro-TRH, as well as examining their
effects on PC biosynthesis. Peripheral levels of T3,
T4, or TSH may also influence NE/E effects on TRH
biosynthesis and/or release.
The PVN also receives prominent dopamine (DA) inputs from the posterior
and dorsal areas of the hypothalamus, the zona incerta of the
subthalamic region, and the A14 region of the anterior hypothalamus
(124 ). Mesencephalic A9 and A10 dopaminergic neurons also project to
the PVN. In turn, large terminal fields to the ME originate in the
arcuate nucleus and periarcuate nucleus regions of the hypothalamus. In
contrast to the NE/E system, DA inputs appear to inhibit TRH secretion,
mainly at the level of the ME (139 ). Augmentation of DA
neurotransmission inhibits basal and/or cold-stimulated TSH
release, while DA antagonism has the opposite effect, although some
studies have failed to replicate these findings (124 ). In addition, TRH
release may be indirectly inhibited by DA-stimulated secretion of SRIF
(140 ). Conversely, DA stimulates TRH release from isolated hypothalamic
fragments (134 140 ), again reinforcing the need to examine TRH
biosynthesis in both in vitro and in vivo
systems. Within the HPT axis, thyroid hormones appear to modulate DA
levels in the ME, and TSH increases the ability of DA to inhibit TRH
(141 ) (Table 1).
A wide array of neuropeptides, including NPY, TRH itself, SRIF, EOPs,
neurotensin (NT), and vasoactive intestinal polypeptide (VIP), have
inputs to the PVN and/or external layer of the ME (142 143 144 145 146 ). Other
mediators, including -aminobutyric acid (GABA) and various
cytokines, also appear to regulate TRH or TSH secretion, but there is
as yet no anatomical evidence to support a direct action on TRH neurons
in the PVN and/or ME (124 ). Anatomically, NPY appears most prominent in
its inputs to the periventricular and medial parvocellular divisions of
the PVN (147 ). NPY cell bodies principally reside in the medulla, often
coexisting with NE and E (148 ), but other sources come from throughout
the brain, including the arcuate nucleus of the hypothalamus itself.
Indeed, the arcuate nucleus is the major source for NPY fibers
innervating the TRH neurons in the PVN (149 ). Few NPY-containing axons
project to the ME. The effects of NPY on tuberinfundibular TRH are not
yet well understood. NPY neurons also innervate SRIF neurons in the
PVN, which would allow indirect regulation of TRH biosynthesis or
secretion (124 ). Central administration of NPY reduces NE utilization
in the PVN, as well as TSH release, indicating an inhibitory
influence. In vivo, NPY also increases hypothalamic DA
content, as well as DA turnover in the ME, the net result of which
would reduce TRH release as well (150 ). Physiologically, NPY is
critical to integrating thyroid function, food intake, and
thermoregulation (151 ) (Table 1).
Inputs containing EOPs represent a second rich innervation to the PVN. These originate in the arcuate nucleus, periarcuate area, and amygdala (144 ). The dorsal raphe projects 5-HT/enkephalin (ENK) axons, and the posterior hypothalamus-mammallary bodies send GABA/histamine/ENK projections, to the PVN. The ME contains numerous ENK, dynorphin, and endorphin (END) synapses originating from the PVN and arcuate nucleus. Both END and ENK inhibit TRH release from the hypothalamus, and ENK and morphine inhibit TRH secretion from the ME (152 ). There is additional evidence that ENK indirectly inhibits tuberoinfundibular TRH via DA release (153 ).
Recent data indicate that pro-TRH processing is regulated by opiate withdrawal (27 ). Opiate withdrawal increases prepro-TRH mRNA, and the N-terminal prepro-TRH53–74 and prepro-TRH83–106 peptides, in the rat PAG, whereas the level of TRH is unaltered (27 154 ). New data also show suckling increases the production of prepro-TRH178–199 and prepro-TRH186–199 (see Section IVC and E). (91 ). These results demonstrate that levels of various products derived from pro-TRH can be posttranslationally regulated in an independent fashion under altered physiological conditions. Thus, it is logical that neuroendocrine inputs into the PVN can affect pro-TRH processing as well.
Finally, we note that while the genomic organization of the rat prepro-TRH gene is well described (155 156 ), the molecular mechanisms regulating the expression of this gene are incompletely understood. The 5'-region of the prepro-TRH gene contains TATA and GC box sequences, also present in the promoter region of other neuropeptide genes (155 ). In addition, sequences similar to a cAMP response element (CRE), and negative thyroid response elements (TREs), are present. The region between -47 and +6 of the rat prepro-TRH gene is active in CA77 TRH-secreting medullary thyroid carcinoma cells (155 157 ), but not in transgenic mice (158 ). Inclusion of most of exon 1 (bp -47 to +84) increases promoter activity in CA77 cells and activates the promoter in transgenic mice, principally in prepro-TRH gene-producing tissues. Thus, cis element(s) located within exon 1 are necessary for the expression of the rat prepro-TRH gene in vivo (158 ). In CA77 cells, the human prepro-TRH gene is regulated by thyroid hormone through two distinct classes of negative TREs (157 ), similar to other neuropeptide genes such as prepro-SRIF (159 ).
F. Glucocorticoids modulate the biosynthesis and processing of
pro-TRH
Glucocorticoids evoke a broad spectrum of responses in many
eukaryotic cells by stimulating or repressing the transcription of
glucocorticoid-regulated genes, including those of peptide hormones
(160 ). The primary effect of glucocorticoids on gene transcription can
occur by specific binding of the steroid receptor complex to DNA at the
site of glucocorticoid response elements. Glucocorticoids can also
interfere with the action of other transcription factors through
protein-protein interactions and may elicit secondary effects at the
posttranscriptional, translational, and posttranslational levels
(161 162 163 ). For example, glucocorticoids stimulate processing of the
precursors for atrial natriuretic factor and neurotensin (NT) (162 163 ). Glucocorticoids also regulate the posttranslational maturation,
the intracellular trafficking, and the extracellular release of the
mouse mammary tumor virus (164 ).
Glucocorticoids enhance TRH gene expression in several in vitro cell systems, including hypothalamic neurons, anterior pituitary cells, and thyroid C cells, an effect that occurs, at least in part, through transcriptional activation (165 ). Dexamethasone substantially elevates biosynthesis of the 26- kDa TRH prohormone and its intermediate products in cultured anterior pituitary cells, consistent with an overall up-regulation of both the biosynthesis and processing of the TRH precursor (161 ). This explains why glucocorticoids act not only at the transcriptional level, but also at the translational/post-translational level. This question can be addressed in experiments with AtT20 cells transfected with prepro-TRH cDNA driven by a CMV-IE promoter not responsive to physiological signals. Dexamethasone causes a 75% increase in newly synthesized 26-kDa pro-TRH without altering prepro-TRH mRNA levels, suggesting that glucocorticoids raise translation rates and/or slow processing of pro-TRH. In fact, dexamethasone treatment accelerates TRH precursor processing.
Interestingly, processing of the N- vs. the C-terminal
intermediates in the AtT20 cells is influenced
differentially by glucocorticoids. Levels of the N-terminally derived
peptide prepro-TRH25–50 are enhanced while levels of the
5.4-kDa C-terminally derived peptide are reduced. TRH content is
increased (Fig. 7) (161 ). How could
dexamethasone differentially affect the processing of N- vs.
the C-terminal intermediates? Glucocorticoids may alter pro-TRH
processing through changes in the expression of processing enzymes, as
well as morphological alterations in AtT20 cells. For
example, GC volume is obviously enlarged in AtT20 cells
treated with dexamethasone. Although speculative, these changes may
slow down the normal transport of intermediate products from the TGN to
ISGs, thereby altering the accumulation or degradation of intermediate
forms through changes in processing enzyme exposure (84 161 ).
|
In conclusion, glucocorticoids induce changes in the biosynthesis and processing of pro-TRH by affecting both transcription and translation rates, and by differentially influencing the processing of N- vs. C-terminal intermediates of pro-TRH. At the translational and posttranslational level, these effects result in an increase in TRH production, with more complicated differential effects on the accumulation of other N- and C-terminal pro-TRH-derived peptides. It is clear that control over the diverse range of pro-TRH-derived peptides within a specific cell is accomplished mostly from the regulation at the posttranslational level rather than the translational or transcriptional levels. Three examples supporting this hypothesis are presented in this review: 1) pro-TRH processing in the PAG is regulated during the opiate withdrawal, so that levels of TRH remain unchanged, but other pro-TRH-derived peptides are induced (Section II.B.6); 2) pro-TRH processing is regulated during suckling, where a selective, yet dramatic, increase in prepro-TRH178–199 and prepro-TRH186–199 peptides is observed (Section IV.D and E); and 3) in the absence of transcriptional effects, glucocorticoids induce differential processing of pro-TRH in both primary cultures of pituitary cells and transfected AtT20 cells encoding prepro-TRH cDNA (this section).
G. Leptin regulates pro-TRH biosynthesis
Food deprivation in animals and humans results in endocrine and
metabolic changes including decreases in circulating thyroid hormones,
TSH, insulin, GH, gonadal hormones, and gonadotropins. Previous work in
starved rats has shown a decrease in hypothalamic, but not thalamic,
reticular, prepro-TRH mRNA, as well as decreased circulating TRH. This
supports the concept that hypothyroidism produced after starvation is
of hypothalamic origin (168 ). Leptin is a recently discovered peptide
hormone that is synthesized and released by adipose tissue. Leptin also
is decreased in starvation. Absence of leptin is responsible for the
obese phenotype of ob/ob mice, and administration of this
hormone to these animals decreases plasma corticosterone, suggesting
that leptin is capable of inhibiting the hypothalamic-pituitary adrenal
axis. In normal rats and mice, leptin inhibits hypothalamic CRH release
(169 ).
Leptin may have an important role in the neuroendocrine regulation of
the HPT axis (170 ). During prolonged fasting in rats, low levels of
T3 and T4 are observed, and TSH is in the low
to normal range. As described above, this is due in part to suppression
of prepro-TRH gene expression in PVN neurons. Since the decrease in
thyroid hormone levels is blunted in mice and rats by systemic leptin,
it has been proposed that the decrease in leptin detected during
fasting alters the set point for feedback inhibition by thyroid
hormones on the biosynthesis of prepro-TRH mRNA (170 ). The mechanism of
such leptin regulation of prepro-TRH biosynthesis is unknown. It is
hypothesized that leptin has direct actions on cell bodies in the
arcuate nucleus, positively regulating POMC, and thus -MSH, and
negatively regulating NPY and the Agouti-related peptide (151 ). NPY
afferents on TRH neurons are inhibitory (see Section II.E).
In preliminary studies done in this laboratory, both leptin and -MSH
elevate prepro-TRH mRNA, pro-TRH, and TRH secretion in primary
hypothalamic cultures (our unpublished results). Using the same
primary cultures of hypothalamic neurons, leptin dose-dependently
increases pro-TRH synthesis and TRH secretion. Immunocytochemical
analysis reveals that approximately 40–50% of the hypothalamic cells
are positive for the leptin receptor. Of these, approximately 10–15%
colocalize with TRH (Fig. 8). These data
suggest that the regulation of pro-TRH biosynthesis and TRH release in
response to starvation includes direct regulatory actions of leptin and
-MSH on hypothalamic TRH neurons involved in HPT axis homeostasis
(171 ).
|
-MSH
release from the arcuate nucleus, which may stimulate TRH release from
the PVN; and 3) a direct action of leptin on TRH neurons located in the
PVN. |
III. Function of TRH |
|---|
|
TopAbstractI. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
TSH is synthesized and secreted by the thyrotrophic cells in the anterior pituitary and is the major regulator of the thyroid gland. TSH secretion is primarily regulated by negative feedback from circulating thyroid hormone and by stimulatory input from the hypothalamus through TRH action on thyrotrophs. There are other factors known to regulate TSH secretion, including glucocorticoids in the systemic circulation and SRIF and DA from the hypothalamus. There is evidence supporting the view that glucocorticoids in man suppress endogenous hypothalamic TRH secretion (173 ). Further support for this hypothesis was demonstrated in adrenalectomized rats in which pro-TRH mRNA levels increase 1.6-fold, an increase reversible with dexamethasone (166 ). The role of glucocorticoids in vivo and in vitro has been described in the previous sections. Both SRIF and DA inhibit TSH release by direct effects on thyrotrophs. As discussed in Section II.C., SRIF perikarya in the POA, periventricular parts of the anterior hypothalamus, and a few in the PVN, and TRH neurons in the PVN, project to the ME. TRH and SRIF are the two main TSH-regulating hypophysiotropic neurohormones released into the hypophysial portal vasculature. The maintenance of euthyroidism is dependent on a highly regulated balance of neuropeptides and neurotransmitters, where the dominant positive hypothalamic control for TSH is TRH, and the principal feedback control is through thyroid hormones. However, thyroid hormones also provide direct negative feedback on prepro-TRH gene expression as well. The relationship of thyroid hormone regulation to pro-TRH processing is undefined.
Even though TRH is the major regulator of the synthesis and secretion of TSH, and thus plays a pivotal role in the HPT axis, in a recent study, homozygous TRH gene knockout mice were shown to be viable, fertile, and exhibit normal development (174 ). Whereas the TRH-/- mice showed normal serum PRL and GH levels, thyroid hormone levels were significantly reduced as compared with the wild-type heterozygous mice. The targeted disruption of the prepro-TRH gene caused a characteristic tertiary hypothyroidism, and a substantial decrease in insulin secretion resulting in a profound hyperglycemia. These authors suggested that in addition to abnormalities of the thyroid function, TRH may be involved in the pathogenesis of diabetes mellitus (174 ).
B. Extrahypophysiotropic TRH
More than two-thirds of iTRH in the brain is found outside of the
traditional "thyrotrophic zone" of the hypothalamus (175 176 ).
This extrahypophysiotropic TRH is believed to function as a
neuromodulator of known neurotransmitters (177 178 ). Indeed, it might
act as a neurotransmitter itself; it is present in secretory granules
whose exocytosis is responsive to membrane depolarization, it acts
through specific receptors that are widely distributed throughout the
CNS, and it is rapidly cleared through specific catabolic pathways
(179 ).
While the following discussion focuses on TRH, many other neuropeptides and neurotransmitters play critical roles in the biological functions discussed below. In several areas of the brain, TRH is colocalized with other neurotransmitters and/or neuromodulators, including 5-HT and SP in the bulbospinal pathway, DA in the olfactory bulb, and histamine, ENK, and NPY in various loci of the hypothalamus (128 ). Where TRH is directly affected by, or directly affects, other neuroactive agents, they have been discussed, but otherwise their roles are left for review elsewhere. Further, in the following we describe many effects of TRH. In fact, these have been demonstrated using TRH and/or TRH analogs. Due to space constraints, we will not distinguish effects by TRH vs. its analogs. Available TRH analogs have higher affinities for the TRH receptor, longer half-lives, etc. and are reviewed elsewhere (180 181 ).
Finally, for several TRH effects, the metabolite histidyl-proline
diketopiperazine, or cyclo (His-Pro) (CHP) also has agonist or
antagonist effects. CHP is present in the CNS and peripheral tissues in
levels that often exceed levels of TRH (182 ). While CHP is a known
breakdown product of TRH, there are data that CHP is also derived from
precursors other than TRH (183 ). Further, high-affinity binding sites
for CHP have not been identified (184 ) (but see Ref. 185 ). Thus, the
precise biological meaning of CHP effects is unknown. Figure 9 summarizes proposed physiological roles
for extra-hypophysiotropic TRH.
|
a. TRH in arousal and sleep: Systemic (188 189 ) and central (190 ) TRH increase wake time and/or decrease sleep time in multiple species. More dramatic is the ability of TRH to arouse animals from drug narcosis induced by alcohol (191 192 ), ß-endorphin (193 ), tetrahydrocannabinol (THC) (194 ), benzodiazepines (195 ), and barbituates (196 197 ). icv Anti-TRH antiserum doubles anesthetic-induced sleep time, supporting an endogenous role for TRH in arousal (198 ). Of unclear significance, CHP is more effective than TRH in decreasing sleep time and reversing ethanol-induced narcosis, but does not affect TRH modulation of barbiturate-induced sleep time in mice (199 200 ).
TRH analepsis is most strongly induced by infusions into the medial septal area, the diagonal band of Broca, or the nucleus basalis of Meynert (186 201 ). TRH levels rise in the medial septum as rats recover from ethanol-induced sedation (202 ). Further, TRH enhances cholinergic activity in the septo-hippocampal and nucleus basalis-cortical systems (203 204 205 ), pathways that play a central role in reversal of drug-induced narcosis (206 ). Atropine blocks both analepsis and cholinergic enhancement when TRH is infused into these areas. However, antagonism of the analeptic response to systemic TRH requires both ACh and NE blockade, supporting mediation by other neuroanatomical sites (192 207 ). The posterior lateral hypothalamus, the midbrain reticular formation, and TRH-containing fibers passing through, or synapsing in, the dorsal septum are additional neuroanatomical substrates that might mediate arousal through noradrenergic mechanisms (203 ).
There are limited data that TRH may be useful to enhance arousal in pathological states. TRH prevents the postconcussive elevation of cortical ACh and reduction of cortical NE seen in mice (208 ) and decreases time of unconsciousness in head-injured mice (209 ). Canine models of narcolepsy show improvement after treatment with TRH (189 ).
b. TRH in cognition. The ergotrophic effects of TRH on consciousness and arousal are often detected along with improvements in measures of learning and memory, consistent with the important role of ACh in both processes (210 ). TRH improves performance in learning-impaired mice, an effect that is blocked by scopolamine (211 ). TRH restores learning and/or memory deficits in rats made cognitively deficient by anticholinergic treatment, electrochemical shock treatment (ECS), or surgical lesions (212 213 214 ). These effects are largely mediated by enhanced cortical ACh release with TRH infusion into the nucleus basalis of Meynert and increased hippocampal ACh with TRH infusion into the medial septum-diagonal band (215 ). Facilitated release of NE (216 ) and N-methyl-D-aspartate (NMDA) receptor activation (217 ) also are implicated in cognitive enhancement by TRH. Human trials have demonstrated only modest cognitive improvements in patients with alcoholic dementia (218 ), in the ECS postictal state (219 ), and in Alzheimer’s disease (220 221 ).
c. TRH in locomotor activation. Systemic TRH elicits a number of motoric and behavioral activating effects in a wide range of species (178 222 ). Further, TRH antagonizes locomotor depression induced by alcohol (223 ) and ß-endorphin (193 ). TRH elicits locomotor activation when injected into the NAc (224 225 ), the ventral tegmental area (VTA), the caudate (222 ), the septal nuclei (225 ), and the ventromedial hypothalamus (226 ).
Locomotor activation by TRH principally is mediated by DA. Repeated TRH treatment in rats elevates DA in the cerebral cortex and increases tyrosine hydroxylase activity, the rate-limiting enzyme in DA biosynthesis in the brain stem. These effects correlate with dose- and time-dependent increases in locomotor activity (227 ). DA antagonists and/or DA depletion block locomotor activation by systemic TRH (228 ). Consistent with TRH activity on the mesocorticolimbic system, systemic TRH increases DA release in the NAc (224 229 ), and bath application of TRH induces DA release from NAc (230 ) and septal (231 ) brain slices.
Intraaccumbens TRH increases DA metabolism in the NAc, while icv TRH or
intra-VTA TRH do not, although these latter treatments elicit locomotor
excitation (222 ). Further, TRH can stimulate locomotion at doses that
have no effect on NAc DA (232 ). These data support redundant but
distinct TRH mechanisms, in addition to those in the mesocortical DA
system, that induce locomotion (233 ). Other studies indicate that
opiate antagonists and -adrenergic antagonists can attenuate
TRH-induced locomotion (234 ). A single study reports that CHP is able
to stimulate locomotion (200 ).
d. TRH and antidepressant effects: In a number of behavioral
assays used to screen compounds for antidepressant efficacy, TRH tests
positively (235 ). This potentiation is independent of effects on TSH or
thyroid hormones; however, PRL similarly can potentiate desipramine
effects in the forced swim test (236 ). For the most part, DA receptor
antagonists block the antidepressant effects of TRH (237 ), although it
is difficult to tease out locomotor effects from antidepressant effects
in some behavioral tests. TRH action in some antidepressant screens is
blocked by opioid receptor antagonists or 2-adrenergic blockade
(237 ). Further, antidepressant treatments alter TRH levels in rat
brain, but not in a clear "antidepressant pattern" (238 ). In human
trials, promising early results (239 240 ) have given way to larger
studies that indicate TRH is of limited benefit in depression (241 242 ). Recent studies using intrathecal TRH reveals significant
reductions in symptomatology of patients with refractory depression,
although tachyphylaxis to the effects develops rapidly (243 244 ).
Recently, a compelling model has been put forward that clinical
depression results from pathologically overdriven glutamatergic
circuits in the limbic forebrain that have escaped inhibitory
regulation by TRH (245 ).
2. TRH and autonomic nervous system function. The brainstem distribution of TRH supports a prominent role in autonomic nervous system (ANS) function. Fully 65% of medullary TRH is associated with dorsal vagal complex (DVC) neurons of the nucleus tractus solitaris (NTS), nucleus intercalatus and commisuralis, the dorsal motor nucleus (DMN) of the vagus, and, to a lesser extent, the nucleus ambiguus (246 ). Injection of TRH onto DMN neurons is uniformly excitatory, while applications onto NTS neurons are inhibitory (247 ). The majority of DVC iTRH derives from fibers arising from the medullary raphe nuclei that pass through the DVC (93 248 ). However, cells within the DMN express prepro-TRH mRNA and pro-TRH (95 249 ), consistent with some endogenous TRH production.
A descending bulbar-spinal pathway, in particular, from the nucleus interfascicularis hypoglossi and the nucleus paragigantocellularis lateralis, projects to the intermediolateral (IML) column of the spinal cord (250 ). Fibers and terminals of this tract are closely apposed to preganglionic sympathetic neurons. TRH fibers and preganglionic sympathetic neurons are also found around the central canal and in the intermediate gray matter of lamina VII (93 251 ). While some studies indicate that more than 90% of TRH immunoreactive neurons also stain positively for 5-HT, and 75% express immunoreactive SP (252 ), only 43% of IML TRH is ablated by 5-HT neurotoxins (250 ). Indeed, more than 90% depletion of 5-HT in the spinal cord reduces spinal cord TRH by only 66% (253 ). Thus, a sizable pool of TRH-containing neurons are not serotonergic.
a. TRH and gastrointestinal function. TRH inhibits food and water intake: TRH inhibits food and water intake, consistent with its high levels in the ventromedial hypothalamus (176 254 ), a center important to regulation of food intake (255 ), and its interaction with NPY and NE, both important to intake behavior (93 ). Systemic TRH reduces food intake less effectively than icv TRH, arguing for a central effect (256 ). Parenteral TRH can suppress eating without altering blood glucose levels (257 ) and without affecting TSH (256 ). icv TRH also reduces water intake (258 ), although others have reported icv TRH reduces food intake far more than water intake (259 ). icv TRH suppression of stress-induced eating is antagonized by D-ala-met-enkephalin (260 ), although DA transmission in the nigrostriatal pathway and lateral hypothalamus also affects stress-induced eating (261 ).
The hypothalamus serves as a principal brain substrate to coordinate hunger and satiety; it is generally held that the ventromedial hypothalamus serves to signal satiety, and the lateral hypothalamus, hunger (255 ). Injection of TRH into the ventromedial hypothalamus is most potent in producing adipsia and anorexia, and lateral hypothalamus injection is selectively potent for adipsia (262 ). Iontophoretic application of TRH onto ventromedial hypothalamic neurons results in facilitation of glucoreceptors, and hence, decreased feeding drive (263 ). However, others argue that the lateral, not ventromedial, hypothalamus is most critical for TRH-induced anorexia (226 ). TRH administration into the NAc is mildly anorexic (262 ). Systemic and icv CHP also suppress spontaneous food and water intake and stress-induced feeding, although it is less potent than TRH in feeding (264 ). In water-deprived rats, CHP is equipotent with TRH in reducing drinking (265 ). Thus, it cannot be ruled out that the TRH effects discussed above are actually the result of CHP as a TRH metabolite.
TRH enhances gastric acid secretion: Vagal preganglionic neurons arising from the rat DVC and nucleus ambiguus terminate in the gastrointestinal tract (266 267 ), in close proximity to nerve efferents of the greater curvature and pylorus of the stomach (268 ). These comprise the major medullary projections to the stomach. In the cat, retrograde tracing does not support a descending tract from the nucleus ambiguus, but the remainder of the DVC participates in the bulbogastric projection (268 ). The high concentration of TRH receptors in the DVC (269 270 ) in close proximity to NTS vagal afferents (266 ) is consistent with modulation by TRH afferents from the NTS and the medullary raphe nuclei. As well, peripheral signals of gastric distention and gastric secretion are carried by vagal afferents back to the medulla to activate DVC neurons. The codistribution of afferent and efferent pathways in the vagus provide a means for bulbogastric TRH neurons to modulate gastrointestinal responses to physiological signals, such as gastric distension, the cephalic phase of gastric acid secretion, etc. (271 272 ).
Central TRH is far more potent than iv TRH in inducing gastric acid secretion (18 273 ). icv Anti-TRH inhibits gastric acid secretion in pylorus-ligated (274 ) and cold-restrained (18 ) rats, supporting an endogenous role for the peptide. TRH stimulates gastric acid secretion independent of hypophysiotropic effects or effects on gastrin (275 276 ). Further, TRH injection into the DVC is 10 times more potent than icv TRH in stimulating gastric acid secretion (18 277 278 ). Bilateral DVC injection of anti-TRH antiserum significantly reduces gastric acid secretion in response to icv TRH (279 ), or chemical or electrical activation of medulla raphe pallidus (RPa) neurons (280 ), supporting the central role of the DVC in TRH effects on gastric acid secretion.
Atropine injection into the DMN does not completely block
TRH-stimulated gastric acid secretion (279 ), because other loci,
including the nucleus ambiguus, lateral hypothalamus, and the
ventromedial hypothalamus, can mediate TRH-induced gastric acid
secretion (18 281 ). TRH action also is partly mediated through
2-adrenergic receptors and enhanced sympathetic outflow that
modulates the vagus (275 277 282 ). By unknown mechanisms a number of
peptides in the DVN, including CRF, bombesin, calcitonin gene-related
peptide (CGRP), calcitonin (CT), endogenous opiates, and, curiously,
gastrin releasing peptide, inhibit TRH-induced gastric acid secretion
(18 ). CHP and TRH-OH have no such activity (277 283 ).
Kainic acid stimulation of afferent nucleus raphe obscurus (ROb) neurons mimics the induction of gastric acid secretion by TRH injection into the DVC (278 ). In addition, the caudal raphe nuclei-DVC pathway mediates cold-induced vagal stimulation of gastric acid secretion and erosion formation (284 ). Further, anti-TRH antisera injected into the DMN abolishes the ability of excitatory amino acid injection into the RPa to enhance indomethacin-induced gastric erosion formation (285 ). Thus, excitation of the raphe nuclei enhances DVC outflow, and one of the mediators of this effect is TRH. 5-HT (286 ) and SP (287 ) afferents from the raphe nuclei to the DVC modulate TRH effects.
icv TRH administration indirectly affects gastric acid secretion by increasing pepsin secretion and gastric mucosal blood flow and secretion (18 ). This effect is partly inhibited by DVC injection of anti-TRH antiserum, and surprisingly, is independent of increased gastric acid secretion (279 ). Thus, TRH may provide a means to regulate pepsinogen secretion without altering acid production. Intracerebral (ic) TRH-stimulated gastric mucosal blood flow is vagally mediated, via stimulation of an L-arginine-nitric oxide (NO) pathway independent of histamine H1 receptors or capsaicin-sensitive afferents (288 ). In addition, ic TRH enhances gastric secretion of 5-HT and 5-HT entry into the portal vasculature, an effect that again is vagally mediated (289 ).
TRH effects on gastrointestinal contractility and transit: ic, But not iv, TRH increases gastric contractions and gastric emptying in most species (290 291 292 ). Enhanced gastric motility is reproduced by direct infusion into the DMN but not the nucleus ambiguus (268 ), an effect completely blocked by vagotomy (293 ). Systemic morphine (294 ), or DVC injections of CRF (295 ), bombesin (296 ), and interleukin-1ß (IL-1ß) (297 ) inhibit the TRH effect. Since gastric contractility is inhibited by excitatory amino acid injection into the DVC (298 ), it is likely that TRH is inhibitory to DVC neurons controlling gastric contractility. Indeed, TRH injections onto NTS neurons reduce their spontaneous activity (299 ).
The ROb, RPa, and nucleus raphe magnus (RMg) provide afferents to the DVC (300 301 ). In the cat, the DMN receives its strongest inputs from the caudal RPa and ROb, where TRH neurons are enriched (301 ). Glutamate or electrical excitation of the caudal RPa and ROb, but not rostral RPa or RMg, results in enhanced gastric contraction. This effect is abolished by vagotomy and anti-TRH antibody injection into the DVC (302 ). Surprisingly, TRH stimulates ROb and RPa TRH afferents to the DVC. This effect is completely abolished by vagotomy or atropine into the ROb, markedly attenuated by atropine into the RPa (303 ), and antagonized by SP or VIP into the ROb (304 ). ic Antisense oligonucleotides to the TRH receptor block the increase in gastric motility seen with TRH injection into either the ROb or the downstream DVC, while glutamate excitation is unaffected. Thus, TRH activates both TRH and cholinergic afferents to the DVC, which in turn increase gastric motility (305 ). Finally, gastric contractility also is increased by TRH injection into the hypothalamic paraventricular nucleus or the central nucleus of the amygdala, an effect abolished by subdiaphragmatic vagotomy. However, unlike medullary effects, the frequency of gastric contractions after these injections is attenuated (306 307 ).
Motility in the proximal small intestine and ascending colon and cecum
is also mediated by a central effect on vagal outflow (291 308 ).
However, central depletion of brain catecholamines blocks the
contractile response in the duodenum, indicating a critical role for
catecholamines as well as ACh in TRH central regulation of bowel
motility, at least in some regions of the gut (309 ). Acceleration of
small intestinal transit appears to occur through a separate pathway
from that described for the stomach (310 ). TRH (
100 ng) increases
small intestine transit only when injected into the medial septum, or
lateral and anterior hypothalamus, in anesthetized rats. icv, But not
iv TRH, also reverses net water absorption in the jejunum and ileum, an
effect entirely abolished by vagotomy (311 ). Large colon transit in
rabbits is increased by iv or icv TRH (312 ) and is associated with
accumulations of fluid in the colon (289 ). TRH effects on colonic
transit are mediated by vagal and sacral cord parasympathetic outflow,
as well as serotonergic transmission (18 ). In humans, iv TRH retards
glucose and xylose absorption by the gut (313 ).
TRH effects on pancreas and liver.In normoglycemic rats, acute systemic TRH will induce hypoglycemia, with little effect on peripheral pancreatic hormones (314 ). Central TRH potently blocks epinephrine-induced hyperglycemia, presumably via combined parasympathetic/sympathetic induction of insulin secretion (315 ). Central TRH antagonizes hyperglycemia induced by treatments other than epinephrine, including central injections of CRF, ENK, and glucagon, as well as systemic 2-deoxyglucose, foot shock, immobilization, or endotoxin (315 ).
Pancreatic effects of TRH are most likely paracrine. TRH is synthesized in the insulin-producing ß-cells (316 ). In neonatal pancreas, TRH and insulin appear to be secreted via the same potassium-, cAMP-, and protein kinase C-responsive pathways (317 ); in adult pancreas, TRH secretion is inversely related to insulin secretion (318 ). While TRH does not affect insulin release (319 ), TRH and CHP inhibit 2-deoxyglucose-stimulated pancreatic secretion in a dose-dependent manner (320 ), and TRH enhances arginine-stimulated glucagon release (319 ). In isolated rat pancreas perfusates, anti-TRH antiserum reduces glucose- and arginine-stimulated glucagon secretion, and exogenous TRH enhances basal glucagon secretion if endogenous TRH is first cleared (318 ).
TRH also mediates central effects on pancreatic secretion. CSF injection of TRH, and microinjection into the DVC, stimulates exocrine pancreatic volume, protein, and bicarbonate secretion via vagal outflow (321 322 ). VIP is also an important nonmuscarinic mediator of TRH-stimulated pancreatic secretion, while CGRP, via noradrenergic mechanisms, opposes TRH pancreatic stimulation (321 ). Curiously, in isolated rat pancreatic acinar cells, TRH inhibits carbachol- and ceruletide-stimulated, but not OAG- or CCK-stimulated amylase secretion (320 323 ). Thus, certain pancreatic secretory pathways may show opposite peripheral and central effects by TRH. Finally, TRH has a trophic effect on the pancreas (324 ). Chronic administration of TRH for 10 days via gastric fistula significantly increases pancreatic weight, DNA content, and protein content, although enzyme concentrations are not proportionally elevated, so that their final concentrations are reduced.
ic RX77368, a TRH analog, stimulates hepatic DNA synthesis 24–72 h post injection in a dose-dependent manner (325 ). iv Administration is ineffective. The effect is abolished by hepatic vagotomy or atropine. Further, ic RX77368 enhances hepatic blood flow 15–90 min post injection. This regulation is abolished by hepatic vagotomy, atropine, indomethacin, or the NO synthesis inhibitor, NG-nitro-L-arginine methyl ester (326 ).
b. TRH and cardiovascular function: TRH reverses shock of varying etiologies in a number of animal species (327 ). However, the precise cardiovascular changes induced by TRH vary markedly with dosage, species, and experimental state, in particular, whether the animal is anesthetized or conscious, and normotensive vs. hypotensive. While TRH displays many cardiovascular effects, CHP and TRH-OH have minimal cardiovascular activity (328 ).
In rabbits, both anesthetized and conscious, iv TRH increases blood pressure and causes peripheral vasoconstriction (328 329 ). Effects in anesthetized rats are the same, except peripheral vasodilatation is seen (330 ); this peripheral vasodilatation is reversed by cholinergic blockade (328 ). Overall, TRH modulates blood pressure through combined parasympathetic and sympathetic effects (328 330 331 ). DA plays a lessor role (332 ), and naloxone is ineffective in altering these effects (333 ). As well, iv TRH in rats completely reverses systemic NT-induced hypotension and attenuates the central pressor effect of NT (334 ).
In anesthetized rabbits, but not conscious rabbits, cerebral
vasodilatation is induced by iv TRH, bringing cerebral blood flow back
to the level observed in conscious animals. Once cerebral blood flow is
normalized, TRH has little effect (329 ). Cerebral vasodilatation is not
the result of alterations in peripheral blood pressure, blood gases,
loss of autoregulation, change in cerebral metabolism, or change in
oxygen saturation gradients (328 ). Cerebral vasodilatation in
anesthetized rabbits is partially blocked by the 2-adrenergic
antagonist yohimbine, although yohimbine has no effect on TRH-induced
elevation of mean arterial pressure (335 ). Neither vagotomy nor
cholinergic blockade reduces these effects (328 ). Transection at the
mesencephalic pons abolishes TRH-induced cerebral vasodilatation
without affecting its systemic pressor effect (336 ). Cordotomy at the
C1 level abolishes the pressor effect of TRH, but not its effect on the
cerebral vasculature. Thus, peripheral pressor and vasoconstriction
effects are mediated more caudally than those increasing cerebral blood
flow.
icv TRH significantly elevates blood pressure and heart rate in anesthetized (337 338 ) and conscious (339 ) rats. In anesthetized rats, icv TRH increases blood flow to most organs, and this is abolished by bilateral vagotomy (328 ). In conscious rats, much more peripheral vasoconstriction is found (340 ), although this effect is reversed in hypovolemic states (341 ). icv TRH elevates plasma levels of NE and E independently of plasma renin activity or vasopressin. The vascular effects of icv TRH in rats are mimicked by TRH activation of sympathetic nuclei within the brain (342 ); intrathecal TRH induces its pressor response in rats and humans via increased sympathetic activity to the peripheral vasculature and adrenals (343 ), an effect mediated in rats by TRH receptors in the IML (344 ). Interestingly, sympathetic nerve responses in rats are partially attenuated by reduced thyroid activity, providing a second route for TRH regulation of cardiovascular function (345 ). Taken together, the data in rabbits and rats suggest that central TRH regulation of cardiac functions and organ blood flow distribution principally are mediated through the sympathetic nerves and the adrenal medulla (339 ).
Selective electrolytic lesions have identified the dorsal raphe nuclei as mediating the pressor response induced by icv TRH. Reductions in blood pressure induce prepro-TRH mRNA in the RPa (346 ). These changes, in turn, effect the descending bulbospinal tracts and ascending tracts to the PVN (342 ). TRH infusion directly into the dorsal raphe nuclei reproduces the pressor, tachycardic, and sympatho-excitatory effects of icv TRH, and these effects are blocked by ganglioplegia with pentolinium. Further, chemical lesioning by the 5-HT-preferring toxin 5,7-dihydroxytryptamine (5,7-DHT) obliterates TRH-induced tachycardia, while the NE/DA-selective toxin 6-hydroxydopamine does not. These data strongly implicate the descending bulbospinal pathway to the IML column to boutons on preganglionic sympathetic neurons as the neuroanatomical pathway mediating the peripheral cardiovascular effects of TRH (347 ).
TRH in experimental CVS disease: In experimental
models of anaphylactic shock, hemodynamic parameters in mice (348 ) and
guinea pigs (349 ) are improved by iv TRH, largely through elevation of
plasma E and NE (350 ). icv TRH mimics the protective effect of iv TRH
(351 ). TRH action in endotoxic shock also is mediated through the
sympathetic nervous system (350 ). In hemorrhagic shock, TRH elevates
blood pressure either by improvement in cardiac output, or by
increasing peripheral vascular resistance (352 ). However, after effects
of the latter may be harmful and could explain why TRH efficacy for
survival after hemorrhagic shock has varied among species (353 354 ).
Although principally mediated by catecholamines, icv TRH-induced
hemodynamic improvements in shock are partially blocked by 
-opiate
receptor antagonism (354 ) and vagotomy or atropine sulfate (355 ).
Few studies have examined the converse effect of shock on TRH. In rats, hemorrhagic shock decreases TRH in frontal cortex, septum, hippocampus, and hindbrain (356 ). TRH receptor binding is significantly decreased in septum and hindbrain. However, other experiments indicate that elevations of TRH in the medulla, midbrain, cortex, striatum, and cerebellum after hemorrhage are associated with better survival (357 ). Thus it appears that reductions in TRH neurotransmission in certain parts of the brain may contribute to the pathophysiology of shock. Further, TRH effects may be mediated, in part, by downstream effects on the thyroid axis (358 ) and plamsa vasopressin (204 ).
In spontaneously hypertensive rats (SHR), CSF TRH, prepro-TRH mRNA, TRH, and TRH receptor binding in the POA are all significantly elevated compared with Wistar-Kyoto (WKY) control rats (359 ). In SHR, elevated TRH receptor binding in striatum and hypothalamus correlates with the development of hypertension (360 ). Further, iv or icv TRH antiserum lowers arterial blood pressure in SHR rats, but not WKY rats. Chronic enalapril, a vasodilator, decreases blood pressure and reduces POA TRH levels, although another vasodilator, diltiazem, has no effect (359 ).
Animal models of stroke have provided a testbed for potential therapeutic application of TRH. In rats with middle cerebral artery (MCA) occlusion-induced infarcts, icv TRH given at 15 min and 24 h post surgery significantly improves survival, protects against ischemic damage, and reduces infarct size 10 days after surgery (361 ). Also in MCA infarcts, TRH increases blood flow to the infarct area (362 ). ip And oral TRH improve neurological deficits in MCA stroke (363 ) and improve recovery in cerebral damage induced by experimental hematoma (363 ). However, in gerbil and dog models of stroke, TRH fails to display efficacy (352 ).
c. TRH and respiration: icv TRH, in doses as low as 3 ng, significantly elevates blood pressure and heart rate in anesthetized rats, but a minimum 16 ng is required to raise the respiratory rate (364 ). Respiratory frequency is increased much more than tidal volume (16 ). Increases of respiratory frequency greater than tidal volume are also seen in conscious rats (365 ) and in rabbits (366 ). icv TRH raises respiratory rates in anesthetized rats partly through a DA D2 receptor mechanism (332 ). Further, icv TRH-induced respiratory stimulation is potentiated by pretreatment with naloxone, methylatropine, or low doses of GABA, but is unaffected by ß-adrenergic blockade and is independent of TSH (367 ). TRH antagonism of opiate-induced respiratory depression (368 ) is described in more detail below in Subsection 7.
In isolated brain stem-spinal cord preparations from rat neonates, bath application of TRH induces respiratory rhythmic neural discharges (369 ). icv TRH produces rhythmic bursting activity in neurons of the NTS (370 ), and local injection of TRH into the NTS induces tachypnea, although with a slower onset than seen with icv TRH (367 ). The respiratory effect occurs in the absence of any change in blood pressure or heart rate. Shortening of inspiratory times, but not tachypnea, results from TRH injections into the ROb (371 ). Tachypnea, without cardiovascular or locomotor effects, is seen with microinjection into the interpeduncular nucleus of the reticular activating system (372 ).
There is significant anatomical support for a role of TRH in respiratory control. Botzinger neurons in the medulla, which inhibit respiratory motoneurons, and the more caudal ventral respiratory group, form close associations with TRH-immunoreactive boutons (373 ). These connections appear to be functional, since TRH injected into the pre-Botzinger complex in neonatal rat medullary slices increases respiratory discharge frequency (369 ). TRH-immunoreactive boutons are also prominent near nucleus ambiguus motoneurons that display rhythmic fluctuations with phrenic nerve discharges (374 ).
Developmental studies support a direct effect for TRH on hypoglossal motoneurons in the caudal medulla. These neurons innervate tongue muscles critical for airway inspiratory control and display respiratory-related activity. TRH increases hypoglossal neuron discharge frequency, duration, and amplitude in neonatal mouse slices. In adult rat brainstem slice preparations, high doses of TRH depolarize hypoglossal neurons and reduce their firing threshold (375 ). More rostral to the hypoglossal nucleus, TRH enhances the responsiveness of inspiratory neurons in the ventrolateral medulla (376 ).
iv Or ic TRH significantly stimulates diaphragmatic activity and antagonizes morphine depression of diaphragmatic activity (377 ). TRH potentiates the excitability of diaphragmatic motor nerve terminals (378 ). Further, injections of TRH as low as 1 ng into the retrotrapezoid nucleus of anesthetized rats increase phrenic nerve firing frequency and amplitude. Only at 5 ng does TRH raise blood pressure. Both CHP and TRH-OH increase phrenic nerve firing frequency, but not amplitude, starting at 5 ng (379 ).
Preclinical literature describes the importance of thyroid hormones and steroid treatment on fetal lung development (380 ). However, in human trials TRH coadministration in steroid and surfactant therapy does not reduce newborn respiratory distress syndrome, chronic lung disease or associated neonatal complications, or death (381 382 ).
3. TRH and seizure modulation. TRH was first reported to potentiate the anticonvulsant actions of phenobarbital in mice in 1975 (383 ) and has since been shown to be anticonvulsant in multiple animal models of seizures (384 385 386 ). Despite extensive preclinical data indicating that TRH is likely to serve as an anticonvulsant, or a potentiator of known anticonvulsants, few large trials with this agent have been conducted. For intractable epilepsy, modest results have been achieved (387 388 ).
a. TRH and electroconvulsive seizures: One seizure paradigm commonly used to test the efficacy of anticonvulsant drugs is electroconvulsive seizure (ECS) (389 ). A single stage 5 seizure, induced after five ECS treatments given on alternate days (ECS x 5), elevates TRH in hippocampus, amygdala/pyriform cortex, whole cortex, and striatum, 48 h post seizure (17 ). Subconvulsive shocks given on alternate days for 5 days result in regulation only in the striatum. A separate study reported that ECS x 5 decreased NAc and lumbar spinal cord TRH 24 h after the last shock (390 ). In sum, ECS effects center on the hippocampus, amygdala and surrounding cortex, and the dorsal and ventral striatum.
The hippocampus most consistently demonstrates TRH induction after chronic ECS (391 ). A significant percentage of hippocampal TRH derives from extrinsic sources. ECS induction of prepro-TRH mRNA in the entorhinal cortex presumably leads, via the perforant pathway, to TRH increases in the dentate gyrus (17 ). In the hilar subregion of the hippocampus, which contains few or no perforant path terminals, fimbrae-fornix lesions do not block ECS induction of TRH, consistent with endogenous biosynthesis (392 ). There is no difference in basal TRH release from hippocampal slices dissected from ECS-treated vs. sham-treated animals. However, potassium-stimulated TRH release increases linearly 12, 24, and 48 h post seizure, and tissue content remains uniformly elevated throughout the postseizure period (393 ). Thus, in addition to elevations in steady state levels, there is a time-dependent shift of intracellular TRH into a potassium-responsive pool (394 ), which may enhance TRH release in response to afferent signaling.
Given the documented effect of TRH to increase cholinergic transmission in the hippocampus (395 ) and cortex (396 397 ), it is logical to speculate that TRH could be used to reverse ECS-induced neurochemical and behavioral deficits (398 399 ). Indeed, in rats post-ECS performance deficits are reversed by TRH.
b. TRH in kindled seizures: A second paradigm used to model epilepsy is kindling, where repeated electrical or chemical stimulation of limbic structures progressively lowers seizure threshold until an initial subthreshold stimulation becomes capable of reliably stimulating generalized (stage 5) seizures (400 ). In fully amygdala-kindled rats (five consecutive stage 5 seizures), prepro-TRH mRNA levels are significantly elevated 24 h following a stage 5 seizure in the dentate gyrus granular layer and the pyriform, entorhinal, and perirhinal cortices (401 402 ). More detailed time course studies in fully kindled animals report significant elevations of prepro-TRH mRNA in the dentate gyrus granular layer, several nuclei of the amygdala, and layers II and III of the pyriform and entorhinal cortices. Increased levels are detected 3 h post seizure, peaks occur at 6–12 h post seizure, and levels return to baseline 24–48 h post-seizure (246 392 ). The time course is similar in all regions, although slightly delayed in entorhinal cortex. The induction in prepro-TRH mRNA seen after full kindling may be observed unilaterally 24 h after partial kindled (stage 1–4) seizures, but is reliably and bilaterally observed only after fully kindled, generalized (stage 5) seizures (401 ).
Kindled seizures induce c-fos mRNA and Fos-related peptides, which in turn are postulated to induce prepro-TRH transcription via the AP-1 site in the prepro-TRH gene promoter (155 403 ). Fos-like immunoreactivity and prepro-TRH mRNA are extensively colocalized, in some cases in up to 70% of cells, in the pyriform cortex, entorhinal cortex, and dentate gyrus granule cells (404 ). A second potential transcriptional regulator of prepro-TRH gene expression, corticosterone, also is rapidly induced during kindled seizures (166 405 ).
Carbamazepine given contingently with kindling treatments attenuates prepro-TRH mRNA increases in the dentate gyrus, pyriform cortex, and ipsilateral entorhinal cortex. No attenuation is seen when carbemazepine is given noncontingently, i.e., after the kindling treatment (406 ). These results are intriguing, indicating that the carbamazepine-TRH interaction might be altered by behavioral or other nonpharmacological interventions.
Regulation of TRH peptide levels, measured 48 h post seizure, correlates with the progression of amygdala kindling. Partial kindling induces TRH in pyriform cortex, with greater regulation in stage 3–4 seizures than stage 2 seizures (17 407 ). In fully kindled rats, TRH is increased even more in pyriform cortex 48 h after stage 5 seizure. Increases are also seen in cingulate and frontal cortex, hippocampus, amygdala, and ventral striatum (408 409 ). Similarly, chemically kindled rats show TRH induction in hippocampus, amygdala, pyriform cortex, and anterior cortex 48 h after stage 5 seizures (17 ). Both kindling and TRH regulation after kindling persist for 6 months after kindling.
In fully kindled animals, all subregions of the hippocampus show reduced levels of TRH 1 h after seizure, consistent with synaptic release and rapid degradation. Levels rise to control levels at 6 h, are elevated at 24 and 48 h, and again return to control levels at 144 h after seizure. Increases are largest in the dorsal hippocampus, including the dentate gyrus, hilus, and CA3 region. It is hypothesized (17 409 ) that TRH elevations may mediate the postictal refractory period (410 ). In fully kindled rats, bilateral hippocampal TRH infusions decrease seizure after-discharge duration and overt seizure duration in a dose-dependent manner, consistent with an anticonvulsant action in the hippocampus (394 ).
TRH receptor binding in the dentate gyrus and perirhinal cortex is decreased in amygdala-kindled rats compared with sham-kindled animals (402 406 ). After a single stage 5 seizure in electrically kindled rats, hippocampal membrane TRH receptor binding is reduced 23–29%, and amygdaloid membrane binding is reduced by 21–22% (409 411 ). Curiously, in amygdala-kindled rats, dorsal striatal receptor binding is increased 24 h after seizure and persists significantly elevated at 21 days, although no regulation in striatal TRH is reported. Thus, only in certain brain regions do receptor adaptations appear to compensate for elevated TRH levels.
c. TRH in chemically induced seizures: TRH regulation in other types of seizures differs somewhat from that seen in ECS and kindling. Limbic seizures induced by systemic kainic acid substantially increase TRH in the posterior cortex and in the underlying dorsal and ventral hippocampus (412 ). Smaller increases are detected in the anterior cortex, amygdala/pyriform cortex, and corpus striatum. The increases in TRH are longer lasting than described for ECS, peaking at 2–4 days and resolving by 14 days, except for the dorsal hippocampus, where TRH elevations persist beyond 2 weeks. TRH is rapidly elevated in the septum, hippocampus, and thalamus/midbrain after a single pentylenetetrazol-induced seizure. Pyriform cortex was not tested (384 ). Pentylenetetrazol-induced tonic-clonic seizures in dogs increase TRH in pyriform and frontal cortex, hippocampus, and amygdala 48 h after seizure (413 ). Soman-induced seizures mediated by excessive cholinergic activity result in particularly high induction of TRH in frontal cortex, hippocampus, pyriform cortex, and entorhinal cortex, and lower induction in the amygdala (414 ).
d. Mechanisms of TRH anticonvulsant action: Irrespective of the precise mechanism of seizure induction, repeated seizures ultimately induce prepro-TRH mRNA in pyriform cortex, amygdala, and hippocampus. These areas contain well characterized TRH receptor binding (415 416 ). Further, the pyriform cortex region is a primary site for initiation of grand mal seizure activity, and the prepyriform region is exquisitely sensitive to direct application of chemical convulsants (417 ). One mechanism of TRH anticonvulsant effects may be through inhibition of L-glutamate excitation of neurons (418 419 ), especially neurons of the perforant pathways synapsing with dentate gyrus granule cells (420 ). This inhibition would increase after treatments such as kindling that elevate TRH, providing a feedback control for further sensitization in phenomena such as kindling and long-term potentiation. Knoblach and Kubek (407 ) suggest that TRH may be coreleased with excitatory neurotransmitters at these sites as a means to modulate neuronal response. If TRH is inhibitory to calcium influx secondary to reducing excitatory amino acid neurotransmission, it may also serve a neuroprotective role. Also it is speculated that hippocampal TRH may interact with coexpressed endogenous opioid peptides in seizure-involved pathways to modulate seizure activity (93 ).
4. TRH and motor control.
a. TRH stimulates ventral horn motoneurons: Bulbospinal
neurons that express prepro-TRH mRNA and iTRH descend from the
medullary raphe nuclei, and the parapyramidal and paraolivar regions,
to end in close apposition to motor neurons in lamina IX, and sparsely
in lamina VIII, of the ventral horn of the spinal cord (421 422 ). The
raphe projections provide dense innervation of spinal motoneurons and
are likely to enhance motor excitability, principally of proximal
muscle groups (423 424 ). In rat (425 ), rabbit (426 ), and human (427 )
spinal cord, the highest concentrations of TRH are found in the ventral
horn. TRH is present in large granules within terminal boutons that
synapse with dendrites, supporting a synaptic role (422 ). Some 60–90%
of these bulbospinal neurons coexpress 5-HT with TRH (428 429 ).
Immunocytochemical and ablation studies also support coexpression of
TRH with SP (422 ), although SP appears to have a more prominent role in
autonomic nervous system function than in voluntary motor control
(430 ). Surprisingly, in most species the ventral horn of the spinal
cord is not enriched in TRH receptor binding (431 ). However, the human
spinal cord displays elevated TRH receptor densities in laminae IX,
which contains -motoneurons (432 ).
An extensive literature describes TRH excitation of spinal cord ventral horn motoneurons (423 433 434 ) and hypoglossal motor neurons (435 ) by suppression of a distinct K+ current and development of an associated Ca++-sensitive inward current (436 ). In addition, TRH enhances motor neuron firing in response to excitatory amino acids (423 ), increases motoneuron recruitment by antidromic stimulation (437 ), and depolarizes ventral roots (438 ). The net effect of this excitation is augmentation of muscle tone, contractility, and spinal reflexivity (438 ). It should be noted that while TRH, SP, and 5-HT each can enhance excitatory amino acid activation of motoneurons, TRH excitation of ventral horn motoneurons is slower and less reliable than that observed with application of 5-HT or SP (423 439 ). Neither TRH-OH nor CHP have any demonstrated effect on spinal motoneurons, so motoneuron effects are presumed to be direct actions of TRH (434 436 ).
Denervation of the plantar foot muscles by botulinum toxin injection
reveals reinnervation deficiencies in adult rats that have undergone
ablation of the descending bulbospinal 5-HT/TRH neurons (440 ). However,
gross motor performance, muscle cell count, electrophysiological
properties, or -motoneuron counts are not made abnormal by this
ablation, arguing for an insignificant role for TRH in adult animals.
White et al. (439 ) argue that TRH function is more
significant in developing animals or on damaged motoneurons. TRH
trophic effects (441 ) and enhancement of contractility (442 ) are
demonstrated best in embryonic/neonatal preparations. More importantly,
TRH-induced depolarization shows significantly less tachyphylaxis in
isolated neonatal rat spinal cord preparations than in adult
preparations (443 ). NE inputs in the ventral horn enhance both
microiontophoretically applied TRH-induced excitation of motoneurons
(444 ) and behavioral excitation elicited by intrathecal TRH (426 ).
Thus, TRH may function only under certain physiological states,
such as stress or healing, that were not well tested in previous
paradigms.
In preclinical studies, TRH displays significant beneficial effects in the Rolling mouse Nagoya (RMN), a mouse model of ataxia (445 446 ). TRH is also effective in ameliorating ataxia induced by 3-acetylpyridine degeneration of the inferior olive in rats (446 ) and by cytosine arabinoside treatment in mice (447 ). Immobility and fall index scores for other ataxic mice models, including staggerer, reeler, weaver, and mice with Purkinje cell degeneration, are reduced by TRH (448 449 ). The NMDA antagonist MK-801 completely blocks TRH improvement of ataxia induced by 3-acetylpyridine, providing one of the few clues as to how TRH might mediate its antiataxic effects (450 ).
Several clinical studies support potential utility for TRH in the treatment of inherited ataxias such as spinocerebellar degeneration. Several studies demonstrate improvements by TRH of motor, occulomotor, and electrophysiological abnormalities in inhereted ataxias such as spinocerebellar degeneration (451 452 ). For the motor neuron disease amyotrophic lateral sclerosis (ALS), studies using higher doses of TRH demonstrate some transient imporvement in symptoms, particularly speech, swallowing, and respiratory function. Unfortunately, longer term results with systemic or intrathecal TRH (453 454 ) indicate that TRH fails to slow progressive motor neuron loss and provides only temporary symptomatic relief.
b. TRH promotes recovery in experimental spinal cord and brain injury: TRH accumulates superior to the site of traumatic or ischemic spinal cord injuries (455 456 ). In both cervical and lumbar injuries, TRH elevations are accompanied by modest but detectable down-regulation of TRH receptor binding proximal to the injury, in lamina X and the ventral horn gray matter, but not in the dorsal gray (457 ). The decrease in TRH receptor binding is detectable 48 h after injury and recovers by 3 weeks. Recovery of TRH receptor binding parallels the functional neurological recovery that occurs late after CNS injury (458 ). In mouse models of spinal cord motoneuron degeneration, TRH metabolism and levels are increased at sites of damage (459 460 ).
Whether increased TRH is a neuroprotective adaptation or a damaging mediator of the disease process is unknown. However, it appears likely that TRH is beneficial. TRH is superior to naloxone or high-dose steroid treatment in promoting recovery in experimental CNS trauma models (461 ). In most studies, sc or iv TRH improves electrophysiological recovery of damaged spinal cord tissue (462 463 ). Improvement may result from trophic effects of TRH on spinal motoneurons (204 464 465 ), from the ability of TRH to increase spinal cord blood flow (466 467 ), or from the ability to reduce edema (468 ) at the site of spinal cord injury. TRH efficacy in spinal trauma recovery requires continuous intrathecal infusion of native TRH, presumably due to its short half-life. Further, while TRH and all its analogs activate spinal motoneurons (434 ), only analogs that preserve the C terminus of native TRH are beneficial in spinal trauma, indicating the healing and activating effects of TRH are separately mediated (469 ).
TRH and its analogs are also efficacious in promoting recovery in animal models of head injury. Cats with brain stem compression injury show improved neurological and EEG parameters (470 ), and mice with head impact injuries demonstrate less behavioral disturbance (471 ), after TRH treatment. In rats with fluid percussion-induced brain injury, the TRH analog NS-3 given 30 min after injury improves survival, neurological parameters, and motor function at 24 h post injury, and improvements persist for at least 4 weeks (472 ). There is significant interest in TRH as a treatment for human traumatic brain injury and spinal cord injury, either to prevent damage progression or to speed neuronal recovery (473 ). However, TRH effects on head trauma, or recovery in spinal cord trauma, have been modest at best (474 ).
5. TRH and antinociception. The presence of TRH and TRH receptors in the midbrain PAG, the raphe nuclei, and, to a limited extent, in the dorsal horn of the spinal cord is highly suggestive of a role in pain modulation (93 ). Antinociception to chemical stimuli has been demonstrated in mice with iv and sc TRH (475 ) and for icv TRH against visceral chemical and mechanical pain (476 ). In the rat, icv TRH increases reaction latencies to visceral acetic acid (477 ). TRH displays potencies comparable to morphine in some of these studies. Thermal analgesia with icv TRH has been demonstrated in the hot-plate test (476 ) and less frequently in the tail-flick test (477 ). icv TRH also potentiates stress-induced analgesia, including foot shock-induced analgesia and swim-induced analgesia (478 ). In general, the antinociceptive effects of TRH are short lived, typically lasting for less than 15 min (479 ).
icv CHP in mice is variably reported to induce antinociception to mechanical, thermal, and chemical stimuli (480 ), or to mechanical stimuli but not chemical stimuli (481 ). This effect is significantly less potent, but longer lasting, than that of TRH and is antagonized to a greater extent by sc naloxone than is TRH. TRH-OH also is reported to be antinociceptive to mechanical and chemical stimuli in mice (481 ). Thus, it is likely that both TRH and its metabolites are active in TRH-induced antinoception in its various forms.
The PAG is critical for integration of pain perception and the
behavioral response to pain (482 ). The PAG expresses moderately high
levels of prepro-TRH mRNA (28 249 ), and significant amounts of iTRH
are detected in cell perikarya and neuronal fibers of the PAG (93 421 483 ), and TRH receptor binding is detected at moderate levels (484 )
throughout the PAG. The neuroanatomical distribution of TRH and TRH
receptor binding in all regions of the PAG implicates it in both
opiate-dependent and opiate-independent pain mechanisms. The
dorsolateral PAG mediates antinociception that is not modified by
naloxone and is presumed to be opiate independent, while the
ventrolateral PAG is more commonly associated with opioid-dependent
antinociception (482 ). This may relate to the finding that TRH
infusions into the ventral PAG decrease cold-water swim-induced
antinociception in the rat in a dose-dependent manner, while dorsal PAG
infusion has the opposite effect (485 ). Curiously, TRH in both
placements antagonizes morphine antinociception and reduces
swim-induced hypothermia and morphine-induced hyperthermia, while the
converse, analgesia produced by TRH infusion into the ventral PAG, is
blocked by opiate antagonists (486 487 ). TRH has no demonstrable
effect on µ-, -, or 
-opiate receptor binding, or receptor
occupancy, arguing against direct TRH regulation of opioid peptide
release. It is postulated that TRH activates one or more types of
inhibitory interneuron, which in turn reduce excitation of pain-excited
opiate-responsive neurons in the PAG.
Complete blockade of TRH-induced antinociception by opiate antagonists occurs in some PAG infusion paradigms, but not with systemic TRH, which presumably activates nonopiate pain pathways in which TRH functions. Separate studies indicate that TRH injections into the NAc (488 ) and amygdala (477 ) produce antinociception. TRH antagonizes NT-induced antinociception (489 ) and both TRH and CHP antagonize THC-induced antinociception (194 ).
Clearly then, the relationship of TRH antinociception to endogenous and
exogenous opiate systems is complex. As opposed to opiate-induced
hypothermia, locomotor depression, catalepsy, and respiratory
depression, systemic TRH does not antagonize acute opiate-induced
analgesia (490 ), although hyperalgesia induced by high-dose naloxone is
antagonized by TRH (491 ). In contrast to the case with drug-naive mice,
TRH does not display antinociceptive effects to chemical stimuli in
morphine-tolerant mice (487 ). In particular, Bhargava and colleagues
(490 ) argue that TRH interacts preferentially with the -opiate
system in mice. Other studies (487 ) indicate that tolerance either to
morphine (µ- and -specific) or ethylketocyclazocine
methanesulfonate (-preferring) equally antagonize TRH
antinociception. Most studies indicate that TRH-opiate interactions do
not occur via TRH modulation of opiate receptor binding, or through
TRH-stimulated release of endogenous opioids. Rather, TRH acts through
intermediary systems that, in turn, modulate opiate-mediated pain
transmission, and perhaps vice versa. In the spinal cord, likely
intermediary systems involve 5-HT and SP (492 ).
PAG-mediated antinociception is mediated via outputs to the ROb, RPa, and RMg (482 ), nuclei that express high levels of prepro-TRH mRNA (249 ). Further, antinociception by excitation of the nucleus reticularis paragiganticellularis (RPGi) is mediated, in large part, through reciprocal connectivity to the RMg (493 ). TRH into either the RMg or RPGi is antinociceptive (477 ). icv TRH also inhibits pain-excited neurons in the mesencephalic reticular formation (MAF) (494 ). The MAF is believed to form a second integration circuit for autonomic response to painful stimuli (495 ).
In addition to the descending bulbospinal pathway, TRH is also found in an intrinsic system of cell bodies in laminae II and the lamina II/III border of the dorsal horn (251 496 ). Dorsal horn iTRH is not depleted by 5-HT neurotoxins, unlike ventral horn TRH (250 ). Laminae II also exhibit a high level of TRH binding in many species (250 432 484 ). Intrathecal TRH antagonizes morphine analgesia in the tail-flick test at most doses, indicating that the net action of TRH within the dorsal horn is to enhance transmission of nociceptive somatosensory information (492 ). iv TRH is reported to facilitate nociceptive transmission through the dorsal horn via positive modulation of NMDA receptor-mediated transmission (497 ).
6. TRH in thermoregulation. TRH plays a prominent role in integrating a number of thermogenic responses to cold (509 ). CNS injection of TRH elevates body temperature (498 ), and TRH antagonizes the hypothermic effects of a number of agents, including barbiturates, ethanol, chlorpromazine, bombesin, NT, and ß-endorphin (204 ). Systemic TRH antagonizes morphine-induced hypothermia, while having little effect on analgesia (499 ). icv Anti-TRH antibodies in rats produce hypothermia, supporting an endogenous role for TRH in body temperature elevation (500 ). A principal site of TRH thermoregulation is the anterior hypothalamic POA (498 ). TRH into the POA inhibits heat-sensitive neurons and activates cold-sensitive neurons (501 ), which results in increased body temperature through peripheral vasoconstriction, increased metabolic heat production, and shivering (502 ). These effects require intact catecholamine neurotransmission (502 ).
Ablation of the POA does not eliminate TRH antagonism of pentobarbital-induced hypothermia, indicating that sites other than the POA can mediate TRH thermoregulation (503 ). Cold exposure elevates prepro-TRH mRNA levels (504 ) and TRH secretion (505 ) in the PVN. These changes elevate thyroid hormones and increase heat generation in brown adipose tissue (506 ). Systemic TRH has similar effects, e.g., systemic TRH improves thermoregulation in neonatal lambs through increased fat oxidation (507 ). We note that cold exposure elevates TSH levels before TRH levels, probably because SRIF, which tonically inhibits TSH secretion, is rapidly down-regulated in the PVN by cold (508 ).
Cold-induced increases in prepro-TRH mRNA also are seen in the DMN
(509 ) and caudal raphe nuclei (510 ). The raphe nuclei, which receive
sensory information from the skin, project to spinal cord preganglionic
sympathetic neurons. Further, the raphe nuclei provide TRH afferents to
the NTS (511 ), which, in turn, projects to the DMN, and then the spinal
cord preganglionic neurons. 5-HT projections from the NTS to the
PVN provide feedback regulation to this stimulation (512 ). Excitation
of spinal cord preganglionic sympathetic neurons results in
postganglionic NE release and increased facultative thermogenesis via
ß- and 1-adrenoreceptors on brown adipocytes (513 ). Additional
stimulation comes from direct projections of TRH neurons from the
dorsal cap of the PVN to preganglionic sympathetic neurons in the
thoracic and sacral spinal cord (514 ).
In mice, icv TRH and CHP antagonize ip THC-induced hypothermia (517 ). However, CHP elicits hypothermia when injected into the cerebral ventricles, an effect antagonized by TRH (515 ). The POA is believed to be the sole site mediating this action of CHP (516 ). Thus, the hypothermic response to icv TRH seen under certain conditions such as warm environments may result from TRH catabolism to CHP (501 ).
7. TRH and drugs of abuse. The psychomotor theory of addiction states that a common biological mechanism mediates both positive reinforcement and motor activation by drugs of abuse (518 ). Given the ergotrophic effect of TRH as a locomotor activator, interaction between TRH and drugs of abuse is likely. While the main focus of drug abuse research has centered on two principal loci in the mesolimbic DA pathway, the NAc and the VTA (519 ), a more extensive network, the "limbic-motor circuit," with inputs to the NAc coming from many limbic areas and outputs going to both limbic areas and motor areas, is now appreciated (520 ).
In rats, systemic TRH and intra-NAc TRH mimic cocaine by inducing locomotor activation via release of DA and 5-HT in the NAc and striatum (222 521 522 ). Conversely, DA D2 agonists increase TRH release from striatal and NAc slices (523 ). Downstream components of the HPT axis act to reinforce psychostimulant effects (524 ). TRH neurons of the PVN receive DA and NE inputs that are regulated by cocaine (124 ). Further, DA and cocaine both activate the HPA axis; stress (or CRF) and cocaine elevate NAc and medial prefrontal cortex DA and cause similar neuronal adaptions (525 526 ).
Few studies have directly examined the link between TRH and psychostimulants. Acute amphetamine lowers TRH in the caudate, NAc, and lateral septum (527 528 ). Over time, TRH levels show some adaptation to chronic amphetamine, and TRH receptor binding increases (528 ). Acute cocaine significantly decreases prepro-TRH mRNA levels in the amygdala and hippocampus, 45 min after injection (529 ). Chronic cocaine regulates prepro-TRH mRNA in the NAc, amygdala, hippocampus, and hypothalamus. Prepro-TRH mRNA regulation is strongly dependent on the length of time after cocaine cessation and persists beyond 72 h post injection in the amygdala.
The role of TRH in morphine actions also is not well understood. TRH
antagonizes a number of morphine’s depressant effects, including
sedation, hypothermia, and catalepsy (530 531 ). Chronic TRH inhibits
the development of tolerance to opiate-induced hypothermia and
catalepsy (499 532 ). - And -opiate receptor activation reduces
TRH receptor binding, but TRH does not effect opiate binding (533 ).
More directly, morphine reduces cortical and diencephalic TRH
concentrations (534 ). Within the HPT axis, a clearer relationship
between TRH and opioids exists. Morphine and opioid peptides reduce
plasma TSH (535 ) and blunt cold-induced TSH release (536 ). Further,
exogenous morphine at pharmacological doses inhibits TRH release via
opiate receptors on TRH-secreting hypothalamic nerve terminals (509 ).
TRH is more strongly implicated in opiate withdrawal. While cessation of chronic cocaine use induces relatively little physical withdrawal (537 ), chronic morphine results in the development of physical dependence and the aversive state of withdrawal upon cessation of morphine use (537 ). The expression of the physical symptoms of withdrawal is mediated principally by the LC and PAG (538 ). A large body of data indicates that the intrinsic NE neurons of the LC undergo an up-regulation of their cAMP second messenger system in response to chronic morphine. When unopposed during morphine withdrawal, the up-regulated cAMP system drives increased firing by LC neurons (539 ). Extrinsic excitatory amino acid inputs from the nucleus paragiganticellularis lateralis contribute an additional 50% to LC neuronal excitability during withdrawal (540 541 ). The medial hypothalamus, medial thalamus, amygdala, frontal cortex, hippocampus, and RMg also are implicated in withdrawal (542 543 ).
The PAG expresses high levels of prepro-TRH mRNA (28 ). Mature TRH (483 ) and TRH receptor binding (484 ) are present in moderate levels throughout the PAG. Prepro-TRH mRNA is strongly induced in the PAG during opiate withdrawal (28 ). Fos-like immunoreactivity is greatly increased in the ventrolateral PAG during withdrawal (544 ) and may mediate induction of prepro-TRH mRNA (401 ).
While TRH levels in the PAG remain unchanged during opiate withdrawal (154 ), it has been found that ic TRH prevents withdrawal-induced hypothermia and decreases jumping during withdrawal in morphine-dependent mice (545 ). In contrast, ic TRH induces wet-dog shakes in normal animals, arguing that it augments withdrawal-like symptomatology (546 ). Opiate withdrawal increases TRH in the lateral hypothalamus, suggesting this region may also play a physiological role in opiate withdrawal (154 ). Thus, much remains to be learned about TRH effects on opiate withdrawal at sites other than the PAG.
The TRH analog TA0910 reduces alcohol-intake in alcohol-preferring rats (547 ) and in primates (548 ) in a dose-dependent manner. This appears mediated by DA D2 receptors (549 ). Behavioral reward to alcohol, as measured by punished responding rates, is enhanced by iv TRH (550 ). Alcohol alters TRH receptor binding (551 ). In long-sleep (LS) and short-sleep (SS) mice that display differential CNS sensitivity to ethanol, SS mice have greater sensitivity to TRH than LS mice during postnatal days 8–14 (552 ). It is hypothesized that a TRH receptor-mediated alteration results in enhanced development of the thyroid gland in SS mice. Alcohol-preferring rats compared with nonpreferring rats have significantly lower TRH levels in the medial and lateral septum. Upon exposure to alcohol, preferring rats are able to right themselves earlier than nonpreferring rats, and this correlates with elevations of medial septal TRH. However, these findings may be nonspecifically related to the analeptic action of TRH in this region (202 ). Of unclear relationship to alcohol preference, chronic ethanol in rats partially "uncouples" PVN TRH expression from peripheral thyroid response (553 554 ) and, like opiates, ethanol blocks the TSH response to cold (553 554 ).
A wide range of other addictive substances alter TRH receptor binding, including THC and chlordiazepoxide (551 555 ). Furthermore, as described above, behavioral reward to pentobarbital and chlordiazepoxide, as well as alcohol, as measured by punished responding rates, is enhanced up to 3.5-fold by iv TRH (550 ). Goeders et al. (556 ) recently presented evidence that levels of benzodiazepine receptor binding are affected by the development of either behavioral tolerance or sensitization to cocaine. Conversely, benzodiazepines specifically attenuate cocaine self-administration (557 ). Since a number of benzodiazepines displace 3H-methyl-TRH from TRH receptors (558 559 ), this provides another suggestive link between TRH and drugs of abuse.
In summary, there are multiple pathways through which TRH affects virtually all classes of abused drugs. Understanding these interactions is likely to advance our understanding of addiction in general. More importantly, this understanding may provide new pharmacological approaches for the clinical treatment of substance abuse.
8. TRH outside of the CNS. TRH is phylogenetically old, present in invertebrates such as the lamprey that lack TSH or the snail that lacks a pituitary (560 561 ). It appears that mammalian endocrine functions for TRH have been "co-opted" (562 ) for a peptide already functioning in more basic ways. TRH is detected in many nonneural vertebrate tissues, although its functions in these tissues are not well understood. In most cases, peripheral TRH is not regulated coordinately with the HPT axis (563 ).
Prominent among the TRH-containing tissues are the gastrointestinal organs, including the stomach, duodenum, small intestine, colon, and rectum (562 ), where TRH may have peripheral effects to modulate gastrointestinal contractility. The pancreas is a rich source of TRH. Indeed, in neonates a significant portion of circulating TRH is derived from pancreas (561 ). Prepro-TRH mRNA is expressed in ß-cells of the pancreatic islet (564 ), and TRH and somatostatin have opposing paracrine effects on glucagon secretion (565 ).
TRH is present at high levels in the genitourinary system including the ventral prostate, Leydig cells of the testes, the epididymis, and seminal vesicles (566 567 ). Interestingly, propylthiouracil-induced hypothyroidism increases TRH in prostate and testis but reduces TRH in epididymis (566 568 ). TRH receptor mRNA is expressed in the ovary and uterus (569 ), and TRH is present in placenta, amniotic fluid, and breast milk (177 570 ). Again, reproductive TRH may act as a paracrine regulator (571 ).
TRH is also present in retina (572 573 ), where its levels are light entrained (572 ). icv TRH raises intraocular pressure and induces marked mydriasis via combined sympathetic and parasympathetic effects (574 ). TRH receptor mRNA is found in human peripheral blood monocytes (PBMCs) and rat splenocytes (575 ). It affects secretion of TSH and immunoglobulins from blood elements (576 577 ) and may be a trophic factor for certain blood elements (578 ). In the heart, TRH is expressed and has direct ionotopic effects (579 580 ).
|
IV. Function of non-TRH pro-TRH-Derived Peptides |
|---|
|
TopAbstractI. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
). Thus, while in situ immunohistochemical studies
remain of value, one must be aware that they are beset with
difficulties in variable immunoreactivity in different loci of the
brain, and they lack precise definition of the immunoreactive peptide
moieties that are being detected. In comparison to the known roles of TRH reviewed in Section III.B, there is precious little known about the biological activities of the other pro-TRH-derived peptides. In the hypothalamus and testis, both these non-TRH peptides and TRH are regulated within HPT parameters, while in other tissues, both TRH and non-TRH peptides are not (563 583 ). Both TRH-Gly and prepro-TRH178–199 are regulated by dexamethasone in the hypothalamus, but not in cerebellum, brain stem, retina, and stomach (583 ). Thus, clues to the roles of pro-TRH-derived peptides other than TRH must come from an examination of their regional distribution or evidence of regulation under specific physiological or pathological conditions.
A. prepro-TRH160–169 (pST10)
Prepro-TRH160–169 is the best characterized of the
non-TRH pro-TRH-derived peptides. It is released from rat hypothalamic
slices and the ME, thus making a hypophysiotropic role likely (20 ).
Prepro-TRH160–169 (also known as Ps4 and TRH-potentiating
peptide) enhances TRH-stimulated TSH release from the anterior
pituitary and stimulates TSHß gene promoter activity (21 ). Thus, it
acts in an opposite manner to feedback by T3, which
decreases TSH secretion (172 ) and inhibits TSH subunit gene expression
(584 ). The peptide has been isolated from bovine hypothalamus and its
amino acid sequence confirmed by Edman degradation (585 ).
Prepro-TRH160–169 is also unique in that a receptor for
this peptide has been characterized (see Section VI) (586 ).
Prepro-TRH160–169 receptor binding is developmentally
regulated, with an increase from birth to weaning, and then a gradual
decline to adult levels at postnatal day 60 (587 ).
Within the CNS, prepro-TRH160–169 is most enriched in hypothalamus, with lesser amounts in the spinal cord and olfactory bulb. The pituitary and striatum contain moderate levels. Its receptor binding is highest in the pituitary, hypothalamus, spinal cord, olfactory bulb, and hippocampus (588 ). Prepro-TRH160–169 is rich in rat testis, but trace levels are detected in urinary bladder, vas deferens, and heart. Receptor binding is high in urinary bladder and vas deferens, heart, and testis (588 ). In adrenal extracts, RIA detects lesser amounts of prepro-TRH160–169, which has been confirmed further in this tissue by chromatographic fractionation. The pancreas contains prepro-TRH160–169 within ß-cell secretory granules (318 ). Finally, prepro-TRH160–169 function is not restricted to a hypophysiotropic one. In parallel fashion to its role in the pituitary, prepro-TRH160–169 does not influence basal gastric acid secretion when injected into the DMN, but does potentiate the ability of TRH to do so. Prepro-TRH178–199 has no effect when coinjected with TRH into the DMN, and neither does prepro-TRH160–169 when coinjected with TRH into the nucleus ambiguus (22 ).
ECS elevates prepro-TRH160–169 in hippocampus, amygdala, pyriform cortex, and anterior cortex, but not corpus striatum, motor cortex, LC, or ventrolateral medulla (589 ). In these studies, elevations of prepro-TRH160–169 correlated with elevations in TRH-Gly and TRH in the hippocampus, amygdala, and pyriform cortex. Thus, prepro-TRH160–169 may share a role with TRH in seizure modulation (17 ). More recently it has been reported that prepro-TRH160–169 levels in the hippocampus and amygdala correlate with immobility times in the Porsolt forced-swim test, leading to speculation that the peptide may act independently or in concert with TRH to affect mood, learning, or memory (589 ). A review of the distribution and postulated functions of this peptide recently has been presented (590 ).
B. prepro-TRH178–199 (pFE22)
The second most studied non-TRH pro-TRH-derived peptide,
prepro-TRH178–199, is also released from rat hypothalamic
slices and the ME (20 591 ) and is localized in dense-core granules of
PVN neurons that project down to the ME. prepro-TRH178–199
has been reported to be a corticotropin-inhibiting factor, acting to
reduce POMC mRNA and inhibit ACTH release (592 593 ). These data fit
the clinical phenomenon of increased TSH in isolated ACTH deficiency
(594 ). Further, 2-adrenergic inputs stimulate release of both
pro-TRH products and ACTH (595 ), and thus,
prepro-TRH178–199 and ACTH may be coreleased into the ME,
allowing the pro-TRH product to modulate ACTH release. Despite this
logic, the relationship between the HPA and HPT axes remains
incompletely understood, and other investigators have been unable to
reproduce corticotropin inhibition by prepro-TRH178–199
(25 ). The pancreas produces prepro-TRH178–199, within the
ß-cell secretory granules (316 ), suggesting a potential involvement
for this peptide in the regulation of glucose metabolism. Finally,
prepro-TRH178–199 acts as a PRL secretagogue in primary
pituitary cultures. However, as described in the next section, it may
be broken down to peptide products that also induce PRL secretion, so a
direct effect remains to be proven. Prepro-TRH178–199
levels rise during the early phases of suckling in rat pups (91 ).
Again, a precise role for the peptide in lactation, suckling, etc.
remains to be determined.
C. prepro-TRH178–185 and
prepro-TRH186–199 (pFQ7 and pSE14)
Suckling increases prepro-TRH mRNA in PVN and markedly increases
TRH release during the first period of lactation (596 ). In experiments
where we coexpressed rat prepro-TRH cDNA with PC1, PC2, and 7B2 in
GH4C1 cells (Section II.D), we
detected two novel peptides, prepro-TRH178–185
(pFQ7) and prepro-TRH186–199
(pSE14). These peptides are generated by cleavage of
prepro-TRH178–199 (pFE22) by PC2 (see
Section II.D and Figs. 1 and 6). We subsequently determined
that these peptides are present in the rat PVN. In examining whether
pro-TRH processing is altered by suckling, we found that in addition to
prepro-TRH178–199, the products
prepro-TRH178–185 and prepro-TRH186–199 also
increase release of PRL from primary pituitary cultures.
Prepro-TRH178–185 was the most active PRL secretagogue. In
suckling experiments, where pups were separated from their mothers for
6 h and then reunited for 45 min to suckle, a 5-fold increase in
PVN prepro-TRH178–199 and prepro-TRH186–199
and a 6-fold increase in serum PRL were observed over nonsuckling
controls. While these data implicate these novel peptides in suckling,
or a response to PRL, further experiments are required to rule out
nonspecific effects of stress (91 ).
D. prepro-TRH53–74 (pFT22)
Prepro-TRH53–74 displays a unique localization in the
rostral two-thirds of the ventrolateral PAG (95 597 ). Electrical
stimulation of, or injection of excitatory amino acids into, this
region of the PAG produces analgesia (598 599 ), and this region of the
PAG is most sensitive to production of morphine-induced antinociception
(600 ). Although the lateral PAG is more commonly associated with
non-opioid-mediated antinociception (482 ), there is limited evidence
that nonopioid antinociception is also mediated within the
ventrolateral PAG (601 ). Thus, prepro-TRH53–74 may be a
candidate molecule to mediate nonopiate pain perception in this region
or play a modulatory role in opiate-mediated pain mechanisms.
As described in Section III.B.5, the expression of prepro-TRH mRNA in the reticular nucleus of the thalamus, the PAG, the raphe nuclei, and, to a limited extent, in the dorsal horn of the spinal cord is highly suggestive of a role in pain modulation (93 ). Prepro-TRH53–74 similarly is prominent in the ventrolateral PAG, RMg, and thalamic reticular nucleus, suggesting a role in opiate-dependent pain perception (95 597 ). The reticular nucleus serves in gating of peripheral somatic sensory information from the dorsal thalamus to the cortex (602 ). Opiate withdrawal induces prepro-TRH53–74 in the rat PAG, while TRH levels are unaltered (27 ), suggesting that the peptide may interact with opiate pain mechanisms in these pathways (406 597 ).
E. prepro-TRH83–106 (pEH24) and
prepro-TRH208–255
As described above (Section III.B.10), during opiate
withdrawal prepro-TRH mRNA is increased in the PAG (27 28 ). Peptide
analysis in the PAG demonstrates an accumulation of the N-terminal
peptides prepro-TRH53–74 (see subsection 7 above) and
prepro-TRH83–106, a reduction in the C-terminal peptide
prepro-TRH208–255, and no change in
prepro-TRH178–199 or TRH, in opiate-withdrawal
vs. control animals (154 ). Opiate withdrawal also increases
prepro-TRH83–106 in the lateral hypothalamus. We speculate
that during opiate withdrawal, pro-TRH processing may be altered in
several brain regions, resulting in increased levels of N-terminally
derived peptides (prepro-TRH53–74 and
prepro-TRH83–106) and decreased levels of some
C-terminally derived peptides (prepro-TRH208–255).
F. TRH-Gly
Immunoreactivity of the immediate precursor of TRH, TRH-Gly, is
widespread and detectable throughout the CNS in a similar distribution
to TRH and is also present in the prostate, serum, spleen, adrenals,
kidney, and gastrointestinal organs. Levels of TRH-Gly are up-regulated
by hypothyroidism and by thermal stress, following the pattern for TRH
itself (603 ). The pancreas contains TRH-Gly within ß-cell secretory
granules (603 ). The ratio of TRH/TRH-Gly is highest in pituitary and
hypothalamus and much lower elsewhere in neural tissue. TRH-Gly is much
better characterized than other pro-TRH-derived peptides discussed
above. TRH-Gly is increased in several limbic regions after chronic
ECS, including hippocampus and pyriform cortex (391 ). These authors
hypothesize that the increase correlates with effectiveness in
increasing swim times in the forced-swim test, and thus it may serve an
antidepressant and/or anticonvulsant function. TRH-Gly, independent of
conversion to TRH, stimulates gastric acid secretion in a
dose-dependent manner, although with a potency 100-fold less than TRH
(18 ). Interestingly, TRH-Gly can directly activate TRH receptors in
high concentrations (100 and 1000 nM). The IC50
of TRH-Gly for displacement of MeTRH (12 µM) is
significantly higher than the TRH receptor dissociation constant
(Kd) of 1.7 nM for MeTRH (604 ).
TRH in pathological conditions may induce secretion of GH (124 ). In normal controls TRH-Gly does not increase secretion of TSH, PRL, or GH (605 ). However, iv TRH-Gly can induce secretion of GH in patients with acromegaly and stimulate PRL and TSH release in women with anorexia nervosa. Preclinical studies indicate estrogen/progesterone treatment, as well as starvation, can enhance the ability of TRH-Gly to stimulate TSH and PRL release, although at a potency far below that for TRH (606 ). In most cases, TRH-Gly effects are difficult to separate from effects resulting from subsequent conversion to TRH. However, TRH-Gly may prove useful as a pharmacological challenge agent, even if an endogenous role in pituitary release is not confirmed.
|
V. Non-TRH pro-TRH-Derived Peptides Outside of the CNS |
|---|
|
TopAbstractI. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
In CA77 thyroid parafollicular cells, prepro-TRH mRNA, 7-kDa and 3-kDa species recognized by anti-prepro-TRH53–74, and TRH are detected (86 ). In thyroid tissue, prepro-TRH mRNA, the 7-kDa and 3-kDa species of immunoreactive prepro-TRH53–74, and immunoreactive prepro-TRH115–151 are found. Part of this immunoreactivity comigrates with synthetic prepro-TRH115–151 standard on gel filtration and reversed-phase HPLC. In addition, immunohistochemical studies localize prepro-TRH53–74 to parafollicular cells in thyroid tissue (610 ). Thus, within the thyroid, significant biological functions for non-TRH peptides remain to be deciphered.
TRH-Gly is present in high levels in the ventral prostate of the rat, as well as in the testis, epididymis, and seminal vesicles (566 ). Both TRH and TRH-Gly in the epididymis and prostate are regulated along with the HPT axis under certain conditions (566 ). Prepro-TRH160–169 is also detected in high levels in the testis. Further, prepro-TRH160–169 receptor binding is high in urogenital organs, second only to CNS tissues (586 ). Human placenta contains appreciable quantities of the human TRH progenitor octa-TRH and a larger non-TRH peptide, human prepro-TRH192–222 (611 ).
|
VI. TRH and Other pro-TRH-Derived Peptide Receptors |
|---|
|
TopAbstractI. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
TRH-R1 is highly conserved between species, e.g., at the nucleotide level the human receptor is 90.3% and 89.2% homologous to the mouse and rat receptors, respectively; the three receptors are approximately 95% conserved at the amino acid level. The receptors are members of the G protein-coupled receptor superfamily. Intracellular signal transduction is principally mediated by coupling to Gq and G11. Ligand binding results in activation of phosphoinositide-specific phospholipase C (PPI-PLC) (618 ), resulting in PIP2 hydrolysis, and subsequent production of inositol 1,4,5-triphosphosphate (IP3) and 1,2-diacylglycerol. This stimulates increased intracellular calcium, although the exact contributions of increased calcium influx vs. mobilization of intracellular stores is controversial (619 ). Downstream activation of protein kinase C (620 ), calcium/calmodulin-dependent protein kinase (621 622 ), and mitogen-activated protein kinase (623 624 ) also occurs. Under specific conditions in certain cell types, the TRH receptor also couples Gi2 and Gi3 and to a Gs-like protein that does not activate adenylate cyclase (625 ). These complexes are less well studied and represent the minority of TRH signal transduction.
Of particular interest to the potential clinical usefulness of TRH or TRH analogs is the phenomeneon of TRH receptor desensitization. Pituitary TRH receptors after several hours exposure to TRH display markedly reduced TSH, but not PRL release. The IP3 response to TRH displays homologous desensitization in as little as 10 sec of TRH exposure in transfected HEK 293 cells (626 ). This occurs by rapid uncoupling of the receptor and a decrease in PPI-PLC activity. In the same model system, intracellular calcium mobilization displays heterologous acute desensitization, with effects on other receptors whose signal transduction also depends on calcium elevation (627 ). While most G protein-coupled receptors undergo acute desensitization by phosphorylation (628 ), this has not been demonstrated for TRH receptors. Neither a specific protein kinase nor calcium concentration have been clearly implicated in TRH receptor desensitization (625 ). Because G11, Gq, and PPI-PLC have not been shown to be targets for desensitization, it still is believed some form of modification of the TRH receptor may explain acute desensitization. Not to be overlooked, if TRH exhibits slowed dissociation from the TRH-receptor complex, reactivation of the receptor cannot occur (629 ). Acute desensitization of the TRH receptor is also dependent on cell type, with pituitary cells displaying the most desensitization (630 ).
A second principal mechanism for TRH receptor desensitization is agonist-induced internalization (631 ). Up to 80% of TRH receptors are internalized by pituitary cells, with a half-time of 2–3 min (631 ). Thus, this mechanism is used by TRH receptors to a greater extent than for many other G protein-coupled receptors and may represent a significant mechanism for clearance of secreted TRH. However, recent studies indicate that desensitization does not depend upon internalization, i.e., if internalization is blocked, receptor uncoupling can still mediate desensitization (629 ). This is similar to the angiotensin II and muscarinin M3 receptors that are also coupled to Gq and G11. While the TRH receptor undergoes internalization and recyclization without ligand binding, this "housekeeping" function is slow relative to TRH-induced endocytosis (632 ). THR receptor-ligand complexes are internalized in clathrin-coated vesicles (632 ). A portion of the receptor is targeted to lysosomes, while the remainder is recycled to the cell surface. Similarly, the ligand may remain associated with the receptor to return to the cell surface, or if it dissociates intracellularly, will be degraded in lysosomes or, possibly, reach the cell surface as well (633 ). Internalization of the TRH receptor is dependent upon sequence motifs within its C terminus (634 ), as well as sequences within the second transmembrane region or third intracellular loop that are necessary for G protein coupling (633 ). Further, optimal rates of internalization appear to require coupling to Gq/11 and PPI-PLC (635 ).
Over longer periods of time, TRH receptor binding is also reduced by down-regulation of receptor number. TRH receptor down-regulation occurs in response to TRH, thyroid hormones, and agents that raise cAMP levels (625 636 ). Like acute desensitization, receptor down-regulation is dependent upon the cell of expression, as determined in transfected cell experiments (637 ). The best characterized mechanism of reduced TRH receptor number is reduction in TRH-R1 mRNA levels in GH3 and GH4C1 cells treated with TRH (638 639 ). TRH-R1 mRNA regulation is tightly coupled to activation of the TRH signal transduction elements, protein kinase C, IP3, and intracellular calcium (638 639 640 ). Desensitization of TRH-R1 by cAMP is also mediated by protein kinase A. Conversely, a number of conditions, including hypothyroidism (641 ), and treatment with dexamethasone (639 ), estradiol (642 ), and cycloheximide elevate TRH-R1 mRNA. Both for TRH-R1 mRNA down- and up-regulation, direct effects on gene transcription rates can be demonstrated.
There is also considerable evidence for regulation of TRH receptor mRNA stability. In GH3 cells transfected with the TRH-R1 coding sequence under control of a cytomegalovirus promoter, TRH increases degradation of TRH-R1 mRNA (643 ). TRH-R1 mRNA degradation is induced in pituitary cell types by phorbol esters (644 ). Narayanan and co-workers (645 ) present evidence in GH3 pituitary cells that TRH-R1 mRNA degradation is controlled both by cis-acting elements within the mRNA 3'-untranslated region and by induction of RNases. Cell transfection studies indicate that regulation of TRH-R1 mRNA degradation does not occur in nonpituitary cells (644 646 ). Estradiol-induced increases in TRH receptor mRNA also appear to be mediated through reduction in TRH receptor mRNA degradation (642 ).
B. The prepro-TRH160–169 (pST10) receptor
Of the other pro-TRH-derived peptides, only
prepro-TRH160–169 has characterized receptor binding
(647 ). Receptors for prepro-TRH160–169 seem to be of a
single class, with a higher affinity for [Tyr0]
pST10 (IC50 = 8.3 ± 1.2
nM) than the native pST10
(IC50 = 9.3 ± 1.2 µM). Recent
studies indicate that pST10 receptors cosegregate with
S-100 protein-positive cells in pituitary cultures, supporting their
expression in the folliculo-stellate cells of the anterior pituitary.
Binding sites for pST10 are developmentally regulated, with
an increase from birth to weaning, and then a gradual decline to adult
levels at postnatal day 60 (587 ). Signal transduction by these
receptors is not yet characterized. Within the CNS, pST10
receptor binding is highest in pituitary, the hypothalamus, spinal
cord, and olfactory bulb, as well as the hippocampus (648 ). Its
receptor binding is very high, two-thirds of that in the pituitary, in
urinary bladder, and vas deferens and in the heart and testis, at a
level equivalent to the hypothalamus (588 ).
|
VII. TRH Degradation |
|---|
|
TopAbstractI. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
Four key enzymes breakdown TRH: PAP I, PAP II, and thyroliberinase give
rise to the stable cyclized metabolite CHP (also known as
histidyl-proline-diketopiperazine or His-Pro-DKP), and prolyl
endopeptidase gives rise to the deamidated free acid, TRH-OH (182 ).
These enzyme pathways are shown schematically in Fig. 10. In the CNS, the soluble PAP I and
prolyl endopeptidase, and the membrane-bound PAP II, are the principal
enzymes acting to metabolize TRH (649 ). TRH degradation in serum and
many peripheral tissues is through the serum enzyme thyroliberinase
(172 ). Each enzyme is described below in more detail.
|
Displaying greater substrate specificity is the ectoenzyme pyroglutamyl aminopeptidase II (PAP II) (EC 3.4.19.6) (650 ). Like PAP I, this 260-kDa metalloenzyme removes pyro-Glu from TRH. The distinguishing features of PAP II are its greater substrate specificity and being membrane bound. PAP II is present in CNS synaptosomal fractions, in adenohypophyseal plasma membrane, and liver and serum particulate fractions. Highest activity is observed in the hippocampus and cerebral cortex. PAP II has been identified in many species, being highest in rabbit CNS, and in most cases retains similar features. However, its substrate specificity in bovine synaptosomes, where it has been extensively characterized, is not as narrow as in other species (651 ). The localization of PAP II in the CNS is consistent with its proposed role in the degradation of synaptic TRH (652 ).
In the hypothalamus, PAP II activity is maximal at day 8 after birth, decreasing to adult values at day 45, while in the adenohypophysis it appears at day 8, peaks at day 30, and then decreases to adult values. In addition, thyroid hormone regulates the PAP II in the anterior pituitary but not in the brain (653 654 ). In the PVN, TRH levels and PAP II activity do not correlate during pregnancy and lactation, indicating that PAP II is not the principal determinant of TRH levels (655 ). PAP II activity does vary with the estrous cycle (655 ). Furthermore, in brain areas other than the hypothalamus, PAP II activity decreases from days 9–20 coincident with increases in TRH and decreases in CHP (656 ), indicating PAP II activity can be a critical determinant of TRH steady state levels in some tissues. In sum, it appears that PAP II in areas under prominent endocrine control, such as the pituitary and PVN, subserves a different role than that in nonendocrine tissues.
Thyroliberinase, a fourth TRH-degrading enzyme, present in serum, is similar to PAP II but does not have the transmembrane anchor of PAP II. Like PAP II, thyroliberinase displays greater substrate specificity than PAP I or prolyl endopeptidase (657 658 ). Thyroliberinase may be regulated by thyroid hormone; TRH half-life ranges from about 2 min in the plasma of thyrotoxic animals to 6 min in hypothyroid animals. In humans, the half-life of TRH is similar (172 ).
In studies examining the ontogeny of TRH catabolizing enzymes in pancreas, PAP I and prolyl endopeptidase are detected at early stages of rat pancreatic development, while PAP II remains undetectable. PAP I-specific activity increases until day 3 and decreases after day 5, and prolyl endopeptidase levels peak at 20 days. Because this development does not parallel that seen for TRH levels, it appears that TRH levels in neonatal rat pancreas are principally determined by biosynthetic rates (659 ).
The physiological significance of the soluble enzymes PAP I and prolyl endopeptidase within the brain and spinal cord are unclear, since in the case of neurotransmitter inactivation, TRH is probably degraded outside the neuron by ectoenzymes located on the cell surface, or within lysosomes after endocytosis. Membrane-bound ectoenzymes that are specific for TRH, such as PAP II, are more logically located for hydrolysis of synaptically released peptides. Soluble enzymes are better situated to control degradation of TRH during its transport in the hypophyseal portal vessels and systemic TRH. The exact mechanisms that control the amount of TRH that ultimately reaches the pituitary remain to be elucidated.
|
VIII. Concluding Remarks |
|---|
|
TopAbstractI. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
Understanding of the role of non-TRH pro-TRH-derived peptides represents an exciting new frontier in pro-TRH research. During biosynthesis, these sequences within the precursor may function as structural or targeting elements that guide the folding and sorting of pro-TRH and its larger intermediates so that subsequent processing and secretion are properly regulated. The unique anatomical distribution of the pro-TRH end products, as well as regulation of their levels by neuroendocrine or pharmacological manipulations, described in this review, argues that these peptides will have unique biological roles. Some of these roles, such as for prepro-TRH160–169, will be within the HPT axis, while many others will be unrelated to traditional thyroid function. This review also gathers together an extensive array of data indicating that TRH can function far beyond the HPT axis and should command significant future effort as a focus to develop new therapeutics. These therapeutics, in the form of TRH analogs or nonprotein peptidomimetics, and perhaps using novel delivery systems, will advance our ability to develop TRH and non-TRH pro-TRH-derived peptide agonists and antagonists that can target pro-TRH-derived peptides functioning in specific tissues or brain loci. It is hoped that these new drugs might provide novel treatment approaches for some of today’s most difficult health and societal issues, including drug abuse, depression, chronic pain disorders, and sequelae of CNS injury.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 This work was supported by National Science Foundation Grant
IBN-9507952 (to E.A.N.) and the National Institute on Drug Abuse Grant
1-R01-DA-10762 (to K.A.S.). 
2 Abbreviations: iTRH, immunoreactive TRH; TRH-OH,
deamidated TRH; CHP, cyclo (His-Pro); PTU, propylthiouracil; HPT,
hypothalamic-pituitary-thyroid; DVC, dorsal vagal complex; DMN, dorsal
motor nucleus of the vagus; NTS, nucleus tractus solitaris; IML,
intermediolateral column of the spinal cord; SP, substance P; NPY,
neuropeptide Y; CGRP, calcitonin gene-related peptide; CT, calcitonin;
IL1, interleukin 1; VIP, vasoactive intestinal peptide; OAG, the
diacylglycerol analog 1-oleoyl-2-acetyl-sn-glycerol; CCK,
cholecystokinin; NT, neurotensin; THC, tetrahydrocannabinol; ACh,
acetylcholine; ECS, electroconvulsive shock treatment; NE,
norepinephrine; E, epinephrine; DA, dopamine; 5-HT, serotonin; NO,
nitric oxide; NMDA, N-methyl-D-aspartate; 6-HODA,
6-hydroxydopamine; 5,7-DHT, 5,7-dihydroxytryptamine; GABA,
-aminobutyric acid; EOPs, endogenous opioid peptides; CA,
catecholamines; ENK, enkephalins; DYN, dynorphin; SRIF, somatostatin;
CP, carboxypeptidases; CPE, carboxypeptidase E; CSF, cerebrospinal
fluid; NAc, nucleus accumbens; VTA, ventral tegmental area; PVN,
periventricular nucleus of the hypothalamus; RPa, nucleus raphe
pallidus; RMg, nucleus raphe magnus; ROb, nucleus raphe obscurus; icv,
intracerebroventricular; sc, subcutaneous, ic, intracerebral; SHR,
spontaneously hypertensive rats; WKY, Wistar-Kyoto rats; MCA, middle
cerebral artery; ALS, amyotrophic lateral sclerosis; RMN, Rolling mouse
Nagoya model of ataxia; CNS, central nervous system; EEG,
electroencephalogram; PAG, periaqueductal gray; RPGi, nucleus
reticularis paragigantocellularis; MAF, mesencephalic reticular
formation; POA; preoptic nucleus of the hypothalamus; LC, locus
coeruleus; LPGi, nucleus paragigantocellularis lateralis; LS,
long-sleep; SS, short-sleep; PBMC, peripheral blood monocyte; ME,
median eminence; RER, rough endoplasmic reticulum; PCs, proconverting
enzymes; RSP, regulated secretory pathway; GC, Golgi complex; SG,
secretory granules; TGN, trans-Golgi network; ICC immunocytochemistry;
TRE, thyroid response element; CRE, cAMP response element.
|
References |
|---|
|
TopAbstractI. IntroductionII. Biosynthesis of TRH...III. Function of TRHIV. Function of non-TRH...V. Non-TRH pro-TRH-Derived...VI. TRH and Other...VII. TRH DegradationVIII. Concluding RemarksReferences |
|---|
-lipotropic hormone from sheep pituitary
glands. Can J Biochem 45:1163–1174[Medline]
-globin from mammalian cells. J Cell Biol 108:1647–1655-9-tetrahydrocannabinol-induced hypothermia. Life Sci 26:845–850[CrossRef][Medline]
