| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Biology Indiana University-Purdue University Indianapolis Indianapolis, Indiana 46202-5132
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-glycoprotein subunit
promoter and a reporter construct containing a high-affinity Lhx3
binding site more effectively than the hLhx3b isoform. In addition,
hLhx3a synergized with the pituitary POU domain factor, Pit-1, to
strongly induce transcription of the
TSHß-subunit gene, while hLhx3b did not. We
demonstrate that the differences in gene activation properties between
hLhx3a and hLhx3b correlate with their DNA binding to sites within
these genes. The short hLhx3b-specific amino-terminal domain inhibits
DNA binding and gene activation functions of the molecule. These data
suggest that isoforms of Lhx3 may play distinct roles during
development of the mammalian pituitary gland and other neuroendocrine
systems. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Proper formation of Rathke’s pouch and the subsequent differentiation of pituitary cell types require the secretion of inductive signaling molecules such as BMP-2, BMP-4, and FGF-8. These molecules are secreted from the ventral diencephalon and the developing pituitary to initiate cellular differentiation cascades within embryonic pituitary cells in a spatial and temporal fashion (4, 5, 6). The protein signals induce subsequent pituitary cellular specification by activating multiple transcription factors. Based on their expression patterns, gene knockout experiments, and the analysis of naturally occurring mutants, many transcription factors have been implicated in pituitary organogenesis. They include Lhx3 (also known as P-Lim or LIM3), Lhx4/Gsh-4, Isl-1, Krox-24, T/ebp/TTF-1, Hesx1/Rpx, Pitx-1/P-Otx, Pitx2, Prop-1, Pit-1/GHF-1, Brn-4, Six-3, SF-1, Pax-6, Msx-1, Nkx 3.1, GATA-2, P-Frk, thyrotrope embryonic factor (TEF), and Zn-16 (Ref. 5 ; reviewed in Refs. 1, 2, 3).
Mutations in pituitary transcription factors have been shown to cause pituitary disease. Humans with mutations in the Pit-1 or Prop-1 genes exhibit combined pituitary hormone deficiency (CPHD). Pit-1 is a well characterized POU domain pituitary transcription factor that directly activates the GH, PRL, and TSHß genes, as well as its own gene (reviewed by Refs. 1, 2, 3, 7). Patients with mutated Pit-1 genes have deficiencies in GH, PRL, and TSH. Many types of Pit-1 gene mutations have been identified, including those with autosomal dominant and recessive patterns of inheritance (see Refs. 7, 8 for review). The Prop-1 (or Prophet of Pit-1) transcription factor is upstream of Pit-1 in the pituitary developmental cascade (9). In addition to deficiencies of GH, PRL, and TSH, patients with mutations in the Prop-1 gene may lack LH and FSH (8). Much effort continues to be devoted to the goal of characterizing additional genetic lesions associated with human pituitary disease.
During pituitary development, the actions of the LIM homeodomain transcription factors Lhx3, Lhx4, and Isl-1 are essential for the establishment of Rathke’s pouch and the subsequent differentiation of specialized hormone-secreting cell types. LIM homeodomain proteins contain two amino-terminal LIM motifs and interact with DNA using a characteristic homeodomain. The LIM domain is a conserved, zinc finger-like structure that mediates interactions with other proteins, and LIM homeodomain proteins have been demonstrated to be critical to many developmental pathways (reviewed in Refs. 10, 11, 12). During early embryogenesis, the Lhx3 gene is expressed in several regions of the brain and spinal cord, and then it becomes restricted to the primordial pituitary cells of Rathke’s pouch and their descendents in the adult gland (13, 14, 15, 16, 17). In the nervous system, the Lhx3, Lhx4, and Isl-1 gene products are required for the specification of motor neurons that emerge from the developing neural tube (e.g. Refs. 18, 19, 20, 21). Elegant studies of mice with ablated Lhx3 genes have revealed that pituitary development is arrested after the formation of Rathke’s pouch and that Lhx3 is required for differentiation of the hormone-secreting cells of the pituitary (22, 23). Lhx4 also is required for complete development of Rathke’s pouch; but unlike Lhx3, this factor is not essential for the determination and specification of differentiated pituitary cell types (23). Mice lacking both Lhx3 and Lhx4 genes do not develop a rudimentary Rathke’s pouch, indicating that at least one of these genes is required during the initial stages of pituitary development (23). Our laboratory and others have demonstrated that Lhx3 can activate pituitary trophic hormone genes, acting both alone and with other pituitary transcription factors such as Pit-1 and Pitx1/P-Otx (15, 24, 25). Additional functional analyses are required to fully understand the role of products of the Lhx3 gene in the developing nervous system, the early establishment of Rathke’s pouch, and in later pituitary gland function and maintenance.
To further characterize the molecular actions of Lhx3, we identified
two alternate forms of human Lhx3, hLhx3a and hLhx3b, which share the
LIM domains, homeodomain, and carboxyl terminus of Lhx3, but possess
distinct amino-terminal protein sequences. Both isoforms are nuclear
proteins and are detected in the pituitary gland, but
trans-activation assays revealed different abilities to
activate anterior pituitary hormone gene regulatory regions. Human
Lhx3a activated a reporter gene containing the -glycoprotein
subunit (GSU) promoter and a minimal reporter gene
containing consensus Lhx3 binding sites. Further, Lhx3a synergized with
Pit-1 in induction of the TSH ß-subunit
(TSHß) gene promoter. By contrast, Lhx3b was either
inactive or only weakly activated trophic hormone genes. We demonstrate
that the differences in trans-activation ability between
hLhx3a and Lhx3b correlate with their DNA binding to sites within these
target genes. Our results indicate that the amino terminus of hLhx3b
inhibits the ability of this factor to bind DNA and
trans-activate target genes compared with Lhx3a. To our
knowledge, this is the first description of different functional
properties for alternate forms of a LIM homeodomain class
transcriptional regulator. The hLhx3a-specific and hLhx3b-specific
amino-terminal domains may represent novel functional motifs derived
throughout evolution to confer properties unique to Lhx3 that are
important in mammalian development. These findings suggest that Lhx3
isoforms may perform distinct roles in the development of the pituitary
gland and during motor neuron differentiation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). We named
these protein isoforms hLhx3a and hLhx3b. The GenBank accession numbers
of the reported sequences are: hLhx3a (AF156888), hLhx3b
(AF156889). Some hLhx3b cDNAs contained an approximately
200-bp insert in the 3'-untranslated region (nucleotide sequences
deposited in GenBank).
|
). By
contrast, the shorter amino-terminal sequence of hLhx3a is only 58%
similar to the amino terminus of murine Lhx3a (Fig. 1B). The a- and
b-specific domains of hLhx3 do not significantly match other sequences
in the databases. The common regions of the hLhx3 proteins display
strong overall similarity to murine (94% identity) and porcine (95%
identity) Lhx3 sequences (Fig. 1B). The LIM and DNA binding domains of
Lhx3 are strongly conserved in the mammalian species, with complete
identity in the LIM 2 domain and the homeodomain (Fig. 1B). Comparison
of mammalian Lhx3 sequences with Lhx3/LIM3 family proteins from
nonmammalian species also reveals conservation of the LIM and
DNA-binding domains and the carboxyl-terminal LIM3-specific domain
(LSD) noted by Glasgow et al. (17) and Thor et
al. (18) (Fig. 1C). The amino termini of nonmammalian Lhx3/LIM3
proteins display more similarity to the amino terminus of hLhx3a than
to that of hLhx3b, but overall are diverged in length, composition, and
sequence.
Analysis of the Human Lhx3 Gene and Its
Products
To determine whether the identified hLhx3 cDNAs
represented a single genomic locus, Southern analyses were performed.
Human Lhx3 cDNA probes hybridized to single DNA fragments (Fig. 2A; and data not shown), indicating that
hLhx3 is encoded by a single gene. In Northern analyses, hLhx3 cDNA
probes hybridized to a rare mRNA of approximately 2.4 kb in samples of
adult human pituitary poly A+ RNA (Fig. 2B). Signals were
not detected in samples of total pituitary RNA (Fig. 2B). As controls,
RNA blots also were hybridized to a human Pit-1 cDNA probe (Fig. 2B).
In vitro transcription/translation was performed using
rabbit reticulocyte lysates to generate radiolabeled hLhx3a and hLhx3b
proteins from the identified cDNAs. Analysis of these proteins by
electrophoresis revealed apparent molecular masses of approximately 60
kDa for both isoforms (Fig. 2C). As we have previously described for
porcine Lhx3 (25), this apparent molecular mass is slightly larger than
that predicted from the hLhx3a/b open reading frames, suggesting
modification of the Lhx3 protein in these preparations or aberrant
migration during electrophoresis due to composition.
|
, A and B). By contrast,
constructs containing hLhx3a or hLhx3b were detected in the nucleus
(Fig. 3, C–F), indicating that they are nuclear proteins.
|
, inset). The
standard curve was linear in a range from 100 to 10 million copies of
hLhx3. Seven normal adult pituitary samples were assayed: hLhx3 was
present at levels ranging from 2,000 to 11,000 copies/150 ng RNA with a
mean value of 7,200 copies/150 ng RNA (Fig. 4A). This value is
consistent with hLhx3 being a rare message. In previous studies, we
have demonstrated that ubiquitously expressed genes such as ß-actin
are expressed at approximately 500,000 copies/150 ng RNA (27). In
specialized tissues such as adipose, we have shown that mRNAs for
hormones such as leptin are found at approximately 300,000 copies/150
ng RNA, and transcription factors such as CCAAT/enhancer binding
protein- (C/EBP) are found at about 150,000 copies/150 ng RNA
(27).
|
). To investigate whether the Lhx3
isoforms were expressed in specific cell types, we determined
expression of Lhx3 isoforms in established mouse pituitary cell lines.
In the -TSH thyrotrope tumor cell line (28), both mLhx3a and mLhx3b
isoforms were readily detected (Fig. 4C). In AtT20 corticotrope cells,
low levels of mLhx3b were detected, and in some preparations, trace
levels of mLhx3a were also observed. Neither isoform was detected in
the GHFT-1 cell type: this Pit-1-expressing cell line represents an
intermediate stage of pituitary development before differentiation and
expression of trophic hormone genes (29). These data are consistent
with observations of the overall levels of Lhx3 gene
expression by RNA analyses of these cell types (Refs. 14, 15, 16 ; K.
W. Sloop and S. J. Rhodes, unpublished data).
Differential Activation of Pituitary Hormone Genes by Human Lhx3
Isoforms
We and others have previously demonstrated that Lhx3 can activate
anterior pituitary trophic hormone gene promoters (15, 24, 25). For
example, we have demonstrated that Lhx3 can induce transcription from
the GSU gene by specifically binding to a site located at
-350 to -323 bp (known as the pituitary glycoprotein basal
element, PGBE) within the proximal region of the promoter (15, 25).
This gene encodes the common subunit of the LH, FSH, and TSH anterior
pituitary hormones. Other laboratories have shown that the Lhx2/LH-2
LIM homeodomain transcription factor also can recognize this element
(30), and that this element is required to correctly restrict
expression of the GSU gene to pituitary gonadotropes and
thyrotropes in transgenic mice (31). To test the ability of the hLhx3
isoforms to activate the GSU gene, we transiently
cotransfected human embryonic 293 cells with GSU
luciferase reporter genes and expression vectors containing full-length
hLhx3a and hLhx3b cDNAs. 293 cells were used because they are of
human origin, efficiently transfected, and do not express Lhx3. In
these assays, hLhx3a activated the GSU promoter (Fig. 5A). Surprisingly, hLhx3b did not induce
transcription from this promoter (Fig. 5A). Similar experiments using
expression vectors encoding hLhx3a or hLhx3b with carboxyl myc epitope
tags gave the same results: hLhx3a-myc activated the GSU
promoter and hLhx3b-myc was inactive (Fig. 5A). In these experiments,
the epitope-tagged constructs were generally more active: this is
likely due to the fact that, in these constructs, the 3'-untranslated
region of the cDNA is absent. The murine Lhx3 3'-untranslated
region contains ATTTA sequence motifs that may confer instability to
the RNA (32). These motifs are conserved in the human Lhx3 sequences
(data not shown), and the levels of both isoforms are therefore likely
to be somewhat lower in the full-length cDNA experiments. An alternate
explanation is that the single myc epitope confers an activation
function to the Lhx3 molecules. This has been noted in experiments
where multimers of five myc epitope tags are used (33). It is important
to note, however, that the relative activities observed for the two
Lhx3 isoforms are consistent for all types of expression constructs. To
confirm expression of the hLhx3 proteins in the transfected cells,
Western analysis was performed using a specific anti-myc antibody. Both
isoforms were detected at similar levels as protein species of
approximately 60 kDa (Fig. 5B). This apparent molecular mass is similar
to that of the in vitro translated hLhx3 proteins described
above. Western blots were quantified to determine relative expression
levels of the hLhx3 isoforms. The observed expression levels of the two
isoforms were similar; on average, the hLhx3b isoform was expressed at
slightly higher levels (1.2-fold the level observed for hLhx3a).
|
GSU promoter, we and others have demonstrated that Lhx3
can induce pituitary trophic hormone gene promoters in synergy with
pituitary transcription factors such as Pit-1 and Pitx1/P-Otx (15, 24, 25). To examine whether the hLhx3 isoforms also had distinct abilities
to cooperate with other factors in transcriptional induction, we
transfected a TSHß reporter gene with hLhx3a and hLhx3b
expression vectors. Pit-1 and hLhx3a moderately activated the
TSHß promoter and together strongly induced transcription
(Fig. 6A). Human Lhx3b only weakly
activated the TSHß promoter and did not effectively
synergize with Pit-1 (Fig. 6A). As controls, similar experiments were
performed with TEF. TEF is a PAR-bZIP transcription factor that
strongly induces the TSHß promoter (34). In our
experiments, TEF activated expression from the TSHß
promoter, and expression of hLhx3a or hLhx3b did not affect this
activity (Fig. 6B), demonstrating that the observed synergy between
hLhx3a and Pit-1 was specific.
|
GSU and TSHß promoters are complex
regulatory structures containing recognition sites for multiple
transcription factors. For example, the GSU PGBE Lhx3
binding site appears to cooperate with other elements in gene
regulation (31, 35). We therefore extended our studies by comparing the
abilities of hLhx3a and hLhx3b to activate a defined synthetic
luciferase reporter gene containing three copies of a consensus Lhx3
DNA binding element cloned upstream of a minimal promoter. Again,
expression vectors for hLhx3a and hLhx3a-myc activated transcription
from this reporter more effectively than the corresponding hLhx3b
vectors (Fig. 7). These data further
support the hypothesis that hLhx3a is a more potent activator of
transcription than hLhx3b for target genes containing this type of
recognition element.
|
and 5).
Mechanisms for the different activities include the possibilities that
the a-specific and b-specific domains are trans-activation
or trans-repression domains, respectively; or that the a-
and b-specific domains may mediate interaction with distinct regulatory
factors; or that the a- and b-specific domains may differently modulate
the DNA binding properties of hLhx3. We have previously demonstrated
that the LIM domains of Lhx3 repress DNA binding by the homeodomain
(15, 25). To test the hypothesis that the a- and b-specific domains
conferred distinct DNA binding properties upon the hLhx3 molecule, we
performed electrophoretic mobility shift assays (EMSAs) to test binding
to the GSU -350/-323 bp Lhx3 site and to the Lhx3
consensus binding site. Experiments using in vitro
translated native hLhx3 proteins revealed that hLhx3a bound to these
sites more effectively than hLhx3b (Fig. 8B). Similar results were obtained with
hLhx3 proteins containing carboxyl-terminal myc epitope tags (Fig. 8B).
|
).
EMSA reactions using the recombinant purified proteins again indicated
that hLhx3a bound more effectively than hLhx3b to both DNA sites (Fig. 8B). Addition of excess unlabeled binding site DNA to hLhx3a EMSA
reactions reduced the amount of bound complex, whereas it abolished the
binding of hLhx3b (Fig. 7B; and data not shown). These experiments
demonstrate that the b-specific domain inhibits the binding of hLhx3b
to this class of DNA site. Further, the b-specific domain can perform
this function when positioned at the amino terminus of the native
protein or when placed in an internal context such as in the GST fusion
proteins. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, the LIM domains, the homeodomain, and the Lhx3/LIM3-specific
domain are conserved in Lhx3-type proteins from Drosophila
to humans. In this study, we have demonstrated that proteins with distinct functional properties are generated from the human Lhx3 gene. During evolution, the increased complexity of organisms has required an extended number of regulatory factors with sufficient capacity to control the development of organ systems comprised of multiple, differentiated cell types. This increased molecular diversity has been accomplished by mechanisms such as the formation of additional genes by gene duplication; the generation of multiple proteins from single genes by the use of distinct promoters; and posttranscriptional processes. Posttranscriptional mechanisms include the use of alternate translation initiation codons, alternative RNA splicing, the production of protein isoforms from single RNAs by RNA editing, and proteolytic processing. For Lhx3-type proteins, it appears that both gene expansion (to produce genes encoding related proteins such as Lhx4) and RNA-mediated mechanisms generating isoforms such as hLhx3a and hLhx3b have been used to provide a battery of proteins with important roles in the development of both neural and endocrine structures, including the anterior pituitary.
Alternate isoforms of LIM homeodomain proteins, nonhomeodomain LIM
proteins, and several other types of transcription factors are known.
For example, the LIM homeodomain transcription factors, Lhx6.1a and
Lhx6.1b, appear to be generated by an alternative splicing event that
generates proteins with different carboxyl terminal amino acids (36).
These factors are expressed at different times in the developing mouse
brain, but unique functional properties of the isoforms have not been
demonstrated (36). Alternate forms of LMO7 and LIM-kinase proteins
lacking either the LIM domain or the kinase domain, respectively, have
also been described (37, 38). The isoforms of these proteins have
different expression patterns and distinct functional properties.
Interestingly, Skn-1a and Skn-1i are alternatively spliced POU domain
factors expressed in the epidermis that, like Lhx3, have different
amino-terminal domains (39). Similar to the hLhx3b-specific
domain, the i-specific domain of Skn-1 inhibits DNA binding and gene
trans-activation by this factor. However, primary
amino acid sequence comparison between these inhibitory domains reveals
no similarity. Alternate isoforms with distinct functions and/or
expression patterns are also seen in the nuclear receptor protein
superfamily. Examples include the distinct expression patterns of the
isoforms of the peroxisome proliferator-activated receptor 
developmental regulatory protein (40, 41) and the unique
transcriptional properties of the ß2 thyroid receptor isoform
(42).
In the mouse, isoforms of Lhx3 also have been described (16, 32). We
suggest that the human Lhx3b clone described in this study is
equivalent to murine Lhx3b, because the two amino-terminal domains are
nearly identical (Fig. 1). The human and murine Lhx3a amino-terminal
domains are much less alike, but some sequence features are similar
(Fig. 1). We hypothesize that the Lhx3a forms are functionally related,
but this function does not require conservation of the Lhx3a-specific
domain. To date, other LIM homeodomain proteins with alternate
functional domains have not been described. The hLhx3a- and
hLhx3b-specific amino-terminal motifs may be novel functional domains
derived throughout evolution to confer properties unique to Lhx3a and
Lhx3b that allow specific roles in the development of the pituitary
gland and neural structures.
We have demonstrated that hLhx3a is able to activate specific target
genes, whereas hLhx3b is inactive and that these functional differences
correlate with a reduced binding of hLhx3b to DNA elements within these
genes (
Figs. 4–8). These data, and the observation that the
Lhx3a-specific domain is poorly conserved in sequence and length
(Fig. 1), together suggest that the Lhx3b-specific domain confers
intramolecular inhibition of DNA binding and
trans-activation functions. We have previously demonstrated
that the LIM domains of mammalian Lhx3 proteins inhibit DNA binding
(15, 25), and others have made similar observations for several LIM
homeodomain proteins (reviewed in Ref. 12). Therefore, the
Lhx3b-specific domain may inhibit DNA binding by configuring the LIM
domains such that they exert a greater inhibitory function upon the
homeodomain. Alternatively, the Lhx3b-specific domain may function
independently of the LIM domains, or it may exert both LIM-dependent
and independent effects.
Our data suggest that the alternate hLhx3 isoforms may have distinct
functions within the developing pituitary gland and neural structures
such as the embryonic spinal cord. Our results suggest that the
described Lhx3 isoforms are critical to thyrotrope differentiation and
maintenance because both are expressed in the -TSH cell line, and
hLhx3a activated the GSU and TSHß genes. The
low level of Lhx3 expression in AtT20 cells is consistent with
observations of Lhx3 knockout mice that have some corticotropes, but
lack other pituitary cell types (22). It is tempting to speculate that
the more abundant Lhx3b isoform in AtT20 cells plays a unique role in
differentiation of this cell type.
The described hLhx3 isoforms may have distinct target genes or, dependent on its expression profile, hLhx3b may play a direct or indirect dominant negative role. Experiments in the mouse have suggested that in this species Lhx3a may be expressed earlier than Lhx3b during development (32). Specification of differentiated pituitary cell phenotypes appears to be controlled by the combinatorial actions of multiple tissue-specific transcription factors and signaling proteins (reviewed in Refs. 1, 2, 3). The differential activities or expression patterns of alternate forms of essential members of this program, such as Lhx3 and Pit-1, may play critical roles in the determination of pituitary cell fates. Indeed, we and others have described isoforms of Pit-1 that have distinct properties and expression patterns (e.g. Refs. 43, 44 ; reviewed in Ref. 7). In addition, the described Lhx3 isoforms may interact differently with broadly expressed regulatory proteins or transcriptional coactivators/corepressors. For example, a conserved family of nuclear LIM domain-interacting proteins, known as NLI/Ldb1/CLIM/Chip, has been characterized (Refs. 45, 46, 47 and references therein). These proteins appear to be regulatory partners for LIM homeodomain factors and can mediate homo- and heterodimerization of these factors. Recently, a RING-H2 zinc finger protein, RLIM, was identified and also described as a regulatory partner for LIM homeodomain factors (48). This coregulatory protein appears to recruit the Sin3a/histone deacetylase corepressor complex to function as a LIM-associated inhibitory factor (48). Coregulators such as NLI proteins and RLIM may differentially modulate the ability of Lhx3 isoforms to regulate target genes.
The detection of hLhx3 in the pituitary gland and the demonstration of its ability to regulate pituitary trophic hormone gene promoters suggest a continued role for Lhx3 in the adult human pituitary gland in maintenance of hormone gene expression. This study also provides tools for future investigations examining the role of hLhx3 in pituitary diseases such as combined pituitary hormone deficiency and pituitary tumor disease. Further studies of hLhx3 and related factors will extend our understanding of the developmental program that establishes the hormone-releasing cells of the human pituitary gland and may facilitate future protocols designed to treat pituitary diseases.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
DNA Cloning/DNA Sequencing/Plasmid Construction
Database searches revealed that a human expressed sequence
tag (EST) sequence displayed similarity to murine and porcine
Lhx3. The bacterial culture containing this plasmid could not be
retrieved by any of the distributors of EST clones. PCR was then used
to amplify this sequence from adult human pituitary cDNA using the
following primers: (5'-cggaattctacaacacctcgcccaagccgg-3',
5'-cggaattcggaacgaggggcccttgac-3'). RACE was then performed using the
5'-RACE 2.0 system (Life Technologies, Inc.) and adult
human pituitary cDNA. Finally, partial hLhx3 cDNAs were used as high
stringency probes (final wash = 0.5x SSC, 65 C) of a human
pituitary cDNA library (CLONTECH Laboratories, Inc., Palo
Alto, CA). From 1 million plaques, 12 positive bacteriophage were
identified, and cDNAs were subcloned into pBluescript KS II (-)
(Stratagene, La Jolla, CA) as described (25). Seven clones
were completely sequenced on both strands by automated DNA sequencing
using a Perkin Elmer Corp. (Norwalk, CT) DNA sequencer
(Biochemistry Biotechnology Facility, Indiana University School of
Medicine). Sequence analysis was performed using Wisconsin
Genetics/GCG and DNASIS (Hitachi, South San
Francisco, CA) software.
Expression vectors for hLhx3a and hLhx3b were generated by directly cloning full-length cDNAs into pcDNA3 (Invitrogen). Myc epitope-tagged hLhx3 expression vectors (hLhx3a-myc, hLhx3b-myc) were made by cloning compatible cDNA fragments into pcDNA3.1/Myc-His(-)C (Invitrogen). DNA fragments were generated by PCR using the following primers: 5'-cgggatccgatcgcttcggcagcagctg-3' (5', hLhx3a-myc); 5'-cgggatccttgatatttaccccggaggc-3' (5', hLhx3b-myc); and 5'-gcgaagcttggaactgagcgtggtctacctca-3' (3', hLhx3a/b-myc). An expression vector containing four copies of GFP was constructed by removing the 4xGFP cassette from plasmid p713 (Ref. 26 ; a generous gift of Dr. Ursula Stochaj, McGill University, Montreal, Quebec, Canada) by digestion with KpnI and EcoRI and cloning of this fragment into pcDNA3 to generate pcDNA3–4xGFP. Compatible hLhx3 DNA fragments were generated by PCR using the following primers: 5'-tacaagcttcgcgatgctgctggaaacgg-3' (5', hLhx3a); 5'-tacaagcttaccatggaggcgcgcgggga-3' (5', hLhx3b); and 5'-cccggtaccaactgagcgtggtctacctc-3' (3', hLhx3a/b). These fragments then were digested with HindIII and KpnI and cloned into pcDNA3–4xGFP to generate vectors expressing hLhx3a-4xGFP and hLhx3b-4xGFP.
Southern and Northern Analyses
Human genomic DNA was extracted from peripheral blood using a
QIAamp Blood Maxi Kit (Qiagen, Chatsworth, CA). DNA was
digested to completion with restriction enzymes, electrophoresed
through 0.7% agarose gels, and transferred to nylon membranes
(Hybond-N+, Amersham Pharmacia Biotech,
Piscataway, NJ). Membranes were hybridized to radiolabeled hLhx3 cDNA
probes in ExpressHyb buffer (CLONTECH Laboratories, Inc.)
at 1–2 x 106 cpm/ml for 2–4 h. DNA fragments were
labeled by random priming using the Klenow fragment enzyme (Life Technologies, Inc.) and 32P-dCTP (Amersham Pharmacia Biotech) to a specific activity of >1 x
109 cpm/µg. After hybridization, membranes were washed in
0.5x SSC at 65 C, followed by exposure to MR film (Eastman Kodak Co., Rochester, NY) with intensifying screens.
Total human pituitary RNA was extracted as described above. Poly A+ RNA was purchased from CLONTECH Laboratories, Inc. RNA was separated on denaturing, formaldehyde agarose gels followed by transfer to Nytran Plus membranes using the Turboblotter system (Schleicher & Schuell, Inc., Keene, NH). Probes were a 0.4 kb hLhx3 cDNA fragment encoding the LIM domains of the protein and a 1.2 kb cDNA encoding human Pit-1. Membranes were hybridized to radiolabeled cDNA probes as described above.
In Vitro Transcription/Translation
Radiolabeled hLhx3 proteins were synthesized in vitro
from pcDNA3 expression vector substrates using T7 RNA polymerase, TnT
rabbit reticulocyte lysates (Promega Corp., Madison, WI),
and 35S-methionine (Amersham Pharmacia Biotech, Arlington Heights, IL). Proteins were analyzed using
SDS-PAGE followed by treatment with Amplify fluorography reagent
(Amersham Pharmacia Biotech) and exposure to MR film
(Eastman Kodak Co.) at -80 C.
Confocal Microscopy
Human 293T cells (1 x 105) were grown in
chamber slides (Nunc) and transfected with 4xGFP, hLhx3a-4xGFP, or
hLhx3b-4xGFP expression vectors as described below. After 48 h,
cells were washed with 1x PBS, and then either directly visualized
live or fixed with 2% paraformaldehyde, washed with 70% ethanol/30%
PBS, and stored in 1x PBS before visualization. Fluorescence and
trans-illumination images were collected using a MRC 1024
laser scanning confocal microscope (Bio-Rad Laboratories, Inc., Richmond, CA) with a 60x water immersion objective
(Nikon, Melville, NY) and a krypton-argon laser. Images
were captured with Metamorph software (Universal Imaging, West Chester,
PA).
Quantitative Assay of Gene Expression
"Real time" quantitative PCR utilizing the ABI PRISM 7700
Sequence Detection System (Perkin Elmer Corp.) was
performed as previously described (27, 49). Total RNA was isolated from
human pituitary tissue as described above. Reverse transcription of
total RNA with a hLhx3-specific reverse primer
(5'-ctcccgtagaggccattg-3') was performed using SuperScript II reverse
transcriptase (Life Technologies, Inc.). Quantitative PCR
amplification reactions were performed in triplicate and included: 2
µl of cDNA synthesis reaction, 1x TaqMan Buffer A, 300
nM dATP, dCTP, dGTP, and 600 nM dUTP, 3.5
mM MgCl2, 1.25 U AmpliTaq Gold DNA polymerase
(Perkin Elmer Corp.), 0.5 U of AmpErase uracil
N-glycosylase, 300 nM forward primer
(5'-ggacaaggacagcgttcag-3') and reverse primer, and 200 nM
hLhx3 fluorogenic probe (5'-ttccccgatgagccttccttggcggaa-3'). Reaction
parameters were as follows: 50 C, 2 min; 94 C, 10 min; and then 35
cycles of 94 C, 30 sec; 60 C, 1 min. Serial dilutions of hLhx3 cDNA
were amplified simultaneously with patient samples to generate a
standard curve. Values reported are averages of four independent
experiments.
RT-PCR Analysis
cDNA was synthesized from adult pituitary RNA as described
above using oligo-d(T) as a primer. PCR then was performed using
the following primers: 5'-cgggatccatgctgctggaaacggggct-3' (5', hLhx3a);
5'-cgggatccatggaggcgcgcggggagct-3' (5', hLhx3b); and
5'-cggaattctcagaactgagcgtggtcta-3' (3', hLhx3a/b). Cycling parameters
were as follows: 94 C, 30 sec; 60 C, 30 sec; 72 C, 1 min; for 30
cycles. Reaction products were analyzed on 1% agarose Tris-borate
gels. For analysis of mouse pituitary cell lines, cDNA was synthesized
from total RNA using 5'-tggtcacagcctgcacacat-3'. PCR then was performed
using the following primers: 5'-aaccactggattagtgactg-3' (5', mLhx3a);
5'-gaagttcagggtcggaggg-3' (5', mLhx3b); and 5'-tggtcacagcctgcacacat-3'
(3', mLhx3a/b). Cycling parameters were as follows: 94 C, 30 sec; 60 C,
30 sec; 72 C, 30 sec; for 30 cycles. Reaction products were analyzed on
11% acrylamide Tris-borate gels.
Cell Culture, Transfection Assays, Statistical Analysis
Human embryonic kidney 293 and 293T cells were cultured in DMEM
(Life Technologies, Inc.) with 10% FBS (Irvine
Scientific, Santa Ana, CA), 100 U/ml penicillin, and 100 µg/ml
streptomycin (Irvine Scientific); 1.5 x 105 cells per
60-mm dish were transfected with calcium phosphate/DNA precipitates
using the CalPhos system (CLONTECH Laboratories, Inc.).
Reporter plasmid (0.5 µg) and expression vector (0.1–1.0 µg of
expression vector) were added per 60-mm dish, and all groups received
equal final DNA concentrations. The murine GSU promoter
luciferase plasmid (30) was a generous gift of Dr. Richard Maurer
(Oregon Health Sciences University, Beaverton, OR). The murine
TSHß -1.2-kb promoter luciferase plasmid was as described
(34). The Lhx3 consensus binding site reporter gene was constructed by
cloning three copies of 5'-cagaaaattaattaattgtaa-3' upstream of a
minimal PRL (-36 bp) promoter luciferase reporter gene (J. L.
Bridwell, J. R. Price, G. E. Parker, K. W. Sloop, A. McCutchan
Schiller, and S. J. Rhodes, in preparation). Control
cultures received empty expression vector DNA. Luciferase activity was
measured 48 h after transfection as described (25). All assays
were performed in triplicate. Total cell protein was determined by the
Bradford method (Bio-Rad Laboratories, Inc.) and
luciferase activity was normalized to protein concentration. Data
points were compared using a one-tailed Student’s t test
for paired samples using Sigma Plot 2.0 (Jandel Scientific, San Rafael,
CA). Values were considered significantly different when
P < 0.05.
Western Analysis
Western analysis of 293 or 293T cells transfected with
hLhx3a-myc, hLhx3b-myc, or control expression vectors was performed as
we described previously (25). The mouse anti-myc monoclonal antibody
9E10, developed by Evan et al. (50), was obtained from the
Developmental Studies Hybridoma Bank developed under the auspices of
the National Institute of Child Health and Human Development and
maintained by the Department of Biological Sciences, University of Iowa
(Iowa City, IA). Ascites fluid was used at a 1:5000 dilution.
The secondary antibody was a goat antimouse/horseradish peroxidase
(Sigma, St. Louis, MO) at 1:15,000. Results were
visualized using Supersignal chemiluminescence reagents (Pierce Chemical Co., Rockford, IL) and MR film (Eastman Kodak Co.). Data were quantified using a Bio-Rad Laboratories, Inc. imaging densitometer and Molecular Analyst software
(Bio-Rad Laboratories, Inc.).
Recombinant Protein Preparation/EMSA
Bacterial expression vectors for GST-hLhx3a and GST-hLhx3b
fusion proteins were generated by cloning
BamHI/EcoRI compatible fragments of the hLhx3a
and hLhx3b cDNAs into pGEX-KT (51). cDNA fragments were generated by
PCR using the following oligonucleotides:
5'-cgggatccatgctgctggaaacggggct-3',5'-cggaattctcagaactgagcgtggtcta-3'
(hLhx3a) and
5'-cgggatccatggaggcgcgcggggagct-3',5'-cggaattctcagaactgagcgtggtcta-3'
(hLhx3b). Recombinant proteins were expressed in E. coli
BL21 (DE3) pLysS and affinity-purified as we have previously described
(25). Proteins were analyzed on 12% SDS-PAGE gels. EMSAs were
performed as described (25). Oligonucleotides representing the -350 to
-323 bp region of the murine GSU promoter
(5'-acattaggtacttagctaattaaatgtg-3' and
5'-cacatttaattagctaagtacctaatgt-3') were used or the Lhx3 consensus
binding site described above. In competition experiments, 1000-fold
molar excess of unlabeled binding site DNA was added to EMSA
reactions.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
This work was supported by grants to S.J.R. from the National Science Foundation, the National Research Initiative Competitive Grants Program/US Department of Agriculture, and the Indiana University-Purdue University Indianapolis Office of Faculty Development. We dedicate this paper to the memory of Dr. Raymond Russo.
Received for publication June 28, 1999. Revision received August 24, 1999. Accepted for publication September 8, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
(C/EBP-). Biochem Biophys Res
Commun 251:142–147[CrossRef][Medline]
-subunit promoter by a
LIM-homeodomain transcription factor. Mol Cell Biol 14:2985–2993 subunit gene requires
multiple interacting DNA elements in transgenic mice and cultured
cells. Mol Endocrinol 12:622–633 subunit gene to
pituitary and placenta. J Biol Chem 270:26497–265042
versus PPAR1 and activation with retinoid X receptor agonists and
antagonists. J Biol Chem 272:8071–8076This article has been cited by other articles:
 |
|
 |
|
  R. W. Pfaeffle, C. S. Hunter, J. J. Savage, M. Duran-Prado, R. D. Mullen, Z. P. Neeb, U. Eiholzer, V. Hesse, N. G. Haddad, H. M. Stobbe, et al. Three Novel Missense Mutations within the LHX4 Gene Are Associated with Variable Pituitary Hormone Deficiencies J. Clin. Endocrinol. Metab., March 1, 2008; 93(3): 1062 - 1071. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. W. Pfaeffle, J. J. Savage, C. S. Hunter, C. Palme, M. Ahlmann, P. Kumar, J. Bellone, E. Schoenau, E. Korsch, J. H. Bramswig, et al. Four Novel Mutations of the LHX3 Gene Cause Combined Pituitary Hormone Deficiencies with or without Limited Neck Rotation J. Clin. Endocrinol. Metab., May 1, 2007; 92(5): 1909 - 1919. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Granger, C. Bleux, M.-L. Kottler, S. J. Rhodes, R. Counis, and J.-N. Laverriere The LIM-Homeodomain Proteins Isl-1 and Lhx3 Act with Steroidogenic Factor 1 to Enhance Gonadotrope-Specific Activity of the Gonadotropin-Releasing Hormone Receptor Gene Promoter Mol. Endocrinol., September 1, 2006; 20(9): 2093 - 2108. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. P. S. Bhangoo, C. S. Hunter, J. J. Savage, H. Anhalt, S. Pavlakis, E. C. Walvoord, S. Ten, and S. J. Rhodes A Novel LHX3 Mutation Presenting as Combined Pituitary Hormonal Deficiency J. Clin. Endocrinol. Metab., March 1, 2006; 91(3): 747 - 753. [Abstract] [Full Text] [PDF]
|
|
