| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Departments of Pediatric Endocrinology (B.I.H.-S., M.J.) and Physiological Chemistry (K.D.A.) and Center for Biomedical Genetics (K.D.A., P.C.v.d.V.), University Medical Center, 3508 AB Utrecht, The Netherlands; and Department of Pediatric Endocrinology, Emma Children’s Hospital Academic Medical Center (B.B.), 1105 AZ Amsterdam, The Netherlands
Address all correspondence and requests for reprints to: Maarten Jansen, M.D., Ph.D., Department of Pediatric Endocrinology, University Medical Center Utrecht, Room KC 03.063.0, P.O. Box 85090, 3508 AB Utrecht, The Netherlands. E-mail: m.jansen{at}wkz.azu.nl
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-helix of the POU-specific domain (W193R),which causes a
500-fold reduction in the ability to bind to DNA and activate
transcription. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The DNA-binding POU domain of POU1F1 is located in the C-terminal part of the molecule [amino acids (aa) 119–273]. It consists of a 60-aa-long POU homeodomain (POUHD) and a 75-aa POU-specific (POUS) domain, connected by a 15-aa flexible linker (5). Both domains contribute to the specific and high affinity binding of the POU1F1 molecule to its recognition sequence, (A/T)(A/T)TTATNCAT (6). The crystal structure of the POU1F1 POU domain bound to DNA shows that both subdomains contain helix-turn-helix motifs and form a dimer (7). DNA binding by POU1F1 as well as interaction with other nuclear proteins are required for specific trans-activation of its target genes (8).
Two strains of dwarf mice have been shown to harbor structural defects
in the POU1F1 gene resulting in combined pituitary hormone deficiency
(CPHD) with pituitary hypoplasia and absence of somatotrophs,
lactotrophs, and thyrotrophs. The Snell dwarf mouse carries a G
T
missense mutation at nucleotide 783 of the POUHD,
which replaces a tryptophan residue at position 261 with a cysteine
(W261C) (9). The Jackson dwarf mice, on the other hand,
have a genomic rearrangement resulting in a truncated POU1F1 protein
that has lost its DNA-binding capacity (9). To date, 12
mutations (9 missense, 2 nonsense, and 1 deletion) have been described
in the human POU1F1 gene (10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22). Two of these are
located in the trans-activation domain, 6 in
POUS, and 4 in POUHD. They
result in partial or total deficiency of GH and PRL and, to a variable
degree, of TSH. In some patients anterior pituitary hypoplasia is
evident on radiographic imaging of the hypothalamic-pituitary area.
Dependent on their localization in the POU1F1-coding sequence, the
mutations may interfere with either DNA binding or the
trans-activation process. Most are transmitted as an
auto-somal recessive trait, but four of them, two in the
trans-activation domain (P14L and P24L) and the other two
located at the borders of the POU homeodomain (K216E and R271W), result
in a dominant negative phenotype with a highly variable level of
penetrance.
In this report we describe a boy with CPHD who was found to be a compound heterozygote for two novel mutations in the POU1F1 gene. Both parents, who have a normal phenotype, harbor these mutations in the heterozygous state. The maternal allele carries a missense mutation in the POUS domain resulting in complete abolishment of DNA binding, whereas the paternal allele harbors a 1-bp deletion frameshift mutation, the first described to date in the POU1F1 gene, resulting in a truncated POU1F1 molecule missing helix 3 of the POUHD.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Messenger ribonucleic acid was isolated from Epstein-Barr virus-transformed lymphocytic cell lines from the proband, his parents, and his brother. This messenger ribonucleic acid was reverse transcribed, and the POU1F1 POU domain was amplified by nested PCR, essentially using the procedure and oligonucleotide primers described by Pfäffle et al. (10). The PCR products were sequenced bidirectionally using the Amplicycle sequencing kit (Perkin-Elmer Corp., Norwalk, CT). Genomic DNA was isolated from the lymphocytic cell lines, and all six POU1F1 exons were amplified separately by PCR using the pairs of oligonucleotide primers corresponding to the intron/exon boundaries described by Ohta et al. (14), with modifications essentially as described by Pellegrini-Bouiller et al. (19). The PCR products were analyzed by single strand conformation polymorphism (Genephor, Amersham Pharmacia Biotech, Arlington Heights, IL) and used for direct sequencing.
Plasmids
For the construction of POU1F1 POU domain expression vectors,
wild-type and mutant complementary DNA (cDNA) carrying the W193R
mutation in the POU1F1 POUS domain were obtained
by RT-PCR as described above. The PCR products were ligated into the
original TA cloning vector, pCR 2.1 (Invitrogen, San
Diego, CA), and transformed to Escherichia coli DH5.
Subsequently, both cDNAs were cloned into an
NdeI/BamHI-digested pET15b expression vector
(Novagen, Madison, WI), yielding
His6-tagged POU1F1 POU domain open reading
frames. The DNA sequence was checked by dideoxy chain termination
sequencing (Invitrogen).
The GH320-luc and PRL DE/P-luc reporter/luciferase constructs, containing the rat (r) GH promoter sequence and the rPRL distal enhancer and promoter sequences, respectively, were reported previously (10) and were gifts from Dr. Rosenfeld, Howard Hughes Medical Institute, University of California-San Diego (La Jolla, CA). For construction of a TSHß-luc reporter/luciferase construct, the GH320 insert was replaced by a 280-bp insert derived from the human (h) TSHß promoter [nucleotides (nt) -206 to +74) (23) containing three putative POU1F1-binding sites (24). This fragment was obtained by one round of PCR (36 cycles of 45 s at 95 C, 1 min at 57 C, and 2 min at 72 C) on 100 ng genomic DNA as template, using 200 pmol each of the oligonucleotide primers 5'-GAGAGGAAAATGCATGCTTT-3' and 5'-TATCATTTCACAGAGCCTTC-3'. The fragment was cloned into the pCRII-TOPO TA cloning vector (Invitrogen) and sequenced.
For use in the transfection assays, plasmid cytomegalovirus (pCMV)-POU1F1 expression plasmids were constructed as follows. POU1F1 wild-type and mutant cDNAs, encompassing the entire POU1F1-coding sequence, were obtained by a 1-side nested RT-PCR using 2 different upstream primers (5'-TGATTTGGGGAGCAGCGGTT-3' and 5'-CTACTCTCTTGTGGGAATGAG-3', respectively) and 1 downstream primer (5'-ATACAATAGAAAACTTTATCTGCACTC-3') in 2 consecutive rounds of PCR, each consisting of 36 cycles of 30 s at 95 C, 1 min at 58 C, 2 min at 72 C. We constructed a total of 4 different pCMV constructs: pCMV-POU1F1 wild-type cDNA, pCMV-POU1F1 cDNA containing the W193R mutation, pCMV-POU1F1 cDNA containing the 747delA mutation, and a pCMV vector without insert to be used as a control effector plasmid. All constructs were cloned into pTargeT (Promega Corp., Madison, WI) and sequenced bidirectionally. A Rous sarcoma virus-ß-galactosidase construct was used as an internal control for transfection efficiency in transient transfection experiments, as described previously (25).
Protein expression and purification
Wild-type and W193R mutant POU1F1 POU domain expression vectors were transformed to strain BL21 (pLYS). Strains containing the wild-type and mutant vectors were grown in 1-L cultures at 37 C and room temperature, respectively. At OD600 0.5, expression was induced by adding 1 mL 1 mol/L isopropylthio-ß-D-galactoside. Wild-type POU1F1POU expression was continued for 3 h at 37 C, whereas W193R POU1F1POU was expressed overnight at room temperature. Cells were pelleted, resuspended in 20 mL sonification buffer (50 mmol/L NaPO4, 300 mmol/L NaCl, 0.5 mmol/L phenylmethylsulfonylfluoride, 1 µg/mL aprotinin, and 10 mmol/L ß-mercaptoethanol) and lysed by freeze-thawing and mild sonification. Insoluble components were removed by centrifugation in an SW41 rotor at 35,000 rpm for 45 min. Wild-type and W193R POU1F1 POU domain proteins were partially purified on nickel-nitrilotriacetic acid agarose (QIAGEN, Chatsworth, CA) columns. Samples were estimated to be approximately 70–80% pure on a Coomassie-stained gel.
Gel retardation assay
For preparation of the probe, the GH320-luciferase construct
described above was digested by HindIII and end-labeled
using the Klenow fragment of DNA polymerase I (Amersham Pharmacia Biotech) and [-32P]deoxy-CTP
(Amersham Pharmacia Biotech; 10 mCi/mL, 3000 Ci/mmol),
followed by XhoI digestion. The resulting 300-bp fragment
containing the rGH promoter was purified by PAGE. POU1F1 dilutions were
made in 50 mmol/L Tris-HCl (pH 7.5), 1 mmol/L ethylenediamine
tetraacetate (EDTA; pH 8.0), 100 mmol/L NaCl, 10% glycerol, and 10
mmol/L ß-mercaptoethanol. Approximately 2 fmol DNA were incubated
with POU1F1 dilutions in a reaction mixture containing 20 mmol/L
HEPES-KOH (pH 7.5), 100 mmol/L NaCl, 1 mmol/L EDTA (pH 8.0), 1 mmol/L
dithiothreitol, 0.025% Nonidet P-40, 1 µg poly(dI-dC) competitor
DNA, and 4% Ficoll for 30 min at room temperature. Before loading, 2
µL 0.02% bromophenol blue and 0.02% xylene cyanol were added. The
samples were run on a 6% polyacrylamide gel containing 0.01% Nonidet
P-40 at 4 C for 3 h at 30 mA. The gel was dried, and the DNA was
visualized by autoradiography.
Deoxyribonuclease I (DNase I) footprint assay
POU1F1 dilutions were prepared as described above. Approximately 10 fmol DNA were incubated with POU1F1 as in the gel retardation assay, with an additional 1 µg BSA, 10 mmol/L MgCl2, and 3 mmol/L CaCl2/reaction. After 30 min at room temperature, samples were incubated with 0.01 U DNase I for 5 min. Reactions were quenched by adding 11 µL 10 µg/mL herring sperm DNA, 200 mmol/L NaAc (pH 8.0) and 75 mmol/L EDTA (pH 8.0). The DNA was purified by phenol/chloroform extraction before ethanol precipitation. Dried samples were resuspended in loading buffer (80% formamide, 0.1% bromophenol blue, and 0.1% xylene cyanol) and loaded on a 6% polyacrylamide sequencing gel (8 mol/L urea). The gel was run for 1.5 h at 30 mA and dried, and the DNA was visualized by autoradiography.
Cotransfection assays
The adenovirus-transformed human embryonic kidney (HEK) 293 cells (26) were cultured as a monolayer in DMEM containing 10% FCS, 3.5 µmol L-glutamine, 100 U penicillin, and 100 µg streptomycin/mL.
Cells were transfected when they were approximately 50% confluent using the calcium phosphate-DNA coprecipitation technique (27) in N,N,bis[2-hydroxyethyl]-2-aminoethanesulfonic acid (Sigma, St. Louis, MO)-buffered saline. Each 25-cm2 flask was transfected with 3 µg reporter plasmid and 500 ng plasmid Rous sarcoma virus-lacZ to normalize for transfection efficiency. The optimal quantity of wild-type POU1F1 effector plasmid to be used in combination with each reporter construct was determined by titration in such a way that the transfection potency was still in the linear range; for the hTSHß and rPRL reporter constructs a total of 500 ng effector were used, whereas for the rGH reporter 25 ng sufficed. Four hours after adding the precipitate to the cells, the medium was changed, and the cells were harvested 22 h thereafter. Luciferase and ß-galactosidase assays were performed as previously described (25, 28). Luciferase data were divided by the galactosidase activity to correct for transfection efficiency. All transfections were performed in duplicate in at least three separate experiments.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our patient, a boy, was born in India and came to The Netherlands at the age of 4 months. He is the second child of unrelated healthy parents with normal stature; he has a healthy and normally growing brother. As dictated by the Dutch neonatal screening program for congenital hypothyroidism (CH) he was screened as yet for CH shortly after arrival in our country; a heel puncture T4 of only 8 nmol/L was found, together with undetectable TSH and a normal T4-binding globulin concentration. Physical examination at the age of 4.5 months revealed typical signs of CH, with a low nasal bridge, macroglossia, facial myxedema, and wide open fontanels. He exhibited generalized hypotonia, slight peripheral myxedema, constipation, and hypothermia. His length at that point was 51 cm (-9 SD), and his body weight was 4100 g.
The first laboratory results showed the combination of a very low free
T4 (2.8 pmol/L) and TSH (<0.1 mU/L)
concentrations, low insulin-like growth factor I (5 ng/mL) and
insulin-like growth factor-binding protein-3 (0.3 mg/L) levels, and
undetectable PRL concentrations (<1 µg/L). Plasma cortisol (580
nmol/L) and ACTH concentrations (53 ng/L) were normal, and testosterone
was appropriate for age (2.3 nmol/L). After an iv injection of 30 µg
TRH, all TSH levels remained below 0.05 mU/L, and PRL remained below
1.0 µg/L. ACTH and cortisol plasma concentrations rose normally after
iv administration of 40 µg CRH. After reaching euthyroidism with
appropriate T4 treatment, an arginine provocation
test was performed. Basal GH levels were undetectable and remained
below 1 mU/L after an iv infusion of 0.5 mg/kg arginine. A magnetic
resonance imaging scan of the hypothalamic-pituitary region showed a
hypoplastic anterior pituitary but otherwise normal anatomy (Fig. 1
). The boy was started on daily GH
injections to which he responded well; he is now 3.5 yr of age and has
attained a height of 98.3 cm (-1.1 SD). He is in good
health and shows no signs of neurodevelopmental delay.
|
Two novel mutations in the POU1F1 gene. Sequence analysis of
lymphocyte-derived POU1F1 cDNA encompassing the
POUS and POUHD regions
(10) revealed heterozygosity in the proband and his mother
for a missense mutation at position 577 in exon 4, a TC transition
that changed amino acid 193 from Trp to Arg (W193R). As the mother had
a normal phenotype as well as a normal hormonal profile, a dominant
negative effect of this mutation was considered unlikely. Further
investigations therefore aimed at identifying an additional mutation in
the proband’s paternal POU1F1 allele. To this end, direct genomic
sequencing of all six exons of the POU1F1 gene from both parents, the
proband, and his brother was performed, which confirmed the
heterozygosity in exon 4 in the proband and his mother, and identified
an additional heterozygosity in the DNA of the proband and his father
for a 1-bp deletion at position 747 in exon 6, codon 249 (747delA, the
A nucleotide of the ATG codon of the primary translation product being
taken as position +1) The shift in the reading frame resulting from
this deletion changes the subsequent codon from Glu to Asn and
introduces a translational stop codon immediately thereafter.
Thus, our patient is heterozygous for a missense mutation in exon 4 (W193R) inherited from the mother and a 1-bp deletion frameshift mutation in exon 6 (747delA) inherited from the father. The effect of the frameshift mutation, a loss of helix 3 of POUHD, is very similar to the E250X nonsense mutation described by Irie et al. (20), which in the homozygous state led to CPHD in their patient. As the W193R missense mutation had not been described previously, the properties of the resulting mutant POU1F1 protein with respect to DNA binding and trans-activation were further analyzed.
W193R binds DNA with approximately 500-fold reduced affinity in
vitro. DNA binding affinity of the wild-type and W193R POU1F1 POU
domain were tested in vitro using the POU1F1-binding site of
the rGH promoter as probe. As shown in Fig. 2, the bacterially expressed wild-type
POU1F1 POU domain bound with high affinity, whereas binding of the
W193R POU1F1 POU domain to the same site was about 500-fold reduced.
Similar results were obtained on the PRL proximal enhancer (data not
shown). Residual DNA binding affinity can be attributed to the POU
homeodomain, which is still intact in the W193R mutant.
|
, the wild-type POU1F1 POU domain was
able to protect a region of 17 bp of the rGH promoter, whereas the
W193R mutant did not show any protection, even at high protein
concentrations. We can therefore conclude that DNA binding of the W193R
mutant is seriously impaired, which explains its lack of
trans-activation capability.
|
, when either one of the mutant expression constructs was transfected
together with the wild-type construct, the trans-activation
potency declined to approximately 60–70% of the wild-type level. This
situation reflects the genetic background of the parents, and the
results are consistent with the fact that neither parent showed an
abnormal phenotype. When both mutants were cotransfected,
trans-activation dropped to background level. From these
results we conclude that neither mutant is able to activate
transcription, which explains the minimal levels of GH, PRL, and TSH
found in the patient.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our patient was a compound heterozygote for two novel mutations in the POU1F1 gene. The 1-bp deletion mutation in the POU homeodomain at position 747 (747delA), transmitted by the paternal allele, is the first frameshift mutation in the POU1F1 gene described to date. The frameshift causes a change in the codon following E249 from Glu to Asn and introduces a translational stop codon immediately thereafter. This mutation is therefore almost identical to the E250X nonsense mutation described by Irie et al. (20). In either case the translational stop codon is located at the C-terminal end of helix 2 of the POU homeodomain. In the homozygous state, they will undoubtedly result in a severe loss of DNA binding, as the entire helix 3 with the DNA recognition domain of POUHD is deleted.
The TC missense mutation at position 577, conferred by the maternal
allele, predicts a tryptophan to arginine substitution at codon 193
located in the C-terminal end of the fourth -helix of the
POU-specific domain. The natural occurrence of this W193R mutation has
not been reported previously, but it has been reported in a yeast
in vivo screening model for DNA-binding negative POU1F1
mutants (32), in which it demonstrated only 2% of the
DNA-binding activity of the wild-type protein.
Figure 5 shows the crystal structure of
POU1F1 (7), with tryptophan 193 highlighted. This figure
clearly shows that W193 is one of the amino acid residues that make up
the hydrophobic core of the POU-specific domain. The -helix
harboring this codon can be considered a structural helix, as it does
not contact the DNA directly. However, changing W193 to arginine
introduces a positively charged residue into the hydrophobic core. Our
bandshift and footprint assays show that the W193R POU domain is no
longer able to bind to DNA with sufficient affinity, probably due to
improper protein folding of the POU-specific domain. As a result, the
mutant W193R POU domain is unable to activate transcription, as shown
in our transient transfection experiments. The residual POU1F1 activity
of approximately 60–70% in the heterozygous state as measured in our
transient transfection assays can be expected to ensure a normal
phenotype, as others have observed a clinically recessive phenotype
with residual activities of approximately 50% in comparable assays
(22).
|
). Study of other patients with CPHD
will undoubtedly result in a further increase in this number of
mutations in POU1F1 or in related genes such as PROP-1 (33, 34). However, as a significant number of patients still turn out
to have CPHD despite normal POU1F1- and PROP-1-coding sequences
(Pfäffle, R., personal communication), the molecular unraveling
of these defects remains a challenge for the future.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 B.I.H.-S. and K.D.A. contributed equally to this study.
Received January 13, 2000.
Revised May 3, 2000.
Accepted December 21, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  I. Miyata, S. Vallette-Kasic, A. Saveanu, M. Takeuchi, H. Yoshikawa, A. Tajima, K. Tojo, R. Reynaud, M. Gueydan, A. Enjalbert, et al. Identification and Functional Analysis of the Novel S179R POU1F1 Mutation Associated with Combined Pituitary Hormone Deficiency J. Clin. Endocrinol. Metab., December 1, 2006; 91(12): 4981 - 4987. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R A Sporici, J S Hodskins, D M Locasto, L B Meszaros, A L Ferry, A M Weidner, C A Rinehart, J C Bailey, I M Mains, and S E Diamond Repression of the prolactin promoter: a functional consequence of the heterodimerization between Pit-1 and Pit-1 {beta} J. Mol. Endocrinol., October 1, 2005; 35(2): 317 - 331. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. van Tijn, J. J. M. de Vijlder, B. Verbeeten Jr., P. H. Verkerk, and T. Vulsma Neonatal Detection of Congenital Hypothyroidism of Central Origin J. Clin. Endocrinol. Metab., June 1, 2005; 90(6): 3350 - 3359. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Hashimoto, M. Cisternino, and L. E. Cohen A Novel Nonsense Mutation in the Pit-1 Gene: Evidence for a Gene Dosage Effect J. Clin. Endocrinol. Metab., March 1, 2003; 88(3): 1241 - 1247. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kishimoto, Y. Okimura, K. Yagita, G. Iguchi, M. Fumoto, K. Iida, H. Kaji, H. Okamura, and K. Chihara Novel Function of the Transactivation Domain of a Pituitary-specific Transcription Factor, Pit-1 J. Biol. Chem., November 15, 2002; 277(47): 45141 - 45148. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. E. Cohen and S. Radovick Molecular Basis of Combined Pituitary Hormone Deficiencies Endocr. Rev., August 1, 2002; 23(4): 431 - 442. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
