0021-972X/99/$03.00/0 The Journal of Clinical Endocrinology & Metabolism Copyright © 1999 by The Endocrine Society
Vol. 84, No. 5 Printed in U.S.A.
“Hot Spot” in the PROP1 Gene Responsible for Combined Pituitary Hormone Deficiency* ¨ Y, CHRISTA FLU ¨ CK, ATILLA BU ¨ YU ¨ KGEBIZ, JOHNNY DELADOE ´ ´ BEATRICE V. KUHLMANN, ANDREE EBLE, PETER C. HINDMARSH, WEI WU, PRIMUS E. MULLIS
AND
Division of Pediatric Endocrinology, University Children’s Hospital (J.D., C.F., B.V.K., A.E., P.E.M.), 3010 Bern, Switzerland; Dokuz Eylu¨l Faculty of Medicine (A.B.), 35340 Izmir, Turkey; Cobbold Laboratories, Middlesex Hospital (P.C.H.), London, United Kingdom W1N 8AA; and Howard Hughes Medical Institute, University of California-San Diego (W.W.), La Jolla, California 92093-0648 ABSTRACT As pituitary function depends on the integrity of the hypothalamicpituitary axis, any defect in the development and organogenesis of this gland may account for a form of combined pituitary hormone deficiency (CPHD). Although pit-1 was 1 of the first factors identified as a cause of CPHD in mice, many other homeodomain and transcription factors have been characterized as being involved in different developmental stages of pituitary gland development, such as prophet of pit-1 (prop-1), P-Lim, ETS-1, and Brn 4. The aims of the present study were first to screen families and patients suffering from different forms of CPHD for PROP1 gene alterations, and second to define possible hot spots and the frequency of the different gene alterations found. Of 73 subjects (36 families) analyzed, we found 35 patients, belonging to 18 unrelated families, with CPHD caused by a
C
OMBINED pituitary hormone deficiency (CPHD) denotes impaired production of GH and one or more of the other five anterior pituitary-derived hormones. Although pit-1 was one of the first factors identified as a cause of CPHD in mice, many other homeodomain and transcription factors, such as prophet of pit-1 (prop-1), P-Lim, ETS-1, and Brn 4, have been characterized as being involved in different stages of pituitary gland development and therefore, if altered, might cause CPHD (1– 6). For instance, a mutation in the prop-1 gene, a tissue-specific, paired-like, homeodomain transcription factor, was found to cause the Ames dwarf (df) mouse phenotype, and subsequently in humans, different PROP1 gene alterations (mutations and deletions) have been reported that are responsible for CPHD, with deficiencies in GH, PRL, TSH, LH, and FSH (7–11). The aims of the present study were first to screen families and patients suffering from different forms of CPHD for PROP1 gene alterations and to determine the frequency of the different gene alterations, and second to define possible “hot spots” within the gene associated with this phenotype. Of 73 subjects analyzed, we found 35 patients with CPHD
Received December 3, 1998. Revision received February 4, 1999. Accepted February 8, 1999. Address all correspondence and requests for reprints to: Prof. Dr. Primus E. Mullis, Department of Pediatrics, Pediatric Endocrinology, Inselspital, CH-3010 Bern, Switzerland. E-mail:
[email protected]. * This work was supported by Pharmacia & Upjohn (Du¨bendorf, Switzerland), Novo-Nordisk (Ku¨snacht, Switzerland), and Swiss National Science Foundation Grant 32.053714.98.
PROP1 gene defect. The PROP1 gene alterations included 3 missense mutations, 2 frameshift mutations, and 1 splice site mutation. The 2 reported frameshift mutations could be caused by any 2-bp GA or AG deletion at either the 148-GGA-GGG-153 or 295-CGA-GAG-AGT-303 position. As any combination of a GA or AG deletion yields the same sequencing data, the frameshift mutations were called 149delGA and 296delGA, respectively. All but 1 mutation were located in the PROP1 gene encoding the homeodomain. Importantly, 3 tandem repeats of the dinucleotides GA at location 296 –302 in the PROP1 gene represent a hot spot for CPHD. In conclusion, the PROP1 gene seems to be a major candidate gene for CPHD; however, further studies are needed to evaluate other genetic defects involved in pituitary development. (J Clin Endocrinol Metab 84: 1645–1650, 1999)
caused by a PROP1 gene defect. There were 3 different missense, 2 frameshift, and 1 splice site mutations resulting in the disorder. Although the occurrence of hormonal deficiency varies from patient to patient, even among those with the same gene mutation, the affected patients as adults were not only GH, PRL, and TSH deficient, but were also gonadotropin deficient (9). Subjects and Methods Patients Thirty-six families with a total of 73 affected patients (41 males and 32 females) diagnosed with CPHD were studied. The parents of the children studied were unaffected and were of normal height and weight for age and sex (12). The criteria for CPHD included impaired production of GH and, in addition, a lack of 1 or more of the other hormones derived from the pituitary anterior gland. Although all CPHD patients were GH, PRL, TSH, and gonadotropin deficient in adulthood, there was a high phenotypic variability among these patients during infancy, childhood, and adolescence. Therefore, the inclusion criteria for this study, as far as the occurrence of the different hormonal deficiencies was concerned, were relatively open. Standard auxological assessment was performed (13). None of the affected patients had evidence of an organic disease, psycho-social deprivation, or any eating disorder, and all had normal renal and hepatic function. Informed consent was obtained from parents and all family members studied.
Hormonal assays Study patients received GH provocative testing using only arginine stimulation before the age of 2 yr; thereafter, both arginine (0.5 mg/kg; Pharmacia & Upjohn, Stockholm, Sweden) and insulin-induced hypoglycemia (ITT; 0.15 U/kg, iv; Novo-Nordisk, Gentofte, Denmark), and
1645
1646
¨ Y ET AL. DELADOE
adequate hypoglycemia (blood glucose, ,2.0 mmol/L) were achieved in all cases. The stress of ITT normally results in an increased secretion of ACTH and, hence, cortisol. Basal cortisol and ACTH were measured before ITT, and peak values were assessed after the response to ITT. This test was combined with the exogenous administration of TRH (200 mg; Ferring, Malmø, Sweden) and GnRH (100 mg; Ferring) to test the remainder of the hypothalamic-pituitary axis. The criteria for GH deficiency included a peak GH level of less than 2 ng/mL after stimulation (arginine stimulation and ITT) and a height velocity below 22.5 sd score (12).
Stimulation tests The stimulation tests were performed as described previously (14). Over the years, GH, TSH, LH, FSH, and PRL were measured using various assays, as previously described (9). However, we retested some of the samples using the methods that are now in use in our laboratory and compared the data with the results obtained using the “old” test procedures. The correlations among the different tests were between r 5 0.81 and 0.96. Therefore, we state in detail the actual test procedures only.
GH assays GH was measured using an immunoradiometric assay, HGH MAIAclone (Biodata Diagnostics, Freiburg, Germany), which incorporates two high affinity monoclonal antibodies. The interassay coefficients of variation (CVs) were 2.3%, 2.4%, and 2.2% at 2, 9, and 24 ng/mL, respectively. The intraassay CVs were 2%, 1.7%, and 1.7% at 2, 9, and 24 ng/mL.
JCE & M • 1999 Vol 84 • No 5
10 min at 68 C. Thereafter, PCR products were purified by gel electrophoresis followed by agarose gel DNA extraction. Direct sequencing of the double-stranded PCR fragments was carried out according to the thermal cycle sequencing protocol (PE Applied Biosystems, 373 DNA Sequencer, Perkin Elmer, Rotkreuz, Switzerland) using a 59-sense primer (59-TCTGGCCATGCTGAGAAG-39) and a 39-antisense primer (59TTCTAGTCGCTGAGCTGAC-59). To sequence the exons individually, exon-specific primers were designed (Table 1).
Screening method for a 2-bp deletion (GA or AG) at nucleotide (nt) location 296 –302 A diagnostic screening test was used to define a 2-bp deletion (GA or AG) at nt position 295-CGA-GAG-AGT-303. As this location belongs to exon 2, exon 2 was PCR amplified (Table 1). Thereafter, samples of the PCR-amplified product were digested to completion with the restriction enzyme BcgI under the conditions recommended by the commercial supplier (BioLabs, Bioconcept, Allschwil, Switzerland). Any GA or AG deletion within that region introduces a BcgI restriction site, and therefore, digestion of exon 2 PCR products confirms the presence of an altered allele. After electrophoresis in ethidium bromide-stained 2% (wt/vol) Metaphor gel (BioLabs), the DNA fragments were photographed by UV transillumination. Having found an altered PROP1 allele, the deletion was confirmed by sequencing as described above. Furthermore, the frameshift mutations were called 296delGA.
Controls The PROP1 genes of 28 unrelated normal control individuals were directly sequenced to evaluate nt variations.
TSH assay TSH was measured by a TSH assay using automated direct chemiluminometric methodology (Chiron Diagnostics Corp., East Walpole, MA). Inter- and intraassay CVs were 6.1% and 4.3% at 0.3 mIU/L and 5.2% and 5.8% at 4.7 mIU/L, respectively.
LH and FSH assays FSH and LH were measured using Dade fluorometric enzyme immunoassays (Stratus, Dade, Miami, FL). For FSH, inter- and intraassay CVs were 1.6% and 3.8% at 4 IU/L and 2.7% and 2.9% at 20 IU/L. Cross-reactivities between LH and TSH were 0.5% and 0.01%, respectively. For LH, inter- and intraassay CVs were 7.5% and 5.8% at 2.5 IU/L and 2% and 2.7% at 20 IU/L. Cross-reactivities between TSH and FSH were 0.02% and 0.001%, respectively.
PRL assay PRL was measured using a fluorometric enzyme immunoassay (Stratus, Dade). Inter- and intraassay CVs were 5% and 4.3% at 4.5 mg/L and 4.8% and 3.2% at 59 mg/L, respectively.
ACTH and cortisol assays ACTH measurements were performed using a immunometric assay (Nichols Institute Diagnostics, San Juan Capistrano, CA). The interassay CVs were 4.6% and 7.0% at 8.3 and 380 pg/mL, respectively. The intraassay CVs at 6.6 and 293 pg/mL were 3.4% and 3.8%. Cortisol was measured by fluorometric enzyme immunoassay (Stratus). The interassay CVs were 4.6%, 4.1%, and 4.9% at 115, 300, and 1000 nmol/L, respectively. The intraassay CVs at 115, 300, and 100 nmol/L were 7.2%, 5.1%, and 4.0%.
Genomic analysis of PROP1 gene DNA was extracted from the peripheral lymphocytes as previously reported (15). One hundred nanograms of human genomic DNA were used as template in a 20-mL PCR. The coding sequence of PROP1 was PCR amplified with a 59-sense primer (59-CGAACATTCAGAGACAGAGTCCCAGA-39) and a 39-antisense primer (59-GAATTCACCATGATCTCCCA-39) to generate a 3.5-kb fragment. The reaction consisted of 1 min at 94 C, followed by 35 cycles of 30 s at 94 C, 30 s at 56 C, and
Results PROP1 gene studies
Of the 36 families with 73 CPHD-affected patients, 35 patients (18 males and 17 females; 35 of 73; 48%) belonging to 18 unrelated families (18 of 36; 50%) were found to have PROP1 gene defects responsible for the disorder. The locations of the PROP1 gene alterations are shown in Fig. 1, and the frequencies are summarized in Table 2. The frequencies of given PROP1 gene alterations among CPHD-affected subjects are stated as percentages of the number of these patients (n 5 35) and their families (n 5 18). However, if one affected family with a PROP1 gene defect is large and the others small, then the frequency of the mutation in the large family is made artificially high; therefore, our results are also expressed as a percentage of the total number of patients (n 5 73), considering several affected patients from the same family as 1. For example, three affected patients in 1 family would be considered as 1 of a now decreased total number of patients (n 5 71; 1 of 71 5 2%; Table 2). TABLE 1. Sequence of PCR primers used for amplification of the 59- and 39-untranslated and intronic sequences flanking each of the three coding exons of PROP1 gene Region
Exon 1 Sense Antisense Exon 2 Sense Antisense Exon 3 Sense Antisense
Sequence
TTCAGAGACAGAGTCCCAGA CTCCTAACCTTCTTCATGGA GACAGGCACATGTGGTCCA AGGCCTGTGTCTGGTGACCA TCTGGCCATGCTGGAGAAG TTCTAGTCGCTGAGCTGAC
PROP1 GENE “HOT SPOT”
1647
FIG. 1. Schematic representation of the genomic organization of the human PROP1 gene. Exons are depicted as hatched (translated) and open (untranslated) boxes. The homeodomain is shown in dark shading. Translated nt and corresponding amino acids are numbered. The mutations/ deletions found are represented at their locations, and the corresponding changes in the PROP1 protein are shown. The exons and introns are drawn to scale. TABLE 2. Frequencies of PROP1 gene alterations Frequencies Type of mutation
Location
Nucleotide change
Amino acide change
Missense
Exon 2 Exon 3 Exon 3
CGC 3 TGC; C217T TTC 3 ATC; T349Ae CGC 3 TGC; C358Tf
R73C F117I R120C
Frame shift
Exon 2 Exon 2
149 del GAg 296 del GAh
S109X S109X
Splice site
Intron 2
Nucleotide 343-2 (A to T)
None
Patientsa (n 5 35)
Patientsb (n 5 73)
Familiesc (n 5 18)
Familiesd (n 5 36)
2 (5.7) 3 (8.6) 5 (14.3)
1/72 (1.4) 1/71 (2) 2/70 (2.9)
1 (5.5) 1 (5.5) 2 (11)
1 (2.7) 1 (2.7) 2 (5.5)
3 (8.6) 21 (60)
1/71 (2) 12/64 (18.8)
1 (5.5) 12 (66)
1 (2.7) 12 (33)
1 (2.8)
1/73 (1.4)
1 (5.5)
1 (2.7)
Frequencies of given PROP1 gene alterations among CPHD-affected subjects are stated as percentages of the number of these patients (a) and their families (c). However, if one affected family with a PROP1 gene defect is large and the others are small, then the frequency of the mutation in the large family is made artificially high; therefore, our results are also expressed as a percentage of the total number of patients (n 5 73) considering several affected patients from the same family as one (b; see Results). Percentages are given in parentheses. d Frequency among all the CPHD families studied. e This missense mutation has been previously reported (7). f This mutation was reported by Wu et al. (7, 9). g,h The precise location of the 2-bp GA or AG deletion in the 148-GGA-GGG-153 or 295-CGA-GAG-AGT-303 sequence is ambiguous. Any combination of a GA or AG deletion could yield the same results. Therefore, the deletions are called either 149delGA (8) or 296delGA (7, 8).
easily be used to screen CPHD patients. All data obtained were confirmed by direct sequencing. New mutations
FIG. 2. Photograph of an ethidium bromide-stained 2% Metaphor gel after electrophoresis of BcgI-digested PCR products following amplification of exon 2. Data of a family with a homozygous 296delGA are shown. Two normal subjects (one female and one male) were added as normal controls (C). Squares, Males; circles, females; filled, homozygosity; half-filled, heterozygosity for the deletion. 296delGA introduces a BcgI restriction site that allows use of this procedure as a screening method. The PCR amplification product of exon 2 is 410 bp in size. The amplification products of a 296delGA-deleted PROP1 alleles are digested to yield 255- and 155-bp fragments. A marker (BioLab., marker V) is shown on the left.
As indicated in Table 2, some of the PROP1 gene alterations have been reported before. However, in this study two new mutations were found. First, a nt C to T transition (CGC3 TGC; C217T) resulted in the substitution of Arg3 Cys at codon 73 (exon 2). Second, a nt A to T transition at the nt 2–343 position introduced an intronic point mutation. Therefore, the invariant GA nt of the splice-acceptor site was destroyed. These two new mutations are depicted in Fig. 3. Clinical evaluation
In patients presenting with any form of PROP1 gene alteration, the development of hormonal deficiencies was studied in great detail. Based on a great variability in phenotype, the secretion of pituitary-derived hormones (GH, TSH, LH, and FSH) declined gradually with age, following a different pattern and time scale in each individual (Fig. 4).
Screening for 296delGA
Hormonal deficiencies and ACTH reserve
As depicted in Fig. 2, BcgI digestion of exon 2 PCR amplification products revealed the 296delGA of the PROP1 based on the restriction digest pattern. This method can
Whereas patients with the PROP1 gene defect presented eventually with a lack not only of GH, PRL, and TSH, but also of LH and FSH (Fig. 4), the cortisol production was normal
1648
¨ Y ET AL. DELADOE
JCE & M • 1999 Vol 84 • No 5
FIG. 3. Genomic DNA sequence analysis of the two PROP1 gene mutations are shown. a, Exon 2: missense mutation; a nt C to T transition (CGC3 TGC, C217T) resulted in the substitution of Arg3 Cys at codon 73. b, Intron 2: spice site mutation; a nt A to T transition at the nt 2–343 position yielded an intronic point mutation that destroyed the invariant AG nt of the splice-acceptor site.
FIG. 4. The development of the hormonal deficiencies in patients with PROP1 gene defects is summarized. On the left, the data are given in percentages; on the right, the numbers of patients are shown. The time of the occurrence of the specific hormonal deficiency (GH, TSH, and LH/FSH) is indicated. Although as adults all patients were not only GH and TSH, but also gonadotropin, deficient, the phenotype of CPHD developed gradually over the years.
(mean, 635 nmol/L; range, 423–931 nmol/L) after insulininduced hypoglycemia (mean, 1.2 mmol/L; range, 0.5–1.7 mmol/L). It is important to stress, however, that in some of the patients (n 5 7; three males and four females), low levels of cortisol (mean, 120 nmol/L; range, 64 –178 nmol/L) and ACTH (mean, 5 pg/mL; range, 5–21 pg/mL) were measured at the beginning of the test. Polymorphic sites
In addition, 28 unrelated normal control individuals were screened for nt variations within the PROP1 gene. None of the 56 chromosomes analyzed carried a mutation/deletion as reported in subjects with CPHD. However, 3 polymorphic sites in the PROP1 gene were detected, 1 located in intron I which is untranslated and the other 2 located in exons I and III (Table 3). The polymorphic site in exon 1 is located at codon 9 (nt 27, GCT3 GCC), yielding an identical substitution of alanine3alanine (A9A). The site in exon 3 is at codon 142 (G to A transition at position 424; GCC3 ACC) and replaces alanine with threonine (A142T). This mutation is
located 14 amino acids outside of the homeodomain region (Fig. 1). In intron 1, the third nt can be either adenine or guanine and remains untranslated. The frequencies of the polymorphisms have been assessed and are presented in Table 3. Discussion
The development of the pituitary gland appears to share with many organs developmental control by the combinatorial actions of diverse classes of transcription factors, exerting positive and negative controls over gene expression, and sequentially dictating organ determination, cell type differentiation, and terminal differentiation of specific cell precursors (16 –18). Several tissue-specific factors have been identified, such as pit-1 (mouse, POU1F1 in human) and prop-1 (mouse; PROP1 in human). In the Snell (dw/dw) and Ames (df/df) dwarf mice, two recessive murine mutations causing dysorganogenesis of the pituitary gland have been described (5, 19). The Snell dwarf phenotype results from mutations in a gene for a pituitary-specific transcription fac-
PROP1 GENE “HOT SPOT” TABLE 3. Allele frequency of PROP1 gene polymorphisms in the control population Location
Nucleotide change
Codon change
Allele frequencya
Exon I Exon III Intron I
nt 27 (T to C; GCT 3 GCC) nt 424 (G to A; GCC 3 ACC) nt 109 1 3
A9A A142T None
26/56 (46.4) 12/56 (21.4) 26/56 (46.4)
nt, Nucleotides. Numbers of alleles, with percentages in parentheses.
a
tor, called pit-1 (19). pit-1, expressed in thyrotropes, somatotropes, and lactotropes, is required for continued expression of the pit-1 gene itself and for the proliferation and survival of these three cell types (20, 21). In the Ames mouse, Sornson et al. cloned and described a gene, prophet of pit-1 (prop-1), a pituitary-specific, paired-like homeodomain transcription factor that was shown to be important for determination of the pit-1 lineages as well as gonadotrope differentiation in the pituitary gland (5, 22). Therefore, patients with a PROP1 defect have a lack not only of GH, PRL, and TSH, but also of LH and FSH (7, 15). It is most important to stress that the clinical phenotype of CPHD varies considerably, especially with respect to the occurrence and the severity of the different pituitary-derived hormone deficiencies (9, 23). This variability impedes the identification of genes that could be used for screening, presenting a difficulty in defining the gene defects causing the disorder. Molecular studies have revealed that only a minority of patients presenting with CPHD can be accounted for by structural defects in the POU1F1 gene. PROP1, because of its broad phenotype, is more likely a candidate gene. The aim of this study was to define the prevalence and frequency of PROP1 gene alterations among patients suffering from different phenotypes of CPHD. To characterize the patients as CPH deficient, the time scale of the development of the hormonal deficiencies was not taken into consideration. All patients, however, were eventually GH, TSH, PRL, and gonadotropin deficient. The prevalence of PROP1 gene alterations among patients with CPHD was high. A PROP1 gene defect could be found in a total of 35 of 73 patients (48%), representing 18 of 56 (32%) nonrelated CPHD-affected subjects. Furthermore, these 35 patients with a PROP1 gene defect belonged to 18 of 36 nonrelated CPHD-affected families (50%). This specification of nonrelated patients and/or families is of importance, because if 1 family is very large and the others small, then the frequency of the reported mutation in a large family is made artificially high. However, these data suggest that the PROP1 gene is an important etiological factor in CPHD. There was no correlation between phenotype and genotype. The manifestation of the different deficiencies varied substantially, as has been reported even in patients with the same C to T transition resulting in the substitution of arginine3cysteine at codon 120 (R120C) (9). Moreover, all of the patients became symptomatic by exhibiting severe growth retardation and failure to thrive, mainly caused by GH deficiency (n 5 28; 80%); TSH deficiency was the first symptom in 20% (n 5 7). Thirteen patients (6 males and 7 females) entered puberty spontaneously. Although 6 girls experienced menarche, they were rather late, at a mean age
1649
of 15.2 yr (range, 14.8 –17.3 yr). The others, both males and females, needed hormonal induction of pubertal development. Seven patients (3 males and 4 females) presented low basal levels of cortisol and ACTH; however, the stimulated levels after insulin-induced hypoglycemia revealed no abnormality. Therefore, a slight impairment of basal ACTH secretion might be present in some of these patients, although none of the patients ever presented with any sign of hypocortisolism, nor was any replacement therapy necessary. As presented in Fig. 1 and Table 2, we found three missense mutations (n 5 10 belonging to 4 of 36 unrelated families; 11%), 2 frameshift mutations within well defined regions (n 5 24 belonging to 13 of 36 families; 36%), and 1 splice site mutation (n 5 1; 1 of 36; 2.7%) causing the CPHD phenotype. The missense mutation, C217T (R73C), and the splice site mutation, nt 343–2 (A to T), are new, whereas the other mutations/deletions have been previously reported (7–9). Importantly, all but 1 mutation (149delGA) were located in the gene fragment encoding the homeodomain of the PROP1 protein. In detail, one family presented a homozygous C to T transition at position 217 (C217T) at codon 73 in exon 2, predicting a missense mutations (R73C). This missense mutation replaces a charged residue by a neutral amino acid in a region highly conserved among all members of the paired-like class of homeodomains (24, 25). Moreover, this arginine residue, which is invariant in all of these proteins, is conserved in 95% of the more than 450 homeodomain proteins known to date (24, 25). In another family, a homozygous T to A transversion resulted in an F117I substitution (7). It has been reported that the phenyl-alanine (F) residue at codon 117 is part of the core DNA binding motif of homeodomains and that it is almost invariant within the homeodomain family (7). In two families, five patients presented the arginine3cysteine substitution at codon 120 (R120C, amino acid 52 in the third helix of homeodomain). None of the homeodomain proteins in this family have a cysteine residue at this position (7, 25). Furthermore, all of the frameshift mutations (149delGA; 296delGA) were caused by a 2-bp deletion in exon 2, yielding an aberrant translation product with a premature stop codon TAG at position 109 (S109X) in the same exon (7, 8). Therefore, in all of these patients the PROP1 protein is lacking 118 carboxyl-terminal amino acids. Preceding exon 3, an intronic point mutation (A to T transition) involving the spliceacceptor site was found. This mutation destroyed the invariant AG nt of the splice-acceptor site, which is crucial for normal splicing. In these patients, there is no splicing at the end of intron 2, but a cryptic splice site within exon 3 may be activated. Therefore, the translated protein lacks part of the homeodomain region, leaving it without any effect. Moreover, as it is well known that mutations occur disproportionately in CpG dinucleotides, two of the three missense mutations were caused by a CG3 TG transition (26). Furthermore, our data suggest that there is a hot spot region for PROP1 gene alterations. In 12 of 36 unrelated CPHD affected families (33.3%), the PROP1 gene defects found were located in the region nt 296 –302, which contains a series of three tandem repeats (GAGAGAG). In addition, as in 66% of all affected unrelated families (12 of 18) the same
¨ Y ET AL. DELADOE
1650
molecular defect, involving a deletion of 2 bp, namely AG or GA, within this region has been identified, it is suggested that this region is likely to represent a mutational hot spot. This hypothesis is underlined by the fact that such repeats, like classical microsatellite loci, are prone to mutations by slipped strand mispairing. Further, in the 28 unrelated normal controls screened for PROP1 gene polymorphisms, none of the 56 chromosomes analyzed presented any alteration of the PROP1 gene as reported in the affected patients. Polymorphisms were found in these normal controls as detailed in Table 3, but appear to be without functional impact, at least as far as CPHD is concerned. Although the PROP1 gene seems to be an important candidate gene for CPHD, the reason for the disorder remains unknown in many of the patients. As most of our understanding of pituitary development has come from rodent animal studies, it is likely that more information on the characterization of CPHD will be derived from the recent studies of CPHD phenotypes in mice generated by targeted disruption of other pituitary-specific transcription factors (27–32). There are recent data on the Hesx-1 gene causing forebrain midline defects with pituitary dysplasia in Hesx-1 null mutant mice as well as septo-optic dysplasia in patients identified as having a Hesx-1 mutation (33). Such animal models adapted and applied to human studies will contribute to define more and more genetic factors involved in pituitary development. In conclusion, we report six different PROP1 gene alterations that were found among 35 patients suffering from CPHD derived from 18 unrelated families. All but 1 mutation were located in the PROP1 gene encoding the homeodomain, which plays a major role in transcriptional regulation. Furthermore, three tandem repeats of the dinucleotide GA at the location nt 296 –302 appear to represent a hot spot for combined pituitary hormone deficiency. Although the PROP1 gene seems to be an important candidate gene for CPHD, further studies are needed to evaluate other genetic defects involved in pituitary development. Phenotypes of murine strains generated by targeted disruption of pituitary transcription factors might help to further elucidate the mechanisms resulting in CPHD in humans.
6. 7. 8. 9.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.
References 1. Bodner M, Karin M. 1987 A pituitary-specific trans-acting factor can stimulate transcription from the growth hormone promoter in extracts of non-expression cells. Cell. 50:267–275. 2. Voss JW, Rosenfeld. 1992 Anterior pituitary development: short tales from dwarf mice. Cell. 70:527–530. 3. Bach I, Rhodes SJ, Pearse RV, et al. 1995 P-Lim, a Lim homeodomain factor, is expressed during pituitary organ and cell commitment any synergizes with Pit-1. Proc Natl Acad Sci USA. 92:2720 –2724. 4. Szeto DP, Ryan AK, O’Connell SM, Rosenfeld MG. 1996 P-OTX: a Pit-1 interacting homeodomain factor expressed during anterior pituitary gland development. Proc Natl Acad Sci USA. 93:7706 –7710. 5. Sornson MW, Wu W, Dasen JS, et al. 1996 Pituitary lineage determination by
29. 30. 31. 32. 33.
JCE & M • 1999 Vol 84 • No 5
the Prophet of Pit-1 homeodomain factor defective in Ames dwarfism. Nature. 384:327–333. Bradford AP, Wasylyk C, Gutierrez-Hartmann A. 1997 Interaction of ETS-1 and the POU-homeodomain protein GHF/Pit-1 reconstitutes pituitary-specific gene expression. Mol Cell Biol. 17:1065–1074. Wu W, Cogan JD, Pfa¨ffle RW, et al. 1998 Mutations in PROP-1 cause familial combined pituitary hormone deficiency. Nat Genet. 18:147–149. Fofanova O, Takamura N, Kinishita E, et al. 1998 Compound heterozygous deletion of the PROP-1 gene in children with combined pituitary hormone deficiency. J Clin Endocrinol Metab. 83:2601–2604. Flu¨ck Ch, Deladoe¨y J, Rutishauser K, et al. 1998 Phenotypic variability in familial combined pituitary hormone deficiency caused by a PROP1 gene mutation resulting in the substitution of Arg3 Cys at codon 120 (R120C). J Clin Endocrinol Metab. 83:3727–3734. Brown MR, Lenartowska I, Oltatzewski M, Wu W, Rosenfeld M, Parks JS. PROP-1 mutations and hypopituitarism in Poland [Abstract P3– 431]. Proc of the 80th Annual Meet of The Endocrine Soc. 1998. Pfa¨ffle R, Kim C, Wu¨ller S, et al. Two PROP-1 mutations are prevalent among patients with combined pituitary hormone deficiency [Abstract O16]. Proc of the Annual Meet of ESPE. 1998. Prader A, Largo RH, Molinari L, Issler C. 1989 Physical growth of Swiss children from birth to 20 years of age. Helv Paediatr Acta. 52(Suppl):1–125. Brook CGD. 1982 Growth assessment in childhood and adolescence. Oxford: Blackwell. Trainer PJ, Besser M. 1995 The Bart’s endocrine protocols. London: Churchill Livingstone. Mullis PE, Patel M, Brickell PM, Brook CGD. 1990 Isolated growth hormone deficiency: analysis of the growth hormone (GH) releasing hormone gene and the GH gene cluster. J Clin Endocrinol Metab. 70:187–191. Watanabe YG. 1982 Effects of brain and mesenchyme upon the cytogenesis of rat adenohypophysis in vitro. I. Differentiation of adenocorticotropes. Cell Tissue Res. 227:257–266. Watanabe YG. 1982 An organ culture study on the site of determination of ACTH and LH cells in the rat adenohypophysis. Cell Tissue Res. 227:267–275. Couly G, Le Douarin NM. 1988 The fate map of the cephalic neural primordium at the presomitic to the 3-somite stage in the avian embryo. Development. 103(Suppl):101–113. Li S, Crenshaw III B, Rawson EJ, Simmons DM, Swanson LW, Rosenfeld MG. 1990 Dwarf locus mutants lacking three pituitary cell types result from mutations in the POU-domain of Pit-1. Nature. 347:528 –533. Rhodes SJ, Dimattia GE, Rosenfeld MG. 1994 Transcriptional Mechanism in anterior pituitary cell differentiation. Curr Opin Cell Biol. 5:488 – 498. Theill LE, Karin M. 1993 Transcriptional control of GH expression and anterior pituitary development. Endocr Rev. 14:670 – 689. Tang K, Bartke A, Gardiner CS, Wagner TE, Yun JS. 1993 Gonadotropin secretion, synthesis, and gene expression in human growth hormone transgenic mice and in Ames dwarf mice. Endocrinology. 132:2518 –2524. Pfa¨ffle RW, DiMattia GE, Parks JS, et al. 1992 A mutation in the POU-specific domain of Pit-1 and hypopituitarism without pituitary hypoplasia. Science. 257:1118 –1121. Kappen C, Schughart K, Ruddle FH. 1993 Early evolutionary origin of major homeodomain sequence classes. Genomics. 18:54 –70. Gehring WJ, Affolter M, Burglin T. 1994 Homeodomain proteins. Annu Rev Biochem. 63:487–526. Cooper DN, Youssoufian H. 1988 The CpG dinucleotide and human genetic disease. Hum Genet. 78:151–155. Li H, Zeitler PS, Valerius MT, Small K, Potter SS. 1996 Gsh-1, an orphan Hox gene, is required for normal pituitary development. EMBO J. 15:714 –724. Sheng HZ, Zhadanov AB, Mosinger BJ, et al. 1996 Specification of pituitary cell lineages be the LIM homeobox gene Lhx3. Science. 272:1004 –1007. Parks ME, Brown MR. 1997 Genes regulating hypothalamic and pituitary development. Acta Paediatr. 423(Suppl):28 –32. Sheng HZ, Moriyama K, Yamashita T, Li H, Potter SS, Mahon KA, Westphal H. 1997 Multistep control of pituitary organogenesis. Science. 278:1809 –1812. Topilko P, Schneider-Maunoury S, Levi G, et al. 1997 Multiple pituitary and ovarian defects in Krox-24 (NGFI-A, Egr-1)-targeted mice. Mol Endocrinol. 12:107–122. Acampora D, Mazan S, Tuorto F, et al. 1998 Transient dwarfism and hypogonadism in mice lacking Oxt1 reveal prepubescent stage-specific control of pituitary levels of GH, FSH and LH. Development. 125:1229 –1239. Dattani MT, Martinez-Barbera JP, Thomas PQ, et al. 1998 Mutations in the homeobox gene HESX1/Hesx1 associated with septo-optic dysplasia with septo-optic dysplasia in human and mouse. Nat Genet. 19:125–133.
