Selective luteinizing hormone deficiency due to mutations in the luteinizing hormone beta-subunit gene (LHB) is a rare cause of hypogonadism. We describe the clinical features of a consanguineous family in which three siblings, two men and one woman, had hypogonadism related to isolated luteinizing hormone deficiency. These subjects have a newly discovered homozygous mutation of a 5' splice site in LHB: IVS2+1G
C. This mutation disrupts the splicing of messenger RNA (mRNA), generating a gross abnormality in the processing of the luteinizing hormone beta-subunit mRNA, which abrogates the secretion of luteinizing hormone. We also determined that the female phenotype of this LHB mutation is characterized by normal pubertal development, secondary amenorrhea, and infertility.
Inactivating mutations of the human luteinizing hormone beta-subunit gene (LHB) were previously described in three men with hypogonadism who had normal genitalia at birth but had no pubertal development and had infertility due to selective luteinizing hormone deficiency.3,4,5 To our knowledge, no female phenotype of such inactivating mutations in the LHB gene has been described. We describe the clinical and hormonal characteristics of three siblings, two men and a woman, all of whom had hypogonadism associated with a newly discovered mutation of the LHB gene.
Case Reports
The proband (Subject II-5) (Figure 1), a 38-year-old Brazilian man born to consanguineous parents, was referred to the Endocrinology Unit of the University Hospital of Brasília for hypogonadism. He presented with a eunuchoid habitus, a juvenile voice, and bilateral gynecomastia, with scant axillary hair and no facial hair (Tanner stage, genitalia 1 and pubic hair 4). He was 181 cm tall, weighed 87 kg, and had an arm span of 193 cm. He had a micropenis (4.5 cm in length; mean [±SD] normal length, 13.3±3.8) and underdeveloped, though descended, testes.
|
|
The proband's 30-year-old brother (Subject II-10) had been treated with intramuscular testosterone for hypogonadism from the age of 25 years. He reported considerable penile and testicular growth after treatment, but he continued to have azoospermia. On admission to the endocrine unit, he had normal male genitalia with no gynecomastia; the Tanner stage was genitalia 5 and pubic hair 5, with a penile length of 9 cm. Ultrasonography revealed testicular volumes of 10.7 cm3 (right testis) and 8.2 cm3 (left testis). The hormonal profile was similar to that of his brother (Table 1). A testicular-biopsy specimen showed interstitial fibrous thickening, hypoplastic seminiferous tubules with a predominance of Sertoli cells, spermatogenic arrest, and an absence of Leydig cells.
The proband's 29-year-old sister (Subject II-13) presented with secondary amenorrhea and infertility. She had had spontaneous, normal pubertal development and menarche at the age of 13 years, followed by oligomenorrhea. Her weight was 56 kg, and her height 166 cm. She had normal breast and pubic-hair development (Tanner stage, mammary glands 5 and pubic hair 5). Ultrasonography of the pelvis revealed a normal uterus (volume, 50 cm3; normal range, 30 to 90), with atrophic endometrium (3 mm in thickness). The ovaries were normal in size (right ovary, 10 cm3; left ovary, 7 cm3) and contained multiple antral follicles (up to 13 mm in diameter) not restricted to the periphery (Figure 2A). Serial ultrasonography of the pelvis over a 2-week period showed no changes in endometrial thickness or ovarian appearance. Repeated measurement of serum estradiol and progesterone showed levels within the low-to-normal range for the follicular phase. Inhibin B and free alpha-subunit levels were elevated.
|
In addition to these 3 subjects, we studied 9 asymptomatic family members and 100 unrelated, ethnically matched controls (two alleles from each). Written informed consent was obtained from all participants. The study was approved by the Research Ethics Committee of the Faculty of Medicine, University of Brasília, Brazil.
Methods
Hormonal Assays
An immunofluorometric assay (AutoDELFIA, Wallac Oy) was used to measure luteinizing hormone, follicle-stimulating hormone, estradiol, testosterone, progesterone, and free alpha-subunit levels. Serum luteinizing hormone, estradiol, and progesterone levels were confirmed by a chemiluminescent immunometric assay (Immulite 2000, Euro/DPC), which was also used to measure androstenedione and dehydroepiandrosterone sulfate levels, as well as the testosterone levels in female Subject II-13. Inhibin B was detected with the use of an enzyme-linked immunoassay (Active Inhibin B ELISA, Diagnostics Systems Laboratories). The intraassay and interassay variations were less than 8% and less than 10%, respectively, for all hormones. The gonadotropin-releasing hormone stimulation test was performed in all three affected siblings. Serum luteinizing hormone and follicle-stimulating hormone levels were measured 0, 15, 30, 60, and 90 minutes after intravenous administration of 100 µg of gonadotropin-releasing hormone. Each result was compared with established normal values.6
DNA Sequencing and Analysis
Genomic DNA samples from subjects and controls were extracted from whole-blood specimens by the Chelex-100 method.7 The region spanning exon 2, intron 2, and exon 3 of the LHB gene was amplified with the use of a polymerase-chain-reaction (PCR) assay, with primers designated as LH23F and LH23R, under previously described conditions.8 PCR products were sequenced in both the sense and antisense orientations, with the use of an automated sequencer (ABI-377, Perkin-Elmer). The primer LH23R has specific 3'-end mismatches in order to discriminate between LHB and the highly homologous human chorionic gonadotropin beta-subunit gene (hCGB) and pseudogenes. To investigate whether the identified mutation represented a polymorphism, the same LHB-gene fragment from 100 ethnically matched controls was amplified and digested with the NcoI restriction enzyme (Fermentas). This procedure, involving PCR and restriction-fragment–length polymorphism assays, was also used to analyze genomic DNA from family members of the three affected subjects.
Amplification and Analysis of Leukocyte LHB mRNA
Since LHB mRNA, but not hCGB mRNA, was previously shown to be detectable by a reverse-transcriptase–PCR (RT-PCR) assay of unstimulated human blood leukocytes,9 we extracted total RNA from a 5-ml buffy-coat sample from each study participant, using Trizol (Invitrogen Life Technologies). To prevent contamination of genomic DNA, each sample was treated with RNase-free DNase I (Fermentas). Synthesis of complementary DNA was performed with the use of the LH23R primer and Moloney murine leukemia virus reverse transcriptase (Promega). The complementary DNA (5-µl samples) was subsequently amplified in advance of the PCR assay (35 cycles of 1 minute at 94°C, 1 minute at 60°C, and 1 minute at 72°C) with primers (3F5'-GCACCAAGGATGGAGATGCTCCAG-3' and LH23R) and GoTaq DNA Polymerase (Promega). Half-nested PCR was then performed with the preamplification product (1-µl samples), the internal primer 1F5'-GGCGGGGCATGGGCATCCAG-3', and the LH23R primer, under the preamplification conditions. The sequences and the specificity of the assay were verified by direct sequencing.
Results
Automated DNA sequencing of exon 2, intron 2, and exon 3 of the LHB gene in the male proband, his affected brother, and his sister revealed a homozygous GC substitution at the +1 position of intron 2 (IVS2+1GC), a 5' splice-donor site (Figure 3). The effect of this substitution on splicing of the mRNA transcript is shown in Figure 4.
|
|
C mutation was not identified in any of the 200 alleles from the controls. We detected RT-PCR products of the expected size in both the control and the normal heterozygotes (Figure 5). In the affected homozygotes, a PCR product approximately 250 bp larger was consistently obtained. Direct sequencing of these amplicons revealed that the observed difference in size was due to the presence of the entire intron 2, containing 236 nucleotides. The larger PCR product was not detected in the heterozygotes (Figure 5).
|
C mutation disrupted the splicing of intron 2 of the LHB mRNA, resulting in the insertion of 236 nucleotides in the mutant transcript (Fig. S1 in the Supplementary Appendix). Discussion
We report a homozygous mutation of the LHB gene associated with selective luteinizing hormone deficiency in three siblings, two men and one woman, who were members of a consanguineous Brazilian family. The IVS2+1GC mutation is located in a noncoding intronic sequence affecting a conserved 5' splice-donor site.
Subject II-13, the woman with a luteinizing hormone deficiency, underwent normal pubertal feminization and thelarche, as have several women with a homozygous inactivating mutation in the luteinizing hormone–receptor gene.1,2,10 However, in contrast to those women, Subject II-13 had a uterus of normal size, allowing sporadic menses to occur during the first 14 years after menarche. Although the amount of estradiol produced throughout her life could not be quantified with certainty, it was sufficient for the development of normal breasts and trophic uterine layers. In the absence of luteinizing hormone activity, the amount of androgenic substrates for aromatization to estradiol in granulosa cells would probably be limited,11 which may have accounted for the low-to-normal serum estrogen levels and endometrial atrophy in our subject. Accordingly, in mice lacking the luteinizing hormone beta-subunit expression of the majority of steroidogenic enzymes is significantly reduced, although differentiation of the ovarian thecal layer is not impaired.12
The amenorrhea and repeatedly low progesterone levels in Subject II-13 corroborate the essential roles of luteinizing hormone in ovulatory function. The finding of multicystic ovaries is consistent with follicular growth induced by follicle-stimulating hormone, as reflected by the elevated inhibin B levels, which may represent an increased number of healthy, early antral follicles.11,13 Taken together, these findings suggest that granulosa cells fail to maintain sufficient estradiol production, probably in relation to decreased luteinizing hormone–dependent secretion of androgenic precursors by thecal cells, whereas inhibin B production is spared. The high inhibin B levels, in turn, provide negative feedback for the secretion of follicle-stimulating hormone, accounting for the normal follicle-stimulating hormone levels in this subject.13
After supplementation with estrogen, the development of a dominant follicle (18 mm in diameter) was evident on ultrasonography, although ovulation remained impaired and no corpus luteum was observed. Given this woman's normal follicle-stimulating hormone level, these findings may indicate that estrogen or estrogen-related factors were the primary determinants of the follicular growth until the early preovulatory stage, whereas luteinizing hormone activity was not necessary until that stage. We conclude that the LHB IVS2+1GC mutation resulted in impairment of ovulation and corpus luteum function. As expected, this LHB mutation did not affect pubertal feminization, thereby permitting normal breast and uterine development.
The two affected men presented with hypogonadism and selective luteinizing hormone deficiency, in association with the absence of mature Leydig cells and spermatogenic arrest, corroborating previous descriptions of hypogonadism.3,4,5 This phenotype is similar to that of male LHB-knockout mice.12 Surprisingly, both affected men had an increase in testicular volume after testosterone-replacement therapy. The key determinant of adult testicular size is the number of mature Sertoli cells, though immature cells may persist in patients with pathologic conditions, such as isolated hypogonadotropic hypogonadism.14 Since both follicle-stimulating hormone and testosterone increase the rate of proliferation and maturation of these cells,14 the administration of exogenous testosterone may have contributed to the proliferation of Sertoli cells and to testicular growth in these men by acting synergistically with elevated follicle-stimulating hormone levels (a late compensatory mechanism).
The inhibin B levels in the two affected men were inappropriately normal, given the elevated follicle-stimulating hormone levels. In normal adults, inhibin B levels are positively correlated with the function of Sertoli cells and spermatogenic status.15 Although unusual, our findings are in keeping with those in previous studies showing that in boys, during the first pubertal years, basal inhibin B levels increase because of stimulation by follicle-stimulating hormone, in parallel with proliferation of Sertoli cells.16 Moreover, in patients with hypogonadotropic hypogonadism, baseline inhibin B levels have been shown to increase with increasing testicular volume.16 Two men previously described as having a luteinizing hormone deficiency also had elevated follicle-stimulating hormone levels,3,4 and at least one had normal inhibin B levels.4
Previously described gonadotropin mutations are located in the coding region of the genes.3,4,5,17,18,19,20,21 However, mutations in exon-flanking intronic sequences have frequently been associated with disease.22 Our analysis of the LHB mRNA isolated from peripheral-blood leukocytes showed that the IVS2+1GC mutation resulted in the inclusion of the entire intron 2 and, consequently, in disruption of the exon 3 reading frame. Furthermore, the mutation induced a gross abnormality in the LHB mRNA (Fig. S1 in the Supplementary Appendix). However, the mechanism accounting for the failure to detect abnormal transcripts in the heterozygotes, but not in the homozygotes, is unknown.
Common variants identified in the LHB gene are sometimes associated with pathologic conditions such as the polycystic ovary syndrome and infertility.23 The absence of the IVS2+1GC mutation in the 200 control chromosomes reinforces the fact that it is not a polymorphism.
On the basis of the mutant nucleotide sequence, the hypothetical aberrant protein would have an insertion of 79 residues encoded by intron 2, which would severely affect its tertiary structure. In addition, the exon 3 frame shift results in a lack of domains required for heterodimer stability and receptor binding, such as the conserved "seat belt" region (amino acids 90 to 110) and disulfide bonds from the cysteine knot (Fig. S1 in the Supplementary Appendix).3,24,25 These observations strongly suggest that the mutant luteinizing hormone beta subunit would be incapable of correct assembly with the alpha subunit and therefore would not be secreted.
We conclude that the IVS2+1GC mutation induced a gross abnormality in the processing of LHB mRNA, causing familial selective hypogonadotropic hypogonadism. The male phenotype was similar to those described previously, whereas the affected woman underwent apparently normal pubertal development, followed by secondary amenorrhea and chronic anovulation.
Supported by grants from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (479403/2003-1, to Dr. Neves, and 300469/2005, to Dr. Latronico); CNPq Programa de Apoio ao Desenvolvimento Científico e Tecnológico, Subcomponente de Biotecnologia (620003/02-2, to Dr. Neves); Fundação de Apoio ao Desenvolvimento Científico e Tecnológico na Área da Saúde (to Dr. Neves); and Fundação de Empreendimentos Científicos e Tecnológicos (to Dr. Lofrano-Porto).
No potential conflict of interest relevant to this article was reported.
We thank Dr. Lawrence C. Layman, Medical College of Georgia, Augusta, for his helpful comments on an earlier version of this manuscript; Dr. Ceres Resende, University Hospital of Brasília, Brasília, Brazil, for performing transvaginal ultrasonography; Dr. Marcelo Serra, Centro de Diagnóstico por Imagem, Brasília, Brazil, for performing testicular ultrasonography; Dr. Luciana Naves, University of Brasília, Brazil, for helpful discussions about the affected subjects and hormonal assays; and staff at the Sabin Institute and Laboratory of Clinical Analysis, Brasília, Brazil, for providing technical assistance.
Source Information
From the Section of Endocrinology, University Hospital of Brasília, Faculty of Medicine (A.L.-P., L.A.G., P.P.N., L.A.C.), and the Molecular Pharmacology Laboratory, Faculty of Health Sciences (G.B.B., F.A.R.N.), University of Brasília, Brasília; the Sabin Institute and Laboratory of Clinical Analysis, Brasília (G.B.B.); and the Developmental Endocrinology Unit, Hormone and Molecular Genetic Laboratory, São Paulo University Medical School, São Paulo (A.C.L.) — all in Brazil.
Drs. Lofrano-Porto and Barra and Drs. Casulari and Neves contributed equally to this article.
Address reprint requests to Dr. Lofrano-Porto at the Hospital Universitário de Brasília, Laboratório de Farmacologia Molecular, Campus Universitário Darcy Ribeiro, Caixa Postal 04473, CEP 70.919-970, Brasília DF, Brazil, or at adlofrano{at}unb.br.
References
subunit of luteinizing hormone. N Engl J Med 1992;326:179-183. [Web of Science][Medline]-subunit gene. N Engl J Med 1997;337:607-611. -subunit of follicle-stimulating hormone. N Engl J Med 1998;338:1729-1732.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
