Case Report: A Detailed Phenotypic Description of Patients and Relatives with Combined Central Hypothyroidism and Growth Hormone Deficiency Carrying IGSF1 Mutations

In recent years, variants in immunoglobulin superfamily member 1 (IGSF1) have been associated with congenital hypopituitarism. Initially, IGSF1 variants were only reported in patients with central hypothyroidism (CeH) and macroorchidism. Later on, IGSF1 variants were also reported in patients with additional endocrinopathies, sometimes without macroorchidism. We studied IGSF1 as a new candidate gene for patients with combined CeH and growth hormone deficiency (GHD). We screened 80 male and 14 female Dutch patients with combined CeH and GHD for variants in the extracellular region of IGSF1, and we report detailed biomedical and clinical data of index cases and relatives. We identified three variants in our patient cohort, of which two were novel variants of unknown significance (p.L570I and c.1765+37C>A). In conclusion, we screened 94 patients with CeH and GHD and found variants in IGSF1 of which p.L570I could be of functional relevance. We provide detailed phenotypic data of two boys with the p.C947R variant and their large family. The remarkable phenotype of some of the relatives sheds new light on the phenotypic spectrum of IGSF1 variants.

Patients with congenital hypopituitarism show a wide variation in phenotype, depending on which hormones are deficient. Symptoms can range from neonatal hypoglycemia, prolonged neonatal jaundice, underdeveloped genitals, short stature, truncal obesity, delayed puberty, and, if the pituitary-adrenal axis is affected, it can sometimes even lead to death [4]. Understanding the etiology of congenital hypopituitarism is important for Genes 2022, 13, 623 2 of 12 anticipation of clinical problems, for genetic counseling and to develop future treatment strategies for the disease.
Genetic variations in genes, such as POU1F1, PROP1, HESX1, LHX3, LHX4, TRHR and TSHB, have been associated with congenital hypopituitarism. In recent years, immunoglobulin superfamily member 1 (IGSF1), a new candidate gene located on the X-chromosome, has been the subject of extensive study. Initially, variants in IGSF1 were reported to be associated with X-linked central hypothyroidism (CeH) in combination with adult macroorchidism in hemizygous male patients [5]. Subsequent large case series showed that affected males can also present with disharmonious pubertal development (normal or early timing of testicular growth, but delayed rise of testosterone), variable prolactin deficiency, increased body mass index, and decreased attentional control [5][6][7][8][9] and partial and transient growth hormone deficiency (GHD) in childhood (16% of cases). Sporadically, patients may show normal testicular volume [10,11]. All previously described pathogenic mutations in IGSF1 led to loss of function in the protein. This phenomenon is called the IGSF1 deficiency syndrome, which is associated with variable endocrine and phenotypic characteristics [8,[11][12][13].
Since 2003, The Dutch HYPOPIT study has investigated the genetic and non-genetic causes of congenital hypopituitarism. As IGSF1 variants have been described in patients with GHD with normal testicular volume, we studied IGSF1 as a possible candidate gene in a cohort of 94 patients with a combination of CeH and GHD. To give a complete overview, we describe all IGSF1 variants found in this cohort and provide a detailed description of the phenotypes seen in the individuals carrying these variants. In addition, we provide the detailed phenotypic data of a large family including two boys with the pathogenic p.C947R variant.

Cohort and Patient Selection
The Dutch HYPOPIT study is a multi-center study that includes six Dutch university hospitals and two regional hospitals. We used the Dutch National Registry of Growth Hormone Treatment to select all recombinant human growth hormone (rhGH)-treated children and adults that were registered between 1992 and 2003. From this group, we selected patients with combined GHD and CeH of unknown cause. To obtain an unbiased representation of the population of patients with combined GHD and CeH, each participating center had to include all eligible patients treated between 1992 and 2003, otherwise the center could not participate in the study. Approval was obtained from the Medical Ethics Committees of all participating hospitals. We obtained informed consent from all participating patients, as well as their parents if the patients were younger than 18 years. All patients were of Western European origin.

DNA Isolation
Genomic DNA of the patients was extracted from peripheral whole blood samples according to standard procedures.

Mutation Screening
All previously described intragenic pathogenic IGSF1 mutations, except for (p.L569Ffs*16) are located in the extracellular domain of the IGSF1 protein (Figure 1), encoded by exons 10 to 17 [14]. Hence, we screened those exons of the IGSF1 gene (NM_001170961.1) in 80 male and 14 female patients with combined CeH and GHD. Primer sequences and polymerase chain reaction (PCR) conditions were kindly provided by the Leiden University Medical Center. Exons 10-17 were amplified using PCR with TAQ polymerase (Qiagen, Venlo, The Netherlands).

Sequencing
Sequencing was outsourced and performed by BaseClear B.V (Leiden, The Netherlands). For this purpose, 80 ng of DNA together with 25 pmol/µL of the primers was supplemented with sterile H2O. Forward and reverse primers were sequenced separately. The results were analyzed with Sequencher 4.2 [15]. The IGSF1 sequence with accession number NM_001170961.1 was used as a reference sequence.

Biomedical Examination
We retrospectively collected patients' clinical data and laboratory results which were performed as part of regular patient care. The same assays were used to carry out all measurements.

Sequencing
Sequencing was outsourced and performed by BaseClear B.V (Leiden, The Netherlands). For this purpose, 80 ng of DNA together with 25 pmol/µL of the primers was supplemented with sterile H 2 O. Forward and reverse primers were sequenced separately. The results were analyzed with Sequencher 4.2 [15]. The IGSF1 sequence with accession number NM_001170961.1 was used as a reference sequence.

Biomedical Examination
We retrospectively collected patients' clinical data and laboratory results which were performed as part of regular patient care. The same assays were used to carry out all measurements.

Screening of the Extracellular Region of IGSF1
We found a total of seven variants in IGSF1 in our patients with combined CeH and GHD. Two novel variants (p.L570I and intronic variant c.1765+37C>A), a known pathogenic mutation (p.C947R) and four previously described benign variants. Genotypic details of these variants are shown in Table 1. The two novel variants (p.L570I and intronic variant c.1765+37C>A) were identified in a female patient. She was born by breech delivery and showed neonatal hypoglycemia. Birth weight was 2440 g after 36 weeks of pregnancy. At presentation, she had a highpitched voice, frontal bossing, strabismus and very poor vision. At the age of 3 years, her height was 81 cm (−5.1 SD), her bone age 1.0 years, and her weight 10 kg. IGF-1 was undetectable (<1.0 nmol/L) and IGF-BP3 was 1.3 mg/L (−3.8 SDS).
GH stimulation with arginine and clonidine revealed a peak GH level of 5.5 and 6.2 mIU/L, respectively (reference men > 20 mIU/L). MRI revealed a very small sella turcica with a small anterior pituitary and an ectopic posterior pituitary. RhGH treatment was started at the age of 3 years. As her free thyroxine (FT4) was also low 7.6 pmol/L (reference range 9-22 pmol/L), levothyroxine treatment was started. Cortisol levels were normal and there was no diabetes insipidus. Her family history was unremarkable, and both her parents showed normal height.
We analyzed the variant with the VEP analysis tool from ensembl.org, accessed on 21 November 2021. In silico analysis predicted this variant to have a moderate modifying effect on the IGSF1 protein [Sift predicted this variant to be tolerated (p = 0.3); PolyPhen predicted this variant to be probably damaging (p = 0.9); LofToot predicted this variant to be possibly damaging (p = 0.38); CADDphred (score = 17.95); ClinPred predicted this variant to be possibly/probably damaging (see Table 2). Hence, we classified this as a variant of unknown significance (VUS), of which the exact relation with clinical phenotype is unknown. This variant is located 10 bp from a splice site.
The intron variant c.1765+37C>A was also analyzed using VEP analysis tool from ensemble.org, accessed on 21 November 2021. Results for this analysis are shown in Table 2. In silico analysis with Mutation Taster V2 predicted this variant to be deleterious (p-value = 0.99, p-value close to 1 indicates a high 'security' of the prediction). This variant is located in a regulatory region 37 bp from a splice site, but its effect could not be predicted.

Pathogenic Mutation p.C947R
The p.C947R was found in two brothers, of which one (III-5) was previously mentioned briefly as individual H.III-3 in the publication by Sun et al., 2012 [5]. Here we describe detailed clinical, phenotypic and biochemical features of the two brothers and their large family who carried the pathogenic variant p.C947R. The family's pedigree is shown in Figure 2 and a summary of clinical and biochemical features of the two boys are provided in Tables 3 and 4. The clinical data of the affected and non-affected relatives is shown in Table 5. The most remarkable observations found in the two brothers and their relatives are described below.
The p.C947R was found in two brothers, of which one (III-5) was previously mentioned briefly as individual H.III-3 in the publication by Sun et al., 2012 [5]. Here we describe detailed clinical, phenotypic and biochemical features of the two brothers and their large family who carried the pathogenic variant p.C947R. The family's pedigree is shown in Figure 2 and a summary of clinical and biochemical features of the two boys are provided in Tables 3 and 4. The clinical data of the affected and non-affected relatives is shown in Table 5. The most remarkable observations found in the two brothers and their relatives are described below.     The normal range for males and females is shown in the column on the right. ↓ value below the lower limit of normal; ↑ = value above the upper limit of normal.

Clinical Features of the Two Brothers with a p.C947R Variant
Index case III-5 is the third child of Dutch non-consanguineous parents. He was identified by the T4-TSH-TBG based Dutch neonatal screening program. His serum FT4 was low (10 pmol/L; reference range 13.5-19 pmol/L), with inappropriately low-normal serum TSH (2.6 mIU/L; ref 1.75-5 mIU/L). In a TRH stimulation test, he had a diminished increase of TSH in response to TRH, indicating a central cause of hypothyroidism. Thyroid ultrasound showed a severely hypoplastic thyroid gland [22]. Testicular volume was 6 mL at the age of eleven years (1.5 SDS [21]). He had an elevated serum FSH of 13.2 mIU/L (ref 4-5 mIU/L), low LH and testosterone, low SHBG, elevated AMH (18.3 µg/L; reference value 9 µg/L) and high inhibin-B (424 pg/L; reference range 131-227 pg/mL) [23]. He showed prepubertal responses to GnRH throughout infancy. His α-subunit concentration was normal. At the age of 12 years, his testicular volume was 10 mL (1.3 SDS). Testicular volume was last measured at 15 years old (30 mL, 1.8 SDS) [21]. Apart from CeH and borderline macroorchidism, he had short stature and partial GHD (GH peak 7.7 mIU/L at arginine test and 14.3 mIU/L at clonidine test). His pituitary MRI was normal. He was treated with rhGH and levothyroxine.
His brother, III-4, was not detected by neonatal screening but was diagnosed with and treated for CeH at 7 years of age. His serum FT4 was low (8 pmol/L; reference range 13.5-19 pmol/L), with inappropriately normal serum TSH (3.6 mIU/L; reference range 1.75-5 mIU/L). TSH response to TRH during a stimulation test was at the lower limit of normal ( Figure 3). Thyroid ultrasound revealed hypoplastic thyroid tissue (0.28 mL; reference value 2-5 mL), without nodules or cysts [24]. MRI showed no pituitary abnormalities. Like his brother, he had above-average testicular growth. He had prepubertal levels of FSH, LH, testosterone and responses to GnRH. At 12.5 years, his testicular volume was 12 mL (1.2 SDS), reaching > 30 mL at 14.5 years (>2.0 SDS). His α-subunit levels were normal, but inhibin-B and AMH were strongly elevated. He did not have GHD.

Clinical Features of the Two Brothers with a p.C947R Variant
Index case III-5 is the third child of Dutch non-consanguineous parents. He was iden tified by the T4-TSH-TBG based Dutch neonatal screening program. His serum FT4 wa low (10 pmol/L; reference range 13.5-19 pmol/L), with inappropriately low-normal serum TSH (2.6 mIU/L; ref 1.75-5 mIU/L). In a TRH stimulation test, he had a diminished in crease of TSH in response to TRH, indicating a central cause of hypothyroidism. Thyroi ultrasound showed a severely hypoplastic thyroid gland [22]. Testicular volume was mL at the age of eleven years (1.5 SDS [21]). He had an elevated serum FSH of 13.2 mIU/ (ref 4-5 mIU/L), low LH and testosterone, low SHBG, elevated AMH (18.3 µg/L; referenc value 9 µg/L) and high inhibin-B (424 pg/L; reference range 131-227 pg/mL) [23]. H showed prepubertal responses to GnRH throughout infancy. His α-subunit concentratio was normal. At the age of 12 years, his testicular volume was 10 mL (1.3 SDS). Testicula volume was last measured at 15 years old (30 mL, 1.8 SDS) [21]. Apart from CeH an borderline macroorchidism, he had short stature and partial GHD (GH peak 7.7 mIU/L a arginine test and 14.3 mIU/L at clonidine test). His pituitary MRI was normal. He wa treated with rhGH and levothyroxine.
His brother, III-4, was not detected by neonatal screening but was diagnosed wit and treated for CeH at 7 years of age. His serum FT4 was low (8 pmol/L; reference rang 13.5-19 pmol/L), with inappropriately normal serum TSH (3.6 mIU/L; reference rang 1.75-5 mIU/L). TSH response to TRH during a stimulation test was at the lower limit o normal ( Figure 3). Thyroid ultrasound revealed hypoplastic thyroid tissue (0.28 mL; ref erence value 2-5 mL), without nodules or cysts [24]. MRI showed no pituitary abnorma ities. Like his brother, he had above-average testicular growth. He had prepubertal level of FSH, LH, testosterone and responses to GnRH. At 12.5 years, his testicular volume wa 12 mL (1.2 SDS), reaching > 30 mL at 14.5 years (>2.0 SDS). His α-subunit levels wer normal, but inhibin-B and AMH were strongly elevated. He did not have GHD.

Genotype and Phenotype Description of the Relatives
We collected the phenotypic data from the brother, parents and uncle of the two affected boys. The older brother of the two boys, III-3 with wild-type (WT) genotype, was healthy and had a normal physical appearance. The father of the brothers, II-5 (WT), and the mother (Carrier) were healthy, with a normal appearance. They did not report any symptoms of hypothyroidism. The mother's pregnancies were unremarkable, with no miscarriages, but she had hyperemesis gravidarum in all three pregnancies. She had a normal age of breast development, although the exact date of menarche was unknown. She voluntarily chose not to give breastfeeding, not because lactation failed. However, she experienced an early menopause at the age of 44 years. She has two brothers and one sister.
The mother's youngest brother II-9 (Affected), uncle of the two boys and hemizygous for the same p.C947R mutation, had a remarkable phenotype. As a child, he had delayed psychomotor development, growth retardation and slow speech. Hypothyroidism (without goiter) was diagnosed, for which he received levothyroxine. After the start of LT4, his school results improved drastically. His ACTH test was normal, and X-sella showed a normalsized sella turcica. Abdominal ultrasound was unremarkable. His adult height is 190 cm without rhGH treatment. As a child, he had undescended testes and at the age of 16 years he was still prepubertal. His testosterone levels were low (0.3 ug %, ref 0.4-1.2 ug %). After he received hCG treatment, his testes descended and, like his nephews, he reported aboveaverage testicular growth. Although not measured by ultrasound or orchidometer, his self-reported final testicular volume was large. He reported a remarkable decrease in testicular volume over the years. At the age of 50, he consulted the GP due to fatigue, after which primary hypogonadism was diagnosed. He had two sons (III-11 and III-12), without endocrine anomalies. The phenotypes of the mother's other brother and sister were unremarkable. The maternal grandmother of the two boys (I-9) was healthy, with a normal appearance. Like all her sisters, she entered puberty late (aged 16 year). She had no fertility issues and entered menopause at the age of 46 y. She had no symptoms of hypo-or hypothyroidism and no goiter.
Apart from the two boys and uncle II-9, there were no thyroid problems in the family. However, a remarkable finding in this pedigree was that according to the grandmother of the boys, all her sisters (I-9,10,11 and 12) had delayed puberty.

Discussion
Although the classical IGSF1 deficiency syndrome includes macro-orchidism and central isolated hypothyroidism, IGSF1 variants have also been described in patients with GHD and normal testicular volume. Therefore, we studied IGSF1 as a possible candidate gene in a unique cohort of 94 Dutch patients with the combination of CeH and GHD.
To give a complete overview, we describe all IGSF-1 variants found in this cohort and provide a detailed phenotypic description of the individuals carrying these variants. In addition, we provide detailed phenotypic data of a large family including two boys with the pathogenic p.C947R variant. We describe a novel missense variant (p.L570I) combined with a novel intronic variant c.1765+37C>A of unknown significance and a pathogenic mutation p.C947R in a family with a remarkable phenotype.
Loss-of-function variants in IGSF1 result in CeH, either isolated or in combination with additional pituitary hormone deficiencies [9]. IGSF1 is a highly polymorphic gene and, to date, 136 genetic variants have been reported. An overview of variants associated with endocrinopathies is shown in Figure 1. Nearly all intragenic variants described to date are present in the C-terminal domain of IGSF1, and therefore we focused our screening on that part of the protein. Even though IGSF1 has been the subject of extensive study, the exact function of IGSF1 is still not determined [14].
Apart from the known variants and SNPs we found, there were several interesting findings. In one female patient with CeH and GHD, we found a novel missense variant (p.L570I) combined with the novel intronic variant c.1765+37C>A. The novel variant, p.L570I, is located in the hydrophilic linker domain of IGSF1. Two other variants in the same domain, p.A532V and p.L569Ffs*16, have been associated with endocrinopathies. Furthermore, p.L569Ffs*16 has been shown to impair plasma membrane trafficking of the protein [25]. In silico analysis of this variant was inconclusive; the fact that p.L570I is located next to L569Ffs*16 suggests that p.L570I could possibly affect plasma membrane trafficking.
The male relatives who were hemizygous for the p.C947R variant showed a variable and remarkable phenotype. The index case and his brother presented with CeH, nearmacroorchidism, and elevated inhibin-B and AMH levels. Their uncle, who had the same mutation, had hypothyroidism, delayed psychomotor development, growth retardation and slow speech as a child. Although not measured by ultrasound or orchidometer, his self-reported final testicular volume was very large. He reported a remarkable decrease in testicular volume over the years and was diagnosed with primary hypogonadism at the age of 50.
The increased AMH/InhB in the two brothers is remarkable, since AMH and inhibin levels are reported to be generally normal in patients carrying IGSF-1 mutations [6]. This might indicate an increased number or functioning of Sertoli cells, and therefore possibly explain the morphologic characteristics of the large testes.
There are some limitations to this study. As we describe a historical cohort, we were unable to perform additional genetic tests in some relatives and we were also unable to perform functional testing for some interesting variants. However, the detailed phenotypic data that we provide about the large family, including two boys with the pathogenic p.C947R variant, and the fact that we provide a complete overview of all IGSF-1 variants found in this unique cohort of patients with GHD and CeH, adds relevant information to the existing literature

Conclusions
We screened 94 patients with CeH and GHD and found several variants, of which p.L570I, based on previously reported data, could be of functional impact. We provide detailed phenotypic data of two boys with the p.C947R variant and their large family. The remarkable phenotype of some of the relatives sheds new light on the phenotypic spectrum of IGSF1 variants.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki 1964 declaration and its later amendments or comparable ethical standards.
Informed Consent Statement: Informed consent was obtained from all individuals participating in this study and their parents if they were less than 18 years old. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. We obtained approval from the medical ethics committees of all participating hospitals.

Data Availability Statement:
The data presented in this study are available in this article.