Der GLI2-Transkriptionsfaktor ist eines der Haupt Effektor-Proteine des Sonic Hedgehog (SHH)-Signalweges und hat vermutlich eine Schlüsselfunktion in der Entwicklung der Hypophyse. Genomische GLI2-Veränderungen welche zu abgeschnittenen Proteinen führten, wurden beschrieben als Ursache für Holoprosenzephalie (HPE) oder HPE-ähnliche Veränderungen, teilweise in Verbindung mit einer Hypophysenunterfunktion. Ziel dieser Arbeit war die Ermittlung der Frequenz von GLI2-Mutationen in Patienten mit multiplem hypophysärem Hormonausfall (multiple pituitary hormone deficiency, MPHD) und eine funktionelle Untersuchung der gefunden Mutationen mittels Transkriptionsaktivitäts-Untersuchungen durch funktionelle Luciferase assays. Hierfür wählten wir Teilnehmer der GeNeSIS (Genetics and Neuroendocrinology of Short Stature International Study)-Studie. Patienten bei denen bereits Mutationen eines der etablierten Gene für MPHD nachgewiesen wurde, wurden ausgeschlossen. Insgesamt haben wir 168 Patienten mit MPHD untersucht. Bei allen Patienten waren mindestens ein GH- und ein TSH-Mangel dokumentiert, Auffälligkeiten in der zentralen Bildgebung mittels cMRT wurden bei 96 Patienten angegeben. In fünf Studienteilnehmern wurden vier verschiedene heterozygote missense Varianten nachgewiesen, hiervon wurden zwei bislang noch nicht in der Literatur beschrieben. Eine Variante, pR516P, führte in den in-vitro Experimenten zu einem kompletten Verlust der Proteinaktivität. Zusätzlich zu einem Wachstumshormonmangel hatte der Träger dieser Mutation einen Mangel an TSH und der Gonadotropine, sowie einen nichtdeszendierten Hypophysenhinterlappen und eine Polydaktylie, aber keine ersichtlichen Mittelliniendefekte. Anhand der funktionellen Untersuchung konnten wir erstmalig nachweisen, dass ein heterozygoter Aminosäuren-Austausch im GLI2-Protein zu einer möglichen Funktionseinschränkung der Transkriptionsaktivität führen kann und somit die Ursache für MPHD mit milden extrahypophysären Auffälligkeiten sein könnte. Der Phänotyp von GLI2-Mutationen ist variabel und die Penetranz ist unvollständig. GLI2-Mutationen sind assoziiert mit einer Hypoplasie des Hypophysenvorderlappens und treten gehäuft mit einem ektopen Hypophysenhinterlappen auf.

Der GLI2-Transkriptionsfaktor ist eines der Haupt Effektor-Proteine des Sonic Hedgehog (SHH)-Signalweges und hat vermutlich eine Schlüsselfunktion in der Entwicklung der Hypophyse. Genomische GLI2-Veränderungen welche zu abgeschnittenen Proteinen führten, wurden beschrieben als Ursache für Holoprosenzephalie (HPE) oder HPE-ähnliche Veränderungen, teilweise in Verbindung mit einer Hypophysenunterfunktion. Ziel dieser Arbeit war die Ermittlung der Frequenz von GLI2-Mutationen in Patienten mit multiplem hypophysärem Hormonausfall (multiple pituitary hormone deficiency, MPHD) und eine funktionelle Untersuchung der gefunden Mutationen mittels Transkriptionsaktivitäts-Untersuchungen durch funktionelle Luciferase assays. Hierfür wählten wir Teilnehmer der GeNeSIS (Genetics and Neuroendocrinology of Short Stature International Study)-Studie. Patienten bei denen bereits Mutationen eines der etablierten Gene für MPHD nachgewiesen wurde, wurden ausgeschlossen. Insgesamt haben wir 168 Patienten mit MPHD untersucht. Bei allen Patienten waren mindestens ein GH- und ein TSH-Mangel dokumentiert, Auffälligkeiten in der zentralen Bildgebung mittels cMRT wurden bei 96 Patienten angegeben. In fünf Studienteilnehmern wurden vier verschiedene heterozygote missense Varianten nachgewiesen, hiervon wurden zwei bislang noch nicht in der Literatur beschrieben. Eine Variante, pR516P, führte in den in-vitro Experimenten zu einem kompletten Verlust der Proteinaktivität. Zusätzlich zu einem Wachstumshormonmangel hatte der Träger dieser Mutation einen Mangel an TSH und der Gonadotropine, sowie einen nichtdeszendierten Hypophysenhinterlappen und eine Polydaktylie, aber keine ersichtlichen Mittelliniendefekte. Anhand der funktionellen Untersuchung konnten wir erstmalig nachweisen, dass ein heterozygoter Aminosäuren-Austausch im GLI2-Protein zu einer möglichen Funktionseinschränkung der Transkriptionsaktivität führen kann und somit die Ursache für MPHD mit milden extrahypophysären Auffälligkeiten sein könnte. Der Phänotyp von GLI2-Mutationen ist variabel und die Penetranz ist unvollständig. GLI2-Mutationen sind assoziiert mit einer Hypoplasie des Hypophysenvorderlappens und treten gehäuft mit einem ektopen Hypophysenhinterlappen auf.:1. Bibliographische Beschreibung S. 4 2. Abkürzungen (alphabetisch sortiert) S. 6 3. Hintergrund S. 8 3.1 Minderwuchs S. 8 3.2 GeNeSIS-Programm S. 10 3.3 Entwicklung, Aufbau und Funktion der Hypophyse S. 11 3.4 Molekularbiologie der Hypophysenentwicklung S. 12 3.4.1 Signalmoleküle, die die Stratifizierung der Rathke-Tasche und die Festlegung der hypophysären Zelllinien bestimmen S. 12 3.4.2 Transkriptionsfaktoren, welche die frühen Phasen der Anlage kontrollieren S. 13 3.5 Hedgehog-Gen Familie S. 15 3.5.1 SHH-Prozessierung und Freisetzung S. 16 3.5.2 SHH-Rezeptorbindung S. 17 3.5.3 SHH-Signaltransduktion S. 17 3.6 GLI-Proteine S. 18 3.6.1 GLI2 S. 19 3.6.2 GLI2-Mutationen bei Menschen S. 20 3.7 Rationale für die Promotionsarbeit S. 20 4. Publikation S. 22 4.1. Druckversion S. 22 4.2 Supplemental Material S. 32 5. Zusammenfassung und Interpretation S. 40 5.1 Screening S. 41 5.2 GLI2-Varianten S. 41 5.3 Experimentelle Untersuchungen S. 42 5.4 Interpretation und Diskussion S. 43 6. Referenzen S. 45 7. Anlagen S. 54 7.1 Erklärung über die eigenständige Abfassung der Arbeit S. 54 7.2 Lebenslauf S. 55 7.3 Publikationen und Auszeichnungen S. 56 7.3.1 Publikationen S. 56 7.3.1 Auszeichnungen S. 56 7.5 Danksagung S. 57

The Hedgehog (HH) signalling pathway and its primary transducer, GLI2, regulate cardiomyogenesis in vivo and in differentiating P19 embryonal carcinoma (EC) cells. To further assess the role of HH signalling during mouse embryonic stem (mES) cell differentiation, we studied the effects of GLI2 overexpression during mES cell differentiation. GLI2 overexpression resulted in temporal enhancement of cardiac progenitor genes, Mef2c and Nkx2-5, along with enhancement of Tbx5, Myhc6, and Myhc7 in day 6 differentiating mES cells. Mass spectrometric analysis of proteins that immunoprecipitate with GLI2 determined that GLI2 forms a complex with BRG1 during mES cell differentiation. Furthermore, modulation of HH signalling during P19 EC cell differentiation followed by chromatin immunoprecipitation with an anti-BRG1 antibody determined that HH signalling regulates BRG1 enrichment on Mef2c. Therefore, HH signalling accelerates cardiac progenitor gene expression during mES cell differentiation potentially by recruiting a chromatin remodelling factor to at least one cardiac progenitor gene.

No description available.

The lung is an organ that contains a vast system of airways carefully constructed to achieve maximal surface area in a confined space, requiring guidance from a multitude of developmental factors. The Shh pathway is one such signaling mechanism that is critical to proper lung formation, guiding branching morphogenesis and cellular proliferation through its downstream Gli transcription factors. Additionally, Foxf1 has been shown to be a key developmental factor required for proper lung formation during embryogenesis. Although theorized that the Gli transcription factors are responsible for regulating foxf1 levels, their exact relationship has yet to be revealed. Using five different models for Shh signaling (gli2 null, gli2 over-expressor [hVER-Gli2], gli3 null, Gli3 constitutive repressor [Gli3Δ699] and cyclopamine treated lung explants), I compared and contrasted the role of Gli2 and Gli3 in terms of their effect on cell cycle regulation, and on the expression levels of foxf1 and its potential downstream target genes tbx4, tbx5 and fgf10. I found that ectopic over-expression of gli2 resulted in increased Shh pathway activation, and increased expression of G1/S phase cyclins, which was associated with increased cellular proliferation and lung growth. However, no change in the levels of G1/S phase cyclins due to altered Gli3 signaling was observed. Foxf1 levels positively correlate with the levels of gli2, and appear to be independent of Gli3 activity. The amount of tbx4, tbx5, and fgf10 transcripts were observed to follow the levels of gli2 in the different gli2 mouse models, however, there was no significant change in gli3 null or Gli3Δ699 mice. Finally, by analyzing gene expression at different time points during gestation, I found that while gli2 levels affect foxf1 throughout gestation, the relationship to tbx4, tbx5 and fgf10, occurs only during the latter stages of lung development. I conclude, that Gli2 and not Gli3 appears to be the primary transducer of Shh signaling influencing cyclin regulation, leading to changes in embryonic lung growth. Furthermore, that Gli2 and not Gli3 appears to regulate foxf1 expression levels, and that this may extend downstream to influence tbx4, tbx5 and fgf10 expression.

The lung is an organ that contains a vast system of airways carefully constructed to achieve maximal surface area in a confined space, requiring guidance from a multitude of developmental factors. The Shh pathway is one such signaling mechanism that is critical to proper lung formation, guiding branching morphogenesis and cellular proliferation through its downstream Gli transcription factors. Additionally, Foxf1 has been shown to be a key developmental factor required for proper lung formation during embryogenesis. Although theorized that the Gli transcription factors are responsible for regulating foxf1 levels, their exact relationship has yet to be revealed. Using five different models for Shh signaling (gli2 null, gli2 over-expressor [hVER-Gli2], gli3 null, Gli3 constitutive repressor [Gli3Δ699] and cyclopamine treated lung explants), I compared and contrasted the role of Gli2 and Gli3 in terms of their effect on cell cycle regulation, and on the expression levels of foxf1 and its potential downstream target genes tbx4, tbx5 and fgf10. I found that ectopic over-expression of gli2 resulted in increased Shh pathway activation, and increased expression of G1/S phase cyclins, which was associated with increased cellular proliferation and lung growth. However, no change in the levels of G1/S phase cyclins due to altered Gli3 signaling was observed. Foxf1 levels positively correlate with the levels of gli2, and appear to be independent of Gli3 activity. The amount of tbx4, tbx5, and fgf10 transcripts were observed to follow the levels of gli2 in the different gli2 mouse models, however, there was no significant change in gli3 null or Gli3Δ699 mice. Finally, by analyzing gene expression at different time points during gestation, I found that while gli2 levels affect foxf1 throughout gestation, the relationship to tbx4, tbx5 and fgf10, occurs only during the latter stages of lung development. I conclude, that Gli2 and not Gli3 appears to be the primary transducer of Shh signaling influencing cyclin regulation, leading to changes in embryonic lung growth. Furthermore, that Gli2 and not Gli3 appears to regulate foxf1 expression levels, and that this may extend downstream to influence tbx4, tbx5 and fgf10 expression.

Με την εργασία έγινε μελέτη του σηματοδοτικού μονοπατιού Sonic Hedgehog σε 96 περιστατικά καρκίνου πνεύμονα με τη μέθοδο της ανοσοϊστοχημείας. Επίσης μελετήσαμε την πιθανή συμμετοχή του μεταγραφικού παράγοντα FoxM1 στο καρκίνωμα του πνεύμονα και την πιθανή συσχέτισή του με το μονπάτι του Hedgehog. Έγινε μελέτη της έκφρασης των μορίων Shh, Ptch1, Smo, Gli1, Gli2 και FoxM1. Τα αποτελέσματα αποκάλυψαν μια έντονη έκφραση των μορίων του μονοπατιού και αυξημένα ποσοστά ενεργοποίησής του. Επίσης βρέθηκε στατιστικά σημαντική συσχέτιση με τα πλακώδη καρκινώματα και με τα χαμηλού grade καρκινώματα. Ανάλογη σημαντική συσχέτιση βρέθηκε και με το φύλο,συχνότερα ενεργοποιημένο μονοπάτι στους άντρες. Ακόμα ανιχνεύτηκε μια συσχέτιση της έκφρασης του FoxM1 με το ενεργοποιημένο μονοπάτι. / The hedgehog (HH)-signaling pathway is implicated in developmental processes and its aberrant activation in adult tissues has been associated with malignancy. The aim of this study was to determine the expression pattern of HH-signaling molecules in lung carcinomas, as well as the involvement of the transcription factor FOXM1, that controls cell proliferation, in this process. Paraffin-embedded tissue sections of 96 lung cancer cases and adjacent non-neoplastic lung parenchyma were immunohistochemically analyzed with anti-SHH, anti-Patched1 (PTCH1), anti-Smoothened (SMO), anti-GLI1, anti-GLI2 and anti-FOXM1 antibodies. Correlations of HH molecules with clinicopathological parameters and FOXM1 expression were evaluated. All the HH-signaling molecules examined were overexpressed in lung cancer compared with the adjacent non-neoplastic lung parenchyma. HH pathway activity and expression of PTCH1 and SMO were significantly higher in squamous cell carcinomas compared to other histological types. Activation of HH pathway and PTCH1 expression were correlated with tumor grade being higher in low grade tumors. There was a significant correlation of lymph node metastases with expression of SMO in all histological types and with the gender higher in men. Overexpression of FOXM1 in lung cancer was also significantly correlated with PTCH1, SMO and GLI1 expression. In conclusion, HH-signaling pathway is activated in lung cancer and correlates with histological type, prognostic parameters of the tumors as well as with the increased expression of FOXM1.

No description available.

Introdução: Alterações em genes relacionados com a secreção de GH ou a organogênese hipofisária foram identificadas em pacientes com deficiência de hormônio do crescimento (DGH) congênita. Entretanto, poucos casos de DGH têm sua etiologia esclarecida. O GHRH é um candidato óbvio para explicar a deficiência isolada de GH (DIGH). Na literatura, os estudos de análise do GHRH não conseguiram identificar mutações, porém são antigos e utilizaram uma metodologia com limitações. A maioria dos pacientes com deficiência hipotálamo-hipofisária múltipla (DHHM) apresenta neuroipófise ectópica sugerindo a importância do estudo de genes que atuam no início do desenvolvimento hipofisário, com expressão inclusive no infundíbulo. O GLI2 é um fator de transcrição na sinalização Sonic Hedgehog, envolvido com o início da embriogênese hipofisária, expresso na bolsa de Rathke primordial e no diencéfalo ventral. Previamente, mutações no GLI2 foram encontradas em pacientes com holoprosencefalia, e também alterações hipofisárias. Objetivos: Analisar o GHRH em 151 pacientes com DIGH (42 brasileiros e 109 encaminhados de centros internacionais) e analisar o GLI2 em 180 pacientes brasileiros com DIGH ou DHHM por PCR e sequenciamento automático dos genes; e descrever o fenótipo dos pacientes com mutações identificadas. Resultados: No GHRH foram identificadas seis variantes em heterozigose com previsão benigna pelas análises in silico. A análise do GLI2 identificou três mutações novas em heterozigose com códon de parada prematuro (p.L788fsX794, p.L694fsX722 e p.E380X), e geração de proteínas truncadas, com perda do domínio responsável pela ativação transcricional. A mutação p.L788fsX794 foi identificada numa paciente com baixa estatura, polidactilia, epilepsia e hipoglicemias. Apresentava deficiência de GH, TSH, ACTH, prolactina, LH e FSH. Na investigação familiar foi diagnosticada DIGH em dois tios e DHHM numa prima. Estes familiares, além de sua mãe e outros parentes maternos também apresentaram a mutação e polidactilia. A mutação p.L694fsX722 foi identificada num menino com baixa estatura por deficiência de GH, além de lábio leporino e fenda palatina. Seu pai, embora saudável, também apresentou a mutação. A mutação p.E380X foi identificada numa lactente com retardo no desenvolvimento, hipoglicemias, poliúria e polidipsia. Apresentava deficiência de GH, ACTH, TSH e ADH. Sua mãe aparentemente normal também apresentou a mutação. Todos os pacientes com DGH e mutação no GLI2 apresentaram neuroipófise ectópica (não visualizada na paciente com p.E380X), adenoipófise hipoplásica e ausência de holoprosencefalia na ressonância magnética. Dezoito variantes não-sinônimas também foram identificadas no GLI2 em 24 pacientes. Dezesseis dessas variantes foram consideradas deletérias em pelo menos um programa de predição in silico e dez delas não foram encontradas em população controle. O fenótipo dos pacientes foi predominante de DHHM e com neuroipófise ectópica e sem holoprosencefalia. Variantes silenciosas, intrônicas e polimorfismos foram identificados no GLI2, mas aparentemente sem alteração funcional. Conclusão: Não identificamos mutação no GHRH e se realmente existe mutação neste gene como causa de DGH, deve ser muito rara. Variantes no GLI2 são frequentes (15%), indicando seu importante papel na etiologia da DGH congênita. Além disso, ampliamos o espectro fenotípico dos pacientes com mutações no GLI2, que foi caracterizado por DIGH ou DHHM, inclusive com diabetes insipidus, neuroipófise ectópica (maioria) e ausência de holoprosencefalia. Outras características observadas foram polidactilia, defeito de linha média facial e herança autossômica dominante com penetrância incompleta / Introduction: Alterations in genes related to GH secretion and pituitary organogenesis have been identified in patients with congenital GH deficiency (GHD). However, in only few cases of GHD the etiology has been established. GH-releasing hormone (GHRH) is an obvious candidate to explain isolated GH deficiency (IGHD). Previous reports in the literature did not identify mutations in GHRH, however, the methodology used was limited. Most patients with combined pituitary hormone deficiency (CPHD) have an ectopic posterior pituitary lobe (EPP) suggesting the study of genes involved in early pituitary development and also expressed in the infundibulum. GLI2 is a transcription factor in Sonic hedgehog signaling expressed in the primordial Rathkes pouch and ventral diencephalon during early pituitary development. Previously, GLI2 mutations were found in patients with holoprosencephaly and pituitary abnormalities. Aim: Analyse GHRH in 151 patients with IGHD (42 Brazilian and 101 from international centers) and GLI2 in 180 Brazilian patients with IGHD or CPHD by PCR and automatic sequencing, and describe the phenotype of patients with mutations. Results: In GHRH, six heterozygous variants that are benign according to in silico analysis were identified. GLI2 study revealed three novel heterozygous mutations leading to premature stop codons (p.L788fsX794, p.L694fsX722 e p.E380X) and truncated proteins, without the transcriptional activator domain. p.L788fsX794 was identified in a girl with short stature, polydactyly, epilepsy and hypoglycemia. She had GH, TSH, ACTH, prolactina, LH and FSH deficiencies. Two uncles had IGHD and one cousin CPHD. These relatives, the mother and other maternal relatives had polydactyly and carried the mutation. p.L694fsX722 was identified in a boy with short stature due to GHD who also had cleft lip and palate. His healthy father also carried the mutation. p.E380X was identified in an infant with delayed development, hypoglycemia, polyuria and polydipsia. She had GH, ACTH, TSH and ADH deficiencies. Her apparently normal mother also carried the mutation. All patients with GHD and GLI2 mutations had an EPP (not visualized in the patient with p.E380X), hypoplastic anterior pituitary lobe and absence of holoprosencephaly on MRI. Eighteen non-synonymous variants in GLI2 were identified in 24 patients. Sixteen of these were considered deleterious in at least one in silico prediction program and ten of these were not found in the control population. The phenotype was mainly of CPHD and EPP without holoprosencephaly. Several synonymous and intronic GLI2 variants and polymorphisms apparently without functional consequences were identified. Conclusions: No mutations in GHRH were identified and if mutations in this gene exist as a cause of IGHD, they must be very rare. Variants in GLI2 are frequent (15%) indicating its important role in the etiology of GHD. Furthermore, we expanded the clinical spectrum of patients with GLI2 mutations characterized by IGHD or CPHD including diabetes insipidus, ectopic posterior pituitary lobe (in most patients) and absence of holoprosencephaly. Additional features were polydactyly and midline facial defects and the inheritance was autosomal dominant with incomplete penetrance

Introdução: Alterações em genes relacionados com a secreção de GH ou a organogênese hipofisária foram identificadas em pacientes com deficiência de hormônio do crescimento (DGH) congênita. Entretanto, poucos casos de DGH têm sua etiologia esclarecida. O GHRH é um candidato óbvio para explicar a deficiência isolada de GH (DIGH). Na literatura, os estudos de análise do GHRH não conseguiram identificar mutações, porém são antigos e utilizaram uma metodologia com limitações. A maioria dos pacientes com deficiência hipotálamo-hipofisária múltipla (DHHM) apresenta neuroipófise ectópica sugerindo a importância do estudo de genes que atuam no início do desenvolvimento hipofisário, com expressão inclusive no infundíbulo. O GLI2 é um fator de transcrição na sinalização Sonic Hedgehog, envolvido com o início da embriogênese hipofisária, expresso na bolsa de Rathke primordial e no diencéfalo ventral. Previamente, mutações no GLI2 foram encontradas em pacientes com holoprosencefalia, e também alterações hipofisárias. Objetivos: Analisar o GHRH em 151 pacientes com DIGH (42 brasileiros e 109 encaminhados de centros internacionais) e analisar o GLI2 em 180 pacientes brasileiros com DIGH ou DHHM por PCR e sequenciamento automático dos genes; e descrever o fenótipo dos pacientes com mutações identificadas. Resultados: No GHRH foram identificadas seis variantes em heterozigose com previsão benigna pelas análises in silico. A análise do GLI2 identificou três mutações novas em heterozigose com códon de parada prematuro (p.L788fsX794, p.L694fsX722 e p.E380X), e geração de proteínas truncadas, com perda do domínio responsável pela ativação transcricional. A mutação p.L788fsX794 foi identificada numa paciente com baixa estatura, polidactilia, epilepsia e hipoglicemias. Apresentava deficiência de GH, TSH, ACTH, prolactina, LH e FSH. Na investigação familiar foi diagnosticada DIGH em dois tios e DHHM numa prima. Estes familiares, além de sua mãe e outros parentes maternos também apresentaram a mutação e polidactilia. A mutação p.L694fsX722 foi identificada num menino com baixa estatura por deficiência de GH, além de lábio leporino e fenda palatina. Seu pai, embora saudável, também apresentou a mutação. A mutação p.E380X foi identificada numa lactente com retardo no desenvolvimento, hipoglicemias, poliúria e polidipsia. Apresentava deficiência de GH, ACTH, TSH e ADH. Sua mãe aparentemente normal também apresentou a mutação. Todos os pacientes com DGH e mutação no GLI2 apresentaram neuroipófise ectópica (não visualizada na paciente com p.E380X), adenoipófise hipoplásica e ausência de holoprosencefalia na ressonância magnética. Dezoito variantes não-sinônimas também foram identificadas no GLI2 em 24 pacientes. Dezesseis dessas variantes foram consideradas deletérias em pelo menos um programa de predição in silico e dez delas não foram encontradas em população controle. O fenótipo dos pacientes foi predominante de DHHM e com neuroipófise ectópica e sem holoprosencefalia. Variantes silenciosas, intrônicas e polimorfismos foram identificados no GLI2, mas aparentemente sem alteração funcional. Conclusão: Não identificamos mutação no GHRH e se realmente existe mutação neste gene como causa de DGH, deve ser muito rara. Variantes no GLI2 são frequentes (15%), indicando seu importante papel na etiologia da DGH congênita. Além disso, ampliamos o espectro fenotípico dos pacientes com mutações no GLI2, que foi caracterizado por DIGH ou DHHM, inclusive com diabetes insipidus, neuroipófise ectópica (maioria) e ausência de holoprosencefalia. Outras características observadas foram polidactilia, defeito de linha média facial e herança autossômica dominante com penetrância incompleta / Introduction: Alterations in genes related to GH secretion and pituitary organogenesis have been identified in patients with congenital GH deficiency (GHD). However, in only few cases of GHD the etiology has been established. GH-releasing hormone (GHRH) is an obvious candidate to explain isolated GH deficiency (IGHD). Previous reports in the literature did not identify mutations in GHRH, however, the methodology used was limited. Most patients with combined pituitary hormone deficiency (CPHD) have an ectopic posterior pituitary lobe (EPP) suggesting the study of genes involved in early pituitary development and also expressed in the infundibulum. GLI2 is a transcription factor in Sonic hedgehog signaling expressed in the primordial Rathkes pouch and ventral diencephalon during early pituitary development. Previously, GLI2 mutations were found in patients with holoprosencephaly and pituitary abnormalities. Aim: Analyse GHRH in 151 patients with IGHD (42 Brazilian and 101 from international centers) and GLI2 in 180 Brazilian patients with IGHD or CPHD by PCR and automatic sequencing, and describe the phenotype of patients with mutations. Results: In GHRH, six heterozygous variants that are benign according to in silico analysis were identified. GLI2 study revealed three novel heterozygous mutations leading to premature stop codons (p.L788fsX794, p.L694fsX722 e p.E380X) and truncated proteins, without the transcriptional activator domain. p.L788fsX794 was identified in a girl with short stature, polydactyly, epilepsy and hypoglycemia. She had GH, TSH, ACTH, prolactina, LH and FSH deficiencies. Two uncles had IGHD and one cousin CPHD. These relatives, the mother and other maternal relatives had polydactyly and carried the mutation. p.L694fsX722 was identified in a boy with short stature due to GHD who also had cleft lip and palate. His healthy father also carried the mutation. p.E380X was identified in an infant with delayed development, hypoglycemia, polyuria and polydipsia. She had GH, ACTH, TSH and ADH deficiencies. Her apparently normal mother also carried the mutation. All patients with GHD and GLI2 mutations had an EPP (not visualized in the patient with p.E380X), hypoplastic anterior pituitary lobe and absence of holoprosencephaly on MRI. Eighteen non-synonymous variants in GLI2 were identified in 24 patients. Sixteen of these were considered deleterious in at least one in silico prediction program and ten of these were not found in the control population. The phenotype was mainly of CPHD and EPP without holoprosencephaly. Several synonymous and intronic GLI2 variants and polymorphisms apparently without functional consequences were identified. Conclusions: No mutations in GHRH were identified and if mutations in this gene exist as a cause of IGHD, they must be very rare. Variants in GLI2 are frequent (15%) indicating its important role in the etiology of GHD. Furthermore, we expanded the clinical spectrum of patients with GLI2 mutations characterized by IGHD or CPHD including diabetes insipidus, ectopic posterior pituitary lobe (in most patients) and absence of holoprosencephaly. Additional features were polydactyly and midline facial defects and the inheritance was autosomal dominant with incomplete penetrance