Article Text


Novel FOXG1 mutations associated with the congenital variant of Rett syndrome
  1. M A Mencarelli1,
  2. A Spanhol-Rosseto1,
  3. R Artuso1,
  4. D Rondinella1,
  5. R De Filippis1,
  6. N Bahi-Buisson2,
  7. J Nectoux2,
  8. R Rubinsztajn2,
  9. T Bienvenu2,
  10. A Moncla3,
  11. B Chabrol3,
  12. L Villard4,
  13. Z Krumina5,
  14. J Armstrong6,
  15. A Roche6,
  16. M Pineda6,
  17. E Gak7,
  18. F Mari1,
  19. F Ariani1,
  20. A Renieri1
  1. 1Medical Genetics, Department of Molecular Biology, University of Siena, Siena, Italy
  2. 2Université Paris Descartes, Institut Cochin, Inserm U567, Paris, France
  3. 3Inserm U910, Université de la Méditerranée, Assistance Publique Hôpitaux de Marseille, Hôpital de La Timone, Marseille, France
  4. 4Inserm U910, Université de la Méditerranée, Faculté de Médecine de La Timone, Marseille, France
  5. 5Medical Genetics Clinic of Latvian State, Children's University Hospital, Latvia
  6. 6Hospital Sant Joan de Déu, Esplugues, Barcelona, Spain
  7. 7Sagol Neuroscience Center, Sheba Medical Center, Tel Hashomer affiliated to the Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
  1. Correspondence to Alessandra Renieri, Medical Genetics, Molecular Biology Department, University of Siena, Viale Bracci, 2, 53100 Siena, Italy; renieri{at}


Background Rett syndrome is a severe neurodevelopmental disorder representing one of the most common genetic causes of mental retardation in girls. The classic form is caused by MECP2 mutations. In two patients affected by the congenital variant of Rett we have recently identified mutations in the FOXG1 gene encoding a brain specific transcriptional repressor, essential for early development of the telencephalon.

Methods 60 MECP2/CDKL5 mutation negative European Rett patients (classic and variants), 43 patients with encephalopathy with early onset seizures, and four atypical Rett patients were analysed for mutations in FOXG1.

Results and conclusions Mutations have been identified in four patients, independently classified as congenital Rett variants from France, Spain and Latvia. Clinical data have been compared with the two previously reported patients with mutations in FOXG1. In all cases hypotonia, irresponsiveness and irritability were present in the neonatal period. At birth, head circumference was normal while a deceleration of growth was recognised soon afterwards, leading to severe microcephaly. Motor development was severely impaired and voluntary hand use was absent. In contrast with classic Rett, patients showed poor eye contact. Typical stereotypic hand movements with hand washing and hand mouthing activities were present continuously. Some patients showed abnormal movements of the tongue and jerky movements of the limbs. Brain magnetic resonance imaging showed corpus callosum hypoplasia in most cases, while epilepsy was a variable sign. Scoliosis was present and severe in the older patients. Neurovegetative symptoms typical of Rett were frequently present.

  • Rett syndrome
  • congenital Rett variant
  • FOXG1
  • mental retardation
  • clinical genetics
  • neurology

Statistics from


Rett syndrome (RTT) is characterised by a serious and global developmental disorder affecting the central nervous system. First described by Andreas Rett 40 years ago, the syndrome has been the object of extensive investigations, revealing a wide spectrum of clinical phenotypes including the classic form, the early onset seizure variant, the Zappella variant (Z-RTT), the congenital variant, the ‘forme fruste’ variant, and the late regression variant.1–3 Mutations in the MECP2 gene, located in Xq28, are responsible for 95% of classic RTT and for 50% of Z-RTT4; while the early onset seizure variant results from mutations in the CDKL5 gene, in Xp22.5 We have recently used a candidate gene approach to demonstrate that the FOXG1 gene, located in 14q12, is responsible for the congenital variant of RTT.3 In this variant, initially described by Rolando in 1985, the affected girls present the same clinical features as in classic RTT, but in addition they are floppy and retarded from the very first months of life.6

We report the identification of FOXG1 mutations in four additional congenital RTT girls, through mutation screening of a cohort of 107 European patients: 60 RTT patients (classic and variants), 43 patients with epileptic encephalopathy, and four RTT-like patients. Clinical data of these four patients were compared with the two previously reported girls in order to improve the characterisation of the phenotype associated with FOXG1 mutations.3

Patients and methods

Patients and phenotype definitions

We collected a cohort of 107 European patients(56 patients from France, 49 from Spain and 2 from Latvia) with the following clinical classification: 60 RTT girls (33 classic, 16 congenital, 7 with early onset seizures, 2 late regression, 1 Z-RTT and 1 ‘forme fruste’), 43 patients with encephalopathy with early onset seizures (40 females and 3 males), and 4 RTT-like patients (1 female and 3 males with microcephaly, hand stereotypies and autistic features). Patients with classic and variant RTT were diagnosed according to the international criteria.7 RTT-like cases are patients who show some RTT clinical features but who do not fulfil all diagnostic criteria for classic or variant RTT. Phenotypic scores have been calculated using the severity score system previously reported by Renieri et al.2 All patients have been tested negative for MECP2 and CDKL5 mutations by a combination of denaturing high performance liquid chromatography (DHPLC) and multiplex ligation dependent probe amplification/quantitative polymerase chain reaction (MLPA/qPCR) analysis.

Molecular analysis

Blood samples were obtained after informed consent. DNA was extracted from peripheral blood using a QIAamp DNA Blood Kit (Qiagen GmbH QIAGEN strasse 1, 40724 Hilden, Germany). To determine the appropriate DNA concentration, we used the OD260/280 method on a photometer.8 DNA samples were screened for mutations in FOXG1 gene using Transgenomic WAVE DHPLC. The entire coding portion of FOXG1 was analysed as previously reported.3 PCR products resulting in abnormal DHPLC profiles were sequenced on both strands using PCR primers with fluorescent dye terminators on an ABI PRISM 310 genetic analyser (PE Applied Biosystems, Foster City, California, USA).

Clinical score

For each patient a severity score was assigned through the evaluation of 22 different clinical signs.2 Results were compared with the severity score of 128 classic and 25 Z-RTT patients with mutation in the MECP2 gene, and nine patients with the early onset seizures variant of RTT with a CDKL5 mutation.2 9 Statistical analysis was performed using the median test.


Molecular analysis

A mutation in the FOXG1 gene was identified in four patients clinically classified as congenital RTT. The de novo origin of the mutation was confirmed in three cases (cases 2–4), while in one case (case 1) parents were not available. Cases 3 and 4 bear the truncating mutations p. S185fsX454 (c.551_552insC) and p.Y208X (c.624C>G). The remaining two cases have missense mutations that lie within the forkhead domain and affect residues highly conserved in different species (including Tetraodon nigroviridis—CAF93027; Xenopus tropicalis—NN_001116933; and Xenopus laevis—NP_001079165): p.N227K (c.681C>G) in case 1 and p.F215L (c.643T>C) in case 2.

Clinical description

Case 1 (1091) is a female from Latvia, presently aged 17 years; she is in the fourth stage of RTT with spastic paraparesis (table 1; figure 1B). Since the age of 16 months she has grown up in an orphanage and her parents are presently unavailable. The girl was her mother’s first child and was born after a normal pregnancy. Birth weight was 3900 g with an Apgar score 8/9. Psychomotor delay was appreciated from the age of 6–7 months. The time of occurrence of the first seizures is unknown. The girl has been examined only once, at the age of 13 years 2 months.

Table 1

Clinical features of patients with FOXG1 mutations

Figure 1

Pictures of three of the new patients with FOXG1 mutations. Panel A, case 4; panel B, case 1; panel C, case 2. Note in patient 4 the constant and intense hand to mouth stereotypic activities. The severe microcephaly of case 1 is clearly evident. Patient 2, presently aged 6 years 6 months, is able to stand up only with support but she cannot walk. Parental/guardian consent has been obtained.

Case 2 (RTT00967) is a female from France, presently aged 8 years 6 months (table 1, figure 1C). She is the second child of a non-consanguineous couple; another child was born afterwards. She was born at 39 weeks after a normal pregnancy. Auxological parameters at birth were normal. In the neonatal period sleep disturbance and severe distress (crying) were noticed. She was referred to a clinical unit at 6 months because of psychomotor retardation. At this age, examination revealed severe hypotonia with sleep disturbances and no dysmorphic features except strabismus and hypermetropia. A re-evaluation at 2 years showed normal weight and height (height 81 cm, weight 12 kg), while the occipitofrontal circumference (OFC) was at −2 SD (45 cm), with developmental delay, hypotonia, very poor social contact, manual stereotypies and dystonia of the extremities. At 6.5 years, she could only stand up with support and did not walk; could hold an object (feeding bottle) with a simple grasp; and showed trunk rocking and tongue chewing. On last examination, at 8 years of age, the clinical phenotype was unchanged; she was only able to pronounce two disyllabic words (mama, papa) and brain magnetic resonance imaging (MRI) showed microcephaly with abnormal development of the frontal lobes without gyration defect.

Case 3 (RTT01158) is a female from France, presently aged 3 years (table 1). The patient is the first child of non-consanguineous parents from Benin, without a relevant family history. She was born at term and showed normal auxological parameters. She developed severe microcephaly at an early stage. Brain MRI performed in the first months showed isolated ventricular dilatation. Subsequently, she developed severe epileptic encephalopathy. In addition she had a disturbed sleep pattern from 2 years. When last examined, she was able to hold her head steadily but showed asymmetric spastic tetraplegia and scoliosis. Brain MRI performed at 2 years showed delayed myelination with hypoplasia/hypomyelination of the corpus callosum.

Case 4 (60719368) is a female from Spain, presently aged 3 years 2 months (table 1, figure 1A). The mother has a normal boy of 9 years and has had three spontaneous abortions. At birth she showed very severe hypotonia and normal auxological parameters. Subsequently microcephaly became evident. At 4 months OFC was at −2SD, at 9 months at −3.5, and since the age of 2 years at −5SD. She developed hand stereotypies at 12 months: she used to bring her hands to the mouth and pat her fingers on her lips. Protruding tongue movements have been constant from the age of 4 months.

Clinical comparison of the six patients with FOXG1 mutations

We compared the clinical picture of these four additional patients with a FOXG1 mutation to the two previously reported.3 These six patients, with age ranging from 3–22 years, present a distinct clinical phenotype (table 1).

In order to recognise the clinical differences among the RTT phenotypes, we compared the clinical scores of four patients with mutations in FOXG1 for whom we could obtain complete clinical information (patients 3–6) to the scores of classic and Z-RTT patients with MECP2 mutations and those of early onset seizure variant patients with CDKL5 mutations. In patients with a FOXG1 mutation the clinical score median value is 31.5, with total score ranging from 28–38. Among the classic RTT cases the median value is 27, with a range from 13–39.9 In the nine patients with CDKL5 mutation, the median value is 27, with total score ranging from 22–30.9 In the Z-RTT the median value is 13, with values ranging from 4–24.2


This work confirms that the FOXG1 gene is responsible for the congenital variant of RTT. In fact, FOXG1 mutations (figure 2) have so far been identified only in patients originally classified by different clinical centres as affected by this variant of RTT.

Figure 2

FOXG1 mutations and alterations of the functional domains. Schematic representation of FoxG1 protein. The three main functional domains are shown: the DNA binding fork-head domain, the Groucho binding domain and the JARID1B binding domain is red. The numbers at the top refer to the amino acid positions. The frameshift and stop mutations are showed below by zigzag lines. The missense mutations are indicated at the top. The asterisks indicate the two mutations previously reported in Ariani et al.3

The overall clinical phenotype is characterised by normal pregnancy and delivery, normal auxological parameters at birth, followed by hypotonia, irresponsiveness and irritability in the neonatal period. Deceleration of head growth represents one of the most important diagnostic signs: in the four cases for whom the OFC had been recorded in the first months of life, microcephaly was evident already before the fourth month. Apparently microcephaly is more severe than in classic RTT (mean −4.32 SD compared with a mean of −2.4 SD in classic RTT). All five patients bearing a rearrangement on chromosome 14q12 involving FOXG1 presented with microcephaly: in the first reported case the onset was not defined10; in three patients microcephaly was evident between 5–8 months11–13; while in the fifth patient head circumference was below the third centile at the age of 11 months.14

In the two girls previously described with FOXG1 mutations, a retrospective assessment as to whether there had been a period of normal development was not feasible, although the parents noted a delay only at 3 months. In patient 3, who presented with the earliest onset, a real regression period was not identifiable, while the other three cases (patients 1, 2 and 4) presented a regression before 6 months of age, more precocious than in classic RTT, in accordance with the diagnosis of the congenital variant. Motor development is severely impaired: no girl with a mutation could either walk with support or speak, although cases 2 and 6 can stand up with assistance (figure 1C). Voluntary hand use was absent in the majority of patients (3/5) and poor in two (cases 2 and 4). Poor eye contact and absence of response to social interaction were evident in five out of six patients, in contrast with classic RTT where eye contact is intense, and this feature represents a supportive criterion for diagnosis.

Stereotypic hand movements are typical as in classic RTT, with intense and continuous hand washing and hand mouthing activities. In addition, all girls showed constant thrusting of the tongue. These repetitive movements are not so typical of classic RTT, where ineffective chewing movements (seen in one of our patients) are found rather than thrusting of the tongue. Similar tongue protruding movements were present in the girl with the 14q12 deletion, which led to the identification of FOXG1 mutations in the first patients with congenital RTT.12 Also, in 4/5 patients jerky movements are often seen in the upper limbs, that are frequently pushed in different directions, while in classic RTT such movements are rarely reported.

In this small cohort of patients, epilepsy was a variable sign: two girls, aged 3 years and 6.5 years respectively, have never presented with epileptic seizures (cases 2 and 4); in four patients epilepsy was present with an onset between 14 and 30 months. In two the epilepsy was well controlled by antiepileptic drugs (patients 3 and 6), while in the remaining two, seizures recurred despite treatment.

Neurological and neurovegetative symptoms are consistent with a diagnosis of RTT: constipation is reported in 4/6 patients; breathing abnormalities in 4/6; cold extremities in 3/6; bruxism in 4/6; sialorrhea in 4/6. Skeletal alterations such as scoliosis and kyphosis, genu valgu and pes planus are severe in the older cases.

Brain MRI showed corpus callosum hypoplasia in four patients, and this has been excluded in only one of the remaining two cases. Moreover, complete agenesis of the corpus callosum has been identified in patients with chromosomal rearrangements involving FOXG1.11 15 All these findings are in accord with the phenotype of heterozygous Foxg1+/− mice, showing a corpus callosum defect.3

Our results contribute to the clarification of the phenotype associated with FOXG1, confirming its role in the RTT spectrum. In particular, they seem to be associated with the most severe end of this spectrum. In fact, the median of the total clinical score in this group of patients is higher in comparison with patients affected by classic or early onset seizure variant of RTT.

All the patients reported to date with FOXG1 mutations are female. This is probably due to an ascertainment bias and we expect that further mutations will be identified in male cases, given that FOXG1 is an autosomal gene.

In conclusion, we suggest that FOXG1 gene mutation analysis should be performed in female and male patients showing RTT features but lacking the typical early normal period due to the precocious onset of symptoms. Besides other features typical of classic RTT, major signs possibly indicating a FOXG1 mutation are severe psychomotor delay with inability to walk, severe postnatal microcephaly evident before the age of 4 months, poor eye contact, tongue stereotypies, jerky movements of limbs, and corpus callosum hypoplasia.

Key points

  • FOXG1 gene, located in 14q12, is responsible for the congenital variant of RTT.

  • Mutation analysis should be performed in female and male patients showing RTT features but lacking the typical early normal period due to the precocious onset of symptoms.

  • Major signs possibly indicating a FOXG1 mutation are severe psychomotor delay with inability to walk, severe postnatal microcephaly evident before the age of 4 months, poor eye contact, tongue stereotypies, jerky movements of limbs, and corpus callosum hypoplasia.


We would first like to thank the Rett patients and their families. This work was supported by Telethon grants GTB07001, by the EuroRETT E-RARE network and by the Emma and Ernesto Rulfo Foundation to A.R. This work was also supported by A.I.R. (Associazione Italiana Rett) to M.A.M and by PAR 06 of University of Siena to F.M.


View Abstract


  • Competing interest None.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.