Journal of Pediatric Epilepsy 2015; 04(02): 092-094
DOI: 10.1055/s-0035-1555604
Letter to the Editor
Georg Thieme Verlag KG Stuttgart · New York

CDKL5 in Different Atypical Rett Syndrome Variants: Description of the First Eight Patients from Spain

Qingping Zhang
1   Department of Pediatrics, Peking University First Hospital, Beijing, China
,
Xinhua Bao
1   Department of Pediatrics, Peking University First Hospital, Beijing, China
,
Ying Zhao
1   Department of Pediatrics, Peking University First Hospital, Beijing, China
,
Xiaoying Zhang
1   Department of Pediatrics, Peking University First Hospital, Beijing, China
,
Jingjing Zhang
1   Department of Pediatrics, Peking University First Hospital, Beijing, China
,
Guangna Cao
1   Department of Pediatrics, Peking University First Hospital, Beijing, China
,
Jiarui Li
2   Center for Bioinformatics Life Science College, Peking University, Beijing, China
,
Liping Wei
2   Center for Bioinformatics Life Science College, Peking University, Beijing, China
,
Xiru Wu
1   Department of Pediatrics, Peking University First Hospital, Beijing, China
› Author Affiliations
Further Information

Publication History

25 September 2014

08 November 2014

Publication Date:
31 July 2015 (online)

CDKL5 in different atypical Rett syndrome variants: Description of the first eight patients from Spain

We read with great interest the article by Martínez et al[1] describing eight patients with cyclin-dependent kinase-like 5 (CDKL5, OMIM 300203) gene mutations. Among the patients, patient 6 carried a CDKL5 c.2927 C > T (p.Pro976Leu) mutation, and the authors considered this mutation as the pathogenic cause of disease in this individual. In our study, we found three individuals in a family—the proband, her younger brother, and her mother—carrying this variant. After further molecular analysis, we speculated that CDKL5 c.2927 C > T (p.Pro976Leu) may not be a pathogenic mutation.

The proband was a 5-year, 3-month-old girl. She was the first child of nonconsanguineous Chinese parents. She was born after an uneventful pregnancy. Her psychomotor development was approximately normal before 2 years of age. She could raise her head at 2 months, sit unaided at 7 months, and walk unsupported at 14 months. At the age of 14 months, she started to speak meaningful words, such as “mom,” “dad,” and “granddad.” Language development stagnated between the age of 2 and 3 years, and then regressed after 4 years of age. Currently, she can only occasionally speak single words with unclear enunciation and make sounds like “mom” or “dad.” Limited interest in others was noticed when she was 2 years and 4 months. Purposeful use of her hands was gradually lost beginning at 2 years. Unstable walking was observed at 4 years and 6 months of age. Stereotypical hand movements such as wringing and clapping when happy presented at 5 years of age.

So far, she has had no seizures, no respiratory distress, and no kyphosis. Her head circumference was 47 cm at 5 years and 2 months of age. Electroencephalography showed a slow background without epileptiform discharge. Brain magnetic resonance imaging showed diffuse cerebral atrophy. Urine and blood amino acid and organic acid levels were normal. Her family history was unremarkable. The clinical features of our proband and patient 6 of the study of Martínez et al[1] are summarized in [Table 1].

Table 1

Compared our proband with the patient 6 of the study of Martinez et al

Features

Our proband

Patient 6 of Martinez et al[1]

Age

5 y 2 mo

22 y

Gender

Female

Female

Head circumference

47 cm

Normal

Language

Regression, one word with unclear enunciation

Lost at 15 mo, regain at 10 y, two-word phrases at 13 y

Stereotypes

Unobvious, twisting cloths, clapping hands occasionally, began at 5 y

Counting coins, clapping, began at 18 mo

Purposeful hand use

Lost at 2 y

Yes

Walking

Yes, unstable

Yes

Autonomic dysfunction

No

Yes

Respiratory dysfunction

No

No

Sleeping disorder

Yes

Yes

Epilepsy

No

No

Magnetic resonance imaging

Cerebellar and cerebral atrophy

Normal

Electroencephalography

Slow background

Paroxysmal activation on sleep

X chromosome inactivation ratio

75:25

46:54

Genetic testing including methyl CpG binding protein 2 gene (MECP2), CDKL5, forkhead box G1 gene (FOXG1) was performed on the proband. Both MECP2 and FOXG1 gene mutations were negative. A CDKL5 variant c.2927 C > T (p.Pro976Leu) was found. Then CDKL5 gene mutational analysis was performed on all the family members. The same CDKL5 variant was found in the mother and the younger brother ([Fig. 1]). Clinically, the mother and the younger brother were healthy. The younger brother is 2 years old now and his psychomotor development is normal, and he has never suffered from seizures.

Zoom Image
Fig. 1 CDKL5 gene mutational analysis. A CDKL5 variant c.2927 C > T (p.Pro976Leu) was found in the proband (A), her mother (B), and her younger brother (C). Arrows indicate the mutational points.

X chromosome inactivation (XCI) pattern of the proband and her mother was studied using the X-linked androgen receptor (AR) locus.[2] The XCI ratio of the proband and her mother was 75:25 ([Fig. 2A]) and 55:45 ([Fig. 2B]), respectively. The allele carrying the c.2927 C > T variant was 75% inactive in the proband and 55% inactive in her mother. The brother also inherited the allele carrying the c.2927 C > T variant from his mother ([Fig. 2C]).

Zoom Image
Fig. 2 The X chromosome inactivation pattern analyzed by Gene marker. (A) The proband. (B) The mother. The upper row is digested by HpaII undigested, the lower is undigested. (C) The brother. DNA fragment 280 was the mutated allele.

Patients with CDKL5 mutations are characterized by early-onset epilepsy and severe psychomotor developmental delay.[3] In this family, all three CDKL5 variant carriers never suffered from seizures. XCI study showed that 55% of the chromosome with the CDKL5 c.2927 C > T variant was inactive in the mother, while in the proband, 75% of the chromosome (with CDKL5 c.2927 C > T variant) inherited from the mother was inactive. If the variant, CDKL5 c.2927 C > T, was a pathogenic mutation as a hemizygote of the X chromosome, the younger brother with this variant should have more severe neurological abnormalities. As the brother is healthy, the gene mutation is unlikely to be disease causing by itself.

As Williamson et al[4] indicated, CDKL5 has two isoforms. One terminates at intron 18 and has abundant expression in the brain. The other isoform contains 20 exons and its expression is rich in the testes. The variant, c.2927 C > T (p.Pro976Leu), is located in exon 20. It could be just a polymorphism. Ho et al[5] also proposed that this missense mutation had minimal effect on disease due to its position in the CDKL5 gene. Therefore, there must be some other causative explanation for the neurological abnormalities observed in both our patient and patient 6 of the study of Martínez et al.[1]