Severe developmental delay and epilepsy in a Japanese patient with severe congenital neutropenia due to HAX1 deficiency

HAX1 deficiency has recently been identified as a cause of severecongenital neutropenia (SCN), but little is known about thephenotype. We described an SCN patient with a homozygous 256C-to-Ttransition causing an R86X mutation in the HAX1 gene. Notably,the patient has been complicated by epilepsy and severe delayof motor, cognitive, and intellectual development; each developmentalquotient was 21–26 at 7 years old. Growth failure anddental development delay were also noted. Neurodevelopmentaldelay in this patient expands the clinical phenotype of HAX1deficiency and suggests an important role of HAX1 on neuraldevelopment as well as myelopoiesis.

Key words: developmental delay, epilepsy, HAX1 deficiency, severe congenital neutropenia.

Severe congenital neutropenia (SCN) is a rare disorder of myelopoiesisresulting in recurrent life-threatening infections due to alack of neutrophils.¹ Individuals with SCN have characteristicbone marrow findings of myeloid hypoplasia with arrest of myelopoiesisat the promyelocyte/myelocyte stage.¹ This disorder was firstdescribed in Kostmann’s seminal paper in 1956.² Recentevidence has indicated that SCN is a heterogeneous disorderinvolving mutations in various genes, including those encodinggranulocyte colony stimulating factor receptor (G-CSFR) (CSF3R),³neutrophil elastase (ELA2),⁴ Wiskott-Aldrich syndrome protein(WAS),⁵ and GFI-1.⁶ More recently, Klein et al. identified homozygousHAX1 mutations as a cause of an autosomal recessive SCN in 20Middle Eastern patients and 3 classical Kostmann family members.⁷They suggested that HAX1 is critical for maintaining the innermitochondrial membrane potential and protecting against apoptosisin myeloid cells.⁷

HAX1 has been initially described as an interacting partnerwith HS-1, a signal-transduction protein in hematopoietic cells.⁸HAX1 mRNA is ubiquitously expressed,^8,9 and therefore its deficiencycould theoretically result in dysfunction of several organsother than hematopoietic systems. Given that the precise functionof HAX1 is not fully defined, clinical phenotypes of patientswith deficiency of the responsible gene may be helpful. Here,we described an SCN patient caused by homozygous R86X mutationsin HAX1, who presented with severe developmental delay followedby refractory epilepsy. This case may expand the phenotype ofHAX1 deficiency in human and may provide a novel function ofthis gene.

Design and Methods

TOP

Design and Methods

A 9-year-old boy was investigated because of congenital neutropeniawith presentation of fulminant left otomastoiditis at two monthsold. The patient was born without asphyxia from nonconsanguineousparents at 39 weeks’ gestation. There was no family historyof immunodeficiency. Initial investigations demonstrated anabsolute neutrophil count (ANC) of 0–0.09x10⁹/L and markedmonocytosis. Bone marrow examinations showed myeloid hypoplasiawith maturation arrest at the promyelocyte/myelocyte stage.Cytogenetic analysis was normal. Under the diagnosis of SCNand otomastoiditis, the patient was treated with parenteralantibiotics and G-CSF. A higher dose of G-CSF at 10 µg/kg/dayfor 9 days was required to achieve ANC over 1x10⁹/L. He wasdischarged with hearing impairment at a threshold of 50 dB onaudiograms of the left ear. Details on his early clinical coursewere published previously. ¹⁰

After discharge, he was treated with subcutaneous administrationof G-CSF at 5–10 µg/kg two to three times per weekand prophylactic use of sulfamethoxazole/ trimethoprim. Sixmonths after discharge, the frequency of G-CSF administrationwas reduced to once per week. Although ANC has remained lowat 0.05–0.1x10⁹/L under this therapy, there has not beenany recurrent, invasive bacterial infection for more than 9years. Annual follow-up bone marrow examinations have not shownany findings compatible with myelodysplastic syndrome or acutemyelogenous leukemia.

Neurological development of the patient has been delayed. Monthsat which the patient achieved each developmental milestone wereas follows; head control (4 months), sit alone (9 months), standwith support (12 months), stand without support (15 months),walk alone (18 months), and speak single words (24 months).Developmental retardation became more pronounced in later periods.The patient became bowel and bladder trained at 5 years old.He can not construct complete sentences and can not read atthe time of this manuscript preparation (9 years old). Usingthe New Edition of the Kyoto Scale of Psychological Development,a standardized test widely used in Japan,¹¹ we assessed hisneurodevelomental status at 36 months and 93 months of age,indicating severe delay in motor (developmental quotient, DQ;49 and 26, respectively), cognitive (DQ; 62 and 21, respectively),and verbal and social development (DQ; 61 and 23, respectively).Plasma amino acids distributions and plasma concentrations oflactic and pyruvic acids were normal. Computed tomography (2months and 2 years of age) and magnetic resonance images (6and 8 years of age) of the brain did not demonstrate any abnormalities.At 4 years old, the patient demonstrated a brief, repetitiveloss of consciousness. Electroencephalographic findings ledto the diagnosis of symptomatic generalized epilepsy presentingwith atypical absence. Epileptic symptoms were partially controlledwith combined administration of valproric acid, clobazam, andethosuximide, and remained under observation. Physical growthhas failed to thrive. At 9 years old, body height and weightwas 121.7 cm (–1.8 SD) and 21.0 kg (–1.5 SD), respectively.Dental development has also been delayed – first mandibularpermanent molars erupted at 8 years of age, and of 8 permanentincisors, only 3 have erupted at 9 years of age.

Mutational analysis and immunoblot analysis of the genes responsible for SCN

TOP

Mutational analysis and...

We obtained written informed consent for analysis from bothparents. All study centers received approval from the institutionalreview boards to perform genetic analysis. Mutations in CSF3R,ELA2, and WAS were analyzed by previously described procedures.^12–15For sequencing analysis of HAX1, genomic DNA was extracted fromthe peripheral blood leukocytes. Polymerase chain reaction (PCR)was performed with primers that span all exons of HAX1 (5'-CGTTTACGACAGTGTCAGGATCG-3' and 5'-TGACAAACTGACATGGCCCCAG- 3') using the Expand HighFidelity PCR System (Roche Diagnostics GmbH, Mannheim, Germany).The PCR products were sequenced using 9 primers (primer designwill be provided upon request). These were then analyzed bycycle sequencing using a BigDye Terminator v3.1 cycle sequencingkit (Applied Biosystems, Foster City, CA) and an ABI PRISM 310genetic analyzer (Applied Biosystems).

To perform immunoblot analysis of HAX1, cell extracts of peripheralblood leukocytes were subjected into SDS-PAGE. The proteinswere transferred to polyvinyldifluoride membranes (Immobilon-P,Millipore, Bradford, MA) using a semi-dry electroblotter (ATTOCorp., Tokyo, Japan). Following the transfer, immunoblots wereincubated with a monoclonal antibody against HAX1 (BD BiosciencesPharmingen, Franklin Lakes, NJ). After washing the membranes,the bound secondary anti-mouse antibody was labeled for visualizationusing SuperSignal (Pierce Biotechnology Inc., Rockford, IL)by enhanced chemiluminescence (FUJIFILM Corp., Tokyo, Japan).

Results and Discussion

TOP

Results and Discussion

Sequence analysis of the patient demonstrated wild type forCSF3R, ELA2, and WAS. Next, we analyzed HAX1 mutations of thepatient, and identified a homozygous single-base pair substitution(256C-to-T) leading to a stop codon at codon 86 (R86X). Thisdeficiency was confirmed by the absence of HAX1 protein in immunoblotanalysis. The site of this mutation was consistent with thatreported in one of 23 SCN patients in the previous study.⁷ Althoughneither parent showed a detectable phenotype, both parents hada heterozygous R86X mutation with the expression of HAX1 proteins.

HAX1 has recently been identified as a cause of SCN by Kleinet al., who analyzed 23 ELA2 mutation-negative SCN patients.⁷According to the mutation sites, the 23 patients were classifiedinto 3 groups; W44X in 19 patients mainly in Kurdish and Turkishpopulations, R86X in 1 Iranian patient, and Q190X in 3 Swedishpatients from an original Kostmann family.⁷ One implicationof this finding might be that mutation sites are linked to ethnicity.Our patient is the second patient identified as harboring R86Xmutation, and the first patient with HAX1-deficient SCN in theJapanese population.

Although not fully understood, HAX1 seems to play a criticalrole in apoptosis and the organization of cytoskeleton.^8,16–18HAX1 is a ubiquitously expressed human gene,8 and thereforesymptoms other than myelopoietic disorder could be present inpatients with this deficiency. Indeed, some of the 23 Klein’spatients were associated with conditions other than low ANC.⁷These included splenomegaly (n=8), growth failure (n=2), muscularhypotonia (n=1), and tricuspid insufficiency (n=1).⁷ Of these,splenomegaly is a common feature with SCN irrespective of theunderlying genetic defects.¹ Our patient has failed to thrive,a condition concordant with the finding in the first patientshowing the same R86X mutation.⁷ Together with dental developmentdelay, this may be a reflection of a malfunction in cytoskeletonorganization due to HAX1 deficiency.⁸

Of special note is the finding that our patient showed severedelay in motor and intellectual milestones. Several explanationsof such developmental delay exist. One possibility is derivedfrom hearing impairment due to sequelae at the first presentationof otomastoiditis. However, since hearing impairment was moderateand unilateral, it is unlikely that this was the cause of severedevelopmental delay, and is, at least, unrelated to the delayof motor function. The second possibility is that infectiondue to initial otomastoiditis might have extended to the centralnervous system. However, this possibility is unlikely becausethe patient’s general condition had remained well andbrain CT scans at that time did not show any abnormalities.Thus, the most plausible explanation is attributable to HAX1mutation. This explanation is supported by the report¹⁹ by Carlssonand Fasth, who described neurological impairments in 3 of 6SCN patients of the original Kostmann pedigree; some of thesepatients were later demonstrated to be HAX1-deficient.⁷ Of thethree, one had mild mental retardation, the second showed speechand gross motor retardation, and the third demonstrated attentiondeficit hyperactivity syndrome plus epilepsy.¹⁹ Furthermore,an experimental study showed that HAX1 is maximally expressedin brain tissues on Western blot analysis.⁹ N2a cells, a mouseneural cell line, also contained significant amount of HAX1mRNA with subcellular localization in the cytoplasm and mitochondria.⁹Collectively, HAX1 deficiency may potentially affect neuraldevelopment, reflecting severe developmental retardation inour patient. However, whether neurological complication in ourpatient is linked to mutation sites in HAX1 or to HAX1 deficiencyitself is not clear. In addition, since diverse proteins haverecently been identified to interact with HAX1,^8,16–18some of these or other yet to be determined genetic factor(s)may be involved in this condition. Further accumulation of similarpatients is needed to clarify this question.

In conclusion, we presented a sporadic SCN patient with a nonsensemutation in HAX1. Importantly, this is the first case in theJapanese population and the second patient with an R86X mutation.Furthermore, this patient demonstrated severe developmentaldelay and epilepsy.

Acknowledgments

We thank Dr. S. Okamura, First Department of Internal Medicine,Faculty of Medicine, Kyushu University, Fukuoka, Japan, foranalyzing mutations in CSF3R gene. This study was supportedin part by Grants from Japan Intractable Diseases Research Foundation,Kawano Masanori Foundation for Promotion of Pediatrics

Footnotes

Authors’ Contributions: KM was responsible for clinicalmanagement and wrote the manuscript. KI contributed to the designof the study, genetic analysis, and drafting and editing ofthe manuscript. SO, MM, NI, and MT carried out genetic analysisand immunoblotting. TK and OO carried out genetic analysis.SN and MK designed the study and critically reviewed the manuscript.All authors approved the version submitted for publication.

References

TOP