Adult siblings with homozygous G6PC3 mutations expand our understanding of the severe congenital neutropenia type 4 (SCN4) phenotype

Severe congenital neutropenia type 4 (SCN4, OMIM# 612541) is caused by autosomal recessive mutations in G6PC3 on 17q21.31, which encodes one of the three subunits of the enzyme glucose-6-phosphatase. Its features include congenital neutropenia, a prominent superficial skin venous pattern and variable congenital malformations [1, 2]. The syndrome was delineated in 2009 and although 49 other affected individuals have been reported to date, only 8 others were adults [1‐12]. Oculocutaneous albinism type 4 (OCA4, OMIM# 606574) is a rare form of OCA except in Japan. Affected individuals have varying degrees of ocular and cutaneous hypopigmentation with visual acuity ranging from 20/30 to 20/400 [13, 14]. OCA4 is caused by autosomal recessive mutations in the solute carrier 45, member 2 gene (SLC45A2) on 5p13.2 [15]. Here we report two siblings, a sister and a brother, from Newfoundland, Canada who were originally thought to have a novel, undescribed syndrome. The sister was previously published by Cullinane et al., 2011 [16]. A combination of Sanger dideoxy-based sequencing and next generation sequencing uncovered that the complex phenotype in this family is due to G6PC3 and SLC45A2 mutations. We present the clinical features of the proband and her brother, and review the phenotypes of the SCN4 patients published to date. We suggest that the siblings developed several medical problems, including Crohn disease, which up until now have not been appreciated as features of SCN4 due to limited information regarding the adult phenotype of this very rare disorder.

This study was approved by Memorial University’s Human Research Ethics Authority (Reference# 11.060) and written informed consent was obtained from the patients.

NC1 and NC2 were the 3rd and 4th of four children born to second cousin parents who are of Irish descent. The parents’ first born is a healthy 43-year-old male. Their second born was a male who had obvious features of OCA at birth including white hair; he died at age 3 weeks from sepsis. There are two other autosomal recessive disorders (pyridoxine dependent seizures and beta-galactosidase deficiency) present in more distant relatives (Figure 1).

Three homozygous regions common to both siblings were identified, the largest of which was a 17.9 megabase (Mb) region on 17q12-21.33 which contains more than 425 genes. Using a candidate gene approach, both siblings were found to have a homozygous nonsense mutation of the SCN4 gene, G6PC3 (17q21.31). This mutation [c.829C > T, p.Gln277X] was previously reported in a French SCN4 patient [2].

Because the SCN4 mutation did not fully explain the siblings’ phenotype (i.e. the OCA), exome sequencing was performed on the proband. As we were analyzing the data, the proband (NC1) was published by Cullinane and colleagues [16]. These authors also used homozygosity mapping and exome sequencing approach to identify the homozygous SCN4 mutation, as well as a homozygous mutation in the OCA4 gene, SLC45A2 on chromosome 5p13.2. Our exome data showed the same OCA4 mutation [c.986delC, p.T329RfsX68], which we confirmed by Sanger sequencing as homozygous in the proband and heterozygous in her brother. The c.986delC variant was previously reported as a mutation in five German OCA4 patients [14].

Evaluation of the siblings in childhood suggested they had the same provisionally unique autosomal recessive disorder. Given their ocular and skin hypopigmentation, immune problems and CD, the best unifying diagnosis was considered to be a variant of Hermansky-Pudlak syndrome, although the proband’s platelet electron microscopy results made this diagnosis less likely. Undoubtedly the pediatric diagnosis of “mild albinism” in the proband’s brother was influenced by the fact that the siblings had several shared medical features (ASD, neutropenia, abnormal platelets) coupled with the fact that NC1 had classic OCA. Both siblings have a homozygous stop mutation in G6PC3 and SCN4 explains many of their medical problems. NC1 also has a homozygous SLC45A2 frameshift mutation, whereas her brother (NC2) carried only one OCA4 mutation. The OCA4 gene is not located within one of the three homozygous blocks shared between the siblings, but is within one of NC1’s identical by descent (IBD) regions. Following identification of the siblings’ homozygous SCN4 mutation, if we had considered variants from NC1’s exome analysis located only within regions of homozygosity shared between the sibs, the SLC45A2 mutation would have been missed. Thus, this analysis underscores the difficulties related to correctly delineating medical syndromes and identifying the responsible gene when more than one recessive disorder segregates in a family. Although we have not excluded the possibility of digenic inheritance, we suggest that NC2’s mild hypopigmentation was a heterozygote manifestation; this has not to our knowledge been previously reported in any of the recessive forms of OCA.

Apart from NC1 and NC2, 49 other individuals with SCN4 have been described in the literature, most of whom were children [1‐12]. Including NC1 and NC2, 10 adults have been reported with an age range of 20-38 years [1, 3, 6, 8]; NC1 (38 years old) and NC2 (who died at 37) are the oldest individuals with SCN4 published to date. The clinical features of all 51 affected individuals are summarized in Table 1. The most common birth defect is ASD, present in 38 of 51 (74.5%) individuals, followed by cryptorchidism in 52% of males. The Newfoundland siblings developed thrombocytopenia which has been reported in 27 other patients [1‐7, 10‐12], suggesting that this is also a key SCN4 hematologic manifestation. NC2 was diagnosed with mitral valve prolapse in childhood and likely developed progressive mitral valve disease. Cardiac valve abnormalities were present in 11/49 previously reported patients, including six cases with mitral and/or tricuspid insufficiency, two with pulmonic stenosis and one with mitral stenosis [1, 2, 6‐8, 10, 11]. This suggests that valvular abnormalities are part of the cardiac phenotype.

Banka and colleagues [1] speculated that individuals with SCN4 might have recognizable dysmorphic facial features and earlier this year Boztug et al.[7] reported variable facial dysmorphology in 12/16 patients. Only two other photographs of affected individuals have been published [1, 7]. One of these is a 26-year-old male who had midface hypoplasia, full lips and prognathism [1], features also present in NC1 and NC2, suggesting that individuals with SCN4 may have a recognizable facial gestalt.

Of the eight previously reported adults with SCN4, one 24-year-old female was developmentally delayed [6] and another three adults had an intellectual disability (ID) [1] (Table 1). This led Banka et al.[1] to hypothesize that ID is part of the syndrome. The three individuals with ID that they described belonged to a large consanguineous family and may have had a second recessive disorder. NC1 and NC2 successfully completed post-secondary education suggesting that ID is a variable SCN4 manifestation or that it is not part of the syndrome. Although neither NC1 nor NC2 had failure to thrive or growth failure, these are probably variable features. Failure to thrive has been reported in 22/49 (44.9%) of patients [1‐3, 5‐10] (Table 1). Moreover six of eight previously reported adolescents/adults had short stature [1, 6‐8] (Table 1).

G6PC3 is a ubiquitously expressed gene which encodes a subunit of glucose-6-phosphatase [21, 22]. This enzyme catalyzes the final step in the glycogenolytic and gluconeogenic pathways, the conversion of glucose-6-phophate to glucose [23]. Glycogen Storage Disease Ia (GSDIa, OMIM# 232200) is caused by autosomal recessive mutations in G6PC1, which encodes one of the other subunits of glucose-6-phosphatase. Banka et al.[1] suggested that SCN4 patients might develop some of the same long-term complications as individuals with GSDIa including pulmonary hypertension, renal failure and platelet dysfunction. Our patients both developed end stage renal disease in the 3rd decade. As adolescents, they were diagnosed with intermittent thrombocytopenia and a platelet aggregation defect. NC1 developed pulmonary hypertension in her mid-30’s and is the 5th reported SCN4 patient who developed this complication [1, 5, 10] (Table 1). Collectively, these findings support the hypothesis that long-term complications may be shared between individuals with mutations in G6PC3 and G6PC1.

NC1 and NC2 developed severe adolescent-onset Crohn disease, and we hypothesize that the siblings’ CD is SCN4-related. Firstly, 15 SCN4 patients older than age 12 years have been published including two others with CD [8], a 30-year-old Pakistani male who was not treated with G-CSF until at least age 9 years and his sister, about whom few other clinical details were provided. Secondly, Glycogen Storage Disease Ib (GSDIb, OMIM#232220) is due to autosomal recessive mutations in SLC37A4 which encodes a protein that transports glucose-6-phosphatase from the cytosol to the endoplasmic reticulum where the enzyme acts [24]. Some individuals with GSD1b develop neutropenia and there is a high incidence of CD in GSD1b patients, but only in the neutropenic subset [25, 26]. In 2002, Dieckgraefe et al.[26] ascertained 36 North American GSD1b patients of whom 28% (all with neutropenia) had CD with an average diagnosis age of 8-9 years. The gastrointestinal symptoms in all 10 patients improved with G-CSF. Finally, Crohn disease has also been associated with other neutrophil-centered immunodeficiency syndromes including chronic granulomatous disease [27], Chediak-Higashi syndrome [28], leukocyte adhesion deficiency [29], cyclic neutropenia [30] and autoimmune neutropenia [31]. In 2000, Korzenik and Dieckgraefe [32] hypothesized that defects in innate immunity predispose some individuals to CD. They suggested that although standard CD therapy dampens the chronic recruitment of B and T-cells to the gastrointestinal mucosa, in some CD patients an earlier step is failure of adequate mucosal neutrophil function in the face of a microbial challenge. If this pathophysiologic mechanism is correct, then using G-CSF to normalize the ANC of patients with CD and a neutrophil deficiency syndrome should improve the course of their CD.

In summary, our findings expand the natural history of SCN4. Additional reports are needed to determine if other older SCN4 patients have developed Crohn disease, other GSD1a/b complications or valvular heart disease. If CD is indeed a feature of SCN4, then early treatment with G-CSF is likely to ameliorate its course and presymptomatic treatment could prevent its occurrence. Meanwhile we suggest monitoring SCN4 patients for valvular heart disease and pulmonary hypertension with serial echocardiograms.