MAN1B1 Mutation Leads to a Recognizable Phenotype: A Case Report and Future Prospects

Early-onset cognitive impairment, also referred to as intellectual disability (ID), has a prevalence of about 3% in the general population and represents a major socioeconomic burden worldwide [Leonard and Wen, 2002; Musante and Ropers, 2014]. In fact, ID is one of the most common reasons for referral to genetic counseling [Ropers, 2010; own unpubl. observations]; yet, in over 50% of the cases, routine examination does not reveal the cause or clinical entity. Whole-exome sequencing has enabled a rapid progress in this field, and recent studies show that undiagnosed cases are probably due to de novo dominant or inherited recessive point mutations in autosomal genes [Najmabadi et al., 2011; de Ligt et al., 2012; Rauch et al., 2012; Gilissen et al., 2014]. Here, we report on our clinical and scientific experience in this field, leading to the suggestion that whole-exome sequencing should be implemented in a larger scale in clinical genetic practice. A family sought a diagnosis for their 3 affected children with mild ID. We identified a mutation in the MAN1B1 gene and thereby, retrospectively, diagnosed a recently described syndromic form of ID. Further, we discuss the phenotypic aspects in our family in comparison to reported cases in the literature.

A consanguineous Turkish family with 3 children showing mild to moderate ID presented at the Department of Human Genetics of the Ruhr-University Bochum (fig. 1). At the time of the first physical examination, the siblings were 17, 14 and 10 years old. For all 3 children, pregnancy and delivery were uneventful. They concordantly showed muscular hypotonia from the beginning, and the motor milestones were rather late, though in the normal range (e.g. unassisted walking was achieved at ∼18 months of age). All 3 children were rather small (but in the range of the parents' height) and displayed truncal obesity that started at ∼5 years of age. They showed foot deformities (flat feet) and tapering fingers and had common facial features including a prominent nose, short philtrum, a thin upper lip and broad eyebrows; however, these features were considered unspecific at the beginning.

ID was diagnosed in the daughter (III.1) when integration into a regular primary school was not possible. At the age of 17, she was able to read and write a few words as well as understand and follow simple assignments. She attended a special school for intellectually disabled children and did not show any other malformations, medical or behavioral symptoms. Karyotyping as well as analysis of fragile X syndrome yielded normal results. Her brother (III.2) showed comparable psychomotor delay. In addition, he developed pectus excavatum and underwent surgery at the age of 5 years. The youngest brother (III.3) showed comparable psychomotor development and additionally had strabismus convergens. Ultrasounds of muscle and brain in infancy, ophthalmologic evaluation and orienting metabolic tests (lactic acid and ammonia in serum, amino acid analysis in plasma and urine, and organic acid analysis in urine) were inconspicuous in the youngest brother. An MRI at the age of almost 3 years showed slight widening of the outer ventricles. At ∼3 years of age, he had one short episode of non-ketotic hyperglycemia (blood sugar level >350 mg/dl). An autoimmune-mediated diabetes mellitus type 1 was excluded and subsequently blood sugar levels remained within the normal range without further therapy.

Due to the lack of obvious malformations or clearly dysmorphic features, the affected children were classified as having unspecific ID, most probably of autosomal recessive inheritance.

As described by Abou Jamra et al. [2011], autozygosity mapping in the family with DNAs from the 2 affected brothers restricted the candidate regions to 12 with a total length of 170 Mb. Subsequently, exome sequencing using DNA from individual III.2 was performed as described in former studies [Hansen et al., 2013; Murakami et al., 2014], resulting in an average coverage of 58×. 65% of the target sequences were covered with a depth of at least 20×, and 79% were covered with a depth of at least 5×. 466 SNVs and indels were rare (i.e. minor allele frequency of ≤0.01 in the databases of 1000 Genomes and Exome Variant Server (EVS) as well as in 700 in-house exomes) and predicted to affect the protein sequence (nonsynonymous or splicing). Under the hypothesis of a homozygous identical by a descent founder mutation, we considered homozygous variants followed by filtering based on in silico parameters (conservation and prediction programs, see online suppl. table 1; filtering strategy, see online suppl. fig. 1 ; see www.karger.com/doi/10.1159/000371399 for all online suppl. material). This revealed only one candidate mutation: c.1000C>T, p.R334C in exon 7 of MAN1B1(NM_016219). We validated the mutation and confirmed segregation in the family using Sanger sequencing.

The MAN1B1 gene is located on chromosome 9q34.3 and comprises 13 exons (fig. 2). It encodes the protein endoplasmic reticulum mannosyl-oligosaccharide 1,2-alpha-mannosidase (ERManI) that is believed to play an important role in the disposal of misfolded glycoproteins [Karaveg et al., 2005]. Rafiq et al. [2011] first described mutations in this gene, identified by homozygosity mapping and next-generation sequencing, as causative for nonsyndromic ID in 4 consanguineous families from Pakistan and Iran. The affected children in the Iranian family carried the same missense mutation (c.1000C>T, p.R334C) that was found in our Turkish family (fig. 2; table 1). Shortly thereafter, additional MAN1B1mutations were identified in a group of patients with congenital disorder of glycosylation type 2 (CDG2) syndrome [Rymen et al., 2013], which was confirmed in a second cohort [Van Scherpenzeel et al., 2014]. Despite individual variability in clinical phenotype, most described patients shared truncal obesity, a broad nasal bridge, a prominent nose, a thin upper lip, down-slanting palpebral fissures, broad eyebrows with lateral thinning, and muscular hypotonia, suggesting that MAN1B1 deficiency may rather be a syndromic form of ID [Rymen et al., 2013; Van Scherpenzeel et al., 2014].

The mutations described so far include nonsense, missense and splice site mutations as well as deletions of single exons and are located throughout the gene (fig. 2). The p.R334C mutation has been found most frequently so far, being in homozygous state in an Iranian family with 3 children [Rafiq et al., 2011], 2 Turkish families with CDG2 [Rymen et al., 2013], a third family with a CDG2 patient of unknown origin [Van Scherpenzeel et al., 2014] as well as in the 3 affected children in our Turkish family, summing up to 10 individuals from 5 unrelated families, mostly of Turkish origin (table 1). This mutation is located at the base of the active-site pocket and thus is believed to influence enzymatic function and substrate binding [Van Scherpenzeel et al., 2014].

Screening the EVS database (http://evs.gs.washington.edu/EVS/) revealed 30 reliable (coverage of >20×) heterozygous variants in the MAN1B1 gene in 51 individuals that are predicted ‘damaging' by PolyPhen2 and highly conserved (online suppl. table 2), theoretically giving a carrier frequency of 51/6500 (0.0078).

In the present study, we further confirm p.R334C in the MAN1B1gene as a causative mutation in autosomal recessive ID. Taking into account data from the literature, it appears that p.R334C is a founder mutation in the Turkish population. Autosomal recessive ID is highly heterogeneous, and private familial mutations are usually identified [Musante and Ropers, 2014]. Nevertheless, further examples such as a recently reported mutation in CLP1 in 4 independent Turkish pedigrees with neuropediatric disease [Schaffer et al., 2014] show that one should be aware of such founder mutations which may be reported as variants in public databases. Based on EVS data, potentially damaging mutations in MAN1B1 have a carrier prevalence of 0.0078 (online suppl. table 2), and the disease prevalence may be as high as 0.000015, comparable to several well-known metabolic disorders (http://www.orpha.net/).

The phenotype associated with MAN1B1mutations appears comparatively uniform for the patients described so far (summarized in table 2). This includes mild to moderate ID, truncal obesity, muscular hypotonia and modest dysmorphic features (broad nasal bridge, prominent nose, thin upper lip, curved eyebrows) as well as a glycosylation type typical for CDG2 [Rymen et al., 2013; Van Scherpenzeel et al., 2014]. Although mutations in MAN1B1 lead to a syndromic disorder, we have to admit that this was only achieved retrospectively in our patients. This is probably a result of the growing number of genes and mutations that make it a difficult task to establish differential diagnoses. Although for some distinct phenotypes clinical examination will still draw attention to a specific disorder or syndrome, it seems that this is getting more and more difficult. We, therefore, suggest that exome sequencing should be implemented on a larger scale in routine diagnostics for ID, especially in cases suspected of autosomal recessive inheritance. We acknowledge that interpretation of the results may not always be as forthcoming as in our case and especially distinguishing disease-causing mutations from benign sequence variants has proven a difficult task [MacArthur et al., 2014]. However, large-scale application of whole-exome sequencing, and eventually also whole-genome sequencing which can additionally detect noncoding regulatory variants, appears warranted in order to get a more comprehensive picture of autosomal recessive ID in both consanguineous and nonconsanguineous families in the future.

We thank the patients and their family for participation in the study. We thank Katharina Batzke and Farah Radwan for technical assistance and Arif Ekici, Mandy Krumbiegel and Steffen Uebe for support in mapping and NGS. This study was supported by a DFG grant AB393/2-2 to Rami Abou Jamra.