Clinical Phenotype in Individuals With Birk-Landau-Perez Syndrome Associated With Biallelic SLC30A9 Pathogenic Variants

Clinical Description

Pedigrees for the multiplex families (1 and 3) are shown in Figure 1. Family 1 is a British family of Pakistani origin, in which the parents were first cousins. The 2 eldest brothers, F1(II-1) and F1(II-2), presented with a similar neurologic phenotype. Both boys were born at term with unexplained low birth weight, <0.4th centile. Both achieved early neurodevelopmental milestones at the later end of the normal range, walking at 18 months, and both were diagnosed with moderate learning difficulties but were able to continue in mainstream education with specialized support.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Pedigrees and Sequencing Data

Above: pedigrees of families 1 and 2. Key: circle: female; square: male; double horizontal line: consanguineous union; filled shape: affected individual. Below: chromatograms showing Sanger sequencing results for selected family members. Family 1: affected base is shaded in blue; note that parents are heterozygous (double trace) and affected children are homozygous. Family 2: the affected codon in parental traces is shaded in blue; note double trace emerging (to the left) indicating heterozygosity. Representative trace from 1 affected child's shows codon homozygously absent.

Patient F1(II-2) presented with a movement disorder at around age 7 years. Patient F1(II-1) did not develop a movement disorder until age 16 years, and it remains somewhat milder than his brother's. In both cases, the movement disorder consists of continuous, irregular, low-amplitude choreiform movements and low-amplitude myoclonic jerks of all 4 extremities, together with some subtle perioral dyskinesia and mild dystonic posturing on walking (Videos 1–3). Gait was effortful for both, and by age 16 years, patient F1(II-2) was only able to walk a few steps without assistance. Patient F1(II-2), but not F1(II-1), developed a progressive thoracic kyphosis. Neither of the boys displayed any spasticity, focal weakness, or any convincing signs of cerebellar involvement, although speech was moderately dysarthric in both. Both also had musculoskeletal contractures: in F1(II-1), these involved the fingers bilaterally, and in F1(II-2), the elbows and the left foot. Patient F1(II-1) had mild difficulties with saccadic eye movements, and F1(II-2) developed clear oculomotor apraxia. Both had subtle bilateral ptosis. Patient F1(II-1) also has moderate bilateral sensorineural hearing impairment. Neither patient had any impairment of kidney function at the time of assessment, but patient F1(II-2) was known to pediatric nephrology for unexplained small echogenic kidneys, and also under investigation for possible hypertension.

Brain MRI for patient F1(II-2), aged 11 years, was locally reported as normal, but review by a pediatric neuroradiologist identified subtle abnormalities including a shallow pontomedullary sulcus, short clivus with retroslanting odontoid, low cerebellar tonsils, a mildly dysmorphic corpus callosum, and borderline small volume putamen and caudate (Figure 2). Detailed diagnostic metabolic testing was nondiagnostic including normal urea and electrolytes, calcium, magnesium, liver function tests, very-long-chain fatty acids, ammonia, creatine phosphokinase, transferrin glycoforms, biotinidase, plasma amino acids, urine organic acids, glycosaminoglycans, and urine guanidinoacetate. He was found to have autoimmune hypothyroidism treated with levothyroxine. CSF showed normal paired glucose, lactate, glycine, and pterins, with a mild isolated reduction in 5-hydroxyindoleacetic acid (51 nmol/L, range 58–220), which was not thought to be significant. A muscle biopsy showed mitochondrial DNA at borderline levels of 40%, but respiratory chain enzyme levels and sequencing of the mitochondrial genome were normal. Serum levels of zinc, copper, and selenium were normal. Both brothers were treated with a low dose of levetiracetam, which they felt helped reduce their abnormal movements.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 MRIs From Participants

(A) F1(II-2) sagittal, T1 weighted, showing dysmorphic corpus callosum (red arrow) and shallow pontomedullary sulcus (blue arrow). (B) F1(II-2): axial, T2 weighted, showing small-volume basal ganglia (green arrow). (C) F2 sagittal, T2-weighted, showing vertical posterior corpus callosum (red arrow). (D) F2 axial, T2 weighted, showing symmetrical hyperintensity of the globus pallidus and posterior putamen (blue arrows). (E) F3(II-5) sagittal, T1-weighted, showing vertical posterior corpus callosum (red arrow). (F) F3(II-5) axial, T2-weighted, showing small globus pallidus with altered signal intensity (blue arrow).

Family 2 is also British Pakistani. The parents are not consanguineous but are from the same geographic area. F2 has 2 healthy older siblings and was born by induced labor at 38 weeks due to intrauterine growth restriction, with a low birth weight of 2.1 kg. He walked at 14 months, but from age 2 years, his gait became unsteady, and he started to toe-walk. Developmental milestones were slightly delayed in all areas relative to his siblings. On examination, at age 4 years and 8 months, dystonia and low-amplitude myoclonic jerks were evident in all limbs, and orolingual dyskinesia was seen (Video 4). On purposeful movements such as finger-nose pointing and finger opposition, he was bradykinetic. Mild spasticity and brisk deep tendon reflexes were present in the lower but not upper limbs, and Achilles tendons were tight. Plantars were downgoing. He was microcephalic with a Z score of −4.08 for head circumference and had very subtle ptosis on the right only. He had difficulties with both saccadic and smooth pursuit eye movements. Moderate bilateral sensorineural hearing loss was diagnosed at age 5 years.

Brain MRI showed very subtle symmetrical T2 hyperintensity of the globus pallidus and posterior putamen, minimal cerebellar volume loss, and a dysmorphic (posteriorly vertical) corpus callosum (Figure 2). Microarray and initial neurometabolic testing were normal, except that both urea and creatinine were modestly elevated: at age 5 years, he was found to have moderately impaired renal function (estimated glomerular filtration rate of 40 mL/min/1.72 m² [range >90 mL/min/1.72 m²]) with small echogenic kidneys and no proteinuria.

Family 3 is a Palestinian family of Egyptian origin with multilevel consanguinity: the parents are double first cousins. They are not known to have Bedouin heritage. The 4 affected siblings all presented within the first 2 years of life with a progressive movement disorder. Patient F3(II-1) never walked, but the other 3 (F3(II-2), F3(II-3), and F3(II-5)) were able to walk briefly from about 20 months old, losing ambulation at 2 years. All 4 children had evidence of bradykinetic voluntary movements, with generalized dyskinesia and dystonia involving the limb extremities and trunk. Ptosis and oculomotor apraxia were also universally evident. All 4 had a severe intellectual disability with limited verbal communication. None are known to have any renal impairment.

Brain MRI for patient F3(II-5) was reported locally as normal, but pediatric neuroradiologic review again identified subtle abnormalities: minimal cerebellar volume loss; the globus pallidus was small with altered signal intensity; the corpus callosum showed posterior vertical morphology; and there was unusual perirolandic sulcation on the right (Figure 2). Due to resource constraints, neither imaging of the other siblings nor serum zinc testing could be performed.

Family 4 is White Australian, and the parents are not consanguineous. The proband was diagnosed with bilateral sensorineural hearing loss at age 13 months and managed with cochlear implants. She had global developmental delay and a mildly abnormal gait but did acquire the ability to walk and run. From age 5 years, she developed a slowly progressive movement disorder involving her trunk, limbs, face, and eyes. Examination at age 10 years showed generalized dystonia, ataxia, bilateral ptosis, and oculomotor apraxia (Videos 5 and 6). Brain MRI (aged 2 years) was normal. Kidney function and renal ultrasound at age 10 years were normal. Serum zinc and manganese levels were normal.

Molecular Genetic Investigations

Sanger confirmation for the variants in families 1 and 3 is shown in Figure 1. In family 1, 14 rare homozygous variants predicted to change the amino acid sequence or occurring within 10 bases of a splice site were identified, but of these, 10 were consistently predicted by in silico tools to be benign, including Combined Annotation Dependent Depletion scores <20. Of the remaining 4, 2 were associated with clearly irrelevant phenotypes: an autosomal dominant craniofacial disorder in the case of FGFR1 and an autosomal dominant hepatic disorder for SEC63. The SEC63 variant, moreover, although close to a splice site, was not predicted to have a significant effect on splicing. A third variant, in IARS2, was found not to segregate with disease when the similarly affected sibling was tested. This left only a variant in SLC30A9: NM_006345.3:c.1253G>T, p.Gly418Val. In silico predictors of pathogenicity consistently predicted a deleterious effect on protein function (Table 1). Heterozygous and compound heterozygous variants in F1(II-2)'s genome were also checked using the same large gene panel, but no additional likely pathogenic candidates were identified. The same variant was subsequently identified in family 2.

For family 3, the previously published pathogenic variant in SLC30A9 (NM_006345.3:c.1049delCAG, pAla350del) was identified in a region of homozygosity shared between all 4 affected siblings. No other known pathogenic variants in relevant genes were found in these regions. Sanger sequencing confirmed that the variant segregated as expected with disease in the family, with both parents being heterozygous carriers.

In proband 4, compound heterozygous variants in trans were present as follows: NM_006345.3:c.1083dup, p.Val362Cysfs*5 (maternal) and NM_006345.3:c.1413A>G, p.Ser471= (paternal). No other variants likely to be pathogenic were identified. The first variant, as a frameshift, was predicted to result in protein truncation and probably nonsense-mediated decay. The second, synonymous, variant was predicted to affect splicing. As in silico splicing predictors were equivocal, this was verified experimentally (see below).

Modeling

The novel variant identified in families 1 and 2, p.Gly418Val, is present in the loop connecting 2 transmembrane helices (TM5 and TM6, Figure 3) and is a partially conserved residue, as observed from the multiple sequence alignment obtained using ConSurf. Because of glycine's unique nature, it is known to provide much more conformational flexibility than other amino acids. Glycine, when present in the transmembrane regions, is also reported to facilitate helix packing in membrane proteins and is therefore important for the association of transmembrane helices.^26,27 Substitution to valine, which has a larger hydrophobic side chain, will decrease the flexibility of the loop and may affect transmembrane helix packing.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 Structure Model of the Transmembrane Domain of Zn-T9 From the AlphaFold¹⁹ Protein Structure Database (Residues 221–449)

The black discs indicate the predicted position of the membrane. (A) Gly418 is shown as red spheres and is present in the loop connecting 2 transmembrane helices. (B) Gly418Val substitution is shown is red on the structural model and indicated with an arrow: valine is predicted to provide less conformational flexibility than glycine, and the substitution is predicted to decrease the stability of the loop and affect transmembrane helix packing.

Splicing Experiment

The synonymous variant (c.1413A>G, p.Ser471=) occurs 5 base pairs before the end of exon 15 (in transcript NM_006345.3). Unexpectedly, electrophoresis of fibroblast-derived cDNA from both the proband and 2 unaffected controls revealed a double band, indicating 1 splicing product of the predicted size and a second, smaller product in which exon 16 and the last 5 base pairs of exon 15 were skipped. This suggests that an undocumented additional transcript exists in which the variant is exactly on the exon boundary. Sequencing of the larger band identified that in fact, 2 different products were present in the proband (but not the controls), with a double trace beginning at the site of the variant (Figure 4). These 2 products were too similar in size to be resolved by gel electrophoresis, differing only by 5 bp, but sequencing indicated that the trace unique to the proband had lost the final 5 bp of exon 15. This would be expected to result in a frameshift and premature termination as follows: p.Val472Glyfs*4.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4 Results of Splicing Assay for the Synonymous Variant Found in F4

(A) Gel electrophoresis of the cDNA from patient F4. Bands on the far left represent the DNA ladder. For both proband and control, 2 bands are present, (red and blue arrows) indicating (at least) 2 splicing products. (B) Chromatogram showing Sanger sequencing results of cDNA isolated from the upper band (red arrow in A). The control shows a single trace; the proband has a double trace commencing at the exon 15–16 boundary indicating the presence of 2 similarly sized splicing products. One of these is wild type; the other (letters below) is missing the final 5 base pairs from exon 15. (C) Schematic showing the different splicing products: (C.a) is the normal larger product, including all of exons 15, 16, and 17; (C.b) is the normal smaller product, in which exon 16 and the final 5 bp of exon 15 are skipped; (C.c) is the abnormal larger product in which the final 5 bp of exon 15 only are skipped, resulting in a frameshift. bp = base pair; cDNA = complementary DNA.