Supplementary Figure 1 Brain magnetic resonance imaging in patient 9 at age 3.6 years. (a d) Axial T2-weighted images show a marked degree of ventriculomegaly. Note the diencephalic mesencephalic junction dysplasia, characterized by a lack of separation between the hypothalamus and midbrain (arrowheads in a). In the nucleobasal region, bilaterally, there are dilated perivascular spaces (arrowheads in b) and an abnormal morphology of the basal nuclei, with prominent thalami (labeled T in c) and maloriented putamen and caudate heads (labeled P and C, respectively, in c). Note also the bilateral abnormal orientation of the posterior arms of the sylvian fissure (empty arrows in d). (e) An axial T1-weighted image obtained with a volumetric technique shows thickening of the left perisylvian cortex, consistent with polymicrogyria. Cortical abnormality is less prominent controlaterally, despite the abnormal orientation of the sylvian fissure. (f) A left parasagittal T1-weighted image shows abnormal orientation of the sylvian fissure (arrows), which is continuous with the sulci of the convexity without a normal posterior delimitation; this pattern, which was also visible controlaterally (data not shown), is typical of perisylvian polymicrogyria. (g) A midsagittal T1-weighted image shows shortened, hypoplastic corpus callosum with a thin splenium (thin arrows) and lacking a proper rostrum (thick arrow); also note the absence of a well-defined anterior commissure (arrowhead). 1
Supplementary Figure 2 Capillary electrophoresis analysis of NANS cdna retrotranscribed and amplified from RNA extracted from cell cultures. (a) Schematic of the exon intron structure of the NANS gene. Asterisks indicate mutations with a putative effect on exon splicing (see Table 1 for details): EXins: c.385_386inst, exonic single-nucleotide insertion producing a frameshift and creating a premature termination codon; SPLko: c.488+1g>t, a canonical splice donor site mutation; INTindel: c.449 10delGATTACinsATGG, an intronic indel 5 of exon 4. (b) Pedigrees of patients 1, 3, and 4, and corresponding capillary electrophoresis traces of semiquantitative RT- PCRs spanning all analyzed NANS mutations. We assessed unprocessed PCR products obtained with the same primer pairs and allowing the simultaneous ampli-fication of the wild-type and the mutant mrna (cdna) forms, indicated on the right, within the same reaction. See the main text for discussion of the results. 2
Supplementary Figure 3 Expression of NANS in human brain as extracted from the Brainspan database. In brain, NANS is expressed at significantly higher levels in the thalamus than in other brain regions (top) (http://developinghumanbrain.org/). Expression seems to be highest in the dorsal thalamus (bottom), which is an important relay station for the limbic circuit (The Human Nervous System 439 468 Academic Press, 1990) and participates in learning and memory processes (J. Neurosci. 34, 15340 15346, 2014), which might contribute to the developmental delay in patients with NANS deficiency. 3
Supplementary Table 1. Overview of the filtering of exonic and splicing variants observed in the patients. Pat1 Pat2 Pat3 Pat4. Pat.9 Number of exonic and splicing variants Number of nonsynonymous variants Number of rare variants (<1%) Number of rare variants after quality control (QC) Number of genes with 2 heterozygous variants 22737 22670 22600 23017 21921 11396 11358 11349 11578 10878 863 909 868 926 495 281 342 320 345 429 1 11 12 13 18 19 + in common in each sib pair + in common in all four patients In common in all 5 patients 4* 4** - 1*** - 1*** Values refer to number of variants unless specified otherwise. Rare (<1%): Variant frequency = 1% or less in public databases ExAC, ESP, and Wellderly, 1KG (from Complete Genomics). Rare (QC): Rare variants after quality control, i.e. after removal of (1) WES data of poor quality, with less than 15 reads per nucleotide and genotype quality less than 50, (2) variants present in control WES processed by the same pipeline (to remove technical error), and (3) variants which have healthy homozygous individuals in public database (ExAC). See URL section in main manuscript. 1 Family 3 (pat 9): the analysis was based on Vancouver pipeline that incorporates some of the quality control steps earlier in the pipeline
*Family 1 (pat 1 & 2): Genes: NANS, AQP7, SLCO1B7, ZNF429 **Family 2 (pat 3 & 4): Genes: NANS, ATP11A, OR14I1, PCDHB8 *** Gene: NANS
Supplementary Table 2. Abnormal features observed by liquid chromatography- QTOF mass spectrometry in the cerebrospinal fluid of patient 9. ESI positive mode ESI negative mode # Ranking Mass RT # Ranking Mass RT 1 244.07902 0.73 1 256.05916 0.69 2 204.08663 0.72 2 256.09295 0.73 3 123.05511 1.38 3 258.05683 0.68 4 219.11006 5.88 4 240.96941 18.37 5 246.10811 0.78 5 258.07227 0.88 6 168.06524 1.20 6 223.01996 14.79 7 137.08057 0.76 7 147.06612 2.79 8 138.05458 0.71 8 259.09362 0.74 9 316.21164 4.78 9 166.05073 1.25 10 99.04171 1.06 10 256.09391 0.97 11 195.11299 4.80 11 344.94818 0.72 12 221.01043 1.09 12 204.08818 0.83 13 170.08089 1.74 13 211.10854 5.68 14 220.97417 1.05 14 200.05651 3.14 15 130.97859 1.07 15 218.06673 0.91 16 429.31854 18.87 16 202.07188 1.48 17 235.0938 5.77 17 168.06649 1.45 18 206.10187 0.86 18 182.04549 3.01 19 281.08261 1.10 19 100.04102 0.82 20 327.1061 2.93 20 346.08176 0.66 21 317.21478 4.79 21 148.06532 2.76 22 196.13208 4.81 22 167.0543 1.24 23 191.0786 3.81 23 236.05632 2.17 24 184.06022 3.01 24 208.0615 3.07 25 226.06796 1.62 26 191.08086 4.40 27 177.07548 1.82 28 343.31986 1.06 0 The CSF samples of patient 9 and of 15 healthy controls were analyzed by reversed-phase liquid chromatography and high-resolution QTOF mass spectrometry on an Agilent 6540 QTOF. The features (signal with exact m/z (mass-to-charge ratio), RT (retention time) and intensity) obtained were aligned
using the XCMS software (https://metlin.scripps.edu/xcms/), after which statistical analysis by Bonferroni-corrected t-tests identified significantly different features between our patient and controls. Features that represented the same ion (adducts, ion-source fragments) were combined. In ESI-positive mode, 28 significantly different features were identified; in ESI-negative mode, 24 significantly different features were detected. Feature 1 in positive mode (m/z 244.07902 @RT 0.73) represents the Na-adduct of N- acetylmannosamine; Feature 1 in negative mode (m/z 256.05916 @ RT 0.69) represents the Cl-adduct of N-acetylmannosamine. In addition, feature 6 in ESI+ and feature 9 in ESI- represent the [M+H] + - adduct, respectively, the [M-H] - adduct of (iso)pyridoxal; feature 13 in ESI+ relates to the [M+H] + adduct of pyridoxine, which is also detected in negative mode (feature 17, [M-H] - adduct). In negative mode, also the [M-H] - adduct of pyridoxic acid is detected (feature 18). These latter features are secondary to medication with vitamin B6 for seizures. All these annotations have been confirmed by analysis of the chemical standard. Other possible annotations against the HMDB database are not listed as they have not been confirmed by chemical standards and are putative annotations. In general, features in this table were significantly different from control samples. They may partly relate to more distant effects of the metabolic perturbation caused by NANS deficiency, or, alternatively, be secondary to the perturbed development of the brain and/or ongoing epileptic activity. If the first hypothesis were true, others may be able to confirm such effects as m/z values and approximate retention times can be reproduced on other systems.
Supplementary Table 3. Analysis of free sialic acid (Neu5Ac) by mass spectrometry in urine, CSF and fibroblasts (see main manuscript for methods). Urine CSF Fibroblast cultures NeuNAc in µmol/mmol creatinine (age-related reference range) NeuNAc in µmol/mmol creatinine (age-related reference range) NeuNAc in in nmol/mg protein (reference range of 5 controls) Pat. 1 10 (2-11) Pat. 2 8 (2-11) Pat. 4 (Family 2) 0.8 (0.9-2.0) Father (Family 2) 0.7 (0.9-2.0) Pat. 8 11 (2-11) 3.7 (0.9-2.0) Pat. 9 20 (12-34) 8,1 (2-24,4) 1.3 (0.9-2.0)
Supplementary Table 4. Analysis of the glycosylation of plasma proteins. Pat.1 (Fam. 1) Pat.2 (Fam.1) Transferrin N-glycosylation (IEF) Pat.3 (Fam. 2) Pat. 4 (Fam. 2) Father (Fam 2) Mother (Fam 2) Pat. 9 Reference range (in %) TF-0 1,1 1,1 0,9 0,8 0,9 0,8 1,6 0-3.2 TF-1 2,1 2,3 1,5 2,6 1,3 1,4 3,4 0-5 TF-2 5,1 4,7 3,7 4,4 3,9 3,5 4,7 3.3-7.6 TF-3 11,8 12,9 14,1 11,8 8,9 11,7 8,8 4.9-10.6 TF-4 59,7 58,7 62,1 62 63,7 60,8 57,4 47.3-62.7 TF-5 16,2 16,7 15 15,2 17,1 17,7 19,7 18.7-31.5 TF-6 4,1 3,6 2,8 3,1 4,3 4,1 4,6 3.2-7.8 Transferrin N-glycosylation (QTOF) T3-TF 1 5,6 5,9 9,6 6,3 3,8 5,4 2,3 <5 ApoCIII O-glycosylation ApoCIII-0 4,4 2,3 6,7 5,3 4,3 4,9 5,5 1.4-9.5 ApoCIII-1 58,2 64,2 37,4 51,7 50 55,5 61,1 48.5-75.2 ApoCIII-2 37,4 33,6 55,8 43 45,7 39,5 33,3 21-45.8 1 T3-TF (Trisialotransferrin) corresponds to the fully glycosylated transferrin lacking a single sialic acid residue, as a measure for hyposialylation (molecular mass 79266 Da in reference 49). Analysis of transferrin N-glycosylation and apolipoprotein CIII mucin type O-glycosylation was done by isoelectrofocusing, and transferrin N-glycosylation also by nanochip-qtof mass spectrometry; see main manuscript for methods.
Supplementary Table 5. Sequences of PCR primers and morpholino oligonucleotides. Primer pair spanning the exon-exon junctions 1-2 and 5-6 of the human NANS gene (used to selectively amplify cdna retrotranscribed from fibroblasts) Forward: 5 -CAAGGAGTGTGGGGCTGATT-3 Reverse: 5 -CACCACAGACTTGCCCAGCTTCTC-3 D. rerio nansa-e3i3 morpholino 5 - AGGGAAATATTGGACCTTTTTGGGC -3 D. rerio nansb-e1i1 morpholino 5 - AAATACCGCATATCTTTACCTTGGC -3 D. rerio standard control MO 5 - CCTCTTACCTCAGTTACAATTTATA-3 D. rerio nansa primers used to amplify cdna retrotranscribed from 24hpf embryos Forward: 5 - GAACGGCCATACACCTCAAA -3 Reverse 3 - AGTTCTGATTGTGCTCCTTCAC -5
SUPPLEMENTARY NOTE Clinical Patient Reports