Human study participants
Affected individual 1 and parents of affected individuals 2, 3 and 4, and of fetus 5 provided their written consent for genetic testing, analysis of fibroblasts and publication of images. The study was performed according to the declaration of Helsinki and approved by the institutional Ethics Committees of Charité—Universitätsmedizin Berlin, Germany (EA2/101/18) and Necker Hospital Paris, France (IRB: 00011928, 2020-04-06). Information on sex, ethnicity and age of the study participants can be found in Extended Data Table 1. Blood, amniocyte and skin samples were obtained through standard procedures.
Exome and Sanger sequencing
DNA from individuals 1, 2, 3 and 4 and parents of individuals 2, 3, 4 and 5 was extracted from peripheral-blood lymphocytes, and from fetus 5 and 6 from uncultured amniocytes according to standard protocols. Sanger sequencing of all exons of TGDS was applied to DNA of individual 2. Segregation analysis of detected TGDS variants using Sanger sequencing was performed on DNA of the parents of Individual 2, father of individual 3 and parents of individual 4. The primers used for Sanger sequencing of TGDS are listed in Supplementary Data Table 1. Exome sequencing was performed on DNA of individuals 1, 3, 4 and 5 as well as the mother of individual 3 and parents of individuals 4 and 5. Exome sequencing of individual 1 was previously described2. The technical approach for exome sequencing of individuals 353 and 454 were previously described. For trio exome sequencing of individual 5, DNA was enriched using Agilent SureSelect DNA + SureSelect OneSeq 300 kb CNV Backbone + Human All Exon V7 capture, and paired-end sequenced on the Illumina platform (GenomeScan, Leiden, the Netherlands). The aim was to obtain 10 Giga base pairs per exome with a mapped fraction of 0.99. The average coverage of the exome is ~50×. Duplicate and non-unique reads were excluded. Data were demultiplexed with bcl2fastq Conversion Software from Illumina. Reads were mapped to the genome using the BWA-MEM algorithm (reference: http://bio-bwa.sourceforge.net/). Sequence variant detection was performed by the Genome Analysis Toolkit HaplotypeCaller (reference: http://www.broadinstitute.org/gatk/). The detected sequence variants (gene package prenatal, version 2, 26-2-2021 (https://www.erasmusmc.nl/genoomdiagnostiek)) were filtered and annotated with Alissa Interpret software (Agilent Technologies) on quality (read depth ≥10), minor allele frequency (≥0.1% in 200 alleles in dbSNP, ESP6500, the 1000 Genome project, GoNL or the ExAC database) and location (within an exon or first or last 10 bp of introns). Variants were further selected based on three inheritance models (de novo autosomal dominant, autosomal recessive and X-linked recessive) and classified using Alamut Visual (interactive Biosoftware, SOPHiA GENETICS) according to the American College of Medical Genetics and Genomics (ACMG) guideline for sequence variants interpretation55 and ClinGen Sequence Variant Interpretation (SVI) General Recommendation for Using ACMG Criteria (https://clinicalgenome.org/working-groups/sequence-variant-interpretation). These variant classification criteria were also applied on TGDS variants detected in the other individuals (Extended Data Table 1).
Embryonic stem cell targeting and transgenic mouse strains
G4 mES cells were maintained as previously described56. The single guide RNA (sgRNA) targeting mouse Tgds exon 4 (NM_029578.3) was designed using http://crispr.mit.edu/guides/. Complementary sgRNA oligonucleotides were subsequently annealed, phosphorylated and cloned into the BbsI site of dephosphorylated pX459 pSpCas9(BB)-2A-Puro vector57 (Addgene; #62988). For knock-in of the pathogenic mutation Tgds c.298 G > T into mouse embryonic stem cells, single-stranded oligodeoxynucleotides (ssODN) (60 pMol) were designed with asymmetric homology arms (HA) and phosphorothioate (PS) bonds as previously described58,59. Transfection of mouse embryonic stem cells and further processing was performed as previously described60. Potential structural variant and knock-in embryonic stem cell clones were first identified by PCR detection using the same genotyping primers as for the animals later and subsequently confirmed by Sanger sequencing. Primer sequences can be found in Supplementary Data Table 1. Mutant animals were produced through tetraploid or diploid aggregation61.
Mouse models
Mice were maintained by crossing with wild-type C57BL6/J mice. All mice were housed in a centrally controlled environment with a 12 h light and 12 h dark cycle, temperature of 20–22.2 °C, and humidity of 30–50%. Bedding, food and water were routinely changed. All animal procedures were conducted as approved by the local authorities (LAGeSo Berlin) under the license numbers G0247/13 and G0176/19. Ages and developmental stages of mice are indicated in the figure legends.
Skeletal preparations
Mouse embryos at stage E18.5 were collected and stained for bone and cartilage markers as follows. Embryos were eviscerated and fixed in 100% ethanol overnight. Following fixation, cartilage was stained using Alcian Blue staining solution (150 mg l−1 Alcian Blue 8GX in 80% ethanol and 20% acetic acid) overnight. Then embryos were post-fixed and washed in 100% ethanol overnight. Samples were pre-cleared with 0.2% KOH for a day and bones were stained with Alizarin Red (50 mg l−1 Alizarin Red S in 0.2% KOH) until the desired colour had developed. Rinsing and clearing was done using low concentrations (0.2%) of KOH and stained embryos were stored in 25% glycerol and imaged using a Zeiss Discovery V12 microscope and Leica DFC420 digital camera.
µCT analysis
E18.5 embryos were fixed and scanned in 70% ethanol using a SkyScan 1172 X-ray microtomography system (Bruker µCT) at 5-μm resolution. 3D model reconstruction was done with the Bruker Micro-CT image analysis software NRecon and CTVOX.
Cell culture
HAP1 cells were obtained from Horizon and cultured at 37 °C, 5% CO2 in IMDM (Biowest) supplemented with 10% FBS (Cytiva) and antibiotics (100 μg ml−1) penicillin streptomycin (Biowest). 293T and HCT116 cells were obtained from Eric Fearon (University of Michigan, MI, USA) and U2OS cell lines were obtained from Anabelle Decottignies (UCLouvain, Brussels, Belgium). 293T, HCT116, U2OS and human fibroblasts described in the present study were cultured at 37 °C, 5% CO2 in DMEM (Biowest) supplemented with 10% FBS (Cytiva), antibiotics (100 μg ml−1) penicillin streptomycin (Biowest) and 5% Ultraglutamine (Biowest). Cell lines were not further identified but were tested regularly for Mycoplasma.
Generation of plasmids
Primers to generate plasmids are listed in Supplementary Data Table 1, and plasmids are listed in Supplementary Data Table 2. We used lentiviral expression constructs based on the plasmid pLVX-PURO (Clontech). In these constructs, expression is driven by the SV40 promoter (pUB82), the CMV promoter (pUB83) and the EF1α promoter (pUB81) (details available upon request).
We amplified the TGDS open reading frame (ORF) by PCR from a sequence-verified cDNA clone (Horizon discovery MGC clone Id 5175390) using the primers hTGDS_s_NheI and hTGDS_as_Acc65I. The resulting product was digested using the restriction enzymes NheI and Acc65I and inserted in the plasmids pUB82 and pUB83 using the restriction sites XbaI and BsrGI giving rise to the lentiviral plasmid pSP19 and pSP20 respectively. Subsequently, the TGDS open reading frame was transferred from pSP19 into the vector pUB81 via the restriction sites BamHI and EcoRI, leading to the plasmid pJJ45.
To generate plasmids carrying variants, TGDS (NM_014305.4) was amplified from cDNA and subcloned into pCMV6-Entry mammalian vector with and without C-terminal Myc-DDK tag. The variants were introduced using site-directed mutagenesis with Kapa Hotstart HiFi (Roche) and In-Fusion cloning kit (Clontech, Takara).
We amplified the ArnA ORF from an the K12 E. coli strain using the ecArnA_s_BamHI and ecArnA_as_BsrGI. The resulting PCR product was digested using the restriction enzymes BamHI and BsrGI, and inserted into the corresponding sites in the vector pUB82, resulting in the plasmid pJG406.
We amplified the B. cinerea UG46DH ORF from a geneblock (IDT) optimized for E. coli codon usage using the primers bfUG46DH_s_BamHI and bfUG46DH_as_EcorI. The sequence is based on the clone XM_001554921.230. The resulting PCR product was digested and inserted into the plasmid pUB83 using the restriction enzymes BamHI and EcoRI, producing the plasmid pJJ43.
We amplified the human UXS1 ORF from a sequence-verified cDNA clone (Horizon, Id 3843312), using the primers hUXS1_s_NheI and hUXS1_as_BsrGI. The resulting PCR product was digested using the restriction enzymes NheI and BsrGI and inserted in the BsrGI and XbaI sites in the plasmid pUB81 producing the plasmid pSP54.
The bacterial expression vector for TGDS, pJJ60, was generated in several steps using the open reading frame from pSP19, yielding a final construct with a N-terminal hexahistidine-tag fused to amino acid 15 of TGDS in the plasmid pET28a (Merck), via a PCR product with the primers hTGDS_s_G15 and hTGDS_as_XhoI (details and map available upon request). The same primers were used to transfer the open reading frames containing variants observed in affected individuals from the eukaryotic into the bacterial expression vector.
To generate the bacterial expression vector for ArnA, its ORF was amplified from K12 E. coli genomic DNA using primers ecArnA_s_NdeI and ecArnA_as_SacI. The resulting PCR product was digested with NdeI and SacI and inserted into the corresponding site in the plasmid pET28a, yielding the plasmid pJG413.
To generate a bacterial expression vector for B. cinerea UG46DH, its ORF was amplified from a geneblock (see above) using the primers bfUG46DH_s_NcoI and bfUG46DH_as_XhoI, and inserted into the corresponding sites in the plasmid pET28a, yielding the plasmid pJJ40.
To generate an expression vector for H6PD, its open reading framing was amplified from a sequence-verified clone (Horizon Discovery Ltd. Clone ID MHS 6278 202808130) using primers hH6PD_s_MluI and hH6PD_rev_EcoRI, and inserted into the corresponding sites of the plasmid pUB81 yielding the plasmid pIG42562. For transient transfection in mammalian cells, H6PD was amplified by PCR adding a V5 tag using Kapa Hotstart HiFi (Roche) and subcloned into pCMV6-Entry vector using In-Fusion cloning kit (Clontech, Takara).
CRISPR–Cas9 constructs to generate knockout cell lines were produced by ligating annealed oligonucleotides into the BbsI site of the vector pX459 or pX458 (Addgene #48138)63, as previously described (see Supplementary Data Table 1)57,63, or into the vector pLenticrispr V2.0 (Addgene #52961)64.
Generation of knockout cell lines
To generate knockout cell lines, cells were transfected in 6-well plates at 70% confluence with 2 µg of the CRISPR–Cas9 guide RNA expression plasmids and 4 µl Lipofectamine 2000 following the manufacturer’s instructions (Life Technologies). Transfected cells were either transiently selected with puromycin (for pX459 constructs) or selected by flow cytometry sorting gating for GFP fluorescence on a FACSAria III flow cytometer (for pX458 constructs). Clonal populations were grown out and analysed by Sanger sequencing. To generate polyclonal knockout cell lines, recombinant lentiviruses driving expression of both CRISPR–Cas9 and specific guide RNAs were generated as described below.
Lentiviral transduction
To produce recombinant lentiviruses, we transiently transfected 293T cells with lentiviral vectors and second generation packaging plasmids psPAX2 and pMD2.G (kind gifts from D. Trono, Addgene #12260 and #12259) using calcium phosphate precipitation65. The culture medium was changed 6 h after transfection, and recombinant viruses were recovered in the culture supernatant 24 h later. The virus-containing supernatant was then incubated with target cells in the presence of 4 μg ml−1 polybrene (Sigma). Infected cells were selected 24 h later for 4 days with puromycin (Thermofisher).
Preparation of metabolite extracts for LC–MS analysis
For metabolomic analyses, non-fibroblast cell lines were plated in 6-well plates at 350,000 cells per well (500,000 for HAP1 cells). Lysates were obtained after quenching of metabolism as described before66. In brief, after one rapid wash with ice-cold water, culture plates were placed on liquid nitrogen for 5 s and then transferred onto dry ice. Cells were scraped after addition of 250 μl of a solution consisting of 90% methanol (Biosolve) and 10% chloroform, and lysates were transferred into microcentrifuge tubes. After centrifugation for 10 min at 4 °C and 27,000g, the supernatant was recovered, dried in a SpeedVac vacuum concentrating system (Life technologies) and resuspended in 30 μl of water before a final centrifugation of 10 min at 4 °C and 27,000g. Twenty microlitres of supernatant were transferred into LC–MS vials (Verex Vial, 9 mm Screw from Phenomenex) before analysis.
Fibroblasts were cultured in 10 cm dishes until 90% confluence. Cells were washed twice with 10 ml of ice-cold water per plate, and recovered using a cell scraper after addition of 500 μl of 0.5 M perchloric acid. Two plates were pooled into one tube followed by centrifugation at 27,000g at 4 °C for 10 min. The supernatant was recovered and transferred into a microcentrifuge tube. 100 µl of potassium carbonate was added to neutralize the solution. Samples were then spun down at 27,000g, 4 °C for 5 min and the supernatant was purified via solid phase extraction (SPE).
Mice were sacrificed at 8 months of age. Brain, heart, liver and kidney were obtained and immediately frozen in liquid nitrogen. Samples were pulverized with a mortar and pestle in liquid nitrogen. Approximately 50 mg of tissue powder was homogenized in a refrigerated Precellys bead beater (Bertin instruments) 6 times for 10 s at 10,000 oscillations per min in reinforced 2 ml tubes containing 500 µl of 0.5 M perchloric acid as well as seven 1.4 mm ceramic beads (Omni). After homogenization and centrifugation at 27,000g at 4 °C for 10 min, the supernatant was recovered and transferred into a microcentrifuge tube. Forty microlitres of potassium carbonate was added to neutralize the pH of the solution. The samples were then spun down at 27,000g, 4 °C for 5 min, and then purified via SPE.
The supernatants from neutralized perchloric acid extracts were loaded onto SPE columns (250 mg Supelclean ENVI-Carb SPE Tube from Supelco) pre-equilibrated by successive addition of 600 µl of 60% acetonitrile, 400 µl of 0.3% formic acid, and 3 ml of water. After sample application, the column was washed with 3 ml of water and 3 ml of acetonitrile. At all steps, the liquid was aspirated through the column via a Vac-Man Laboratory Vacuum Manifold from Promega. Metabolites were eluted into a microcentrifuge tube by addition of 1 ml 60% acetonitrile/0.3% formic acid. The eluate was dried in a SpeedVac vacuum concentrating system (Life Technologies) and resuspended in 25 μl of water before a final centrifugation of 10 min at 4 °C and 27,000g. Twenty microlitres of supernatant were transferred into LC–MS vials for analysis.
LC–MS analysis
Analyses by LC–MS were performed as previously described67 based on a previously described method68. In brief, 5 μl of sample were analysed with an Inertsil 3 μm particle ODS-4 column (150 ×2.1 mm; GL Biosciences) at a constant flow rate of 0.2 ml min−1 with an Agilent 1290 HPLC system. Mobile phase A consisted of 5 mM hexylamine (Sigma-Aldrich) adjusted to pH 6.3 with acetic acid (Biosolve BV) and phase B of 90% methanol (Biosolve BV)/10% 10 mM ammonium acetate (Biosolve BV). The mobile phase profile consisted of the following steps and linear gradients: 0–2 min at 0% B; 2–6 min from 0 to 20% B; 6–17 min from 20 to 31% B; 17–36 min from 31 to 60% B; 36–41 min from 60 to 100% B; 41–51 min at 100% B; 51–53 min from 100 to 0% B; 53–60 min at 0% B. For analysis of metabolite extracts obtained from fibroblasts and organs, MS acquisition was stopped between minutes 9 and 10 to avoid the peak resulting from the perchloric acid extraction.
When exploring the effect of H6PD inactivation on UDP-4-keto-6-deoxyglucose levels in 293T and HAP1 cell lines, we had to ensure that small increases were not masked by the background noise. For this reason, 4 million cells were plated in 10 cm dishes and metabolites were collected two days later via the protocol described for fibroblasts.
For the analysis of enzymatic reactions, a shorter gradient was used where the mobile phase profile consisted of the following steps and linear gradients: 0–2 min at 0% B; 2–6 min from 0 to 24% B; 6–13 min from 24 to 31% B; 13–21 min from 31 to 50% B; 21–22 min from 50 to 100% B; 22–29 min at 100% B; 29–30 min from 100 to 0% B; 30–37 min at 0% B.
Analytes were identified with an Agilent 6550 ion funnel mass spectrometer operated in negative mode with an electrospray ionization (ESI) source and the following settings: ESI spray voltage 3500 V, sheath gas 350 °C at 11 l/min, nebulizer pressure 35 psig and drying gas 200 °C at 14 l min−1. An m/z range from 70 to 1,200 was acquired with a frequency of 1 per second by adding 8,122 transients. Compound identification was based on their exact mass (<5 ppm) and retention time compared to standards, obtained from Sigma-Aldrich (UDP-GlcNAc U4375, UDP-GlcA U6751, UDP-Glc U4625, CMP sialic acid 233264, GDP-mannose G5131) or Carbosource Services at the Complex Carbohydrate Research Center of the University of Georgia (UDP-Xyl, UDP-Ara). UDP-4k6dG and UDP-6dH were synthesized by ArnA or UG46DH as described below. CDP-ribitol was synthesized as described before69. The area under the curve of extracted-ion chromatograms of the [M-H]- forms were integrated using MassHunter software (Agilent), and normalized to the mean of the areas obtained for a series of 150 other metabolites (total ion current).
Protein purification
Expression plasmids for human UXS1 were transformed into the BL21 Rosetta strain (Merck) using electroporation. A pool of >10 colonies was used to inoculate a 5 ml culture in Lysogeny broth (LB) containing 30 μg ml−1 kanamycin, and incubated overnight at 37 °C, shaking at 200 rpm. The preculture was added to 500 ml LB. Once this culture reached an optical density of 0.5 at 600 nm, we added 1 mM isopropyl-β-d-thiogalactoside (IPTG), and the culture was incubated overnight at 20 °C. Bacteria were collected by a 20 min centrifugation at 6,000g and 4 °C. Bacterial pellets were stored at −20 °C until purification. Cell pellets were resuspended in 25 ml of lysis buffer containing 50 mM HEPES pH 7.5, 500 mM NaCl, 5 mM imidazole, 5% glycerol and protease inhibitors (p-toluenesulfonyl fluoride (TSF) at 1 mM, leupeptin at 1 µg ml−1, and antipain at 1 µg ml−1) and lysed using a French Press (Glen Mills). The lysate was then centrifuged for 20 min at 20,000g and 4 °C. The supernatant was incubated with 1 ml Ni-NTA beads (Cytiva) for 10 min in a 15 ml tube at 4 °C on a rotative device. Beads were collected by centrifugation for 10 min at 400g and 4 °C, and resuspended in 3 ml of lysis buffer. The slurry was added in a 2 ml disposable Column (Pierce) and washed with 15 ml of lysis buffer supplemented with 30 mM imidazole, followed by 5 ml of lysis buffer supplemented with 100 mM of imidazole. Protein was eluted using 5 ml of lysis buffer containing 250 mM imidazole. Fractions containing the protein of interest were pooled, buffer-exchanged using a G25 Sepharose column (Cytiva PD-10) following the manufacturer’s protocol with a buffer containing 20 mM triethanolamine, 250 mM NaCl and 1 mM DTT, and stored at −80 °C. We obtained approximately 5 mg of protein per liter of culture.
Production of human TGDS and its mutants was performed using a similar protocol as for UXS1 with the following differences: After induction with IPTG, cultures were incubated overnight at 20 °C. We used a lysis buffer containing HEPES pH 7.4, 300 mM KCl, 1 mg ml−1 lysozyme (Roche), 10 mM imidazole, 3 mM β-mercaptoethanol, 10% glycerol, protease inhibitors (TSF at 2 mM, leupeptin at 1 µg ml−1, and antipain at 1 µg ml−1), and 5 µM NAD+, and lysed by French Press. The lysate was clarified by centrifugation for 20 min at 20,000g and 4 °C. We used 1 ml of a 50% HisPur Ni-NTA matrix slurry for purification. Washes were performed with five times 2 ml of lysis buffer and the protein was eluted by resuspension in 2× 500 µl of lysis buffer containing 250 mM of imidazole, and 2× 500 µl of lysis buffer containing 500 mM imidazole.
Comparable results were obtained when protein purification was performed by using a liquid chromatography system (Akta, Cytiva) using a 1 ml HisTrap HP Ni-NTA column with a flow rate of 1 ml min−1. The column was equilibrated in buffer A (25 mM Hepes 7.4, 10% glycerol, 300 mM KCl, 3 mM β-mercaptoethanol, 10 mM imidazole and 5 µM NAD+). After sample application, the column was washed with 20 column volumes of 94% buffer A and 6% buffer B (50 mM Hepes 7.4, 10% glycerol, 200 mM KCl, 5 mM β-mercaptoethanol, 500 mM imidazole, 0.2% sodium dodecyl maltoside, 1 mM TCEP) followed by a gradient to 100% over the course of 17 column volumes. We obtained approximately 0.6 mg of purified protein per liter of culture.
Production of ArnA was performed as described for human UXS1 with the following differences: the lysis buffer consisted of 100 mM HEPES pH 7.5, 150 mM KCl, 1 mg ml−1 lysozyme (Roche), 5 mM MgCl2, 5 mM β-mercaptoethanol, 10% glycerol and 1 mM of the protease inhibitor phenylmethylsulfonyl fluoride (PMSF). Lysis was performed by a freeze-thaw cycle in liquid nitrogen. After incubation for 15 min on ice, the concentration of KCl was brought to 500 mM and the lysate was sonicated. The lysate was clarified by centrifugation for 20 min at 20,000g and 4 °C, followed by purification with a liquid chromatography system (Akta, Cytiva) using a 1 ml HisTrap HP Ni-NTA column with a flow rate of 1 ml min−1. The column was equilibrated in buffer A (100 mM Hepes 7.4, 10% glycerol, 500 mM KCl, 5 mM MgCl2, 5 mM β-mercaptoethanol). After sample application, the column was washed with 20 column volumes of 94% buffer A and 6% buffer B (50 mM Hepes 7.4, 10% glycerol, 200 mM KCl, 5 mM β-mercaptoethanol, 300 mM imidazole) followed by a gradient to 100% over the course of 20 column volumes. We obtained more than 50 mg of purified protein per liter of culture.
Purification of B. cinerea UG46DH was performed as described for UXS1 with the following differences. Expression was induced overnight at 20 °C. The lysis buffer consisted of 100 mM Tris-HCl 7.4, 150 mM NaCl, 1 mg ml−1 lysozyme (Roche), 1 mM EDTA, 10% glycerol. Lysis was performed by sonication and the clarified lysate was purified using a liquid chromatography system with a 1 ml HisTrap HP Ni-NTA column. The column was equilibrated in buffer A (50 mM sodium phosphate buffer pH 8, 300 mM NaCl, 5 mM β mercaptoethanol). After sample application, the column was washed with 20 column volumes of 96% buffer A with 4% buffer B (50 mM sodium phosphate pH 8, 300 mM NaCl, 5 mM β-mercaptoethanol, 300 mM imidazole). Protein was eluted in a gradient to 100% buffer B over 20 column volumes. Positive fractions were pooled and buffer-exchanged into 50 mM Tris pH 8, 150 mM NaCl, 10% glycerol, 1 mM DTT, and 10 µM NAD+. We obtained approximately 2.5 mg of purified protein per liter of culture.
Experimental setup for enzymatic assays
To produce UDP-4-keto-6-deoxyglucose via B. cinerea UG46DH, we used 20 µl reactions containing 25 mM triethanolamine, 1 mM UDP-glucose, 1 mM MgCl2, 10 mM DTT, and 2 µM of enzyme. Reactions were incubated for 1 h at 30 °C, heated to 100 °C for 2 min and extracted with 40 µl of CHCl3, followed by centrifugation at 27,700g for 10 min at 4 °C.
To produce UDP-4-keto-6-deoxyglucose via TGDS, a reaction of 500 µl containing 25 mM triethanolamine, 1 mM UDP-glucose, 0.1% BSA, 1 mM MgCl2 and 0.53 µM purified enzyme was incubated for 23 h at 30 °C, followed by 5 min at 85 °C and a centrifugation at 27,700g for 10 min at 4 °C.
The activity of UXS1 was assessed in 20 µl reactions containing 25 mM triethanolamine, 1 mM UDP-glucuronate, 10 mM DTT, 0.1% BSA, and 2 µM purified enzyme. Reactions were incubated for 1 h at 30 °C, followed by deproteinization by addition of 20 µl of methanol and 40 µl of chloroform. The aqueous phase was collected and analysed by LC–MS.
Production of UDP-4-ketoxylose using ArnA was performed in 500 µl reactions containing 25 mM triethanolamine, 1 mM UDP-glucuronate, 10 mM DTT, 0.1% BSA, 20 µM NAD+, 5 mM sodium pyruvate, and 5.5 U ml−1 rabbit muscle LDH (Sigma) to reoxidize NADH to NAD+, and 21.8 µM of ArnA. Reactions were incubated for 6 h at 30 °C followed by a centrifugation using a centrifugal ultrafiltration device (Vivaspin 500, Sartorius) using the manufacturer’s protocol to obtain the reaction product in the flow through.
To inactivate recombinant UXS1 with sodium borohydride, the enzyme was incubated at 2 µM for 45 min on ice in the presence of 5 mM NaBH4 in a reaction containing 25 mM triethanolamine, 1 mM DTT, and 0.1% BSA. 1% acetone was added on ice for another 30 min to destroy borohydride. To assess activity, the enzyme was incubated with 1 mM UDP-glucuronate for 1 h at 30 °C, followed by extraction with 3 volumes of chloroform:methanol 2:1, recovery of the aqueaous phase, and analysis by LC–MS.
In reactivation experiments, we incubated 0.56 µM inactivated recombinant UXS1 with 0.875 µM of UDP-4-keto-6-deoxyglucose produced by TGDS, containing residual UDP-glucose. This corresponds to the estimated concentration in the endoplasmic reticulum and Golgi apparatus and represents approximately a 1.5-fold excess relative to the inactivated UXS1 enzyme. TGDS reactions contained residual UDP-glucose. To exclude a confounding effect of UDP-glucose, control reactions therefore contained 17.5 µM UDP-glucose.
In reactivation experiment using ArnAs product, we used 1.25 µM UDP-4-ketoxylose, corresponding approximately to a 2-fold excess of this metabolite relative to the inactivated UXS1 enzyme.
Given the higher levels of UDP-4-keto-6-deoxyglucose observed in UG46DH overexpressing cells, we used a 100-fold excess of the UG46DH product relative to UXS1. To exclude a confounding effect of residual UDP-glucose from the UG46DH reaction, the control reaction contained 500 µM UDP-glucose.
Laminin overlay
Laminin overlay assays were performed as previously described69. In brief, cells were lysed in PBS containing 1% Triton X-100 and centrifuged at 27,700g for 10 min at 4 °C. Supernatants were incubated under gentle rotation for 16 h at 4 °C with WGA-Agarose beads (Vector Laboratories) using 100 μl of beads per 1400 μg of HAP1 cells protein with PBS 1 containing 0.1% Triton X-100. Beads were washed 3 times with 1 ml of PBS containing 0.1% Triton X-100, proteins were released by incubation for 10 min at 72 °C in the presence of reducing sample buffer and resolved on 3-8% Tris-acetate gels (Life Technologies) for 75 min at 130 V. After overnight transfer at 30 mA onto a PVDF membrane, membranes were blocked with laminin-binding buffer (LBB; 50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM MgCl2 and 1 mM CaCl2) containing 3% BSA (Sigma) for 1 h at room temperature and incubated overnight with 1.15 μg ml−1 Laminin-111 (Sigma, L2020) in LBB. After three 10 min washes with LBB, membranes were incubated for 2 h at room temperature with rabbit anti-laminin antibody (Sigma, L9393) diluted 1:1000 in LBB containing 3% BSA, washed another three times for 10 min with LBB, incubated for 1 h at room temperature with horseradish peroxidase-coupled donkey anti-rabbit IgG antibody (GE healthcare, NA934V) diluted 1:15,000 in LBB containing 3% BSA. The signal for β-dystroglycan obtained with mouse anti-β-dystroglycan antibody (clone 7D11, 33701, Santa-Cruz, 1:1,000) in the same membrane was used to normalize for overall abundance of dystroglycan. Chemiluminescent signals were detected using a Cytiva Amersham ImageQuant 800 western blot imaging systems. Uncropped images are shown in Supplementary Data Fig. 1.
Western blot analysis
U2OS cells were seeded at 5 × 105 cells per well in 6-well format, followed by transfection of 1 µg of plasmid using Lipofectamine 3000 (Invitrogen) and Opti-MEM (Gibco) according to the manufacturer’s instructions. After 48 h proteins were extracted in RIPA buffer (150 mM NaCl, 50 mM Tris, 5 mM EDTA, 1% Triton X-100, 0.25% desoxycholate, 5% SDS) containing protease inhibitors (cOmplete; Roche) and phosphatase inhibitors (PhosSTOP; Roche). Total protein concentration was determined using Pierce BCA assay (Thermo Scientific). Twenty micrograms of protein per lane was separated by SDS–PAGE (12%), transferred to nitrocellulose membrane and probed with primary antibodies. Immunoblot staining was performed for TGDS (rabbit anti-TGDS; HPA040857, Atlas, 1:1,000) and GAPDH (mouse anti-GAPDH; AM4300, ThermoFisher, 1:2,000). Membranes were incubated with IRDye anti-rabbit 800CW and anti-mouse 680RD secondary antibodies (926-32211 and 926-68070, Li-Cor Biosciences, 1:10,000). Signals were detected with OdysseyFc Imaging System and quantification was performed using Image Studio (Li-Cor Biosciences). The signal of TGDS was normalized to its GAPDH signal, and then to the wild type within each experiment. Data are presented as mean ± s.d. obtained in three independent experiments. *P < 0.05 in post hoc testing after one-way ANOVA corrected according to Dunnett.
Western blots for H6PD and β-actin were performed as described above, but using PVDF membrane (Immobilon P, Milipore), horseradish peroxidase-coupled secondary antibodies, chemiluminescence peroxidase substrate (Immobilon Western Blot reagent, Milipore) and a Cytiva IQ600 system. Mouse anti-H6PD (TA501257, Origene, 1:1,000) and anti-β-actin (A5441, Sigma, 1:5,000) were used as primary antibodies.
Uncropped images are shown in Supplementary Data Fig. 1.
Immunofluorescence
U2OS cells were grown on glass coverslips overnight (1.5 × 105 cells per well), followed by transfection of 1 µg of plasmid using Lipofectamine 3000 (Invitrogen) and Opti-MEM (Gibco) according to the manufacturer’s instructions. After 48 h, cells were fixed in cold methanol for 10 min at 4 °C, or for 10 min at room temperature with 4% paraformaldehyde in 1× PBS followed by permeabilization with 0,4% Triton X-100 in 1× PBS for 15 min at room temperature for the staining of TGDS and GM130. Immunofluorescence staining was performed overnight in PBS containing 3% BSA at 4 °C using the following antibodies: TGDS (rabbit anti-TGDS; HPA040857, Atlas, 1:500); PDIA1 (mouse anti-PDIA1 (P4HB); ab2792, Abcam, 1:200); GM130 (mouse anti-GM130; 610823, BD Transduction Laboratories, 1:500); FLAG (mouse anti-FLAG M2; F1804, Sigma, 1:500); CALR (rabbit anti-Calreticulin; ab92516, Abcam, 1:500); V5 (rabbit anti-V5; V8137, Sigma, 1:400); H6PD (mouse anti-H6PD; TA501257, OriGene, 1:100) and GIANTIN (rabbit anti-Giantin; 621352, BioLegend, 1:1,000). Secondary antibody staining was performed using anti-mouse IgG Alexa Fluor 555 (A21422, Invitrogen, 1:1000) and anti-rabbit IgG Alexa Fluor 488 (A21206, Invitrogen, 1:1,000) for 1 h in 1× PBS at room temperature. Coverslips were mounted in Fluoromount G (Invitrogen). Images were taken using either LSM700 or LSM980 with Airyscan 2 (Zeiss).
Flow cytometry
The presence of heparan sulfate on the surface of cells was assessed by flow cytometry using the Heparan sulfate antibody HS10E4 antibody (H1890 USBIOlogical). Cells were washed once with PBS and incubated for 5 min with 1 ml of non-enzymatic cell dissociation solution (Sigma) at 37 °C until detached. Cells were recovered by sequential addition of 2 ml and 1 ml of PBS + 2% FCS, and washed twice with 1 ml of PBS, spinning at 500g for 3 min in between. Cells were then washed twice with 2 ml of PBS + 2% FCS, centrifuging at 1,000g for 3 min in between. For staining, cells were resuspended in 100 μl PBS + 2% FCS containing 1:200 anti heparan sulfate antibody (clone 10E4, H1890 USBIOlogical), and incubated for 60 min on ice. After washes with PBS containing 2% FCS, the cells were stained with a 1:50 dilution of Alexa Fluor 647 (AffiniPure Goat Anti-Mouse IgM, µ chain specific from Jacksonimmuno) in PBS containing 2% FCS for 40 min on ice. Cells were again washed twice and analysed using a Fortessa or FACSverse flow cytometer (BD biosciences). The same protocol was followed for the detection of α-dystroglycan (mouse IgM clone IIH6C4, 05-593, Sigma-Aldrich) with 1% BSA and 0.1% sodium azide in PBS.
Statistical analyses
Investigators were blinded with regard to the genotype of the mice being analysed. No specific randomization was performed but each mouse had a certain probability to be wild-type, heterozygote or KO/KI. No blinding or randomization was performed for in vitro experiments. Mouse sex was not taken into consideration since Catel–Manzke syndrome affects both female and make individuals. Samples sizes for in vivo experiments were based on published studies with similar experimental design and phenotypic analyses. No specific sample size calculation was performed for in vitro experiments. Statistical analyses were performed in Prism 10 (GraphPad) and were two-tailed. Pairwise comparisons were performed using multiple t-tests correcting for multiple testing according to Holm–Sidak70. When more than two conditions were compared we performed a one-way ANOVA followed by post hoc testing corrected for multiple testing according to Dunnett (for comparisons with one control)71 and Sidak (for selected comparisons)70. When one condition lacked detectable signals, no statistical comparisons were performed with this condition, and this condition was not included in the ANOVA (Fig. 3h and Extended Data Figs. 2e,f, 4i–l and 5a–c). When indicated, testing was performed on log-transformed data, leading to comparable variances between sample groups. When results from several independent experiments were performed, the analyses were performed on the means obtained within each experiment (indicated as symbols with strong contrast), but values from individual experiments are presented in partially transparent symbols72. In this setting, paired tests were used. For craniofacial and finger measurements in animals a comparison between the two groups was performed using two-tailed unpaired t-test with Welch’s correction followed by correction for multiple testing according to Holm–Sidak. Figures were generated in GraphPad Prism and Adobe Illustrator, and illustrations were drawn with BioRender (https://www.biorender.com).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
