Abridged pedigrees of four HFS families are shown, numbers of individuals correspond to Table 1. Symbol notations: squares, males; circles, females; open symbols, unaffected individuals; filled symbols, affected mutation carrier of the nucleotide and predicted amino acid change indicated on top; index patients are marked with a red arrow. Electropherograms of index patients are shown to the right of each pedigree, altered nucleotides are marked in pink in both reference sequence (top) and actual read (bottom); underlined nucleotides indicate deletions or the site of insertions in exon 13. Inserted nucleotides in patients' sequences are also underlined, predicted changes in the amino acid sequence given above. All mutations segregated with the disease phenotype, showing compound homozygosity or homozygosity in patients, heterozygosity in parents, and wild-type alleles or heterozygous mutations in unaffected siblings.

Basic representation of the CMG2 protein and its domains, highlighting the positions of the seven cysteines and HFS mutations analysed in this study.Structural model of the extracellular domains of CMG2. The vWA domain has been taken from the crystal structure 1tzn, chain ‘a’, and the predicted Ig-like model (this work) has been grafted onto the crystal structure of the vWA domain (1tzn, chain a). To better show the overall fold, the successive secondary structure elements each have a distinct colour from dark-blue (N-terminal) to red (C-terminal). Cysteines participating in predicted disulphide bridges are marked as black spheres.

A,C-E. CHO cells were transfected for 48 h with WT or mutants CMG2-V5 constructs. Cell lysates were subjected to immunoprecipitation with an anti-V5 antibody and analysed by SDS–PAGE—under reducing or non-reducing conditions—and Western blotting with an anti-V5 antibody.B. HeLa cells were co-transfected for 48 h with mutant CMG2-V5 or CMG2-HA constructs. Cell lysates were subjected to immunoprecipitation with an anti-HA antibody and analysed by SDS–PAGE under non-reducing (WB: V5) and reducing (WB: HA) conditions and Western blotting with anti-V5 or anti-HA antibodies.

A,D. CHO cells were transfected for 48 h with WT or mutants CMG2-V5 constructs. Cell lysates were subjected to immunoprecipitation with an anti-V5 antibody. Samples were analysed by SDS–PAGE—under reducing (A) or non-reducing conditions (D)—and Western blotting with an anti-V5 antibody.B. CHO cell lysates were subjected to immunoprecipitation with an anti-V5 antibody, treated with or without EndoH and analysed by SDS–PAGE and Western blotting with an anti-V5 antibody.C. HeLa cells were transfected with mutant CMG2-V5 cDNAs for 24 h prior to fixation and permeabilization, and labelling with anti-V5 antibodies.E. CHO cells were transiently transfected with WT or mutant CMG2-V5 cDNA. Cells were lysed in immunoprecipitation buffer and 1 µg/ml of PA was added in the buffer for 1 h at 4°C. After immunoprecipitation using an antibody against the V5 tag, samples were analysed by SDS–PAGE followed by Western blotting against PA or the V5 tag.F. HeLa cells were transfected for 48 h with WT or mutant CMG2-V5 cDNA. Surface biotinylation was performed as described in 4B. Error bars represent the standard deviation (n = 3). Asterisks represent significant difference with respect to WT (p < 0.05).

CMG2 was immunoprecipitated from 300 µg of cell lysates from patient-derived fibroblasts and analysed by SDS–PAGE under non-reducing conditions and Western blotting using the 2F6 anti-hCMG2 antibody.mRNA levels of cmg2 in patient fibroblasts were analysed by quantitative real-time RT-PCR. Patient 5 has been previously characterized and harbours the I189T mutation in the vWA domain and a frame-shift mutation in exon 13 (Table 2; Deuquet et al, 2009; Dowling et al, 2003). The cmg2 mRNA levels in this patient was 54 ± 15% of the control (n = 5). Levels and number of experiments for other patients are mentioned in the main text. mRNA levels of the unrelated gene RhoA are shown for comparison. Errors bars represent standard deviations. Paired t-test analysis between control cells and patient-derived cells was performed showing that the observed differences are statistically significant (*p < 0.05).Patient-derived fibroblasts were incubated with 10 µM MG132 or not for 16 h. CMG2 was immunoprecipitated from 300 µg of cell lysates and analysed by SDS–PAGE under non-reducing conditions and Western blotting using the 2F6 anti-hCMG2 antibody.

A,C. HeLa or CHO cells were transfected for 48 h with WT or mutant CMG2-V5 constructs. Cell extracts were analysed by SDS–PAGE and Western blotting with an anti-V5 antibody.B. HeLa cells were transfected for 48 h with WT or mutant CMG2-V5 cDNAs. Confluent cells were pulsed with [³⁵S]-methionine and subsequently incubated with 0.2 mg/ml NHS-SS-biotin. Biotinylated CMG2 proteins and expression levels were quantified using the Typhoon scanner. For each experiment, biotinylated CMG2 values were normalized to the synthesis levels and expressed as a percentage of WT. Error bars represent the standard deviation (n = 3).D. CHO cell lysates were submitted to immunoprecipitation with an anti-V5 antibody, treated with or without EndoH and analysed by SDS–PAGE and Western blotting with an anti-V5 antibody.E. HeLa cells were transfected with WT or mutant CMG2-V5 cDNA for 24 h prior to fixation and permeabilization and labelling with anti-V5 monoclonal as well as anti-calnexin polyclonal antibodies. The inserts illustrate an enlargement of a specific region. Bar: 10 µm.

A,D. CHO cells were transfected for 48 h with WT or mutants CMG2-V5 constructs. Cell lysates were subjected to immunoprecipitation with an anti-V5 antibody. Samples were analysed by SDS–PAGE—under reducing or non-reducing conditions—and Western blotting with an anti-V5 antibody.B. CHO cell lysates were subjected to immunoprecipitation with an anti-V5 antibody, treated with or without EndoH enzyme and analysed by SDS–PAGE and Western blotting with an anti-V5 antibody.C. HeLa cells were transfected with WT or mutant CMG2-V5 cDNA for 24 h prior to fixation and permeabilization, and labelling with anti-V5 monoclonal and anti-calnexin polyclonal antibodies. Bar: 10 µm.E. CHO cells were transiently transfected with WT or mutant CMG2-V5 cDNA. Cells were lysed in immunoprecipitation buffer and 1 µg/ml of PA was added in the buffer for 1 h at 4°C. After immunoprecipitation using an antibody against the V5 tag, samples were analysed by SDS–PAGE followed by Western blotting against PA or the V5 tag.F. HeLa cells were transfected for 48 h with WT or mutant CMG2-V5 cDNA. Surface biotinylation was performed as in 4B. Error bars represent the standard deviation (n = 3). Asterisks represent significant difference with respect to WT (p < 0.05).

Patient-derived fibroblasts were incubated with 10 µM of BZ or not for 16 h. CMG2 was immunoprecipitated using the goat anti-human CMG2 antibody from 300 µg of cell lysates and analysed by SDS–PAGE under non-reducing conditions and Western blotting using the 2F6 anti-hCMG2 antibody.Fibroblasts derived from Patients 1, 2 and 5 and control cells were incubated with 10 µM BZ or not for 16 h and subsequently incubated with 0.2 mg/ml NHS-SS-biotin. Cell lysates (300 µg of total proteins) were subjected to immunoprecipitation with streptavidin–agarose beads and analysed by SDS–PAGE and Western blotting probed with anti-hCMG2, anti-calnexin and anti-hTransferin receptor antibodies. * indicates unspecific band recognized by the 2F6 antibody after surface biotinylation and immunoprecipitation. This band is also observed for Patient 4 fibroblasts that express undetectable levels of CMG2. While calnexin (Cnx) was absent in these immunoprecipitations, it was readily detected in cell extracts (Supplementary Fig 4E). Under non-reducing conditions, human transferrin receptor (hTfR) migrates as a covalent dimer.Fibroblasts derived from Patient 1 and control cells were incubated with 10 µM BZ for 16 h and subsequently incubated with anthrax PA at a 500 ng/mL final concentration. After different time points, cell lysates were immunoprecipitated with an anti-hCMG2 antibody. Samples were analysed by SDS–PAGE under reducing or non-reducing conditions and Western blotting with an anti-hCMG2, an anti-ubiquitin, and anti-PA antibodies.