Fig 1.
Hair phenotype and identification of HEPHL1 mutations.
(A) Patient photographs taken at age 5 years show hair abnormalities. Sparse distribution of hair in the temporal area (left), and coarse hair texture (middle) are shown. Light microscopic examination of hairs showing pili torti (right upper) and trichorrhexis nodosa (right lower). (B) Sanger sequencing chromatograms showing biallelic mutations in the proband. The proband is a compound heterozygous for NM_001098672 (HEPHL1): c.1063G>A; p. Ala355Thr (inherited from the mother, left) and NM_001098672 (HEPHL1): c.3176T>C; p. Met1059Thr (inherited from the father, right).
Fig 2.
Molecular analyses of HEPHL1 mutations.
(A) Analysis of cDNA from the control and proband. An agarose gel image of the PCR products shows both a correctly spliced and a truncated transcript in the proband. Lane 1 is a 1Kb molecular weight DNA ladder. DNA sequencing of the PCR products confirmed both a correctly spliced (exon 4–5) and a truncated (exon 4–6) transcript in the proband. Skipping of exon 5 in the proband leads to an in-frame deletion of 255 nucleotides. (B) Chromatograms showing presence of c.3176T>C variant in cDNA from the proband. (C) Multiple sequence alignment of HEPHL1 confirms conservation of Met 1059 across different species. (D) Stereo view of portions of domains 2 and 6 of the HEPHL1 model. Met 1059 is shown as space-filling, as are the bound copper ions (purple). Each of these ion binding sites involves the side chains of two histidines, one methionine, and one cysteine. The HEPHL1 main chain is colored by sequence identity to the ceruloplasmin template structure, yellow where identical (52% overall) and light blue where different. Additional side chains that bind the copper ions or surround the “labile” and “holding” sites are shown as bonds, with atoms colored conventionally: carbon, gray; oxygen, red; nitrogen, blue; and sulfur, yellow. (E) Quantitative real time PCR analysis shows no significant difference in mRNA expression between control and patient derived iPSc (values represent mean±SD, n = 6). GAPDH was used as a housekeeping gene to normalize the expression of HEPHL1.
Fig 3.
HEPHL1 mutations disrupt copper-dependent ferroxidase activity and glycosylation.
(A) Upper panel, immunoblot analysis of whole cell extracts of HEK293 cells overexpressing WT-HEPHL1, HEPHL1 M1059T, HEPHL1 Δexon 5 with anti-DDK antibody. Lower panel shows immunoblotting of whole cell extracts with anti-vinculin antibody to confirm equivalent sample loading. (B) In-gel ferroxidase activity. Whole cell extracts of HEK293 cells overexpressing WT-HEPHL1, HEPHL1 M1059T, HEPHL1 Δexon 5 were prepared under non-denaturing conditions and separated by native gel electrophoresis. After incubation of the gel in ferrous ammonium sulfate solution for 2 h, color was developed with ferrozine. Purified human CP (3 μg) was used as a positive control. The lower panel shows Coomassie blue staining of the gel to confirm equal loading of cellular extracts in lanes 1–4. (C) In vitro ferroxidase activity. Anti-DDK beads were used to immunoprecipitate HEPHL1 from whole cell extracts of untreated or cells treated with 200 μM BCS for 2 days. Beads containing immunoprecipitated proteins were incubated with ferrous ammonium sulfate solution followed by incubation in ferrozine solution. Fe (II) forms a complex with ferrozine that can be detected by absorbance at 550 nm. The strong ferroxidase activity of WT-HEPHL1 converted most the Fe (II) to Fe (III), leading to significant reduction in absorbance, while M1059T and Δexon 5 did not show any activity (values represent mean±SD, n = 6). Treatment of cells with 200 μM BCS abrogated ferroxidase activity of WT-HEPHL1, and again no activity was observed for M1059T and Δexon 5 (values represent mean±SD, n = 3). Whole cell extracts used for IP were also immunoblotted with anti-DDK antibody to confirm the expression of HEPHL1 (upper blot). The lower blot shows immunoblotting with anti-vinculin antibody to confirm equivalent sample loading. (D) HEPHL1 is glycosylated. Whole cell extracts of HEK293 cells overexpressing WT-HEPHL1 or HEPHL1 M1059T were immunoprecipitated using anti-DDK beads. Beads were treated with deglycosylation mix for the indicated time points and processed for immunoblotting with anti-DDK antibody. The deglycosylation treatment removed the upper band. (E) Copper is required for HEPHL1 post-translational modification. HEK293 cells were transfected with WT-HEPHL1 and treated with the copper chelators ATTM or BCS. Immunoblotting of lysates with anti-DDK antibody shows that the high molecular weight form of HEPHL1 was diminished when copper was chelated.
Fig 4.
Functional consequences of the loss of HEPHL1.
(A) Increased total intracellular iron content in the patient’s fibroblasts compared to control. The control value represents the mean of data from two independent normal cell lines (error bars indicate ± SD, n = 8, * p<0.05). (B) Upper panels showing a representative anti-ferritin immunoblot of whole cell extracts of fibroblasts either untreated or treated with either 100 μM FeCl3 or 100 μM DFO (top). Anti-ß actin immunoblot shows equivalent loading of lysates (bottom). Lower panel shows fold changes in expression, calculated by quantitative analysis of bands using image studio software (Li-COR Biosciences). (C) Reduced lysyl oxidase enzyme activity in the patient’s fibroblast. Lysyl oxidase activity was measured using a fluorescent enzyme assay. Each sample was measured in triplicate, and enzyme activity was normalized to total cellular protein content for each sample. Results are expressed as percentage of lysyl oxidase activity in patient’s fibroblasts relative to control. Control represents mean value of four independent normal cell lines (error bars indicate ± SD, n = 13, **p<0.0001). (D) Reduced levels of secreted lysyl oxidase in patient’s fibroblasts. Equal amounts of cell culture medium from two controls and patient’s fibroblasts grown to confluency were collected and concentrated using a protein concentrator column. Whole cell lysates were prepared by lysis of cells in Triton X-100 buffer. The upper panel shows a representative anti-lysyl oxidase immunoblot of whole cell extracts (lanes 1–3) and cell culture medium (lanes 4–6). The lower panel shows anti-Vinculin immunoblot to confirm equal loading of lysates in lanes 1–3. (E) Reduced lysyl oxidase activity in Zp-/- MEFs. Three Zp+/+ and five Zp-/- MEF cell lines were used to measure lysyl oxidase activity in two independent experiments. Each sample was measured in triplicate, and enzyme activity was normalized to total cellular protein content for each sample. Results are expressed as relative fold changes in lysyl oxidase activity. Data are shown as mean ± SD, n = 6 for Zp+/+ and n = 10 for Zp-/-,* p<0.05.
Fig 5.
Generation and phenotype of Hephl1 knockout mice.
(A) Schematic showing structure of the Hephl1 genomic region and the strategy used to generate the Hephl1 knockout (Zp-/-) allele (for details see Materials and Methods). (B) Quantitative real time PCR analysis of Hephl1 mRNA expression in Zp+/+ and Zp-/- placental tissue. No expression was observed in Zp-/- placental tissue using primers specific for exon 2. Expression was significantly reduced in Zp-/- when primers specific for a region downstream of knockout site (exon 18–19) were used, indicating instability of transcript lacking exon 2. Hprt was used as a housekeeping gene to normalize the expression of Hephl1. (C) Phenotype of the Zp-/- mouse showing short, curled whiskers (vibrissae). A wild-type (Zp+/+) littermate is shown on the left. (D) A close-up photograph of whiskers.