Molecular analysis of DMP1 mutants causing autosomal recessive hypophosphatemic rickets

Abstract

We previously demonstrated that the mutations Met1Val (M1V) and the deletion of nucleotides 1484-1490 (1484-1490del) in Dentin matrix protein-1 (DMP1) cause the novel disorder autosomal recessive hypophosphatemic rickets (ARHR), which is associated with elevated fibroblast growth factor-23 (FGF23). To further understand the role of DMP1 in ARHR, we undertook molecular genetic and in vitro expression studies. First, we examined a kindred with a severe hypophosphatemic rickets phenotype and recessive inheritance. Analyses of this family demonstrated that the affected members had elevated serum FGF23 and carried a large, biallelic deletion that removed the majority of DMP1. At a minimum, this deletion encompassed 49 kb between DMP1 exon 3 and an intergenic region 5' to the next telomeric gene, integrin-binding sialoprotein (IBSP). We next performed immunofluorescent studies in cells to understand the effects of the known ARHR mutations on DMP1 cellular processing. These analyses showed that the M1V DMP1 mutant was not sorted to the trans-Golgi network (TGN) and secretory pathway, but filled the entire cytoplasm. In contrast, the 1484-1490del mutant localized to the TGN and was secreted, similar to wild type DMP1. The 1484-1490del mutation replaces the DMP1 18 C-terminal amino acids with 33 non-native residues. Truncation of wild type DMP1 by these native 18 residues followed by Western blot and confocal microscopic analyses demonstrated a wild type expression pattern when compared with the 1484-1490del mutant, indicating that the last 18 residues are not critical for cellular trafficking, but that the 33 additional residues arising from the 1484-1490del mutation likely compromise DMP1 processing. The relationship between DMP1 and FGF23 is unclear. To test endogenous DMP1 response to serum metabolites that also regulate FGF23, UMR-106 cells were treated with 1,25(OH)(2) vitamin D (1x10(-7) M) and showed a 12-fold increase in DMP1 mRNA and protein at 24 h. In summary, we have identified a novel DMP1 deletion as the cause of ARHR, as well as demonstrated that the ARHR mutations alter DMP1 cellular processing, and that DMP1 can be regulated by vitamin D. Taken together, this work expands our understanding of the genetic and molecular mechanisms associated with DMP1 alterations causing ARHR.

Figure 1. Novel ARHR DMP1 mutation

A.) The PCR amplification of DMP1 exons, intergenic region, and IBSP from unaffected father (I-1) and the ARHR patients (II-1 affected brother and II-3 affected sister) revealed a large deletion in the patients. The family pedigree of individuals with recessive hypophosphatemia assessed in these studies is shown on the right. B.) A schematic of the deletion between DMP1 and IBSP in this kindred. The deletion is a minimum of 49 kb and a maximum of 145 kb (NCBI accession file NT_016354). The known gene exons are labeled as ‘Ex.’

Figure 1. Novel ARHR DMP1 mutation

A.) The PCR amplification of DMP1 exons, intergenic region, and IBSP from unaffected father (I-1) and the ARHR patients (II-1 affected brother and II-3 affected sister) revealed a large deletion in the patients. The family pedigree of individuals with recessive hypophosphatemia assessed in these studies is shown on the right. B.) A schematic of the deletion between DMP1 and IBSP in this kindred. The deletion is a minimum of 49 kb and a maximum of 145 kb (NCBI accession file NT_016354). The known gene exons are labeled as ‘Ex.’

Figure 2. Cellular sorting of ARHR mutants

A.) UMR-106 cells were transfected with a wild type, 1484-1490del, or M1V DMP1 cDNAs. Immunofluorescent confocal analyses with anti-V5 showed that wild type DMP1 (green) was sorted to the TGN (red; red and green overlay is yellow), the M1V mutant (green) fills the cytoplasm and does not enter the TGN (red). Nuclei were stained with DAPI (blue). B.) The results in 2A were confirmed using an antibody specific to the TGN, TGN38 (red). C.) When DMP1 constructs were re-assessed lacking the C-terminal V5 tag using anti-human DMP1, the results are the same as in upper panels, demonstrating that the tag does not interfere with cellular processing

Figure 3. Analyses of conserved DMP1 C-terminal residues

A.) Transient transfection and Western analyses showed that the 94 and 57 kD species in conditioned media originating from the DMP1Δ18 plasmid (right lane), when compared to the 1484-1490del mutant, is processed and secreted similar to wild type DMP1 (left lane). The M1V mutant is not secreted into the media. B.) Immunofluorescent analysis of DMP1Δ18 in HEK293 cells with anti-V5 (green) showed co-localization with the TGN (red). The upper panel is the TGN, the lower panel is DMP1, nuclei (DAPI-blue), and the TGN merged.

Figure 4. Vitamin D regulation of DMP1 in UMR-106 cells

UMR-106 cells were treated with either vehicle or 1×10⁷ 1,25(OH)₂ vitamin D (1,25D) for the times indicated. A.) DMP1 mRNA, assessed by qPCR was maximally elevated 12-fold at 24 hours when compared to vehicle-treated control cells (*; P<0.05). Inset: DMP1 protein was increased in the cell media as assessed by Western analysis when compared to vehicle-treated control cells at 24 h. Beta-actin from the lysates of the same cells is shown as control. FGF23 was also elevated with vitamin D treatment as determined by: B.) qPCR (*; P<0.05); and C.) qualitative RT-PCR with intron-spanning primers to rat FGF23.