Table 2.

Gene and mutation identification Analysis of the single nucleotide polymorphism (SNP) microarray genotypes from six patients with PHID identified a single common 1.4 Mb region of shared homozygosity within cytogenetic band 10q22.1, between SNPs rs10509321 and rs1900515 (Fig. 1A). Inspection of the Ensembl and UCSC Genome Browsers demonstrated that the region included 14 known and predicted genes: ADAMTS14, C10orf27, CDH23, EIF4EBP2, LRRC2, KIAA1274, NODAL, NPFFR1, PCBD1, PPA1, PRF1, SLC29A3, SGPL1, UNC5B. Positional candidate genes were selected for sequencing in order to identify pathogenic mutations (Fig. 1B). Five different homozygous mutations were identified in the SLC29A3 gene encoding the equilibrative nucleoside transporter 3 (Fig. 1C, Table 1). Human ENT3 has been shown recently to be the cause of H syndrome, an autosomal recessive disorder with overlapping skin features to PHID (6), indicating that these are allelic disorders. hENT3 belongs to a structurally unique protein family of nucleoside and/or nucleobase transporters (13,14). The five homozygous mutations identified segregated with the disease within each family and were not present in over 100 Lebanese and Pakistani control chromosomes. Figure 1.Open in new tabDownload slideHomozygosity mapping and mutation results in five PHID families. (A) Homozygosity mapping of chromosome 10. Red bars indicate heterozygous single nucleotide polymorphisms (SNPs); blue bars, an extended run of homozygous SNPs greater than would be predicted by chance; green bars, a run of homozygous SNPs but insufficient to meet criteria for significance. (B) Genes within the 1.4 Mb interval bounded by heterozygous SNPs rs10509321 (CII.3) and rs1900515 (DII.1). (C) SLC29A3 sequence electrophoretograms demonstrating homozygous mutations in five PHID families. (D) Protein structure prediction of hENT3 by SVMTM software. (E) Multispecies alignment of the second transmembrane and the C-terminal cytoplasmic domains of ENT3 protein (accession numbers: Homo sapiens, NP_060814.4; Canis lupus, XP_546152.2; Bos taurus, XP_580640.3; Mus musculus, NP_076085.1; monodelphus domesticus, XP_001344438.1; Gallus gallus, XP_421594.2; Gasterosteus aculeatus, ENSGACP00000006741; Takifugu rubipens, ENSTRUP00000028258; Drosophila melangaster, FBpp0077753; Caenorhabditis elegans, NP_501807.3). Table 1.Mutation summary Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03 Open in new tab Table 1.Mutation summary Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03 Open in new tab Two novel homozygous mutations predicted to cause premature protein truncation were identified, a 1 bp deletion c.940delT and a nonsense mutation c.1330G>T. The deletion causes a frame shift mutation p.Y314Tfs in transmembrane domain 7 and is predicted to result in the substitution of transmembrane domains 8–11 with an irrelevant protein sequence of 91 amino acids derived from an alternate reading frame. The nonsense mutation p.E444X is predicted to lead to premature protein truncation with loss of part of the final cytoplasmic domain and transmembrane domain 11 of hENT3. It would appear unlikely that the either residual protein would be capable of nucleoside transport function.Three homozygous missense mutations of SLC29A3 were identified to involve conserved amino acids (Table 1). The first missense mutation involves Methionine116 in transmembrane domain 2, which is a hydrophobic amino acid in species as distantly related as humans and nematodes, with a large basic amino acid, Arginine (p.M116R; Fig. 1C and E). The Grantham score for this substitution is 91, which is classified as a moderate change (15). Analysis using the SIFT prediction software also classifies this as a likely deleterious change (16). The introduction of a large basic amino acid into a transmembrane domain would likely result in significant disruption of the hydrophobic nature of the domain and result in retention of the transporter protein within the endoplasmic reticulum, as has been demonstrated for other membrane proteins (17).The second missense change is a recently reported Glycine to Arginine substitution at hENT3 amino acid 437 (p.G437R Fig. 1C), which is located in the putative cytoplasmic domain between transmembrane domains 10 and 11. This is the only mutation in the PHID cohort that is shared with the known mutations in H syndrome. Of particular interest is the observation that this is the sole individual in the PHID cohort who has not developed insulin-dependent DM. The p.G437R change involves the substitution of a large basic amino acid at a position where 41 members of the ENT family as distantly related as humans and Arabidopsis thaliana encode a small hydrophobic amino acid (Fig. 1E) (18). The Grantham difference for this change is 125 and is classified as a moderately radical change.The third missense change occurs at hENT3 amino acid residue 449, which is also located in the final putative intracellular domain between transmembrane domains 10 and 11, and involves the substitution of a Threonine by an Arginine (p.T449R). The amino acids encoded at this position in hENT3 orthologs are small polar/uncharged or small hydrophobic residues (Fig. 1E). Substitution by arginine is associated with a Grantham score of 71 and is predicted by SIFT as not to be tolerated.

RESULTS Gene and mutation identification Analysis of the single nucleotide polymorphism (SNP) microarray genotypes from six patients with PHID identified a single common 1.4 Mb region of shared homozygosity within cytogenetic band 10q22.1, between SNPs rs10509321 and rs1900515 (Fig. 1A). Inspection of the Ensembl and UCSC Genome Browsers demonstrated that the region included 14 known and predicted genes: ADAMTS14, C10orf27, CDH23, EIF4EBP2, LRRC2, KIAA1274, NODAL, NPFFR1, PCBD1, PPA1, PRF1, SLC29A3, SGPL1, UNC5B. Positional candidate genes were selected for sequencing in order to identify pathogenic mutations (Fig. 1B). Five different homozygous mutations were identified in the SLC29A3 gene encoding the equilibrative nucleoside transporter 3 (Fig. 1C, Table 1). Human ENT3 has been shown recently to be the cause of H syndrome, an autosomal recessive disorder with overlapping skin features to PHID (6), indicating that these are allelic disorders. hENT3 belongs to a structurally unique protein family of nucleoside and/or nucleobase transporters (13,14). The five homozygous mutations identified segregated with the disease within each family and were not present in over 100 Lebanese and Pakistani control chromosomes. Figure 1.Open in new tabDownload slideHomozygosity mapping and mutation results in five PHID families. (A) Homozygosity mapping of chromosome 10. Red bars indicate heterozygous single nucleotide polymorphisms (SNPs); blue bars, an extended run of homozygous SNPs greater than would be predicted by chance; green bars, a run of homozygous SNPs but insufficient to meet criteria for significance. (B) Genes within the 1.4 Mb interval bounded by heterozygous SNPs rs10509321 (CII.3) and rs1900515 (DII.1). (C) SLC29A3 sequence electrophoretograms demonstrating homozygous mutations in five PHID families. (D) Protein structure prediction of hENT3 by SVMTM software. (E) Multispecies alignment of the second transmembrane and the C-terminal cytoplasmic domains of ENT3 protein (accession numbers: Homo sapiens, NP_060814.4; Canis lupus, XP_546152.2; Bos taurus, XP_580640.3; Mus musculus, NP_076085.1; monodelphus domesticus, XP_001344438.1; Gallus gallus, XP_421594.2; Gasterosteus aculeatus, ENSGACP00000006741; Takifugu rubipens, ENSTRUP00000028258; Drosophila melangaster, FBpp0077753; Caenorhabditis elegans, NP_501807.3). Table 1.Mutation summary Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03 Open in new tab Table 1.Mutation summary Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03 Open in new tab Two novel homozygous mutations predicted to cause premature protein truncation were identified, a 1 bp deletion c.940delT and a nonsense mutation c.1330G>T. The deletion causes a frame shift mutation p.Y314Tfs in transmembrane domain 7 and is predicted to result in the substitution of transmembrane domains 8–11 with an irrelevant protein sequence of 91 amino acids derived from an alternate reading frame. The nonsense mutation p.E444X is predicted to lead to premature protein truncation with loss of part of the final cytoplasmic domain and transmembrane domain 11 of hENT3. It would appear unlikely that the either residual protein would be capable of nucleoside transport function.Three homozygous missense mutations of SLC29A3 were identified to involve conserved amino acids (Table 1). The first missense mutation involves Methionine116 in transmembrane domain 2, which is a hydrophobic amino acid in species as distantly related as humans and nematodes, with a large basic amino acid, Arginine (p.M116R; Fig. 1C and E). The Grantham score for this substitution is 91, which is classified as a moderate change (15). Analysis using the SIFT prediction software also classifies this as a likely deleterious change (16). The introduction of a large basic amino acid into a transmembrane domain would likely result in significant disruption of the hydrophobic nature of the domain and result in retention of the transporter protein within the endoplasmic reticulum, as has been demonstrated for other membrane proteins (17).The second missense change is a recently reported Glycine to Arginine substitution at hENT3 amino acid 437 (p.G437R Fig. 1C), which is located in the putative cytoplasmic domain between transmembrane domains 10 and 11. This is the only mutation in the PHID cohort that is shared with the known mutations in H syndrome. Of particular interest is the observation that this is the sole individual in the PHID cohort who has not developed insulin-dependent DM. The p.G437R change involves the substitution of a large basic amino acid at a position where 41 members of the ENT family as distantly related as humans and Arabidopsis thaliana encode a small hydrophobic amino acid (Fig. 1E) (18). The Grantham difference for this change is 125 and is classified as a moderately radical change.The third missense change occurs at hENT3 amino acid residue 449, which is also located in the final putative intracellular domain between transmembrane domains 10 and 11, and involves the substitution of a Threonine by an Arginine (p.T449R). The amino acids encoded at this position in hENT3 orthologs are small polar/uncharged or small hydrophobic residues (Fig. 1E). Substitution by arginine is associated with a Grantham score of 71 and is predicted by SIFT as not to be tolerated.

Gene and mutation identification Analysis of the single nucleotide polymorphism (SNP) microarray genotypes from six patients with PHID identified a single common 1.4 Mb region of shared homozygosity within cytogenetic band 10q22.1, between SNPs rs10509321 and rs1900515 (Fig. 1A). Inspection of the Ensembl and UCSC Genome Browsers demonstrated that the region included 14 known and predicted genes: ADAMTS14, C10orf27, CDH23, EIF4EBP2, LRRC2, KIAA1274, NODAL, NPFFR1, PCBD1, PPA1, PRF1, SLC29A3, SGPL1, UNC5B. Positional candidate genes were selected for sequencing in order to identify pathogenic mutations (Fig. 1B). Five different homozygous mutations were identified in the SLC29A3 gene encoding the equilibrative nucleoside transporter 3 (Fig. 1C, Table 1). Human ENT3 has been shown recently to be the cause of H syndrome, an autosomal recessive disorder with overlapping skin features to PHID (6), indicating that these are allelic disorders. hENT3 belongs to a structurally unique protein family of nucleoside and/or nucleobase transporters (13,14). The five homozygous mutations identified segregated with the disease within each family and were not present in over 100 Lebanese and Pakistani control chromosomes. Figure 1.Open in new tabDownload slideHomozygosity mapping and mutation results in five PHID families. (A) Homozygosity mapping of chromosome 10. Red bars indicate heterozygous single nucleotide polymorphisms (SNPs); blue bars, an extended run of homozygous SNPs greater than would be predicted by chance; green bars, a run of homozygous SNPs but insufficient to meet criteria for significance. (B) Genes within the 1.4 Mb interval bounded by heterozygous SNPs rs10509321 (CII.3) and rs1900515 (DII.1). (C) SLC29A3 sequence electrophoretograms demonstrating homozygous mutations in five PHID families. (D) Protein structure prediction of hENT3 by SVMTM software. (E) Multispecies alignment of the second transmembrane and the C-terminal cytoplasmic domains of ENT3 protein (accession numbers: Homo sapiens, NP_060814.4; Canis lupus, XP_546152.2; Bos taurus, XP_580640.3; Mus musculus, NP_076085.1; monodelphus domesticus, XP_001344438.1; Gallus gallus, XP_421594.2; Gasterosteus aculeatus, ENSGACP00000006741; Takifugu rubipens, ENSTRUP00000028258; Drosophila melangaster, FBpp0077753; Caenorhabditis elegans, NP_501807.3). Table 1.Mutation summary Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03 Open in new tab Table 1.Mutation summary Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03 Open in new tab Two novel homozygous mutations predicted to cause premature protein truncation were identified, a 1 bp deletion c.940delT and a nonsense mutation c.1330G>T. The deletion causes a frame shift mutation p.Y314Tfs in transmembrane domain 7 and is predicted to result in the substitution of transmembrane domains 8–11 with an irrelevant protein sequence of 91 amino acids derived from an alternate reading frame. The nonsense mutation p.E444X is predicted to lead to premature protein truncation with loss of part of the final cytoplasmic domain and transmembrane domain 11 of hENT3. It would appear unlikely that the either residual protein would be capable of nucleoside transport function.Three homozygous missense mutations of SLC29A3 were identified to involve conserved amino acids (Table 1). The first missense mutation involves Methionine116 in transmembrane domain 2, which is a hydrophobic amino acid in species as distantly related as humans and nematodes, with a large basic amino acid, Arginine (p.M116R; Fig. 1C and E). The Grantham score for this substitution is 91, which is classified as a moderate change (15). Analysis using the SIFT prediction software also classifies this as a likely deleterious change (16). The introduction of a large basic amino acid into a transmembrane domain would likely result in significant disruption of the hydrophobic nature of the domain and result in retention of the transporter protein within the endoplasmic reticulum, as has been demonstrated for other membrane proteins (17).The second missense change is a recently reported Glycine to Arginine substitution at hENT3 amino acid 437 (p.G437R Fig. 1C), which is located in the putative cytoplasmic domain between transmembrane domains 10 and 11. This is the only mutation in the PHID cohort that is shared with the known mutations in H syndrome. Of particular interest is the observation that this is the sole individual in the PHID cohort who has not developed insulin-dependent DM. The p.G437R change involves the substitution of a large basic amino acid at a position where 41 members of the ENT family as distantly related as humans and Arabidopsis thaliana encode a small hydrophobic amino acid (Fig. 1E) (18). The Grantham difference for this change is 125 and is classified as a moderately radical change.The third missense change occurs at hENT3 amino acid residue 449, which is also located in the final putative intracellular domain between transmembrane domains 10 and 11, and involves the substitution of a Threonine by an Arginine (p.T449R). The amino acids encoded at this position in hENT3 orthologs are small polar/uncharged or small hydrophobic residues (Fig. 1E). Substitution by arginine is associated with a Grantham score of 71 and is predicted by SIFT as not to be tolerated.

RESULTS Gene and mutation identification Analysis of the single nucleotide polymorphism (SNP) microarray genotypes from six patients with PHID identified a single common 1.4 Mb region of shared homozygosity within cytogenetic band 10q22.1, between SNPs rs10509321 and rs1900515 (Fig. 1A). Inspection of the Ensembl and UCSC Genome Browsers demonstrated that the region included 14 known and predicted genes: ADAMTS14, C10orf27, CDH23, EIF4EBP2, LRRC2, KIAA1274, NODAL, NPFFR1, PCBD1, PPA1, PRF1, SLC29A3, SGPL1, UNC5B. Positional candidate genes were selected for sequencing in order to identify pathogenic mutations (Fig. 1B). Five different homozygous mutations were identified in the SLC29A3 gene encoding the equilibrative nucleoside transporter 3 (Fig. 1C, Table 1). Human ENT3 has been shown recently to be the cause of H syndrome, an autosomal recessive disorder with overlapping skin features to PHID (6), indicating that these are allelic disorders. hENT3 belongs to a structurally unique protein family of nucleoside and/or nucleobase transporters (13,14). The five homozygous mutations identified segregated with the disease within each family and were not present in over 100 Lebanese and Pakistani control chromosomes. Figure 1.Open in new tabDownload slideHomozygosity mapping and mutation results in five PHID families. (A) Homozygosity mapping of chromosome 10. Red bars indicate heterozygous single nucleotide polymorphisms (SNPs); blue bars, an extended run of homozygous SNPs greater than would be predicted by chance; green bars, a run of homozygous SNPs but insufficient to meet criteria for significance. (B) Genes within the 1.4 Mb interval bounded by heterozygous SNPs rs10509321 (CII.3) and rs1900515 (DII.1). (C) SLC29A3 sequence electrophoretograms demonstrating homozygous mutations in five PHID families. (D) Protein structure prediction of hENT3 by SVMTM software. (E) Multispecies alignment of the second transmembrane and the C-terminal cytoplasmic domains of ENT3 protein (accession numbers: Homo sapiens, NP_060814.4; Canis lupus, XP_546152.2; Bos taurus, XP_580640.3; Mus musculus, NP_076085.1; monodelphus domesticus, XP_001344438.1; Gallus gallus, XP_421594.2; Gasterosteus aculeatus, ENSGACP00000006741; Takifugu rubipens, ENSTRUP00000028258; Drosophila melangaster, FBpp0077753; Caenorhabditis elegans, NP_501807.3). Table 1.Mutation summary Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03 Open in new tab Table 1.Mutation summary Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03Mutation NM_018344.4 . Protein CCDS7310.1 . Protein domain . Predicted . Mutation predictions . Grantham . SIFT . c.347T>Gp.Met116ArgTransmembrane Domain 2Missense Hydrophobic to large basic910.03c.940delTp.Tyr314ThrfsX91Transmembrane Domain 7Frameshift––c.1309G>Ap.Gly437ArgIntra-cytoplasmic DomainMissense Small hydrophobic to large basic12512.0c.1330G>Tp.Glu444XIntra-cytoplasmic DomainNonsense––c.1346C>Gp.Thr449ArgIntra-cytoplasmic DomainMissense Small polar to large basic710.03 Open in new tab Two novel homozygous mutations predicted to cause premature protein truncation were identified, a 1 bp deletion c.940delT and a nonsense mutation c.1330G>T. The deletion causes a frame shift mutation p.Y314Tfs in transmembrane domain 7 and is predicted to result in the substitution of transmembrane domains 8–11 with an irrelevant protein sequence of 91 amino acids derived from an alternate reading frame. The nonsense mutation p.E444X is predicted to lead to premature protein truncation with loss of part of the final cytoplasmic domain and transmembrane domain 11 of hENT3. It would appear unlikely that the either residual protein would be capable of nucleoside transport function.Three homozygous missense mutations of SLC29A3 were identified to involve conserved amino acids (Table 1). The first missense mutation involves Methionine116 in transmembrane domain 2, which is a hydrophobic amino acid in species as distantly related as humans and nematodes, with a large basic amino acid, Arginine (p.M116R; Fig. 1C and E). The Grantham score for this substitution is 91, which is classified as a moderate change (15). Analysis using the SIFT prediction software also classifies this as a likely deleterious change (16). The introduction of a large basic amino acid into a transmembrane domain would likely result in significant disruption of the hydrophobic nature of the domain and result in retention of the transporter protein within the endoplasmic reticulum, as has been demonstrated for other membrane proteins (17).The second missense change is a recently reported Glycine to Arginine substitution at hENT3 amino acid 437 (p.G437R Fig. 1C), which is located in the putative cytoplasmic domain between transmembrane domains 10 and 11. This is the only mutation in the PHID cohort that is shared with the known mutations in H syndrome. Of particular interest is the observation that this is the sole individual in the PHID cohort who has not developed insulin-dependent DM. The p.G437R change involves the substitution of a large basic amino acid at a position where 41 members of the ENT family as distantly related as humans and Arabidopsis thaliana encode a small hydrophobic amino acid (Fig. 1E) (18). The Grantham difference for this change is 125 and is classified as a moderately radical change.The third missense change occurs at hENT3 amino acid residue 449, which is also located in the final putative intracellular domain between transmembrane domains 10 and 11, and involves the substitution of a Threonine by an Arginine (p.T449R). The amino acids encoded at this position in hENT3 orthologs are small polar/uncharged or small hydrophobic residues (Fig. 1E). Substitution by arginine is associated with a Grantham score of 71 and is predicted by SIFT as not to be tolerated.

III-6

Figure 2Haplotype Analysis Performed in Six Families with H SyndromeAffected individuals are marked by gray symbols. The disease-associated haplotypes in the linked interval on chromosome 10q21.3–q22.1 are shown by vertical bars (red and green) in families A–F. Alleles excluded by recombination events are marked in pink. The physical locations of the polymorphic microsatellites are given near the haplotype for individual AI-2.

III-6

Figure 2Haplotype Analysis Performed in Six Families with H SyndromeAffected individuals are marked by gray symbols. The disease-associated haplotypes in the linked interval on chromosome 10q21.3–q22.1 are shown by vertical bars (red and green) in families A–F. Alleles excluded by recombination events are marked in pink. The physical locations of the polymorphic microsatellites are given near the haplotype for individual AI-2.

Dysosteosclerosis (DSS), an extremely rare dense bone disease, features short stature and fractures and sometimes optic atrophy, cranial nerve palsy, developmental delay, and failure of tooth eruption in infancy or early childhood consistent with osteopetrosis (OPT). Bone histology during childhood shows unresorbed primary spongiosa from deficient osteoclast action. Additionally, there is remarkable progressive flattening of all vertebrae and, by adolescence, paradoxical metaphyseal osteopenia with thin cortical bone. Reports of consanguinity indicate autosomal recessive inheritance, yet more affected males than females suggest X-linked recessive inheritance. We investigated a nonconsanguineous girl with DSS. Osteosclerosis was discovered at age 7 months. Our studies, spanning ages 11 to 44 months, showed weight at approximately 50th percentile, and length diminishing from approximately 30th percentile to –2.3 SD. Head circumference was +4 SD. The patient had frontal bossing, blue sclera, normal teeth, genu valgum, and unremarkable joints. Radiographs showed orbital and facial sclerosis, basilar thickening, bone-in-bone appearance of the pelvis, sclerotic long bone ends, and fractures of ribs and extremities. Progressive metaphyseal widening occurred as vertebrae changed from ovoid to flattened and became beaked anteriorly. A hemogram was normal. Consistent with OPT, serum parathyroid hormone (PTH) concentrations reflected dietary calcium levels. Serum bone alkaline phosphatase, osteocalcin, and TRACP-5b were subnormal. The iliac crest contained excessive primary spongiosa and no osteoclasts. No mutations were identified in the splice sites or exons for the genes encoding chloride channel 7, T-cell immune regulator 1, OPT-associated transmembrane protein 1, and monocyte colony-stimulating factor (M-CSF) and its receptor C-FMS, ANKH, OPG, RANK, and RANKL. Genomic copy-number microarray was unrevealing. Hence, DSS is a distinctive OPT of unknown etiology featuring osteoclast deficiency during early childhood. How osteopenia follows is an enigma of human skeletal pathobiology. © 2010 American Society for Bone and Mineral Research.Keywords: bone remodeling, erlenmeyer flask deformity, osteoclast, osteosclerosis, RANKL, skeletal dysplasiaGo to:IntroductionDysosteosclerosis (DSS; OMIM % 224300)(1) refers to an extremely rare disorder characterized in 1968 by Spranger and colleagues(2) that presents in infancy or early childhood and features diffuse osteosclerosis with expanded ends of the tubular bones and platyspondyly. In keeping with a type of osteopetrosis (OPT),(3) bone histology during childhood shows unresorbed primary spongiosa and therefore deficient osteoclast action.(4,5) The skeleton is dense but brittle. Patients develop short stature and fractures and can suffer optic atrophy and sometimes cranial nerve palsy, developmental delay, and failure of tooth eruption,(1,6–9) in keeping with “marble bone disease.”(3) MRI has suggested delayed myelination in the brain,(10) seizures have occurred,(11) and various skin lesions have been reported.(11–14) Consanguinity(2,6,12,13,15) and a greater number of affected males (including in two generations) than females(11) suggest that DSS can be either an autosomal recessive trait(1,15) or an X-linked recessive trait,(11) but the genetic defect(s) is not known.(1)The clinical course and prognosis of DSS are not well understood. In 1978, Houston and colleagues(6) reported a 15-year-old boy whose iliac crest histology at 20 months of age revealed unresorbed primary spongiosa consistent with OPT. Subsequently, however, metaphyseal widening and sclerosis of early childhood changed to greater metaphyseal expansion (extending into the diaphysis) with remarkable osteopenia and cortical thinning by adolescence. In 2008, his follow-up at age 44 years was reported by Lemire and Wiebe,(16) who documented persistence of the widened proximal and distal ends of the long tubular bones with central diaphyses that were not expanded and showed cortical thickening.Here, we describe our clinical, biochemical, radiologic, histopathologic, and genetic investigation of a young girl with DSS and review the pertinent literature.Go to:Materials and MethodsInformed written consent for all studies was obtained as approved by the Human Research Protection Office, Washington University School of Medicine (St Louis, MO, USA).Case ReportThis 44-month-old American girl was admitted to our research center at ages 11, 23, 25, and 35 months. Additional outpatient serum and urine specimens were obtained at ages 27, 30, and 44 months.Medical historyThe patient was born vaginally at 38 weeks' gestation to a 37-year-old G3, P3, A0 woman. Birth weight was low for gestational age at 5 pounds, 14 ounces, but length 19½ inches.At 5 months of age, computed tomography (CT) for rapidly increasing head circumference (25th percentile at birth, 97th percentile at age 5 months, despite growth at 50th percentile for both length and weight) reportedly showed prominent ventricles, sulci, and subarachnoid spaces but no space-occupying lesion or hemorrhage.At 7 months of age, a radiograph (see below) for posttraumatic arm pain prompted a skeletal survey showing metaphyseal sclerosis that was attributed to “rickets.” Vitamin D was prescribed but not taken.From birth to 8 months of age, the patient had recurrent ear and skin infections and abdominal discomfort believed from colic.Between 8 and 11 months of age, several episodes of joint-centered limb pain were thought to be from dislocations. On referral to us at age 11 months, psychomotor development was on time. Tooth eruption was not delayed.At age 20 months, the patient fractured her distal left femur during a fall. Healing was delayed in a full splint.Ophthalmologic exam was normal at age 28 months. Audiometry had not been performed, but she seemed to hear well, had clear speech, and was bilingual.At age 30 months, she fractured through dense metaphyseal bone in her right radius and ulna during a fall. Reportedly, healing was slow (casted for 3 months). Parental recollection of further breaks was negative, but radiographs showed previous fractures. Morning stiffness in her lower limbs lasted 2 to 3 hours each day. She had chronic nasal “stuffiness.” Several episodes of erythematous papules involved her lower abdomen and were always on at least one area of her body but usually her belly, legs, or arms. Episodes lasted 3 to 4 days, were not associated with fever, and did not hurt or itch.At 36 months, she fell and fractured her distal right femur, where dense metaphyseal and more normal-appearing diaphyseal bone joined. Healing was delayed in a long-leg cast for 3 months.She received acetaminophen for nighttime fussiness but no multivitamins or supplements. She drank filtered municipal tap water. There were no factories near her home, but lead paint had been removed from a nearby house.At age 44 months (3⅔ years), interval history revealed a fractured left femur from a fall onto a carpeted floor with delayed healing while casted. Subsequently, however, the patient could run. There was generalized morning stiffness and painful ribs.Her healthy brothers were 6 and 8 years old. The family history was negative for skeletal disease or parental consanguinity. The father was of Turkish descent. The mother's family emigrated to Turkey from Greece 70 years previously.Physical findingsWhen examined for diagnosis at age 11 months, the patient had a high forehead, frontal bossing, blue sclerae, a flattened nasal bridge, downturned mouth (Fig. 1A), a narrow and high-arched palate, and a wide palatine ridge with nine normal-appearing teeth, indicating no delay in tooth eruption. Joint range of motion was full. A dermatosis involved her lower abdomen (Fig. 1B).Open in a separate windowFig. 1Patient. (A) At age 44 months, there is frontal bossing, a small midface, low-set ears, and blue sclerae. (B) At age 3 years, the recurring dermatosis (arrow) is present on the lower abdomen.At age 22 months, she had no palpable cranial sutures, bony tenderness, or genu valgum. Although 50th percentile for weight and length at age 5 months, between 11 and 35 months of age she maintained her weight at approximately 30th percentile of normal, but length decreased from approximately 30th percentile to below the growth curve (Z-score = –2.5). The upper/lower body segment Z-score increased to greater than 2.0. Arm span was less than 5th percentile. Head circumference Z-score decreased from +4.3 to +3.6. Genu valgum was present radiographically as early as 22 months of age.At age 44 months, weight was approximately 59th percentile, and length was –2.3 SD. Arm span versus height had progressed initially along the 50th percentile, but arm span measurements now showed that her arms were disproportionately short.(23)Biochemical testingFasting blood was used for all in-patient studies. Biochemical profiles were obtained from our Dade Behring Dimension Xpand instrument (Siemens Health Care Diagnostics, Inc., Los Angeles, CA, USA) using the reference ranges provided by the manufacturer. However, for all key findings, results were contrasted with values from fasting sera obtained in 2006 and 2007 from 34 healthy children and 24 healthy adults.(17)At age 11 months, the patient was given a diet containing approximately 400 mg of calcium per day to match her home intake estimated from a 7-day food record [Recommended Daily Allowance (RDA) for age = 270 mg]. At age 23 months, intake was approximately 950 mg. At age 35 months, we reduced her calcium intake to 775 mg daily (RDA of 500 mg recommended) to prevent suppressed serum parathyroid hormone (PTH) levels (see “Results”).To explore whether DSS is an OPT, we assayed serum levels of the brain isoenzyme of creatine kinase, BB-CK (Kit K20, Sebia, Norcorss, GA, USA),(18) tartrate-resistant acid phosphatase (TRACP-5b, Kit 8033, Quidel, Los Angeles, CA, USA),(17) and lactate dehydrogenase (LDH) and aspartate aminotransferase (AST).(17)Skeletal remodeling was assessed by quantitating serum bone-specific alkaline phosphatase (BAP) using ELISA (Quidel Corp., San Diego, CA, USA) and osteocalcin (Kit LKON1, Siemens Health Care Diagnostics, Inc.) and urine free deoxypyridinoline (DPD, Immulite 1000 Pyrilinks, D Kit, Siemens Medical Solutions Diagnostics, Ltd., Lianberis, Gwynedd, UK).Radiologic studiesAll radiologic studies, including images obtained elsewhere, were reviewed to assess the nature and evolution of the patient's skeletal disease, including fractures and their rate of healing. Dual-energy X-ray absorptiometry (DXA) of the patient and her parents involved a Hologic QDR 4500-A instrument (Hologic Corp., Waltham, MA, USA).Histopathologic studiesAt 2 years of age, two 3-day oral courses of demeclocycline separated by 10 days were ingested. Two days after the last dose, a transapophyseal iliac crest biopsy was performed using a Trap-Lok needle (Medical Device Technologies; Gainesville, FL, USA). The specimen was fixed in 70% ethanol and 30% water, embedded without decalcification in methyl methacrylate, and cut into 5-µm sections for staining via the hematoxylin and eosin (H&E), toluidine blue, and von Kossa methods. Unstained 10-µm sections were used for fluorescence microscopy.Mutation analysesThree genes (exons and mRNA splice sites) associated with OPT [chloride channel 7 (CLCN7), T-cell immune regulator 1 (TCIRG1), and OPT-associated transmembrane protein 1 (OSTM1)] had been studied at Connective Tissue Gene Tests (Allentown, PA, USA). We purified genomic DNA from blood leukocytes using the Gentra Puregene DNA extraction kit (Invitrogen, Carlsbad, CA, USA) and examined all coding exons and adjacent mRNA splice sites of TNFRSF11A encoding RANK, TNFRSF11B encoding OPG, and TNFSF11 encoding RANK ligand (RANKL) by PCR and sequencing in both directions using published methods and primers.(19–21) Because few osteoclasts were noted on our patient's iliac crest biopsy, the CSF1 and CSF1R genes encoding monocyte colony-stimulating factor (M-CSF) and its receptor C-FMS, respectively, were examined using primers and conditions that we developed (available on request). Furthermore, her bone modeling disturbances and the literature report of resolution of long bone osteosclerosis in DSS(6,16) led us to study the ANKH gene deactivated in craniometaphyseal dysplasia.(22) DNA sequences were evaluated using AlignX software (Vector NTI, Invitrogen) and by inspecting electropherograms.Genomic DNA copy-number microarray analysis of the proband and parents was performed using the Affymetrix SNP 6.0 chip at the Laboratory for Clinical Genomics, Washington University School of Medicine (St Louis, MO, USA) and the Partek Genomics Suite (Partek, St Louis, MO, USA).Literature reviewWe assessed all previous reports of DSS using original publications obtained from the Becker Library of Medicine, Washington University School of Medicine (St Louis, MO, USA), especially the radiographic reproductions.Go to:ResultsBiochemical testingSerum ionized calcium was highest, 5.3 mg/dL (normal 4.7 to 5.3 mg/dL), during the first admission at age 11 months (Table 1), when her estimated dietary intake of calcium significantly exceeded the RDA. With less dietary calcium, serum generally contained lower total and ionized calcium concentrations, but PTH levels fluctuated considerably over ages 11 to 44 months, ranging from 7 to 120 pg/mL (normal 4 to 52 pg/mL; Table 1). Calcium in timed urine collections ranged from 196 to 241 mg/g of creatinine (Cr).Table 1Longitudinal Biochemical FindingsSerumTotal CaCa2+PTHPiALPBAPOsteocalcinTRACP-5bAge (mos.)(normal 8.9–10.2 mg/dL)(normal 4.7–5.3 mg/dL)(normal 4–52 pg/mL)(normal 3.4–5.9 mg/dL)(normal 157–477 IU/L)(normal 88–247 U/L)(normal 37–119 ng/mL)(normal 13–23 U/L)1110.15.3344.7213—982310.25.2335.0162—2—259.14.91204.52108512227*10.1—75.2125———30*10.0—384.6140———359.95.0624.21364112344*9.85.0224.41766363Open in a separate window— = not done.*= non-fasting outpatient serum.Serum 25-hydroxyvitamin D [25(OH)D] was 28 ng/mL (normal 32 to 100 ng/mL) (Lab Corp., Burlington, NC, USA), with normal levels of magnesium and inorganic phosphate.Serum alkaline phosphatase (ALP) decreased to as low as 125 IU/L (normal 157 to 477 IU/L) and was intermittently subnormal for age, especially in view of recurrent fractures. Serum BAP, osteocalcin, and TRACP-5b levels all were consistently low (Table 1). Urine free DPD was 18 nmol/mmol of Cr (23 months) and 14 nmol/mmol of Cr (44 months) (normal 7 to 34 nmol/mmol).Serum BB-CK, LDH, and AST levels were unremarkable, unlike in some OPTs, where elevated values can occur.(17,18) Urine did not contain heavy metals. Serum was negative for lead (Legacy Laboratory Services, Portland, OR, USA).At ages 23, 35, and 44 months, complete blood counts were normal, including monocytes. Erythrocyte sedimentation rate seemed elevated for a child at 18, 15, and 24 mm/h, respectively (normal 0 to 20 mm/h for young adults), and serum C-reactive protein was consistently normal. Serum contained elevated IgG of 1416 mg/dL (normal 453 to 916 mg/dL) and slightly high IgA of 111 mg/dL (normal 20 to 100 mg/dL) (Lab Corp.).Radiologic findingsRadiographic skeletal surveys were available from ages 7, 11, 22, and 33 months. Right hand and wrist films were available at 30 months, left knee at 42 months, and right knee at 47 months.A generalized skeletal disorder was revealed, where every imaged bone was abnormal (Figs. 2 through ​through5).5). Prominent findings included basilar sclerosis of the skull and modeling abnormalities (expansion) in the major long bones resembling OPT but also ovoid vertebral bodies characteristic of DSS. The sclerosis must have occurred postnatally because relatively normal iliac bones had formed in utero that subsequently became sclerotic peripherally. The changes per anatomic site are described below.Open in a separate windowFig. 2Skull. (A) At age 22 months, anteroposterior view of the skull shows sclerosis that affects the periorbital region and skull base. The sinuses are opacified. (B) At age 33 months, the neurocranium appears large on lateral view, with diffuse sclerosis greatest at the base. The midface is small, and the mandibular angle is obtuse.Open in a separate windowFig. 5Hand and foot radiographs. At 22 (A) and 33 (B) months of age the left hand and at 22 (C) months of age the left foot show diffuse sclerosis greatest in the metaphyses of the metacarpals and metatarsals and distal radius and ulna. The amount of sclerosis in the proximal phalanges varies from much to little and is greatest in the metaphyseal regions. The tarsals have some sclerosis greatest in the cuboids. Pathologic fractures are shown through sclerotic bone in the distal radius (arrow) and ulna in the anteroposterior projection (D) at 30 months and near nonsclerotic bone in the right femur (arrows) at 47 months (E).Axial skeletonSkull The neurocranium at ages 22 and 33 months was large, resulting in craniofacial disproportion (Fig. 2). The frontal bone bulged anteriorly; the midface was small. The posterior fossa appeared prominent. Sutures were normal. Closure of the anterior fontanelle seemed delayed (open at 22 months; virtually closed at 33 months). The ascending ramus of the mandible was smaller than expected, and the mandibular angle was obtuse. Some teeth were malformed or abnormally oriented. The skull base was sclerotic, increasing between 7 to 33 months. Periorbital sclerosis was especially prominent along with sclerosis of the greater and lesser wings of the sphenoid and the posterior portion of the sphenoid body. Maxillary and ethmoid sinuses were opaque at ages 22 and 33 months, and the mastoid air cells were small.Head CT At age 5 months, CT showed a sclerotic skull base but no hydrocephalus or intracranial calcifications. The sulci were slightly enlarged. There was a suggestion of abnormal myelination. The optic foramina, middle ears, and internal auditory canals appeared grossly normal, but detailed optic foramina and petrous bone imaging was not obtained. Multiple follow-up CTs elsewhere were essentially unchanged.Chest Chest shape was normal, but the ribs were broad and diffusely sclerotic with multiple lower rib fracture deformities at age 7 months. By age 22 months, the ribs were more sclerotic, the medullary cavities were delayed (because they were just beginning to form), and there were multiple healed fractures that had become more numerous in the fifth through the eleventh ribs bilaterally. The ribs were shortened with a modeling error anteriorly, and wide anteriorly and laterally (Fig. 3A). Fractures had occurred anteriorly, and laterally and were not remodeling. At 33 months, the ribs were still sclerotic and widened, and they had more recent fractures and multiple fracture deformities but were showing further medullary cavity development. The clavicles were sclerotic, and their medial ends widened at ages 22 and 33 months. The sternal segments were broad and sclerotic, with a central lucency, as were the scapulae. The heart and lungs appeared normal.Open in a separate windowFig. 3Spine. This anteroposterior projection (A) at 22 months shows sclerotic vertebrae, dense ribs with healing fractures, sclerotic medial clavicles, and bone-in-bone appearance in the ilium as seen in some forms of OPT. Between 7 (B) and 33 (C) months of age, the lateral spine shows sclerosis and flattening of vertebral bodies that increased with age. The vertebral bodies have a pointed configuration anteriorly (greatest at 7 months). Sclerosis affects the entire vertebra.Pelvis Sclerosis was accompanied by a normal-appearing central portion of the iliac bones causing a “bone in bone” appearance at age 22 months (Fig. 3A), and even more apparent at 33 months. The upper iliac borders were irregular. The ischial and pubic bones were sclerotic. The femoral heads and metaphyses were sclerotic, and the necks were wide with failure of modeling.Spine The vertebral bodies were sclerotic, with the cranial and caudal ends appearing “sandwich-like,” and their anterior margins were pointed anteriorly. Compared with age 7 months, when the lumbar vertebrae had a more oval shape (Fig. 3B), vertebral body flattening was more striking at age 22 months, and at 33 months the vertebral bodies showed rectangular shapes with greater relative flattening (Fig. 3C). There was some anterior vertebral notching. The disk spaces were relatively widened. The neural arches also were sclerotic. A pars defect, in keeping with an OPT,(24) was observed at L4, but healed by 33 months.Appendicular skeletonExtremities The overall pattern was one of sclerosis focused primarily on the metaphyses that also were expanded (Fig. 4).Open in a separate windowFig. 4Appendicular skeleton. At 22 (A) and 33 (B) months of age the lower extremities, and at 7 (C) and 33 (D) months of age the left upper extremity, show sclerosis greatest at the proximal and distal ends of the bones and metaphyseal expansion from tubulation failure greatest in the distal femurs. (E) There are transverse metaphyseal bands of osteosclerosis, as seen in some forms of OPT (arrows). A healing fracture featuring focal bone expansion and periosteal new bone formation is present in the distal left femur at the junction of the sclerotic and nonsclerotic bone (arrows) (A). There is a slight irregularity of the physeal plates and adjacent provisional zones of calcification (C, D, E). Sclerosis in the upper extremities has increased at the ends of the bones between 7 (C) and 33 (D) months of age. In the leg bones and distal right femur, the junction between the sclerotic and nonsclerotic bone is sharp, but not well defined in the proximal femurs or distal left femur. The junction between the sclerotic and nonsclerotic bone can have a transverse or oblique shape.Tubulation failure was best seen in the distal femurs (Fig. 4A, B). By 33 months, this expansion extended almost to the midshafts. The expansion, which was greatest in the metaphyses, extended into the diaphyses and from sclerotic bone into nonsclerotic bone. Between the ages of 7 and 43 months, the dense sclerosis at the ends of the bones increased appreciably in length. Some areas of metaphyseal sclerosis were sharply demarcated at the diaphysis, whereas others appeared to fade into more normal-appearing bone. There was a healing fracture in the distal left femur at the junction of the sclerotic and nonsclerotic bone, shown by periosteal new bone formation and focal expansion of the bone (Fig. 4A). The rate at which extremity fractures healed was delayed. Fractures continued to occur. Transverse lucent areas that were small and faint were seen in the dense sclerosis of some metaphyses, such as the distal femur (Fig. 4E).In the proximal end of the humeral metaphyses and both distal radii and ulnas, between 7 and 33 months of age, there were well-demarcated areas of dense sclerosis that increased in length (Fig. 4C, D). The junctions between the sclerotic and nonsclerotic bone were varied, being transverse, oblique, and rounded. The cortices in the midshafts were thickened. The proximal humeri and the distal radii and ulnas showed best the irregularity of the provisional zones of calcification at the physeal plates (Fig. 4C, D). The epiphyses were sclerotic.Hands and feet Sclerosis was greatest in the metacarpals, metatarsals, and distal radii and ulnas and often most in their metaphyses (Fig. 5). Nevertheless, metacarpals and metatarsals contained nonsclerotic areas. Some phalanges were almost entirely sclerotic, whereas others appeared normal. In some, metaphyseal sclerosis terminated abruptly into more normal-appearing bone, and then into a mixed sclerosis pattern. The cuboids had focal sclerosis, as did some other tarsal bones, but the carpals and distal radial epiphyses had no sclerosis. The distal radii and ulnas showed irregular physeal plates, provisional zones of calcification, and broad undertubulated, densely sclerotic metaphyses where the sclerosis terminated abruptly into more normal-appearing bone (Fig. 5D). In addition, the sclerosis ended in a rounded or oblique shape. The second metacarpals had accessory epiphyses. The bone age at 22 months was delayed to 18 months. At 33 months, the metaphyses of the phalanges were more sclerotic, as were the epiphyses.At age 3 years, DXA L1–L4 spine BMD Z-score was +8.1.Fractures Fractures would occur through radiographically dense bone (Fig. 5D) or at the junction of osteosclerosis and less remarkable bone (Fig. 5E) and were slow to heal.Histopathologic findingsTransapophyseal iliac crest biopsy showed a disordered growth plate consistent with the findings in some appendicular radiographs (see above). Specifically, the physis had zones of chondrocytes with normal proliferative and hypertrophic morphologies (Fig. 6A), but the cells were not arranged in their usual columnar orientation. An area deep to the growth plate, which at the patient's age is normally composed of trabecular bone with modest amounts of centrally located cartilage remnants, consisted almost entirely of calcified cartilage, as shown by H&E, toluidine blue, and von Kossa stains (Fig. 6,B–D). There was replacement of trabecular bone by calcified cartilage and diminished numbers of hematopoietic cells. Fluorescence imaging of unstained sections (not shown) demonstrated diffuse uptake of demeclocycline in areas of calcified cartilage near the growth plate, but there were no linear single or double labels to indicate normal bone formation. The area between the calcified cartilage contained primarily mesenchymal cells (ie, osteoblasts and fibroblasts), with only rare hematopoietic cells. Osteoclasts were not identified morphologically or by TRAP staining (Fig. 6E). Overall, the findings of increased amounts of cartilage, decreased hematopoietic marrow, and an absence of identifiable osteoclasts were consistent with a form of “osteoclast-poor” OPT.Open in a separate windowFig. 6Osteoclast-poor osteopetrosis. Sections of the nondecalcified transapophyseal iliac crest biopsy at age 25 months show a growth plate (A) containing proliferative (Pr) and hypertrophic (Hy) chondrocytes that lack the normal columnar orientation along the axis perpendicular to the bone surface. In this toluidine-stained section, the bone surface is toward the left, and the chondrocytes should be arranged in rows from left to right. Scale bar = 200 µm. H&E-stained (B) and toluidine blue–stained (C) sections demonstrate disorganized trabeculae composed primarily of cartilage (purple in both stains), with only minimal bone matrix (pink in B, pale gray in C). A von Kossa stain (D) shows that the cartilage in these “trabecular” areas is calcified. TRAP staining (E) fails to identify any osteoclasts. Scale bar (B–E) = 100 µm.Family studiesSkeletal survey of the parents revealed no significant abnormalities. The mother, 5 feet, 1 inch tall, had brachycephaly and minimal degenerative changes of L5. The father, a 5 foot, 11 inch postman, had brachycephaly, spondylolysis of L5, anterior wedging of T12, and multiple Schmorl's nodes in his lumbar and thoracic spine. The patient's two brothers and three paternal cousins had several radiographs taken elsewhere following trauma that were essentially unremarkable.DXA of the mother showed BMD Z-scores at the upper half of the age-matched normal range for the L1–L4 spine, total hip, wrist, and whole body. The father's values were average for gender and age.Studies of mineral homeostasis were unremarkable for the mother and father. Respectively, serum TRACP-5b was 3 and 2 U/L (normal 2 to 4 U/L for females and 3 to 5 U/L for males), osteocalcin 8 and 5 ng/mL (normal 3 to 14 ng/mL), and BAP 16 and 15 U/L (normal 12 to 31 U/L for females and 15 to 41 U/L for males). Urine free DPD was 5.0 and 2.9 nmol DPD/mmol of Cr (normal 3.0 to 7.4 nmol/mmol for females and 2.3 to 5.4 nmol/mmol for males). The father's profile suggested low-normal remodeling.Mutation analysesMutation analysis of the patient's CLCN7, TCIRG1, and OSTM genes was notable only for a previously unreported heterozygous missense variation in CLCN7 (p.T198M). This seemed noncontributory to the DSS because it occurred in the healthy father and would affect an amino acid not conserved across species.(25)No mutations were detected in TNFRSF11A, TNFRSF11B, CSF1, CSF1R, or ANKH. However, in TNFSF11, encoding RANKL, we found a heterozygous single-base alteration predicting an amino acid change (c.107C > G, p.Pro36Arg). This missense variation also was identified in the patient's father, but not in her mother, and was not listed in dbSNP,(26) nor had it been reported in several association studies of RANKL polymorphisms.(27,28) However, we did find it (heterozygous) in 4 of 67 patients [4 of 134 alleles (3.0%)] with a variety of metabolic/dysplastic bone diseases (see “Discussion”) and homozygous in a man with an unusual, severe variant of Camurati-Engelmann disease.(29) Nevertheless, using this candidate gene approach, we did not identify a causal gene defect for the DSS that we presume was inherited in this family as an autosomal recessive trait. Microarray-based genomic copy-number analysis of the patient and parents also was unrevealing.TreatmentLowering excessive dietary calcium was advised to ensure that circulating PTH levels were not suppressed and thereby impairing bone resorption. Following her second inpatient evaluation, reduction from approximately 950 to 500 mg calcium (RDA) was recommended, but a food record suggested intake of 775 mg. Thereafter, even small changes seemed to explain wide swings in serum PTH levels (Table 1), in keeping with severe OPT.(30)No recommendation was made for bone marrow transplantation (see “Discussion”). Teriparatide was not prescribed, and PTH seemed ineffective in a previous patient.(6) Particular attention was paid to possible surgical intervention if narrowed cranial foramina

Dysosteosclerosis (DSS), an extremely rare dense bone disease, features short stature and fractures and sometimes optic atrophy, cranial nerve palsy, developmental delay, and failure of tooth eruption in infancy or early childhood consistent with osteopetrosis (OPT). Bone histology during childhood shows unresorbed primary spongiosa from deficient osteoclast action. Additionally, there is remarkable progressive flattening of all vertebrae and, by adolescence, paradoxical metaphyseal osteopenia with thin cortical bone. Reports of consanguinity indicate autosomal recessive inheritance, yet more affected males than females suggest X-linked recessive inheritance. We investigated a nonconsanguineous girl with DSS. Osteosclerosis was discovered at age 7 months. Our studies, spanning ages 11 to 44 months, showed weight at approximately 50th percentile, and length diminishing from approximately 30th percentile to –2.3 SD. Head circumference was +4 SD. The patient had frontal bossing, blue sclera, normal teeth, genu valgum, and unremarkable joints. Radiographs showed orbital and facial sclerosis, basilar thickening, bone-in-bone appearance of the pelvis, sclerotic long bone ends, and fractures of ribs and extremities. Progressive metaphyseal widening occurred as vertebrae changed from ovoid to flattened and became beaked anteriorly. A hemogram was normal. Consistent with OPT, serum parathyroid hormone (PTH) concentrations reflected dietary calcium levels. Serum bone alkaline phosphatase, osteocalcin, and TRACP-5b were subnormal. The iliac crest contained excessive primary spongiosa and no osteoclasts. No mutations were identified in the splice sites or exons for the genes encoding chloride channel 7, T-cell immune regulator 1, OPT-associated transmembrane protein 1, and monocyte colony-stimulating factor (M-CSF) and its receptor C-FMS, ANKH, OPG, RANK, and RANKL. Genomic copy-number microarray was unrevealing. Hence, DSS is a distinctive OPT of unknown etiology featuring osteoclast deficiency during early childhood. How osteopenia follows is an enigma of human skeletal pathobiology. © 2010 American Society for Bone and Mineral Research.Keywords: bone remodeling, erlenmeyer flask deformity, osteoclast, osteosclerosis, RANKL, skeletal dysplasiaGo to:IntroductionDysosteosclerosis (DSS; OMIM % 224300)(1) refers to an extremely rare disorder characterized in 1968 by Spranger and colleagues(2) that presents in infancy or early childhood and features diffuse osteosclerosis with expanded ends of the tubular bones and platyspondyly. In keeping with a type of osteopetrosis (OPT),(3) bone histology during childhood shows unresorbed primary spongiosa and therefore deficient osteoclast action.(4,5) The skeleton is dense but brittle. Patients develop short stature and fractures and can suffer optic atrophy and sometimes cranial nerve palsy, developmental delay, and failure of tooth eruption,(1,6–9) in keeping with “marble bone disease.”(3) MRI has suggested delayed myelination in the brain,(10) seizures have occurred,(11) and various skin lesions have been reported.(11–14) Consanguinity(2,6,12,13,15) and a greater number of affected males (including in two generations) than females(11) suggest that DSS can be either an autosomal recessive trait(1,15) or an X-linked recessive trait,(11) but the genetic defect(s) is not known.(1)The clinical course and prognosis of DSS are not well understood. In 1978, Houston and colleagues(6) reported a 15-year-old boy whose iliac crest histology at 20 months of age revealed unresorbed primary spongiosa consistent with OPT. Subsequently, however, metaphyseal widening and sclerosis of early childhood changed to greater metaphyseal expansion (extending into the diaphysis) with remarkable osteopenia and cortical thinning by adolescence. In 2008, his follow-up at age 44 years was reported by Lemire and Wiebe,(16) who documented persistence of the widened proximal and distal ends of the long tubular bones with central diaphyses that were not expanded and showed cortical thickening.Here, we describe our clinical, biochemical, radiologic, histopathologic, and genetic investigation of a young girl with DSS and review the pertinent literature.Go to:Materials and MethodsInformed written consent for all studies was obtained as approved by the Human Research Protection Office, Washington University School of Medicine (St Louis, MO, USA).Case ReportThis 44-month-old American girl was admitted to our research center at ages 11, 23, 25, and 35 months. Additional outpatient serum and urine specimens were obtained at ages 27, 30, and 44 months.Medical historyThe patient was born vaginally at 38 weeks' gestation to a 37-year-old G3, P3, A0 woman. Birth weight was low for gestational age at 5 pounds, 14 ounces, but length 19½ inches.At 5 months of age, computed tomography (CT) for rapidly increasing head circumference (25th percentile at birth, 97th percentile at age 5 months, despite growth at 50th percentile for both length and weight) reportedly showed prominent ventricles, sulci, and subarachnoid spaces but no space-occupying lesion or hemorrhage.At 7 months of age, a radiograph (see below) for posttraumatic arm pain prompted a skeletal survey showing metaphyseal sclerosis that was attributed to “rickets.” Vitamin D was prescribed but not taken.From birth to 8 months of age, the patient had recurrent ear and skin infections and abdominal discomfort believed from colic.Between 8 and 11 months of age, several episodes of joint-centered limb pain were thought to be from dislocations. On referral to us at age 11 months, psychomotor development was on time. Tooth eruption was not delayed.At age 20 months, the patient fractured her distal left femur during a fall. Healing was delayed in a full splint.Ophthalmologic exam was normal at age 28 months. Audiometry had not been performed, but she seemed to hear well, had clear speech, and was bilingual.At age 30 months, she fractured through dense metaphyseal bone in her right radius and ulna during a fall. Reportedly, healing was slow (casted for 3 months). Parental recollection of further breaks was negative, but radiographs showed previous fractures. Morning stiffness in her lower limbs lasted 2 to 3 hours each day. She had chronic nasal “stuffiness.” Several episodes of erythematous papules involved her lower abdomen and were always on at least one area of her body but usually her belly, legs, or arms. Episodes lasted 3 to 4 days, were not associated with fever, and did not hurt or itch.At 36 months, she fell and fractured her distal right femur, where dense metaphyseal and more normal-appearing diaphyseal bone joined. Healing was delayed in a long-leg cast for 3 months.She received acetaminophen for nighttime fussiness but no multivitamins or supplements. She drank filtered municipal tap water. There were no factories near her home, but lead paint had been removed from a nearby house.At age 44 months (3⅔ years), interval history revealed a fractured left femur from a fall onto a carpeted floor with delayed healing while casted. Subsequently, however, the patient could run. There was generalized morning stiffness and painful ribs.Her healthy brothers were 6 and 8 years old. The family history was negative for skeletal disease or parental consanguinity. The father was of Turkish descent. The mother's family emigrated to Turkey from Greece 70 years previously.Physical findingsWhen examined for diagnosis at age 11 months, the patient had a high forehead, frontal bossing, blue sclerae, a flattened nasal bridge, downturned mouth (Fig. 1A), a narrow and high-arched palate, and a wide palatine ridge with nine normal-appearing teeth, indicating no delay in tooth eruption. Joint range of motion was full. A dermatosis involved her lower abdomen (Fig. 1B).Open in a separate windowFig. 1Patient. (A) At age 44 months, there is frontal bossing, a small midface, low-set ears, and blue sclerae. (B) At age 3 years, the recurring dermatosis (arrow) is present on the lower abdomen.At age 22 months, she had no palpable cranial sutures, bony tenderness, or genu valgum. Although 50th percentile for weight and length at age 5 months, between 11 and 35 months of age she maintained her weight at approximately 30th percentile of normal, but length decreased from approximately 30th percentile to below the growth curve (Z-score = –2.5). The upper/lower body segment Z-score increased to greater than 2.0. Arm span was less than 5th percentile. Head circumference Z-score decreased from +4.3 to +3.6. Genu valgum was present radiographically as early as 22 months of age.At age 44 months, weight was approximately 59th percentile, and length was –2.3 SD. Arm span versus height had progressed initially along the 50th percentile, but arm span measurements now showed that her arms were disproportionately short.(23)Biochemical testingFasting blood was used for all in-patient studies. Biochemical profiles were obtained from our Dade Behring Dimension Xpand instrument (Siemens Health Care Diagnostics, Inc., Los Angeles, CA, USA) using the reference ranges provided by the manufacturer. However, for all key findings, results were contrasted with values from fasting sera obtained in 2006 and 2007 from 34 healthy children and 24 healthy adults.(17)At age 11 months, the patient was given a diet containing approximately 400 mg of calcium per day to match her home intake estimated from a 7-day food record [Recommended Daily Allowance (RDA) for age = 270 mg]. At age 23 months, intake was approximately 950 mg. At age 35 months, we reduced her calcium intake to 775 mg daily (RDA of 500 mg recommended) to prevent suppressed serum parathyroid hormone (PTH) levels (see “Results”).To explore whether DSS is an OPT, we assayed serum levels of the brain isoenzyme of creatine kinase, BB-CK (Kit K20, Sebia, Norcorss, GA, USA),(18) tartrate-resistant acid phosphatase (TRACP-5b, Kit 8033, Quidel, Los Angeles, CA, USA),(17) and lactate dehydrogenase (LDH) and aspartate aminotransferase (AST).(17)Skeletal remodeling was assessed by quantitating serum bone-specific alkaline phosphatase (BAP) using ELISA (Quidel Corp., San Diego, CA, USA) and osteocalcin (Kit LKON1, Siemens Health Care Diagnostics, Inc.) and urine free deoxypyridinoline (DPD, Immulite 1000 Pyrilinks, D Kit, Siemens Medical Solutions Diagnostics, Ltd., Lianberis, Gwynedd, UK).Radiologic studiesAll radiologic studies, including images obtained elsewhere, were reviewed to assess the nature and evolution of the patient's skeletal disease, including fractures and their rate of healing. Dual-energy X-ray absorptiometry (DXA) of the patient and her parents involved a Hologic QDR 4500-A instrument (Hologic Corp., Waltham, MA, USA).Histopathologic studiesAt 2 years of age, two 3-day oral courses of demeclocycline separated by 10 days were ingested. Two days after the last dose, a transapophyseal iliac crest biopsy was performed using a Trap-Lok needle (Medical Device Technologies; Gainesville, FL, USA). The specimen was fixed in 70% ethanol and 30% water, embedded without decalcification in methyl methacrylate, and cut into 5-µm sections for staining via the hematoxylin and eosin (H&E), toluidine blue, and von Kossa methods. Unstained 10-µm sections were used for fluorescence microscopy.Mutation analysesThree genes (exons and mRNA splice sites) associated with OPT [chloride channel 7 (CLCN7), T-cell immune regulator 1 (TCIRG1), and OPT-associated transmembrane protein 1 (OSTM1)] had been studied at Connective Tissue Gene Tests (Allentown, PA, USA). We purified genomic DNA from blood leukocytes using the Gentra Puregene DNA extraction kit (Invitrogen, Carlsbad, CA, USA) and examined all coding exons and adjacent mRNA splice sites of TNFRSF11A encoding RANK, TNFRSF11B encoding OPG, and TNFSF11 encoding RANK ligand (RANKL) by PCR and sequencing in both directions using published methods and primers.(19–21) Because few osteoclasts were noted on our patient's iliac crest biopsy, the CSF1 and CSF1R genes encoding monocyte colony-stimulating factor (M-CSF) and its receptor C-FMS, respectively, were examined using primers and conditions that we developed (available on request). Furthermore, her bone modeling disturbances and the literature report of resolution of long bone osteosclerosis in DSS(6,16) led us to study the ANKH gene deactivated in craniometaphyseal dysplasia.(22) DNA sequences were evaluated using AlignX software (Vector NTI, Invitrogen) and by inspecting electropherograms.Genomic DNA copy-number microarray analysis of the proband and parents was performed using the Affymetrix SNP 6.0 chip at the Laboratory for Clinical Genomics, Washington University School of Medicine (St Louis, MO, USA) and the Partek Genomics Suite (Partek, St Louis, MO, USA).Literature reviewWe assessed all previous reports of DSS using original publications obtained from the Becker Library of Medicine, Washington University School of Medicine (St Louis, MO, USA), especially the radiographic reproductions.Go to:ResultsBiochemical testingSerum ionized calcium was highest, 5.3 mg/dL (normal 4.7 to 5.3 mg/dL), during the first admission at age 11 months (Table 1), when her estimated dietary intake of calcium significantly exceeded the RDA. With less dietary calcium, serum generally contained lower total and ionized calcium concentrations, but PTH levels fluctuated considerably over ages 11 to 44 months, ranging from 7 to 120 pg/mL (normal 4 to 52 pg/mL; Table 1). Calcium in timed urine collections ranged from 196 to 241 mg/g of creatinine (Cr).Table 1Longitudinal Biochemical FindingsSerumTotal CaCa2+PTHPiALPBAPOsteocalcinTRACP-5bAge (mos.)(normal 8.9–10.2 mg/dL)(normal 4.7–5.3 mg/dL)(normal 4–52 pg/mL)(normal 3.4–5.9 mg/dL)(normal 157–477 IU/L)(normal 88–247 U/L)(normal 37–119 ng/mL)(normal 13–23 U/L)1110.15.3344.7213—982310.25.2335.0162—2—259.14.91204.52108512227*10.1—75.2125———30*10.0—384.6140———359.95.0624.21364112344*9.85.0224.41766363Open in a separate window— = not done.*= non-fasting outpatient serum.Serum 25-hydroxyvitamin D [25(OH)D] was 28 ng/mL (normal 32 to 100 ng/mL) (Lab Corp., Burlington, NC, USA), with normal levels of magnesium and inorganic phosphate.Serum alkaline phosphatase (ALP) decreased to as low as 125 IU/L (normal 157 to 477 IU/L) and was intermittently subnormal for age, especially in view of recurrent fractures. Serum BAP, osteocalcin, and TRACP-5b levels all were consistently low (Table 1). Urine free DPD was 18 nmol/mmol of Cr (23 months) and 14 nmol/mmol of Cr (44 months) (normal 7 to 34 nmol/mmol).Serum BB-CK, LDH, and AST levels were unremarkable, unlike in some OPTs, where elevated values can occur.(17,18) Urine did not contain heavy metals. Serum was negative for lead (Legacy Laboratory Services, Portland, OR, USA).At ages 23, 35, and 44 months, complete blood counts were normal, including monocytes. Erythrocyte sedimentation rate seemed elevated for a child at 18, 15, and 24 mm/h, respectively (normal 0 to 20 mm/h for young adults), and serum C-reactive protein was consistently normal. Serum contained elevated IgG of 1416 mg/dL (normal 453 to 916 mg/dL) and slightly high IgA of 111 mg/dL (normal 20 to 100 mg/dL) (Lab Corp.).Radiologic findingsRadiographic skeletal surveys were available from ages 7, 11, 22, and 33 months. Right hand and wrist films were available at 30 months, left knee at 42 months, and right knee at 47 months.A generalized skeletal disorder was revealed, where every imaged bone was abnormal (Figs. 2 through ​through5).5). Prominent findings included basilar sclerosis of the skull and modeling abnormalities (expansion) in the major long bones resembling OPT but also ovoid vertebral bodies characteristic of DSS. The sclerosis must have occurred postnatally because relatively normal iliac bones had formed in utero that subsequently became sclerotic peripherally. The changes per anatomic site are described below.Open in a separate windowFig. 2Skull. (A) At age 22 months, anteroposterior view of the skull shows sclerosis that affects the periorbital region and skull base. The sinuses are opacified. (B) At age 33 months, the neurocranium appears large on lateral view, with diffuse sclerosis greatest at the base. The midface is small, and the mandibular angle is obtuse.Open in a separate windowFig. 5Hand and foot radiographs. At 22 (A) and 33 (B) months of age the left hand and at 22 (C) months of age the left foot show diffuse sclerosis greatest in the metaphyses of the metacarpals and metatarsals and distal radius and ulna. The amount of sclerosis in the proximal phalanges varies from much to little and is greatest in the metaphyseal regions. The tarsals have some sclerosis greatest in the cuboids. Pathologic fractures are shown through sclerotic bone in the distal radius (arrow) and ulna in the anteroposterior projection (D) at 30 months and near nonsclerotic bone in the right femur (arrows) at 47 months (E).Axial skeletonSkull The neurocranium at ages 22 and 33 months was large, resulting in craniofacial disproportion (Fig. 2). The frontal bone bulged anteriorly; the midface was small. The posterior fossa appeared prominent. Sutures were normal. Closure of the anterior fontanelle seemed delayed (open at 22 months; virtually closed at 33 months). The ascending ramus of the mandible was smaller than expected, and the mandibular angle was obtuse. Some teeth were malformed or abnormally oriented. The skull base was sclerotic, increasing between 7 to 33 months. Periorbital sclerosis was especially prominent along with sclerosis of the greater and lesser wings of the sphenoid and the posterior portion of the sphenoid body. Maxillary and ethmoid sinuses were opaque at ages 22 and 33 months, and the mastoid air cells were small.Head CT At age 5 months, CT showed a sclerotic skull base but no hydrocephalus or intracranial calcifications. The sulci were slightly enlarged. There was a suggestion of abnormal myelination. The optic foramina, middle ears, and internal auditory canals appeared grossly normal, but detailed optic foramina and petrous bone imaging was not obtained. Multiple follow-up CTs elsewhere were essentially unchanged.Chest Chest shape was normal, but the ribs were broad and diffusely sclerotic with multiple lower rib fracture deformities at age 7 months. By age 22 months, the ribs were more sclerotic, the medullary cavities were delayed (because they were just beginning to form), and there were multiple healed fractures that had become more numerous in the fifth through the eleventh ribs bilaterally. The ribs were shortened with a modeling error anteriorly, and wide anteriorly and laterally (Fig. 3A). Fractures had occurred anteriorly, and laterally and were not remodeling. At 33 months, the ribs were still sclerotic and widened, and they had more recent fractures and multiple fracture deformities but were showing further medullary cavity development. The clavicles were sclerotic, and their medial ends widened at ages 22 and 33 months. The sternal segments were broad and sclerotic, with a central lucency, as were the scapulae. The heart and lungs appeared normal.Open in a separate windowFig. 3Spine. This anteroposterior projection (A) at 22 months shows sclerotic vertebrae, dense ribs with healing fractures, sclerotic medial clavicles, and bone-in-bone appearance in the ilium as seen in some forms of OPT. Between 7 (B) and 33 (C) months of age, the lateral spine shows sclerosis and flattening of vertebral bodies that increased with age. The vertebral bodies have a pointed configuration anteriorly (greatest at 7 months). Sclerosis affects the entire vertebra.Pelvis Sclerosis was accompanied by a normal-appearing central portion of the iliac bones causing a “bone in bone” appearance at age 22 months (Fig. 3A), and even more apparent at 33 months. The upper iliac borders were irregular. The ischial and pubic bones were sclerotic. The femoral heads and metaphyses were sclerotic, and the necks were wide with failure of modeling.Spine The vertebral bodies were sclerotic, with the cranial and caudal ends appearing “sandwich-like,” and their anterior margins were pointed anteriorly. Compared with age 7 months, when the lumbar vertebrae had a more oval shape (Fig. 3B), vertebral body flattening was more striking at age 22 months, and at 33 months the vertebral bodies showed rectangular shapes with greater relative flattening (Fig. 3C). There was some anterior vertebral notching. The disk spaces were relatively widened. The neural arches also were sclerotic. A pars defect, in keeping with an OPT,(24) was observed at L4, but healed by 33 months.Appendicular skeletonExtremities The overall pattern was one of sclerosis focused primarily on the metaphyses that also were expanded (Fig. 4).Open in a separate windowFig. 4Appendicular skeleton. At 22 (A) and 33 (B) months of age the lower extremities, and at 7 (C) and 33 (D) months of age the left upper extremity, show sclerosis greatest at the proximal and distal ends of the bones and metaphyseal expansion from tubulation failure greatest in the distal femurs. (E) There are transverse metaphyseal bands of osteosclerosis, as seen in some forms of OPT (arrows). A healing fracture featuring focal bone expansion and periosteal new bone formation is present in the distal left femur at the junction of the sclerotic and nonsclerotic bone (arrows) (A). There is a slight irregularity of the physeal plates and adjacent provisional zones of calcification (C, D, E). Sclerosis in the upper extremities has increased at the ends of the bones between 7 (C) and 33 (D) months of age. In the leg bones and distal right femur, the junction between the sclerotic and nonsclerotic bone is sharp, but not well defined in the proximal femurs or distal left femur. The junction between the sclerotic and nonsclerotic bone can have a transverse or oblique shape.Tubulation failure was best seen in the distal femurs (Fig. 4A, B). By 33 months, this expansion extended almost to the midshafts. The expansion, which was greatest in the metaphyses, extended into the diaphyses and from sclerotic bone into nonsclerotic bone. Between the ages of 7 and 43 months, the dense sclerosis at the ends of the bones increased appreciably in length. Some areas of metaphyseal sclerosis were sharply demarcated at the diaphysis, whereas others appeared to fade into more normal-appearing bone. There was a healing fracture in the distal left femur at the junction of the sclerotic and nonsclerotic bone, shown by periosteal new bone formation and focal expansion of the bone (Fig. 4A). The rate at which extremity fractures healed was delayed. Fractures continued to occur. Transverse lucent areas that were small and faint were seen in the dense sclerosis of some metaphyses, such as the distal femur (Fig. 4E).In the proximal end of the humeral metaphyses and both distal radii and ulnas, between 7 and 33 months of age, there were well-demarcated areas of dense sclerosis that increased in length (Fig. 4C, D). The junctions between the sclerotic and nonsclerotic bone were varied, being transverse, oblique, and rounded. The cortices in the midshafts were thickened. The proximal humeri and the distal radii and ulnas showed best the irregularity of the provisional zones of calcification at the physeal plates (Fig. 4C, D). The epiphyses were sclerotic.Hands and feet Sclerosis was greatest in the metacarpals, metatarsals, and distal radii and ulnas and often most in their metaphyses (Fig. 5). Nevertheless, metacarpals and metatarsals contained nonsclerotic areas. Some phalanges were almost entirely sclerotic, whereas others appeared normal. In some, metaphyseal sclerosis terminated abruptly into more normal-appearing bone, and then into a mixed sclerosis pattern. The cuboids had focal sclerosis, as did some other tarsal bones, but the carpals and distal radial epiphyses had no sclerosis. The distal radii and ulnas showed irregular physeal plates, provisional zones of calcification, and broad undertubulated, densely sclerotic metaphyses where the sclerosis terminated abruptly into more normal-appearing bone (Fig. 5D). In addition, the sclerosis ended in a rounded or oblique shape. The second metacarpals had accessory epiphyses. The bone age at 22 months was delayed to 18 months. At 33 months, the metaphyses of the phalanges were more sclerotic, as were the epiphyses.At age 3 years, DXA L1–L4 spine BMD Z-score was +8.1.Fractures Fractures would occur through radiographically dense bone (Fig. 5D) or at the junction of osteosclerosis and less remarkable bone (Fig. 5E) and were slow to heal.Histopathologic findingsTransapophyseal iliac crest biopsy showed a disordered growth plate consistent with the findings in some appendicular radiographs (see above). Specifically, the physis had zones of chondrocytes with normal proliferative and hypertrophic morphologies (Fig. 6A), but the cells were not arranged in their usual columnar orientation. An area deep to the growth plate, which at the patient's age is normally composed of trabecular bone with modest amounts of centrally located cartilage remnants, consisted almost entirely of calcified cartilage, as shown by H&E, toluidine blue, and von Kossa stains (Fig. 6,B–D). There was replacement of trabecular bone by calcified cartilage and diminished numbers of hematopoietic cells. Fluorescence imaging of unstained sections (not shown) demonstrated diffuse uptake of demeclocycline in areas of calcified cartilage near the growth plate, but there were no linear single or double labels to indicate normal bone formation. The area between the calcified cartilage contained primarily mesenchymal cells (ie, osteoblasts and fibroblasts), with only rare hematopoietic cells. Osteoclasts were not identified morphologically or by TRAP staining (Fig. 6E). Overall, the findings of increased amounts of cartilage, decreased hematopoietic marrow, and an absence of identifiable osteoclasts were consistent with a form of “osteoclast-poor” OPT.Open in a separate windowFig. 6Osteoclast-poor osteopetrosis. Sections of the nondecalcified transapophyseal iliac crest biopsy at age 25 months show a growth plate (A) containing proliferative (Pr) and hypertrophic (Hy) chondrocytes that lack the normal columnar orientation along the axis perpendicular to the bone surface. In this toluidine-stained section, the bone surface is toward the left, and the chondrocytes should be arranged in rows from left to right. Scale bar = 200 µm. H&E-stained (B) and toluidine blue–stained (C) sections demonstrate disorganized trabeculae composed primarily of cartilage (purple in both stains), with only minimal bone matrix (pink in B, pale gray in C). A von Kossa stain (D) shows that the cartilage in these “trabecular” areas is calcified. TRAP staining (E) fails to identify any osteoclasts. Scale bar (B–E) = 100 µm.Family studiesSkeletal survey of the parents revealed no significant abnormalities. The mother, 5 feet, 1 inch tall, had brachycephaly and minimal degenerative changes of L5. The father, a 5 foot, 11 inch postman, had brachycephaly, spondylolysis of L5, anterior wedging of T12, and multiple Schmorl's nodes in his lumbar and thoracic spine. The patient's two brothers and three paternal cousins had several radiographs taken elsewhere following trauma that were essentially unremarkable.DXA of the mother showed BMD Z-scores at the upper half of the age-matched normal range for the L1–L4 spine, total hip, wrist, and whole body. The father's values were average for gender and age.Studies of mineral homeostasis were unremarkable for the mother and father. Respectively, serum TRACP-5b was 3 and 2 U/L (normal 2 to 4 U/L for females and 3 to 5 U/L for males), osteocalcin 8 and 5 ng/mL (normal 3 to 14 ng/mL), and BAP 16 and 15 U/L (normal 12 to 31 U/L for females and 15 to 41 U/L for males). Urine free DPD was 5.0 and 2.9 nmol DPD/mmol of Cr (normal 3.0 to 7.4 nmol/mmol for females and 2.3 to 5.4 nmol/mmol for males). The father's profile suggested low-normal remodeling.Mutation analysesMutation analysis of the patient's CLCN7, TCIRG1, and OSTM genes was notable only for a previously unreported heterozygous missense variation in CLCN7 (p.T198M). This seemed noncontributory to the DSS because it occurred in the healthy father and would affect an amino acid not conserved across species.(25)No mutations were detected in TNFRSF11A, TNFRSF11B, CSF1, CSF1R, or ANKH. However, in TNFSF11, encoding RANKL, we found a heterozygous single-base alteration predicting an amino acid change (c.107C > G, p.Pro36Arg). This missense variation also was identified in the patient's father, but not in her mother, and was not listed in dbSNP,(26) nor had it been reported in several association studies of RANKL polymorphisms.(27,28) However, we did find it (heterozygous) in 4 of 67 patients [4 of 134 alleles (3.0%)] with a variety of metabolic/dysplastic bone diseases (see “Discussion”) and homozygous in a man with an unusual, severe variant of Camurati-Engelmann disease.(29) Nevertheless, using this candidate gene approach, we did not identify a causal gene defect for the DSS that we presume was inherited in this family as an autosomal recessive trait. Microarray-based genomic copy-number analysis of the patient and parents also was unrevealing.TreatmentLowering excessive dietary calcium was advised to ensure that circulating PTH levels were not suppressed and thereby impairing bone resorption. Following her second inpatient evaluation, reduction from approximately 950 to 500 mg calcium (RDA) was recommended, but a food record suggested intake of 775 mg. Thereafter, even small changes seemed to explain wide swings in serum PTH levels (Table 1), in keeping with severe OPT.(30)No recommendation was made for bone marrow transplantation (see “Discussion”). Teriparatide was not prescribed, and PTH seemed ineffective in a previous patient.(6) Particular attention was paid to possible surgical intervention if narrowed cranial foramina

DNA from two Syrian families with H syndrome was analyzed through direct sequencing, and the expression of hENT3 in normal human skin was investigated by in situ hybridization and immunostaining. Results: We identified two novel mutations in the SLC29A3 gene: a homozygous splice site mutation IVS1+2T>G predicted to cause a splicing error, and a homozygous missense mutation c.1157G>A (p.R386Q) which substituted highly conserved amino acid residue in a transmembrane domain of hENT3. Furthermore, we demonstrate that hENT3 is expressed in histiocytes as well as in endothelium of blood and lymphatic vessels in normal human skin.

DNA from two Syrian families with H syndrome was analyzed through direct sequencing, and the expression of hENT3 in normal human skin was investigated by in situ hybridization and immunostaining. Results: We identified two novel mutations in the SLC29A3 gene: a homozygous splice site mutation IVS1+2T>G predicted to cause a splicing error, and a homozygous missense mutation c.1157G>A (p.R386Q) which substituted highly conserved amino acid residue in a transmembrane domain of hENT3. Furthermore, we demonstrate that hENT3 is expressed in histiocytes as well as in endothelium of blood and lymphatic vessels in normal human skin.

DNA from two Syrian families with H syndrome was analyzed through direct sequencing, and the expression of hENT3 in normal human skin was investigated by in situ hybridization and immunostaining. Results: We identified two novel mutations in the SLC29A3 gene: a homozygous splice site mutation IVS1+2T>G predicted to cause a splicing error, and a homozygous missense mutation c.1157G>A (p.R386Q) which substituted highly conserved amino acid residue in a transmembrane domain of hENT3. Furthermore, we demonstrate that hENT3 is expressed in histiocytes as well as in endothelium of blood and lymphatic vessels in normal human skin.

DNA from two Syrian families with H syndrome was analyzed through direct sequencing, and the expression of hENT3 in normal human skin was investigated by in situ hybridization and immunostaining. Results: We identified two novel mutations in the SLC29A3 gene: a homozygous splice site mutation IVS1+2T>G predicted to cause a splicing error, and a homozygous missense mutation c.1157G>A (p.R386Q) which substituted highly conserved amino acid residue in a transmembrane domain of hENT3. Furthermore, we demonstrate that hENT3 is expressed in histiocytes as well as in endothelium of blood and lymphatic vessels in normal human skin.