Search tips
Search criteria 


Logo of scirepAboutEditorial BoardFor AuthorsScientific Reports
Sci Rep. 2017; 7: 7786.
Published online 2017 August 10. doi:  10.1038/s41598-017-07341-2
PMCID: PMC5552841

Enpp1 is an anti-aging factor that regulates Klotho under phosphate overload conditions


Control of phosphate metabolism is crucial to regulate aging in mammals. Klotho is a well-known anti-aging factor that regulates phosphate metabolism: mice mutant or deficient in Klotho exhibit phenotypes resembling human aging. Here we show that ectonucleotide pyrophosphatase/phosphodiesterase 1 (Enpp1) is required for Klotho expression under phosphate overload conditions. Loss-of-function Enpp1 ttw/ttw mice under phosphate overload conditions exhibited phenotypes resembling human aging and Klotho mutants, such as short life span, arteriosclerosis and osteoporosis, with elevated serum 1,25(OH)2D3 levels. Enpp1 ttw/ttw mice also exhibited significantly reduced renal Klotho expression under phosphate overload conditions, and aging phenotypes in these mice were rescued by Klotho overexpression, a low vitamin D diet or vitamin D receptor knockout. These findings indicate that Enpp1 plays a crucial role in regulating aging via Klotho expression under phosphate overload conditions.


A fundamental question in human biology is what mechanisms underlie aging. Aging is reportedly defined as age-related deterioration of physiological functions necessary for survival and fertility1. Several phenotypes associated with aging have been identified, such as short life span, osteoporosis, arteriosclerosis, cancer and cataract development. To date, various mouse models showing premature aging syndromes have been established and characterized, such as Klotho (kl/kl)-2, SAM-3, ATM-4, p53-5, FoxOs-6, telomerase7-, and Fetuin A-deficient8 mice; however, none exhibits the full range of aging phenotypes.

Among models of aging, the kl/kl mouse is best known for showing premature aging phenotypes including short life span, osteoporosis and arteriosclerosis2. Klotho is a transmembrane protein that gives rise to a soluble form when cleaved at the cell surface. The transmembrane Klotho (αKlotho) forms a heterocomplex with fibroblast growth factor receptors (FGFRs) required for high affinity FGF23 binding and signaling911. FGF23 gain-of-function mutations reportedly underlie autosomal dominant hypophosphatemic rickets (ADHR)12, 13, evidence that FGF23 regulates phosphate metabolism. FGF23-deficient mice reportedly exhibit aging-related phenotypes similar to kl/kl mice14, suggesting that FGF23 and Klotho co-operate in regulating phosphate metabolism. Klotho and FGF23 are reported required to regulate circulating 1,25(OH)2D3 levels by suppressing expression of cyp27b1 in the kidney15, 16. Indeed, premature aging phenotypes seen in either kl/kl or FGF23-deficient mice are completely rescued by ablation of the Vitamin D receptor (VDR)1719. Loss-of-function mutations in either KLOTHO or FGF23 reportedly cause tumoral calcinosis in humans, a disease characterized by ectopic, vascular calcifications20, 21. FGF23 is mainly produced in bone5, and FGF23 secretion in bone is stimulated by 1,25(OH)2D3 and by increased extracellular phosphate, an activity regulated in a feedback loop between bone and kidney22, 23. In contrast, Klotho expression is downregulated by phosphate24 and reportedly regulated epigenetically25. However, mechanisms underlying Klotho regulation are unclear, and there are as yet no animal models resembling kl/kl mice that have been established by deleting factors regulating Klotho.

Ectonucleotide pyrophosphatase/phosphodiesterase 1 (Enpp1) is a single-pass transmembrane protein and major generator of extracellular pyrophosphate (PPi), which inhibits hydroxyapatite crystal deposition26. ENPP1 mutations cause autosomal recessive hypophosphatemic rickets (ARHR) or Generalized Arterial Calcification of Infancy (GACI) in humans2730. ARHR caused either by ENPP1 mutation or mutations in PHEX or DMP1 genes is marked by high circulating FGF23 levels31, 32. In contrast, GACI is a rare autosomal-recessive disorder characterized by calcification of diffuse vascular and periarticular soft-tissues, and few GACI patients survive the neonatal period33. ENPP1 mutations are also seen in patients with ossification of the posterior longitudinal ligament (OPLL), a disease characterized by ectopic ossification in a spinal ligament, although mechanisms are unknown3436. The Enpp1 mutants Enpp1 ttw/ttw or Enpp1 asj/asj reportedly exhibit OPLL- or GACI-like, respectively, phenotypes37, 38.

Here, we show that Enpp1 acts as an anti-aging factor under phosphate overload by regulating Klotho expression. Enpp1 ttw/ttw or genetically engineered Enpp1-deleted mice (Enpp1 Δ/Δ) exhibited premature aging phenotypes, such as short life span and arteriosclerosis, phenotypes resembling kl/kl mice and human aging, under phosphate overload. Klotho expression in kidney was significantly downregulated in Enpp1 ttw/ttw mice by phosphate overload, and premature aging phenotypes seen in Enpp1 ttw/ttw mice under overload conditions were completely rescued by VDR ablation. Thus, Enpp1-Klotho-VDR signals are required to prevent premature aging phenotypes, particularly under phosphate overload conditions.


Enpp1 mutation causes premature aging phenotypes under phosphate overload

Enpp1 mutations cause several human diseases, among them AHRH, GACI or OPLL1215. Enpp1 ttw/ttw mice are Enpp1 mutation mouse models that exhibit OPLL-like phenotypes38. We confirmed that at 8 weeks of age Enpp1 ttw/ttw mice show significantly lower serum phosphate (Pi) levels relative to controls, an outcome also seen in ARHR patients (Fig. 1a). Although not statistically significant, serum calcium levels were elevated in both WT or Enpp1 ttw/ttw mice by a high phosphate diet (HPD) (Fig. 1a). However, feeding 8-week old Enpp1 ttw/ttw mice a HPD did not elevate serum phosphate levels in as it did in wild-type (WT) mice (Fig. 1a). Ectopic calcification in the ear was previously demonstrated to be accelerated by feeding Enpp1 ttw/ttw mice a HPD39. We found that HPD worsened OPLL phenotypes, as shown by elevated ectopic calcification in the posterior longitudinal ligament (PLL) of Enpp1 ttw/ttw mice (Fig. 1b). Moreover, after phosphate overload, Enpp1 ttw/ttw mice showed body weight loss (Fig. 1c), became inactive and marantic, and died within three weeks (Fig. 1d). Enpp1 ttw/ttw mice fed a phosphate diet also exhibited ectopic calcification in aorta and kidney (Fig. 1e), atrophic skin (Fig. 1f) and osteoporotic reduced bone mass (Fig. 1g), all premature aging phenotypes seen in kl/kl mice and aging humans2.

Figure 1
A high phosphate diet promotes premature aging phenotypes in Enpp1 ttw/ttw mice. Eight-week-old wild-type and Enpp1 ttw/ttw mice were fed a normal (ND) or high phosphate (HPD) diet for two (ac, e– g) to twenty weeks (d). The following ...

Enpp1 ttw/ttw mice showed calcification in kidney and aorta at cellular levels (Supplementary Fig. 1a). Expression of runt-related transcription factor 2 (Runx2), a factor essential for osteoblastogenesis40, was significantly upregulated in aorta of Enpp1 ttw/ttw mice fed a HPD rather than a normal diet (Supplementary Fig. 1b), suggesting that ectopic ossification is likely due to trans-differentiation of cells into an osteoblastic lineage. Although, no differences were detected in senescence-associated beta-galactosidase (SA β-gal) staining, expression of p16, another aging-related factor, increased in kidney of Enpp1 ttw/ttw mice fed a HPD relative to that of Enpp1 ttw/ttw mice fed a ND or wild-type (WT) mice fed either diet (Supplementary Fig. 2).

We also found that serum levels of receptor activator of nuclear factor kappa B ligand (RANKL), a cytokine essential for osteoclastogenesis41, increased in Enpp1 ttw/ttw mice fed a HPD compared with Enpp1 ttw/ttw mice fed a ND or WT mice fed either diet (Supplementary Fig. 3a), but serum osteoprotegerin (OPG), a natural agonist of RANKL42, also increased (Supplementary Fig. 3b), and as a result, the serum RANKL/OPG ratio was comparable in Enpp1 ttw/ttw mice fed a HPD relative to Enpp1 ttw/ttw mice fed a ND or WT mice fed either diet (Supplementary Fig. 3c). However, osteoclast bone resorbing activity, as measured by serum CTx levels, was significantly elevated in Enpp1 ttw/ttw mice fed a HPD, suggesting that decreased bone mineral density is due at least in part to elevated osteoclast bone-resorption (Supplementary Fig. 3d).

Although statistically not significant, serum creatinine and blood urinary nitrogen (BUN) levels were elevated in Enpp1 ttw/ttw mice fed a HPD relative to Enpp1 ttw/ttw mice fed a ND or WT mice fed either diet, and urine volume was significantly decreased in Enpp1 ttw/ttw mice fed a HPD compared to WT mice fed a HPD diet (Supplementary Fig. 4a and b), suggesting that Enpp1 ttw/ttw mice fed a HPD undergo renal dysfunction. Urinary calcium was elevated by HPD in WT mice, a phenotype not seen in Enpp1 ttw/ttw mice fed a HPD (Supplementary Fig. 4c). In contrast, urinary phosphate and creatinine levels were elevated and downregulated, respectively, in both Enpp1 ttw/ttw and WT mice by HPD (Supplementary Fig. 4c).

In contrast, Hyp mice, a different model of hypophosphatemic rickets caused by mutation in the Phex gene, did not exhibit premature aging or OPLL phenotypes even when fed a HPD (Fig. 2). Eight-week-old Hyp mice were fed a HPD for two weeks, but since mice exhibited no obvious phenotypes, feeding of the HPD was extended six weeks longer (Fig. 2). HPD effectively elevated serum phosphate levels in these Hyp mice (data not shown), as previously described43, an effect not seen in Enpp1 ttw/ttw mice fed the same diet for two weeks (Fig. 1a). Unlike Enpp1 ttw/ttw mice, which exhibited visible aging phenotypes after two weeks of phosphate overload, Hyp mice were normal in appearance (Fig. 2a), nor did they exhibit lethality after eight weeks of phosphate overload (Fig. 2b). Soft X-ray images showed that Enpp1 ttw/ttw showed kyphosis by two weeks of HPD feeding, but Hyp mice did not become kyphotic even after eight weeks of phosphate overload (Fig. 2c). Hyp mice at eight weeks of phosphate overload did not show OPLL, renal or aortic calcification (Fig. 2d and e). Enpp1 expression in kidney was significantly higher in Hyp than in WT mice and was significantly upregulated by HPD in mice of either genotype (Fig. 2f). These results suggest that in mice, premature aging phenotypes associated with phosphate overload are not common to all cases of hypophosphatemic rickets mice but rather they require mutation in Enpp1.

Figure 2
Hyp mice fed a high phosphate diet do not exhibit aging phenotypes. Eight-week-old wild-type or Hyp mice were fed with normal (ND) or high phosphate (HPD) diet for eight weeks. Mice were then analyzed for: gross appearance (a); survival rate (each group; ...

Enpp1ttw/ttw mice under phosphate overload show reduced Klotho expression

Aging phenotypes seen in Enpp1 ttw/ttw mice fed a HPD, such as short lifespan and ectopic calcifications, resemble those seen in kl/kl mice, and kl/kl mice reportedly exhibit abnormal correlation between serum FGF23 and 1,25(OH)2D3 levels44. Moreover, both factors are known to regulate serum phosphate levels. Thus, we fed eight-week-old wild-type or mutant mice a normal or HPD for two weeks, since Enpp1 ttw/ttw mice fed a HPD died within three weeks of feeding (Fig. 1), and then analyzed both for serum FGF23 and active vitamin D3, 1,25(OH)2D3, levels. Both were upregulated significantly in Enpp1 ttw/ttw compared with control mice under phosphate overload (Fig. 3a and b). 1,25(OH)2D3 is derived from 25(OH)D3 due to hydroxylation by the enzyme cyp27b145. In accordance, serum 25(OH)D3 levels were significantly lower in Enpp1 ttw/ttw mice fed a HPD compared with those fed a normal diet or wild-type mice fed either diet (Fig. 3b). Serum PTH levels were also significantly elevated in WT or Enpp1 ttw/ttw mice fed a HDP (Fig. 3c). FGF23 is known to down-regulate 1,25(OH)2D3 14; thus it is unusual that FGF23 and 1,25(OH)2D3 levels would be concomitantly elevated, as seen in Enpp1 ttw/ttw mice. However, Klotho mutant mice reportedly show elevated levels of both FGF23 and 1,25(OH)2D3 in sera2, suggesting that Klotho expression is suppressed in Enpp1 ttw/ttw mice under phosphate overload. Furthermore, since aging phenotypes in Enpp1 ttw/ttw mice under phosphate overload phenocopy kl/kl mice, we next analyzed Klotho expression in Enpp1 ttw/ttw mouse kidney (Fig. 3d and Supplementary Fig. 5). As expected, renal Klotho mRNA and protein expression were significantly lower in Enpp1 ttw/ttw than in wild-type mice under phosphate overload conditions (Fig. 3d and Supplementary Fig. 5). Klotho reportedly antagonizes expression of the sodium-phosphate co-transporter NaPi-IIa, also called solute carrier family 34 (sodium phosphate), member 1 (Slc34a1)46. Indeed, we observed elevated NaPi-IIa expression inversely correlated with Klotho expression in Enpp1 ttw/ttw mice fed a HPD (Fig. 3d and Supplementary Fig. 5). Furthermore, Klotho is reportedly required to inhibit cyp27b1 47. We found that cyp27b1 expression in kidney was significantly upregulated in Enpp1 ttw/ttw mice fed a HPD compared with those fed a normal diet (Fig. 3d), potentially due to inhibited Klotho expression.

Figure 3
Dietary phosphate overload decreases Klotho expression in kidney of Enpp1 ttw/ttw mice. Eight-week-old wild-type and Enpp1 ttw/ttw mice were fed a ND or HPD for two weeks, and serum levels of FGF23 (a), 1,25(OH)2D3 (b, left panel), 25(OH)D3 (b, right ...

To confirm that aging phenotypes seen in Enpp1 ttw/ttw mice following phosphate overload are due to reduced Klotho expression, we crossed the Enpp1 ttw/ttw mice with Klotho-overexpressing transgenic (Klotho Tg) mice, an approach that reportedly rescues aging phenotypes in kl/kl mice2 (Supplementary Fig. 6). In resultant mice fed a HPD, shortened life span was rescued in part (Supplementary Fig. 6a), while ectopic calcification in aorta was significantly rescued compared to similarly fed Enpp1 ttw/ttw mice (Supplementary Fig. 6b). These results support the idea that under phosphate overload, decreased Klotho expression due to Enpp1 mutation promotes development of aging phenotypes in Enpp1 ttw/ttw mice. We also observed OPLL-like peri-vertebral bone phenotypes in kl/kl mice (Supplementary Fig. 7), as is seen in Enpp1 ttw/ttw mice (Fig. 1b), suggesting that the Enpp1-Klotho axis is required to prevent ectopic calcification and aging phenotypes under phosphate overload conditions.

Elevated vitamin D levels promote aging phenotypes in Enpp1ttw/ttw mice under phosphate overload

Serum 1,25(OH)2D3 levels were significantly elevated in Enpp1 ttw/ttw mice fed a HPD (Fig. 1a), and high 1,25(OH)2D3 levels reportedly cause premature aging phenotypes in kl/kl mice14. Thus, we asked whether elevated serum 1,25(OH)2D3 levels promote aging phenotypes in Enpp1 ttw/ttw mice under phosphate overload conditions (Fig. 4). To do so, we fed Enpp1 ttw/ttw mice a high phosphate/reduced vitamin D diet (HPLD) starting at 8 weeks of age (Fig. 4). Compared with animals fed a HPD only, serum phosphate and calcium levels were down-regulated in Enpp1 ttw/ttw mice fed a high phosphate/low vitamin D (HPLD) diet (Fig. 4a). Urinary calcium and creatinine levels were elevated in both WT and Enpp1 ttw/ttw mice fed a HPLD compared with those fed a HPD (Supplementary Fig. 8). Urinary phosphate levels were comparable in WT and Enpp1 ttw/ttw mice fed a HPLD (Supplementary Fig. 8). Serum 1,25(OH)2D3 and FGF23 levels were down-regulated in both WT or Enpp1 ttw/ttw mice fed a HPLD (Fig. 4b). Moreover, serum PTH levels were down- and up-regulated in WT and Enpp1 ttw/ttw mice, respectively, fed a HPLD (Fig. 4b). Reduced body weight, shortened life span, reduced bone mass and ectopic calcification in kidney and aorta, all seen in Enpp1 ttw/ttw mice fed a HPD, were all significantly rescued by vitamin D depletion, even under high phosphate loading (Fig. 4c–f). Inactivity and maranic phenotypes were prevented in Enpp1 ttw/ttw mice by depletion of vitamin D from a HPD, suggesting that premature aging phenotypes observed in Enpp1 ttw/ttw mice under phosphate overload require high vitamin D signals, as is the case with kl/kl mice. Reduced Klotho expression in kidney seen following phosphate overload in Enpp1 ttw/ttw mice was significantly rescued by vitamin D depletion (Fig. 4g), suggesting feedback between Klotho and vitamin D signals. Furthermore, correlated with elevated renal Klotho expression, cyp27b1 expression was significantly downregulated by vitamin D depletion under phosphate overload conditions in Enpp1 ttw/ttw mice (Fig. 4g).

Figure 4
A low vitamin D diet antagonizes aging phenotypes seen in Enpp1 ttw/ttw mice under phosphate overload. Eight-week-old wild-type and Enpp1 ttw/ttw mice were fed a ND, HPD or a high phosphate/low vitamin D diet (HPLD) for eight weeks (a, b, eh ...

Fetuin A activity antagonizes ectopic calcification, and circulating Fetuin A levels are down-regulated with age48. Thus, we performed ELISA analysis to measure Fetuin A levels in WT or Enpp1 ttw/ttw mice fed a HPD (Supplementary Fig. 9). Serum Fetuin A protein levels significantly decreased in response to a HPD in both genotypes and were equivalent in each (Supplementary Fig. 9). Although aging phenotypes such as short lifespan were rescued in response to a HPLD, low Fetuin A levels were not (Supplementary Fig. 9). Thus, aging phenotypes seen in Enpp1 ttw/ttw mice fed a HPD are more likely associated with vitamin D rather than with Fetuin A signaling.

Next we crossed Enpp1 ttw/ttw with VDR-deficient (VDR −/−) mice to yield Enpp1/VDR doubly-deficient (Enpp1 ttw/ttw/VDR −/−) mice (Fig. 5). VDR −/− mice die after weaning, but can survive beyond weaning if fed a high calcium diet49, 50. Thus, we fed Enpp1 ttw/ttw/VDR −/− mice a high phosphate/high calcium diet starting at 8-weeks of age. Others have reported that calcium and phosphate overloading worsens OPLL phenotypes in Enpp1 ttw/ttw mice39. However, ankylosis of the fore- and hindlimb seen in Enpp1 ttw/ttw mice was efficiently rescued in Enpp1 ttw/ttw/VDR −/− mice fed a high phosphate/high calcium diet (Fig. 5a). Reduced body weight, shortened life span, ectopic calcification in kidney and aorta and ossification in ligaments seen in Enpp1 ttw/ttw mice were all completely abrogated in Enpp1 ttw/ttw/VDR −/− mice even under phosphate overload and high calcium conditions (Fig. 5b–e). Moreover, Enpp1 ttw/ttw/VDR −/− mice were active and not maranic compared with Enpp1 ttw/ttw mice. In addition, Klotho expression in kidney was significantly high in Enpp1 ttw/ttw/VDR −/− relative to Enpp1 ttw/ttw mice under phosphate overload (Fig. 5f).

Figure 5
Aging phenotypes seen in Enpp1 ttw/ttw mice fed a high phosphate diet are abrogated by deletion of vitamin D receptor. Eight-week-old Enpp1 ttw/ttw or Enpp1 ttw/ttw/VDR −/− mice were fed a HPD for indicated periods (a and b) or eight weeks ...


How to achieve longevity is a fundamental concern for people living all over the world. The development of strategies to promote healthy aging requires understanding molecular mechanisms underlying regulation of aging, and various molecules have been identified as aging-related5155. Here, we show that under phosphate overload, Enpp1 is required for renal Klotho expression, and its activity leads to down-regulation of 1,25(OH)2D3 production by inhibiting cyp27b1 expression and is part of a crucial axis that regulates phosphate and vitamin D metabolism (Fig. 6a). Enpp1 is also required to suppress FGF23 production by osteocytes, and Enpp1 loss elevates serum FGF23 levels owing to increased serum 1,25(OH)2D3 (Fig. 6b). Thus Enpp1 serves as an upstream mediator of the active Klotho/vitamin D3/FGF23 axis to suppress aging phenotypes (Fig. 6).

Figure 6
A schematic showing regulation of aging phenotypes under phosphate overload. (a) Normal Enpp1 maintains Klotho expression in kidney under dietary phosphate overload. Elevated serum phosphate levels promote FGF23 production from bone, and resultant FGF23 ...

Previous studies demonstrate that Klotho plays a pivotal role in regulating aging2, stimulating a search for upstream regulators of Klotho5658. Promoter methylation has been demonstrated to restrict Klotho expression in the kidney25. However, the targeting of potential upstream regulators of Klotho has not yet yielded animal models that resemble kl/kl mice in terms of premature aging. Here, we demonstrate that mutation of Enpp1 phenocopies kl/kl mice under conditions of phosphate overload. How Enpp1 regulates Klotho expression in kidney is unclear, and this remains to be addressed. However, since Enpp1 is reportedly expressed in osteocytes59, Enpp1 expressed in those cells may regulate renal Klotho expression. Further studies are needed to define the regulatory system between bone and kidney in the context of aging. It is also possible that Klotho expression is inhibited by phosphate overload itself. Indeed, in our study Klotho expression was down-regulated by phosphate overload even in WT mice (Fig. 3d), although these changes were not statistically significant. Thus, low phosphate conditions caused by low vitamin D diet or VDR-deficiency likely inhibit Klotho down-regulation and contribute to reversal of aging phenotypes seen in Enpp1 ttw/ttw mice fed a HPD. However, since Klotho expression was lower in Enpp1 ttw/ttw than in WT mice, Enpp1 is likely required for Klotho expression. Nonetheless, VDR could be a therapeutic target to treat Enpp1-deficient patients.

Various Enpp1 gene mutations have been identified in humans, but phenotypes of patients harboring differing Enpp1 mutations vary from hypophosphatemia rickets or GACI to OPLL29, 30, 34. There are several types of hypophosphatemia rickets, among them XLH (Phex mutation), ARHR1 (DMP1 mutation) and ARHR2 (Enpp1 mutation). Currently, patients with any of these conditions are treated with vitamin D3 (calcitriol) and phosphorus supplementation. The effects of such treatment on ARHR2 patients reportedly varies, and some patients are reportedly worsened by treatment60. Here we propose that supplementation of patients harboring an Enpp1 mutation with vitamin D3 (calcitriol) and phosphorus is a potential contraindication, and that by contrast, inhibition of vitamin D signals should be considered.

Among Enpp1 mutations, GACI is associated with the most severe phenotypes, and patients exhibit calcification of the aorta, with a mortality rate of approximately 85% by the age of 6 months18. GACI phenotypes are reminiscent of Enpp1 ttw/ttw and Enpp1 Δ/Δ mice fed a HPD or of kl/kl mice. To date, patients with GACI are treated with bisphosphonate61, 62. In animal models, Enpp1-Fc is reportedly effective in blocking HPD-induced death in Enpp1 mutant mice63. OPLL is a disease characterized by ossification of the PLL of the spine, although its molecular pathogenesis is unclear, and no therapeutic drugs have been established. Several gene mutations are reportedly associated with OPLL64, and Enpp1 mutations are detected in some OPLL patients34.

Recent advances in aging research have revealed that excess dietary phosphate intake accelerates aging and renal dysfunction65. Klotho plays an essential role in regulating phosphate levels, and kl/kl mice exhibit various aging phenotypes seen in humans44. The Enpp1-Klotho/FGF23-VDR axis is considered crucial to regulate human aging by controlling circulating phosphate levels. Moreover, Klotho and Fetuin A, the latter a soluble protein produced in the liver, are reportedly required to regulate circulating phosphate levels, and Fetuin A-deficient mice exhibit systemic ectopic calcification8. Fetuin A reportedly forms “calciprotein particles” to inhibit precipitation of calcium and phosphate to prevent unwanted calcification66, and circulating Fetuin A levels are reportedly down-regulated with age48. Thus, aging and changes in phosphate levels are regulated in a complex manner, and further studies are needed to clarify mechanisms underlying their relationship.

Overall, our study sheds light on the pathogenesis of Enpp1 mutation-related disease. We conclude that inhibition of the vitamin D3-VDR pathway may be a better option to treat patients with Enpp1 mutations, although further clinical studies are required to test this strategy.


Mice and diets

The Enpp1 ttw/ttw mouse is a spontaneous mutant harboring a nonsense mutation in Enpp1 and first characterized in 1998 as an excellent model of ectopic ossification38, 67. We obtained Enpp1 wt/ttw heterozygotes from the Central Institute for Experimental Animals (Kawasaki, Japan) and mated them to obtain Enpp1 ttw/ttw homozygotes. Wild-type mice were obtained from Sankyo Lab Service (Tsuchiura, Ibaraki, Japan). Klotho-mutant (kl/kl)2, Klotho-overexpressing transgenic (Klotho Tg)68, and vitamin D receptor (VDR)-deficient (VDR −/−) mice were established previously69. Klotho Tg or VDR −/− mice were crossed with Enpp1 ttw/ttw mice to yield Enpp1 ttw/ttw/Klotho Tg or Enpp1 ttw/ttw VDR −/− mice, respectively. Mice were fed either a normal phosphate diet (1% phosphate, ND), a HPD (1.5–2% phosphate, HPD), a high phosphate/low vitamin D (HPLD) diet or a high phosphate/high calcium diet starting at eight weeks of age for at least two weeks or indicated periods. An HPLD contains 0 units/100 g Vitamin D units. Other diets contain 240 units/100 g. Components of the HPD and HPLD are shown in the Supplementary Table 1. All animal methods were carried out in accordance with the Guidelines of the Keio University animal care committee. All experimental protocols were approved by that committee.

Quantitative PCR Analysis

Total RNAs were isolated from kidney by TRIzol reagent (Invitrogen Corp.), and cDNA synthesis was done using oligo(dT) primers and reverse transcriptase (Wako Pure Chemicals Industries). Quantitative PCR was performed using SYBR Premix ExTaq II reagent and a DICE Thermal cycler (Takara Bio Inc., Otsu, Shiga, Japan). Samples were matched to a standard curve generated by amplifying serially diluted products using the same PCR reactions. β-actin (Actb) expression served as an internal control. Primers for Klotho, Cyp27b1 and Actb were as follows.







β-actin (Actb)-forward: 5′-TGAGAGGGAAATCGTGCGTGAC-3′

β-actin (Actb)-reverse: 5′-AAGAAGGAAGGCTGGAAAAGAG-3′

Western blotting

Harvested kidneys were homogenized in RIPA buffer (1% Tween 20, 0.1% SDS, 150 nM NaCl, 10 mM Tris-HCl (pH 7.4), 0.25 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml leupeptin, 1 mM Na3VO4, 5 mM NaF (Sigma)). Lysates were collected by centrifugation at 15,000 rpm at 4 °C for 10 min. Equivalent amounts of protein were separated by SDS/PAGE and transferred to a PVDF membrane (EMD Millipore Corporation). Proteins were detected using anti-Klotho (ab98111, abcam, Cambridge, UK), anti-NaPi-IIa (ab182099, abcam) and anti-Actin (A2066, Sigma) antibodies. Bands were visualized using ECL Western Blotting Detection Reagent (GE Healthcare, Uppsala, Sweden).

Analysis of skeletal morphology

Bone mineral density (BMD) of whole femurs was measured by Dual-energy X-ray absorptiometry (DEXA) using a DCS-600R system (Aloka Co. Ltd., Tokyo, Japan). Vertebral bones and surrounding tissues including intervertebral discs and the posterior longitudinal ligament were scanned by a micro-computed tomography (R_mCT2; Rigaku Corp., Tokyo, Japan) at 90 kV and 160 A. Two-dimensional regions of interest were created at the level of the cervical spine and Achilles tendon using TRI/3D-BON software (RATOC Co. Ltd., Tokyo, Japan).

Histopathological analysis

Kidney, aorta and skin from euthanized mice were fixed in 10% phosphate-buffered formalin, and embedded in paraffin. Tissues were sectioned and stained with hematoxylin and eosin (H&E) and using von Kossa methods. Slides were examined by light microscopy (BIOREVO, BZ-9000 (Keyence, Osaka, Japan)) for tissue mineralization. Relative calcification areas in kidney and aorta were calculated using microscopic image analysis software BZ-II analyzer (Keyence). Kidney sections were also stained using anti-p16 INK4A (10883-1-AP, Proteintech, Rosemont, IN, USA) or anti-Klotho (ab98111, abcam) followed by Alexa488-conjugated goat anti-rabbit IgG H&L (Alexa Fluor® (ab150077, abcam), and nuclei were stained with DAPI (Wako Pure Chemicals Industries, Osaka, Japan). Senescence-associated beta-galactosidase was stained using a SA-β-gal kit (#9860, Cell Signaling, Danvers, MA). Then, sections were observed under a fluorescence microscope (Keyence, Osaka, Japan).


Serum levels of FGF23 (full length, KAINOS Lab Inc., Tokyo, Japan), RANKL (R&D, systems, Inc., Minneapolis, MN), OPG (R&D) CTx (Immunodiagnostic Systems Limited, Boldon, UK), urinary Fetuin A (R&D) and α-Klotho (Immuno-Biological Laboratories Co, Ltd, Gunma, Japan) were measured by using an ELISA kit based on the manufacturers’ instructions.

Biochemical analyses

Peripheral blood was obtained from the postorbital vein. Serum was isolated by centrifugation at 6,000 rpm for 15 minutes at 4 °C and stored at −80 °C. 1,25(OH)2D3 and 25(OH)D3 levels in sera were measured using a 25OH-Vitamin D total-RIA-CT kit (DIAsource, Ottignies-Louvain-la-Neuve, Belgium) and an ECLIA kit (Cobas, Roche Diagnostics, Basel, Switzerland), respectively.

Statistical analysis

Data were analyzed using a two-tailed Student’s t-test. For all graphs, data are represented as means ± standard deviation (SD). A p-value less than 0.05 was considered statistically significant (*p < 0.05; **p < 0.01; ***p < 0.001).

Electronic supplementary material


T. Miyamoto was supported by a grant-in-aid for Scientific Research in Japan and a grant from the Japan Agency for Medical Research and Development. Y. Sato and K. Miyamoto were supported by a grant-in-aid for Scientific Research in Japan. This study was supported in part by a grant-in-aid for Scientific Research, a grant from the Translational Research Network Program.

Author Contributions

Author Contributions

R.W. and Y.S. performed animal experiments. T.K. and M.K. prepared animals for experiments. M.M. (Morita), T.O. and K.M. analyzed data. N.F., T.M. (Michigami), S.F., Y.T., M.N., M.M. (Matsumoto) and T.M. (Miyamoto) designed the study. T.M. (Miyamoto) wrote the manuscript with input from all authors. All authors discussed the results and commented on the manuscript.


Competing Interests

The authors declare that they have no competing interests.


Electronic supplementary material

Supplementary information accompanies this paper at doi:10.1038/s41598-017-07341-2

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


1. Masoro EJ. Aging: current concepts. Aging (Milano) 1997;9:436–437. [PubMed]
2. Kuro-o M, et al. Mutation of the mouse klotho gene leads to a syndrome resembling ageing. Nature. 1997;390:45–51. doi: 10.1038/36285. [PubMed] [Cross Ref]
3. Higuchi K. Genetic characterization of senescence-accelerated mouse (SAM) Exp Gerontol. 1997;32:129–138. doi: 10.1016/S0531-5565(96)00060-5. [PubMed] [Cross Ref]
4. Rotman G, Shiloh Y. ATM: from gene to function. Hum Mol Genet. 1998;7:1555–1563. doi: 10.1093/hmg/7.10.1555. [PubMed] [Cross Ref]
5. Harvey M, McArthur MJ, Montgomery CA, Jr., Bradley A, Donehower LA. Genetic background alters the spectrum of tumors that develop in p53-deficient mice. FASEB J. 1993;7:938–943. [PubMed]
6. Furuyama T, et al. Abnormal angiogenesis in Foxo1 (Fkhr)-deficient mice. J Biol Chem. 2004;279:34741–34749. doi: 10.1074/jbc.M314214200. [PubMed] [Cross Ref]
7. Rudolph KL, et al. Longevity, stress response, and cancer in aging telomerase-deficient mice. Cell. 1999;96:701–712. doi: 10.1016/S0092-8674(00)80580-2. [PubMed] [Cross Ref]
8. Schafer C, et al. The serum protein alpha 2-Heremans-Schmid glycoprotein/fetuin-A is a systemically acting inhibitor of ectopic calcification. J Clin Invest. 2003;112:357–366. doi: 10.1172/JCI17202. [PMC free article] [PubMed] [Cross Ref]
9. Kurosu H, et al. Regulation of fibroblast growth factor-23 signaling by klotho. J Biol Chem. 2006;281:6120–6123. doi: 10.1074/jbc.C500457200. [PMC free article] [PubMed] [Cross Ref]
10. Urakawa I, et al. Klotho converts canonical FGF receptor into a specific receptor for FGF23. Nature. 2006;444:770–774. doi: 10.1038/nature05315. [PubMed] [Cross Ref]
11. Goetz R, et al. Conversion of a paracrine fibroblast growth factor into an endocrine fibroblast growth factor. J Biol Chem. 2012;287:29134–29146. doi: 10.1074/jbc.M112.342980. [PMC free article] [PubMed] [Cross Ref]
12. Consortium A. Autosomal dominant hypophosphataemic rickets is associated with mutations in FGF23. Nat Genet. 2000;26:345–348. doi: 10.1038/81664. [PubMed] [Cross Ref]
13. Shimada T, et al. Cloning and characterization of FGF23 as a causative factor of tumor-induced osteomalacia. Proc Natl Acad Sci USA. 2001;98:6500–6505. doi: 10.1073/pnas.101545198. [PubMed] [Cross Ref]
14. Shimada T, et al. Targeted ablation of Fgf23 demonstrates an essential physiological role of FGF23 in phosphate and vitamin D metabolism. J Clin Invest. 2004;113:561–568. doi: 10.1172/JCI200419081. [PMC free article] [PubMed] [Cross Ref]
15. Yoshida T, Fujimori T, Nabeshima Y. Mediation of unusually high concentrations of 1,25-dihydroxyvitamin D in homozygous klotho mutant mice by increased expression of renal 1alpha-hydroxylase gene. Endocrinology. 2002;143:683–689. doi: 10.1210/endo.143.2.8657. [PubMed] [Cross Ref]
16. Shimada T, et al. FGF-23 is a potent regulator of vitamin D metabolism and phosphate homeostasis. J Bone Miner Res. 2004;19:429–435. doi: 10.1359/JBMR.0301264. [PubMed] [Cross Ref]
17. Hesse M, Frohlich LF, Zeitz U, Lanske B, Erben RG. Ablation of vitamin D signaling rescues bone, mineral, and glucose homeostasis in Fgf-23 deficient mice. Matrix Biol. 2007;26:75–84. doi: 10.1016/j.matbio.2006.10.003. [PubMed] [Cross Ref]
18. Anour R, Andrukhova O, Ritter E, Zeitz U, Erben RG. Klotho lacks a vitamin D independent physiological role in glucose homeostasis, bone turnover, and steady-state PTH secretion in vivo. PLoS One. 2012;7:e31376. doi: 10.1371/journal.pone.0031376. [PMC free article] [PubMed] [Cross Ref]
19. Streicher C, et al. Long-term Fgf23 deficiency does not influence aging, glucose homeostasis, or fat metabolism in mice with a nonfunctioning vitamin D receptor. Endocrinology. 2012;153:1795–1805. doi: 10.1210/en.2011-1878. [PubMed] [Cross Ref]
20. Ichikawa S, et al. A homozygous missense mutation in human KLOTHO causes severe tumoral calcinosis. J Clin Invest. 2007;117:2684–2691. doi: 10.1172/JCI31330. [PubMed] [Cross Ref]
21. Araya K, et al. A novel mutation in fibroblast growth factor 23 gene as a cause of tumoral calcinosis. J Clin Endocrinol Metab. 2005;90:5523–5527. doi: 10.1210/jc.2005-0301. [PubMed] [Cross Ref]
22. Quarles LD. Evidence for a bone-kidney axis regulating phosphate homeostasis. J Clin Invest. 2003;112:642–646. doi: 10.1172/JCI200319687. [PMC free article] [PubMed] [Cross Ref]
23. White KE, et al. Autosomal-dominant hypophosphatemic rickets (ADHR) mutations stabilize FGF-23. Kidney Int. 2001;60:2079–2086. doi: 10.1046/j.1523-1755.2001.00064.x. [PubMed] [Cross Ref]
24. Hu MC, et al. Klotho deficiency causes vascular calcification in chronic kidney disease. J Am Soc Nephrol. 2011;22:124–136. doi: 10.1681/ASN.2009121311. [PubMed] [Cross Ref]
25. Azuma M, et al. Promoter methylation confers kidney-specific expression of the Klotho gene. FASEB J. 2012;26:4264–4274. doi: 10.1096/fj.12-211631. [PubMed] [Cross Ref]
26. Johnson KA, et al. Osteoblast tissue-nonspecific alkaline phosphatase antagonizes and regulates PC-1. Am J Physiol Regul Integr Comp Physiol. 2000;279:R1365–1377. [PubMed]
27. Rutsch F, et al. Mutations in ENPP1 are associated with ‘idiopathic’ infantile arterial calcification. Nat Genet. 2003;34:379–381. doi: 10.1038/ng1221. [PubMed] [Cross Ref]
28. Rutsch F, et al. PC-1 nucleoside triphosphate pyrophosphohydrolase deficiency in idiopathic infantile arterial calcification. Am J Pathol. 2001;158:543–554. doi: 10.1016/S0002-9440(10)63996-X. [PubMed] [Cross Ref]
29. Levy-Litan V, et al. Autosomal-recessive hypophosphatemic rickets is associated with an inactivation mutation in the ENPP1 gene. Am J Hum Genet. 2010;86:273–278. doi: 10.1016/j.ajhg.2010.01.010. [PubMed] [Cross Ref]
30. Lorenz-Depiereux B, Schnabel D, Tiosano D, Hausler G, Strom TM. Loss-of-function ENPP1 mutations cause both generalized arterial calcification of infancy and autosomal-recessive hypophosphatemic rickets. Am J Hum Genet. 2010;86:267–272. doi: 10.1016/j.ajhg.2010.01.006. [PubMed] [Cross Ref]
31. A gene (PEX) with homologies to endopeptidases is mutated in patients with X-linked hypophosphatemic rickets. The HYP Consortium. Nat Genet. 1995;11:130–136. doi: 10.1038/ng1095-130. [PubMed] [Cross Ref]
32. Feng JQ, et al. Loss of DMP1 causes rickets and osteomalacia and identifies a role for osteocytes in mineral metabolism. Nat Genet. 2006;38:1310–1315. doi: 10.1038/ng1905. [PMC free article] [PubMed] [Cross Ref]
33. Moran JJ. Idiopathic arterial calcification of infancy: a clinicopathologic study. Pathol Annu. 1975;10:393–417. [PubMed]
34. Nakamura I, et al. Association of the human NPPS gene with ossification of the posterior longitudinal ligament of the spine (OPLL) Hum Genet. 1999;104:492–497. doi: 10.1007/s004390050993. [PubMed] [Cross Ref]
35. Koshizuka Y, et al. Nucleotide pyrophosphatase gene polymorphism associated with ossification of the posterior longitudinal ligament of the spine. J Bone Miner Res. 2002;17:138–144. doi: 10.1359/jbmr.2002.17.1.138. [PubMed] [Cross Ref]
36. Tahara M, et al. The extent of ossification of posterior longitudinal ligament of the spine associated with nucleotide pyrophosphatase gene and leptin receptor gene polymorphisms. Spine (Phila Pa 1976) 2005;30:877–880. doi: 10.1097/ [PubMed] [Cross Ref]
37. Li Q, et al. Mutant Enpp1asj mice as a model for generalized arterial calcification of infancy. Dis Model Mech. 2013;6:1227–1235. doi: 10.1242/dmm.012765. [PMC free article] [PubMed] [Cross Ref]
38. Okawa A, et al. Mutation in Npps in a mouse model of ossification of the posterior longitudinal ligament of the spine. Nat Genet. 1998;19:271–273. doi: 10.1038/956. [PubMed] [Cross Ref]
39. Koshizuka Y, Ikegawa S, Sano M, Nakamura K, Nakamura Y. Isolation of novel mouse genes associated with ectopic ossification by differential display method using ttw, a mouse model for ectopic ossification. Cytogenet Cell Genet. 2001;94:163–168. doi: 10.1159/000048809. [PubMed] [Cross Ref]
40. Komori T, et al. Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell. 1997;89:755–764. doi: 10.1016/S0092-8674(00)80258-5. [PubMed] [Cross Ref]
41. Yasuda H, et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci USA. 1998;95:3597–3602. doi: 10.1073/pnas.95.7.3597. [PubMed] [Cross Ref]
42. Tsuda E, et al. Isolation of a novel cytokine from human fibroblasts that specifically inhibits osteoclastogenesis. Biochem Biophys Res Commun. 1997;234:137–142. doi: 10.1006/bbrc.1997.6603. [PubMed] [Cross Ref]
43. Abe K, et al. The occurrence of interglobular dentin in incisors of hypophosphatemic mice fed a high-calcium and high-phosphate diet. J Dent Res. 1992;71:478–483. doi: 10.1177/00220345920710031101. [PubMed] [Cross Ref]
44. Kuro-o M. Klotho, phosphate and FGF-23 in ageing and disturbed mineral metabolism. Nat Rev Nephrol. 2013;9:650–660. doi: 10.1038/nrneph.2013.111. [PubMed] [Cross Ref]
45. Nature. Volume 228, 1970: pages 764-766. Unique biosynthesis by kidney of a biologically active vitamin D metabolite. Nutr Rev. 1981;39:215–218. [PubMed]
46. Dermaku-Sopjani M, et al. Downregulation of NaPi-IIa and NaPi-IIb Na-coupled phosphate transporters by coexpression of Klotho. Cell Physiol Biochem. 2011;28:251–258. doi: 10.1159/000331737. [PubMed] [Cross Ref]
47. Martin A, David V, Quarles LD. Regulation and function of the FGF23/klotho endocrine pathways. Physiol Rev. 2012;92:131–155. doi: 10.1152/physrev.00002.2011. [PMC free article] [PubMed] [Cross Ref]
48. Laughlin GA, McEvoy LK, Barrett-Connor E, Daniels LB, Ix JH. Fetuin-A, a new vascular biomarker of cognitive decline in older adults. Clin Endocrinol (Oxf) 2014;81:134–140. doi: 10.1111/cen.12382. [PMC free article] [PubMed] [Cross Ref]
49. Li YC, et al. Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor-ablated mice. Endocrinology. 1998;139:4391–4396. doi: 10.1210/endo.139.10.6262. [PubMed] [Cross Ref]
50. Panda DK, et al. Inactivation of the 25-hydroxyvitamin D 1alpha-hydroxylase and vitamin D receptor demonstrates independent and interdependent effects of calcium and vitamin D on skeletal and mineral homeostasis. J Biol Chem. 2004;279:16754–16766. doi: 10.1074/jbc.M310271200. [PubMed] [Cross Ref]
51. Tullet JM, et al. Direct inhibition of the longevity-promoting factor SKN-1 by insulin-like signaling in C. elegans. Cell. 2008;132:1025–1038. doi: 10.1016/j.cell.2008.01.030. [PMC free article] [PubMed] [Cross Ref]
52. Dang W, et al. Histone H4 lysine 16 acetylation regulates cellular lifespan. Nature. 2009;459:802–807. doi: 10.1038/nature08085. [PMC free article] [PubMed] [Cross Ref]
53. Shi Y, et al. ROS-dependent activation of JNK converts p53 into an efficient inhibitor of oncogenes leading to robust apoptosis. Cell Death Differ. 2014;21:612–623. doi: 10.1038/cdd.2013.186. [PMC free article] [PubMed] [Cross Ref]
54. Barascu A, et al. Oxidative stress induces an ATM-independent senescence pathway through p38 MAPK-mediated lamin B1 accumulation. EMBO J. 2012;31:1080–1094. doi: 10.1038/emboj.2011.492. [PubMed] [Cross Ref]
55. Finkel T, Holbrook NJ. Oxidants, oxidative stress and the biology of ageing. Nature. 2000;408:239–247. doi: 10.1038/35041687. [PubMed] [Cross Ref]
56. Tsujikawa H, Kurotaki Y, Fujimori T, Fukuda K, Nabeshima Y. Klotho, a gene related to a syndrome resembling human premature aging, functions in a negative regulatory circuit of vitamin D endocrine system. Mol Endocrinol. 2003;17:2393–2403. doi: 10.1210/me.2003-0048. [PubMed] [Cross Ref]
57. Zhang H, et al. Klotho is a target gene of PPAR-gamma. Kidney Int. 2008;74:732–739. doi: 10.1038/ki.2008.244. [PubMed] [Cross Ref]
58. Mitani H, et al. In vivo klotho gene transfer ameliorates angiotensin II-induced renal damage. Hypertension. 2002;39:838–843. doi: 10.1161/01.HYP.0000013734.33441.EA. [PubMed] [Cross Ref]
59. Ito N, et al. Regulation of FGF23 expression in IDG-SW3 osteocytes and human bone by pro-inflammatory stimuli. Mol Cell Endocrinol. 2015;399:208–218. doi: 10.1016/j.mce.2014.10.007. [PubMed] [Cross Ref]
60. Fukumoto S. FGF23-FGF Receptor/Klotho Pathway as a New Drug Target for Disorders of Bone and Mineral Metabolism. Calcif Tissue Int. 2016;98:334–340. doi: 10.1007/s00223-015-0029-y. [PubMed] [Cross Ref]
61. Rutsch F, et al. Hypophosphatemia, hyperphosphaturia, and bisphosphonate treatment are associated with survival beyond infancy in generalized arterial calcification of infancy. Circ Cardiovasc Genet. 2008;1:133–140. doi: 10.1161/CIRCGENETICS.108.797704. [PMC free article] [PubMed] [Cross Ref]
62. Edouard T, et al. Efficacy and safety of 2-year etidronate treatment in a child with generalized arterial calcification of infancy. Eur J Pediatr. 2011;170:1585–1590. doi: 10.1007/s00431-011-1572-9. [PubMed] [Cross Ref]
63. Albright RA, et al. ENPP1-Fc prevents mortality and vascular calcifications in rodent model of generalized arterial calcification of infancy. Nat Commun. 2015;6:10006. doi: 10.1038/ncomms10006. [PMC free article] [PubMed] [Cross Ref]
64. Wilson JR, et al. Genetics and heritability of cervical spondylotic myelopathy and ossification of the posterior longitudinal ligament: results of a systematic review. Spine (Phila Pa 1976) 2013;38:S123–146. doi: 10.1097/BRS.0b013e3182a7f478. [PubMed] [Cross Ref]
65. McClelland R, et al. Accelerated ageing and renal dysfunction links lower socioeconomic status and dietary phosphate intake. Aging (Albany NY) 2016;8:1135–1149. doi: 10.18632/aging.100948. [PMC free article] [PubMed] [Cross Ref]
66. Heiss A, et al. Structural basis of calcification inhibition by alpha 2-HS glycoprotein/fetuin-A. Formation of colloidal calciprotein particles. J Biol Chem. 2003;278:13333–13341. doi: 10.1074/jbc.M210868200. [PubMed] [Cross Ref]
67. Okawa A, et al. Mapping of a gene responsible for twy (tip-toe walking Yoshimura), a mouse model of ossification of the posterior longitudinal ligament of the spine (OPLL) Mamm Genome. 1998;9:155–156. doi: 10.1007/s003359900707. [PubMed] [Cross Ref]
68. Kurosu H, et al. Suppression of aging in mice by the hormone Klotho. Science. 2005;309:1829–1833. doi: 10.1126/science.1112766. [PMC free article] [PubMed] [Cross Ref]
69. Yoshizawa T, et al. Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nat Genet. 1997;16:391–396. doi: 10.1038/ng0897-391. [PubMed] [Cross Ref]

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group