University of Groningen
Klotho in vascular biology
Mencke, Rik
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Mencke, R. (2018). Klotho in vascular biology. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
21
Chapter 2
The role of the anti-ageing protein Klotho in
vascular physiology and pathophysiology
R. Mencke J.L. Hillebrands on behalf of the NIGRAM Consortium
22
Abstract
Klotho is an anti-ageing protein that functions in many pathways that govern ageing, like regulation of phosphate homeostasis, insulin signaling, and Wnt signaling. Klotho expression levels and levels in blood decline during ageing. The vascular phenotype of Klotho deficiency features medial calcification, intima hyperplasia, endothelial dysfunction, arterial stiffening, hypertension, and impaired angiogenesis and vasculogenesis, with characteristics similar to aged human arteries.
Klotho-deficient phenotypes can be prevented and rescued by Klotho gene expression or protein supplementation. High phosphate levels are likely to be directly pathogenic and are a prerequisite for medial calcification, but more important determinants are pathways that regulate cellular senescence, suggesting that deficiency of Klotho renders cells susceptible to phosphate toxicity. Overexpression of Klotho is shown to ameliorate medial calcification, endothelial dysfunction, and hypertension.
Endogenous vascular Klotho expression is a controversial subject and, currently, no compelling evidence exists that supports the existence of vascular membrane-bound Klotho expression, as expressed in kidney. In vitro, Klotho has been shown to decrease oxidative stress and apoptosis in both SMCs and ECs, to reduce SMC calcification, to maintain the contractile SMC phenotype, and to prevent µ-calpain overactivation in ECs.
Klotho has many protective effects with regard to the vasculature and constitutes a very promising therapeutic target. The purpose of this review is to explore the etiology of the vascular phenotype of Klotho deficiency and the therapeutic potential of Klotho in vascular disease.
23
Introduction
Klotho is an anti-ageing gene that was discovered in 1997 (1). It is expressed mainly in the kidney, parathyroid gland, and choroid plexus and exists as a membrane-bound protein that can be cleaved off, also to be found as a soluble protein in the blood, urine, and cerebrospinal fluid (2, 3). The Klotho protein contains two homologous internal repeats, termed KL1 and KL2. Soluble Klotho contains both KL1 and KL2, but proteolytic cleavage may also occur between KL1 and KL2, producing two additional, smaller soluble Klotho proteins. An alternatively spliced transcript has also been hypothesized to code for a secreted Klotho protein (4). The current paradigm of Klotho protein forms is summarized in Figure 1.
Deficiency of Klotho in mice was found to have profound systemic effects, producing a phenotype markedly reminiscent of human ageing. This phenotype consists of, among other traits, a short lifespan, stunted growth and kyphosis, vascular calcification and atherosclerosis, osteoporosis, pulmonary emphysema, cognitive impairment, deafness, and atrophy of skin, muscles, gonads, and many other organs (1). It has been shown that increasing Klotho levels in mice yields an extended lifespan (120-130% of normal) (5), better cognitive function (6, 7), resistance against induction of renal disease (8-15), cardiac disease (16, 17), pulmonary disease (18, 19), vascular calcification (3), diabetes (20, 21), oxidative stress (22, 23), while also acting as an in vivo tumor suppressor (9, 24-27).
So far, it has been shown that Klotho acts via at least six distinct mechanisms: (1) as a
membrane-bound co-receptor for soluble ligands (as a co-receptor for ligand fibroblast growth factor (FGF)23 with FGFR1c, inducing phosphaturia, down-regulating the vitamin D-producing enzyme 1α-hydroxylase, and regulating renal sodium re-absorption) (28-32), (2) as
a soluble co-receptor for soluble ligands (maintaining endothelial integrity by mediating vascular endothelial growth factor (VEGF)-induced internalization of the Klotho-bound transient receptor potential cation channel (TRPC)1/VEGFR2 complex (33), (3) as a soluble
decoy receptor for soluble factors (inhibiting Wnt signaling by binding to several Wnt factors) (10, 11, 34, 35), (4) as a soluble protein decreasing receptor affinity for ligands (directly
inhibiting insulin growth factor (IGF)1 and transforming growth factor (TGF)β signaling by binding to IGF1R and TGFβR2, respectively) (5, 9, 24), and (5) as a membrane-bound
competitor for binding sites (inhibiting FGF2 signaling by binding to FGFR1c) (14, 25), and (6)
as an enzyme (modifying sugar moieties on calcium channel TRPV5, potassium channel ROMK1, and phosphate transporter NaPi2a, affecting their cell surface abundance through sialidase or β-glucoronidase activity) (36-40).
Klotho has garnered a lot of attention in vascular biology and a number of observations can be linked to underscore its clinical relevance. The vascular phenotype of Klotho deficiency is very similar to both human ageing and “accelated” ageing observed in chronic kidney disease (CKD) (1). CKD is also a state of acquired Klotho deficiency (3, 41, 42). In humans, genetic
24
Klotho deficiency (the H193R missense mutation) also leads to severe vascular calcification (43). It is therefore conceivable that Klotho may play a causal role in the pathogenesis of cardiovascular complications in CKD, the development of which is the leading cause of death in CKD patients (44). Additionally, Klotho gene variants have been found to be protective or detrimental for the development of cardiovascular and cerebrovascular disease (45, 46). Finally, although current serum Klotho measurements may not be reliable, Klotho levels may also be lower in patients with cardiovascular disease (47, 48). It is therefore very important to delineate the effects of Klotho on the cardiovascular system, in order to identify new targets for new therapies. Possible approaches include up-regulation of Klotho, administration of Klotho, or administration of Klotho-based compounds. Structure-function analyses indicate that different domains in the Klotho protein have different functions. FGF23 requires the full-length membrane-bound Klotho protein (31). The KL2 domain is required for binding to TRPC1/VEGFR2 (33), while KL1 can exert tumor suppressor effects and inhibit IGFR and Wnt signaling (26, 34), independently of enzymatic activity, while enzymatic activity in either domain is required for modifying TRPV5 and NaPi2a (37, 39, 49).
Figure 1. Paradigm of Klotho forms. Two Klotho mRNA transcripts are expressed in the kidney, of which one
contains a short intronic sequence after exon 3, giving rise to a stop codon. This mRNA transcript has been hypothesized to code for a secreted Klotho protein. The other transcript codes for membrane-bound Klotho (positive immunohistochemical staining on healthy human kidney). Membrane-bound Klotho is cleaved proteolytically by secretases above the membrane and between the KL1 and KL2 internal repeats, producing three soluble Klotho molecules.
25
These effects, apparently acting in concert in alleviating ageing from the sub-cellular level to the level to the level of the organism, show that different functions of Klotho can be dissected, which may be of consequence for future therapies. consequence for future therapies. Therefore, this comprehensive review will focus on the role of Klotho in vascular physiology and pathophysiology in order to assess its therapeutic potential. We will first describe the phenotype of Klotho deficiency, the effects of interventions on the phenotype of Klotho deficiency, and the experimental effects of Klotho overexpression, proceeding to the topics of endogenous vascular Klotho expression, and the in vitro effects of Klotho on vascular cells.
The vasculature and Klotho – in vivo experimental evidence
Klotho-hypomorphic kl/kl mice, as originally described, exhibit two remarkable vascular histological features: vascular calcification in the tunica media (the contractile, smooth muscle cell layer of arteries) and intima hyperplasia (hyperplasia of the inner, endothelial lining of arteries, which is invaded by migrating and proliferating smooth muscle cells) (1). Functionally, Klotho deficiency causes endothelial dysfunction and arterial stiffening, as well as hypertension and impaired angiogenesis. We will assess the histological, functional, and molecular phenotype of in vivo Klotho deficiency and its resemblance to the phenotype of the aged human vasculature, as well as assess the effects of interventions in Klotho deficiency (summed up in Tables 1 and 2), assess the effects of Klotho overexpression or supplementation on various vascular phenotypes.
Vascular calcification
Vascular calcification in Klotho deficiency
The vascular calcification in kl/kl mice is progressive from 4 weeks of age onward, mostly confined to the media and reminiscent of Mönckeberg’s sclerosis in human chronic kidney disease (CKD). It is present in arteries ranging from aorta to middle-sized muscular arteries, to small renal arteries. Furthermore, extensive ectopic calcification was noted in brain, lung, gastrointestinal tract, testis, skin, and heart, in line with the overt hypercalcemia, hyperphosphatemia, and hypervitaminosis D (50) found in these mice. These features were noted only in animals homozygous for the hypomorphic kl allele. It has been demonstrated that aorta and kidneys from Klotho-deficienct mice indeed have a significantly higher calcium
26
content (3). Both histochemistry and electron microscopy analyses indicate that these calcifications are largely confined to the elastin fibers of the tunica media (51). Furthermore, the medial smooth muscle cells are phenotypically akin to matrix synthesizing cells surrounded by secreted matrix vesicles. The vascular smooth muscle cells in calcifying areas in kl/kl aorta exhibit high Runx2 expression, indicative of trans-differentiation to an osteoblast-like phenotype (52). Expression of matrix Gla protein (MGP), a potent vitamin K-dependent calcification inhibitor known to be highly expressed in calcifying human arteries, is similarly increased at the edges of the calcified areas. Hu et al. later also described higher Runx2 mRNA expression levels in kl/kl aorta, in addition to higher mRNA levels of phosphate transporters Pit1 and Pit2 and lower mRNA levels of smooth muscle cell (SMC) marker SM22α (3). Higher Pit1 mRNA and Runx2 mRNA and protein levels in kl/kl aorta were also found by other authors, in addition to higher Msx2, osterix, tumor necrosis factor (TNF)-α, alkaline phosphatase, osteopontin, receptor activator of nuclear factor kappa-B ligand (RANKL), nuclear factor of activated T cells (NFAT)5, and Sox9 mRNA and/or protein levels (53-56). These findings suggest that medial calcification in Klotho deficiency is an active process similar to osteogenesis, analogous to medial calcification in human CKD and human ageing. This view is supported by the finding that even expression of otherwise osteocyte-exclusive FGF23 was found to arise in kl/kl aortic calcifications (55). Mammalian target of rapamycin (mTOR) has also been found to be activated in kl/kl mouse aorta, which contributes to medial calcification (57). Furthermore, cyclooxygenase 2, involved in prostaglandin synthesis and bone formation, is also up-regulated in Klotho-/- mice, although at a lower level in aortic calcified lesions as
compared to aortic valve calcifications (58). In addition to MGP and osteopontin up-regulation, the finding of up-regulated stanniocalcin 2 expression (but not stanniocalcin 1 expression) in calcifying lesions in Klotho-deficient aorta and renal arterioles is speculated to indicate a protective mechanism at work as well (59). Kl/kl mouse aorta and kidney may also express higher levels of ectonucleotide pyrophosphatase/phosphodiesterase 1 (ENPP1) and ANK, a pyrophosphate generator and transporter, respectively (60). This possibly reflects activation of an additional protective mechanism, in order to prevent calcium phosphate deposition. Furthermore, it was found that the miRNAs miR-135a*, miR-762, miR-714, and miR-712* were highly up-regulated in kl/kl aorta, which was accompanied by down-regulation of target genes: calcium efflux pumps/exchangers NCKX1, PMCA1, and NCKX4 (an effect found to predispose towards calcification) (61). These authors also note that 10% of (heterozygous) kl/+ animals also display minor vascular calcification, a feature that was associated with up-regulated expression of the aforementioned miRNAs, also in kl/+ animals. No differences have been reported between vascular phenotypes of different Klotho mutant mice, which include the original (hypomorphic) kl//kl mice (1), different Klotho-/- mice (30, 62,
63), and β-actin-Cre/KL-LoxP mice (64). In short, the phenotype of Klotho deficiency features medial calcification, associated with altered expression of both phosphate and calcium transporters, activation of mechanisms that aim to inhibit calcification, and osteochondrogenic transdifferentiation of SMCs. These lesions greatly resemble the calcified
27
lesions in human age-related or (exaggeratedly) CKD-related medial calcification, both morphologically and molecularly.
Interventions in the development of vascular calcification
The effect of restoration of Klotho expression in Klotho deficiency
In their original study, Kuro-o et al. describe crossing of kl/kl mice with mice that express Klotho constitutively under the ubiquitous elongation factor 1α promoter, rescuing the phenotype to a large extent (1). Both aortic calcification and ectopic calcification were markedly reduced. The same group then performed an experiment to assess whether the Klotho gene is also able to rescue the developed kl/kl phenotype, rather than prevent it. Using an adenoviral vector containing the Klotho gene via tail vein infusion at 4-5 weeks of age, the development of aortic medial calcification could be halted and was noted, at 27 weeks of age, in the few mice that survived that long, to be less advanced than in 4-10-weeks-old kl/kl mice (65). They then used kl/kl mice that express an ectopic Klotho gene conditionally (under the zinc-dependent mouse metallothionein-I promoter), allowing for free manipulation of Klotho expression. It was also found that there was no development of medial calcification if Klotho expression was induced via zinc water feeding from three weeks of age onwards (66). Furthermore, it was found that inducing Klotho expression for three weeks was already successful in completely reversing medial calcification that had already developed before starting zinc water feeding at 5 or 8 weeks of age, illustrating the potential therapeutic potency of Klotho. It is as of yet unknown how this therapeutic effect can be explained, but a potential to both reverse and prevent age-related disorders (for the phenotypical improvement was systemic) offers tantalizing possibilities. Moreover, after initially preventing development of medial calcification, subsequent zinc withdrawal from 11 weeks of age onwards caused renewed development of medial calcification (assessed at 19 and 27 weeks of age). This illustrates that Klotho is continuously required in order to maintain vascular health, rather than only during a hypothetical critical period after birth. It also underscores how acquired Klotho deficiency later in life, as occurs in CKD, may as a single factor be enough to materialize a phenotype of vascular calcification.
Restoration of Klotho expression in Klotho deficiency affects the phenotype via soluble Klotho While it is evident from the previous discussion that genetic re-introduction of Klotho can rescue the vascular calcification in Klotho deficiency, it is not immediately clear whether these effects are mediated by membrane-bound or soluble Klotho. It could be argued that soluble Klotho is the likely mediator, since the induced non-vascular expression patterns differ greatly. Induced Klotho expression is mostly confined to the gastrointestinal tract in the conditional zinc-dependent model (66), confined to the liver in the adenovirus-mediated
28
model (65), and found in many organs in the ubiquitous overexpression model (1), but all models produce a similar effect. There is, however, more compelling evidence that shows that the rescue of the vascular kl/kl phenotype is at least predominantly dependent on soluble Klotho. First of all, a study by Chen et al. shows that intraperitoneal injections of soluble Klotho (0.02 mg/kg/48 h between 3 and 8 weeks of age) result in marked reduction of vascular calcification (67). Notably, while increasing urinary phosphate excretion, this treatment left serum phosphate and calcium levels unaltered. Secondly, a number of studies by a group that has created a number of organ-specific Klotho knockout models has yielded important insights. Lindberg et al. showed that the vascular kl/kl phenotype is present in Six2-Cre/KL-LoxP mice, a mouse model of selective and complete renal tubular Klotho deficiency, suggesting that vascular calcification is normally prevented by kidney-derived Klotho (68). In further support of this conclusion, this study also shows that the soluble Klotho levels in the blood are largely kidney-derived. Distal tubule Klotho deletion in Ksp-Cre/KL-LoxP mice, however, did not induce vascular calcification (64), suggesting that moderate remaining Klotho levels are enough to prevent the overt phenotype, akin to the lack of vascular calcification in kl/+ mice. Selective deletion of Klotho in arterial smooth muscle cells (in SM22a-Cre/KL-LoxP mice) did not produce vascular abnormalities (69). This suggests that if any endogenous expression of Klotho in smooth muscle cells is present, its deletion alone does not contribute significantly to the development of an overtly aberrant phenotype. Moreover, neither deletion of parathyroid Klotho in PTH-Cre/KL-LoxP mice, nor deletion of proximal tubule Klotho in Kap-Cre/KL-LoxP mice, PEPCK-Cre/KL-LoxP mice, or Scl34a1-Cre/KL-LoxP mice induced an overt vascular phenotype (70, 71). A final important argument for soluble Klotho-mediated vasculoprotection is derived from the parabiosis experiment by Saito et al., showing that a shared circulatory system with WT mice restores acetylcholine-dependent vasodilation in kl/+ mice after 4 weeks (72).
Interventions in mineral homeostasis
The first interventions aimed at rescuing the kl/kl phenotype involved interventions in mineral metabolism. An early study shows that reducing dietary phosphate (0.4% vs 1.03%) in kl/kl mice partially rescues many features of their phenotype, among which reduction of ectopic calcification in kidney (in male mice, and in female mice as well after addition of 0.25% zinc orotate) (73). Vascular calcification was not examined, but it is likely that vascular calcifications were reduced as well. These effects, however, may have been dependent at least in part on a phosphate restriction-induced increase in renal Klotho expression, even in Klotho-hypomorphic kl/kl mice, which are not fully deficient.
Subsequent studies have yielded extensive evidence indicative of a sine qua non role for phosphate toxicity in vascular calcification in complete Klotho deficiency. Lowering phosphate levels through a genetic approach (in NaPi2a-/-/Klotho-/- mice) was shown to prevent vascular
29
re-established the phenotype of vascular calcification (75). These results indicate that high phosphate levels are a requirement (in the setting of Klotho deficiency) for the vascular calcification to develop despite the presence of hypercalcemia and extremely elevated 1,25(OH)2-vitamin D3 levels. NaPi2a expression was found to be increased in kl/kl kidney,
which is thought to cause hyperphosphatemia through increased phosphate re-absorption, a hypothesis that is supported by the lack of hyperphosphatemia in the NaPi2a-/-/Klotho-/- mice.
It is unknown whether other vascular features of Klotho deficiency may also be rescued by normalization of phosphate levels. Curiously, in another experiment, a low phosphate diet (0.2%) in full Klotho-/- mice was unable to rescue renal calcification after 7 weeks (assuming
that vascular calcifications were also still present) (30). The low phosphate diet, however, only caused a decrease in serum phosphate level of 0.7 mg/dL (whereas a 3.6 mg/dL decrease was reported in the genetic study (75) at 6 weeks of age, although the actual phosphate levels and assays may not be comparable). It is therefore possible that despite the low phosphate diet, the serum phosphate levels remained elevated to the point of causing calcifications. This may be mediated via increased NaPi2a activity and without the possibility of compensatory Klotho up-regulation, as was shown to be possible in Klotho-hypomorphic kl/kl mice (73). Interestingly, a low calcium diet (0.02%) was shown to be effective in preventing ectopic calcifications in this study, showing that calcium can also function as a rate-limiting factor in the pathophysiology of Klotho-/- vascular calcification, in addition to phosphate (30). An
attempt to target phosphate homeostasis by ablating secreted frizzled-related protein 4 (Sfrp4) in Sfrp4-/-/Klotho-/- mice did not improve or worsen vascular calcification (76). Since
Sfrp4-/- mice display no mineral homeostasis abnormalities, these data suggest that Sfrp4
does not play a significant role in phosphate regulation. Finally, deletion of both Klotho and dentin matrix protein 1 (DMP1) underlined the importance of DMP1 in phosphate homeostasis(77). Klotho-/-/DMP1-/- mice displayed more severe vascular calcification and
more apoptosis in aorta and arterioles than did Klotho-/- mice, at similar serum phosphate
levels. If DMP1 expression is increased in Klotho-/- arteries as it is in the kidney, this may
signify activation of another local mechanism that protects against Klotho deficiency-induced calcification. Hypophosphatemia in DMP-/- mice is converted to hyperphosphatemia upon
additional knockout of Klotho, which improves bone mineralization, but the lack of DMP1 apparently leaves arteries particularly vulnerable to calcification.
Another approach that has been studied is the modulation of vitamin D levels that have been reported to be extremely elevated in Klotho deficiency (50). Dietary vitamin D restriction was also shown to be effective in reducing vascular calcification in complete Klotho-/- mice (30). In
the vitamin D-deficient diet that was used, however, phosphate content was also decreased from 1.09% to 0.4% and calcium content was decreased from 1.46% to 0.6%. This offers other possible explanations for the phenotypic amelioration, perhaps to be attributed to synergistic lower phosphate, calcium, and vitamin D levels. It was noted that in control experiments with diets with the same phosphate and calcium content, rescue of the phenotype was not observed. However, it is difficult to account for the interactions between these mediators of
30
mineral homeostasis, e.g. for the additional effect of vitamin D deficiency on phosphate levels, which may result in a difference in phenotype. Using a genetic approach, ablation of 1α-hydroxylase expression in Cyp27b1-/-/Klotho-/- mice completely prevented the
development of vascular and ectopic calcifications (78). However, this genetic intervention also caused hypophosphatemia and hypocalcemia, suggesting that vitamin D effects on the vasculature are mediated, at least in part, by phosphate and calcium. These findings concerning serum biochemistry and renal calcifications have been corroborated in another study (79). In a similar study, it was shown that ablation of vitamin D signaling, by mutation of the vitamin D receptor in VDRΔ/Δ/Klotho-/- mice also prevents vascular calcification at least
at 8 weeks of age (80). It should be noted that these mice were on a rescue diet enriched in phosphate (1.25%), calcium (2.0%), lactose (20%), and 600 IU vitamin D/kg, ensuring normocalcemia, normophosphatemia, and normal PTH levels, so again, in this experiment, rescue of the phenotype might be confounded by normalized phosphate and calcium levels. A study by Alexander et al. compared Klotho-/- mice on a control diet (0.9% calcium, 0.63%
phosphate, 1500 IU 1,25(OH)2-vitamin D3) to Klotho-/- mice on a rescue diet (0.34% calcium,
0.22% phosphate, <5 IU 1,25(OH)2-vitamin D3) (40). In this study, while serum vitamin D and
calcium levels normalized, phosphate levels were still elevated and the development of renal calcifications was only halted slightly. Vascular calcification was not assessed.
Moving up a step in regulatory mechanisms to phosphaturic hormones, the identical phenotypes with regard to mineral metabolism in FGF23-/- mice, Klotho-/- mice, and FGF23
-/-/Klotho-/- mice constitute compelling evidence for FGF23 and Klotho acting in a common
pathway (63). FGF23 action was determined to be dependent on Klotho due to the inability of FGF23 to induce phosphaturia in Klotho-/- mice, of which it is capable in wild-type and
FGF23-/- mice. In a similar study, it was found that Hyp/Klotho-/- mice (harboring mutations in
the Klotho and PHEX genes, the latter of which causes high FGF23 levels and hypophosphatemia) basically exhibit a Klotho-/- phenotype with vascular calcification and
hyperphosphatemia(60). This supports the notion that FGF23 regulation of phosphate is fully Klotho-dependent. Finally, the finding that Klotho-/-/FGF23TG mice (lacking Klotho and
overexpressing FGF23) also display vascular calcification similar to Klotho-/- mice
independently confirms this line of evidence of FGF23 signaling effects on mineral metabolism being Klotho-dependent (81). Another group has generated PTH-/-/Klotho-/- mice, in which
ectopic calcifications are still pervasive, in the presence of even more exaggerated hyperphosphatemia and normocalcemia (yielding a Ca x P product similar to Klotho-/- mice)
(82). Although the vasculature was not investigated, it is reasonable to hypothesize that vascular calcifications were also still present. Following the same train of thought, vascular calcified lesions may be alleviated in PTH-infused Klotho-/- mice that were shown to exhibit
normophosphatemia (82, 83). These results suggest that, although there are many interactions between PTH and Klotho, they influence vascular calcification largely independently via regulating phosphate and calcium levels.
31
The previous discussion favors a direct pathogenic role for phosphate and calcium, whereas vitamin D effects are most likely indirect and mediated by phosphate and calcium. However, a lot of evidence is still circumstantial and an experiment in which vitamin D signaling ablation is combined with a high phosphate diet as compared to both interventions alone might provide clearer answers as to whether vitamin D effects on calcification are exclusively mediated by its target electrolytes.
Interventions in osteochondrogenic signaling
In a study exploring the effects of treatment with the aldosterone receptor antagonis spironolactone on kl/kl mice, it was found that aortic calcification was markedly diminished, potentially owing to less osteoinductive signaling due to lower Pit1 levels (53). Levels of Runx2, Msx2, osterix, TNF-α, and alkaline phosphatase mRNA were also lowered by spironolactone treatment. Interestingly, plasminogen activator inhibitor (PAI)-1 mRNA expression was reduced by spironolactone and endothelial nitric oxide synthase (eNOS) mRNA expression was normalized, suggesting that spironolactone may also have improved endothelial function, although this was not assessed in this study. The role of eNOS itself was recently addressed when it was found that treatment of kl/kl mice with homoarginine exacerbated vascular calcification (56). Homoarginine treatment essentially resulted in NOS inhibition and subsequent osteoinductive signaling, as determined by increased levels of Cbfa1, Pai1, Msx2, and alkaline phosphatase mRNA (however, coupled with attenuation of apoptosis). As will be discussed in section 2.3.1, Klotho deficiency entails impaired NO production, but the homoarginine-induced aggravation of vascular calcification indicates that the residual NO production still offers some degree of protection against vascular calcification. In a recent study, it was found that 0.28 M NH4Cl in drinking water greatly
reduced vascular calcification in kl/kl mice, whereas vitamin D, calcium, and phosphate levels were unaltered (54). Acidosis, which is already a feature of Klotho deficiency, was slightly aggravated by NH4Cl treatment, but the difference did not reach significance. A lower blood
pH may have negatively affected calcium and phosphate precipitation slightly, but lysosomal alkalinization and subsequent osmosensitive NFAT5 down-regulation were likely to be a greater contributor. This prevented downstream Runx2-mediated osteochondrogenic signaling in smooth muscle cells and normalized senescence-associated TGFβ, PAI-1, p21, and senescence-associated (SA)-β-galactosidase mRNA levels. Aiming to address to which property of NH4Cl these effects can be attributed, Leibrock et al. have also tested NH4NO3 (a
different NH4+ donor) (84), acetazolamide (which induces acidosis) (85), and NaHCO3 (which
induces alkalosis) (86). Treatment with 0.28 M NH4NO3 until 8 weeks of age prevented the
development of vascular calcification as well, without affecting serum phosphate of calcium levels (84). Acetazolamide-induced aggravation of acidosis was also found to prevent medial calcification in kl/kl mice, an effect indeed likely mediated by increased solubility of osteoinductive calcium phosphate crystals, as well as by a decrease in aldosterone levels. The
32
inhibition of osteochondrogenic signaling was also exemplified by normalization of Klotho deficiency-induced elevations of aortic alkaline phosphatase and of calcification inhibitors osteopontin, osteoprotegerin, and fetuin A (85). Illustrating the effects of alkalosis, however, treatment with 0.15 M NaHCO3 also partially reversed ectopic calcification (arteries were not
investigated, but may also have been less affected). This effect was likely also the result of lower aldosterone levels, as well as the result of alkalosis-induced phosphaturia, concordant with lower serum phosphate (but not calcium) levels (86). Aiming to modulate active calcification by targeting cyclooxygenase 2, it was found that both a genetic approach (in Ptgs2+/-/Klotho-/- and Ptgs2-/-/Klotho-/-mice) and a pharmacological approach (using celecoxib
in Klotho-/- mice) resulted in less calcification of the aortic valve (58). Celecoxib treatment
inhibited osteochondrogenic signaling, as evidenced by decreased Runx2, osteopontin, and alkaline phosphatase mRNA levels. Aortic calcification, however, was not analyzed in the genetic model and was similar in Klotho-/- mice fed a normal or a celecoxib diet, a result that
seems in line with lower aortic cyclooxygenase 2 expression in these animals. Finally, the same group also found that aortic valve calcification in Klotho-/- mice was dependent on bone
morphogenetic protein (BMP) signaling via pSmad1/5/8 in aortic valce interstitial cells (87). However, pSmad1/5/8 was not detected in aortic SMCs, so the relevance of BMP signaling to vascular calcification in Klotho deficiency is yet to be determined.
Interventions in senescence-related pathways
The relevance of the decrease in the expression of PAI-1, a known inducer of cellular senescence, is especially evident in a study on PAI+/-/kl/kl and PAI-/-/kl/kl mice (88). Renal
calcifications were reduced by 41% and 96%, respectively (aortic calcification was not assessed), while mice displayed comparable hyperphosphatemia and hypercalcemia, as compared to kl/kl mice. Klotho deficiency-induced up-regulation of a down-stream target of PAI-1, p16Ink4a, a known tumor suppressor that induces cellular senescence, was found to be
decreased by partial and full PAI-1 deletion. A recent study uncovered an interesting additional link between PAI-1, p16Ink4a, and Klotho, when it was found that the Klotho
deficiency phenotype including ectopic calcification was partially rescued in p16Ink4a-/-/kl/kl
mice. This was due to de-repression of p16Ink4a-induced E2F1- and E2F3-mediated
down-regulation of residual Klotho expression (89). This is in line with p16Ink4a-/-/Klotho-/- mice being
phenotypically identical to Klotho-/- mice, although the vascular phenotype in these mice was
not specifically disclosed. The study by Eren et al. demonstrates that, although high phosphate and calcium levels are probably required for the development of calcifications, other factors are also capable of influencing the phenotype. Apparently, modulating cellular susceptibility to noxious stimuli may impede the development of vascular calcification, without directly altering phosphate and calcium levels. Another argument for this less phosphate-centric view is a study in which a central role was proposed for µ-calpain overactivation in the development of kl/kl phenotypes (55). Their data show that µ-calpain
33
inhibitor BDA-410, administered intraperitoneally at a dose of 100 mg/kg/day between 2 and 6 weeks of age, completely prevented the development of aortic calcification and associated arterial wall thickening at 6 weeks of age. Hyperphosphatemia, hypercalcemia, and hypervitaminosis D were comparable to kl/kl levels. This was apparently also the case if mice were treated between 4 and 6 weeks of age. Aortic mRNA levels of FGF23, Runx2, osteopontin, and RANKL were reduced after BDA-410 treatment, suggesting prevention of osteochondrogenic differentiation of smooth muscle cells. Cortical calcifications in the kidney, however, were still present to a minor extent and were hypothesized to develop due to dysregulated calcium re-absorption, rather than due to a process similar to vascular calcification.
Interestingly, rapamycin-induced inhibition of mTOR signaling (1.2 mg/kg/day between 3 and 4 weeks of age), while blunting induction of Cbfa1 expression, greatly ameliorated vascular calcification in CKD mice, but not at all in Klotho-/- mice (or ex vivo in Klotho-/- aorta rings) (57).
This cements Klotho as a key downstream mediator of rapamycin. Targeting a different pathway, it was reported that partial ablation of insulin receptor substrate (IRS) in IRS
+/-/Klotho-/- mice also prevents the development of vascular calcification, probably due to
inhibition of IGF1- and insulin-induced senescence. The previous discussion endorses the view that dysregulation of mineral homeostasis in Klotho deficiency is only part of the pathogenesis and that Klotho has a profound effect on pathways that regulate apoptosis and cellular senescence. Deficiency of Klotho may render cells susceptible to phosphate toxicity. The emergent paradigm for contributors to Klotho deficiency-induced vascular calcification is depicted in Figure 2.
Miscellaneous interventions
It was noted that leptin-deficient ob/ob/Klotho-/- mice still displayed vascular calcification, but
it is unclear how the phenotype compares to Klotho-/- mice (90). This would be interesting to
address since it could be hypothesized that vascular lesions may be aggravated due to leptin deficiency in addition to Klotho deficiency. High-fat diet-induced vascular calcification in kl/+ mice was ameliorated by AMP-activated protein kinase activator AICAR (91). Treatment of Klotho-/- mice with rikkunshito, aiming to increase ghrelin signaling, did not result in an effect
on vascular calcification(92), suggesting that ghrelin signaling may not be of relevance in vascular calcification in Klotho deficiency.
34
Effects of Klotho overexpression/supplementation on vascular calcification
A study by Hu et al., using the transgenic KL-Tg mice originally described by Kuro-o et al., found that Klotho overexpression causes mice with CKD (induced by nephrectomy + ischemia/reperfusion injury) to display very little or no vascular calcification and lower aortic calcium content than WT CKD mice (3). There are multiple possible mechanistic explanations for this. Firstly, these mice exhibit less severe vascular calcification, because CKD in these mice is relatively mild and creatinine levels do not rise significantly. Secondly, the lack of increase in phosphate serum level and fractional excretion due to phosphaturic actions of Klotho may also render these mice less prone to developing vascular calcification. Thirdly, the higher serum Klotho levels may have a direct protective effect on the vasculature. Corrected for serum phosphate and creatinine, however, the KL-Tg mice still have the lowest calcium content, so there is an additional vasculoprotective effect mediated by Klotho. Overexpression of Klotho down-regulated Pit1, Pit2, and Runx2, while up-regulating SM22α in KL-Tg aorta as compared to WT controls, possibly reflecting Klotho-driven smooth muscle cell differentiation towards a contractile phenotype. Using a
Figure 2. Paradigm of contributing factors in Klotho deficiency-induced vascular calcification. Vitamin D causes
high calcium and phosphate levels, which contribute to the induction of vascular calcification, in smooth muscle cells that have undergone osteochondrogenic transition and are senescent due to plasminogen activator inhibitor (PAI)-1 overexpression, µ-calpain overactivation, and increased insulin signaling. Anti-calcification mechanisms are activated, but are overwhelmed by the effects of Klotho deficiency.
35
Table 1. Effects of interventions on vascular calcification in models of Klotho deficiency
Intervention Animal model Effect on phenotype Reference
Constitutive Klotho gene expression
EF1α-KL/kl/kl mice Prevention of VC Kuro-o, 1997 Re-introduction of Klotho gene
expression
Ad-KL Kl/kl mice Rescue of VC Shiraki-Iida, 2000 Early induction of Klotho gene
expression
MTm-KL Kl/kl mice Prevention of VC Masuda, 2005 Late induction of Klotho gene
expression
MTm-KL Kl/kl mice Rescue of VC Masuda, 2005 Cessation of induction of Klotho
gene expression
MTm-KL Kl/kl mice Re-development of VC
Masuda, 2005 Klotho protein i.p. Kl/kl mice + Klotho i.p. Reduction of VC Chen, 2012 Whole nephron Klotho deletion Six2-Cre/KL-LoxP mice Development of VC Lindberg, 2014 Distal tubule Klotho deletion Ksp-Cre/KL-LoxP mice Prevention of VC Olauson, 2013 Proximal tubule Klotho deletion Kap-Cre/KL-LoxP mice Prevention of VC Ide, 2016 Proximal tubule Klotho deletion PEPCK-Cre/KL-LoxP mice Prevention of VC Ide, 2016 Proximal tubule Klotho deletion SLC4a1-Cre/KL-Loxp mice Prevention of VC Ide, 2016 Arterial Klotho deletion SM22a-Cre/KL-LoxP mice Prevention of VC Lindberg, 2013 Parathyroid Klotho deletion PTH-Cre/KL-LoxP mice Prevention of VC Olauson, 2013 Dietary phosphate restriction Kl/kl mice + low phosphate
diet
Reduction of EC Morishita, 2001 Genetic phosphate reduction NaPi2a-/-/Klotho-/- mice Prevention of VC Ohnishi, 2009
Genetic phosphate reduction NaPi2a-/-/Klotho-/- mice Prevention of VC Ohnishi, 2010
Genetic phosphate reduction + dietary phosphate overload
NaPi2a-/-/Klotho-/- mice +
high phosphate diet
Development of VC Ohnishi, 2010 Dietary phosphate restriction Klotho-/- mice + low
phosphate diet
Development of EC Tsujikawa, 2003 Dietary calcium restriction Klotho-/- mice + low calcium
diet
Prevention of EC Tsujikawa, 2003 Deletion of Sfrp4 Sfrp4-/-/Klotho-/- mice Development of VC Christov, 2011
Deletion of DMP1 DMP1-/-/Klotho-/- mice Exacerbation of VC Rangiani, 2012
Dietary vitamin D restriction Klotho-/- mice + low vitamin
D diet
Reduction of VC Tsujikawa, 2003 Genetic vitamin D reduction Cyp27b1-/-/Klotho-/- mice Prevention of VC Ohnishi, 2009
Genetic vitamin D reduction Cyp27b1-/-/Klotho-/- mice Prevention of EC
Woudenberg-Vrenken, 2012 Genetic ablation of vitamin D
signaling
VDRΔ/Δ/Klotho-/- mice Prevention of VC Anour, 2012
Dietary calcium, phosphate, and vitamin D restriction
Klotho-/- mice + low Ca, P,
vitamin D diet
Reduction of EC Alexander, 2009 Deletion of FGF23 FGF23-/-/Klotho-/- mice Development of VC Nakatani, 2009
Supplementation of FGF23 Klotho-/- mice + FGF23
injection
Unaltered
hyperphosphatemia
Nakatani, 2009 Deletion of PHEX Hyp/Klotho-/- mice Development of VC Nakatani, 2009
36
Deletion of PTH PTH-/-/Klotho-/- mice Development of EC Yuan, 2012
Suppletion of PTH Klotho-/- mice + PTH infusion Normophosphatemia Yuan, 2011
Inhibition of aldosterone signaling Klotho-/- mice +
spironolactone
Reduction of VC Voelkl, 2013 Homoarginine treatment Kl/kl mice + 20 mg/L
homoarginine in drinking water
Exacerbation of VC Alesutan, 2016
Lysosomal alkalinization Kl/kl mice + 0.28 M NH4Cl Prevention of VC Leibrock, 2015
Lysosomal alkalinization Kl/kl mice + 0.28 M NH4NO3 Reduction of VC Leibrock, 2016
Induction of acidosis Kl/kl mice + acetazolamide Prevention of VC Leibrock, 2015 Induction of alkalosis Kl/kl mice + 0.15 M NaHCO3 Reduction of EC Leibrock, 2015
Partial cyclooxygenase 2 deletion Ptgs2+/-/Klotho-/- mice Reduction of AVC Wirrig, 2015
Full cyclooxygenase 2 deletion Ptgs2-/-/Klotho-/- mice Reduction of AVC Wirrig, 2015
Cyclooxygenase 2 inhibition Klotho-/- mice + celecoxib Reduction of AVC,
development of VC
Wirrig, 2015
Partial deletion of PAI-1 PAI+/-/kl/kl mice Reduction of EC Eren, 2014
Full deletion of PAI-1 PAI-/-/kl/kl mice Prevention of EC Eren, 2014
Deletion of p16Ink4a p16Ink4a-/-/kl/kl mice Prevention of EC Sato, 2015
Deletion of p16Ink4a p16Ink4a-/-/Klotho-/- mice Development of EC Sato, 2015
Inhibition of µ-calpain Klotho-/- mice + BDA-410 i.p. Prevention of VC Nabeshima,
2014 Inhibition of mTOR signaling Klotho-/- mice + rapamycin Development of VC Zhao, 2015
Partial ablation of insulin signaling IRS+/-/Klotho-/- mice Prevention of VC Kurosu, 2005
Deletion of leptin Ob/ob/Klotho-/- mice Development of VC Ohnishi, 2011
AMPK activation Kl/+ mice + high-fat diet Prevention of VC Lin, 2016 Activation of ghrelin signaling Klotho-/- mice + 1000
mg/kg/day (p.o.) rikkunshito
Development of VC Fujitsuka, 2016
VC, vascular calcification; EC, ectopic calcification; AVC, aortic valve calcification
pharmacological rather than a genetic approach, concurrently elevated Klotho expression levels induced by rapamycin, vitamin D receptor agonists and the calcitonin gene-related peptide family member intermedin1-53 have also been found to improve vascular calcification
in WT mice. This suggests that the beneficial effects of these compounds are partially Klotho-dependent (57, 93, 94).
37
Intima hyperplasia
Intima hyperplasia in Klotho deficiency
The other important arterial feature of Klotho deficiency is intima hyperplasia (usually the result of de-differentiation of smooth muscle cells, followed by proliferation and migration to the intima). It was described to be present only in middle-sized muscular arteries (1), whereas one group notes hyalinous sclerosis of interlobar arteries in 6-week-old kl/kl mice (95). These hallmarks have gained virtually no attention in later studies and their characteristics have therefore not been reported.
Klotho overexpression and intima hyperplasia
Experiments aiming to influence the development of intima hyperplasia in Klotho deficiency have not been published to date, but Klotho overexpression and KL1 supplementation have been employed in other models. Data from a study in streptozotocin-injected rats on a high-fat diet seem to indicate that Klotho gene delivery prevents intima hyperplasia in this model (96). Furthermore, a recent study, examining the effect of soluble Klotho (0.01 mg/kg, 5 days/week, between 12 and 20 weeks of age) on the development of atherosclerosis in ApoE -/- mice, found no effect of Klotho on atherosclerotic lesion area (97). This may be due to the
timing of the intervention with respect to plaque formation or a difference in pathogenesis of atherosclerosis between the ApoE-/- model and other models, but blood pressure was also unexpectedly unaltered in these mice. On the whole, Klotho effects on intima hyperplasia are still to be investigated more elaborately.
Endothelial dysfunction
Endothelial dysfunction in Klotho deficiency
Klotho deficiency in mice is known to cause overactivation of µ-calpain by 2 weeks of age, leading to depletion of endogenous calpain inhibitor calpastatin by 3 weeks of age (at which point calcifications have not yet developed). Ultimately, this causes cleavage of calpain substrate αII-spectrin by 4 weeks of age, which would predispose cells to cell death (98). Akin to molecular changes in human aged arteries, aortas of kl/kl mice were also found to exhibit more calpain and capase 3 activity in zymography assays and calpain substrate αII-spectrin cleavage was shown to be increased significantly. Furthermore, in kl/kl aorta, a higher Bax/Bcl2 mRNA ratio was found and both endothelial cells (ECs) and SMCs exhibit an
38
increased rate of apoptosis (33, 56). Endothelial apoptosis leads to compromised integrity of the endothelium and it was shown that the kl/kl mouse endothelium is hyperpermeable, as assessed using Evans Blue dye in vivo, already in mice that were only 2 weeks old.
Compromise of endothelial integrity leads to endothelial dysfunction, which is another hallmark of human arterial ageing and leads to intima hyperplasia. This was also demonstrated in Klotho deficiency as impaired vasodilation upon acetylcholine stimulation in aortas (kl/+ mice) and arterioles (kl/kl and kl/+ mice) (72, 99). Aortas of kl/kl mice were too severely calcified to dilate in response to acetylcholine. Also noted were decreased sensitivity (higher ED50) to acetylcholine and lower urinary excretion of NO2- and NO3- in kl/+ mice, a
finding that was later also reported in kl/kl mice (100) and that is unsurprising, given the lower bioavailability of NO in aging. There was no difference in arterial dilatation or ED50 upon
stimulation with sodium nitroprusside, indicating that the impaired vasodilation in Klotho deficiency is endothelium-dependent and at least in part due to impaired endogenous NO production. Further evidence for this conclusion is the lack of a difference in urinary NO2- and
NO3- after NOS inhibitor (L-NAME) treatment. In a follow-up study, these findings were
replicated and expanded upon: after administration of L-NAME, differential Klotho allele status caused no difference in relaxation. This is in line with the notion of Klotho deficiency-induced impairment of endothelial NO production as evidenced by decreased relaxation of kl/+ aortas upon stimulation with lecithinized superoxide dismutase (which protects NO from degradation). In addition, reduced NO synthase and VEGF protein expression levels are found in kl/+ and kl/kl aorta (101). Lower eNOS and higher levels of senescence-associated genes PAI-1, TGF-β, p21, and SA-β-galactosidase were noted to be up-regulated in kl/kl aorta (53, 54). Increased PAI-1 mRNA expression and increased PAI-1 protein levels were also described in both smooth muscle and endothelial cells (and kidney) (95). Increased PAI-1 expression in Klotho deficiency was found to be associated with senescence and increased microvascular thrombosis in kidney, and telomere shortening in aorta, as reported by others (88). The presence of increased PAI-1 serum levels and PAI-1 expression in sclerotic vascular lesions in Klotho deficiency, as in aging human arteries, is hypothesized to contribute to thrombosis and extracellular matrix protein accumulation by inhibition of proteolysis. In short, the phenotype of Klotho deficiency also features endothelial dysfunction due to impaired NO synthesis, increased PAI-1 expression, and vascular hyperpermeability due to increased µ-calpain activity. In terms of vascular senescence and deterioration of endothelial function, the molecular aspects of Klotho deficiency appear to be a valid model for human arterial ageing.
Interventions in the development of endothelial dysfunction
The parabiosis experiment described by Saito et al. is the only experiment in which endothelial function was used as the outcome measure of an intervention in Klotho deficiency (72). In this experiment, it was shown that when the circulatory systems of kl/+ mice and WT mice are connected for 4 weeks, acetylcholine-dependent vasodilation in kl/+ mice is
39
restored. This indicates that soluble factors, differentially present in Klotho deficiency, are capable of rescuing endothelial dysfunction in this model and circulating Klotho itself would be a prime candidate for this hypothesis.
Klotho overexpression and endothelial dysfunction
An early study explored the effects of Klotho gene transfer in OLETF rats, which significantly improved endothelial function after three weeks, as evidenced by an increase in acetylcholine-induced relaxation (62% compared to 82%) (102). This difference was abolished by L-NAME treatment, suggesting that it was due to a difference in endogenous NO production. Indeed, Klotho gene delivery significantly increased the urinary excretion of NO metabolites. Furthermore, systolic blood pressure and coronary perivascular fibrosis and medial hypertrophy were reduced in Klotho-transfected animals. Another group found that 7 days after gene delivery, Klotho leads to an increase in manganese superoxide dismutase (Mn-SOD) protein levels and activity in mice and in spontaneously hypertensive rats (SHR) (103). This effect was abolished in rats by L-NAME treatment, while Klotho also increased NO metabolite production and decreased lipid peroxide production in mice. However, systolic blood pressure in SHR was not affected by Klotho overexpression, possibly due to the short study period. Taken together, Klotho appears to be a promising candidate for the prevention of endothelial dysfunction.
Arterial stiffening
Arterial stiffening in Klotho deficiency
Not wholly unrelated to endothelial dysfunction and hypertension, kl/+ mice develop arterial stiffening, as is evident from the higher arterial pulse wave velocity at least from 14 weeks of age onwards (104). It is likely that arterial stiffening has not yet fully developed at 6 to 9 weeks of age, at which point kl/+ mice are known to still have a normal sodium nitroprusside response (but already an increased norepinephrine-induced contraction in kl/+ aorta, which was abolished by L-NAME treatment) (101). Histologically, the aorta (but not smaller arteries) was found to contain more collagen-I and less elastin, as well as higher matrix metalloproteinase (MMP)-2, MMP-9, and TGF-β1 expression, all features very similar to the development of age-related arterial stiffening in humans (91, 104). Moreover, Klotho deficiency was associated with increased expression of aldosterone-induced collagen transcription factor scleraxis, with autophagy activation (increased LC3-II/LC3-I ratio), and with the presence of more aortic α-SMA-positive cells. These findings, combined with the development of arterial stiffening preceding the development of hypertension, were taken to
40
mean that arterial stiffening was a primary, rather than secondary process. Since the kl/+ mice do not exhibit such severe spontaneous vascular calcifications as found in the full knockout, kl/+ mice may be a suitable model for normal, mild, human ageing, apparently also exhibiting arterial stiffening, which is a major hallmark of human ageing.
Interventions in the development of arterial stiffening
Arterial stiffening in kl/+ mice was found to be improved by eplerenone treatment, which also prevented increased collagen-I deposition, elastin degradation, MMP-2 and MMP-9 expression, and myofibroblast differentiation (104). Arterial stiffening in kl/+ mice was found to develop in as little as 5 weeks on a high-fat diet (HFD) (91). Up-regulation of aortic collagen-I, Runx2, and TGFβ1 levels (but not of tropoelastin levels) was also exacerbated by HFD, while phosphorylated AMP-activated protein kinase (pAMPKα), p-eNOS, and Mn-SOD levels were decreased. Activation of cAMP-dependent AMPK by AMP analogue AICAR reversed the HFD-induced incease in pulse wave velocity, as well as the baseline increase in pulse wave velocity. Phosphorylation of eNOS, urinary nitrate/nitrite, and the expression level of Mn-SOD were also increased by AICAR treatment, and superoxide, collagen-I, Runx2, TGFβ1 levels, and number of elastin breaks were decreased in aorta, without change in tropoelastin levels. It would be very interesting to assess the effect of Klotho overexpression on arterial stiffening, which has not been published to date.
Hypertension
Blood pressure in Klotho deficiency
It was found by Zhou et al. that kl/+ mice develop hypertension from 15 weeks of age onward, as well as increased salt sensitivity, as shown by an additional increase in systolic blood pressure by a 2% saline diet (105). Chen et al. corroborated the increase in blood pressure in kl/+ mice from 16 weeks of age onward, which was preceded by arterial stiffening (104). In another study, the same group re-affirmed the development of hypertension and linked it to higher aldosterone levels, with coincident higher adrenal CYP11B2 expression in kl/+ mice (106). Kl/kl mice on a normal diet were shown to have a lower blood pressure than kl/kl mice after 4 weeks of having been on a low vitamin D, which in turn had a lower blood pressure than WT mice (107). Klotho-/-/VDRΔ/Δ mice on a vitamin D/phosphate/calcium rescue diet
were also shown to have a lower mean arterial pressure and blood volume than WT or VDRΔ/Δ
mice (32). These seemingly contradictory findings can be explained by the Klotho-dependent effect of FGF23 to stimulate Na+/Cl--co-transporter (NCC)-mediated sodium re-absorption in
41
hyperaldosteronism, extracellular volume depletion, and increased ADH levels in Klotho deficiency (107). However, Zhou et al. report higher NCC levels in Klotho+/- mice that also have
hyperaldosteronism (105) and Andrukhova et al. report lower NCC levels in Klotho-/- mice (32).
A few pieces of the puzzle with regard to Klotho, sodium handling, and blood pressure are still to be clarified.Another discrepancy with that mechanism would be that at 8-9 weeks of age, kl/kl mice were still found to have a lower systolic blood pressure than WT mice, without responding to spironolactone treatment (53).
Interventions in the development of hypertension
Zhou et al. detail that the increased salt sensitivity in kl/+ mice, manifested as an additional increase in blood pressure after 4 weeks of a 2% saline diet, could be reversed by selective CC chemokine receptor 2 (CCR2) inhibitors INCB3284 and RS102895 (105). However, baseline hypertension due to partial Klotho deficiency was unaltered. This indicates that the monocyte chemoattractant protein 1 (MCP-1)/CCR2 pathway may be involved in Klotho deficiency-induced increased salt sensitivity, but not in the initial development of hypertension. Targeting the increased aldosterone signaling present in Klotho deficiency using eplerenone, however, does seem to affect the Klotho decrease-induced hypertension, reducing blood pressure to WT levels (106). Also, both baseline hypertension and high-fat diet-induced exacerbation of hypertension in kl/+ mice could be normalized by AMPK activator AICAR (91).
Klotho overexpression and hypertension
Saito et al. first found that systolic blood pressure was reduced in OLETF rats after Klotho gene transfer (102). Klotho gene delivery has also been shown by Wang et al. to prevent the development of hypertension in SHR during the study period of 12 weeks after gene delivery (108). Klotho overexpression in SHR was associated with less oxidative stress and down-regulation of Nox2 protein expression and NADPH oxidase activity, indicative of an anti-oxidative state. In another study, Klotho overexpression normalized hypertension induced by 5/6th nephrectomy as well (109). A study using recombinant Klotho protein, however, did not
find an effect on blood pressure in ApoE-/- mice after 3 weeks of treatment, nor did a study
42 Klotho and blood pressure regulation
Given the complex regulation of blood pressure, the findings of hypertension in (partial) Klotho deficiency and Klotho-induced reduction of blood pressure in spontaneous hypertension warrant a more elaborate background for discussion. Renal Klotho gene expression has been uniformly shown to be decreased in many experimental models of circulatory stress, including SHR (99, 108, 110-112), DOCA salt-sensitive hypertensive rats (99, 110), hypertensive/diabetic OLETF rats (99, 110, 113), angiotensin II infusion in rats (114), and ApoE-/- mice (115). The mechanism, however, behind this down-regulation is not clear and
probably involves many different factors acting in concert, including angiotensin II, aldosterone, oxidative stress, and pro-inflammatory factors. It is unlikely that Klotho is down-regulated in a direct response to hypertension, since norepinephrine infusion in rats for 7 days did not down-regulate Klotho (116). Indirectly, however, long-term hypertension-induced renal damage probably does down-regulate Klotho expression.
Klotho has been shown to modulate many determinants of blood pressure, including the left ventricular cardiac muscle, which is prone to hypertrophy in Klotho deficiency and is protected by Klotho overexpression (16, 17, 117-119). Furthermore, the aortic valve exhibits hinge region calcification and fibrosis in Klotho deficiency (58, 87, 120, 121), and of course arterial resistance is increased in the setting of extensive arterial calcification. The endothelin system has been shown to be downstream of Klotho, since silencing of CNS Klotho using shRNAs in the cerebrospinal fluid potentiated cold-induced hypertension, which was prevented by concurrently silenced endothelin-1 (122). Endothelin plasma levels were later found to be down-regulated by Klotho overexpression, while concurrently preserving ETB
receptor levels (123). The baroreflex blood pressure regulation mechanism in SHR has been shown to be impaired if Klotho is silenced and is restored by recombinant Klotho (124). In a similar study, recombinant Klotho or rosiglitazone administration, reverting down-regulation of Klotho, in streptozotocin-induced diabetes in rats was shown to improve the diabetes-impaired baroreflex (125).
Klotho and cardiovascular risk management
The observation that many of the medicines that are used to reduce the risk of cardiovascular events seem to influence Klotho expression is especially interesting. Many studies have identified a link between the renin-angiotensin-aldosterone system and Klotho. Klotho was shown to regulate CYP11B2 transcription factors SF-1 and ATF2, likely resulting in inhibition of adrenal aldosterone synthesis (106). Klotho also down-regulates the expression of angiotensinogen, renin, angiotensin converting enzyme, and angiotensin II receptor type 1, by blocking Wnt signaling and the consequent Wnt-induced up-regulation of the aforementioned RAAS genes (109). Conversely, angiotensin II has been shown to down-regulate Klotho even in the absence of hypertension (114). Direct or indirect angiotensin
II-43
mediated down-regulation of Klotho has been demonstrated to be restored by ACE inhibitors (111, 112, 126-128), angiotensin II receptor antagonists (111, 112, 114, 126, 127, 129, 130), renin inhibitors (131), and aldosterone receptors antagonists (132). The pressor-independent nature of this effect is further supported by similar findings in renal tubular epithelial cells in vitro, while identifying RhoA activation as a possible pathway by which angiotensin II suppresses Klotho (133, 134). Similarly, statins have been shown to up-regulate Klotho, possibly through RhoA inactivation (134-136), whereas PPAR-γ agonists even cause a direct PPAR-γ-mediated increase in transcriptional activity of the Klotho gene (113, 125, 137-139).
Impaired angiogenesis and vasculogenesis
Impaired angiogenesis and vasculogenesis in Klotho deficiency
As noted before, kl/kl mice exhibit decreased VEGF expression and NO production in aorta (101), raising the question whether angiogenesis and vasculogenesis may be impaired in kl/kl mice. Indeed, Fukino et al. first described that heterozygous kl/+ mice display impaired angiogenesis in response to unilateral hind limb ischemia (140). Laser Doppler perfusion-assessed blood flow was significantly higher in WT mice between 7 and 14 days after induction of ischemia, after which the blood flow became statistically comparable again between groups (although the kl/+ mice never reached WT levels). Kl/+ mice also exhibited decreased capillary density 5 weeks after surgery. A similar experiment was described by Shimada et al., also performing unilateral hind limb ischemia, in WT, kl/+, and kl/kl mice (100). The authors describe significantly reduced blood flow in kl/+ mice as compared to WT mice, as well as in kl/kl mice as compared to kl/+ mice, between 3 days after surgery and the end of the study period (28 days after surgery). Capillary density was also significantly reduced in both kl/+ and kl/kl mice. Ex vivo aortic ring microvessel sprouting was also impaired in kl/kl mice and these mice had fewer EPC-like progenitor cells in both bone marrow and peripheral blood. Concurrent homologous bone marrow transplantation (WT to WT and kl/kl to kl/kl, which would be interesting heterologously) and hind limb ischemia revealed less engraftment of CD31+ or vWF+ donor cells after 14 days, suggestive of impaired vasculogenesis in Klotho
deficiency. Furthermore, spontaneous limb amputation occurred more frequently in kl/kl animals than in WT animals. It should be noted that Klotho does not appear to have uniformly pro-angiogenic properties. For example, the retinal choroid layer of kl/kl mice is severely deformed with dilated blood vessels and Klotho may inhibit VEGF secretion from the retinal pigment epithelium, essentially exerting anti-angiogenic effects (141). In short, Klotho deficiency is associated with impaired angiogenesis and vasculogenesis, but the effects may differ per tissue or organ.
44 Interventions in angiogenesis
There are two reports on interventions aimed at modulating angiogenesis in Klotho deficiency. Shimada et al. have performed unilateral hind limb ischemia in kl/kl mice with and without cerivastatin (s.c., 5 mg/kg/day for 28 days after surgery), thought to induce vasodilation and enhance NO synthase activity (100). Laser Doppler blood flow was increased as compared to untreated animals, as were capillary density, ex vivo aortic sprouting, and the number of EPCs in bone marrow and peripheral blood. Amputation rate was also significantly improved by cerivastatin treatment. Arima et al. found that VEGF stimulation can ameliorate the impaired angiogenesis of Klotho deficiency, which is further improved by antibody-mediated inhibition of pigment epithelium-derived factor (PEDF) (142). PEDF mRNA was found to be up-regulated in response to VEGF and PEDF administration was found to completely abolish VEGF-induced angiogenesis. It is unknown whether PEDF is up-regulated in kl/kl mice as compared to WT mice. Experiments assessing the effect of Klotho overexpression on angiogenesis have not been published to date.
45
Table 2. Effects of interventions on other vascular phenotypes in models of Klotho
deficiency
Intervention Animal model Effect on phenotype Reference
Parabiosis Kl/+ and WT mice Improvement of endothelial function
Saito, 1998
Inhibition of aldosterone signaling
Kl/+ mice + eplerenone Prevention of arterial stiffening Chen, 2015
High-fat diet Kl/+ mice + high-fat diet Acceleration of development of arterial stiffening
Lin, 2016 AMPK activation Kl/+ mice + AICAR Prevention of arterial stiffening Lin, 2016 AMPK activation Kl/+ mice + high-fat diet +
AICAR
Prevention of arterial stiffening Lin, 2016 High-fat diet Kl/+ mice + high-fat diet Acceleration of development
of hypertension
Lin, 2016 AMPK activation Kl/+ mice + AICAR Prevention of hypertension Lin, 2016 AMPK activation Kl/+ mice + high-fat diet +
AICAR
Prevention of hypertension Lin, 2016 Inhibition of aldosterone
signaling
Kl/+ mice + eplerenone Normalization of blood
pressure
Zhou, 2015 Salt loading Kl/+ mice + 2% saline
drinking water
Aggravation of hypertension Zhou, 2015 CCR2 inhibition Kl/+ mice + high salt diet +
INCB3284
Normalization of increased salt sensitivity, unaltered hypertension
Zhou, 2015
CCR2 inhibition Kl/+ mice + high salt diet +
RS102895
Normalization of increased salt sensitivity, unaltered hypertension
Zhou, 2015
Statin treatment Kl/kl mice + unilateral hind
limb ischemia + cerivastatin
Improvement in angiogenesis Shimada, 2004 VEGF treatment Kl/kl mice + unilateral hind
limb ischemia + VEGF
Improvement in angiogenesis Arima, 2010 VEGF treatment + PEDF
inhibition
Kl/kl mice + unilateral hind
limb ischemia + VEGF + PEDF antibodies
Increased improvement in angiogenesis
Arima, 2010
VEGF treatment + PEDF Kl/kl mice + unilateral hind
limb ischemia + VEGF + PEDF
Impaired angiogenesis Arima, 2010
AMPK, 5’ adenosine monophosphate-activated protein kinase; AICAR, 5-aminoimidazole-4-carboxamide ribinucleotide VEGF, vascular endothelial growth factor; PEDF, pigment epithelium-derived factor.
46
Endogenous vascular Klotho expression
Having assessed all reported in vivo effects, Klotho appears to be a very promising factor to target. A logical first step would be to assess whether the vasculature is best targeted by increasing serum levels of Klotho, or by up-regulating endogenous Klotho expression. This endeavor, however, will immediately encounter disputed and conflicting evidence. Having established that the effects of Klotho on the vasculature appear to be mediated predominantly by soluble Klotho, the potential existence of endogenous vascular Klotho expression is still the most controversial topic in the current field of vascular Klotho biology.
Endogenous Klotho in aorta and other arteries
The first hint at possible vascular Klotho expression can be found in the original study by Kuro-o et al. (1). A NKuro-orthern blKuro-ot fKuro-or KlKuro-othKuro-o prKuro-oduced a faint band in mKuro-ouse aKuro-orta that cKuro-ontrasted markedly with high renal and cerebral Klotho expression, being the faintest band of all organs that exhibit Klotho expression and being slightly exceeded by putative expression in colon and thyroid gland. A follow-up study by the same group, however, using RT-PCR on mouse aorta instead of Northern blotting, failed to show expression of alternatively spliced Klotho mRNA (143). The expression of alternatively spliced Klotho is usually predominant over membrane-bound Klotho mRNA in mouse by a factor 10 and the expression pattern is otherwise consistent with the expression pattern of membrane-bound Klotho mRNA. The first protein data consisted of induction of expression in aorta on WB at the same weight as renal Klotho after, but not before, Klotho gene transfer (114). Immunoreactivity with proteins of higher and lower molecular weight, however, was detected in aorta, heart, and liver, which may reflect the existence of different Klotho proteins, but may also reflect aspecific immunoreactivity of anti-rat IgG secondary antibodies on rat tissue lysates. Mice with Klotho promoter-driven X galactosidase expression revealed endogenous Klotho expression in renal distal tubules, parathyroid gland, and brain, as expected (95). Sino-atrial node Klotho expression was an interesting new finding in this study, but arterial Klotho expression was not observed. It is possible however, that the expression level is lower than could be detected by this method. This interpretation is endorsed by three studies from another group, who have only used RT-PCR on human aorta, generally detecting a low expression level of Klotho (144-146). This is in line with our own data, yielding just barely detectable levels of Klotho mRNA in various human arteries (147). However, in various other studies, Klotho protein and mRNA could not be detected in rat or mouse aorta using various techniques (71, 93, 108, 148, 149). Interestingly, RNA-Seq analysis of normal and uremic rat aorta did not yield values for Klotho above the commonly used threshold of 1 RPKM (1 read/kilobase transcript/million mapped
