.
F U L L T E X T S O U R C E : Cell Reports
Highlights
• Zebrafish αklotho mutants display reduced lifespans
• The αklotho phenotype occurs later in zebrafish than in mice
• Zebrafish αklotho mutants display adult-onset vascular calcification
• Calcification coincides with an increase in osteoclast differentiation pathways
Summary
The hormone αKlotho regulates lifespan in mice, as knockouts die early of what appears to be accelerated aging due to hyperphosphatemia and soft tissue calcification. In contrast, the overexpression of αKlotho increases lifespan. Given the severe mouse phenotype, we generated zebrafish mutants for αklotho as well as its binding partner fibroblast growth factor-23 (fgf23). Both mutations cause shortened lifespan in zebrafish, with abrupt onset of behavioral and degenerative physical changes at around 5 months of age. There is a calcification of vessels throughout the body, most dramatically in the outflow tract of the heart, the bulbus arteriosus (BA). This calcification is associated with an ectopic activation of osteoclast differentiation pathways. These findings suggest that the gradual loss of αKlotho found in normal aging might give rise to ectopic calcification.
Introduction
Systemic factors that regulate aging are of interest due to their potential as novel drug targets in preventing or slowing down age-related decline in animal health. αKlotho, a molecular scaffold protein, is considered an anti-aging hormone that regulates mineral homeostasis in mammals (Chen et al., 2018, Kuro-o, 2013, Kuro-o et al., 1997, Kurosu et al., 2006, Kurosu et al., 2005, Lindberg et al., 2014, Shimada et al., 2004). It is one of the few systemic secreted factors whose loss is sufficient to induce premature morbidity and mortality that resembles accelerated aging (Kuro-o et al., 1997), and its overexpression extends lifespans (Kurosu et al., 2005). It is therefore of interest to understand the cellular and molecular mechanisms by which αKlotho regulates the aging process.
The mouse knockout models of αklotho are difficult to study; animals die by ∼8−12 weeks of age and are difficult to maintain (Fernández et al., 2018, Kuro-o et al., 1997; unpublished data). αklotho loss-of-function mice develop normally until about 3 or 4 weeks of age and then begin to display age-related conditions, including ectopic calcification, arteriosclerosis, osteoporosis, and reduced lifespans (Kuro-o et al., 1997). It was suggested that the extended lifespan in mice overexpressing αklotho is due to a suppression of insulin and the insulin-like growth factor-1 signaling (Kurosu et al., 2005), although it is likely that other pathways are involved. Recently, it was shown that an increase in autophagy levels could delay or prevent early mortality in αklotho mutant mice (Fernández et al., 2018). αKlotho acts as a co-receptor for fibroblast growth factor-23 (FGF23) (Urakawa et al., 2006). Knockouts of fgf23 have a similar accelerated aging phenotype to that of αklotho and show a perturbation in vitamin D metabolism (Shimada et al., 2004). αKlotho can also be released from the cell surface. In such settings, it can still heterodimerize with FGF23, functioning as a co-ligand in forming a high-affinity activator of FGF receptor signaling (Erben, 2018).
Fish diverged from tetrapods approximately 400 million years ago (Daeschler et al., 2006, Romer, 1967). There are fundamental differences in physiology between fish and terrestrial vertebrates owing to unique demands of aquatic versus terrestrial environments. In terms of renal function and mineral and fluid homeostasis, they have evolved different physiologies. For example, renal function in freshwater fish serves partly to prevent overhydration, whereas in mammals, it is designed to prevent dehydration. We therefore examined if the roles of αKlotho, a renal hormone, would be conserved in aging and mineral homeostasis. The zebrafish also provide a tractable system for measuring certain behaviors, including physical activity. The zebrafish genome encodes one αklotho and one fgf23 (Mangos et al., 2012, Sugano and Lardelli, 2011). Consistent with mammalian studies, zebrafish αklotho expression is detected in multiple organs, including adult kidneys (Mangos et al., 2012). Zebrafish fgf23 is expressed in corpuscles of Stannius, a teleost-specific, kidney-associated endocrine gland involved in mineral homeostasis (Elizondo et al., 2010, Mangos et al., 2012). We generated knockouts of αklotho and fgf23 in zebrafish in order to understand the mechanistic link between aging and the αKlotho/FGF23 pathway.
Results
Zebrafish Mutants in Both αklotho and fgf23 Have Early-Onset Mortality
We targeted the αklotho gene using the CRISPR/Cas9 method (Irion et al., 2014), generating mutations in two background zebrafish strains, Tü and AB, and identified αklotho alleles carrying frameshift mutations and early stop codons (Table S1). We also targeted fgf23 because the function of αKlotho in mammals depends in large part upon its binding to FGF receptors and recruiting FGF23 to activate FGF signaling (Chen et al., 2018, Kurosu et al., 2006).
We find that αklotho−/− and fgf23−/− mutant zebrafish display essentially indistinguishable phenotypes (Figure 1). As adults of about 5 months of age, they develop emaciated bodies, tattered fins, and an opaque overgrowth on the eyes (αklotho−/− mutant; Figures 1A, 1B, 1D, and 1E; fgf23−/− mutant; Figures 1C, 1F, and S1A–S1F). In addition, female αklotho−/− and fgf23−/− mutants displayed protruding eyes (Figures S1G–S1L).
(A–F) Body condition of αklotho and fgf23 mutant males at 5 mpf: (A) Tü wild-type strain, (B) αklotho (klΔ5), and © fgf23 (fgf23ins1) mutant in Tü background; (D) AB wild-type strain, (E) αklotho (klΔ5), and (F) fgf23 (fgf23Δ11) mutant in AB background.
(G–J) Survival curves for (G) αklotho (n = 36 background controls, 32 mutants; p < 0.0001) and (H) fgf23 mutants (n = 14 wild-type siblings, 14 mutants; p < 0.0001) in Tü background and for (I) αklotho (n = 24 wild-type siblings, 21 mutants; p = 0.0001) and (J) fgf23 mutants (n = 60 background controls, 68 mutants; p < 0.0001) in AB background. Log-rank (Mantel-Cox) test for statistical analysis on survival curves in GraphPad Prism.
(K and L) Analysis of speed (cm) in the (K) circular arena and (L) home-tank arena. Age (m; mpf) on x axis; n = number of fish.
(M and N) αklotho mutants and wild-type controls in (M) circular and (N) home-tank arena.
(O and P) fgf23 mutants and wild-type controls in (O) circular and (P) home-tank arena. Statistical analysis using unpaired t test in GraphPad Prism.
See also Figure S1 and Table S1.
The onset of mortality in mutant colonies began around 4−5 months post-fertilization (mpf; survival curves in Figures 1G–1J; p < 0.001), compared to wild-type strains of zebrafish, which live for 3−5 years (Carneiro et al., 2016, Gerhard et al., 2002). Both αklotho−/− and fgf23−/− mutant fish appear morphologically comparable to wild-type siblings at 2−3 mpf (Figures S1M–S1P) and are fertile as young adults, allowing us to breed homozygotes. Among adult progeny (∼3 mpf) obtained by inbreeding αklotho heterozygotes, we recovered homozygous mutants in ratios consistent with Mendelian inheritance. Among 281 siblings raised together until adulthood, we obtained 70 wild-type siblings (25%), 130 heterozygous siblings (46%), and 79 homozygous αklotho mutants (28%). Among 241 adult zebrafish obtained from breeding parents fgf23 heterozygotes, we obtained 66 wild-type siblings (27%), 128 heterozygous siblings (52%), and 50 homozygous mutants (20%). This indicates that αklotho−/− and fgf23−/− mutants have no survival disadvantage until adulthood, even when raised with wild-type siblings.
In order to probe the timing of more subtle aspects of physical decline, we analyzed the behavior of the αklotho−/− and fgf23−/− zebrafish in two settings: a circular arena (Figure 2K) and an arena resembling their home tank (Figure 2L). Although spontaneous behavior in zebrafish is intrinsically variable, αklotho−/− and fgf23−/− mutants demonstrated reduced activity in both behavioral paradigms at 5 and 6 mpf, corroborating the physical evidence of decline at this time (Figures 1M–1P).
H&E staining on paraffin sections from 5-month-old wild-type control (Tü) and αklotho (klΔ5) males. Shown are (A) skin (arrows indicate mucous cells in wild type); (B) muscle (arrows indicate vascular calcification); © gills; (D) heart (arrow indicates calcification in the BA); (D′ and D″) alizarin red-stained whole-mount hearts (arrows indicate the BA); (E) BA (arrow indicates calcification); (F) liver (arrows indicate bile-duct); and (G) kidney (arrows indicate glomeruli). (H) Bright-field images of 5-mpf wild-type (left) and klΔ5 (right) hearts. TRAP staining on (I) whole mount and (J) cryosection of 5-mpf hearts.
See also Figure S2.
We conclude from these data that there is an adult-onset, age-related decline in the body condition in both αklotho−/− and fgf23−/− mutants. Although it is difficult to align developmental frameworks between species, the adult-onset decline in both αklotho−/− and fgf23−/− mutant zebrafish appears to be proportionally later than described for the mouse αklotho mutants (Kuro-o et al., 1997; unpublished data).
Vascular Calcification and Inflammation across Organs in αklotho−/− Mutant Zebrafish
In order to understand the phenotype at the cell and tissue level, we performed comprehensive histopathological analysis by H&E staining on sections of 5-month-old wild-type and αklotho−/− mutant fish (Figure 2; N = 3 males each). In αklotho−/− fish, there was widespread calcification and inflammation. Within the integument of αklotho−/− fish, there was a reduction in the number of mucosal cells and necrosis in areas of the epidermis and dermis, with a mineralization of the dermal vasculature (Figure 2A). Furthermore, in αklotho−/− mutants, there was calcification of medium- to small-size blood vessels in the skeletal muscles (arrow in Figure 2B), accompanied by a degeneration and fibrosis of adjacent skeletal muscles with immune cell infiltration (Figure 2B). Mineralization was often in a concentric pattern in the affected areas. The gill arch demonstrated bone overgrowth (hyperostosis) with chondrodysplasia of the gill arch (Figure 2C) and a loss of normal architecture of filaments and lamellae due to blunting, fusion, and necrosis of the lamellae epithelium, along with immune cell infiltration.
In the αklotho−/− zebrafish, calcification was particularly striking within the walls of the bulbus arteriosus (BA) (the outflow tract of the heart) (Figures 2D and 2E). The BA is composed of smooth muscles and is lined by the endothelium, and its elasticity is believed to buffer pulsatile blood flow to the thin-walled capillaries of the gills (Farrell, 1979, Grimes and Kirby, 2009). To validate calcification in the BA, we used alizarin red, a stain for calcification (Walker and Kimmel, 2007). The BA in αklotho−/− mutants is prominently stained with alizarin red in contrast to wild-type animals (Figure 2D; insets), confirming calcification. Calcification was also observed in the bile duct of the livers of αklotho−/− fish (arrow in Figure 2F). The kidneys of αklotho−/− mutants appeared comparable to the wild-type controls (Figure 2G). Within the skeletal system, there were multifocal areas of hyperostosis and chondrodysplasia. These changes were prominent in the caudal region (Figure S2A). Frequently, regions of bone overgrowth were accompanied by areas of dystrophic calcification, connective tissue proliferation, and immune cell infiltration (Figure S2A).
Vascular calcification was the most prominent phenotype. In fact, unprocessed and unstained BAs appear opaque white in αklotho−/− (Figure 2H), indicating severe calcification. fgf23−/− mutants phenocopy αklotho−/− mutants—the BAs in fgf23−/− mutants are prominently stained with alizarin red, in contrast to wild-type animals (Figure S2B). It has been shown that calcification in zebrafish is accompanied by an increase in osteoclast activity (Apschner et al., 2014). Consistent with this, we observe strong Tartrate-resistant acid phosphatase (TRAP) staining in the outflow tract of the αklotho−/− mutant hearts (Figures 2I and 2J), indicating the presence of osteoclasts in the BA.
Regulation of Osteogenesis in BAs of αklotho−/− Mutants
In order to understand the molecular mechanisms underlying ectopic calcification in αklotho−/− mutants, we performed an RNA sequencing (RNA-seq) analysis of the kidney, heart (including BA), and gills at 3 mpf, when mutants appeared comparable phenotypically to wild-type siblings, and at 5 mpf, when αklotho−/− mutants became phenotypically distinct from wild-type siblings (n = 8 animals per genotype and condition; 4 males and 4 females). In the wild-type siblings, αklotho expression is highest in the kidneys (Figure 3A[3]). As expected, a significant downregulation of αklotho expression is observed in αklotho−/− mutants at both 3 and 5 mpf (Figure 3A[1]). In wild-type fish, fgf23 expression is detected in the kidneys and, surprisingly, in the gills (Figure 3B[2]). In αklotho−/− mutant gills, fgf23 is the most significantly upregulated gene at 3 mpf, indicating a dysregulation of the αKlotho/FGF23 axis (Figure 3B[2]).
.../...
F O R T H E R E S T O F T H E S T U D Y , P L E A S E V I S I T T H E S O U R C E
