- Top of page
- Materials and methods
- Supporting Information
Retinitis Pigmentosa involves a hereditary degeneration of photoreceptors by as yet unresolved mechanisms. The secretable protein α-Klotho has a function related to ageing processes, and α-Klotho-deficient mice have reduced lifespan and declining functions in several tissues. Here, we studied Klotho in connection with inherited photoreceptor degeneration. Increased nuclear immunostaining for α-Klotho protein was seen in degenerating photoreceptors in four different Retinitis Pigmentosa models (rd1, rd2 mice; P23H, S334ter rhodopsin mutant rats). Correspondingly, in rd1 retina α-Klotho mRNA expression was significantly up-regulated. Moreover, immunostaining for another Klotho family protein, β-Klotho, also co-localized with degenerating rd1 photoreceptors. The rd1 retina displayed reduced levels of fibroblast growth factor 15, a member of the fibroblast growth factor subfamily for which Klotho acts as a co-receptor. Exogenous α-Klotho protein added to retinal explant cultures did not affect cell death in rd1 retinae, but caused a severe layer disordering in wild-type retinae. Our study suggests Klotho as a novel player in the retina, with a clear connection to photoreceptor cell death as well as with an influence on retinal organization.
The inheritable disease Retinitis Pigmentosa (RP) leads to loss of vision via degeneration of rod and cone photoreceptors: typically rods die via mutation-induced mechanisms, after which cones degenerate secondarily (Pierce 2001). Although today over 60 genes have been linked to RP (RetNet: http://www.sph.uth.tmc.edu/RetNet), the mechanisms behind the degeneration are largely unclear, and there is currently no treatment available.
The Klotho protein – named after the Greek goddess Klotho, who spins the thread of life – is a player in longevity, and defective α-Klotho expression provokes rapid ageing and early death in mouse (Kuro-o et al. 1997). Apart from the prototypical α-Klotho, there are other Klotho family members, including β-Klotho and Klotho/lactase-phlorizin hydrolase-related protein [Lctl or γ-Klotho] (Ito et al. 2000). The Klotho gene encodes a 1014 amino acid, transmembrane protein (Matsumura et al. 1998), with homology to β-glucuronidases and is found mainly in kidney distal tubules, parathyroid gland, and brain choroid plexus, but also in other tissues including urinary bladder (Kuro-o et al. 1997) and inner ear (Kamemori et al. 2002). Both membrane and soluble forms (Imura et al. 2004) are known, with shedding of α-Klotho ectodomain (fragments) from the cell membrane via various secretases (Bloch et al. 2009). The connection between Klotho and the demise of various cell types (Kuro-o et al. 1997) warrants a consideration of an involvement also in the retina and RP. Current information on retinal Klotho is sparse. High mRNA expression levels of Lctl have been reported in adult mouse eyes (Fon Tacer et al. 2010), with α-Klotho and β-Klotho expression either absent or limited, respectively. However, a previous, microarray based study suggested elevated levels of β-Klotho mRNA in retinae of the rd1 mouse model for RP (Azadi et al. 2006).
The altered β-Klotho mRNA in rd1 retina and the α-Klotho connection to ageing and death triggered us to investigate Klotho isoforms in different RP animal models. These included the rd1 mouse, which carries a mutation in the gene coding for the β-subunit of phosphodiesterase-6 (Farber and Lolley 1974), giving a rapid degeneration, and the rd2 mouse, with a mutation in the photoreceptor disc protein peripherin-2 gene (Goldberg 2006) and in which the retinal degeneration is slower. Both models represent human RP forms (Kajiwara et al. 1991; Bayés et al. 1995). Rhodopsin mutations are frequent in RP patients, and hence we also used two rhodopsin mutant rats that display fast (S334ter; Liu et al. 1999) and slow (P23H; Sung et al. 1991) rod degeneration respectively. The rd2 mutation affects both rod and cone photoreceptors, while the other mutations are rod specific. The unique characteristics of the four models, from two different species, make them well suited to address the possible involvement of Klotho in inherited photoreceptor cell death.
Here, we demonstrate a strong up-regulation of several Klotho isoforms in degenerating retinae, as well as an effect on retinal morphology by addition of exogenous Klotho to retinal cultures. The findings suggest a link between Klotho and the process of retinal degeneration.
- Top of page
- Materials and methods
- Supporting Information
The Klotho protein is recognized as an important factor for cellular ageing and survival (Kuro-o et al. 1997; Kurosu et al. 2005), but a clear connection between high Klotho expression and neuronal cell death, as shown here, has to our knowledge not been demonstrated before. Our results also suggest that over-expression of Klotho proteins, particularly α-Klotho, serves as a marker for hereditary photoreceptor cell death, and that the protein as such may act towards disorganization of photoreceptors and other retinal cells.
Klotho expression in the retina
Immunostaining using three independent antibodies identified a sub-population of photoreceptors in the rd1 retina with a highly increased nuclear α-Klotho protein expression. Nuclear Klotho has previously been shown in both brain (e.g. Purkinje cells; German et al. 2012) and inner ear sensory cells (Takumida et al. 2009). α-Klotho is known to assist the atypical FGF family in binding to the FGFRs. FGFRs are membrane proteins, although alternative spliced transcripts may lose the hydrophobic transmembrane domain and can then also be found as soluble forms, of which at least the soluble FGFR1 has also been shown to be able to enter retinal cell nuclei (Guilonneau et al. 1998). Since FGFR4 could be detected in photoreceptor and other retinal cell nuclei of wt and rd1 retinae alike (Figure S3), one of the possible Klotho interaction partners during retinal degeneration may reside in this compartment.
The degeneration also affected α-Klotho mRNA, which was higher in rd1 retinal samples than in wt material. This increase may have occurred over a low steady-state transcription of the α-Klotho gene, since a screen for Klotho mRNA showed its expression to be low or not detectable in several (adult) mouse tissues, including the eye (Fon Tacer et al. 2010). In contrast to immunostainings and mRNA measurements, our comparisons at the western blot level did not show an rd1 versus wt difference. However, the western data suggested wt retinal α-Klotho protein levels to be sustained at a detectable level at the PN11 age, and an increase in a small, select number of cells, as in the rd1 situation, would then be difficult to distinguish in a global sample.
Klotho and cell death
Each of the three α-Klotho antibodies revealed an extensive co-localisation with dying, TUNEL positive photoreceptors in the rd1 retina. This confirms that the antibodies identified the same population of cells, and suggests that the α-Klotho increase connects with the rd1 degeneration process. Moreover, this connection was not restricted to a certain model or species, since comparable co-localisation results were readily detected in the rd2 mouse and in the P23H and S334ter rats. Taken together, these data argue for α-Klotho up-regulation as an integrated and general event of photoreceptor degeneration, regardless of how the pathological processes start. Interestingly, this notion extends also to the β-Klotho protein, which showed increased expression, and broad TUNEL co-localisation, in a subset of rd1 photoreceptors. We could not identify a significant increase in β-Klotho mRNA in rd1 samples compared with control tissue, even though this was suggested in a previous microarray study (Azadi et al. 2006). The discrepancy might be attributed to the distinctive technologies used (microarray vs. qRT-PCR) or to other methodological differences. At any rate, Klotho expression is unequivocally associated with photoreceptor cell death and may thus serve as a novel diagnostic marker for RP and related neurodegenerative diseases.
Is Klotho protective or destructive?
α-Klotho is linked to anti-ageing and has potential cellular protection capacities. In this role it may act extracellularly either on its own or as a co-receptor for members of the FGF19 subfamily, such as FGF23 (Wang and Sun 2009), although intracellular effects have also been described (Liu et al. 2011). Lack of α-Klotho appears to promote senescence (Kuro-o et al. 1997) and weaken the oxidative stress defence (Nagai et al. 2002). Conversely, oxidative stress, old age, inflammation, and cellular senescence lead to, or coincide with, reduced α-Klotho levels (Mitani et al. 2002; Mitobe et al. 2005; Takumida et al. 2009; Thurston et al. 2010; Liu et al. 2011). Furthermore, α-Klotho gene over-expression or addition of α-Klotho protein may counteract or reduce oxidative stress, inflammation, cellular senescence, cellular dysfunction, or even cell death in various systems (Saito et al. 2000; Yamamoto et al. 2005; Haruna et al. 2007; Sugiura et al. 2010; Liu et al. 2011). Since oxidative stress, in particular DNA oxidation, is involved in rd1 degeneration (Paquet-Durand et al. 2007) the up-regulation of α-Klotho in photoreceptors could thus be part of a protective response to such insults. However, α-Klotho has been reported to confer resistance to oxidative stress through the expression of manganese superoxide dismutase, led by FoxO forkhead transcription factors, a downstream effector of the insulin intracellular signalling (Yamamoto et al. 2005). The activation of FoxO forkhead transcription factors is negatively regulated by Akt (also known as PKB)-dependent phosphorylation. Akt has been shown to be overactivated in rd1 retinae (Johnson et al. 2005), and it is therefore possible that this has counteracted the ability of α-Klotho to induce expression of manganese superoxide dismutase, and hence to confer protection in the degenerating retina.
As an alternative scenario, α-Klotho could instead be part of the neurodegeneration mechanism as such. This notion appears novel, since many investigations rather suggest α-Klotho to counteract neurodegeneration, in that α-Klotho reduction leads to signs of neurodegeneration and/or loss of cells in several areas of the CNS (Kuro-o et al. 1997; Nagai et al. 2002; Anamizu et al. 2005; Shiozaki et al. 2008; Kosakai et al. 2011). However, these reports identify neurodegeneration in a situation of experimentally altered α-Klotho, whereas we detected altered α-Klotho in a situation of disease-induced neurodegeneration, which represents a completely different setting. Judging from the distinct co-localisation of α-Klotho (and β-Klotho) with TUNEL, that is, a clear connection with cell death, we are tempted to speculate that Klotho proteins may be involved in the later stages of the degeneration process. The fact that photoreceptor degeneration was not promoted by exogenous α-Klotho would then point to a preferential intracellular degeneration involvement of α-Klotho, although the lack of effect might also be a result of Klotho co-effectors (e.g. FGF15) being low in rd1 photoreceptors.
In any event our study implies that degenerating photoreceptors experience a distinct imbalance in the Klotho-FGF system, which is interesting not the least because of the benign effects of FGF19 signalling on photoreceptor survival and maintenance (Siffroi-Fernandez et al. 2008). Perhaps future experimental manipulation of the interaction(s) between α- and/or β-Klotho and FGF15/19 protein family can shed light on the exact importance of the Klotho/FGF axis for the retinal degeneration process.
Klotho and post-natal retinal development
Interestingly, exogenously added α-Klotho protein resulted in a structural disorganization of both inner and outer retinal elements at the edges of the preparations. For one thing this underlines that the lack of effect by Klotho on the degeneration was not likely to be because of methodological problems in those experiments. Furthermore, with respect to the mechanisms behind the disorganization, the treatment of the explants started at a time-point when the retina is still undergoing development in a centre-to-periphery fashion (Young 1985), making it possible that Klotho interfered with normal tissue development. Since retinal precursors are present in the retinal margin (Willbold and Layer 1992), one should also not disregard the possibility that these may have been re-activated and contributed for instance to the loss of markers such as PKCα/βII. However, since the treatment did not increase the size of the treated explants, a re-activation and significant growth of such stem cells seem not to have occurred.
In conclusion, the present report introduces α-Klotho as a novel player in the context of retinal health and disease, and particularly in inherited retinal degeneration. While Klotho proteins are increased in degenerating rd1 photoreceptors, FGF15 is reduced, suggesting an imbalanced α-Klotho-FGF axis as part of the disease characteristics. Furthermore, the disorganization of the developing mouse retina by α-Klotho is compatible with a role for this protein in retinal cell differentiation and/or layer formation.