Abstract
Mutations in the CTNS gene encoding the lysosomal membrane cystine transporter cystinosin are the cause of cystinosis, an autosomal recessive lysosomal storage disease. More than 140 CTNS mutations have been reported worldwide. Recent studies have discovered that cystinosin exerts other key cellular functions beyond cystine transport such as regulation of oxidative state, lysosomal dynamics and autophagy. Here, we review the different mutations described in the CTNS gene and the geographical distribution of incidence. In addition, the characteristics of the various mutations in relation to the functions of cystinosin needs to be further elucidated. In this review, we highlight the functional consequences of the different mutations in correlation with the clinical phenotypes. Moreover, we propose how this understanding would be fundamental for the development of new technologies through targeted gene therapy, holding promises for a possible cure of the kidney and extra-renal phenotypes of cystinosis.
Clinical Presentation of Cystinosis
Cystinosis is an autosomal recessive disease caused by mutations in the CTNS gene encoding the lysosomal membrane transporter cystinosin [1]. Cystinosin is a ubiquitously expressed protein, functioning as a cystine-proton co-transporter and extruding cystine out of the lysosomes; this activity is dependent on the lysosomal acidification conveyed by the H+-ATPase [2]. Cystinosin deficiency results in lysosomal cystine accumulation and cystine crystals formation in virtually all tissues and organs [3]. While cystinosis belongs to a larger group of ~50 lysosomal storage disorders, mostly caused by lysosomal enzyme deficiencies, its clinical phenotype is distinct, with the kidney being the first affected organ.
Depending on kidney disease severity, the cystinosis phenotype is divided in 3 clinical forms [4]. The most severe variant, infantile nephropathic cystinosis (MIM219800) affects ~95% of patients and is characterized by the development of renal Fanconi syndrome during the first months of life followed by glomerular dysfunction, which if untreated, results in end-stage kidney disease (ESKD) around the age of 10 years. Late-onset juvenile nephropathic type (MIM219900) usually presents during childhood or at adolescence with mild or even absent -proximal tubular dysfunction, proteinuria, which can be in the nephrotic range, and a slower rate of progression towards ESKD [5]. Non-nephropathic cystinosis (MIM219750) is a benign variant presenting with photophobia due to cystine accumulation in the cornea but causing no systemic organ damage [6].
All 3 clinical forms of cystinosis are due to bi-allelic mutations in the CTNS gene with a mutation detection rate > 95% [7].
The first two clinical phenotypes of cystinosis are also characterized by extra-renal organ dysfunction gradually developing from childhood to adulthood. The eye is the second affected organ with cystine crystals found in the cornea and retina early in life, and progressive retinopathy eventually leading to blindness at adolescence or adult age [8]. Endocrine organs (thyroid, pancreas, and gonads), GI tract, muscles, bones, central and peripheral nervous systems are also impacted by the disease, with morbidity and mortality frequently being associated with the swallowing dysfunction and aspiration due to muscle weakness [9].
Long-term prognosis of patients with cystinosis has dramatically improved during the last 30 years by the availability of the cystine-depleting drug cysteamine and advances of kidney transplantation. Cysteamine, a small amino thiol, was first introduced for the treatment of cystinosis in 1976 [10] and later was shown to prolong kidney function survival, improve growth, and delay or even prevent extra-renal disease complications [11, 12]. The results of kidney transplantation in cystinosis are excellent with no disease recurrence and long-term graft survival being better in cystinosis compared to other renal disease patients [13, 14]. Hence, the oldest cystinosis patients passed now the age of 55, and a new generation of well-treated patients is expected to have even better life expectancy. Nevertheless, the severity of renal Fanconi syndrome with daily urine volume sometimes exceeding 10 L and resistance to cysteamine therapy, multiorgan involvement in cysteamine-treated patients, and the burden of cysteamine administration causing unpleasant body and breath odour, GI and bone complaints in some patients, point out the necessity of developing novel, less toxic and more efficient treatment strategies tackling all disease aspects.
Genetic Basis of Cystinosis
Overview of Mutations in the CTNS Gene and Geographical Distribution
Over 140 pathogenic CTNS mutations have been reported in cystinosis patients worldwide. These include 57 missense and nonsense mutations, 23 intronic mutations, 45 deletions, 13 small insertions, 4 indels and 3 promoter region mutations (Table 1) [15]. Overall, there is a considerable genotype-phenotype correlation of the CTNS mutations, with the absolute majority of pathogenic mutations causing the infantile phenotype, while only 15 and 4 mutations have been attributed to juvenile and ocular cystinosis respectively (Table 1).
Most reported CTNS mutations were detected in countries of Europe and North America, with only a minority of developing nations conducting genetic studies, mostly due to lacking funds to perform molecular diagnosis in potential patients [16]. Notably over the last decade, an increasing number of reports have been published on CTNS gene mutations in some of the developing nations’ populations [17-20]. Still, large geographic regions, such as sub-Saharan Africa, South-East Asia and the Far East are underrepresented in the genetic spectrum of their cystinosis patients. Although accurate statistics are lacking in the developing world [21], recent studies from the Middle East [17-20], Mexico [22] and South -Africa [23] may indicate that the incidence of cystinosis in many of these countries is expected to be higher than that of Europe and North America. Moreover, many cystinosis patients in poor countries remain undiagnosed and die at a young age due to complications of the disease [24].
The genetic landscape of cystinosis varies widely based on the ethnic and genetic makeup of each reporting population. The most common pathogenic mutations in Northern Europe and North America is a large deletion (57-kb del) [12] affecting the promoter region and the first 10 exons of the CTNS gene together with 2 upstream genes (CARKL and TRPV1) [25]. This deletion represents over 50% of mutant alleles in cystinosis patients of Northern European ancestry; however, it is completely absent in all reported patients from the Middle East, Asia and Africa [16], suggesting a founder mutation. Other important founder mutations can be encountered in certain populations or certain geographic loci, such as the c.1015G>A in the Amish population in Western Ontario, Canada [26], the c.971–12G>A in the black population of South Africa [23] and the c.681G>A in the Middle East [17-19]. A more detailed description of the genetic map of cystinosis is needed to elucidate the similarities and the differences between patients all over the world and to prepare for individualized disease treatment targeting gene repair.
Functional Characteristics of CTNS Mutations: Genotype-Phenotype Correlations
CTNS is a lysosomal 7-transmembrane protein (Fig. 1), which functions as H+-driven cystine transporter [2]. It is predicted to have 7 N-glycosylation sites located at the N-terminal tail, residing in the lysosomal lumen. Two lysosomal targeting motifs, YFPQA (5th inter-transmembrane loop) and GYDQL (C-terminal tail), ensure localization to the lysosome [27]. An alternative splice variant termed CTNS-LKG exists, resulting in a -longer protein (400 vs. 367 amino acids) with a plasma membrane targeting signal SSLKG replacing the -C-terminal GYDQL [28, 29].
Kalatzis et al. [30] have studied the functional consequences of 31 CTNS mutations using constructs that encoded CTNS lacking the GYDQL lysosomal targeting signal, termed CTNS-∆GYDQL. This truncation leads to a translocation of the cystinosin protein to the plasma -membrane, thus enabling the monitoring of extracellular [35S]-cystine influx. In particular, they measured the transport activity compared with WT CTNS in a COS-7 cell transfection model. Surprisingly, it was found that specific CTNS mutations (e.g., CTNSS298N and -CTNSW182R) causing infantile nephropathic cystinosis, retained cystine transport activity. Conversely, mutations (e.g., CTNSN323K and CTNSK280R) resulting in juvenile nephropathic cystinosis, lost cystine transport activity [30]. These findings opposed the well-accepted paradigm that cystine accumulation is the primary pathogenic cause of cystinosis and suggested that CTNS exerts other key functions beyond cystine transport [2, 30]. Supporting these findings, cysteamine treatment does not alleviate the complete phenotype of cystinosis.
Indeed, other CTNS functions are being investigated, involving cell-survival pathways and the mammalian target of rapamycin complex 1 (mTORC1) activity. mTORC1 is the master regulator of cell growth, which is recruited to the lysosomal membrane in presence of nutrients, activating protein translation and inhibiting autophagy. Conversely, upon starvation, mTORC1 released from the lysosomal surface is inactivated, leading to the upregulation of autophagy and inhibition of the main anabolic pathways [31, 32]. Recently, CTNS was found to be part of the V-ATPase-Ragulator-Rag complex, which controls mTORC1 activation [33]. Both mutations CTNSN323K and CTNSK280R did not show any alteration in the interaction of mTORC1 with cystinosin. However, this was not true for CTNSN288K, which specifically resulted in infantile nephropathic cystinosis with loss of cystine transport activity [33] (Table 2). Moreover, mouse -Ctns–/– proximal tubular cells showed disturbed mTORC1 signalling and delayed docking of mTORC1 to the lysosomal surface following the re-introduction of the complete medium after starvation [5]. In addition, human cystinotic proximal tubular cells carrying the 57 kb deletion and a compound heterozygous (c.Y173X+p.G339R) showed abnormal lysosomal mTORC1 localization upon starvation, and cysteamine supplementation did not rescue the phenotype [34].
When mTORC1 is recruited to the lysosomal membrane, it also phosphorylates and inhibits transcription factor EB (TFEB). TFEB normally activates lysosomal biogenesis and autophagy-related genes, in response to starvation [32]. Human CTNS–/– proximal tubular epithelial cells (PTEC) showed lower TFEB expression, which was mainly located in the nucleus. Overexpressing TFEB in CTNS–/– PTEC rescued some of the CTNS WT characteristic such as (i) delayed lysosomal cargo processing, (ii) improved morphological irregularities of the lysosomal compartments (after > 24 h) and (iii) promoted clearance of lysosomal cystine by exocytosis [35]. It would be interesting to investigate whether overexpressing (mutant and WT) CTNS in CTNS–/– PTEC could rescue this defect, as this could further confirm the altered mTORC1 signalling shown in Ctns–/– proximal tubular cells.
Another role of CTNS, in addition to its potential involvement in mTORC1 signalling, is the altered chaperone mediated autophagy, as shown by two studies from Napolitano et al. [37] and Zhang et al. [36]. They observed impaired chaperone mediated autophagy (CMA) in cystinotic mouse skin fibroblasts presenting dislocation of LAMP2A (CMA receptor) and accumulation of GAPDH (CMA substrate). Interestingly, the CTNSK280R mutation rescued the localization of LAMP2A in contrast to cysteamine treatment. This was also true for CTNS-LKG, which is an alternative splice variant of CTNS, primarily targeted to the plasma membrane and other lysosomal/endosomal vesicles. Based on these results, it was suggested that CTNS is a necessary cofactor for LAMP2A trafficking [36, 37].
The lack of cystine transport activity of the CTNSK280R and CTNSN288K mutants may be explained by the fact that both mutations are located at the 5th inter-transmembrane loop containing the PQ motif, which is required for H+ and cystine co-transport [38]. The difference in clinical presentation, juvenile versus infantile cystinosis for CTNSK280R and CTNSN288K, respectively, can thus be only explained by effects beyond cystine transport.
As mentioned earlier, CTNS is predicted to have 7 -N-glycosylation sites at the N-terminal tail (Fig. 1). -CTNSN323K and CTNSN288K do not affect glycosylation of CTNS. However, a deletion of 7 amino acids at the -N-terminus, referred to as CTNSDel.67–73 (juvenile nephropathic cystinosis with severely inhibited cystine transport), results in the loss of 1 consensus N-glycosylation site. Furthermore, this mutant protein degraded threefold faster than WT CTNS and was partially retained at the endoplasmic reticulum. Nevertheless, CTNSDel.67–73 still interacts with the V-ATPase complex, which could explain the less severe phenotype caused by this mutation [39].
Animal Models of Cystinosis
Mouse
Sixteen years ago, Antignac’s group generated the first model of cystinosis in FVB/N mice by using promoter trap approach to eliminate the CTNS gene [40]. Although the model showed ocular abnormalities, bone defects, behavioural anomalies and partial response to cysteamine treatment, it presented a moderate cystine accumulation compared to the affected humans without any sign of morphological or functional alterations of the proximal tubule. As the genetic background can influence the renal phenotype, the same group developed a second model in C57BL/6 mouse, and observed mild proximal tubulopathy (without the urinary loss of amino acids, bicarbonates or sodium) and focal tubular lesions. However, no podocyte damage was present [41]. The latter mouse model has been widely used by different research groups for studying the pathogenesis and treatment strategies in cystinosis [42, 43].
Yeast
Saccharomyces cerevisiae is a well-established model organism, mainly used for genetic manipulations and biochemical analyses. ERS1, the yeast orthologous counterpart of the human CTNS gene, encodes for the Ers1 protein (28% identical/46% similar to human CTNS), which localizes to the endosomes and the vacuole. ers1Δ mutant shows hygromycin B (hygB) sensitivity, which could be reversed only by the complementation of a functional CTNS human gene [44] or by overexpressing MEH1, which encodes a protein involved in vacuolar acidification and general amino acid permease (Gap1p) localization [44, 45]. Interestingly, Simpkins et al. [46] showed that Ers1 acts as a cystine transporter; however, the ers1Δ mutant does not show any detectable defect in growth and has no cystine accumulation [44], thereby supporting the evidence that other genes could compensate for its lost function [46]. Another relevant study performed in yeast confirmed that the GYQDL signature at the C-terminal end is essential for the trafficking of cystinosin; indeed, deleting the GYQDL C-terminal region of CTNS expressed in yeast delocalized cystinosin from the vacuole to the plasma membrane, restoring the ability to grow on cystine [47]. As the developed CTNS-ΔGYDQL transformants showed low cystine uptake, vacuolar protein-sorting deletions have been developed to increase the protein concentration at the plasma membrane, from which 2 deletions stand out: vsp1Δ, vsp17Δ enhancing the uptake levels. Several gain-of-function mutants were isolated, including 1 patient mutation, G197R. These data demonstrate that cystine-uptake assay in yeast cells is useful to decipher the functionality of mutant cystinosin proteins.
Zebrafish
Zebrafish is the most recent model organism to study cystinosis. Carrying the homozygous nonsense mutation in exon 8 of the ctns, zebrafish larvae demonstrated early cystine accumulation, enhanced deformity, apoptosis and increased mortality, which are partially responsive to cysteamine treatment [48]. Furthermore, the model is characterized by impaired glomerular permselectivity and defective tubular reabsorption. Unlike the murine model, the adult ctns–/– zebrafish kidney accumulates the highest concentration of cystine. In line with functional abnormalities, increased lysosomal size and numbers characterize larvae’s PTEC, while podocytes present partial foot process effacement and narrowed slit diaphragmatic space. This last phenotype is lacking in the mouse model, but is present in humans. The abundance and localization of the megalin receptor, is altered in ctns–/– larvae compared to WT, a marker for defective tubular reabsorption, which is not restored by cysteamine treatment. Hence, the zebrafish larval model closely copies the human kidney phenotype for cystinosis and is superior to all model organisms currently available. Therefore, it will be useful to further unravel the pathophysiological aspects of cystinosis and for the in vivo screening of novel therapeutic agents [48].
A comparison of cystinosis in humans, mice, yeast and zebrafish is presented in Table 3 (adapted from [48]).
Future Perspective: Potential for Gene Repair
To date, cysteamine is the only drug available to treat cystinosis. Although cysteamine can alleviate symptoms and delay disease progression, it does not prevent the Fanconi syndrome and renal transplantation is still necessary. In addition, compliance to cysteamine is low and side effects are frequently observed [49]. Thus, there is a need for novel therapies, which not only treat, but could potentially cure cystinosis.
Since cystinosis is a multi-systemic disease, developing gene therapy is challenging as all body cells require correction. In line with the strategy applied for other lysosomal storage disorders, Syres and colleagues transplanted bone marrow-derived cells and hematopoietic stem cells (HSC) from Ctns+/+ to Ctns–/– mice [50]. The rationale underlying HSC therapy is that healthy donor cells migrate into the recipient’s organs and release the missing protein locally, thereby correcting the metabolic defect [51]. However, for cystinosis it was still a question whether a transmembrane protein like CTNS could be taken up by the diseased cells [50]. In mice, bone marrow-derived cells and HSC-derived cells engrafted efficiently in the interstitial compartments of the kidney and other organs, decreased cystine accumulation and averted development of kidney dysfunction [50]. Moreover, if sufficient donor-derived blood cell engraftment took place (> 50% of blood cells), the cystinotic phenotype was corrected up till 7–15 months post transplantation [52]. All together these findings provide a proof-of-principle that cystinotic mice can be treated by HSC transplantation. Elmonem et al. [53] reported the first human case which underwent allogenic HSC transplantation from a full HLA-matched unrelated donor. Although mRNA and cystinosin protein transfer from HSC to epithelial cells did occur and was able to reduce cystine crystal load, the patient still developed ESKD and died from severe graft-versus-host disease, as a complication of transplantation. Because of the inherent risks associated with allogenic transplantation, autologous Ctns–/– HSC transplantation of HSC corrected with a lentiviral vector encoding a functional copy of CTNS were evaluated in mice. This approach reduced cystine content in all tissues and improved kidney function in a mouse model of cystinosis [54]. In addition, cross correction of diseased cells was shown to be mediated by tunnelling nanotubes produced by the gene-corrected macrophages [54]. Both in vitro and in vivo, these tunnelling nanotubes provided diseased cells with functional CTNS bearing lysosomes. The transfer was found to be bi-directional as cystine loaded lysosomes from -cystinotic cells were transferred to the macrophage as well [55].
With several gene therapeutic approaches reaching the market in the past years, gene therapy also holds promise to cure cystinosis. In line with this, the company Avrobio has planned to start Phase 1/2 human trials in 2019 (http://www.avrobio.com/pipeline/). However, the importance of CTNS expression levels should be noted as CTNS is regulated both at the transcriptional and the posttranscriptional level [56, 57]. The promoter driving CTNS expression in lentiviral constructs should be evaluated carefully as overexpression of CTNS has been reported to enlarge lysosomal structures of which the long-term consequence is currently unknown [27, 54]. Despite the promising results of gene therapy for other lysosomal storage disorders [58] long-term safety and efficacy of such therapy still require to be studied in more detail before it becomes a realistic option as a first-line treatment for patients affected by lysosomal storage disorders.
Due to the risks associated with HSC transplantation, especially in young children, researchers set out to find novel gene therapeutic approaches. Arcolino et al. [59] showed that urine of preterm neonates contained kidney stem/progenitor cells that had regenerative paracrine effect and could differentiate into podocytes and proximal tubule epithelial cells. Therefore, we hypothesize that, ex vivo gene-corrected autologous cystinosis kidney progenitors could also be used as source for cell therapy that could potentially cure the kidney phenotype. As cystinosis is a multisystemic disease, this therapy should be combined with drug therapy protecting extra-renal -organs.
Conclusion
Worldwide pathogenic CTNS mutations have been reported in cystinosis patients, but still, well-documented incidence and geographical distribution has been poorly described. Analysis of mutant cystinosin has provided the research field with a new scientific consensus in which CTNS has other critical functions beyond cystine transport. However, the extent to which defects in these other functions contribute to the overall clinical phenotype still needs further investigation. Preclinical studies using different animal models might elucidate functional consequences of specific mutations and might pave the way for a more personalized treatment depending on the correlations between genotype and phenotype.
The results of treatment with cysteamine vary widely and studies showed that cysteamine improves the outcome of cystinosis but does not provide cure. Hence, there is a need to find alternative therapeutic strategies. Over the last decades, cystinosis evolved from a paediatric lethal to a treatable disorder, with the advent of gene- and cell therapy, cystinosis might even become curable.
Ethics Statement
The authors have no ethical conflicts to disclose.
Disclosure Statement
The authors have no conflicts of interest to declare.
Funding Sources
E.L. is supported by FWO (1801110N) and E-RARE (JTC2014). R.G is supported by FWO (G0B3516N). E.L., R.G., and S.P.B. are supported by C1 grant from KU Leuven. F.O.A. is supported by FWO (12Q9917N).
Authors Contribution
D.D., S.P.B., M.A.E., F.O.A., N.S., B.V.H., R.G., and E.L. searched the literature and wrote the review text. D.D., S.P.B., and M.A.E. prepared the figure and tables.
References
D.D. and S.P.B. equal contribution.