Molecular Mechanisms of Hypopigmentation in Cystinosis

Christine CHIAVERINI, Robert BALLOTTI,   INSERM U597

MOLECULAR MECHANISMS OF HYPOPIGMENTATION IN CYSTINOSIS

Summary of the proposal

Patients with infantile cystinosis have hypopigmentation with, for Caucasian subjects, blond hair, blue eyes and a clear skin. However it seems that some patients, in particular African American patients, but also few Caucasian patients have not hypopigmentation. Unfortunately no correlation between cutaneous phenotype, severity of renal disease and genotype was carried out. The causes of hypopigmentation have not been so far elucidated. In humans, pigmentation results from the synthesis and distribution of melanin in skin, hair bulbs, and eyes. Melanin synthesis or melanogenesis is an enzymatic process, catalyzed by tyrosinase, tyrosinase-related protein 1 (Tyrp1) and dopa chrome tautomerase (DCT), which convert tyrosine to melanin pigments. This process takes place in melanocytes within lysosomes-related vesicles named melanosomes.

Two types of melanin are produced by melanocytes. Eumelanins are black/brown melanins with high photoprotective properties and pheomelanins are red/yellow sulfur containing pigments that provide no protection against the noxious effects of solar irradiation. Melanogenesis proceeds in three distinctive steps. The initial step is the production of cysteinyldopa, which requires dopa (the product of tyrosine hydroxylation by tyrosinase) and sulfur containing compounds such as cystine. The second step is the oxidation of cysteinyldopa to give pheomelanin, which continues as long as cysteinyldopa are present. The last step is the production of eumelanin, which begins only after most of the cysteinyldopa are depleted.

Further, melanin synthesis is also regulated by the pH of melanosomes: alkalinization of melanosomes by an inhibitor of vacuolar ATPase increases melanin synthesis and stimulation of melanin synthesis by _MSH is accompanied by an increase in melanosome pH. These data indicate that melanin synthesis take place at neutral or alkaline pH. Since CTNS is a cystine/H+ co-transporter, alteration of CTNS function in cystinotic cells would also lead to an acidification of melanosomes and an inhibition of melanin synthesis.

Taking into account that cystinosin transports cystine out of lysosome and that melanosomes are lysosome-related vesicles, it is tempting to propose that cystinosin is involved in the active melanosomal efflux of cystine and regulates thereby melanogenesis. In cystinosis, cystinosin dysfunction might lead to an intramelanosomal cystine accumulation that could favor pheomelanin synthesis compared to eumelanogenesis or to a decreased intramelanosomal pH, which is known to be an important parameter in melanin synthesis. Alternatively, accumulation of cystine in melanosome might toxic for melanocyte leading to a decreased cell growth or increased apoptosis sensitivity.

The aim of this project is to explain the molecular mechanisms of hypopigmentation in cystinosis. To reach this objective:

A.     We will determine the dermatologic phenotype of patients with infantile cystinosis to have an objective and quantitative evaluation of pigmentation disorders in these disease.  This part of the project will start in January 2008.

B.     We will study the role of cystinosin in melanogenesis as well as in melanocyte growth and apoptosis

Results

Expression and subcellular localization of cystinosin

First, real time PCR analysis of cystinosin gene (Ctns) expression in mouse adipocytes, liver, skeletal muscle, kidney and B16 mouse melanoma cells showed a strong expression of Ctns in mouse melanoma cells that is comparable at those found in kidney. Next we studied the localisation of cystinosin within B16 mouse melanoma. In absence of antibodies that recognize mouse Ctns, we transfected B16 melanoma cells with a CTNS-GFP construct and studied the co-localization of CTNS-GFP with melanosomes marker such as Tyrp 1 (B8G3) and pmel17 (HMB45) (Fig 1A). CTNS-GFP showed a vesicular pattern that co-localized with Tyrp1 and to a lesser extent with pmel17, mainly at the dendrite extremity. Confocal analysis, with anti-Tyrp1, at higher magnification (Fig 1B) also showed discrete co-localization at the level of intra cellular melanosome. When transfected with a GFP construct, cell were uniformly marked and no co-localization with melanosome markers was observed. To confirm the localisation of cystinosin in melanosomes, detergent free cell lysat from B16 melanoma cells transfected with CTNS-GFP, were immunoprecipitated with an anti-GFP antibody. In the immune complex, we detected by western blot CTNS-GFP and tyrosinase, indicating the presence of CTNS-GFP in the melanosomal membrane fraction (Fig 1C). Taken together, these data suggest that CTNS is a melanosomal protein.

CTNS silencing inhibits melanin synthesis

In order to investigate the role of Ctns in melanogenesis, we have designed a specific si-RNA directed against murine Ctns and then evaluated by RT-PCR its efficiency in pigmented B16 melanoma cells. In these cells, transfection of si-Ctns inhibits more than 75% of Ctns RNA expression compared to control si-RNA transfected cells (see Fig 3B). Treatment of B16 melanoma cells by forskolin (20_M) for 48h, increased melanin synthesis, as shown by visualization of the cells pellets (Fig 2A) and microscopy using direct light (Fig 2B) and by spectrophotometric measurement of melanin (Fig 2C). Interestingly, Ctns silencing inhibited cells pigmentation and reduced melanin synthesis by more than 50% compared to control cells.

CTNS silencing inhibits tyrosinase expression

Then, we have evaluated the effect of Ctns silencing on melanogenesis enzyme expression.Western blot analysis showed that Ctns silencing led to strong inhibition of tyrosinaseexpression, but did not affect significantly Typ1 and DCT expression (Fig 3A). Furthermorewe observed that forskolin treatment led to an increase of tyrosinase and Ctns mRNA level.In cells transfected with si-Ctns we observed a decrease of Ctns mRNA by more than 75%.However in these cells tyrosinase mRNA is increased compared to cells transfected withcontrol si-RNA. (Fig 3B).These data suggest that Ctns silencing in B16 melanoma cells decreases the tyrosinase levelthrough a posttranscriptional process.

Ctns silencing leads to acidification of intracellular vesicles

Since CTNS is a cystine/H+ co-transporter, it is tempting to propose that Ctns silencing would affect melanosomal pH.

Thus we evaluated the pH of intracellular vesicles by using Lyso-tracker® to label acidic compartments. In cells transfected by si-control we observed that forskolin treatment lead to a decrease of immunofluorescence labeling indicating an increase of the intracellular vesicles pH. Interestingly in cells transfected with si-Ctns we observed an increase in immunofluorescence labeling with Lyso-tracker® indicating that Ctns silencing led to an acidification of the intracellular vesicular compartments (Fig 4). Further experiments are required to identify this vesicular compartment as melanosomes.

CTNS expression is regulated by cAMP

We have shown in figure 3 that forskolin increase Ctns mRNA in B16 melanoma cells. Next we verified in NHM the regulation of CTNS by forskolin. Q-PCR analysis showed that forskolin induced a three-fold increase of the CTNS mRNA level in NHM (Fig 5A). Further western blot analysis confirm at protein level the stimulation of CTNS expression by forskolin (Fig 5B). The specificity of this antibody (clone 5G6) needs to be verified.

Next we wonder how cAMP regulates Ctns gene expression. Analysis of Ctns promoter showed the presence of four Eboxes susceptible to bind MITF, the transcription factor that mediate most of the cAMP transcriptional effects in melanocytes.  Interestingly, the comparison of the sequence of the human (NC_000017.9) and murine (NC_000077.4) CTNS promoter by BLAST (NCBI) showed, in the proximal region, an highly conserved region with 3 conserved potential biding sites for MITF suggesting that CTNS might be a new MITF target gene (Fig. 5C). Thus we cloned a 600 bp fragment of the human CTNS gene upstream the luciferase reporter gene and we performed a functional analysis of this promoter. We showed that the CTNS promoter was responsive to cAMP, since forskolin treatment led to a 3 fold increase in the luciferase activity. However co-transfection of the CTNS promoter reporter construct with an expression plasmid encoding MITF did not increase the promoter activity (Fig 5D). However we have shown that infection of NHM by an adenovirus encoding for MITF, induces a six fold increase of Ctns RNA level compared to NHM infected with empty adenovirus. Conversely, the silencing of Mitf with specific si-RNA in B16 melanoma cells induced a two fold decrease in Ctns RNA level (Figure 5E). These data confirm the role of MITF in the regulation of Ctns expression, but the lack of effect of MITF on Ctns promoter remain to be elucidated.

Figure Legends

Figure 1: Subcellular localization of Ctns. (A) Immunofluorescence of B16 melanoma cells, treated for 24hours with forskolin, transfected with CTNS-GFP (green) and labeled with an anti-tyrp1 or anti-pmel17 antibody (red). Co-localization of CTNS-GFP with Tyrp1 or Pmel17 (yellow). (B) Same experiment at higher magnification shows colocalization (yellow) of CTNS-GFP (green) and Tyrp1 (red). (C) Co-immunoprecipitation of melanosome using an anti-GFP antibody in B16 melanoma cells transfected with CTNS-GFP or GFP alone. Western blot of total (tot), and immunoprecipitated (IP) proteins using anti GFP, Rab4 and tyrosinase antibodies.

Figure 2: Effects of Ctns silencing on B16 melanoma cells pigmentation. (A) pellets of cells treated (FSK) or not (20_M) for 48 hours by forskolin and transfected by si-Ctns or control (si-Sc). (B) direct light microscope visualization of cells treated 48h by FSK and transfected by si-Ctns or control. (C) ratio of spectrophotometric measure of melanin level and level of total proteins in cells transfected by si-Ctns or control and treated or not by FSK.

Figure 3: Effects of Ctns silencing on enzymes of melanogenesis. (A) western blot analysis on B16 cells melanoma cells transfected by si-Ctns or control (Sc) and stimulated by FSK for 48 hours. (B) Q-PCR Ctns and tyrosinase mRNA levels in B16 cells melanoma cells transfected by si-Ctns or control (Sc) and stimulated or not by FSK for 48 hours.

Figure 4: Effects of Ctns silencing on melanosomal pH. B16 melanoma cells were transfected by si-Ctns or control (si-Sc), treated (FSK) or not by forskoline for 48h and then labeled with Lysotracker TR® for 4h (150nM).

Figure 5: (A) RNA levels of Ctns in MHN treated (FSK) or not (0) by forskolin (20_M) for 48hours, (B) cystinosin (cyst) expression in MHN treated (FSK) or not (0) by forskolin (20_M) for 48h, (C) Alignment of human and mouse of the ctns promoter performed using BLAST program (NCBI). Exon 1 of the human and murine ctns gene is shown by black arrow. E boxes are indicated in red (D) B16 cells were transfected in lipofectamine with 0.3 µg of pctns-wt, 0.05 µg of pCMVGAL and 0.2 µg of pMITF or not. After 24 h with forskolin, luciferase activity was normalized by the _-galactosidase activity and the results were expressed as fold stimulation of the basal luciferase activity from un-stimulated cells. Data are means ± SE of five experiments performed in triplicate. (E) RNA levels of Ctns in HNM infected for 72 h by adenovirus encoding for MITF or control (pcdna3) and in B16 melanoma cells and transfected by si-mitf or control (si-Sc).