Development, and In Vivo Testing, of Novel Therapies for Cystinosis

This 2 yr research project is a joint proposal between:

Pr. Corinne Antignac (Inserm U574, Hôpital Necker-Enfants Malades, Paris, France) - Project 1 and

Dr. Vasiliki Kalatzis (CNRS UMR 5535, Institut de Génétique Moléculaire, Montpellier, France) - Project 2

The money allotted to Project 2 is currently being used to finance Claire Hippert, a Ph.D. student.

Progress report Project 1 – Cysteamine trial

i) Characterization of the renal phenotype in Ctns-/- mice backcrossed onto pure genetic backgrounds

Prior to treat Ctns-/- mice with cysteamine to verify its efficiency, we have performed extensive analysis of the murine model of cystinosis. The identification of the CTNS gene enabled us to generate a mouse model of cystinosis by disrupting the murine homologous of CTNS (Cherqui et al., 2002). This model has been shown to present some of the findings in patients with cystinosis, such as cystine accumulation, cystine crystal deposition in tissues and ocular anomalies. However, these mice do not develop any renal abnormalities, aside from rare vacuolised renal tubules in a small number of mice.

Knowing that the genetic background can strongly influence the phenotypic severity, we transferred the Ctns mutation from our mixed C57BL6/129sv background mice onto pure C57BL/6 and FVB/N backgrounds. We then verified that the mice were accumulating cystine in their organs. In addition, we performed various histological studies, which interestingly showed that whereas FVB/N Ctns -/- mice do not have renal lesions, C57BL/6 Ctns -/- mice present several histological findings. Chronologically, the first lesions we observed were vacuolised tubules, which appeared around 6 months of age. At 9 months, proximal tubular atrophy or cellular infiltration was found. Notably, these lesions were focal; hence, both normal and affected areas were seen in the same kidney. We initially measured serum and urine parameters and failed to see any global differences in renal function between Ctns +/+ and Ctns -/- mice. Given that the lesions do not involve the entire kidney in Ctns -/- mice, these measures of global renal function may fail to detect a mild disorder in renal function. Thus, we plan to study tubular function more precisely in these mice.

This murine model of cystinosis, which develops renal lesions, will be a relevant model on which the cysteamine trial can be performed.

(ii) Immortalised cell lines

Last year, we mated Ctns mice with the Immortomouse, bearing a thermosensible mutant of the SV40 T antigen in their genome (Jat et al., 1991), and isolated proximal tubular cells from 6 Ctns-/- and 5 Ctns+/+ mice. The cells we isolated were conditionally immortalized and could either proliferate or differentiate when grown at 33° or 39°, respectively. We then demonstrated that the Ctns-/- cell lines accumulate cystine and that all the cell lines exhibit γ﷓glutamyl transpeptidase activity consistent with a proximal tubular phenotype. This year, we further characterised these cell lines. First, we detected alkaline phosphatase activity in these cells. Furthermore, when the cells were grown on filters, transepithelial resistance (around 50 Ω/cm²) was measurable, demonstrating that the cells are polarised. Additionally, we observed that α-tubulin, β-catenin and zona occludens 1 were properly localised in these cells; β-catenin and zona occludens 1 were detected at the plasma membrane, whereas α-tubulin was found in the submembranous region. The results of all these studies are consistent with the cells possessing a proximal tubular phenotype. At present, we are performing biochemical analysis to evaluate the redox status of these cells.

The human cell lines we previously generated were permanently immortalised and these cells mildly express tubular markers. Thus, we aimed to generate human tubular cells that are conditionally immortalised. We obtained cells from the urine of 3 cystinosis patients and our collaborators at the Centre de Thérapie Génique (Nantes, France) transduced these with a thermosensitive SV40 T antigen. We have recently received these new cell lines and we will now perform the same sets of experiments to verify the proximal tubular phenotype of these cells. In addition, we also generated 2 control tubular lines from normal individuals.

(iii) Antibodies against cystinosin

Since polyclonal antibodies generated against cystinosin-derived peptides were highly batch-dependant and since no reliable antibodies directed against the N-terminal part of cystinosin could be generated by this approach, we decided to use the Baculovirus expression system to produce recombinant cystinosin proteins in insect cells. This system has the unique advantage of producing a high level of recombinant proteins that can undergo several post-translational modifications, which is crucial for cystinosin as the N-terminal part of the protein is highly glycosylated.

We have already generated a CTNS transfer vector by inserting the 5’ part of the CTNS gene (encoding amino acids 24 to 123) preceded by sequences encoding a signal peptide and a 6XHIS tag in the multiple cloning site of pFastBac1. This vector is currently used to produce the recombinant protein with the Bac-to-Bac® Baculovirus Expression System (Invitrogen), which will then be injected to rabbits to produce polyclonal antibodies [in collaboration with Dr Y. Boublik, at the CRBM (Montpellier, France)].

Progress report Project 2 – Gene transfer studies

We have made significant progress in two parts of our original research proposal.

i) Validation of in vitro gene transfer studies

To validate our in vitro gene transfer studies showing that the efficiency of cystine reduction is age-dependent in primary murine hepatocyte cultures, we began tail vein injections of human adenovirus vectors expressing CTNS (hAd-CTNS) or CTNS fused to the green fluorescent protein GFP (hAd-CTNS-GFP) to target the liver in different aged Ctns-/- mice. Initially, we performed liver biopsies on Ctns-/- mice on day 1 (n = 9), injected the mice with hAd-CTNS (n = 3), hAd-CTNS-GFP (n = 3) or hAd-GFP (n=3) on day 2, and sacrificed the mice on day 9. However, the liver biopsy itself caused a reduction in cystine levels likely, due to a regeneration of the liver. We thereby altered our experimental plan to compare cystine levels post-transduction with those of non-injected mice: we injected mice (n=9) with hAd-CTNS (n = 4), hAd-CTNS-GFP (n = 4) or hAd-GFP (n = 4) on day 1, and sacrificed the animals on day 8; we used 4 non-injected age-matched Ctns-/- mice and 3 Ctns+/+ as controls.

Following sacrifice, we evaluated the liver transduction efficiency by FACS analysis (GFP expression) and by immunofluorescence studies (GFP expression or using an anti-cystinosin antibody). We evaluated cystine clearance using the radio-competition cystine binding protein assay. Following injection of 9 month-old mice, we did not reduce cystine levels with any vector 1 week post-injection despite a 70% transduction efficiency. In contrast, following injection of 2 month-old mice, we reduced cystine levels with the hAd-CTNS vector despite a lower transduction efficiency. We are currently repeating the injections of 2 month-old mice to confirm these data and we will also inject an intermediate age (6 months).

ii) Eye gene transfer

Our in vitro data suggests that cystine reduction is age-dependent, thus a spatial and temporal guide of the ocular anomalies in the Ctns-/- mice is a necessary prelude to in vivo gene transfer studies. We recently finished the biochemical, histological and clinical characterisation of the ocular anomalies in Ctns-/- mice.

Biochemical assay: We dissected the mouse eye into cornea, iris (plus ciliary body), neural retina and lens, and analysed the cystine levels in each tissue versus age (1, 3, 5, 7, 9, 11, 13, 15-17, 19-21 & 23-25 months). We detected elevated cystine levels in the iris (including the ciliary body) of Ctns-/- mice from 1 month compared to wild-type mice, and in the cornea and retina from 3 months. For the lens, a significant difference was seen from 5 months. The comparison of cystine accumulation for different Ctns-/- tissues showed that cystine levels in the cornea and iris were globally the highest and increased at a greater rate with age. Cystine levels increased dramatically between 5 and 7 months, peaked at 9 months with levels 90- and 260-fold over wild-type for the cornea and iris, respectively; cystine levels were relatively stable from thereon. By comparison, retinal cystine levels increased less dramatically with age and at 9 months were 70-fold over wild-type. Finally, the lens contained the lowest cystine levels: 6-fold over wild-type at 9 months and reaching a maximum of 30-fold at 23 months.

Histological study: We screened for the presence of cystine crystals and lesions versus age (1, 3, 5, 7, 9, 11, 13, 15, 19 and 22 months). We detected rare cystine crystals in the iris and ciliary body from 1 month. The number of crystals increased with age and became abundant from 7 months. Moreover, in 19 and 22 month-old Ctns-/- mice, we noted a fusion of the iris and cornea. In contrast to the iris and ciliary body, we did not detect corneal crystals at 1 month of age. The corneal crystals appeared from 3 months and were located within the keratocytes and distributed throughout the corneal stroma. We did not detect cystine crystals in the epithelium, Bowman’s layer, endothelium or Descemet’s membrane. Finally, in addition to the anterior synechiae, a vascularization of the cornea was also visible from 19 months.

In the choroid and sclera, we first detected cystine crystals from 3 and 5 months, respectively, and their number also increased with age, but at a slower rate. In addition, from 7 months the choroid of Ctns-/- mice was thinner and less homogenous than that of Ctns+/+ controls and continued to degenerate with age. Up to at least 15 months, we could not detect crystals in the retina and both the neural retina and the retinal pigment epithelium (RP) retained a homogeneous appearance in Ctns+/+ and Ctns-/- mice. In contrast, in the 19 and 22 month-old Ctns-/- mice, the RP showed signs of degeneration. Moreover, the photoreceptor segments and nuclei were absent and the inner nuclear layer constituted the most posterior layer of the neural retina and contained rare crystals. Finally, we could not detect cystine crystals in the lens at any age.

Taken together, the appearance of cystine crystals is consistent with the data we obtained from the cystine assay: i) we did not detect cystine crystals in the lens, consistent with the low cystine levels in this tissue (maximum of 2 nmol 1/2 cystine/mg protein); ii) we found rare retinal crystals in the oldest mice when cystine levels reached ~30 nmol 1/2 cystine/mg protein, and iii) we detected cystine crystals in the iris and ciliary body prior to 5 months, however, these became abundant from 7 months, consistent with the sharp 3-fold increase from 30 to 80 nmol 1/2 cystine/mg protein seen at this age. Furthermore, the biochemical data provides an insight into the cystine levels of different eye tissues, which to our knowledge is not known in cystinosis patients, whereas the histological data are consistent with those reported in patients.

Clinical tests: We used four different tests to determine the biological effect of cystine accumulation.

i) Slit lamp photography: We examined the evolution of cystine crystals in the cornea and iris with age (1, 3, 5, 7, 9 & 13 months). We first detected rare crystals at 3 months in both tissues and, consistent with the histological data, crystals were easily detectable and abundant from 7 months. Thus, although corneal crystals appeared slightly later, their evolution paralleled that of the iris, consistent with the overlapping profile of cystine levels in these tissues.

ii) Intra-ocular pressure (IOP): Differences in IOP can be due to obstructed flow of aqueous humour (increased IOP) or defective production by the ciliary body (decreased IOP). Therefore, we indirectly examined the effect of ciliary body crystals on humour production by measuring IOP. We measured the IOP of 6 Ctns+/+ and Ctns-/- mice aged 15-19 months using a Tonopen XL and found that the IOP of the Ctns-/- mice was significantly decreased. These results suggest that the presence of ciliary body crystals could disrupt aqueous humour production and account for this relative decrease in IOP.

iii) Electroretinogram (ERG) profiles: We studied the retinal function of Ctns+/+ and Ctns-/- mice as a function of age (3, 5, 7, 9 & 11 months). We did not detect a significant difference in a-wave or b-wave amplitude between the Ctns+/+ and Ctns-/- response at any age. Taken together, these data suggest that retinal function is preserved in Ctns-/- mice up to 11 months, consistent with the low cystine levels and the absence of retinal cystine crystals.

iv) Light/dark exploration test: We assayed 17-19 month-old Ctns-/- mice for signs of photophobia. Both the Ctns+/+ and Ctns-/- mice exhibited a clear preference for the dark compartment, where they spent almost two thirds of the duration of the session. The Ctns-/- mice, however, showed an increase in the total time spent in the light compartment, but not in the number of entries into the light section or the mean duration of each visit. Interestingly, the increased total time the Ctns-/- mice spent in the light section was due to an increased number of short-duration (inferior to 10 s) visits, which, on average, lasted 5.1 ± 0.3 s. Thus, although a classic photophobic behaviour was not detected, our results highlight a need for the Ctns-/- mice to rapidly re-enter the dark compartment significantly more often than the wild-type mice, which could suggest a mild photophobic state.

Taken together, his temporospatial guide of the ocular anomalies of the Ctns-/- mice, in addition to validating the mouse model, provides a foundation for the future testing of novel therapeutics to efficiently restore a normal ocular phenotype. We are now in a position to continue with the next part of our research proposal, eye targeted in vivo gene transfer studies.

Please note that this work is documented in our manuscript: “A temporospatial guide to the ocular anomalies in a cystinosis mouse model” by V. Kalatzis, N. Serratrice, C. Hippert, O. Payet, C. Arndt, C. Cazavieille, T. Maurice, C. Hamel, F. Malecaze, C. Antignac, A. Müller and E.J. Kremer, which is submitted for publication.

back to top