Synthetic 9-cis-beta-carotene inhibits photoreceptor degeneration in cultures of eye cups from rpe65rd12 mouse model of retinoid cycle defect

The retinoid cycle enzymes regenerate the visual chromophore 11-cis retinal to enable vision. Mutations in the genes encoding the proteins of the retinoid cycle are the leading cause for recessively inherited retinal dystrophies such as retinitis pigmentosa, Leber congenital amaurosis, congenital cone-rod dystrophy and fundus albipunctatus. Currently there is no treatment for these blinding diseases. In previous studies we demonstrated that oral treatment with the 9-cis-β-carotene rich Dunaliella Bardawil algae powder significantly improved visual and retinal functions in patients with retinitis pigmentosa and fundus albipunctatus. Here we developed a convenient and economical synthetic route for biologically active 9-cis-β-carotene from inexpensive building materials and demonstrated that the molecule is stable for at least one month. Synthetic 9-cis-β-carotene rescued cone photoreceptors from degeneration in eye cup cultures of mice with a retinoid cycle genetic defect. This study suggests that synthetic 9-cis-β-carotene may serve as an effective treatment for retinal dystrophies involving the retinoid cycle.

the retinal function in one third of the patients 14 . Since 9CBC is a known precursor of retinal 15 , we hypothesized that the therapeutic effects of 9CBC may have been mediated via production of 9-cis-retinal in vivo. No adverse effects were observed in these studies, suggesting that 9CBC may potentially be used for treatment of retinal dystrophies. Furthermore, preclinical and clinical studies suggest that 9CBC may be used for treatment of other diseases. Thus, oral treatment with the 9CBC-rich alga reduced the severity of chronic plaque psoriasis in a recent clinical trial and inhibited atherogenesis and fatty liver formation in LDL knockout mice 16,17 . In spite of the importance of 9CBC the current experimental therapy is solely based on natural alga products. In order to allow a therapeutic use of the compound there is a need for controlled production of the compound under industrial pharmaceutical settings. Challenges are related not only to the production of the compound but also for its formulation and stabilization.
Here we developed a synthetic route for 9CBC from readily available low cost materials and demonstrated that synthetic 9CBC can rescue retinal photoreceptor cells from degeneration in eye cup cultures from RPE65rd12 mouse model, that lacks the expression of the retinoid cycle RPE65 enzyme. The stability of 9CBC throughout synthesis, purification, storage and packing is vital for preserving its biological activity. In the present study, a challenging matter was ensuring the stability of the 9CBC by development of suitable preserving formulation. A number of experiments shown that oxidation was the main reason causing 9CBC degradation, especially in its pure form (the crude material was significantly less affected by oxidation). Initial stability tests indicated that synthetic 9CBC is sensitive to both oxygen and UV degradation due to the highly conjugated nature of the carbon chain, resulting in full decomposition in less than 24 hours. Attempt to stabilize 9CBC by the addition of antioxidant(s), as suggested in the literature 18 were unsuccessful, as such a composition has quite rapidly degraded. As has been surprisingly found and shown herein, a 9CBC formulation containing, as non-active ingredients, a viscosity enhancing agent such as a lecithin, e.g., a soybean lecithin, a medium chain triglyceride such as miglyol, or a polysorbate such as polysorbate 20 and polysorbate 40 as antioxidants, are highly stable for a long period of time. In a further aspect, the present formulation thus provides a pharmaceutical composition, comprising an active agent selected from 9CBC or a derivative thereof of the formula C as defined above, a thickening-or solidifying-, i.e., viscosity enhancing, agent, and an antioxidant. Such pharmaceutical formulations are useful for treatment of any biological indication in which administration of 9CBC is useful, e.g., neuro-retinal degeneration.

Synthesis of ethyl 4-bromo-3-methylbut-2-enoate. As depicted in
Reformatsky reaction (lactone ring formation) (step a). As depicted in Fig. 1, step a, in a 100 mL round-bottomed flask fitted with a reflux condenser, nitrogen inlet, and addition funnel, 2.5 grams of zinc dust (38.28 mmol) and 4 mL of dry tetrahydrofuran were added. The β-cyclocitral (36.45 mmol, 5.88 mL) and ethyl 4-bromo-3-methylbut-2-enoate (36.51 mmol, 5.52 mL) were added to the addition funnel with 15 mL of dry tetrahydrofuran. The reactants were added dropwise following an initial addition of 1-2 mL to initiate the reaction. The rate of addition was adjusted to maintain a gentle reflux, with ca. 20-30 min for complete addition. The reaction solution was then heated to reflux for 30 minutes. After cooling to room temperature, 15 mL of saturated ammonium chloride solution was added and the solution was stirred for 30 min. The reaction solution was worked up with an addition of another 30 mL of saturated ammonium chloride solution to a 500 mL separatory funnel and diethyl ether extraction (3 × 25 mL). The combined organic layers were washed with brine and dried over MgSO4. The organic phase was evaporated to give 8.95 g of crude product. The product was dissolved in hot hexane and allowed to recrystallize with cooling. The white solid was filtered and collected to yield 4.68 g lactone (55%). 1  Lactone reduction (step b). As depicted in Fig. 1, step b, to 24 mL of tetrahydrofuran (THF) lactone was added (14.08 mmol, 3.3 g). The resulting solution was cooled in a −78 °C bath and with constant stirring DIBAL (18.4 mL, 1.0 M in CH 2 Cl 2 ) was added dropwise during 2 hours, never allowing the solution to warm above −40 °C. After 15-20 minutes the reaction was completed, by TLC monitoring (Et 2 O/Hexane; 40:60). Five hundred ml of 10% H 2 O/THF were added to the cooled solution as a quenching solution. The solution was removed from the bath and allowed to warm to room temperature, while monitoring the rate of warming. After reaching 0°−15 °C, the solution was cooled by using of the cooling bath again. At this point, 2 g of Na 2 SO 4 was added and the resulting suspension was stirred for 0.5 hours. The solution was filtered through a bed of celite, the

Ring opening of lactol (aldehyde formation) (step c).
To a solution of 2.76 g of the lactol in 23 mL of 1,2-dichloroethane 11.5 mL of 1 N HCl solution were added. The two phase solution was heated to 40 °C, and monitored by TLC. The solution turned orange after 30 min. The reaction was found to be complete after 12.5 hours (formed an orange-red solution). Then, 50 mL of saturated Na 2 CO 3 was carefully added to the organic layer with stirring. The aqueous layer was washed with two portions of dichloromethane and the collective organic layers were washed once with brine. The organic solution was then dried over K 2 CO 3 (anhydrous) and concentrated in vacuo to afford a red-orange oil to afford 2.274 grams (60% for steps b and c).

Horner-Emmons reaction (9-cis retinyl ester formation). A solution of triethyl-3-methyl-
4-phosphono-2-butenoate (14.42 mmol, 3.51 mL) in anhydrous THF (21.7 mL) was cooled to 0 °C and treated with anhydrous DMPU (29.7 mmol, 3.57 mL) and n-BuLi in hexanes (6.08 mL of 2.5 M solution, 15.21 mmol) 21 . The mixture was stirred at this temperature for 20 min and then cooled to −78 °C. A solution of aldehyde (7.93 mmol 1.73 g) in THF (21.7 mL) was slowly added and the reaction mixture and stirred at −78 °C for an additional 60 min. The mixture was allowed to warm to 0 °C. A saturated solution of ammonium chloride (15 mL) was added and the mixture extracted with EtOAc (3 × 20 mL). The organic layer was washed with water (2 × 10 mL) and brine (20 mL), dried over MgS0 4 , and concentrated to give 6.33 g red oil. The residue was purified on a short SilicaGel column (95:5 hexane: diethyl ether) to give 1.85 g (71% yield) of the desired ester. 9-cis-retinyl ester was separated from its trans isomer by preparative HPLC (cosmosil cholester packed column 20.0 mm I.D × 250 mm) using elution of CH 3 OH / H 2 O, 90:10 for 4 min then, 90:10 to 95:05 for 4 min, and then 95:05 for 17 min (flow rate 25 mL/min, λ = 402 nm). Retention time was 13.7 min. Analytic HPLC analysis was performed using cosmosil cholester packed column (4.6 mm I.D × 250 mm) using the following method: isocratic elution of CH 3 OH / H 2 O 95:05, in 12 min (flow rate 2 mL/min). Retention time was 8 min (99.6% purity). 1  Ester reduction (9-cis-retinol formation) (step e). To a solution of ethyl 9-cis-retinyl ester (1.674 mmol 0.55 g,) in THF (5 mL) at −78 °C, DIBAL-H (7 mL, 1 M in hexane, 7 mmol) was added and the resulting suspension was stirred for 1 h at −78 °C 22 . After careful addition of 10% aqueous NH 4 Cl, the mixture was extracted with ethyl ether (3 × 10 mL). The combined organic layers were dried over MgSO 4 and concentrated using rotary evaporator. The residue was used in the next step without further purification. Analytic HPLC analysis was done using cosmosil cholester packed column Wittig reaction 9CBC formation (step g). To a stirred solution of 86 mg of 9-cis-retinyl triphenylphosphonium bromide (0.14 mmol) in 2 ml of dry ethanol at 70 °C, all-trans retinal (19.7 mg, 0.069 mmol) in 17 ml of dry ethanol was added dropwise 19 . KOH (0.43 g) dissolved in dry ethanol (5 ml) was added slowly to the mixture, and the solution was stirred at 70 °C for 15 min and then at room temperature for 2 h. The reaction was monitored by TLC (8:2 hexane:ethyl acetate). When no more carotene formation was noted after 2 h, water (10 ml) was  Figure 4 demonstrates the analytic chromatogram of the main fraction obtained (λ = 447 nm) using isocratic elution of 80:20 CH 3 OH:CHCl 3 .
It is also possible to substitute CHCl 3 with methylene chloride for preventing possible damage during separation. Likewise, the final separation can be done with methyl-tert-butyl ether and acetonitrile, instead of the solvents mentioned above.
Stable formulations of 9-cis-β-carotene. Purified 9CBC was found to be unstable under vacuum or nitrogen flushing. In this study, various 9CBC formulations involving a natural or synthetic commercially available antioxidant(s), a viscosity enhancing agent, or their mixture, were prepared, and the chemical stability of some of these formulations was tested. The specific formulations prepared were as following:    Table 1 shows spectral tests confirming the stability of 9CBC in the formulations tested within the specified time period. As clearly indicated, lecithin as an excipient and binder plays a crucial role in stabilizing 9CBC. Interestingly, although a formulation comprising an antioxidant only as the anti-degradative measure has somewhat of a positive effect during the first 14 days, the formulation has rapidly degraded during the next days, wherein the measurable 9CBC by day 22 was almost as low as in the blank formulation. In fact, as shown in Table 1, the amount of 9CBC in an antioxidant-based formulation was significantly lower than that in the formulation that further contained a viscosity enhancing agent (lecithin), already after 14 days, suggesting that 9CBC degradation starts during the first 14 days and intensifies later. In sharp contrast, the measurable amount of 9CBC in the formulation containing both antioxidant and a viscosity enhancing agent was almost 100% compared to the initial amount, indicating the stability of formulation.

Examination of the toxicity of vehicles of synthetic 9CBC in cultures of RPE65rd12 mouse eye cups.
To examine the biological activity of the synthetic 9CBC we chose a mouse model of retinoid cycle defect, RPE65rd12 23 . The RPE65rd12 mice carry a null mutation in the gene encoding the retinoid cycle RPE65 isomerohydrolase and are a widely-used model for RPE65 deficiency. RPE65 is essential for regenerating the rod and cone chromophore 11-cis retinal after it is bleached during light absorption 24 . Loss of RPE65 function results in death of photoreceptors and causes Leber congenital amaurosis (LCA), a severe form of RP 25 . We have recently shown that feeding these mice with an extract of Dunaliella Bardawil algae rich in 9CBC rescued retinal function and preserved cone structure 26 , suggesting that this model may appropriate for testing the biological activity of the synthetic 9CBC. Since only small amount of synthetic 9CBC was available, we examined the biological activity of 9CBC in vitro in eye cup cultures of these mice. Since the BCMO1 enzyme is expressed in the retina in the RPE cells 12 we chose to use whole eye cups that maintain the neuro-retina attached to the underlying RPE tissue.
First, we examined the toxicity of different vehicles to determine which vehicles are compatible for these in vitro experiments. To this aim, eye cups from RPE65rd12 mice at age of 21days were cultivated in vitro for 18 hours in the presence of different vehicles presented in Table 2. Toxicity of compounds was examined following staining for cell death (apoptosis) by TUNEL assay and scored by two masked observers as indicated in the "Methods" section. As shown in Table 2, addition of Miglyol 810 for enhanced viscosity and the "antioxidant mix" at a volume equivalent to addition of 1 μM 9CBC resulted in minimal apoptosis in the neuroretina in these cultures. Based on these findings it was decided to progress with Step f (Fig. 1) -9-cis-retinyl phosphonium salt formation and use Miglyol 810+ "antioxidant mix" as carriers for 9CBC as these reagents demonstrated low toxicity in vitro.

Examination of the biological effects of synthetic 9CBC in eye cup cultures of RPE65rd12 mice.
Next, we examined the biological activity of this formulation of the synthetic 9CBC in retinal eye cup cultures of RPE65rd12 mice. As shown in Fig. 5, in control cultures that contained the carriers Miglyol 810 and "antioxidant mix", hardly any positive staining of s-and m-opsin was evident throughout the eye cups sections, suggesting that both S-and M-cones degenerated (Fig. 5A,C and E). By contrast, in eye cups incubated with 1μM synthetic 9CBC formulated with the Miglyol 810 and the"antioxidant mix",over 2-and 9-fold higher m-opsin and s-opsin positive staining were observed, respectively, compared with control cultures [mean ± SE: 4.53 ± 1.34 vs. 1.83 ± 0.32 m-opsin/mm retina, p = 0.05; 6.33 ± 2.11 vs. 0.77 ± 0.39 s-opsin/mm retina, p = 0.034, respectively, Fig. 5B

,D and E]
There was no significant difference in rhodopsin staining between control cultures and cultures incubated with 9CBC (data not shown).

Discussion
In this study we demonstrated a practical synthetic route for biologically active 9CBC.
Synthetic 9CBC rescued both M-and S-cones from degeneration in cultures of RPE65rd12 mouse eye cups, increasing by 2-and 9-fold the number of cones in the retina, suggesting that the synthetic 9CBC compound may have a therapeutic effect in the degenerating retina. These findings are in accordance with the biological activity of the natural form of 9CBC. Thus, previous clinical trials demonstrated retinal function rescue by oral treatment with 9CBC rich Dunaliella Bardawil algae extract in patients with retinitis pigmentosa and defects in the retinoid cycle 13,14 . In addition in vivo animal studies suggested that feeding with 9CBC rich Dunaliella Bardawil algae extract rescued retinal function and M-cone cell structure in vivo in RPE65rd12 mice 26 and 9CBC extracted from the algae rescued M-cone photoreceptors from degeneration in cultures of RPE65 knockout mice eyes 27 . Our findings that synthetic 9CBC had no significant effect on rhodopsin expression, are in agreement with the biological activity of the algae 9CBC in previous clinical studies and in vivo animal studies as well as with the results of Ozaki et al. 13,14,27 . In future studies we aim to upscale the production of the synthetic 9CBC molecule and test the therapeutic activity and safety of the synthetic molecule in vivo in the RPE65rd12 mouse model as a step towards clinical trials. The motivation of the presented study was to develop the robust experimental protocol for synthetic preparation of 9CBC from inexpensive building materials. We have synthetized 9CBC by step-by-step method from 3,3-dimethylacrylate, but we also can use the alternative approach to start from a commercially available β-Ionone as a starting platform for preparation of the target 9CBC product. Thus, the main focus of next stage of this study is development of a practical cost-effective synthetic route to 9CBC, including the process optimization and further up-scaling leading both to the significant lowering of the 9CBC cost. Nevertheless the main obstacle in the cost effectiveness of 9CBC synthesis is preservation of the final product from degradation, that we successfully overcame using appropriate formulations. 9CBC intermediates were characterized by NMR, MS and HPLC. The final product was found extremely unstable under drying (evaporation of HPLC solvents) by both: vacuum or nitrogen flushing. Most examined formulations proved to be unstable during storage in the presence of light and air, with 9CBC degraded by photodegradative/oxidative etc. Several tests were performed in HPLC (cosmosil cholester packed column 4.6 mm I.D × 250 mm) in order to determine its stability conditions. A number of stabilizers were used in order to preserve 9CBC: butylated hydroxytoluene (BHT) and tert-butylhydroquinone (TBHQ) soybean oil, L-α-lysophosphatidylcholine from glycine max, lecithin, miglyol, tween-40, dimethyl sulfoxide (DMSO) etc. The Miglyol 810 Volume equivalent to addition of 10 μM 9CBC 1 2 "antioxidant mix"* Volume equivalent to addition of 10 μM 9CBC 1 6 Tween 40 and "antioxidant mix"* Volume equivalent to addition of 10 μM 9CBC 4 3 Miglyol 810 and "antioxidant mix"* Volume equivalent to addition of 1 μM 9CBC 1 3 Miglyol 810 and "antioxidant mix"* Volume equivalent to addition of 10 μM 9CBC 3 3 Table 2. Toxicity of vehicles in RPE65/rd12 mouse eye cup cultures. * "antioxidants mix" -a pellet content from the commercially available product CarotenALL, Mixed Carotenoid Complex (Jarrow Formulas) and 9CBC (10:5:1 w/w/w). efficacy of a range of synthetic and natural antioxidants in preventing or at least significantly minimizing degradation of 9CBC during HPLC purification, evaporation and storage were determined. The formulations were spectrally tested for active agent stability within time frame of 1 month. Some of the antioxidant formulations were toxic in the eye cup cultures in vitro. However as these compounds are commonly used as food supplements (e,g. BHT) we plan in future studies to test these formulations in vivo in the RPE65rd12 mouse model where their toxicity is predicted to be significantly reduced.
In conclusion, this study presents a novel, convenient and economical synthetic route for biologically active 9-cis-β-carotene. Up scaling the production of 9CBC may pave the way to development of treatment for incurable retinal dystrophies and other diseases including psoriasis and atherogenesis.

Methods
Materials. Ethyl 3,3-dimethylacrylate was purchased from Sigma Aldrich. b-Cyclocitral was purchased from Alfa Aeser. All trans retinal and triethyl 3-Methyl-4-phosphono-2-butenoate were purchased from TRC. Triphenylphosphonium bromide was purchased from Acros Organics. alpha,alpha-Azoisobutyronitrile (AIBN) was purchased from Molekula. Soybean, Mw = 537.7 gr/mol, purchased from Sigma. All reactions were carried out under argon/nitrogen atmosphere. The reactions (except for steps a and b in Fig. 1) were performed in the dark under red dim light.
Animals. RPE65/rd12 mice were used in this study 23  Eye cup cultures. Fifty two eyes were enucleated from 21 day old RPE65/rd12 mice, and the posterior segment was separated from the cornea, the lens, the iris and the ciliary body under a surgical microscope. The eye cups were placed onto a microporous membrane (30 mm in diameter; Millicell-CM; Millipore, Bedford, MA) with the ganglion cell layer (GCL) facing up and the sclera facing the filter in a six-well culture plate. Each well contained 1 ml of culture medium consisting of 50% minimum essential medium/HEPES (Sigma, St. Louis, MO), 25% HBSS (Invitrogen, USA), and 25% heat inactivated horse serum (Invitrogen, USA) supplemented with 200 μM L-glutamine and 5.75 mg/ml glucose and 1μM synthetic 9CBC or vehicle as indicated in the text. The eye cups were maintained in culture at 37 °C, 5% CO 2 incubator for 18 hours.
Histology analysis. Tissues were fixed in formalin and embedded in paraffin as previously described 28 . Cross sections (4 µm) were cut along the vertical meridian of the eye through the optic nerve and were stained with hematoxylin and eosin. TUNEL staining was performed following manufacturer instructions (Roche, Germany). For detection of rhodopsin and cone opsins, paraffin sections were incubated following citrate buffer antigen retrieval with anti-rhodopsin antibody (RET-P1, Abcam), anti S-(Millipore) and M-(Millipore) opsin and Alexa Fluor ® 488-AffiniPure Donkey Anti-Mouse IgG antibody (Jackson Immuno Research). Following counterstaining with DAPI, sections were examined and photographed by a fluorescent microscope (Olympus BX51).
The Kruskal-Wallis test was used for comparison of m-and s-opsin staining in 9CBC treated vs. control eye cups. Statistical analysis was performed using SPSS for windows version 20.0 (IBM).
Scoring of apoptosis rate. To score the apoptosis rate in the eye cups, sections of eye cups were stained as indicated above and examined by two masked observers. Sections were assigned a score on the scale of 0-4: 0 -no cell death in neuro-retina, 1 -less than 10% of photoreceptors are TUNEL positive, 2-10-50% of photoreceptors are TUNEL positive, 3 -more than 50% of photoreceptors are TUNEL positive, 4-massive cell death in all neuro-retinal layers.