Retinal-chitosan conjugates effectively deliver active chromophores to retinal photoreceptor cells in blind mice and dogs

Article abstract of DOI:10.1124/mol.117.111294

The retinoid (visual) cycle consists of a series of biochemical reactions needed to regenerate the visual chromophore 11-cis-retinal and sustain vision. Genetic or environmental factors affecting chromophore production can lead to blindness. Using animal models that mimic human retinal diseases, we previously demonstrated that mechanism-based pharmacological interventions can maintain vision in otherwise incurable genetic diseases of the retina. Here, we report that after 9-cis-retinal administration to lecithin:retinol acyltransferase-deficient (Lrat²/²) mice, the drug was rapidly absorbed and then cleared within 1 to 2 hours. However, when conjugated to form chitosan-9-cis-retinal, this prodrug was slowly absorbed from the gastrointestinal tract, resulting in sustainable plasma levels of 9-cis-retinol and recovery of visual function without causing elevated levels, as occurs with unconjugated drug treatment. Administration of chitosan-9-cis-retinal conjugate intravitreally in retinal pigment epithelium-specific 65 retinoid isomerase (RPE65)-deficient dogs improved photoreceptor function as assessed by electroretinography. Functional rescue was dose dependent and maintained for several weeks. Dosing via the gastrointestinal tract in canines was found ineffective, most likely due to peculiarities of vitamin A blood transport in canines. Use of the chitosan conjugate in combination with 11-cis-6-ring-retinal, a locked ring analog of 11-cis-retinal that selectively blocks rod opsin consumption of chromophore while largely sparing cone opsins, was found to prolong cone vision in Lrat²/² mice. Development of such combination low-dose regimens to selectively prolong useful cone vision could not only expand retinal disease treatments to include Leber congenital amaurosis but also the age-related decline in human dark adaptation from progressive retinoid cycle deficiency.

Full text of DOI:10.1124/mol.117.111294

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
TITLE PAGE
Retinal-chitosan conjugates effectively deliver active chromophores to retinal
photoreceptor cells in blind mice and dogs
Songqi Gao, Shirin Kahremany, Jianye Zhang, Beata Jastrzebska, Janice Querubin, Simon M.
Petersen-Jones, and Krzysztof Palczewski
Department of Pharmacology and Cleveland Center for Membrane and Structural Biology, School
of Medicine, Case Western Reserve University, Cleveland, OH 44106, USA (S. G., S. K., J. Z.,
B. J., K. P.)
Department of Small Animal Clinical Sciences, College of Veterinary Medicine, Michigan State
University, 736 Wilson Road, D-208., East Lansing, MI 48864, USA (J. Q., S. M. P-J.)
1

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
RUNNING TITLE PAGE
RUNNING TITLE: Retinal-chitosan conjugates for treatment of retinal dystrophies
CORRESPONDENCE
Krzysztof Palczewski, Dept. of Pharmacology, School of Medicine, Case Western Reserve
University, 10900 Euclid Ave., Cleveland, OH 44106. Tel.: 216-368-4631; Fax: 216-368-1300; E-
mail: kxp65@case.edu
MANUSCRIPT STATISTICS
Text Pages: 41
Tables: 1
Figures: 10
References: 46
Abstract Length: 249 words
Introduction Length: 738 words
Discussion Length: 957 words
2

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
ABBREVIATIONS: BSA, bovine serum albumin; BTP, Bis-tris propane; CM-chitosan,
carboxymethyl-chitosan; CWRU, Case Western Reserve University; DDM, dodecyl-β-D-
maltoside; ERG, electroretinography; EDTA, ethylenediaminetetraacetic acid; GI,
gastrointestinal; HPLC, high performance liquid chromatography; LCA, Leber congenital
amaurosis; Lrat, lecithin:retinol acyltransferase; MES, 2-(N-morpholino)ethanesulfonic acid;
MOPS, 4-morpholinepropanesulfonic acid; MWCO, molecular weight cut-off; ONH, optic nerve
head; ONL, outer nuclear layer; RAL, retinal; Rho, rhodopsin; ROL, retinol; ROS, rod outer
segment(s); RPE, retinal pigment epithelium; RPE65, retinal pigment epithelium-specific 65-kDa
protein or retinoid isomerase; SD-OCT, spectral domain-optical coherence tomography; UV,
ultraviolet, Vis, visible.
3

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
ABSTRACT
The retinoid (visual) cycle consists of a series of biochemical reactions needed to regenerate the
visual chromophore, 11-cis-retinal and sustain vision. Genetic or environmental factors affecting
chromophore production can lead to blindness. Using animal models that mimic human retinal
diseases, we previously demonstrated that mechanism-based pharmacological interventions can
maintain vision in otherwise incurable genetic diseases of the retina. Here, we report that after 9-
-/-
cis-retinal administration to lecithin:retinol acyltransferase-deficient (Lrat ) mice, the drug was
rapidly absorbed and then cleared within 1-2 h. However, when conjugated to form chitosan-9-
cis-retinal, this pro-drug was slowly absorbed from the gastrointestinal tract resulting in
sustainable plasma levels of 9-cis-retinol and recovery of visual function without causing elevated
levels, as occurs with unconjugated drug treatment. Administration of chitosan-9-cis-retinal
conjugate intravitreally in retinal pigment epithelium-specific 65 retinoid isomerase (RPE65)-
deficient dogs improved photoreceptor function as assessed by electroretinography. Functional
rescue was dose dependent and maintained for several weeks. Dosing via the gastrointestinal
tract in canines was found ineffective, most likely due to peculiarities of vitamin A blood transport
in canines. Use of the chitosan conjugate in combination with 11-cis-6-ring-retinal, a locked ring
analog of 11-cis-retinal that selectively blocks rod opsin consumption of chromophore while
-/-
largely sparing cone opsins, was found to prolong cone vision in Lrat mice. Development of such
combination low-dose regimens to selectively prolong useful cone vision could not only expand
retinal disease treatments to include Leber congenital amaurosis but also the age-related decline
in human dark adaptation from progressive retinoid cycle deficiency.
4

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
INTRODUCTION
Production of the active chromophore, 11-cis-retinal (11-cis-RAL) is essential for sustaining vision
in all mammalian species (von Lintig et al., 2010). This process, termed the retinoid (visual) cycle,
takes place in photoreceptor cells and the retinal pigment epithelium (RPE), a tightly connected
two-cell system (Kiser et al., 2012; Kiser et al., 2014). An additional cycle involving cone pigments
and Müller cells has also been postulated (Mata et al., 2002; Wang and Kefalov, 2009). The
enzymology of the retinoid cycle has been elucidated in detail at genetic biochemical, and
structural levels (Kiser and Palczewski, 2016; Lamb and Pugh, 2004). Alterations in genes
encoding retinoid-binding proteins and enzymes can result in blinding diseases of varied severity
depending on whether an alternative pathway can fulfill a required function (Kiser and Palczewski,
2016; Travis et al., 2007). For example, defects in 11-cis-RAL dehydrogenase usually cause
rather mild disease, whereas lack of either lecithin:retinol acyltransferase (LRAT) or retinal
pigment epithelium-specific 65 retinoid isomerase (RPE65) lead to the severe early onset
blindness Leber congenital amaurosis (LCA) (Kiser and Palczewski, 2016; Travis et al., 2007).
Thus, a major effort has been made to devise experimental approaches that alleviate lack of
chromophore production, including pharmacological (Batten et al., 2005; Koenekoop et al., 2014;
Scholl et al., 2015; Van Hooser et al., 2000) and genetic (Bainbridge et al., 2008; Cideciyan et al.,
2008; Cideciyan et al., 2013; Maguire et al., 2009) interventions. These initial attempts to correct
retinoid cycle deficiencies require further improvement.
We previously demonstrated that mechanism-based pharmacological interventions can maintain
and even restore vision in otherwise incurable genetic diseases. Rapid improvement of visual
function was observed in patients with RPE65 or LRAT mutations after oral administration of
synthetic 9-cis-retinyl acetate (Batten et al., 2005; Van Hooser et al., 2000). Pharmacological
doses of 9-cis-RAL also dramatically restored visual function in animal models (mice and dogs)
5

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
that lack efficient regeneration of 11-cis-RAL due to dysfunctional or suppressed RPE65 and/or
LRAT (Batten et al., 2004; Gearhart et al., 2010; Maeda et al., 2009a; Van Hooser et al., 2000;
Van Hooser et al., 2002). Systemic pharmacological intervention with activating 9-cis-retinoids, a
substitute for much less stable 11-cis-retinoids, has now moved from preclinical studies to clinical
trials in patients suffering from LCA caused by inactivating mutations in either LRAT or RPE65
(
Koenekoop et al., 2014; Scholl et al., 2015). Moreover, cis-retinoid treatment prevented
deteriorating vision in aging mice (Maeda et al., 2006), raising hope for a similar outcome in
humans.
Two further stages of drug development should be considered to further improve this experimental
treatment. First, retinoid drugs should be designed so they are retained safely after unconjugated
drug administration to lower their potentially toxic systemic exposure during prolonged treatment.
Second, oral delivery of retinoids for treatment of ocular disease is a specialized process involving
their gastrointestinal (GI) absorption, carrier-mediated transport in the blood and specific transfer
into the eye through the blood-retina barrier (Kiser et al., 2014). An additional factor for
consideration is that artificial illumination in modern life renders rod vision in humans less critical
than cone-mediated color vision. So, could a therapy be devised that delivers retinoids in a more
sustainable way preferentially to cone photoreceptor cells? To safely prolong drug exposure,
possibilities include slow-release formulations delivered either systemically or by intraocular
injection. To help preserve cone vision, rod cells could be selectively blocked with a synthetic
retinoid to avoid their consumption of 11-cis-RAL while simultaneously providing a photoactive
compound more specifically to cone cells. This is relevant since rhodopsin (Rho) in rods is present
in much higher quantities than cone visual pigments in the retina.
Here, we found that after oral administration of unconjugated 9-cis-RAL, the drug was rapidly
absorbed resulting in high circulating levels and then cleared within 1-2 h, as published previously
(
Maeda et al., 2013; Maeda et al., 2009b). But when conjugated to chitosan, a linear
6

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
polysaccharide frequently used as a carrier in medical applications (Upadhyaya et al., 2014), to
form chitosan-9-cis-RAL, the pro-drug was slowly absorbed from the GI tract resulting in a
continuous sustainable level of 9-cis-RAL without attaining the high levels that occurred with
unconjugated drug treatment. In the Rpe65-deficient dog model, intravitreal delivery of chitosan-
9-cis-RAL resulted in improvement of both rod and cone function as assessed by
electroretinography. By delivering 9-cis-RAL to cone opsin via chitosan conjugation in
combination with a locked 11-cis-6-ring analog of RAL to block rod opsin, cone vision was found
to be prolonged in mouse and dog models of retinal degeneration.
MATERIALS AND METHODS
-/-
Animals ˗ Male and female C57BL/6J Lrat mice at 4 to 8 weeks of age were used in all
experiments. Mice were housed and maintained in a 12 h light (≤10 lux)/12 h dark cyclic
environment or in a dark room in the Animal Resource Center at the School of Medicine, Case
Western Reserve University (CWRU). All animal handling and protocols were approved by the
Institutional Animal Care and Use Committee at CWRU and conformed to recommendations of
both the American Veterinary Medical Association Panel on Euthanasia and the Association for
Research in Vision and Ophthalmology.
Rpe65-deficient dogs were maintained as a colony at Michigan State University. All procedures
adhered to the Association for Research in Vision and Ophthalmology (ARVO) guidelines for
the use of animals in ophthalmic and vision research and were approved by the institutional animal
care and use committee.
7

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
Materials ˗ Carboxymethyl-chitosan (CM-chitosan) was purchased from Santa Cruz
Biotechnology (Dallas, TX). All other chemicals were purchased from Sigma Aldrich (St. Louis,
MO).
Synthesis of 9-cis-retinyl acetate (1) ˗ All-trans-retinyl acetate (400 g) was dissolved in a
solution of hexanes (800 ml) and trimethylamine (0.22 ml). After addition of 0.4 mg of
o
bis(benzonitrile)palladium chloride, the mixture was stirred overnight in the dark at 65 C under
nitrogen (Supplemental Fig. 1). The resulting solution was cooled to room temperature for 1 h
o
and then further cooled to -80 C. The all-trans-retinyl acetate that crystallized was filtered by
suction using a filter pre-cooled with dry ice. The filtrate was concentrated and further crystallized.
This process was repeated to obtain a 70:30 mixture of 9-cis-/all-trans-retinyl acetate, which was
used for the next steps without further purification.
Synthesis of 3,7-dimethyl-9-(2,6,6-trimethylcyclohexen-1-yl)nona-2,4,6,8-tetraen-1-ol (9-
cis-retinol) (2) ˗ To a solution of retinyl acetate 1 (~55 ml) in ethanol (181 ml), a solution of NaOH
o
(
20 g) in water (84 ml) was added dropwise at 40 C under nitrogen in the dark. After stirring for
o
3
0 min, the mixture was cooled to 0 C and then extracted with hexanes (3 x 200 ml). The
combined organic layers were washed with ice water twice, dried over anhydrous MgSO , filtrated
4
and the filtrate was concentrated under vacuum. The resultant yellow oil was dissolved in a
o
solution of pyridine (0.198 ml) in methyl formate (99 ml), stirred at 0 C for 2 h and then cooled to
o
-
20 C for overnight crystallization to obtain compound 2 (Supplemental Fig. 1).
Synthesis of 3,7-dimethyl-9-(2,6,6-trimethylcyclohexen-1-yl)nona-2,4,6,8-tetraenal (9-cis-
retinal) (3) (Supplemental Fig. 1) ˗ To a solution of 2 (~30 g) in dry methylene chloride (300 ml),
manganese dioxide (180 g) was added in the dark and the mixture was stirred overnight at room
temperature. The mixture was then filtered, and the filtrate evaporated; both steps were
conducted under vacuum. The resulting oil was dissolved in hexanes (90 ml) and cooled to -80
o
1
C. Crystals from the oil were filtered by suction using a filter pre-cooled with dry ice. H NMR
8

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
1
1
(
400 MHz, CDCl ) δ 10.09 (d, J = 8.2 Hz, H), 7.22 (dd, J = 15, 11.6 Hz, H), 6.66 (d, J = 15.8 Hz,
3
1
1
1
1
H), 6.33 (d, J = 15.8 Hz, H), 6.30 (d, J = 15 Hz, H), 6.09 (d, J = 11.6 Hz, H), 5.97 (d, J = 8.2
1
3
2
3
3
2
2
Hz, H), 2.31 (s, H), 2.05 (m, H), 2.02 (s, H), 1.75 (s, H), 1.64 (m, H), 1.48 (m, H), 1.04 (s,
13
6
H); C NMR (100 MHz, CDCl ) δ 191.4, 155.2, 140.4, 138.2, 134.0, 131.5, 131.3, 130.6, 129.5,
3
129.2, 128.1, 39.64, 34.45 33.28, 29.20, 22.09, 21.21, 19.38, 13.40.
11-cis-6-ring-RAL synthesis ˗ A mixture of (E)-2-((Z)-4-((E)-4-(2,6,6-trimethylcyclohex-1-en-1-
yl)but-3-en-2-ylidene)cyclohex-2-en-1-ylidene)acetaldehyde (9,11-dicis), (Z)-2-((E)-4-((E)-4-
(
(
2,6,6-trimethylcyclohex-1-en-1-yl)but-3-en-2-ylidene)cyclohex-2-en-1-ylidene)acetaldehyde
11,13-dicis), (E)-2-((E)-4-((E)-4-(2,6,6-trimethylcyclohex-1-en-1-yl)but-3-en-2-ylidene)cyclohex-
2-en-1-ylidene)acetaldehyde (11-cis), and (Z)-2-((Z)-4-((E)-4-(2,6,6-trimethylcyclohex-1-en-1-
yl)but-3-en-2-ylidene)cyclohex-2-en-1-ylidene)acetaldehyde (9,11,13-tricis) was prepared by the
method reported by Bhattacharya et al. (Bhattacharya et al., 1992). Characterization of these
compounds was reported previously (Alexander et al., 2017; Gulati et al., 2017).
Synthesis and characterization of 9-cis-RAL and 11-cis-6-ring-RAL chitosan conjugates ˗
For synthesis of 9-cis-RAL-chitosan, CM-chitosan, 200 mg in 9 ml of 2% 2-(N-
morpholino)ethanesulfonic acid (MES), pH 6.3, was mixed with 13 mg or up to 87 mg of 9-cis-
RAL in 7 ml ethanol. The mixture was stirred at room temperature for 24 h in a dark room and
then precipitated with ethanol. The color of this mixture changed from light yellow to red. The
precipitate was dried in a Speed-Vac. 9-cis-RAL in the supernatant was measured with a UV-
visible (UV-Vis) spectrophotometer (Cary 50, Varian, Palo Alto, CA). 9-cis RAL content in the
conjugate was determined by subtracting the content of 9-cis-RAL in the supernatant from the
-
1
-1
original loading amount using the extinction efficient ε = 36,068 M cm at 373 nm (Robeson et
al., 1955). Molecular weights (MWs) of the CM-chitosan conjugates were estimated by gel
filtration chromatography. Samples were loaded onto a Superdex 200 gel filtration column (GE
9

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
Healthcare, Little Chalfont, United Kingdom) equilibrated and eluted with PBS (137 mM NaCl, 2.7
mM KCl, 4.3 mM Na HPO , 1.4 mM KH PO ) buffer, pH 7.0, at flow rate of 0.5 ml/min.
2
4
2
4
To synthesize the 11-cis-6-ring-RAL conjugate, CM-chitosan, 245 mg in 10 ml 2% MES, pH 6.3,
was mixed with 10 mg of 11-cis-6-ring-RAL in 7 ml ethanol. The mixture was then processed and
its content calculated as described above for 9-cis-chitosan except that the extinction coefficient
-1
-1
used for 11-cis-6-ring-RAL upon gel filtration chromatography was ε = 24,935 M cm at 376.5
nm (Brown and Wald, 1956).
Synthesis and characterization of 9-cis-RAL chitosan conjugate (a large scale) ˗ To a
solution of 13.2 g CM-chitosan in 660 ml of 0.25X PBS buffer and 400 ml of ethanol, a solution of
1
g 9-cis-RAL in 62 ml ethanol was added dropwise in the dark. The mixture was stirred at room
temperature for 24 h and then concentrated under vacuum. The resultant was dried by
lyophilization to obtain a yellow powder. The concentration of 9-cis-RAL in the conjugate was
calculated as described above.
Retinoid administration ˗ All retinoids were handled under dim red light. 9-cis-RAL or 11-cis-6-
ring-RAL was dissolved in ethanol and mixed with soybean oil, to achieve a final concentration of
10% ethanol except when 9-cis-RAL was dissolved in ethanol prior to oral gavage for drug-in-
plasma profile experiments. Chitosan conjugates were ground into small particles (<120 mesh)
that were suspended in water. Both the soybean oil mixtures and conjugate water suspensions
-/-
were administrated to Lrat mice by oral gavage with a 20 G feeding needle.
Retinoid analysis and pigment measurements ˗ Retinoids in the eyes and plasma were
extracted under dim red light using previously published methods (Palczewski et al., 1999).
Briefly, samples were homogenized in a buffer containing 50 mM 4-morpholinepropanesulfonic
acid (MOPS), 10 mM methoxylamine, pH 7.0, and 50% ethanol, extracted by hexanes, dried in a
Speed-Vac, and dissolved in acetonitrile before injection into an HPLC column as previously
1
0

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
described (Batten et al., 2005). Extracted retinoids dissolved in 75% acetonitrile were separated
on a C18 reverse−phase HPLC column (Agilent Zorbax ODS C18, 5 μm, 4.6 x 250 mm) with a
linear gradient of acetonitrile (75-100%) in water and 0.1% formic acid at a flow rate of 1 ml/min
for 20 min. Elution of retinoids was monitored by the absorbance at 360 nm at room temperature.
Visual pigment levels in the mouse eyes were determined by extraction of a homogenate of two
retinas in 1% dodecyl-β-D-maltopyranoside (DDM) in 50 mM sodium phosphate buffer, pH 7.3.
Homogenized samples were shaken at 4 °C for 2 h, centrifuged at 40,000 g for 30 min, and the
spectrum of the supernatant was measured with a UV-Vis spectrophotometer. Freshly prepared
hydroxylamine (20 mM, pH 7.0) was added to the supernatant prior to determining the spectrum.
Difference spectra were obtained from spectral measurements made before and after bleaching
with white light for 5 min. Concentrations of iso-Rho were calculated based on the extinction
-
1
-1
coefficient, ε 487 nm = 43 000 M cm (Spalink et al., 1983).
Preparation of opsin membranes ˗ Bovine rod outer segment (ROS) membranes were prepared
from fresh retinas under dim red light (Papermaster, 1982). Isolated ROS membranes were
washed free of membrane associated proteins with hypotonic buffer composed of 5 mM Bis-tris
propane (BTP) and 1 mM ethylenediaminetetraacetic acid (EDTA), pH 7.5 by gentle
homogenization and subsequent centrifugation at 25,000 g for 30 min. This procedure was
o
repeated four times at 4 C. The final membrane pellet was suspended in 10 mM sodium
phosphate, pH 7.0, and 50 mM hydroxylamine to a 3 mg/ml concentration of Rho, placed on ice
and illuminated with a 150 W bulb for 30 min. Membranes were pelleted by centrifugation at
16,000 g for 5 min and then washed four times with 10 mM sodium phosphate, pH 7.0, and 2%
bovine serum albumin (BSA) followed by 4 washes with 10 mM sodium phosphate, pH 7.0, and
o
2
washes with 20 mM BTP, pH 7.5, and 100 mM NaCl at 4 C. Rho and opsin concentrations
were measured with a UV-Vis spectrophotometer and quantified by using the absorption
1
1

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
-1
-1
-1
-1
coefficient ε500 nm = 40,600 M cm (Wald and Brown, 1953) and ε280 nm = 81,200 M cm , (Surya
et al., 1995), respectively.
In vitro 9-cis-RAL uptake from the 9-cis-RAL conjugate ˗ Fresh rod opsin membranes were
suspended in 0.5 ml of 20 mM BTP, pH 7.0, to achieve 0.4 mg/ml protein concentration, placed
in a 1.5 ml Eppendorff tube and covered with a 12,000 molecular weight cut-off (MWCO) dialysis
membrane before being dialyzed against 5 ml of 9-cis-RAL conjugate solution or against CM-
chitosan at 4 °C in the dark for 16 h or 5 days, respectively. Next, membranes were solubilized
by addition of DDM to a final concentration of 10 mM for 15 min at room temperature followed by
a 10-min centrifugation at 16,000 g at 4 °C. The supernatant fraction was used to measure the
absorbance spectra. The control experiment involved the use of BSA alone at 0.4 mg/ml. To
quantify the uptake of 9-cis-RAL from the 9-cis-RAL conjugate by rod opsin and BSA, the
absorption spectra were measured at 20 °C with a UV-Vis spectrophotometer.
-/-
Rho purification by 1D4 immunoaffinity chromatography ˗ Eyes collected from Lrat mice
either untreated or treated with 9-cis-RAL or 11-cis-6-ring-RAL were homogenized in 2 ml of 10
mM BTP, pH 7.5, 150 mM NaCl, 1 mM EDTA and protease inhibitors and centrifuged at 13,000
o
g for 20 min at 4 C. The supernatant was discarded, and the pellet was solubilized in 10 mM
o
BTP, pH 7.5, 150 mM NaCl and 20 mM DDM for 1 h at 4 C. Insoluble material was removed by
o
centrifugation for 1 h at 100,000 g and 4 C. The cleared lysate was used to purify Rho by affinity
chromatography on a customized 1D4 antibody-conjugated CNBr resin (Jastrzebska et al., 2009).
The antibody (6 mg/ml resin) was coupled to the CNBr-activated Sepharose 4B according to the
manufacturer’s protocol (GE Healthcare Bio-Sciences Corp, Piscataway NJ). Samples were
eluted by addition of the 1D4 peptide (TETSQVAPA) and their spectra were subsequently
measured with a UV-Vis spectrophotometer.
Electroretinography (ERG) ˗ ERG responses were recorded as previously described (Maeda et
al., 2012). Briefly, dark-adapted mice were anesthetized with a combination of ketamine (6 mg/ml)
1
2

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
and xylazine (0.44 mg/ml) at a dose of 10 µl/g body-weight after their pupils were dilated with 1%
tropicamide. Contact lens electrodes, a reference electrode, and a ground electrode, were
positioned on the corneas, the head between two ears and the tail, respectively. For single-flash
recording, durations of white-light flash stimuli (from 20 μs to 1 ms) were adjusted to provide a
−
2
range of stimulus luminance (from ~3.7 to 1.6 log cd∙s∙m ). Three to five recordings were made
at sufficient intervals between flash stimuli (from 3 s to 1 min) to allow recovery from any
photobleaching effects. A UTAS E-3000 (LKC Technologies, Inc., Gaithersburg. MD) was used
for ERG recording.
Spectral domain-optical coherence tomography (SD-OCT) imaging ˗ Ultra−high resolution
SD−OCT (Bioptigen, Durhan, NC) was used for in vivo imaging of mouse retinas as previously
described (Sundermeier et al., 2014). Briefly, mice were anesthetized by intraperitoneal injection
of an anesthetic cocktail consisting of ketamine (6 mg/ml) and xylazine (0.44 mg/ml) at a dose of
1
0 µl/g body-weight, after their pupils were dilated with 1% tropicamide. Then four frames of OCT
images were acquired in the B−scan mode and averaged. Thicknesses of the outer nuclear layers
ONL) were measured in both the superior and inferior retina with the optic nerve head (ONH)
(
serving as the point of origin.
Histological evaluation ˗ Histological slides were prepared as previously described
(
Sundermeier et al., 2014). Briefly, mouse eyes were fixed for 24 h in a solution containing 4%
paraformalaldehyde and 1% glutaraldehyde in PBS, and then processed through a series of
ethanol, xylenes, and paraffin in a Tissue-Tek VIP automatic processor according to the
manufacturer’s instructions (Sakura, Torrance, CA). Sections (5-μm) were cut and stained with
hematoxylin and eosin (H&E).
Treatment of RPE65-deficient dogs ˗ A pilot study showed no effect on ERG recordings after
oral dosing with 9-cis-RAL-CM-chitosan conjugate. This was anticipated because of the species
specific vitamin A metabolism of dogs (Raila et al., 2002). So, for testing in this model, intravitreal
1
3

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
administration was performed. 9-cis-RAL-CM-chitosan conjugate (124 mg chitosan containing 3
mg of 9-cis-RAL in 3.5 ml water/oil [50:50], 3 mM final concentration of active compound) was
warmed to room temperature and aliquoted in the dark into foil wrapped syringes with the aid of
night vision goggles and injected into the vitreous under dim light conditions. One hundred µl of
full strength conjugate or 100 µl of a 50% or 25% dilution (dilution series 1, 2 and 4) in balanced
salt solution were injected. Successful injections were performed in 2 eyes of 8-month-old Rpe65-
deficient dogs for each concentration (2 female dogs were treated unilaterally for 100%
concentration and 2 males dog bilaterally with 50% dilution in one eye and 25% in the contralateral
eye). Dogs were maintained initially in the dark for 3 days and then under a very dim light. ERGs
2
were performed as previously described (Annear et al., 2011) except that an Espion e visual
electrophysiology system with ColorDome Ganzfeld (Diagnosys, Lowell, MA) was used and only
a limited number of stimuli were tested following International Society for Clinical
Electrophysiology of Vision (ISCEV) protocols (McCulloch et al., 2015).
Statistical analyses ˗ Results were expressed as the mean ± SD. Statistical analyses were
performed with Student t test for two-group comparisons, and two-way ANOVA is used for more
than two group comparisons. P values ≤ 0.05 were considered statistically significant.
RESULTS
The scheme of the synthetic reactions is illustrated in Fig. 1 and Supplementary Fig. 1. 9-cis-
RAL-CM-chitosan conjugate (compound b) was synthesized by reacting CM-chitosan (compound
a) with 9-cis-RAL in MES buffer, pH 6.3. The Schiff-base 9-cis-RAL-CM-chitosan conjugate then
could either be cleaved by hydroxylamine to produce CM-chitosan (compound d) and 9-cis-RAL
oxime or the imine group (compound c) in the Schiff base could be reduced by sodium
borohydride (NaBH ) to form an N-C bond that connects CM-chitosan to 9-cis-retinoid in the new
4
1
4

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
conjugate (compound e). Pre-incubation of CM-chitosan with NaBH did not affect the formation
4
of the 9-cis-RAL CM-chitosan conjugate (compound b) from CM-chitosan and RAL.
The 9-cis-RAL-CM-chitosan conjugate and CM-chitosan were characterized by their UV-Vis
spectra, and their MW was estimated by gel filtration chromatography. The UV-Vis spectra and
gel filtration profiles are shown in Supplemental Fig. 2A and 2B, respectively. The maximum
absorbance of the 9-cis-RAL-CM-chitosan conjugate (compound b) was 350 nm. After reduction,
the maximum absorbance of the CM-chitosan conjugate (compound e) shifted to the shorter
wavelength of 315 nm. There was no absorbance peak between 250-600 nm for CM-chitosan
(
compound a, d). Dextran standards were used to estimate the MWs of the 9-cis-RAL-CM-
chitosan conjugates. The dextran structure is similar to that of chitosan, and both are
polysaccharides. As shown in Supplemental Fig. 2B, the 80 kDa dextran standard peak and the
9-cis-RAL-CM-chitosan conjugate peak were similar. Therefore, the MW of the 9-cis-RAL-CM-
chitosan conjugate was estimated at about 80 kDa.
Release of 9-cis-RAL from the 9-cis-RAL conjugate and its uptake by opsin in vitro ˗ To test
the release of 9-cis-RAL from the 9-cis-RAL conjugate and its ability to regenerate iso-Rho, a
suspension of opsin rod outer segment (ROS) membranes was dialyzed against the 9-cis-RAL
conjugate solution. This allowed only released 9-cis-RAL to penetrate the dialysis membrane and
associate with opsin. Regeneration of iso-Rho was determined by measurement of the UV-Vis
absorption spectra either after 16 h or 5 days of dialysis. A small but distinct absorption peak was
detected at 485 nm after 16 h of incubation (Fig. 2A, red spectrum) that increased after 5 days of
dialysis (Fig. 2B, red spectrum). This indicates that both the release of 9-cis-RAL from the 9-cis-
RAL conjugated to CM-chitosan and its uptake by opsin resulted in the regeneration of iso-Rho
as compared with the sample incubated with CM-chitosan only (Fig. 2A and B, black spectra).
BSA containing a hydrophobic pocket in its structure that can accommodate small hydrophobic
molecules was used for the control experiment. The uptake of RAL released from the RAL-
1
5

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
conjugate CM-chitosan was determined by UV-Vis spectroscopy as an increase in absorption at
360 nm. Although only a negligible change was observed after 16 h (Fig. 2C), a significant
~
spectral change was detected after 5 days (Fig. 2D), indicating again the ability of 9-cis-RAL
released from the 9-cis-RAL conjugate to associate with the hydrophobic binding pocket of certain
proteins.
9-cis-Retinoid levels in plasma after treatment ˗ To evaluate the capability of the CM-chitosan
conjugates to maintain sustained release of 9-cis-retinoid after oral administration, 9-cis-retinoid
-/-
levels in plasma were measured by HPLC. Lrat mice received the CM-chitosan conjugates or
free drug by oral gavage at a dose of 50 mg/kg of either 9-cis-RAL conjugate or 9-cis-RAL. Their
plasma was collected at scheduled time points and analyzed by HPLC. As shown in Fig. 3, 9-cis-
retinoid levels in plasma of the CM-chitosan conjugate treated mice increased gradually, reaching
a peak at about 4 h after oral gavage and then decreased slowly. Even 24 h later, 9-cis-retinoid
was detected in the plasma of CM-chitosan conjugate-treated mice. In contrast, 9-cis-retinoid
levels in the plasma of unconjugated drug treated mice increased rapidly, reaching a peak within
1
h after oral gavage, and then sharply dropped, being barely detectable 10 h later. The 9-cis-
retinoid peak in plasma of unconjugated drug treated mice was 3.2-fold higher than that of CM-
chitosan conjugate treated mice. This higher drug peak, above the therapeutic window, could
potentially cause toxicity and the rapid drug clearance noted in the unconjugated drug treated
group would certainly result in lower drug bioavailability. No 9-cis-retinoid was detected in mice
administered only with chitosan or ethanol in the control experiment.
-/-
Eyes of Lrat mice accumulated higher levels of 9-cis-retinoid after oral administration of
-
the 9-cis-RAL conjugate ˗ HPLC was used to determine 9-cis-retinoid levels in the eyes of Lrat
/-
-/-
mice as shown by a representative chromatogram in Supplemental Fig. 3. Lrat mice were
gavaged with the CM-chitosan conjugate or free drug at a dose of 45 mg/kg of 9-cis-RAL
conjugate or 9-cis-RAL 24 h prior to analysis. Eye samples were collected, homogenized in a
1
6

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
buffer containing 50 mM MOPS and 50 mM O-methylhydroxylamine, pH 7.0, extracted, and
examined by reverse-phase HPLC. The 9-cis-RAL (O-methyl)oxime peaks from the eye samples
were identified based on their elution times and absorbance spectra. The maximum UV
absorbance of 9-cis-RAL (O-methyl)oxime was at 358 nm. The chromatogram reveals that the 9-
cis-retinoid peak from the eyes of the CM-chitosan conjugate treated mice was larger than that of
unconjugated drug treated mice, whereas there was no 9-cis-retinoid peak from the eyes of
-/-
untreated Lrat mice. These higher ocular 9-cis-retinoid levels detected by HPLC indicate that
more visual chromophore is delivered to the eyes of CM-chitosan conjugate-treated mice
compared to unconjugated drug treated mice.
Improvement of retinal function after treatment with the 9-cis-RAL conjugate ˗ To test retinal
-/-
function, we recorded the ERG responses of Lrat mice treated with the CM-chitosan conjugate
or free drug and the a-wave and b-wave amplitudes of the ERG responses are shown in Fig. 4.
-/-
Lrat mice were gavaged with either the CM-chitosan conjugate or free drug at a dose of 35
mg/kg of 9-cis-RAL conjugate or 9-cis-RAL, and kept in the dark room for 3 days prior to single-
flash scotopic ERG recording. Delivery of 9-cis-retinoid from the CM-chitosan conjugate to the
eyes caused significantly increased ERG responses as shown in both a-wave and b-wave
-2
amplitudes beginning at a stimulus intensity of 0.34 and - 2.08 log cd∙s∙m respectively, as
compared with those of unconjugated drug treated mice. Since both the CM-chitosan conjugate
-/-
and free drug treatment can deliver visual chromophore to the eyes of Lrat mice, the ERG
responses of both groups were higher than those of the control group. This ERG response pattern
suggests greater improvement of retinal function in the CM-chitosan conjugate treated group than
-/-
in the free drug treated Lrat mice.
Restored visual pigment iso-Rho in eyes after oral gavage with 9-cis-RAL conjugate ˗ In
some inherited retinal diseases, mutations in retinoid cycle genes such as Lrat cause a decreased
supply of the visual chromophore, 11-cis-RAL (Batten et al., 2004; Batten et al., 2005). Thus,
1
7

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
pharmacological replacement of the missing chromophore is an efficient way to treat this type of
disease. An artificial chromophore, 9-cis-RAL, was selected to replace the missing chromophore
because it is easier to synthesize and thermodynamically more stable than 11-cis-RAL. 9-cis-RAL
binds with opsin to form iso-Rho, which when bleached, undergoes conformational changes
through the same photoproducts as 11-cis-RAL-regenerated Rho. To confirm the formation of
endogenous visual pigment, we measured iso-Rho levels at different time points and dosages
after oral gavage with either 9-cis-RAL or 9-cis-RAL conjugate. The findings are shown in Fig. 5.
With CM-chitosan conjugates or free drug given at a dose of 150 mg/kg of 9-cis-RAL conjugate
-/-
or 9-cis-RAL, iso-Rho gradually accumulated in the Lrat mouse retina, achieving a plateau level
of 230 ± 20 pmol/eye at 14 h for the free drug group and at 24 h for the CM-chitosan conjugate
group (Fig. 5A). This level corresponds to ~50% of the level of visual pigment found in age-
-/-
matched WT mice (400-600 pmol/eye). There was no iso-Rho accumulation in the eyes of Lrat
mice that received only soybean oil as a control. The iso-Rho accumulation rate in the CM-
chitosan conjugate treated group was slower than that of the free drug treated group because 9-
cis-RAL had to be first released from the conjugate before its systemic delivery could occur.
However, slower but sustained drug release from the longer lasting CM-chitosan conjugates is
desired to prevent its toxicity and prolong its effect. In fact, our in vitro experiment shown in Fig.
2
indicated that a longer time is needed to regenerate iso-Rho when opsin membranes are
incubated with 9-cis-RAL conjugate compared to free retinoid (Jastrzebska et al., 2013; McKibbin
-/-
et al., 2007; Tian et al., 2017). To test the dosage effect on pigment accumulation, Lrat mice
were gavaged with the CM-chitosan conjugate or free drug at doses ranging from 0 to 180 mg/kg
of 9-cis-RAL conjugate or 9-cis-RAL two days prior to analysis. Iso-Rho accumulated gradually in
the retinas with increasing drug doses, reaching a plateau level of 235 ± 21 pmol/eye at 70 mg/kg
for the CM-chitosan conjugate treated group and 100 mg/kg for the unconjugated drug treated
group (Fig. 5B). Before reaching the plateau level, more iso-Rho accumulated in the retinas of
CM-chitosan conjugate treated mice than in free drug treated mice at the same dosage. The
1
8

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
presence of iso-Rho was confirmed by UV-Vis spectroscopy wherein its maximum absorbance
wavelength was 487 nm (Spalink et al., 1983) (Fig. 5C); this is blue-shifted by 12 nm as compared
to Rho at 498 nm .
Percentages of 9-cis-retinoid contained in the CM-chitosan conjugates affect ocular iso-
Rho levels and ERG responses ˗ To examine the influence of water-solubility of the CM-
chitosan conjugates on visual pigment formation and retinal function, we used four CM-chitosan
conjugates containing 6% (water-soluble), 10% (partially water-soluble), 20% (water-insoluble),
-/-
and 30% (water-insoluble) 9-cis-retinoid (w/w). Lrat mice were gavaged with the CM-chitosan
conjugates at the same dose of 90 mg/kg of 9-cis-RAL equivalent two days prior to analysis of
iso-Rho levels and ERG recording. Oral administration of the CM-chitosan conjugates at the same
dose of 9-cis-retinoid led to different levels of ocular iso-Rho due to variations in the water-
solubility of the CM-chitosan conjugates (Supplemental Fig. 4A). A water insoluble conjugate
with 30% 9-cis-retinoid content led to a significantly lower ocular level of iso-Rho, 142 ± 28
pmol/eye, as compared to 236 ± 29 pmol/eye for a more water-soluble conjugate (6% 9-cis-
retinoid content, P < 0.001). In addition, ERG responses were recorded to assess the impact of
9-cis-retinoid content in the CM-chitosan conjugates on retinal function. Amplitudes of b-waves at
-2
1
.6 log cd·s·m under scotopic conditions are shown in Supplemental Fig. 4B, wherein a 9-cis-
retinoid content-dependent ERG response was recorded. Here the water-soluble and lower 9-cis-
retinoid-content (6%) conjugate produced significantly higher b-wave amplitudes (700 ± 110 µV)
compared to the water-insoluble and higher 9-cis-retinoid-content (30%) conjugate (390 ± 80 µV)
(
P < 0.001). These data also indicate that the pattern of ERG responses is similar to that of ocular
iso-Rho accumulation.
Effect of 9-cis-RAL-CM-chitosan conjugate treatment on retinal pathology ˗ To assess the
effect of 9-cis-RAL-CM-chitosan conjugate treatment on retinal morphology, we gavaged 4-week-
-/-
old Lrat mice with the CM-chitosan conjugate at a dose of 80 mg/kg of 9-cis-RAL equivalent
1
9

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
every 3 days for 4 weeks. Fig. 6A and 6B display H&E-stained paraffin sections from the eyes of
-/-
8
-week-old Lrat mice treated with or without the CM-chitosan conjugate. Histological analysis
showed that multiple gavages effectively improved retinal morphology. The ROS of the conjugate-
treated mice was thicker and more tightly packed than that of untreated control mice. The
thickness of the ROS layer was substantially improved from 9.0 ± 1.0 μm in control mice to 12.3
±
1.0 μm for the conjugate-treated mice (Fig. 6C). Since ROS structural morphology depends on
functional Rho, regeneration of visual pigments preserves retinal morphology. Further, visual
pigment formation requires 9-cis-retinoid that can be supplied efficiently with sustained release of
this retinoid from CM-chitosan conjugates. Our UV-Vis spectra showed that opsin in the retina
was regenerated and formed iso-Rho (Fig. 5C), suggesting this opsin was properly folded.
-/-
We also analyzed retinal morphology in 8-week-old Lrat mice using ultra-high resolution SD-
OCT to assess the treatment efficacy of CM-chitosan conjugates. At a younger age, the retinal
-/-
layers of Lrat mice are similar to those of WT mice. But at 8 weeks of age, the ONL thickness of
-/-
Lrat mice is 45 ± 3 μm, compared to 50 ± 3 μm for WT mice. SD-OCT images revealed that the
-/-
ONL morphology of Lrat mice treated with 9-cis-RAL-CM-chitosan conjugate appears
unchanged as compared to untreated mice. Supplemental Fig. 5A and 5B show that SD-OCT
-/-
images of retinas from 8-week-old Lrat mice treated with or without the conjugate. Quantification
-/-
of ONL thicknesses in Lrat mice demonstrated that there was no significant difference between
conjugate-treated and untreated control groups (Supplemental Fig. 5C).
The 11-cis-6-ring-RAL-CM-chitosan conjugate ˗ Previously, we found that 6-ring-locked RAL
11
12
with a ring structure that prevents isomerization around the C =C double bond can specifically
bind to rod opsin in vitro and in vivo. Rod cells were only marginally active with 11-cis-6-ring-
retinal even when exposed to bright light (Jang et al., 2001; Kuksa et al., 2002). Green cone
pigment does not regenerate with this chromophore (Alexander et al., 2017).
2
0

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
The 11-cis-6-ring-RAL-CM-chitosan conjugate (compound C1) was synthesized by a method
similar to that used for the 9-cis-RAL-CM-chitosan conjugate. The scheme of synthetic reactions
is shown in Fig. 7A. CM-chitosan reacted with 11-cis-6-ring-RAL in MES buffer, pH 6.3, to
produce the 11-cis-6-ring-RAL-CM-chitosan conjugate (compound C1) that is a Schiff-base. The
imine group in the Schiff-base compound C1 can be reduced by reaction with NaBH to form a
4
new CM-chitosan conjugate, compound C2. Both these 11-cis-6-ring-RAL-CM-chitosan
conjugates were characterized by their UV spectra (Fig. 7B). The maximum absorbance of the
11-cis-6-ring-RAL-CM-chitosan conjugate (compound C1) was 363 nm. Reduction of the Schiff-
base led to a blue-shift of 46 nm documented by UV-Vis spectroscopy of the maximum
absorbance of the CM-chitosan conjugate (compound C2, λ = 317 nm) as compared to compound
C1 (λ = 363 nm). Moreover, a red-shift of 11-cis-6-ring-RAL-CM-chitosan conjugate (compound
C1, λ = 363 nm) was observed compared to 9-cis-RAL-CM-chitosan conjugate (λ = 350 nm). We
attribute this finding to the red-shift of the 11-cis-6-ring-RAL monomer (λ = 374 nm) relative to the
9-cis-RAL monomer (λ = 365 nm).
Regeneration of visual pigment with both 9-cis-RAL and 11-cis-6-ring-RAL chromophore
analogs in vivo and in vitro ˗ Visual pigment was regenerated with 9-cis-RAL or 11-cis-6-ring-
-/-
RAL chromophore analogs in vivo in Lrat mice lacking production of the endogenous
chromophore 11-cis-RAL and in vitro in opsin containing ROS membranes isolated from bovine
-/-
retinas. To determine pigment regeneration in vivo, eyes from treated and untreated Lrat mice
were collected and pigments were purified by 1D4 immunoaffinity chromatography. In parallel,
pigments were regenerated with chromophore analogs in bovine ROS membranes. UV-Vis
spectra obtained from the purified proteins indicated formation of iso-Rho. In both proteins purified
from mouse eyes and bovine ROS, the appearance of an absorption maximum peak at 485 nm
after regeneration with 9-cis-RAL and an absorption maximum peak at 505 nm after regeneration
with 11-cis-6-ring-RAL were detected, whereas no comparable absorption maximum peaks were
2
1

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
observed in protein samples purified from untreated mice and untreated bovine ROS (Figs. 8A
and 8B).
11-cis-6-ring-RAL-CM-chitosan conjugate and 9-cis-RAL-CM-chitosan conjugate improve
-/-
RAL function in Lrat mice ˗ To compare 11-cis-6-ring-RAL and its chitosan conjugate with 9-
-/-
cis-RAL and its conjugate in rescuing of visual function, 4-week-old Lrat mice were reared in the
-/-
dark and divided into 8 groups. For the Lrat control group, mice were gavaged with soybean oil
or CM-chitosan. For the 11-cis-6-ring-RAL group, mice were gavaged with 11-cis-6-ring-RAL or
11-cis-6-ring-RAL-CM-chitosan conjugate at a dose of 40 mg/kg of 11-cis-6-ring-RAL or its
equivalent 48 h prior to ERG evaluation. For the 9-cis-RAL group, mice were gavaged with 9-cis-
RAL or 9-cis-RAL-CM-chitosan conjugate at a dose of 40 mg/kg of 9-cis-RAL or its equivalent 48
h prior to ERG evaluation. For the 11-cis-6-ring-RAL plus 9-cis-RAL group, mice were gavaged
with 11-cis-6-ring-RAL or 11-cis-6-ring-RAL-CM-chitosan conjugate at a dose of 40 mg/kg of 11-
cis-6-ring-RAL or its equivalent at day 1 and day 3, and 9-cis-RAL or 9-cis-RAL-CM-chitosan
conjugate at a dose of 40 mg/kg of 9-cis-RAL or its equivalent at day 5, and the ERG evaluation
was performed at day 7. ERG responses were recorded under both scotopic and photopic
-2
conditions, and amplitudes of the b-waves at 1.6 log cd·s·m are summarized in Fig. 9A. For
scotopic ERG b-waves, treatment with 9-cis-RAL or 11-cis-6-ring-RAL produced significant
increases in amplitudes compared to the vehicle control group; treatment with conjugates
produced significant increases in amplitude compared to the unconjugated drug treatment groups;
treatment with 9-cis-RAL produced significant increases in amplitudes compared to the 11-cis-6-
ring-RAL treatment group; treatment with 11-cis-6-ring-RAL plus 9-cis-RAL produced significant
increases in amplitudes compared to the 9-cis-RAL treatment group. UV-Vis absorption spectra
of rod pigment purified from mice treated with 9-cis-RAL or 11-cis-6-ring-RAL indicated pigment
regeneration with a peak of absorption at 485 nm for 9-cis-RAL regenerated iso-Rho and a smaller
peak at 505 nm for 11-cis-6-ring-RAL regenerated pigment (Fig. 9B). Pigment isolated from mice
2
2

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
treated with 9-cis-RAL and 11-cis-6-ring-RAL had the highest peak of absorption among all 4
groups, in parallel with the highest scotopic ERG responses. Scotopic ERGs reflect the response
primarily from rod cells, and their increase in amplitudes indicates restoration of rod pigment and
improvement of rod cell function. For photopic ERG b-waves, treatment with 11-cis-6-ring-RAL or
9-cis-RAL produced significant increases in amplitudes compared to the vehicle-treatment control
group. However, there was no significant difference in the amplitudes of the photopic ERG b-
waves between all treatment groups. Because photopic ERGs reflect the response primarily from
cone cells, similar amplitudes of photopic ERG b-waves indicate formation of cone pigment at
similar levels, resulting in the similar improvement of cone cell function in all treatment groups.
This suggests that 11-cis-6-ring-RAL does not affect cone pigment regeneration as published
previously (Alexander et al., 2017), but may block rhodopsin photochemistry.
9-cis-RAL-CM-chitosan conjugate improves photoreceptor function in RPE65-deficient
dogs ˗ The scotopic and photopic ERG responses of RPE65-deficient dogs prior to treatment
were below detectable limits to most stimuli (<1 µV in amplitude) but in some dogs a very small
response to stronger stimuli was noted (1 – 3.5 µV in amplitude, Supplemental Tables 1-6).
Following intravitreal injection of 9-cis-RAL-CM-chitosan, scotopic and photopic ERG responses
were present indicating improvement in both rod and cone responses (Fig. 10A and B, Table 1,
and Supplemental Tables 1-6). There also appeared to be a dose effect (Fig. 10C). Table 1
-2
shows the scotopic ERG a- and b-wave amplitudes to stimuli of 3 and 10 cd·s·m for RPE65-
deficient dogs treated with intravitreal 9-cis-RAL-CM-chitosan conjugate with statistical
comparison of the mean ERG amplitudes pre- and post-treatment. The peak recorded responses
were at 10 days post injection after which there was a slow and progressive decline in amplitudes
such that at 54 days post injection the mean ERG amplitudes in the eyes injected with 9-cis-RAL-
CM-chitosan at 100% concentration had returned to almost pre-injection levels (Fig. 10D). There
was a more rapid decline in the photopic ERG rescue and at lower doses (data not shown).
2
3

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
DISCUSSION
Many retinal degenerative diseases are known to be associated with defects in the retinoid
(
visual) cycle (Travis et al., 2007). Thanks to painstaking biochemical reconstitution studies
supported by genetically engineered animal models and genetic/phenotypic studies of humans
with specific blinding diseases, a molecular understanding of the retinoid cycle in the mammalian
retina has advanced appreciably over the past few years (Kiser et al., 2014). Understanding the
fundamental biochemical processes underlying diseases that involve the retinoid cycle is then
essential for the development of effective therapeutics. A successful clinical trial recently reported
that pharmacological doses of 9-cis-RAL could dramatically restore visual function in humans with
genetic disease involving the retinoid cycle (Koenekoop et al., 2014). Because artificial
illumination renders rod vision less critical than cone vision we need to develop improved methods
for sustained delivery of retinoids predominantly to improve the function and preserve the
structure of cone photoreceptors.
We already have demonstrated that mechanism-based pharmacological interventions can restore
vision in otherwise incurable genetic retinal dystrophies and further improvements are possible
(
Batten et al., 2005; Van Hooser et al., 2000; Van Hooser et al., 2002). The current study centers
on two straightforward findings: first, we found that conjugation of 9-cis-RAL with chitosan results
in a sustained 9-cis-RAL release profile. This finding can be translated into improvements in
retinoid-based therapeutic strategies. Second, we also found that ring-locked retinal that prevents
11
12
isomerization around the C =C double bond, specifically binds and blocks rod opsin but not
cone opsins, thereby permitting a specific reduction of chromophore consumption by rod cells
and increasing retinoid availability for the regeneration of cone pigments. The simple chemistry
and direct conjugation of retinoids to CM-chitosan could be a potentially outstanding way for
sustained delivery of these compounds. Though this complicates the treatment and approval
2
4

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
process, a major benefit could be lower doses of therapeutic retinoids, thereby minimizing their
potential side effects. While no toxicology profile is available for 11-cis-6-ring-RAL, specifically
how this compound could affect retinoid acid signaling when oxidized, our data support further
work in this area of therapeutic intervention. With approaches in gene therapy recently approved
by the FDA for LCA patients, chromophore supplementation therapy as described in his study
could serve in the capacity as a combination treatment. One limitation, however, should be noted,
that retinoid therapy would not be recommended for pregnant women because it may interfere
with visiatin A in the developing fetus (Chassaing et al., 2009; Mahan and Vallet, 1997).
Chitosan, obtained from chitin shells of shrimps by base hydrolysis, and its carboxymethylated
derivative CM-chitosan are used as carriers for drug delivery. Both appear to be safe as they are
also used as part of antibacterial preparations, cosmetics, food preservatives and included in
bandages to reduce bleeding. Chitosan, composed of randomly distributed β-(1→4)-linked D-
glucosamine and N-acetylated D-glucosamine has little if any toxicity, but is highly biodegradable
(
Hu et al., 2017; Kalliola et al., 2017; Paolicelli et al., 2009; Upadhyaya et al., 2014). The free
amino group of chitosan has a pKa of about 6.5, making it ideal for Schiff base conjugation of
functional aldehydes (like RAL). Chitosan, modified by carboxymethylation, displays a marked
increase in solubility at neutral and alkaline pH without adversely affecting its other properties.
However, it has yet to be approved by the FDA as a vehicle for either drug or non-viral gene
delivery.
The water solubility of a 9-cis-RAL-CM-chitosan conjugate depends on its content of 9-cis-
retinoid. Because CM-chitosan is a water-soluble polysaccharide, whereas 9-cis-retinoid is
hydrophobic, the higher the content of 9-cis-retinoid in the CM-chitosan conjugate, the lower its
water solubility. The 9-cis-RAL-CM-chitosan conjugate is produced by attaching 9-cis-RAL to CM-
chitosan via a covalent bond. The higher 9-cis-retinoid content means that more 9-cis-retinoid is
attached to a given amount of CM-chitosan, making the CM-chitosan conjugate more
2
5

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
hydrophobic. Also, hydrophobic portions of the CM-chitosan conjugate can aggregate, and thus
lower aqueous solubility. Moreover, water-insoluble CM-chitosan conjugates can block cleavage
of the covalent bond that connects the drug to CM-chitosan, and thereby hinder drug release from
the CM-chitosan conjugate, allowing for a more sustained drug delivery.
To demonstrate the ability of the 9-cis-RAL-CM-chitosan conjugate to supply 9-cis-RAL and
improve photoreceptor function in a large animal model, the RPE65-deficient dog was chosen.
Our previous studies with this model, which lacks visual cycle function, showed that direct
intravitreal administration of 9-cis-RAL temporarily improved retinal function (Gearhart et al.,
2010). The dog, and other carnivores, differ from other mammals in their circulatory vitamin A
transport (Raila et al., 2002). In this respect, vitamin A metabolism in mice more closely resembles
the conditions in humans (Blaner et al., 2016). A pilot study of oral administration of the 9-cis-
RAL-CM-chitosan conjugate failed to rescue photoreceptor function in the RPE65-deficient dog,
most likely because of the vitamin A transport peculiarities in this species. Therefore, to test the
bioavailability of 9-cis-RAL from the conjugate, intravitreal administration was used. This resulted
in rescue of rod and cone photoreceptor function, further confirming that 9-cis-RAL was available
to the photoreceptors. The ERG amplitudes were markedly improved over those of untreated
RPE65-deficient dogs, and achieved ~ 25% of wild-type levels in response to the ISCEV standard
flash stimulus. Our previous study using unconjugated intravitreal 9-cis-RAL had resulted in a
similar degree of ERG rescue (Gearhart et al., 2010).
In summary, defects in most retinoid cycle genes can cause retinal degeneration in humans,
ranging from childhood diseases such as LCA to late-onset disease. Both gene therapy and
pharmacological approaches have been translated into the clinic where promising first-in-human
data were obtained, but limitations of these therapies also were noted (Zhang et al., 2015). Due
to better understanding of the retinoid cycle, improved interventions are now feasible. For
example, the interplay between rod and cone photoreceptor regeneration described here
2
6

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
promises to make it possible to deliver smaller amounts of active chromophore exclusively to
cone cells, while blocking rod bleaching and regeneration cycles during day-time vision.
Moreover, the higher levels of active agent following the delivery of the unconjugated drug can
cause toxic side effects, but this problem can be alleviated by the slower delivery of chromophore
through the digestive tract using simple conjugated pro-drugs. As demonstrated in two animal
models deficient in the production of active chromophore, administration of conjugated 9-cis-RAL
is an effective approach to improve retinal function and when it is possible to deliver it in a
sustained manner via the GI tract has advantageous properties over simple dosing with free
retinoid.
ACKNOWLEDGEMENTS
We thank Dr. Leslie T. Webster, Jr. and the members of the Palczewski laboratory for their
valuable comments regarding this manuscript. We are also grateful to Dr. Muneto Mogi (Novartis
Inc.) for 11-cis-6-ring-RAL. We thank Heather Defore for assistance with the dog studies.
AUTHORSHIP CONTRIBUTIONS
Participated in research design: S. G., S. K., J. Z., K. P., B. J., J. Q., S. M. P-J.
Conducted experiments: S. G., S. K., J. Z., B. J., J. Q., S. M. P-J.
Contributed new reagents or analytical tools: S. K.
Performed data analysis: S. G., S. K., J. Z., K. P., B. J., J. Q., S. M. P-J.
Wrote or contributed to writing of the manuscript: S. G., S. K., J. Z., K. P., B. J., S. M. P-J.
2
7

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
REFERENCES
Alexander NS, Katayama K, Sun W, Salom D, Gulati S, Zhang J, Mogi M, Palczewski K
and Jastrzebska B (2017) Complex binding pathways determine the regeneration
of mammalian green cone opsin with a locked retinal analogue. J Biol Chem
2
92(26): 10983-10997.
Annear MJ, Bartoe JT, Barker SE, Smith AJ, Curran PG, Bainbridge JW, Ali RR and
Petersen-Jones SM (2011) Gene therapy in the second eye of RPE65-deficient
dogs improves retinal function. Gene Ther 18(1): 53-61.
Bainbridge JW, Smith AJ, Barker SS, Robbie S, Henderson R, Balaggan K, Viswanathan
A, Holder GE, Stockman A, Tyler N, Petersen-Jones S, Bhattacharya SS, Thrasher
AJ, Fitzke FW, Carter BJ, Rubin GS, Moore AT and Ali RR (2008) Effect of gene
therapy on visual function in Leber's congenital amaurosis. N Engl J Med 358(21):
2
231-2239.
Batten ML, Imanishi Y, Maeda T, Tu DC, Moise AR, Bronson D, Possin D, Van Gelder
RN, Baehr W and Palczewski K (2004) Lecithin-retinol acyltransferase is essential
for accumulation of all-trans-retinyl esters in the eye and in the liver. J Biol Chem
2
79(11): 10422-10432.
Batten ML, Imanishi Y, Tu DC, Doan T, Zhu L, Pang J, Glushakova L, Moise AR, Baehr
W, Van Gelder RN, Hauswirth WW, Rieke F and Palczewski K (2005)
Pharmacological and rAAV gene therapy rescue of visual functions in a blind
mouse model of Leber congenital amaurosis. PLoS Med 2(11): e333.
2
8

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
Bhattacharya S, Ridge KD, Knox BE and Khorana HG (1992) Light-stable rhodopsin. I. A
rhodopsin analog reconstituted with a nonisomerizable 11-cis retinal derivative. J
Biol Chem 267(10): 6763-6769.
Blaner WS, Li Y, Brun PJ, Yuen JJ, Lee SA and Clugston RD (2016) Vitamin A
Absorption, Storage and Mobilization. Subcell Biochem 81: 95-125.
Brown PK and Wald G (1956) The neo-b isomer of vitamin A and retinene. J Biol Chem
2
22(2): 865-877.
Chassaing N, Golzio C, Odent S, Lequeux L, Vigouroux A, Martinovic-Bouriel J, Tiziano
FD, Masini L, Piro F, Maragliano G, Delezoide AL, Attie-Bitach T, Manouvrier-Hanu
S, Etchevers HC and Calvas P (2009) Phenotypic spectrum of STRA6 mutations:
from Matthew-Wood syndrome to non-lethal anophthalmia. Hum Mutat 30(5):
E673-681.
Cideciyan AV, Aleman TS, Boye SL, Schwartz SB, Kaushal S, Roman AJ, Pang JJ,
Sumaroka A, Windsor EA, Wilson JM, Flotte TR, Fishman GA, Heon E, Stone EM,
Byrne BJ, Jacobson SG and Hauswirth WW (2008) Human gene therapy for
RPE65 isomerase deficiency activates the retinoid cycle of vision but with slow rod
kinetics. Proc Natl Acad Sci U S A 105(39): 15112-15117.
Cideciyan AV, Jacobson SG, Beltran WA, Sumaroka A, Swider M, Iwabe S, Roman AJ,
Olivares MB, Schwartz SB, Komaromy AM, Hauswirth WW and Aguirre GD (2013)
Human retinal gene therapy for Leber congenital amaurosis shows advancing
retinal degeneration despite enduring visual improvement. Proc Natl Acad Sci U S
A 110(6): E517-525.
2
9

Molecular Pharmacology Fast Forward. Published on February 16, 2018 as DOI: 10.1124/mol.117.111294
This article has not been copyedited and formatted. The final version may differ from this version.
MOL #111294
Gearhart PM, Gearhart C, Thompson DA and Petersen-Jones SM (2010) Improvement
of visual performance with intravitreal administration of 9-cis-retinal in Rpe65-
mutant dogs. Arch Ophthalmol 128(11): 1442-1448.
Gulati S, Jastrzebska B, Banerjee S, Placeres AL, Miszta P, Gao S, Gunderson K,
Tochtrop GP, Filipek S, Katayama K, Kiser PD, Mogi M, Stewart PL and
Palczewski K (2017) Photocyclic behavior of rhodopsin induced by an atypical
isomerization mechanism. Proc Natl Acad Sci U S A 114(13): E2608-E2615.
Hu L, Zhang P, Wang X, Cheng X, Qin J and Tang R (2017) pH-sensitive carboxymethyl
chitosan hydrogels via acid-labile ortho ester linkage for potential biomedical
applications. Carbohydr Polym 178: 166-179.
Jang GF, Kuksa V, Filipek S, Bartl F, Ritter E, Gelb MH, Hofmann KP and Palczewski K
(
2001) Mechanism of rhodopsin activation as examined with ring-constrained
retinal analogs and the crystal structure of the ground state protein. J Biol Chem
76(28): 26148-26153.
2
Jastrzebska B, Golczak M, Fotiadis D, Engel A and Palczewski K (2009) Isolation and
functional characterization of a stable complex between photoactivated rhodopsin
and the G protein, transducin. FASEB J 23(2): 371-381.
Jastrzebska B, Orban T, Golczak M, Engel A and Palczewski K (2013) Asymmetry of the
rhodopsin dimer in complex with transducin. FASEB J 27(4): 1572-1584.
Kalliola S, Repo E, Srivastava V, Heiskanen JP, Sirvio JA, Liimatainen H and Sillanpaa
M (2017) The pH sensitive properties of carboxymethyl chitosan nanoparticles
cross-linked with calcium ions. Colloids Surf B Biointerfaces 153: 229-236.
3
0