Parallel visual cycles in the zebrafish retina
Valerie C. Fleisch1 and Stephan C.F. Neuhauss2
1
Department of Biological Sciences & Centre for Prions and Protein Folding Diseases
Biological Sciences, University of Alberta, Edmonton, Canada,
2
University of Zurich, Institute of Molecular Life Sciences; Neuroscience Center Zurich and
Center for Integrative Human Physiology, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland
Corresponding author:
Stephan C.F. Neuhauss
Telephone: +41 446356040
Telefax: +416356897
[email protected]
Abstract
Vertebrate vision necessitates continuous recycling of the chromophore 11-cis retinal (RAL). The
classical (or canonical) visual cycle employs a number of enzymes located in the photoreceptor
outer segment and RPE (retinal pigment epithelium) of the retina to regenerate 11-cis RAL from
all-trans RAL. Cone-dominant species are believed to utilize a second, intra-retinal, pathway for
11-cis RAL generation, involving retinal Müller glia cells. This review summarizes the efforts
made in zebrafish to gain a better understanding of the role of these two visual cycles for rod and
cone
photoreceptor
chromophore
recycling.
1
Article Outline
1. Introduction .................................................................................................................................. 3
1.1.
Vertebrate duplex retina .................................................................................................... 3
1.2.
Canonical retinoid recycling in vertebrate eyes ................................................................ 4
2.
Alternative cone visual cycle in cone-dominant species .......................................................... 5
3.
Model system zebrafish ............................................................................................................ 8
4.
3.1.
Zebrafish as a model organism .......................................................................................... 8
3.2.
Zebrafish visual system ..................................................................................................... 9
Cone visual cycle in zebrafish ................................................................................................ 10
4.1.
Vision in canonical cycle-deficient zebrafish .............................................................. 10
4.2.
Müller cells are implicated in the intra-retinal visual cycle ............................................ 12
4.3.
Subfunctionalization of zebrafish Cralbp provides evidence for the existence of two
pathways for visual pigment recycling in the zebrafish ............................................................. 14
4.4.
4.5.
Cone photoreceptors use both, the canonical and the intra-retinal visual cycle .......... 16
A novel enzymatic coupling reaction for the formation of 11-cis RAL in cone
photoreceptors in the carp (Cyprinus carpio) ............................................................................. 18
4.6.
5.
What may be the rationale behind having two vertebrate visual cycles .......................... 19
Future directions ..................................................................................................................... 20
5.1.
Cone visual cycle in higher mammals ............................................................................. 20
5.2.
Transport of 11-cis ROL from Müller glia cells to cone photoreceptors ......................... 22
5.3.
Implication for therapy of retinal disease (rod-cone, cone dystrophies) ......................... 24
2
6. References .................................................................................................................................. 25
1. Introduction
1.1.
Vertebrate duplex retina
Our visual perception of the world relies on the ability to detect differences in luminance,
contrast and color with high temporal and spatial resolution. To achieve this remarkable feat,
light perceiving eyes have evolved from quite simple organs, being capable to detect differences
in luminance (for phototaxis) into highly complex image-forming vertebrate eyes (Lamb et al.,
2007). Photoreceptors, the light-receptive cells in the neural retina have existed throughout
evolution, however the “modern” vertebrate duplex retina, containing both rod and cone
photoreceptors, has only been established by the time of the last common ancestor that humans
share with jawless fish over 500 million years ago (Lamb et al., 2009). Our early ancestors
exclusively relied on photopic (cone photoreceptor-mediated) vision. Scotopic vision, mediated
through rod photoreceptors, has only evolved from cone photoreceptors as a specialization to low
light-level environments later in vertebrate history. The majority of extant vertebrates are teleost
species with most of them living in deep water under mainly scotopic conditions.
Rod and cone photoreceptors are morphologically distinct and show substantial differences in
light sensitivity and overall kinetics (Kawamura and Tachibanaki, 2008; Tachibanaki et al.,
2007). Whereas rod photoreceptors are extremely sensitive to light and are thus used under dim
light (scotopic) conditions, cone photoreceptors allow for high-resolution color vision under
daylight (photopic) conditions, during which rods become saturated. The opsins, mediating
photon capture, also come in rod- (rod opsin) and cone (cone opsins)-specific variants.
3
For continuous vision, two main enzymatic pathways are required in the retina:
phototransduction, the pathway converting light into a neuronal response, and the visual (or
retinoid) cycle, regenerating bleached visual pigment. The vertebrate visual cycle (Wald, 1968)
has been extensively studied, and components as well as enzymatic reactions including their
kinetics are well-understood. It had been assumed that vertebrates utilize the same pathway for
regenerating rod and cone visual pigment (11-cis RAL). Interestingly, rods and cones do partly
utilize different subtypes of enzymes in phototransduction (for example different subunits of the
transducin protein (Larhammar et al., 2009)). Considering this specialization and the apparent
differences in cone versus rod light sensitivities and kinetics, it has been tempting to speculate
that cone and rod photoreceptors might use cell type-specific enzymes/pathways for processes
involving the retinoid cycle as well.
1.2.
Canonical retinoid recycling in vertebrate eyes
The so-called canonical visual cycle (Wald, 1968) is a series of enzymatic reactions resulting in
the regeneration of the visual pigment 11-cis RAL from all-trans RAL, which is the end-product
of phototransduction. Thus, it is fundamental for the restoration of vision after light exposure.
The vertebrate visual cycle involves two cellular compartments - photoreceptor outer segments
and the RPE (retinal pigment epithelium) (Crouch et al., 1996; Lamb and Pugh, 2004; McBee et
al., 2001a; McBee et al., 2001b; Rando, 2001; Saari, 2000) (Figure 1).
Retinoids are highly suited to capture photons, but their chemical properties also make them
prone to degradation potentially leading to toxic products. This may well be the rationale to
separate the regeneration of the visual pigment to another cell type to prevent photoreceptor
toxicity. Even in cephalopods, where pigment regeneration takes place in the photoreceptors,
4
photon capture and pigment regeneration is intracellularly separated (Hara and Hara, 1976;
Molina et al., 1992; Terakita et al., 1989).
In the photoreceptor outer segments absorption of a photon results in the cis to trans conversion
of the Vitamin A derivative 11-cis retinal (RAL). The resulting all-trans RAL is first reduced to
all-trans ROL by an NADPH-dependent dehydrogenase (Cideciyan et al., 2000; Lion et al., 1975;
Rattner et al., 2000; Suzuki et al., 1993), and then transported from the photoreceptor to the
second compartment involved in visual pigment recycling, the RPE. Upon arrival in RPE cells,
all-trans ROL is esterified into retinylesters (REs) by the lecithin:retinol acyltransferase LRAT
(Mondal et al., 2000; Ruiz et al., 1999; Saari and Bredberg, 1989; Saari et al., 1993), and then
isomerized to 11-cis ROL by the isomerase RPE65 (or Isomerase I) (Gollapalli and Rando, 2003;
Jin et al., 2005; Moiseyev et al., 2005; Redmond et al., 2005). Finally, 11-cis ROL is oxidized to
11-cis RAL by a retinol dehydrogenase (Cideciyan et al., 2000; Driessen et al., 1997; Gamble et
al., 2000; Haeseleer et al., 1998; Lion et al., 1975; Simon et al., 1999; Suzuki et al., 1993). The
recycled 11-cis RAL is transported back to photoreceptor outer segments where it re-assembles
with the opsin to form functional visual pigment (Bok, 1985; Pepperberg and Clack, 1984)
(Figure 1).
2. Alternative cone visual cycle in cone-dominant species
In recent years researchers have begun to challenge the concept of the existence of a single (the
canonical) visual cycle mediating pigment regeneration in rod as well as in cone photoreceptors.
In vivo data thus far had mainly stemmed from studies in nocturnal (and consequently roddominant) rodent animal models such as mouse and rat. These studies might have been biased,
5
since the vast majority of photoreceptors in the mouse retina are rods (more than 95%), and
mouse vision primarily relies on rods.
Evidence supporting the existence of a second (cone-specific) retinoid cycle can be found in the
older literature. Goldstein and co-workers had shown in the 60s and 70s that in the isolated frog
neural retina (separated from the RPE) the cone but not the rod photoreceptor early receptor
potential (ERP) recovered after bleaching and subsequent dark adaptation (Goldstein, 1967,
1968; Goldstein and Wolf, 1973; Hood and Hock, 1973). Moreover, it had been shown in the
1980s that in the retina of the cone-dominant chicken the major source of retinylesters are 11-cis
REs localized inside the neural retina, whereas rod-dominant species, such as frogs and cows
show a pre-dominance of all-trans REs in the RPE (Bridges and Alvarez, 1987; Rodriguez and
Tsin, 1989). Bleached salamander rod outer segments are only capable of recovering their visual
sensitivity upon administration of exogenous 11-cis RAL, but not ROL, whereas cones were able
to utilize both retinoid sources (Jones et al., 1989).
Thus, the idea that rod and cone photoreceptors may utilize distinct pathways for regenerating
their visual pigment has been around for a couple of decades. However, only fairly recent mostly
biochemical studies conducted in cone-dominant species such as chicken and ground squirrel
have confirmed this concept (Muniz et al., 2007).
Light exposure of the chicken retina leads to the accumulation of two pools of retinylesters – a
pool of 11-cis REs in the neural retina itself, and a second pool of all-trans REs in the RPE.
Importantly, light exposure was tightly associated with rising levels of 11-cis REs and lowering
levels of 11-cis RAL in the retina and vice versa, reflecting the basic action of a visual cycle
(Trevino et al., 2005; Villazana-Espinoza et al., 2006). Biochemical studies in chicken and
ground squirrel have then provided ground-breaking new insights into how the suspected cone
visual cycle might work. It was Mata and co-workers who demonstrated the existence of three
6
novel enzymatic activities localized to the retina of the cone-dominant species chicken and
ground squirrel (Mata et al., 2002; Mata et al., 2005). These include a retinylester synthase which
is thought to be the equivalent of LRAT in the canonical visual cycle, a retinol dehydrogenase
able to oxidize 11-cis ROL, and last and most importantly, a novel isomerase, termed Isomerase
II. This isomerase converts all-trans ROL into 11-cis ROL in the presence of CRBP and
CRALBP. It is evidently different from RPE65, the isomerase of the canonical visual cycle, as
its substrate is all-trans ROL as opposed to all-trans REs, and it is able to directly create 11-cis
ROL without the formation of REs (Mata et al., 2002) (Figure 2).
However, actual in vivo data are scarce and data presented from studies performed in the mouse,
the most widely used model system for retina research, have remained controversial. The mouse
visual cycle has been mainly assessed in RPE65 knock-out mouse models. All studies have
unequivocally shown that loss of the retinoid isomerase RPE65 leads to impaired rod function.
Initial studies report that cones were unaffected by the absence of RPE65 protein (Redmond et
al., 1998), whereas subsequent studies have argued that both rod and cone photoreceptors were
affected in the knock-out. Thus, cone photoreceptors were thought to rely entirely on the
canonical visual cycle for chromophore regeneration (Fan et al., 2003; Feathers et al., 2008;
Seeliger et al., 2001; Wenzel et al., 2007). This hypothesis has been challenged by the recent
findings of Wang and Kefalov, who provided clear evidence for an intra-retinal visual cycle in
the rod-dominant mouse and salamander, which will be discussed in more detail below (Wang et
al., 2009; Wang and Kefalov, 2009).
Additional in vivo studies, preferably in diurnal cone-dominant animal models, will likely
contribute to a more in-depth understanding of the functional roles of the intra-retinal and the
RPE-localized visual cycles in the vertebrate retina.
7
3. Model system zebrafish
3.1.
Zebrafish as a model organism
The zebrafish has earned its status as one of the most widely used vertebrate animal models
besides the mouse for a number of reasons. In contrast to larger vertebrate models, zebrafish are
easy to maintain in large numbers, and a great amount of offspring can be bred weekly. In vivo
studies in zebrafish embryos and larvae are facilitated by the transparency of the embryos in
combination with their extraordinary rapid embryonic development.
Resources for genetic engineering in zebrafish are vast compared to other cone-dominant models,
such as the chicken or ground squirrel. Gain- as well as loss of function studies can be conducted
by using a variety of transgenic and knock-down as well as novel knock-out approaches. In
addition, the zebrafish genome is fully sequenced and largely annotated.
One peculiarity of the zebrafish genome that it shares with all teleosts, is that frequently two
orthologs of a given mammalian gene are found. Vertebrates have experienced 2 successive
rounds of genome duplication (1R, 2R) during their evolution (Garcia-Fernandez and Holland,
1994; Lundin, 1993; Putnam et al., 2008). In addition to this tetraploidization, the teleost lineage
underwent a 3rd round (R3) of genome duplication (Meyer and Van de Peer, 2005; Taylor et al.,
2003). Whereas usually a large part of duplicated genes are being lost during the course of
evolution, a considerable number of genes will generate either genes with novel functions (neofunctionalization) or genes with specific sub-functions (sub-functionalization) (Force et al.,
1999). Through mutations in regulatory sequences, the expression domains of an ancestral gene
may be split among the two paralogs arising through a duplication event. In such an event both
8
genes perform a kind of job sharing arrangement to fulfill the essential ancestral gene function
and thereby evade mutational elimination by being under selective pressure (Force et al., 1999).
3.2.
Zebrafish visual system
Zebrafish rely on a functional visual system already very early in development. As soon as
embryos hatch from their chorions at around 3dpf (days post-fertilization), free-swimming larvae
need to track food and avoid predators, necessitating a fully developed retina (Gahtan et al.,
2005).
The overall structure and function of the retina is conserved to a high degree across vertebrate
species. Accordingly, the zebrafish retina contains all the cell types found in any other vertebrate
– retinal ganglion cells, bipolar cells and amacrine cells, photoreceptors and Müller glia cells.
These cell types develop in an inside-out progression. After evagination of the optic lobes,
ganglion cells are the first retinal cells to differentiate at around 32hpf (Burrill and Easter, 1995;
Hu and Easter, 1999; Schmitt and Dowling, 1994, 1999). In succession first amacrine and
horizontal cells develop around 50hpf, and the photoreceptor cells are the last cells to
differentiate around 55hpf (Schmitt and Dowling, 1999). Functional synapses only develop after
cell type differentiation, and thus signal transmission from photoreceptors to second-order
neurons gets only fully functional around 5dpf (Biehlmaier et al., 2003). At 5dpf the larval
zebrafish retina has developed synaptic connections between all retinal layers, and robust
behavioral as well as electrophysiological responses (ERGs) can be measured (Baier, 2000;
Branchek, 1984; Brockerhoff et al., 1995; Easter and Nicola, 1996, 1997; Neuhauss, 2003;
Neuhauss et al., 1999).
Like humans, zebrafish have a duplex retina containing two types of photoreceptors – rods and
cones. Zebrafish are a day-active diurnal species and hence predominantly use cone
9
photoreceptors for vision, just like humans. ERG recordings in the larval zebrafish have shown
that rod-driven responses can only be measured at around 15dpf, and thus it is assumed that
zebrafish rod photoreceptors do not significantly contribute to vision before this developmental
stage (Bilotta et al., 2001; Branchek, 1984). As a result, the larval zebrafish retina can be
considered functionally cone-dominant.
4. Cone visual cycle in zebrafish
4.1.
Vision in canonical cycle-deficient zebrafish
The 11-cis ROL isomerase RPE65 is a key enzyme in the vertebrate visual cycle (Redmond et al.,
2005; Redmond et al., 1998). It is highly expressed in the RPE of all vertebrates and is vital for
regenerating the visual pigment 11-cis RAL. A variety of different mutations in RPE65 have been
isolated in human patients presenting with mild-to severe retinal dystrophies (Thompson et al.,
2000). RPE65 and its effect on chromophore recycling has been studied extensively in the mouse.
However due to conflicting data in different studies, the question if an intra-retinal visual cycle
exists in rod-dominant species remained controversial until recently (Fan et al., 2003; Feathers et
al., 2008; Redmond et al., 2005; Redmond et al., 1998; Seeliger et al., 2001; Wang et al., 2009;
Wang and Kefalov, 2009; Wenzel et al., 2007).
In 2007, Schonthaler et al. were the first to tackle this controversy using the cone-dominant
zebrafish (Schonthaler et al., 2007). Two orthologs of mammalian RPE65, which are likely to
have arisen during the 3rd round of genome duplication in teleosts, were identified in the
zebrafish genome. Only one paralog, RPE65a, is expressed in the zebrafish eye (in the RPE),
10
whereas RPE65b is localized to the ventricular zone, the upper and lower jaw and the pectoral
fins during early development.
Schonthaler et al. took a dual approach to create a zebrafish model of RPE65a loss of function.
First a morpholino-mediated knock-down approach was employed to block translation of Rpe65a
protein. This technique reduced the protein level by at least 88%. Since the enzymatic activity of
RPE65 is quite low, necessitating surprisingly high RPE65 protein levels in RPE cells, this
reduction is expected to block any RPE65 dependent pathway. They additionally took advantage
of Ret-NH2 (retinylamine), a potent inhibitor of RPE65 (Golczak et al., 2005), and thus
additionally blocking the canonical visual cycle by pharmacological means.
The impact of severely decreased RPE65a levels onto retinoid metabolism was assessed by
HPLC analysis of retinoids in the larval zebrafish eye. Baseline levels of 11-cis RAL were
decreased in RPE65a MO-injected larvae, as compared to wildtype siblings. This finding could
be explained by immunostainings of rod photoreceptors, which showed deterioration of rod outer
segments. Intense light treatment was used to bleach visual pigment in larvae, and pigment
regeneration was assessed after a subsequent period of dark adaptation. Surprisingly, RPE65adeficient larvae were able to regenerate the full amount of 11-cis RAL (Figure 3A).
To exclude that the regenerated 11-cis RAL did not stem from the remaining low levels of
RPE65a due to only partial knock-down, Schonthaler et al. blocked the canonical visual cycle
using the RPE65 inhibitor Ret-NH2 (Golczak et al., 2005). Initial studies in cell culture revealed
that Ret-NH2 directly blocks RPE65 enzymatic activity, as cells incubated with the compound as
well as all-trans ROL were unable to synthesize 11-cis ROL in contrast to untreated cells.
Bathing zebrafish larvae in a Ret-NH2 solution and subsequent bleaching and re-dark adaptation
resulted in the same level of 11-cis RAL reduction as had been measured in MO-injected larvae.
Moreover, combining the MO and Ret-NH2 did not further exacerbate the phenotype (Figure3A).
11
As the zebrafish visual system is fully functional by 5dpf, MO-treated larvae could be assessed
for visually-mediated behavior. The OKR (optokinetic response) assays tracking eye movements
of zebrafish larvae in response to a moving stimulus, and is a highly robust behavioral test for
visual system function (Huang and Neuhauss, 2008; Neuhauss, 2003; Rinner et al., 2005b).
Consistent with the previous result, RPE65a deficiency did not result in decreased contrast
sensitivity/visual resolution, as tested by the OKR.
Whereas the majority of researchers have found that in the RPE65 knock-out mouse, both rod
and cone photoreceptors were severely impaired (Fan et al., 2003; Feathers et al., 2008; Seeliger
et al., 2001; Wenzel et al., 2007), Redmond et al. had shown that although rod-driven ERG
responses were abolished, cone function still remained intact. This finding led them to speculate
that cones might use a different pathway for regenerating their pigment. Thus the question if an
intra-retinal visual cycle existed in rod-dominant species remained unanswered until recently.
Kefalov and colleagues provided direct in vitro as well as in vivo evidence for the existence of an
intra-retinal visual cycle in the rod dominant salamander and mouse, as well as in primates
including humans (Wang et al., 2009; Wang and Kefalov, 2009). As to the question why previous
studies had concluded that the RPE is the only source of 11-cis RAL in the mouse retina, Wang et
al. speculate that 11-cis ROL generated in Müller cells is only supplied to a tiny fraction of
mouse photoreceptors. In the mouse retina only about 3% of all photoreceptors are cones. Thus,
increasing the mouse cone population by the nrl-/- knock-out would not increase the
chromophore recycling capacity. Moreover, RPE65 might be expressed in cone photoreceptors
and a knock-out would thus directly impact the intra-retinal visual (Wang et al., 2009; Wenzel et
al., 2007; Znoiko et al., 2005).
4.2.
Müller glia cells are implicated in the intra-retinal visual cycle
12
RPE65 is highly expressed in the RPE of all vertebrates, where the canonical visual cycle is
located. The second, intra-retinal visual cycle is independent of the RPE, and thus some of the
enzymatic steps for conversion of all-trans ROL to 11-cis RAL must be localized to a
compartment different than the RPE. It had long been assumed that this cellular compartment
might be the Müller glia cells of the retina. Theses specialized astrocytes perform an impressive
number of different functions in the retina, including structural and metabolic support of cells,
ion buffering, and neurotransmitter homeostasis (Bringmann et al., 2006), but also might have yet
undiscovered additional roles. Das et al. have shown that cultured Müller cells can isomerize and
esterify retinoids in cell culture (Das et al., 1992). It has been shown later that these cells possess
an 11-cis ARAT activity, which catalyzes the synthesis of 11-cis REs from 11-cis ROL and
whose activity is enhanced in the presence of CRALBP (Muniz et al., 2006). Very recently Wang
et al. have provided direct in vivo evidence that Müller glia cells are essential for regeneration of
visual pigment for cone photoreceptors in the salamander and the mouse (Wang et al., 2009;
Wang and Kefalov, 2009). They showed that if physical contact between Müller cells and cone
photoreceptors is disturbed (by physical separation using a suction electrode) or when the
gliotoxin L-alpha-AAA is administered to the retina, cone chromophore cannot be regenerated
(Wang et al., 2009; Wang and Kefalov, 2009).
Müller glia cells have also been shown to express components involved in the canonical visual
cycle, such as RGR (retinal G-protein coupled receptor) (Chen et al., 1996; Shen et al., 1994),
and the retinoid binding proteins CRBP (cellular retinol binding protein) and CRALBP (cellular
retinalaldehyde binding protein) (Bunt-Milam and Saari, 1983; Eisenfeld et al., 1985; Fleisch et
al., 2008).
13
4.3.
Subfunctionalization of zebrafish Cralbp provides evidence for the existence of two
pathways for visual pigment recycling in the zebrafish
In zebrafish Cone visual pigment regeneration seems to be independent of RPE65, the retinoid
isomerase of the canonical visual cycle (Schonthaler et al., 2007). In addition, data from other
cone-dominant species suggested the existence of a second, cone-specific pathway for 11-cis
RAL regeneration.
Retinoid binding proteins such as CRALBP have a variety of functions in the vertebrate visual
cycle. Mutations in human CRALBP lead to different types of photoreceptor dystrophies, such as
Retinitis Pigmentosa and Bothnia dystrophy (Burstedt et al., 1999; Maw et al., 1997; Morimura et
al., 1999), OMIM #180090. CRALBP function has been investigated in a mouse knock-out
model (Saari et al., 2001). Lack of Cralbp protein in these animals led to delayed rhodopsin
regeneration and dark adaptation after light exposure. Whereas levels of 11-cis RAL decreased,
all-trans REs increased, suggesting that Cralbp plays a role in the isomerization step of the
canonical visual cycle (Saari et al., 2001).
Intriguingly Cralbp is not only expressed in the RPE, where it is involved in the canonical cycle,
but also in Müller glia cells (Saari and Crabb, 2005).
Whereas rodents (and humans) possess only one gene encoding CRALBP, zebrafish possess two
variants of the mammalian CRALBP gene – cralbp a and cralbp b (Collery et al., 2008; Fleisch
et al., 2008). Notably, these two paralogs are expressed in the two compartments believed to be
involved in chromophore regeneration – cralbp a in the zebrafish RPE, and cralbp b in Müller
glia cells, in both the larval and adult retina (Collery et al., 2008; Fleisch et al., 2008). This
specific expression pattern suggests a sub-functionalization of the two cralbp genes which had
arisen during teleost evolution and allowed us to separately assess the role of each paralog in
retinoid recycling in the zebrafish ( Fleisch et al., 2008). For this purpose morpholino (MO)
14
oligonuleotides directed against each paralog were designed and injected into larval zebrafish.
HPLC analysis of retinoids was used to determine the implication of loss of cralbp a and/or
cralbp b on retinoid recycling. Control and MO-injected larvae were dark adapted, bleached by
high-intensity light and finally again dark adapted. In line with the findings published by
Schonthaler et al., knock-down of the RPE-expressed Cralbp a, led to decreased levels of 11-cis
RAL (Figure 3B). Knocking-down Müller cell Cralbp b resulted in almost identically reduced
(about 50%) 11-cis RAL levels in the zebrafish retina. Thus, both zebrafish CRALBP orthologs
are clearly implicated in visual pigment regeneration in the zebrafish larva. To further prove this
point, a combination of both morpholinos was administered. Measurements of 11-cis RAL levels
showed a synergistic effect, as loss of both Cralp proteins led to a reduction in retinal 11-cis RAL
to about 25% of control levels. Residual 11-cis RAL in the retina could be due to either residual
Cralbp protein expression caused by incomplete morpholino penetrance and/ or to incomplete
bleaching of the visual pigment by the light treatment applied.
To exclude that photoreceptor degeneration (as opposed to defective chromophore recycling)
caused the observed phenotype, we conducted standard histology as well as immunostainings
using rod (1D1) and cone (zpr1)-specific antibodies and did not detect any signs of photoreceptor
outer segment degeneration (Fleisch et al., 2008). Collery et al. additionally used an antibody
against cone blue-opsin to test for cone integrity after Cralbp b knock-down. In concordance with
our results, they did not detect any abnormalities in cone morphology or opsin expression
(Collery et al., 2008; Fleisch et al., 2008). Our data in zebrafish mirror the data published in the
Cralbp knock-out mouse model. Saari et al. showed that lack of Cralbp resulted in delayed dark
adaptation as well as impaired 11-cis RAL regeneration, which could be attributed to a defective
visual cycle, as photoreceptors did not show any sign of degeneration (Saari et al., 2001).
15
4.4.
Cone photoreceptors use both, the canonical and the intra-retinal visual cycle
Visually-mediated behavior (OKR assay) and ERG recordings were used to assess the effects of
reduced 11-cis RAL levels on visual performance in cralbp MO-injected larvae (Fleisch et al.,
2008). Studies have shown that rod photoreceptors do not contribute to larval vision until 15dpf
(Bilotta et al., 2001; Branchek, 1984). Thus, the data obtained from these experiments
exclusively reflect cone photoreceptor output.
ERG responses were measured in both single (cralbp a or cralbp b) MO- as well as double MOinjected larvae, and the ERG b-wave amplitude was used as a measure for outer retina function
(Rinner et al., 2005a) (Figure 4A). Injection of either cralp a or b MO induced a highly
significant reduction in the ERG b-wave amplitude, and co-injection further decreased the
measured response to levels lower than 25% of controls (Figure 4B). Remarkably, Cralbp b
deficiency led to a more pronounced effect on ERG b-wave amplitude than Cralbp a deficiency,
suggesting that cone photoreceptors are more reliant on the intra-retinal (Müller cell) pathway for
chromophore regeneration. One could speculate that this difference might be due to differences in
remaining Cralbp protein levels. Due to the lack of a Cralbp antibody reliably recognizing
zebrafish Cralbp in Western Blots on larval tissue, residual Cralbp a and b protein levels could
not be measured. We have provided a partial answer by measuring cralbp a and b mRNA levels
after MO injection. RT PCR experiments showed that cralbp a mRNA was reduced to 22% of
control levels, whereas cralbp b mRNA was reduced to 29% of control levels (Fleisch et al.,
2008). This suggests that the cralbp a and b knock-downs were of comparable efficiency. Thus,
the more severe decline in cone function in the cralbp b knock-down indicates that the intraretinal visual cycle is more relevant for zebrafish cone photoreceptors than the canonical visual
cycle in the RPE.
16
In addition to recordings of retinal currents, the visually-mediated OKR (optokinetic response)
assay was used to assess visual system performance in cralbp-deficient zebrafish larvae. Collery
et al. reported that knock-down of both zebrafish Cralbp proteins resulted in a significant
reduction in the number of saccades in the OKR (Collery et al., 2008). In contrast to that, we
found that visual performance, measured as a contrast sensitivity function, was significantly
impaired in zebrafish larvae injected with the Müller cell-specific cralbp b MO, but cralbp a
MO-injected fish did not differ from wildtype controls (Fleisch et al., 2008) (Figure 5).
Interestingly, Schonthaler et al. had similarly not observed an impaired OKR response in rpe65a
MO-injected larvae, despite severely reduced 11-cis RAL levels as assessed by HPLC analysis
(Schonthaler et al., 2007) (Figure 3B). These data support our hypothesis that the Müller cellspecific pathway mediated by cralbp b is more important for cone photoreceptor function.
In conclusion, our data and the data of Collery et al. suggest that two pathways for regenerating
visual pigment exist in the zebrafish retina – one located to the RPE and one located to Müller
glia cells. Cralbp seems to be indispensable for efficient 11-cis RAL production in both
pathways. The cralbp b (Müller cell) dependent pathway is not reliant on rpe65a (Schonthaler et
al., 2007), and thus most likely utilizes a different isomerase activity (similar to the Isomerase II
catalytic activity identified in chicken/ground squirrel by Mata et al. (Mata et al., 2005)).
Importantly, we have provided evidence that zebrafish cone photoreceptors utilize both sources
of regenerated pigment. Knocking-down either cralbp paralog did not result in complete absence
of 11-cis RAL, as cone photoreceptors are capable of utilizing the 2nd source of recycled visual
pigment (from the RPE or Müller cells). It has been shown that cone photoreceptors possess a 11cis ROL dehydrogenase activity, capable of oxidizing 11-cis ROL into 11-cis RAL (Jones et al.,
17
1989; Mata et al., 2002), and are able to regenerate visual pigment from 11-cis ROL or 11-cis
RAL (Jones et al., 1989).
Chromophore regeneration in mouse cone photoreceptors has been shown to be significantly
delayed in the absence of Cralbp (Saari et al., 2001). This can be readily explained by the fact
that the Cralbp protein is expressed in the RPE as well as Müller glia cells of the mouse retina
(Saari and Crabb, 2005). Thus, loss of Cralbp function will affect both, the canonical as well as
the intra-retinal visual cycle.
4.5.
A novel enzymatic coupling reaction for the formation of 11-cis RAL in cone
photoreceptors in the carp (Cyprinus carpio)
Mata et al. had identified three novel catalytic activities implicated in the intra-retinal visual cycle
in the chicken and ground squirrel retina (Mata et al., 2002; Mata et al., 2005) - a retinylester
synthase, a retinol dehydrogenase and a novel isomerase. The molecular identity of these
enzymes has not been elucidated yet.
Cone but not rod photoreceptors seem capable of oxidizing 11-cis ROL to 11-cis RAL (Jones et
al., 1989). Such an 11-cis retinol dehydrogenase (RDH) activity preferring NADP+ as a cofactor
had then been isolated in chicken and ground squirrel microsomes, and suggested to be
implicated in this cone-specific oxidation step (Mata et al., 2002; Mata et al., 2005).
Miyazono et al. investigated the metabolism of all-trans RAL and 11-cis RAL in purified carp
(Cyprinus carpio) rod and cone photoreceptors. In contrast to the data from chicken and ground
squirrel, the oxidation of 11-cis ROL into 11-cis RAL by RDHs was not effective. The authors
speculated that the co-factor might be all-trans RAL rather than NADP+ (Miyazono et al., 2008).
To test their hypothesis, they added all-trans RAL to the carp cone lysate. Addition of 11-cis
18
ROL plus all-trans RAL, but not NADP+, resulted in formation of 11-cis RAL as well as alltrans ROL. Miyazono et al. concluded that this ALOL coupling reaction, which they detected
exclusively in carp cone but not rod photoreceptors, might be the enzymatic activity underlying
11-cis ROL oxidation in cone photoreceptors. Heat treatment of cone lysates abolished the ALOL
reaction, suggesting that it is mediated by an enzyme (complex).
The identity of this enzyme and the molecular mechanisms underlying the ALOL coupling
reaction are currently still unknown. Moreover it remains unclear why different retinol
dehydrogenase enzymatic activities have been identified in the carp versus the chicken and
ground squirrel retina.
4.6.
What may be the rationale behind having two vertebrate visual cycles?
The discussed recent data indicate that the canonical visual cycle can regenerate both rod and
cone pigments. This begs the question why two separate visual cycles have emerged during the
course of evolution. One reason may be found in the high demand of regenerating capacity of the
more numerous rods under photopic conditions.
This competition is aggravated by the rods being driven into saturation under photopic conditions
with a maximal demand for pigment regeneration. As rod and cone photoreceptors compete for
11-cis RAL, the high demand for visual pigment by rod photoreceptors under photopic conditions
might therefore restrict the availability of chromophore to cone photoreceptors in their working
range (Mata et al., 2002; Muniz et al., 2007). Samardzija et al. created an RPE65 knock-in mouse
model carrying the R91W mutation to investigate the consequences of chromophore insufficiency
on cone function (Samardzija et al., 2009). Their data revealed that under conditions of limited
11-cis RAL supply, rod and cone opsins compete for the remaining visual pigment. In their
19
mouse model, rod photoreceptors out-competed cone photoreceptors, which ultimately led to
cone degeneration.
On the other hand, one could speculate that the intra-retinal visual cycle might be evolutionary
older than the canonical visual cycle localized to the RPE. Rod photoreceptors have evolved from
cone photoreceptors during vertebrate evolution, and thus a chromophore recycling pathway
supplying 11-cis RAL to cone photoreceptors must have existed before the emergence of rod
photoreceptors. Most recent data from zebrafish, salamander, mouse and primates have shown
that cones most likely are able to utilize both visual cycles for pigment regeneration (Fleisch et
al., 2008; Wang et al., 2009; Wang and Kefalov, 2009). Therefore it is likely that the Müller cellbased pathway has evolved at a later step in evolution, at the time when rod photoreceptors
started to populate the vertebrate retina. This led to reduced availability of chromophore for cone
photoreceptors, as rods are saturated under photopic conditions, but still consume 11-cis RAL.
Consequently, cone photoreceptors needed another additional source of chromophore supply.
5. Future directions
5.1.
Cone visual cycle in primates
Until very recently it was claimed by the majority of researchers working with rodent models that
rod-dominant species entirely rely on the canonical visual cycle for chromophore recycling (Fan
et al., 2003; Feathers et al., 2008; Seeliger et al., 2001; Wenzel et al., 2007). The data arguing in
favor of a second, cone-specific visual cycle, acquired in the cone-dominant models chicken,
ground squirrel and zebrafish had been acknowledged, but it had been speculated that this
discrepancy might be due to species differences. Primates, including humans, highly rely on cone
20
photoreceptors for daylight vision. However rod photoreceptors outnumber cones, which are
concentrated in the fovea of the retina, by far and in consequence primates and humans could be
considered rod-dominant (with the exception of the foveal region). Hence the existence of an
intra-retinal visual cycle in humans has been questioned.
The studies of Wang et al. are an important contribution towards settling the question if all
vertebrates have two pathways for 11-cis RAL recycling (Wang et al., 2009; Wang and Kefalov,
2009). In their first study they investigated if the rod-dominant salamander and mouse retina
promotes cone visual pigment regeneration (Wang et al., 2009). Microspectrophotometry as well
as electrophysiological recordings (single-cell and ERG) revealed that visual pigment is
regenerated in the salamander retina independent of the RPE (and thus the canonical visual
cycle). Similarly, ERG recordings in the isolated mouse retina indicated substantial pigment
regeneration.
In a subsequent study they expanded their study to include primates, including humans (Wang
and Kefalov, 2009). Eyecups were isolated from the RPE, and recovery of ERG responses
measured after bleaching of photoreceptors. The primate, including the human retina showed
robust dark adaptation in cones but not rods. This cone response recovery could be blocked by
the gliotoxin L-alpha-AAA and could subsequently be reversed by addition of exogenous 11-cis
ROL.
In summary, these data convincingly demonstrate that rod- as well as cone-dominant species
possess an intra-retinal visual cycle, which acts to regenerate 11-cis RAL for cone
photoreceptors, independent of the RPE. Thus, this Müller glia cell-dependent pathway is
evolutionally conserved among vertebrates.
21
5.2. Transport of 11-cis ROL from Müller glia cells to cone photoreceptors
Both, the canonical as well as the proposed intra-retinal visual cycle, are each localized to two
different retinal compartments. Retinoids are largely water insoluble and potentially toxic to the
cell, and thus need to be in complex with a chaperone protein for transportation between different
cellular compartments. IRBP (interphotoreceptor retinoid-binding protein) is localized to the subretinal space (Loew and Gonzalez-Fernandez, 2002) and is known to be able to bind retinoids
(Shaw and Noy, 2001). It therefore has been hypothesized to play a role in the transport of 11-cis
RAL from the RPE to rod outer segments (Bunt-Milam and Saari, 1983; Fong et al., 1984; Lamb
and Pugh, 2006; Pepperberg et al., 1993). Surprisingly, no evidence of an impairment of the
canonical visual cycle has been found in irbp-/- mice (Palczewski et al., 1999; Ripps, 2001).
This obvious discrepancy has been revisited by two recent studies (Jin et al., 2009; Parker et al.,
2009). Both studies suggest that the lack of a detectable visual cycle phenotype in those mice
could have been due to the genetic background of the mouse strain used in the original studies
(which had been uncontrolled for an rpe65 mutation). They document a clear reduction in cone
function in the novel irbp-/- model, however differ greatly in a number of experimental results,
and present two quite contrary hypothesis of IRBP function. Whereas Jin et al. demonstrate rod
and cone photoreceptor degeneration, and postulate that the severe cone outer segment
deterioration is due to opsin mistrafficking caused by 11-cis RAL deficiency, Parker et al. show
normal cone opsin localization and no photoreceptor degeneration. Recovery of 11-cis RAL after
photobleach was delayed in irbp-/- mice, and accumulation of 11-cis RAL in the RPE and alltrans ROL in the retina was observed (Jin et al., 2009). These data suggest that IRBP has a role in
retinoid trafficking in the RPE. Parker et al. specifically examined the impact of IRBP loss on
cone photoreceptor function. ERG recordings in irbp-/- mice revealed decreased cone responses,
22
which could be rescued by the administration of exogenous 11-cis ROL (Parker et al., 2009).
Consequently, IRBP is likely to be implicated supplying regenerated chromophore to mouse
photoreceptors.
The two datasets of Jin and Parker et al. conflict in one major point – whereas Jin et al. argue that
degeneration of photoreceptor outer segments underlies reduced cone function, Parker et al. state
that it is caused by a deficiency in retinoid recycling.
The function of zebrafish Irbp has not been addressed yet. It is noteworthy that zebrafish possess
two IRBP orthologs, – irbp, and irbp-like that are arranged head-to-tail in the zebrafish genome
(Nickerson et al., 2006). It has been shown that irbp is expressed in the zebrafish RPE as well as
in cone and rod photoreceptors (Stenkamp et al., 1998). Whereas Nickerson et al. confirmed that
Irbp is expressed in the RPE and photoreceptors, Irbp-like was found to be expressed in the inner
nuclear layer of the zebrafish retina (Nickerson et al., 2006). This specific expression pattern
makes it tempting to speculate that zebrafish Irbp might be involved in transporting 11-cis RAL
from the RPE to rod photoreceptors, whereas Irbp-like is localized perfectly to transport 11-cis
ROL from Müller glia cells to cone photoreceptors.
Müller cell processes do not reach photoreceptor outer segments, but do surround the inner
segment (Sarantis and Mobbs, 1992). Thus, in the intra-retinal visual cycle 11-cis ROL must be
transported from Müller glia cells to the cell bodies of cone photoreceptors. It has been shown
that chromophore can diffuse from the inner to the outer segment in cone but not rod
photoreceptors (Jin et al., 1994). Thus, if IRBP provides a retinoid transport mechanism between
the Müller glia cell to the cone photoreceptor cell body, 11-cis RAL could then readily diffuse to
the outer segment to recombine with opsin to form functional pigment.
23
5.3.
Implication for therapy of retinal disease (rod-cone, cone dystrophies)
Human patients suffering from photoreceptor dystrophies exhibit mutations in a great number of
genes involved in the canonical visual cycle (RPE65, RDHs, LRAT, CRALBP; for a review see
(Travis et al., 2007)) Hence, it is very likely that mutations in the genes encoding for the yet
unidentified enzymatic components of the Muller glia cell-dependent pathway, are implicated in
human retinal disease. Identification of these genes should be of prior importance to the retina
research community. Moreover, the molecular mechanisms of the intra-retinal visual cycle need
to be understood in order to develop appropriate therapeutics.
It has been shown that cone photoreceptors are affected by defective canonical as well was intraretinal chromophore regeneration. It is still not clear to which extent cone photoreceptors utilize
each respective pathway. Thus mutations in genes that affects specifically one or the other
pathway, might affect cone vision differently.
24
6. References
Baier, H. (2000). Zebrafish on the move: towards a behavior-genetic analysis of vertebrate vision.
Curr Opin Neurobiol 10, 451-455.
Biehlmaier, O., Neuhauss, S.C., and Kohler, K. (2003). Synaptic plasticity and functionality at
the cone terminal of the developing zebrafish retina. J Neurobiol 56, 222-236.
Bilotta, J., Saszik, S., and Sutherland, S.E. (2001). Rod contributions to the electroretinogram of
the dark-adapted developing zebrafish. Dev Dyn 222, 564-570.
Bok, D. (1985). Retinal photoreceptor-pigment epithelium interactions. Friedenwald lecture.
Invest Ophthalmol Vis Sci 26, 1659-1694.
Branchek, T. (1984). The development of photoreceptors in the zebrafish, brachydanio rerio. II.
Function. J Comp Neurol 224, 116-122.
Bridges, C.D., and Alvarez, R.A. (1987). The visual cycle operates via an isomerase acting on
all-trans retinol in the pigment epithelium. Science 236, 1678-1680.
Bringmann, A., Pannicke, T., Grosche, J., Francke, M., Wiedemann, P., Skatchkov, S.N.,
Osborne, N.N., and Reichenbach, A. (2006). Muller cells in the healthy and diseased retina. Prog
Retin Eye Res 25, 397-424.
Brockerhoff, S.E., Hurley, J.B., Janssen-Bienhold, U., Neuhauss, S.C., Driever, W., and
Dowling, J.E. (1995). A behavioral screen for isolating zebrafish mutants with visual system
defects. Proc Natl Acad Sci U S A 92, 10545-10549.
Bunt-Milam, A.H., and Saari, J.C. (1983). Immunocytochemical localization of two retinoidbinding proteins in vertebrate retina. J Cell Biol 97, 703-712.
Burrill, J.D., and Easter, S.S., Jr. (1995). The first retinal axons and their microenvironment in
zebrafish: cryptic pioneers and the pretract. J Neurosci 15, 2935-2947.
25
Burstedt, M.S., Sandgren, O., Holmgren, G., and Forsman-Semb, K. (1999). Bothnia dystrophy
caused by mutations in the cellular retinaldehyde-binding protein gene (RLBP1) on chromosome
15q26. Invest Ophthalmol Vis Sci 40, 995-1000.
Bustamante, J.J., Ziari, S., Ramirez, R.D., and Tsin, A.T. (1995). Retinyl ester hydrolase and the
visual cycle in the chicken eye. Am J Physiol 269, R1346-1350.
Chen, X.N., Korenberg, J.R., Jiang, M., Shen, D., and Fong, H.K. (1996). Localization of the
human RGR opsin gene to chromosome 10q23. Hum Genet 97, 720-722.
Cideciyan, A.V., Haeseleer, F., Fariss, R.N., Aleman, T.S., Jang, G.F., Verlinde, C.L., Marmor,
M.F., Jacobson, S.G., and Palczewski, K. (2000). Rod and cone visual cycle consequences of a
null mutation in the 11-cis-retinol dehydrogenase gene in man. Vis Neurosci 17, 667-678.
Collery, R., McLoughlin, S., Vendrell, V., Finnegan, J., Crabb, J.W., Saari, J.C., and Kennedy,
B.N. (2008). Duplication and divergence of zebrafish CRALBP genes uncovers novel role for
RPE- and Muller-CRALBP in cone vision. Invest Ophthalmol Vis Sci 49, 3812-3820.
Crouch, R.K., Chader, G.J., Wiggert, B., and Pepperberg, D.R. (1996). Retinoids and the visual
process. Photochem Photobiol 64, 613-621.
Das, S.R., Bhardwaj, N., Kjeldbye, H., and Gouras, P. (1992). Muller cells of chicken retina
synthesize 11-cis-retinol. Biochem J 285 ( Pt 3), 907-913.
Driessen, C.A., Winkens, H.J., Kuhlmann, L.D., Janssen, B.P., van Vugt, A.H., Deutman, A.F.,
and Janssen, J.J. (1997). Cloning and structural analysis of the murine GCN5L1 gene. Gene 203,
27-31.
Easter, S.S., Jr., and Nicola, G.N. (1996). The development of vision in the zebrafish (Danio
rerio). Dev Biol 180, 646-663.
Easter, S.S., Jr., and Nicola, G.N. (1997). The development of eye movements in the zebrafish
(Danio rerio). Dev Psychobiol 31, 267-276.
26
Eisenfeld, A.J., Bunt-Milam, A.H., and Saari, J.C. (1985). Immunocytochemical localization of
retinoid-binding proteins in developing normal and RCS rats. Prog Clin Biol Res 190, 231-240.
Fan, J., Rohrer, B., Moiseyev, G., Ma, J.X., and Crouch, R.K. (2003). Isorhodopsin rather than
rhodopsin mediates rod function in RPE65 knock-out mice. Proc Natl Acad Sci U S A 100,
13662-13667.
Feathers, K.L., Lyubarsky, A.L., Khan, N.W., Teofilo, K., Swaroop, A., Williams, D.S., Pugh,
E.N., Jr., and Thompson, D.A. (2008). Nrl-knockout mice deficient in Rpe65 fail to synthesize
11-cis retinal and cone outer segments. Invest Ophthalmol Vis Sci 49, 1126-1135.
Fleisch, V.C., Schonthaler, H.B., von Lintig, J., and Neuhauss, S.C. (2008). Subfunctionalization
of a retinoid-binding protein provides evidence for two parallel visual cycles in the conedominant zebrafish retina. J Neurosci 28, 8208-8216.
Fong, S.L., Liou, G.I., Landers, R.A., Alvarez, R.A., and Bridges, C.D. (1984). Purification and
characterization of a retinol-binding glycoprotein synthesized and secreted by bovine neural
retina. J Biol Chem 259, 6534-6542.
Force, A., Lynch, M., Pickett, F.B., Amores, A., Yan, Y.L., and Postlethwait, J. (1999).
Preservation of duplicate genes by complementary, degenerative mutations. Genetics 151, 15311545.
Gahtan, E., Tanger, P., and Baier, H. (2005). Visual prey capture in larval zebrafish is controlled
by identified reticulospinal neurons downstream of the tectum. J Neurosci 25, 9294-9303.
Gamble, M.V., Mata, N.L., Tsin, A.T., Mertz, J.R., and Blaner, W.S. (2000). Substrate
specificities and 13-cis-retinoic acid inhibition of human, mouse and bovine cis-retinol
dehydrogenases. Biochim Biophys Acta 1476, 3-8.
Garcia-Fernandez, J., and Holland, P.W. (1994). Archetypal organization of the amphioxus Hox
gene cluster. Nature 370, 563-566.
27
Golczak, M., Imanishi, Y., Kuksa, V., Maeda, T., Kubota, R., and Palczewski, K. (2005).
Lecithin:retinol acyltransferase is responsible for amidation of retinylamine, a potent inhibitor of
the retinoid cycle. J Biol Chem 280, 42263-42273.
Goldstein, E.B. (1967). Early receptor potential of the isolated frog (Rana pipiens) retina. Vision
Res 7, 837-845.
Goldstein, E.B. (1968). Visual pigments and the early receptor potential of the isolated frog
retina. Vision Res 8, 953-963.
Goldstein, E.B., and Wolf, B.M. (1973). Regeneration of the green-rod pigment in the isolated
frog retina. Vision Res 13, 527-534.
Gollapalli, D.R., and Rando, R.R. (2003). Molecular logic of 11-cis-retinoid biosynthesis in a
cone-dominated species. Biochemistry 42, 14921-14929.
Haeseleer, F., Huang, J., Lebioda, L., Saari, J.C., and Palczewski, K. (1998). Molecular
characterization of a novel short-chain dehydrogenase/reductase that reduces all-trans-retinal. J
Biol Chem 273, 21790-21799.
Hara, T., and Hara, R. (1976). Distribution of rhodopsin and retinochrome in the squid retina. J
Gen Physiol 67, 791-805.
Hood, D.C., and Hock, P.A. (1973). Recovery of cone receptor activity in the frog's isolated
retina. Vision Res 13, 1943-1951.
Hu, M., and Easter, S.S. (1999). Retinal neurogenesis: the formation of the initial central patch of
postmitotic cells. Dev Biol 207, 309-321.
Huang, Y.Y., and Neuhauss, S.C. (2008). The optokinetic response in zebrafish and its
applications. Front Biosci 13, 1899-1916.
Jin, J., Jones, G.J., and Cornwall, M.C. (1994). Movement of retinal along cone and rod
photoreceptors. Vis Neurosci 11, 389-399.
28
Jin, M., Li, S., Moghrabi, W.N., Sun, H., and Travis, G.H. (2005). Rpe65 is the retinoid
isomerase in bovine retinal pigment epithelium. Cell 122, 449-459.
Jin, M., Li, S., Nusinowitz, S., Lloyd, M., Hu, J., Radu, R.A., Bok, D., and Travis, G.H. (2009).
The role of interphotoreceptor retinoid-binding protein on the translocation of visual retinoids
and function of cone photoreceptors. J Neurosci 29, 1486-1495.
Jones, G.J., Crouch, R.K., Wiggert, B., Cornwall, M.C., and Chader, G.J. (1989). Retinoid
requirements for recovery of sensitivity after visual-pigment bleaching in isolated photoreceptors.
Proc Natl Acad Sci U S A 86, 9606-9610.
Kawamura, S., and Tachibanaki, S. (2008). Rod and cone photoreceptors: molecular basis of the
difference in their physiology. Comp Biochem Physiol A Mol Integr Physiol 150, 369-377.
Lamb, T.D., Arendt, D., and Collin, S.P. (2009). The evolution of phototransduction and eyes.
Philos Trans R Soc Lond B Biol Sci 364, 2791-2793.
Lamb, T.D., Collin, S.P., and Pugh, E.N., Jr. (2007). Evolution of the vertebrate eye: opsins,
photoreceptors, retina and eye cup. Nat Rev Neurosci 8, 960-976.
Lamb, T.D., and Pugh, E.N., Jr. (2004). Dark adaptation and the retinoid cycle of vision. Prog
Retin Eye Res 23, 307-380.
Lamb, T.D., and Pugh, E.N., Jr. (2006). Phototransduction, dark adaptation, and rhodopsin
regeneration the proctor lecture. Invest Ophthalmol Vis Sci 47, 5137-5152.
Larhammar, D., Nordstrom, K., and Larsson, T.A. (2009). Evolution of vertebrate rod and cone
phototransduction genes. Philos Trans R Soc Lond B Biol Sci 364, 2867-2880.
Lion, F., Rotmans, J.P., Daemen, F.J., and Bonting, S.L. (1975). Biochemical aspects of the
visual process. XXVII. Stereospecificity of ocular retinol dehydrogenases and the visual cycle.
Biochim Biophys Acta 384, 283-292.
29
Loew, A., and Gonzalez-Fernandez, F. (2002). Crystal structure of the functional unit of
interphotoreceptor retinoid binding protein. Structure 10, 43-49.
Lundin, L.G. (1993). Evolution of the vertebrate genome as reflected in paralogous chromosomal
regions in man and the house mouse. Genomics 16, 1-19.
Mata, N.L., Radu, R.A., Clemmons, R.C., and Travis, G.H. (2002). Isomerization and oxidation
of vitamin a in cone-dominant retinas: a novel pathway for visual-pigment regeneration in
daylight. Neuron 36, 69-80.
Mata, N.L., Ruiz, A., Radu, R.A., Bui, T.V., and Travis, G.H. (2005). Chicken retinas contain a
retinoid isomerase activity that catalyzes the direct conversion of all-trans-retinol to 11-cisretinol. Biochemistry 44, 11715-11721.
Maw, M.A., Kennedy, B., Knight, A., Bridges, R., Roth, K.E., Mani, E.J., Mukkadan, J.K.,
Nancarrow, D., Crabb, J.W., and Denton, M.J. (1997). Mutation of the gene encoding cellular
retinaldehyde-binding protein in autosomal recessive retinitis pigmentosa. Nat Genet 17, 198200.
McBee, J.K., Palczewski, K., Baehr, W., and Pepperberg, D.R. (2001a). Confronting complexity:
the interlink of phototransduction and retinoid metabolism in the vertebrate retina. Prog Retin
Eye Res 20, 469-529.
McBee, J.K., Van Hooser, J.P., Jang, G.F., and Palczewski, K. (2001b). Isomerization of 11-cisretinoids to all-trans-retinoids in vitro and in vivo. J Biol Chem 276, 48483-48493.
Meyer, A., and Van de Peer, Y. (2005). From 2R to 3R: evidence for a fish-specific genome
duplication (FSGD). Bioessays 27, 937-945.
Miyazono, S., Shimauchi-Matsukawa, Y., Tachibanaki, S., and Kawamura, S. (2008). Highly
efficient retinal metabolism in cones. Proc Natl Acad Sci U S A 105, 16051-16056.
30
Moiseyev, G., Chen, Y., Takahashi, Y., Wu, B.X., and Ma, J.X. (2005). RPE65 is the
isomerohydrolase in the retinoid visual cycle. Proc Natl Acad Sci U S A 102, 12413-12418.
Molina, T.M., Torres, S.C., Flores, A., Hara, T., Hara, R., and Robles, L.J. (1992).
Immunocytochemical localization of retinal binding protein in the octopus retina: a shuttle
protein for 11-cis retinal. Exp Eye Res 54, 83-90.
Mondal, M.S., Ruiz, A., Bok, D., and Rando, R.R. (2000). Lecithin retinol acyltransferase
contains cysteine residues essential for catalysis. Biochemistry 39, 5215-5220.
Morimura, H., Berson, E.L., and Dryja, T.P. (1999). Recessive mutations in the RLBP1 gene
encoding cellular retinaldehyde-binding protein in a form of retinitis punctata albescens. Invest
Ophthalmol Vis Sci 40, 1000-1004.
Muniz, A., Villazana-Espinoza, E.T., Hatch, A.L., Trevino, S.G., Allen, D.M., and Tsin, A.T.
(2007). A novel cone visual cycle in the cone-dominated retina. Exp Eye Res 85, 175-184.
Muniz, A., Villazana-Espinoza, E.T., Thackeray, B., and Tsin, A.T. (2006). 11-cis-AcylCoA:retinol O-acyltransferase activity in the primary culture of chicken Muller cells.
Biochemistry 45, 12265-12273.
Neuhauss, S.C. (2003). Behavioral genetic approaches to visual system development and function
in zebrafish. J Neurobiol 54, 148-160.
Neuhauss, S.C., Biehlmaier, O., Seeliger, M.W., Das, T., Kohler, K., Harris, W.A., and Baier, H.
(1999). Genetic disorders of vision revealed by a behavioral screen of 400 essential loci in
zebrafish. J Neurosci 19, 8603-8615.
Nickerson, J.M., Frey, R.A., Ciavatta, V.T., and Stenkamp, D.L. (2006). Interphotoreceptor
retinoid-binding protein gene structure in tetrapods and teleost fish. Mol Vis 12, 1565-1585.
Palczewski, K., Van Hooser, J.P., Garwin, G.G., Chen, J., Liou, G.I., and Saari, J.C. (1999).
Kinetics of visual pigment regeneration in excised mouse eyes and in mice with a targeted
31
disruption of the gene encoding interphotoreceptor retinoid-binding protein or arrestin.
Biochemistry 38, 12012-12019.
Parker, R.O., Fan, J., Nickerson, J.M., Liou, G.I., and Crouch, R.K. (2009). Normal cone function
requires the interphotoreceptor retinoid binding protein. J Neurosci 29, 4616-4621.
Pepperberg, D.R., and Clack, J.W. (1984). Rhodopsin photoproducts and the visual response of
vertebrate rods. Vision Res 24, 1481-1486.
Pepperberg, D.R., Okajima, T.L., Wiggert, B., Ripps, H., Crouch, R.K., and Chader, G.J. (1993).
Interphotoreceptor retinoid-binding protein (IRBP). Molecular biology and physiological role in
the visual cycle of rhodopsin. Mol Neurobiol 7, 61-85.
Putnam, N.H., Butts, T., Ferrier, D.E., Furlong, R.F., Hellsten, U., Kawashima, T., RobinsonRechavi, M., Shoguchi, E., Terry, A., Yu, J.K., et al. (2008). The amphioxus genome and the
evolution of the chordate karyotype. Nature 453, 1064-1071.
Rando, R.R. (2001). The biochemistry of the visual cycle. Chem Rev 101, 1881-1896.
Rattner, A., Smallwood, P.M., and Nathans, J. (2000). Identification and characterization of alltrans-retinol dehydrogenase from photoreceptor outer segments, the visual cycle enzyme that
reduces all-trans-retinal to all-trans-retinol. J Biol Chem 275, 11034-11043.
Redmond, T.M., Poliakov, E., Yu, S., Tsai, J.Y., Lu, Z., and Gentleman, S. (2005). Mutation of
key residues of RPE65 abolishes its enzymatic role as isomerohydrolase in the visual cycle. Proc
Natl Acad Sci U S A 102, 13658-13663.
Redmond, T.M., Yu, S., Lee, E., Bok, D., Hamasaki, D., Chen, N., Goletz, P., Ma, J.X., Crouch,
R.K., and Pfeifer, K. (1998). Rpe65 is necessary for production of 11-cis-vitamin A in the retinal
visual cycle. Nat Genet 20, 344-351.
32
Rinner, O., Makhankov, Y.V., Biehlmaier, O., and Neuhauss, S.C. (2005a). Knockdown of conespecific kinase GRK7 in larval zebrafish leads to impaired cone response recovery and delayed
dark adaptation. Neuron 47, 231-242.
Rinner, O., Rick, J.M., and Neuhauss, S.C. (2005b). Contrast sensitivity, spatial and temporal
tuning of the larval zebrafish optokinetic response. Invest Ophthalmol Vis Sci 46, 137-142.
Ripps, H. (2001). The rhodopsin cycle: a twist in the tale. Prog Brain Res 131, 335-350.
Rodriguez, K.A., and Tsin, A.T. (1989). Retinyl esters in the vertebrate neuroretina. Am J
Physiol 256, R255-258.
Ruiz, A., Winston, A., Lim, Y.H., Gilbert, B.A., Rando, R.R., and Bok, D. (1999). Molecular and
biochemical characterization of lecithin retinol acyltransferase. J Biol Chem 274, 3834-3841.
Saari, J.C. (2000). Biochemistry of visual pigment regeneration: the Friedenwald lecture. Invest
Ophthalmol Vis Sci 41, 337-348.
Saari, J.C., and Bredberg, D.L. (1989). Lecithin:retinol acyltransferase in retinal pigment
epithelial microsomes. J Biol Chem 264, 8636-8640.
Saari, J.C., Bredberg, D.L., and Farrell, D.F. (1993). Retinol esterification in bovine retinal
pigment epithelium: reversibility of lecithin:retinol acyltransferase. Biochem J 291 ( Pt 3), 697700.
Saari, J.C., and Crabb, J.W. (2005). Focus on molecules: cellular retinaldehyde-binding protein
(CRALBP). Exp Eye Res 81, 245-246.
Saari, J.C., Nawrot, M., Kennedy, B.N., Garwin, G.G., Hurley, J.B., Huang, J., Possin, D.E., and
Crabb, J.W. (2001). Visual cycle impairment in cellular retinaldehyde binding protein (CRALBP)
knockout mice results in delayed dark adaptation. Neuron 29, 739-748.
Samardzija, M., Tanimoto, N., Kostic, C., Beck, S., Oberhauser, V., Joly, S., Thiersch, M., Fahl,
E., Arsenijevic, Y., von Lintig, J., et al. (2009). In conditions of limited chromophore supply rods
33
entrap 11-cis-retinal leading to loss of cone function and cell death. Hum Mol Genet 18, 12661275.
Sarantis, M., and Mobbs, P. (1992). The spatial relationship between Muller cell processes and
the photoreceptor output synapse. Brain Res 584, 299-304.
Schmitt, E.A., and Dowling, J.E. (1994). Early eye morphogenesis in the zebrafish, Brachydanio
rerio. J Comp Neurol 344, 532-542.
Schmitt, E.A., and Dowling, J.E. (1999). Early retinal development in the zebrafish, Danio rerio:
light and electron microscopic analyses. J Comp Neurol 404, 515-536.
Schonthaler, H.B., Lampert, J.M., Isken, A., Rinner, O., Mader, A., Gesemann, M., Oberhauser,
V., Golczak, M., Biehlmaier, O., Palczewski, K., et al. (2007). Evidence for RPE65-independent
vision in the cone-dominated zebrafish retina. Eur J Neurosci 26, 1940-1949.
Seeliger, M.W., Grimm, C., Stahlberg, F., Friedburg, C., Jaissle, G., Zrenner, E., Guo, H., Reme,
C.E., Humphries, P., Hofmann, F., et al. (2001). New views on RPE65 deficiency: the rod system
is the source of vision in a mouse model of Leber congenital amaurosis. Nat Genet 29, 70-74.
Shaw, N.S., and Noy, N. (2001). Interphotoreceptor retinoid-binding protein contains three
retinoid binding sites. Exp Eye Res 72, 183-190.
Shen, D., Jiang, M., Hao, W., Tao, L., Salazar, M., and Fong, H.K. (1994). A human opsinrelated gene that encodes a retinaldehyde-binding protein. Biochemistry 33, 13117-13125.
Simon, A., Romert, A., Gustafson, A.L., McCaffery, J.M., and Eriksson, U. (1999). Intracellular
localization and membrane topology of 11-cis retinol dehydrogenase in the retinal pigment
epithelium suggest a compartmentalized synthesis of 11-cis retinaldehyde. J Cell Sci 112 ( Pt 4),
549-558.
34
Stenkamp, D.L., Cunningham, L.L., Raymond, P.A., and Gonzalez-Fernandez, F. (1998). Novel
expression pattern of interphotoreceptor retinoid-binding protein (IRBP) in the adult and
developing zebrafish retina and RPE. Mol Vis 4, 26.
Suzuki, Y., Ishiguro, S., and Tamai, M. (1993). Identification and immunohistochemistry of
retinol dehydrogenase from bovine retinal pigment epithelium. Biochim Biophys Acta 1163, 201208.
Tachibanaki, S., Shimauchi-Matsukawa, Y., Arinobu, D., and Kawamura, S. (2007). Molecular
mechanisms characterizing cone photoresponses. Photochem Photobiol 83, 19-26.
Taylor, J.S., Braasch, I., Frickey, T., Meyer, A., and Van de Peer, Y. (2003). Genome
duplication, a trait shared by 22000 species of ray-finned fish. Genome Res 13, 382-390.
Terakita, A., Hara, R., and Hara, T. (1989). Retinal-binding protein as a shuttle for retinal in the
rhodopsin-retinochrome system of the squid visual cells. Vision Res 29, 639-652.
Thompson, D.A., Gyurus, P., Fleischer, L.L., Bingham, E.L., McHenry, C.L., Apfelstedt-Sylla,
E., Zrenner, E., Lorenz, B., Richards, J.E., Jacobson, S.G., et al. (2000). Genetics and phenotypes
of RPE65 mutations in inherited retinal degeneration. Invest Ophthalmol Vis Sci 41, 4293-4299.
Travis, G.H., Golczak, M., Moise, A.R., and Palczewski, K. (2007). Diseases caused by defects
in the visual cycle: retinoids as potential therapeutic agents. Annu Rev Pharmacol Toxicol 47,
469-512.
Trevino, S.G., Villazana-Espinoza, E.T., Muniz, A., and Tsin, A.T. (2005). Retinoid cycles in the
cone-dominated chicken retina. J Exp Biol 208, 4151-4157.
Villazana-Espinoza, E.T., Hatch, A.L., and Tsin, A.T. (2006). Effect of light exposure on the
accumulation and depletion of retinyl ester in the chicken retina. Exp Eye Res 83, 871-876.
Wald, G. (1968). The molecular basis of visual excitation. Nature 219, 800-807.
35
Wang, J.S., Estevez, M.E., Cornwall, M.C., and Kefalov, V.J. (2009). Intra-retinal visual cycle
required for rapid and complete cone dark adaptation. Nat Neurosci 12, 295-302.
Wang, J.S., and Kefalov, V.J. (2009). An alternative pathway mediates the mouse and human
cone visual cycle. Curr Biol 19, 1665-1669.
Wenzel, A., von Lintig, J., Oberhauser, V., Tanimoto, N., Grimm, C., and Seeliger, M.W. (2007).
RPE65 is essential for the function of cone photoreceptors in NRL-deficient mice. Invest
Ophthalmol Vis Sci 48, 534-542.
Znoiko, S.L., Rohrer, B., Lu, K., Lohr, H.R., Crouch, R.K., and Ma, J.X. (2005). Downregulation
of cone-specific gene expression and degeneration of cone photoreceptors in the Rpe65-/- mouse
at early ages. Invest Ophthalmol Vis Sci 46, 1473-1479.
36
Figure 1. Canonical visual cycle in vertebrates
37
Figure 1. Canonical visual cycle in vertebrates
Rod photoreceptors exclusively use the RPE for recycling their visual pigment through a series of
enzymatic reactions. For a detailed description of the pathway refer to (Lamb and Pugh, 2004).
Enzymatic activities are depicted by numbers: 1, all-trans ROL dehydrogenase; 2, retinylester
synthase; 3; retinylester hydrolase; 4, isomerase; 5, 11-cis ROL dehydrogenase. Transport of
retinoids between the rod outer segment and the RPE (indicated by dashed arrows) is thought to
involve the retinoid-binding protein IRBP (Bunt-Milam and Saari, 1983; Fong et al., 1984; Jin et
al., 2009; Lamb and Pugh, 2006; Parker et al., 2009; Pepperberg et al., 1993).
38
Figure 2
39
Figure 2. Vertebrate cone photoreceptors utilize two retinoid cycles to recycle their visual
pigment
Cone photoreceptors use two parallel visual cycles: the canonical cycle localized to the RPE
(depicted in green arrows), as well as an intra-retinal cycle involving Müller glia cells (depicted
in red arrows). IRBP is hypothesized to play a role in transporting retinoids between the
photoreceptors and the RPE, as well as Müller glia cells (Bunt-Milam and Saari, 1983; Fong et
al., 1984; Jin et al., 2009; Lamb and Pugh, 2006; Parker et al., 2009; Pepperberg et al., 1993). For
a comprehensive review of the intra-retinal visual cycle please refer to (Muniz et al., 2007).
Enzymatic activities are depicted by numbers: 1, all-trans ROL dehydrogenase; 2, retinylester
synthase; 3; retinylester hydrolase; 4, isomerase; 5, 11-cis ROL dehydrogenase. Novel enzymatic
activities unique to the intra-retinal visual cycle (in red arrows) include an isomerase (Mata et al.,
2005), a RE hydrolase (Bustamante et al., 1995), a retinylester synthase (Mata et al., 2002; Muniz
et al., 2006), and a retinol dehydrogenase (Mata et al., 2002; Miyazono et al., 2008).
40
Figure 3. HPLC analysis of ocular retinoids in visual cycle-impaired zebrafish larvae
(A) Injection of rpe65 MO as well as administration of the Rpe65 inhibitor retinylamine (RetNH2) lead to equally reduced levels of 11-cis RAL after bleaching and subsequent dark
adaptation, without showing a synergistic effect. (B) Knock-down of both zebrafish cralbp a and
b severly impact chromophore regeneration; the combined knock-down exacerbates the effect.
For experimental details see (Schonthaler et al., 2007) and (Fleisch et al., 2008)
41
Figure 4
42
Figure 4. Electrophysiological assessment of visual system function in visual cycle-impaired
zebrafish larvae
(A) ERG recordings of zebrafish larvae with reduced cralbp a and b levels illustrate impaired
vision as a consequence of reduced chromophore levels; interestingly Cralbp b (Müller cellspecific paralog) seems to be more important for visual pigment recycling. (B) A comparison of
the averaged saturated b-wave amplitudes of cralbp a, b and double MO-injected larvae
highlights these observations: lack of either cralbp paralog leads to significantly reduced visual
function; the effect of cralbp b knock-down is more severe than the effect of cralbp a knockdown. For experimental details see (Schonthaler et al., 2007) and (Fleisch et al., 2008)
43
Figure 5. Behavioral assessment of visual system function in visual cycle-impaired zebrafish
larvae
cralbp a (RPE-specific paralog) as well as rpe65 (expressed in the zebrafish RPE) knock-down
have no effect on visual system performance, measured by the OKR, whereas cralbp b (Müller
cell-specific paralog) knock-down results in decreased contrast sensitivity.
For experimental details see (Schonthaler et al., 2007) and (Fleisch et al., 2008)
44