Experimental Eye Research 111 (2013) 17e26
Contents lists available at SciVerse ScienceDirect
Experimental Eye Research
journal homepage: www.elsevier.com/locate/yexer
Controlling the number of melanopsin-containing retinal ganglion cells by early
light exposure
Jie Hong a,1, Qiang Zeng b,1, Huaizhou Wang a, Debbie S. Kuo c, William H. Baldridge d, Ningli Wang a, *
a
Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University, Beijing Ophthalmology & Visual Sciences Key Laboratory,
No 1# Dong Jiao Min Xiang Dongcheng District, Beijing 100730, China
b
State Key Laboratory of Brain and Cognitive Sciences, Institute of Biophysics, Chinese Academy of Sciences, 15 Da-tun Road, Beijing 100101, China
c
Department of Ophthalmology, University of California, San Francisco School of Medicine, San Francisco, CA 94143, USA
d
Department of Anatomy & Neurobiology, Dalhousie University, Halifax, Nova Scotia, Canada
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 August 2012
Accepted in revised form 14 March 2013
Available online 27 March 2013
A small percentage of retinal ganglion cells (RGCs) express melanopsin and are intrinsically photosensitive (ipRGCs). Whether light can affect the development of ipRGCs is not clear. In the rat retina, we
found constant light exposure during the first postnatal week significantly increased the number of
melanopsin immunopositive ipRGCs. This increase was durable and specific for melanopsin immunopositive ipRGCs. BrdU labeling showed no proliferation of the melanopsin immunopositive ipRGCs
during constant light exposure. Retrograde labeling from the superior colliculus showed that no other
types of RGCs were induced to express melanopsin. Light exposure was effective in increasing melanopsin immunopositive ipRGCs only when it coincided with the apoptotic phase of RGC development.
However, daily intravitreous injection of tetrodotoxin, blocking action potentials, abolished the light
induced increase of melanopsin immunopositive ipRGCs. These findings indicate that early light exposure can increase the number of melanopsin immunopositive ipRGCs through a process dependent on
intrinsic photosensitive spiking activity. Furthermore, the increase of melanopsin immunopositive
ipRGCs is potentially induced by apoptosis suppression in ipRGCs or enhanced expression of melanopsin.
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
development
melanopsin-containing retinal ganglion cell
constant light
1. Introduction
The discovery of a subtype of retinal ganglion cells (RGCs) utilizing a novel opsin, melanopsin, as a photopigment to generate
intrinsic light responses has allowed for rapid progress in the past
decade toward understanding the non-image forming visual system (Provencio et al., 1998; Provencio et al., 2000; Berson et al.,
2002; Hannibal et al., 2002; Hattar et al., 2002; Panda et al.,
2002; Provencio et al., 2002; Hattar et al., 2003; Chen et al.,
2011). It has been reported that constant light exposure suppresses melanopsin expression in the adult retina (Hannibal, 2006;
Hannibal et al., 2007). Melanopsin-containing RGCs exhibit light
responses at birth and therefore are the earliest light responsive
neurons in the retina (Sekaran et al., 2005; Tu et al., 2005). Similar
to many other types of RGCs, ipRGCs are overproduced and the
population decreases in early development. We postulated that
* Corresponding author. Tel./fax: þ86 10 5826 9920.
E-mail addresses: [email protected], [email protected] (N. Wang).
1
These authors contribute equally to this work.
0014-4835/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.exer.2013.03.011
light affects the development of ipRGCs and investigated the effects
of constant light exposure on developing and mature ipRGCs.
2. Methods
2.1. Animals and retina preparation
Albino Sprague Dawley rats obtained from the animal care facility of the Institute of Biophysics, Chinese Academy of Sciences,
were used in this study. Use and handling of animals were strictly in
accordance with the institutional guidelines and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.
During constant light exposure (LL) treatments, rats were exposed
to a continual luminance of 400 lux. Control rats were maintained
on a 12 h light/dark cycle (LD) with the same (400 lux) luminance
during the day.
Animals were deeply anaesthetized with an i.p. injection of
ketamine (50 mg/kg) and xylazine (10 mg/kg) and perfused with
warm saline transcardially. Eyes were enucleated, marked at the
nasal pole, and fixed for 10 min in 4% paraformaldehyde in 0.01 M
18
J. Hong et al. / Experimental Eye Research 111 (2013) 17e26
PBS. Then retinas were isolated from the pigment epithelium while
maintaining the reference at the nasal pole and were post-fixed for
1 h at room temperature.
2.2. Immunohistochemistry
After rinsing with 0.01 M PBS, retinas were incubated with
primary antibodies in medium containing 0.01 M PBS with 1% BSA
and 0.5% Triton-X 100 for 2 days at 4 C. Then retinas were rinsed
with 0.01 M PBS and incubated with secondary antibodies in the
same medium with 0.05% DAPI overnight at 4 C. Retinas were
rinsed and whole-mounted with Vectashield (Laboratories, H1000, Vector, Burlingame, CA). Images were acquired on an
Olympus FV500 confocal microscope with a 10objective (UPlanApo, N.A. ¼ 0.40).
For retinal sections, retinas were cryoprotected in 30% sucrose in
0.01 M PBS overnight and frozen in OCT. Tissue was sectioned to a
thickness of 18 mm using a cryostat (Leica CM, 1950). The sections
were incubated with primary antibodies overnight, rinsed three
times and incubated with secondary antibodies and 0.05% DAPI for
2 h at room temperature. After rinsing, the sections were coverslipped with Vectashield. Images were acquired on an Olympus
FV500 confocal microscope with a 60 oil-immersion objective
(PlanApo, N.A. ¼ 1.40).
2.3. Antibodies
Primary antibodies used in this study were: rabbit antimelanopsin (1:400, PA1-781, Affinity BioReagents, Golden, CO),
rabbit anti-PKCa (1:10,000, P4334, SigmaeAldrich, St Louis, MO),
mouse anti-neurofilament (1:200, N1042, SigmaeAldrich, St Louis,
MO) and mouse anti-BrdU (1:200, B8434, SigmaeAldrich, St Louis,
MO).
Secondary antibodies were all obtained from Jackson ImmunoResearch Laboratories (West Grove, PA): donkey anti-rabbit FITC
(1:200, 711-095-152), donkey anti-mouse TRITC (1:200, 711-025151) and donkey anti-rabbit Cy5 (1:200, 711-175-152).
2.4. In situ hybridization
A 539 bp rat PACAP fragment was amplified by RT-PCR
(307 bp to 845 bp, primers: 50 -GTAGCGGAGCAAGGTTGGCCC-30
and 50 -ACGGTAGCTGGCAACTCATCGCT-30 ) from rat retinal total
RNA and cloned into pGM-T (VT202-01, Tiangen, Beijing, China).
Digoxigenin-labeled riboprobes were prepared by in vitro transcription reactions (11277073910, Roche, Mannheim, Germany)
with linearized vector DNA as the template using SP6 (for antisense
probe) or T7 (for sense probe) polymerase (10810274001 and
10881767001, Roche, Mannheim, Germany) following the manufacturer’s instructions.
In situ hybridization of the whole mount retina was performed
as described by Z Bao et al.(Bao and Cepko, 1997). Briefly, after
fixation overnight in 4% PFA in 0.01 M PBS at 4 C, rat retinas were
rinsed and treated with 10 mg/ml proteinase K (1092766, Roche,
Mannheim, Germany) for 15 min at room temperature. After 1 h
pre-hybridization at 70 C, the retinas were hybridized with
PACAP riboprobes (0.2 mg/ml) overnight at 70 C. Then retinas were
washed and incubated overnight at 4 C with sheep antidigoxigenin-AP (1:10,000, 11093274910, Roche, Mannheim, Germany). After rinsing, the retinas were incubated with NBT/BCIP
reagents (11383213001 and 11383221001, Roche, Mannheim, Germany) for color detection following product instructions. Images
were acquired by a Nikon E800 microscope using a 4 (Plan Fluor,
N.A. ¼ 0.13) or 10 (Plan Apo, N.A. ¼ 0.45) objective.
2.5. BrdU labeling
Rats reared in LL condition were injected i.p. daily with 50 mg/g
body weight BrdU (B5002, SigmaeAldrich, St Louis, MO) from P0 to
P7 under hypothermia. Rats were returned to normal condition
from P8 and sacrificed at P20. The retinas were prepared and cryosectioned. Sections were rinsed with 0.01 M PBS, incubated in 4 N
HCl for 15 min at room temperature, stained with mouse anti-BrdU
antibody (1:200, B8434, SigmaeAldrich, St Louis, MO) and counterstained with rabbit anti-melanopsin antibody and rabbit antiPKCa antibody.
2.6. Intraocular injection
From P0 to P7, rats were anaesthetized by hypothermia, the
eyelid was opened with a small incision, and 0.5 ml tetrodotoxin
(0.4 mM, T8024, SigmaeAldrich, St Louis, MO) in 3.5 mM citrate
buffer (pH ¼ 4.8) was injected into the vitreous body with a glass
micropipette daily. Rats in the control group were injected with
citrate buffer only.
2.7. TUNEL
Whole-mount retinas were fixed for 20 min in 4% paraformaldehyde in 0.01 M PBS at room temperature. After incubation
in 0.01 M PBS containing 2% Triton X-100 at 37 C for 1 h, the retinas
were stained with TUNEL reaction mixture at 37 C for 1 h following
the manufacturer’s instructions (Roche, In Situ Cell Death Detection
Kit, 12156792001, Boehringer Mannheim, Mannheim). After TUNEL
staining, whole-mount retinas were counterstained with antibodies against melanopsin.
2.8. Retrograde labeling
Animals were anesthetized and placed in a stereotaxic device
(Narishige Scientific Instruments, Tokyo). A craniotomy was
preformed above the injection sites (SC: 6 and 7 mm AP,
1.4 mm ML, 3.5 and 4.5 mm DV). Then 0.1 ml 4% FluoroGold
(80014, Biotium, Hayward, CA) was injected at each point using a
Hamilton syringe, which was kept in place for 10 min after injection. The animals recovered for 5 days before being sacrificed for
examination of RGCs.
2.9. Quantification of the retinal ganglion cells
The micrographs of whole mount retinas were acquired by a
Nikon E800 microscope with a 4 objective (Plan Fluor,
N.A. ¼ 0.13) and individual images were stitched together to form a
whole-retina montage. Cell counts were made by at least two individuals blind to the experimental condition and the average of
the two observers was used.
To account for area difference, whole-mount retina was divided
into four quadrants; micrographs were taken along the midline of
each quadrant from the optic disc to the periphery at an interval of
500 mm. The density (cells/mm2) of RGCs in each photograph was
calculated and assigned to that site.
2.10. Statistical analysis
RGC density (cells/mm2) or number is presented as mean standard error of the mean (SE), unless stated otherwise. One-way
ANOVA (Origin 7.5) or Independent-samples T test (SPSS 11.5) was
used for the analysis of the results. Statistical significance was set at
p < 0.025 based on the Bonferroni correction for multiple comparisons when the t-test was applied or p < 0.05 for ANOVA analysis.
J. Hong et al. / Experimental Eye Research 111 (2013) 17e26
3. Results
3.1. Constant light exposure down regulates melanopsin expression
in adult retinas
We reared adult rats (LD30) in an environment with constant
light exposure for 20 days (LL20). Immunohistochemical analysis
showed a clear reduction in melanopsin labeling. The meshwork of
melanopsin positive dendrites visible in the control retinas (Fig. 1A)
was notably absent in light treated retinas, which only displayed
few primary dendrites (Fig. 1B). The number of somas positive for
melanopsin was also significantly reduced in light treated retinas
19
compared to controls, especially in the superior and temporal
quadrants (Fig. 1E and F). These findings are consistent with earlier
studies of melanopsin expression in adult rat retinas (Hannibal
et al., 2005).
It is not clear whether the reduction in number of melanopsin
immunopositive ipRGCs is due to the down regulation of melanopsin expression to a level below antibody detection as interpreted in the previous study or to the loss of ipRGCs. To address this
question, we returned animals exposed to constant light for 20 days
to normal lightedark condition for 10 days (LL20 þ LD10). The
number of melanopsin immunopositive ipRGCs and the intensity of
melanopsin immunostaining recovered to normal (Fig. 1CeG),
Fig. 1. Constant light exposure reversibly suppressed the level of melanopsin immunostaining in adult retina. (AeC) Representative micrographs of temporal retinas 2.5 mm from
the optic nerve head illustrating melanopsin immunostaining of ipRGCs. (A) In control retinas (LD50), melanopsin positive somas and processes were clearly visible. (B) Fewer
melanopsin immunopositive ipRGCs were noted in retinas receiving a 20-day constant light exposure (LD30 þ LL20). (C) Following ten days of normal light/dark cycle conditions,
after a 20-day constant light exposure, the number of melanopsin immunopositive somas and intensity of melanopsin immunostaining recovered (LD20 þ LL20 þ LD10). Scale bar:
200 mm (D, E, F and G) reduction and recovery in mean densities (cells/mm2) (SE) of melanopsin immunopositive ipRGCs in different retinal quadrants. Eccentricity is listed
relative to the optic disc. Significant differences were observed in the superior, temporal, and inferior quadrants. (***: p < 0.001, n for each group listed in D).
20
J. Hong et al. / Experimental Eye Research 111 (2013) 17e26
supporting that the apparent reduction in melanopsin immunopositive ipRGCs was due to the down regulation of melanopsin
expression rather than the loss of melanopsin immunopositive
ipRGCs.
3.2. Constant light exposure in the early postnatal period increases
the number of melanopsin immunopositive ipRGCs
When P0 animals were subjected to constant light for 20 days, a
similar reduction in the intensity of melanopsin immunostaining
in the retina was also observed at the end of the treatment
period (Fig. 2B, E and F). When we returned the LL20 animals to the
normal environment for 10 days, the melanopsin positive somas
and dendrites became clearly visible (Fig. 2C). Surprisingly, the
number of melanopsin immunopositive ipRGCs in the LL20 þ LD10
retinas did not simply return to the control level, as we observed in
adult retinas, but greatly exceeded the number of melanopsin
immunopositive ipRGCs in the control retinas, especially in the
superior and temporal quadrants (Fig. 2C, E and F). This finding is
consistent with data from a recent report studying the effects of
different light conditions on melanopsin-expressing ipRGCs in albino mice (González-Menéndez et al., 2010a,b).
To check whether constant light exposure induced increases in
other type of RGCs, we examined the population of neurofilamentpositive RGCs. The number of neurofilament-positive RGCs with
large somas was not affected by exposure to constant light in early
Fig. 2. Constant light exposure in the early postnatal period significantly increased the number of melanopsin immunopositive ipRGCs. (AeC) Representative micrographs of
temporal retinas 2.5 mm from the optic nerve head illustrating the effect of constant light on melanopsin immunostaining. (A) Melanopsin antibody staining of ipRGCs in wholemount control retina (LD30) revealed prominent cell bodies and dendrites. (B) A reduction of melanopsin positive somas and processes were found after 20 days exposure to
constant light (LL20 in 400 lux environment). (C) A melanopsin positive meshwork of dendrites and many more melanopsin immunopositive ipRGCs were observed after 10 days
recovery (12 h light/dark cycle) following a 20-day constant light exposure from P0 (LL20 þ LD10). Scale bar: 200 mm (D, E, F, and G) mean densities (cells/mm2) (SE) of melanopsin
immunopositive ipRGCs were calculated along the midline in each four quadrants of the rat retina at 1 mm intervals from the optic nerve. Following LL20 þ LD10, the mean density
of melanopsin immunopositive ipRGCs increased in the superior (E) and temporal (F) quadrants. (***: p < 0.001; n for each group listed in D).
J. Hong et al. / Experimental Eye Research 111 (2013) 17e26
postnatal stage (Fig. 3AeD), indicating the increase in number of
melanopsin immunopositive ipRGCs was specific.
To determine if the increase of melanopsin immunopositive
ipRGCs in LL20 þ LD10 retinas was a sustained change, we extended
the recovery period to 30 days after a 20-day exposure to constant
light. The increase of melanopsin immunopositive ipRGCs persisted
in temporal and superior quadrants of retinas after 30 days LD recovery (Fig. 4B and C), indicating this was not a transient change.
These findings contrast with the data presented by GonzálezMenéndez et al., (2010a,b) showing the number of ipRGCs was
increased by light exposure but then returned to normal after an
extended recovery period in mice .
3.3. An increased population of ipRGCs contributes to the increased
number of melanopsin immunopositive ipRGCs
21
exposed to constant light in the early postnatal period (LL8 þ LD12,
Fig. 5), supporting that the increased melanopsin immunopositive
ipRGCs number is independent of melanopsin immunostaining.
To summarize, we showed exposure to constant light reversibly
suppressed melanopsin expression in adult and neonatal rat retinas. However, exposure to constant light in the early postnatal
period led to an increase in the number of melanopsin immunopositive ipRGCs that was not observed in adult retinas and that was
specific and sustained for at least 30 days. An increase in the
number of melanopsin immunopositive ipRGCs was confirmed by
PACAP in situ hybridization. In the following sections, we consider
several potential mechanisms that could account for the increase in
number of melanopsin immunopositive ipRGCs: proliferation,
conversion, overexpression and suppression of apoptosis.
3.4. Light exposure does not induce proliferation
Since increased immunostaining with melanopsin could result
from both increased melanopsin expression and increase in number
of melanopsin immunopositive ipRGCs, we examined the number of
melanopsin immunopositive ipRGCs with a melanopsin independent marker. It has been shown that pituitary adenylate cyclaseactivating polypeptide (PACAP) is selectively expressed in ipRGCs
(Hannibal et al., 2002). Similar to LL20 þ LD10 treated retinas, we
found LL8 þ LD12 treated retinas demonstrated significantly
increased numbers of melanopsin immunolabeled cells compared
to control retinas (data not shown). We confirmed the number of
melanopsin immunopositive ipRGCs by performing in situ hybridization of PACAP. More PACAP positive cells were detected in retinas
RGCs are among first neurons to differentiate in the retina and
the proliferation of RGCs ends around birth (Cepko et al., 1996;
McNeill et al., 2011). We injected BrdU daily during the first postnatal week under constant light exposure to determine if proliferation of ipRGCs was induced. BrdU positive cells were observed in
the outer nuclear layer (ONL) and the inner nuclear layer (INL)
(Fig. 6A). Some cells were double-labeled with PKCa, showing that
dividing photoreceptors and rod bipolar cells were reliably labeled.
Very few BrdU positive cells were observed in the ganglion cell
layer. None of the 127 melanopsin immunopositive ipRGCs we
examined were labeled with BrdU, indicating that the light-
Fig. 3. Constant light exposure in the early postnatal period did not change the number of neurofilament positive RGCs with large somas. (A and B) Representative micrographs
illustrating the effect of constant light on neurofilament positive RGCs with large somas. (A) Neurofilament antibody staining of RGCs in whole-mount control retina (LD30). (B)
Neurofilament antibody staining of RGCs in whole-mount retina (LL20 þ LD10). Scale bar: 200 mm (C and D) the mean density (cells/mm2) (SE) of neurofilament positive RGCs
with large somas was not influenced by treatment with 20-day constant light exposure from P0 followed by10 days recovery (12 h light/dark cycle) (LL20 þ LD10) in either nasal (C)
or temporal (D) quadrants of the retina compared to controls (LD30). (n for each group listed in C).
22
J. Hong et al. / Experimental Eye Research 111 (2013) 17e26
Fig. 4. The effect of constant light exposure on the number of melanopsin immunopositive ipRGCs was specific and sustained. (A, B, C and D) The increase in the number of
melanopsin immunopositive ipRGCs in retinas receiving a 20-day exposure to constant light persisted after 30 days recovery in normal light/dark cycle (LL20 þ LD30). The mean
(SE) increase remained significant in the superior (B) and temporal (C) quadrants of the retina. (**: p < 0.01, ***: p < 0.001; n listed in the bar for each group).
induced increase of melanopsin immunopositive ipRGCs was not
due to increased proliferation of ipRGCs.
3.5. Light exposure does not induce non-melanopsin containing
cells to express melanopsin
Another possible explanation for the increase in the number of
melanopsin immunolabeled cells is that non-ipRGCs were
induced to express melanopsin by constant light exposure. The
majority of RGCs project to the superior colliculus (SC) (Vaney
et al., 1981; Linden and Perry, 1983), but only about 10% of
ipRGCs project to the SC (Hattar et al., 2006; Baver et al., 2008;
Pickard and Sollars, 2010). The projection of ipRGCs appears to be
independent of the presence or absence of melanopsin (Lucas
et al., 2003). Therefore, if constant light exposure induces melanopsin expression in non-ipRGCs, one would expect an increased
percentage of SC projecting ipRGCs. We investigated this possibility by retrograde labeling of RGCs from the SC. Despite the
increase in the number of melanopsin immunopositive ipRGCs in
the retina, the percentage of SC projecting melanopsin immunopositive ipRGCs remained unchanged after constant light exposure (Fig. 6B), demonstrating that the increase of melanopsin
immunopositive ipRGCs was not due to the conversion of nonipRGCs to ipRGCs.
Fig. 5. Constant light exposure in the early postnatal period increased the number of PACAP positive RGCs. (A1, A2, B1, B2) Representative high magnification (A1 and A2) and low
magnification (B1 and B2) micrographs of in situ hybridization of PACAP mRNA on whole-mount rat retina. The density (cells/mm2) of PACAP positive cells in control group (A1 and
B1) was lower than the density (cells/mm2) in LL8 þ LD12 (A2 and B2) group in a similar region of superior retina. Scale bars: 200 mm. (C) Mean (SE) total number of PACAP
positive cells in the control and LL8 þ LD12 rat retinas. The number of PACAP positive cells increased significantly after early exposure to constant light from P0 to P7 and recovery in
normal light/dark cycle for 12 days, consistent with the melanopsin immunostaining results (Fig. 2.). (*: p < 0.001, n listed in the bar for each group).
J. Hong et al. / Experimental Eye Research 111 (2013) 17e26
23
Fig. 6. Constant light exposure did not induce proliferation of ipRGCs or non-ipRGCs to express melanopsin. (A) Representative micrograph showing BrdU labeling of rat retina. Rats
were injected daily with BrdU from P0 to P7 during the period of constant light exposure, and then returned to normal conditions (12 h light/dark cycle) for 12 days (LL8 þ LD12).
Retinal cryosections were quadruple labeled with melanopsin (cyan), BrdU (green), DAPI (blue) and PKCa (red). BrdU positive cells were observed in the outer and inner nuclear
layers, and some colocalized with rod bipolar cells (PKCa, red). Only a few BrdU positive cells were detected in the ganglion cells layer (GCL), and they never colocalized with any of
the 127 ipRGCs examined, supporting the absence of ipRGCs proliferation after birth. Scale bar: 40 mm (B) RGCs in rats from control (LD30) group and light treatment (LL20 þ LD10)
group were subjected to retrograde labeling with fluorogold (FG) applied to the superior colliculus (SC) and counter-staining with melanopsin antibody. The mean number (SE) of
melanopsin positive cells and FG þ melanopsin positive cells were sampled in the region 3 mm from optic disc in the temporal and nasal quadrants of the retina. The percentage of
SC projecting ipRGCs did not change in animals reared in constant light conditions despite the dramatic increase in number of melanopsin immunopositive ipRGCs, demonstrating
that conversion of non-ipRGCs to ipRGCs did not occur in RGCs projecting to the SC.
3.6. TTX blocks the increase of melanopsin immunopositive ipRGCs
induced by constant light exposure
To examine the possible role of spiking activity resulting from
activation of melanopsin by light exposure, we performed daily
injection of a voltage gated Naþ channel blocker, tetrodotoxin (TTX).
We previously showed that daily injection of TTX is sufficient to
block spiking activities of RGCs in the retina throughout the treatment period (Sun et al., 2011). TTX completely blocked the lightinduced increase in melanopsin immunopositive ipRGCs (Fig. 7).
As repeated intraocular injection of TTX may induce changes in eye
diameter, we showed the total number of melanopsin immunopositive ipRGCs rather than the density in specific regions.
3.7. Apoptosis of melanopsin immunopositive ipRGCs occurs during
early postnatal development
One of the remaining possibilities is that constant light exposure
increases the number of melanopsin immunopositive ipRGCs
Fig. 7. Daily intraocular injection of TTX blocked the light exposure induced increase of
melanopsin immunopositive ipRGCs. Mean (SE) total number of melanopsin
immunopositive ipRGCs calculated from P0, LD20, LL8 þ LD12 retinas, as well as
LL8 þ LD12 retinas receiving daily intravitreal injection of TTX to block spiking activity
from P0 to P7. TTX blocked the light induced increase of melanopsin immunopositive
ipRGCs, indicating this effect was activity dependent. (*: p < 0.05 by ANOVA).
through suppression of apoptosis that normally occurs during
development (Potts et al., 1982; Perry et al., 1983). It is difficult to
demonstrate the reduction of apoptotic melanopsin immunopositive ipRGCs directly because the level of melanopsin expression
is greatly down regulated by constant light exposure, as we have
shown earlier. Attempts to use antibodies against PACAP, a
melanopsin-independent ipRGC marker, never worked in our experiments. This is likely due to low transcription or expression of
PACAP during the early postnatal period. In situ hybridization only
showed very weak PACAP signal in early postnatal retinas (data not
shown). While we were unable to directly test apoptosis in our
population of melanopsin immunopositive ipRGCs, we have obtained some evidence addressing the hypothesis of apoptosis
suppression.
Many TUNEL positive cells were identified in P3 and P5 retinas
and some were clearly double labeled with antibodies against
melanopsin (Fig. 8A and B). Very few apoptotic cells were seen at P8
and they were found only in the far peripheral retina. Therefore
ipRGCs undergo apoptosis and the apoptotic phase is very similar to
that of other types of RGCs (Young, 1984; Linden and Pinto, 1985).
Next we maintained animals under one of two paradigms:
LL8 þ LD22 and LD8 þ LL10 þ LD12. Eight days light exposure from
P0 significantly increased the number of melanopsin immunopositive ipRGCs, whereas 10 days exposure from P7 did not
(Fig. 8CeF). As previously discussed, twenty days constant light
exposure after P20 down regulated the intensity of melanopsin
immunostaining but did not change the number of melanopsin
immunopositive ipRGCs either (Fig. 1DeG). Therefore, light exposure was effective in increasing melanopsin immunopositive
ipRGCs only when it coincided with the apoptotic phase of RGC
development.
Given that there is no proliferation of melanopsin immunopositive ipRGCs or conversion of other RGCs into ipRGCs and that
apoptosis of ipRGCs occurs around P3 and P5, one would expect
that the peak number of ipRGCs appears at birth. If light exposure
suppresses apoptosis of ipRGCs during the first postnatal week,
then retinas receiving constant light exposure should have more
ipRGCs than the control retinas, with the upper limit approaching
the number of ipRGCs at P0. Therefore, we compared the total
number of melanopsin immunopositive ipRGCs in neonatal, normal
P20, and LL8 þ LD12 retinas. Interestingly, while more melanopsin
24
J. Hong et al. / Experimental Eye Research 111 (2013) 17e26
Fig. 8. Constant light exposure was effective in increasing melanopsin immunopositive ipRGCs only during the developmental period of RGC apoptosis. (A and B) Representative
micrographs illustrating TUNEL positive cells at P3 (A) and P5 (B), with some cells double labeled with melanopsin antibody (arrowheads). Red: TUNEL. Green: melanopsin. Scale
bar: 40 mm (C, D, E and F) the mean (SE) densities (cells/mm2) of melanopsin immunopositive ipRGCs were calculated in different retinal quadrants from LD30, LL8 þ LD22 and
LD8 þ LL10 þ LD12 groups. No significant difference was observed between LD30 and LD8 þ LL10 þ LD12 retinas, indicating that constant light exposure after P7 had little effect on
melanopsin immunopositive ipRGC number. Significantly more melanopsin immunopositive ipRGCs were found in retinas from LL8 þ LD22 group compared to controls in the
superior and temporal quadrants. Thus light exposure was only effective between P0 and P7, coinciding with the period of RGCs apoptosis during development. (***: p < 0.001, n for
each group listed in C).
immunopositive ipRGCs were found in LL8 þ LD12 retinas than P20
retinas as expected, the number of melanopsin immunopositive
ipRGCs was very similar between P0 and P20 retinas reared under
normal conditions (Fig. 7). These data seems to contradict our
finding that apoptosis occurs in ipRGCs in normal P3 and P5 retina,
however the population of melanopsin immunolabeled cells represents only a subset of ipRGCs (Ecker et al., 2010) and it has been
shown that P0 retina have twice as many intrinsically photosensitive RGCs than adult retina (Sekaran et al., 2005). Thus, these data
are likely a byproduct of the limitations of our antibody’s ability to
label ipRGCs and the dynamic process of melanopsin expression in
these cells rather than an absence of apoptosis.
4. Discussion
In this study, we showed that constant light exposure reversibly
suppressed melanopsin expression in adult retina, consistent with
a few earlier studies (Hannibal, 2006; Hannibal et al., 2007).
Notably, constant light exposure in the early postnatal period followed by return to normal light conditions increased the number of
melanopsin immunopositive ipRGCs, and this effect was sustained
for at least 30 days. A transient increase in melanopsin immunopositive ipRGCs following early postnatal constant light exposure
has been reported in mice, however this is the first report of a
sustained effect (González-Menéndez et al., 2010a,b). Furthermore,
the mechanism(s) for this observation has never been explored. We
found that the increase of melanopsin immunopositive ipRGCs was
not induced by proliferation of existing ipRGCs or conversion of
other RGCs to ipRGCs. Further experiments showed that melanopsin immunopositive ipRGCs were increased by constant light
only if treatment occurred before P8. This effect was shown to be
dependent on intrinsic photosensitive spiking activity. One possible
mechanism for this light-sensitive increase in the number of melanopsin immunopositive ipRGCs is that constant light suppresses
J. Hong et al. / Experimental Eye Research 111 (2013) 17e26
melanopsin immunopositive ipRGCs apoptosis through light sensitive spiking activity.
Many lines of evidence suggest that the light induced increase of
melanopsin immunopositive ipRGCs may be due to suppression of
apoptosis. First, ipRGCs are overproduced early in development.
Calcium imaging and multielectrode array recordings showed four
times as many neurons exhibit intrinsic light responses in P0 retinas compared to adult retinas, and twice as many when gap
junctions were blocked (Sekaran et al., 2005; Tu et al., 2005). In
adult retinas, ipRGCs are only coupled with amacrine cells (Perez de
Sevilla Müller et al., 2010). If this is the same early in development,
and the gap junction blocker suppresses intrinsic light responses
from coupled amacrine cells (Sekaran et al., 2005), then ipRGCs are
overproduced by at least two fold early in development, similar to
many other types of RGCs (Williams et al., 1990). Second, as we
showed in this study, ipRGCs did undergo apoptosis at a similar
period as other types of RGCs. Third, constant light exposure was
only effective in increasing the number of ipRGCs when it coincided
with the period of apoptosis during RGC development.
Suppression of apoptosis could be mediated in part by activation of ipRGCs. As mentioned above, the discrepancy between
the number of melanopsin immunopositive cells and intrinsically
photosensitive cells, suggests that low levels of melanopsin
expression, below antibody detection, are sufficient to produce
photosensitivity. Thus, ipRGCs with decreased melanopsin labeling
during constant light exposure could continue to be light
responsive. Spiking activity has been reported to suppress
apoptosis, as reviewed by Mennerick and Zorumski (2000). Constant light exposure may result in constant but low level activation
of ipRGCs due to their intrinsic light sensitivity. This hypothesis is
consistent with our experiments showing that blocking spiking
activity with TTX during constant light exposure abolished the
light induced increase of ipRGCs. Light-induced activity appears to
rescue some ipRGCs that normally undergo apoptosis.
Despite the evidence supporting suppression of apoptosis as the
mechanism for the phenomenon we observed in our study, other
mechanisms cannot be ruled out. Another possibility is that melanopsin is expressed in more cells following early exposure to
constant light. An interesting finding in our study was that number
of melanopsin immunopositive ipRGCs was similar between P0 and
P20 retinas from rats housed under normal lighting conditions and
significantly increased in retinas from rats treated with constant
light in the early postnatal period. A recent paper reported similar
numbers of ipRGCs in newborn and adult mice (GonzálezMenéndez et al., 2010a,b), which is consistent with our data.
These authors concluded that there was no apoptosis occurring in
ipRGCs, but our results show that ipRGCs undergo apoptosis around
P3 and P5. To reconcile this discrepancy, one needs consider a
second process: increasing melanopsin expression. As previously
discussed, calcium imaging showed twice as many ipRGCs in P0
retinas compared to adult retinas (Sekaran et al., 2005), while immunostaining showed similar numbers between the two groups in
both our data and those of González-Menéndez et al., (2010a,b).
Therefore, a portion of ipRGCs that showed intrinsic light responses
are not detectable by antibodies. This conclusion is supported by a
recent finding that intrinsic light responses can be detected in
several subtypes of ipRGCs which are negative for antibody staining
(Ecker et al., 2010). Taken together, it is possible that the increase of
melanopsin immunopositive ipRGCs following constant light
exposure and recovery results from enhanced expression of melanopsin is in a subset of ipRGCs in which the expression of melanopsin was initially undetectable.
Regardless of the underlying mechanism, this novel finding that
stimulation by constant light during a formative period leads to a
sustained increase in the population of melanopsin immunopositive
25
ipRGCs brings up many important and unanswered questions. An
increase in this population of cells has now been demonstrated in
both neonatal mice and rats subjected to constant light conditions.
Does this phenomenon occur in other species, and if so, what is the
durability of the effect? Whether this response is adaptive or maladaptive remains to be determined and could be explored by testing
functions requiring input from ipRGCs such as pupillary response
and circadian rhythm. While constant light conditions do not exist
naturally, they may be found in places like hospitals and homes
where newborn babies reside and potentially affect ipRGCs in
humans. Our data show constant light exposure affects the development of ipRGCs, and understanding this phenomenon may help
us better understand the processes regulating early retinal development of intrinsically photosensitive RGCs.
Acknowledgment
We thank Yiming Sun for the help with TUNEL staining, Zhiping
Liu for immunohistochemistry and in situ hybridization, Hui Yang,
Wei Shi and Yan Yin for the quantification of ipRGCs, and Yan Yin for
assistance with TTX injection experiment. This project was supported by an NSFC key project grants (30530280), two National Basic
Research Program of China (2007CB512208 and 2006CB911003)
to SH, an NSFC-CIHR China/Canada Joint Health Research Initiative
Program (30611120529) to SH and WHB, NSFC project grants
(30672283) to NW and NSFC project grants (30973261)to HW.
References
Bao, Z.-Z., Cepko, C.L., 1997. The expression and function of notch pathway genes in
the developing rat eye. The Journal of Neuroscience 17 (4), 1425e1434.
Baver, S.B., Pickard, G.E., et al., 2008. Two types of melanopsin retinal ganglion cell
differentially innervate the hypothalamic suprachiasmatic nucleus and the
olivary pretectal nucleus. European Journal of Neuroscience 27 (7), 1763e1770.
Berson, D.M., Dunn, F.A., et al., 2002. Phototransduction by retinal ganglion cells
that set the circadian clock. Science 295 (5557), 1070.
Cepko, C.L., Austin, C.P., et al., 1996. Cell fate determination in the vertebrate retina.
Proceedings of the National Academy of Sciences of the United States of
America 93 (2), 589.
Chen, S.K., Badea, T.C., et al., 2011. Photoentrainment and pupillary light reflex are
mediated by distinct populations of ipRGCs. Nature 476 (7358), 92e95.
Ecker, J.L., Dumitrescu, O.N., et al., 2010. Melanopsin-expressing retinal ganglioncell photoreceptors: cellular diversity and role in pattern vision. Neuron 67
(1), 49e60.
González-Menéndez, I., Contreras, F., et al., 2010a. No loss of melanopsin-expressing
ganglion cells detected during postnatal development of the mouse retina.
Histology and Histopathology 25 (1), 73.
González-Menéndez, I., Contreras, F., et al., 2010b. Postnatal development and
functional adaptations of the melanopsin photoreceptive system in the albino
mouse retina. Investigative Ophthalmology & Visual Science 51 (9), 4840.
Hannibal, J., 2006. Regulation of melanopsin expression. Chronobiology International 23 (1), 159e166.
Hannibal, J., Georg, B., et al., 2007. Melanopsin changes in neonatal albino rat independent of rods and cones. Neuroreport 18 (1), 81.
Hannibal, J., Georg, B., et al., 2005. Light and darkness regulate melanopsin in the
retinal ganglion cells of the albino Wistar rat. Journal of Molecular Neuroscience 27 (2), 147e155.
Hannibal, J., Hindersson, P., et al., 2002. The photopigment melanopsin is exclusively present in pituitary adenylate cyclase-activating polypeptide-containing
retinal ganglion cells of the retinohypothalamic tract. Journal of Neuroscience
22 (1), 191.
Hattar, S., Kumar, M., et al., 2006. Central projections of melanopsin expressing
retinal ganglion cells in the mouse. The Journal of Comparative Neurology 497
(3), 326e349.
Hattar, S., Liao, H.W., et al., 2002. Melanopsin-containing retinal ganglion cells: architecture, projections, and intrinsic photosensitivity. Science 295 (5557), 1065.
Hattar, S., Lucas, R.J., et al., 2003. Melanopsin and rod-cone photoreceptive systems account for all major accessory visual functions in mice. Nature 424
(6944), 75e81.
Linden, R., Perry, V.H., 1983. Massive retinotectal projection in rats. Brain Research
272 (1), 145e149.
Linden, R., Pinto, L.H., 1985. Developmental genetics of the retina: evidence that the
pearl mutation in the mouse affects the time course of natural cell death in the
ganglion cell layer. Experimental Brain Research 60 (1), 79e86.
Lucas, R.J., Hattar, S., et al., 2003. Diminished pupillary light reflex at high irradiances in melanopsin-knockout mice. Science 299 (5604), 245.
26
J. Hong et al. / Experimental Eye Research 111 (2013) 17e26
McNeill, D., Sheely, C., et al., 2011. Development of melanopsin-based irradiance
detecting circuitry. Neural Development 6 (1), 8.
Mennerick, S., Zorumski, C.F., 2000. Neural activity and survival in the developing
nervous system. Molecular Neurobiology 22 (1), 41e54.
Panda, S., Sato, T.K., et al., 2002. Melanopsin (Opn4) requirement for normal lightinduced circadian phase shifting. Science 298 (5601), 2213.
Perez de Sevilla Müller, L., Do, M.T.H., et al., 2010. Tracer coupling of intrinsically
photosensitive retinal ganglion cells to amacrine cells in the mouse retina. The
Journal of Comparative Neurology 518 (23), 4813e4824.
Perry, V.H., Henderson, Z., et al., 1983. Postnatal changes in retinal ganglion cell and
optic axon populations in the pigmented rat. The Journal of Comparative
Neurology 219 (3), 356e368.
Pickard, G.E., Sollars, P.J., 2010. Intrinsically photosensitive retinal ganglion cells.
Science China Life Sciences 53 (1), 58e67.
Potts, R.A., Dreher, B., et al., 1982. The loss of ganglion cells in the developing retina
of the rat. Developmental Brain Research 3 (3), 481e486.
Provencio, I., Jiang, G., et al., 1998. Melanopsin: an opsin in melanophores, brain,
and eye. Proceedings of the National Academy of Sciences of the United States
of America 95 (1), 340.
Provencio, I., Rodriguez, I.R., et al., 2000. A novel human opsin in the inner retina.
Journal of Neuroscience 20 (2), 600.
Provencio, I., Rollag, M.D., et al., 2002. Anatomy: photoreceptive net in the
mammalian retina. Nature 415 (6871), 493.
Sekaran, S., Lupi, D., et al., 2005. Melanopsin-dependent photoreception provides
earliest light detection in the mammalian retina. Current Biology 15 (12),
1099e1107.
Sun, L., Han, X., et al., 2011. Direction-selective circuitry in rat retina develops
independently of GABAergic, cholinergic and action potential activity. PLoS ONE
6 (5), e19477.
Tu, D.C., Zhang, D., et al., 2005. Physiologic diversity and development of intrinsically photosensitive retinal ganglion cells. Neuron 48 (6), 987e999.
Vaney, D.I., Peichl, L., et al., 1981. Almost all ganglion cells in the rabbit retina project
to the superior colliculus. Brain Research 212 (2), 447e453.
Williams, M.A., Pinon, L.G.P., et al., 1990. The pearl mutation accelerates the
schedule of natural cell death in the early postnatal retina. Experimental Brain
Research 82 (2), 393e400.
Young, R.W., 1984. Cell death during differentiation of the retina in the mouse. The
Journal of Comparative Neurology 229 (3), 362e373.