Light Inputs to Dopaminergic Amacrine Cells of the Mammalian Retina

Home , Melanopsin

LIGHT INPUTS TO DOPAMINERGIC AMACRINE

CELLS OF THE MAMMALIAN RETINA

Víctor Pérez-Fernández

A thesis submitted in fulfilment of the requirements for the degree of

Doctor of Philosophy

School of Medicine, Western Sydney University, Australia

November 2018

Supervisor: Dr. Morven A. Cameron

Co-supervisors: Dr. Yossi Buskila and Professor John W. Morley

Statement of Authentication

The work presented in this thesis is, to the best of my knowledge and belief, original except as acknowledged in the text. I hereby declare that I have not submitted this material, either in full or in part, for a degree at this or any other institution.

Víctor Pérez-Fernández

November, 2018.

Acknowledgements

During the Middle Age a peregrine spotted three builders working on huge blocks of stone.

The man approached and asked what they were doing. “Cutting stone” said one with indifference. “Earning some money” answered the second. The third stopped cutting for a moment and full of enthusiasm responded: “I am building the most amazing cathedral the entire region has ever seen.”

Along this experience, I felt like the three builders at different time points: just cutting stone

(or pipetting), earning some money or contributing to the creation of something remarkable: knowledge. I really want to thank everyone that pushed me towards the latter thought because science –and any daily activity- should be taken with enthusiasm. I will not name them all because probably most of them will never know how they inspired me, since I would even include the will of a tiny ant carrying a branch of unproportioned dimensions to my list…

However, I would like acknowledge my main supervisor Dr. Morven Cameron for helping me through the whole experience, she cannot hide her enthusiasm about retinal biology. Second, a big recognition to Dr. David Harman and Meena Michael for supporting me on a regular basis in the Mass Spectrometry Facility, where I spent most of my time. My previously tangled relationship with organic chemistry truly improved thanks to them. Third, thanks to all colleagues and co-workers from the School of Medicine, their daily teaching and advice served me to improve scientifically and personally.

Last, but by no means least, I want to express my sincerest gratitude to my family and friends for their tenacious support, even with half of a planet between us (most of the time).

Life without them would be an unforgivable mistake.

iii

Table of Contents

Table of Contents ...... i

List of Tables ...... v

List of Figures ...... vi

List of abbreviations ...... ix

Abstract ...... xii

Publications, conferences and awards ...... xiv

CHAPTER 1: General Introduction ...... 1

1.1. Vision ...... 2

1.2. Photoreceptors ...... 6

1.2.1. Visual pigments ...... 6

1.2.2. Phototransduction cascade ...... 7

1.2.3. Morphology, properties and functions ...... 10

1.2.4. “Earliest photoreception” ...... 15

1.3. Retinal pathways ...... 17

1.3.1. Cone pathways ...... 17

1.3.2. Rod pathways ...... 22

1.4. Dopamine in the retina ...... 26

1.4.1. Light inputs to dopaminergic amacrine cells ...... 31

1.5. Aims ...... 37

CHAPTER 2: Materials and Methods ...... 38

2.1. Animals ...... 39

2.2. Dopamine release experiments...... 41

2.2.1. Light sources...... 41

2.2.2. Light pulsing, sample preparation and analysis...... 45

2.2.3. Viral injections ...... 47

2.2.4. Statistical Analysis ...... 47

2.3. Microglia characterisation ...... 49

2.3.1. Immunohistochemistry ...... 49

2.3.2. Cell Counting and 3D Reconstruction ...... 49

2.3.3. Statistical Analysis ...... 50

CHAPTER 3: Optimisation of a method to accurately quantify dopamine turnover in the retina ...... 52

3.1. INTRODUCTION ...... 53

3.2. Extended RESULTS ...... 64

3.3. Extended DISCUSSION ...... 69

CHAPTER 4: Light-induced dopamine release in wild-type mice ...... 73

4.1. INTRODUCTION ...... 74

4.2. RESULTS...... 75

4.2.1. Dim light suppresses dopamine release ...... 75

4.2.2. Light-induced dopamine release is high threshold ...... 75

4.2.3. Dopamine release is modulated via excitatory and inhibitory inputs ...... 79

4.2.4. Gap junction coupling is required to drive the majority of light-induced dopamine

release ...... 81

4.2.5. Dopamine release is enhanced with depolarisation ...... 81

4.2.6. Light-induced dopamine release occurs as early as P18 ...... 82

4.3. DISCUSSION ...... 90

4.3.1. Dopamine is released under both “dark” and bright light ...... 90

4.3.2. In vivo and ex vivo approaches provide complementary results ...... 93

4.3.3. Dopamine release does not match electrophysiological inputs ...... 95

CHAPTER 5: Photoreceptive input driving light-induced dopamine release ...... 102

5.1. INTRODUCTION ...... 103

5.2. RESULTS...... 105

5.2.1. Rod input supresses dopamine release under dim light ...... 105

5.2.2. Rods are both sufficient and necessary to drive light-induced dopamine release 106

5.3. DISCUSSION ...... 123

5.3.1. Rods define light-induced dopamine release and set its threshold ...... 123

iii

CHAPTER 6: Effect of inflammation on retinal microglia and dopamine release ...... 130

6.1. INTRODUCTION ...... 131

6.2. RESULTS...... 134

6.2.1. Microglial density increases with age ...... 134

6.2.2. Microglia tend towards an inflammatory morphological phenotype with

inflammation and age ...... 134

6.2.3. Aging, but not IL-6 upregulation, causes retraction of microglial dendrites ...... 136

6.2.4. Bivariate correlations of morphological characteristics with convex hull area .... 137

6.2.5. Light-induced dopamine release is normal in GFAP-IL6 mice ...... 138

6.3. DISCUSSION ...... 146

6.3.1. Chronic neuroinflammation diminishes the hallmarks of the microglial reactive

phenotype in aged GFAP-IL6 mice ...... 146

6.3.2. Chronic neuroinflammation does not impair light-induced dopamine release..... 150

CHAPTER 7: General Discussion ...... 152

References ...... 157

List of Tables

CHAPTER 1: General Introduction

Table 1.1 Physiological properties of ipRGC subtypes……………..…………………...13

Table 1.2 Dopamine receptor subtypes across the retina....………..…………………….30

Table 1.3 Pharmacological modulation of dopamine release………….………………...36

CHAPTER 2: Materials and Methods

Table 2.1 Intensity of light sources expressed in different units………………………..42

CHAPTER 3: Optimisation of a method to accurately quantify dopamine turnover in the retina

Table 3.1 Light-pulsing methods across the literature…….…………………………….71

Table 3.2 Analytical methods across the literature……………….……………………..72

CHAPTER 5: Photoreceptive input driving light-induced dopamine release

Table 5.1 No difference between in vivo and ex vivo thresholds for dopamine release.108

CHAPTER 6: Effect of inflammation on retinal microglia and dopamine release

Table 6.1 R2 values and Pearson r test statistics for bivariate correlations ……………..138

List of Figures

CHAPTER 1: General Introduction

Figure 1.1 Vertical section of a human retina……………………………………………...5

Figure 1.2 Possible routes conveying light signals to dopaminergic amacrine cells……...25

Figure 1.3 Proposed cone pathways conveying cone signals to dopaminergic amacrine

cells…………………………………………………………………………...34

CHAPTER 2: Materials and Methods

Figure 2.1 Light sources relative spectrograms…………………………………………..43

Figure 2.2 Opsin nomograms of the murine retina……………………………………….44

Figure 2.3 3D reconstruction and analysis using Neurolucida 360 software……………..51

CHAPTER 3: Optimisation of a method to accurately quantify dopamine turnover in the retina

Figure 3.1 “Popping” the retina from the eye-cup provides increased dopamine and

DOPAC counts instead of scope dissection …………………………………..66

Figure 3.2 Both sodium metabisulfite and EDTA cause a significant reduction in dopamine

counts ………………………………………………………………………...67

Figure 3.3 Anaesthesia reduces behavioural variation…………………………………...68

CHAPTER 4: Light-induced dopamine release in wild-type mice

Figure 4.1 Dim light supresses dopamine turnover………………………………………83

Figure 4.2 In vivo light-induced dopamine release occurs only at bright intensities……...84

Figure 4.3 Ex vivo light-induced dopamine release occurs only at bright intensities and

matches the in vivo approach………………………………………………….85

Figure 4.4 Light-induced dopamine release is reduced in retinae without RPE………….86

Figure 4.5 Light-induced dopamine release occurs within 1 minute from a dark-adapted

state…………………………………………………………………………...87

Figure 4.6 Pharmacological modulation of retinal pathways…………………………….88

Figure 4.7 Significant light-induced dopamine release occurs as early as P18………..…89

CHAPTER 5: Photoreceptive input driving light-induced dopamine release

Figure 5.1 Dopamine turnover is suppressed under dim light by rod input……………...114

Figure 5.2 Light-induced dopamine release is defined by rod input…………………….115

Figure 5.3 Rd/rd mice are capable of releasing dopamine……………………………....116

Figure 5.4 Aged rd/rd mice have higher basal dopamine turnover levels compared to

young animals but cannot release dopamine in response to light…………….117

Figure 5.5 Rd/rd mice lack light-induced dopamine release at all ages during

development…………………………………………………………… …...118

Figure 5.6 Chemogenetic activation of ipRGCs does not drive dopamine release……...119

Figure 5.7 Theoretical chromatic contributions of photoreceptors to light-induced

dopamine release…… ………………………………………………………120

Figure 5.8 Rods are both sufficient and necessary to drive light-induced dopamine

release……………………………………………………………………….121

Figure 5.9 Dopamine release is driven by gap-junction dependent and independent

pathways…………………………………………………………………….122

Figure 5.10 Widespread dopamine release is driven by rods……………………………..129

vii

CHAPTER 6: Effect of inflammation on retinal microglia and dopamine release

Figure 6.1 GFAP-IL6 mice have a significantly higher density of microglia compared to

wild-type mice at 4 but not at 24 months…………………………………….138

Figure 6.2 Microglia from all GFAP-IL6 cohorts and aged wild-type mice present reactive-

like morphology...... ……………………………………...………………..139

Figure 6.3 Age causes retraction towards the soma ………………………………….....140

Figure 6.4 Representative image of microglial retraction with age and inflammation….142

Figure 6.5 Aged microglia retract dendrites…………………………………………….143

Figure 6.6 Bivariate morphological correlations to Convex Hull Area in GFAP-IL6 and

wild-type mice……………………………………………………………….144

Figure 6.7 Light-induced dopamine release is not altered in GFAP-IL6 mice…………..145

viii

List of abbreviations

AAV Adeno-associated virus

AC Amacrine cell aCSF Artificial cerebrospinal fluid

AMPA α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic

AP5 Amino-5-phosphovaleric acid

ATP Adenosine triphosphate

BAC Bacterial artificial chromosome

BC Bipolar cell cGMP Cyclic guanosine monophosphate

CNO Clozapine N-oxide

CNQX 6-Cyano-7-nitroquinoxaline-2,3-dione

CT Circadian Time

DA Dopamine

DAC Dopaminergic amacrine cell

DAG Diacylglycerol

DAT Dopamine transporter

DIO Double-floxed inverse open reading frame

DOPAC 3,4-Dihidroxyphenylacetic acid

DREADD Designer receptors exclusively activated by designer drugs

E Embryonic day

ECD Electrochemical detection

EDTA Ethylenediaminetetraacetic acid

ELISA Enzyme-linked immunosorbent assay

ERG Electroretinogram

GABA γ-Aminobutyric acid

GC Ganglion cell

GCL Ganglion cell layer

GFAP Glial fibrillary acidic protein

GFP Green fluorescent protein

HCN Hyperpolarization-activated cyclic nucleotide-gated

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid hM3Dq human M3 muscarinic receptor

HPLC High performance liquid chromatography hSyn Human synapsin 1 gene promoter

Iba1 Ionised calcium-binding adapter molecule 1 iGluR Ionotropic glutamate receptors

IL-6 Interleukin 6

INL Inner nuclear layer

IP3 Inositol 3-phosphate

IPL Inner plexiform layer ipRGC Intrinsically photoresponsive retinal ganglion cell kV Kilovolt

L-AP4 L-2-amino-4-phosphonobutyric acid

L-DOPA L-3,4-dihydroxyphenylalanine

Log Logarithm base 10

NMDA N-methyl-D-aspartate mbar Millibar

MFA Meclofenamic acid

x mGluR Metabotropic glutamate receptors

MS/MS Mass spectrometry tandem

ONL Outer nuclear layer

OPL Outer plexiform layer

OPN Olivary pretectal nucleus

P Post-natal day

PBS Phosphate base saline

PCR Polymerase chain reaction

PLAP Placental alkaline phosphatase

PLR Pupillary light reflex

RFP Red fluorescent protein r.p.m. Revolutions per Minute

RPC Retinal Progenitor Cell

RPE Retinal Pigmented Epithelium

SCN Suprachiasmatic nucleus

S.E.M. Standard error of the mean

TH Tyrosine hydroxylase

TRITC Tetramethylrhodamine

TRP Transient receptor potential

TSPO Translocator protein

UHPLC Ultra-high performance liquid chromatography

UV Ultra-violet

WT Wild-type

λmax Wavelength of maximum absorption

Abstract

Background: The retina responds to light over a wide range of operational conditions, surpassing 10 units in a logarithmic scale. Adaptation of the retina to the particular presenting light conditions relies considerably on modulation of retinal pathways by dopamine, which is released in response to light or circadian rhythms exclusively from dopaminergic amacrine cells. Rods, cones and intrinsically photoresponsive retinal ganglion cells (ipRGCs) have all been shown to input into dopaminergic amacrine cells. However, the pathways that these photoreceptors employ to ultimately trigger dopamine release in response to light remain unclear.

Methods: Ultra-high performance liquid chromatography separation and tandem mass spectrometry detection was used to quantify dopamine, and its primary metabolite 3,4- dihidroxyphenylacetic acid (DOPAC). Retinal dopamine release was assessed under various conditions, in a variety of mouse models, using two complementary experimental designs: in vivo anaesthetised mice and ex vivo explanted retinae.

Results: In wild-type mice, significant light-induced dopamine release compared to dark- adapted retinae occurred only at bright light intensities, within one minute and after postnatal day 18. Conversely, dopamine turnover was suppressed by dim red light when compared to infrared light dissection. These results occurred in both in vivo and ex vivo preparations with analogous results, although in vivo provided a more “instantaneous” measurement when time courses were compared. Under pharmacological modulation, a reduction in light-evoked responses occurred when sodium channels or gap-junctions were blocked, and the response was nearly abolished by inhibiting glutamate receptors (mGluR type III, AMPA/kainate or

NMDA). In the dark, dopamine turnover was upregulated by blocking GABA receptors,

xii increasing overall retinal excitability (with high K+ solution), or activating NMDA receptors.

The use of transgenic mice models (Gnat2A518G, Opn4Cre/Cre, Opn1mwR and rd/rd) of altered photoreception revealed that rods were responsible for both, driving the suppression of dopamine release under dim red light and triggering light-induced dopamine release under bright light conditions. The contribution of ipRGCs was tested in i) rd/rd animals, ii) by chemogenetic activation in Opn4Cre/+ mice, and iii) in Opn4Cre/Cre mice. The contribution of cones was tested i) before cone degeneration in rd/rd animals, ii) in Gnat2A518G mice, and iii) by spectrally segregating the activation of different photoreceptors in Opn1mwR mice.

Surprisingly, despite the electrophysiological input from cones and ipRGCs into dopaminergic amacrine cells, these photoreceptors did not elicit widespread dopamine release in any experimental approach, and using mouse models associated with their inactivation did not show differences in dopamine release compared to wild-types. Lastly, density and morphology of microglia were characterised for the first time in a mouse model of chronic neuroinflammation,

GFAP-IL6, exhibiting increased numbers and reactive phenotype. Surprisingly, light-induced dopamine release in GFAP-IL6 animals was not different compared to wild-types.

Conclusions: This thesis provides novel evidence about dopamine dynamics in a variety of light conditions, transgenic mouse models and presence of pharmacological agents.

Surprisingly, I found that rod input is both necessary and sufficient to evoke light-induced release dopamine across a wide range of light intensities, without quantifiable contribution from cones or ipRGCs, suggesting that electrophysiological inputs do not match dopamine release. Further, this data suggests that the main pathway that drives this increase in light- induced dopamine release at light intensities where rods should be saturated is the primary rod pathway (with smaller contributions from the secondary and tertiary pathways) and involves bleaching adaptation of rods.

xiii

Publications, conferences and awards

Publications

• Pérez-Fernández, V., Harman, D. G., Morley, J. W. & Cameron, M. A. Optimized

method to quantify dopamine turnover in the mammalian retina. Analytical Chemistry

89, 12276–12283 (2017)

• Pérez-Fernández, V., Milosavljevic, N., Allen, A. E., Vessey, K. A., Jobling, A. I.,

Fletcher, E. L., Morley, J. W. & Cameron, M. A. Rod photoreceptor activation alone

defines the release of dopamine in the retina. Current Biology (under review)

Conferences

• A natural evolution for dopamine quantification. NSW Inter-University Neuroscience

and Mental Health Conference. Sydney, NSW, Australia, 2016.

• A new technique to measure dopamine release in the mammalian retina. Australasian

Neuroscience Society (ANS). Hobart, Tasmania, Australia, 2016.

• Meriva curcumin downregulates glia activation in both brain and retina in a mice model

of chronic neuroinflammation. UNSW Symposium of Neuroinflammation. Sydney,

NSW, Australia, 2017.

• A high-threshold rod input drives retinal dopamine release in response to light.

Association for Research in Vision & Ophthalmology (ARVO). Baltimore, Maryland,

United States of America, 2017.

• Determination of minimum effective dose and optimal dosing schedule for Meriva

curcumin in a mouse model of chronic neuroinflammation. NSW Inter-University

Neuroscience and Mental Health Conference. Campbelltown, NSW, Australia, 2017.

xiv

• Pathways leading to dopamine release in the mammalian retina. NSW Inter-University

Neuroscience and Mental Health Conference. Campbelltown, NSW, Australia, 2017.

• Pathways leading to dopamine release in the mammalian retina. Australasian

Neuroscience Society (ANS). Sydney, NSW, Australia, 2017.

• Effect of curcumin on retinal microglia in GFAP-IL6 mice. Association for Research

in Vision & Ophthalmology (ARVO). Honolulu, Hawaii, United States of America,

2018.

• Rod input drives, and supresses, dopamine release in the mouse retina. Association for

Research in Vision & Ophthalmology (ARVO). Honolulu, Hawaii, United States of

America, 2018.

• Light inputs to dopaminergic amacrine cells of the mammalian retina. Health Beyond

Research & Innovation Showcase. Liverpool, NSW, Australia, 2018.

• The allegory of retinal dopamine. 3-minute thesis (3MT). School of Medicine,

Campbelltown Campus, Western Sydney University, NSW, Australia, 2018.

Awards

• Retinal Neurophysiology and Tuition Fee Waiver Scholarship. Australian Government,

Department of Education, Science and Training, 2015.

• 1st place Poster Presentation Award. Health Beyond Research & Innovation Showcase.

Liverpool, NSW, Australia, 2018.

CHAPTER 1:

General Introduction

1.1. Vision

The earliest evidence of life on Earth can be traced nearly 4 billion years ago1. Since then, sunlight has been a primordial force leading the evolution of living organisms. Fossil record, embryological development and comparative genetics on extant species allows traceability from the simplest to the most sophisticated mechanisms to detect light2. Simply, light sensing is advantageous to the survival of organisms in most environments3, even in the deepest and

“darkest” places in the ocean4. Thus, the ability to respond to light is spread across all biological kingdoms. In many species light sensing occurs in the eye, a direct channel that allows the brain to interact with a portion of the electromagnetic spectrum.

The eye is often considered as an optical instrument similar to a camera, it is capable of detecting and transforming photons contained in light into interpretable biochemical signals under a wide range of operational conditions. Anatomically, the eye is mostly covered by the sclera, a protective and connective tissue that helps to maintain the shape of the organ. The cornea is a curved and transparent tissue continuous to the sclera that transmits light, and protects the iris and the pupil. The pupil is a hole by which light passes through and whose size is regulated by the iris, a set of pigmented muscles that also provides our eye colour. The crystalline lens is located beneath the iris, suspended by ligaments attached to ciliary muscles that change the shape of the lens to alter the refractive index of the eye. This allows accommodation of the eye to view objects in focus over a wide range of distances. On the posterior part of the eye lays the retina, which is nourished by the retinal pigmented epithelium

(RPE) and oxygenated by the choroid.

The retina is a multi-layered photoreceptive tissue that triggers a cascade of chemical and electrical events in response to light (or its absence) and sends the information contained in light patterns to higher areas of the brain through the optic nerve. In adult humans the retina

2 spreads ~1,200 mm2 in surface area and ~400 µm in thickness5 (~15 mm2 and ~200 µm in mice6), which covers two-thirds of the posterior inner part of the eye. Retinal circuits are built from five types of neurons: photoreceptors, bipolar cells, ganglion cells, horizontal cells and amacrine cells; and three types of glia: Müller cells, astrocytes and microglia.

The macula is a pigmented and rounded area (~20 mm2 in humans) located near the centre of the retina and responsible for central vision. This area provides great visual acuity compared to the rest of the retina that contributes to peripheral vision. The centre of the macula is called the fovea, and it consists of a depression in the inner retina with a high population of cones

(~200,000 per mm2)7 and an absence of rods8, in contrast to the rest of the retina which is rod- dominated. Cones are the photoreceptors responsible for colour, high-temporal and high- resolution vision under bright light, whereas rods specialise in low-light conditions. This strategic adaptation of the eye allows humans to exhibit high acuity day-time vision, despite possessing a nocturnal retina by definition with ~95% rods and ~5% cones5. Rodents exhibit a similar retinal composition (~97% rods; ~3% cones) but lack any foveal specialisation9.

As shown in Figure 1.1, the retina is organized in five distinct layers: three nuclear, and two synaptic layers. Bordering the innermost edge of those layers resides the RPE, which is a pigmented (melanin) cuboidal monolayer of cells essential for the maintenance and survival of the adjacent photoreceptor cell layer. The RPE performs a wide range of visual-maintaining functions, including: recycling of photoreceptor pigments, light absorption and protection against free radicals, transepithelial transport, spatial buffering of ions, production of growth factors, and immune modulation10.

In the outer nuclear layer, rods and cones (also known as classical photoreceptors) convert photons from light into synaptic outputs that can be passed to second and third order neurons of the retina. In humans, three types of cones can be found depending on the wavelength

3 sensitivity range: short (blue), medium (green) and long (red); and only one type of rods

(green). In the outer plexiform layer, photoreceptors synapse onto specific types of bipolar cells, rods generally synapse with rod bipolar cells whereas cones contact with cone bipolar cells. Horizontal cells laterally interact with photoreceptors to regulate and integrate their signals.

Bipolar cell bodies are located in the inner nuclear layer, and transmit the signals from photoreceptors to retinal ganglion cells directly, or indirectly via amacrine cells, which are interneurons that mainly project their dendrites into the inner plexiform layer and interact with bipolar cells, ganglion cells and other amacrine cells. Further, a small percentage of ganglion cells (~1%) function as photoreceptors and are named intrinsically photoresponsive retinal ganglion cells (ipRGCs).

Lastly, the axons of retinal ganglion cells form the optic nerve and send the visual information from the retina to higher areas of the brain11.

Figure 1.1. Vertical section of a human retina. Five main layers constitute the thickness of the retina, three nuclear and two synaptic or plexiform layers; from top to bottom: outer nuclear layer, outer plexiform layer, inner nuclear layer, inner plexiform layer, ganglion cell layer. Five types of neurons are discretely distributed across the retina, and the information flows vertically from classic photoreceptors (rods and cones) to bipolar cells, and then to ganglion cells. Lateral regulatory synapses occur at the outer and inner synaptic layers from horizontal and amacrine cells respectively. The retinal pigmented epithelium (RPE) is adjacent to the outermost part of the retina performing supportive functions. Image adapted from Mader, S & Windelspecht, M. Human Biology 13th edition, McGrawHill (2016).

1.2. Photoreceptors

Photoreceptors are the specialised neurons that convert photons into biochemical signals, and three different types of photoreceptors are found in the retina: rods, cones and ipRGCs.

Photoreceptors differ from each other in the visual pigment they contain, distribution across the retina, morphology and light-response properties. These variations reflect that each photoreceptor is specialised for different aspects of vision.

1.2.1. Visual pigments

Visual pigments are light-absorbing molecules that mediate vision, made of an opsin apoprotein (membrane-bound G-protein coupled receptor) covalently bound (through a Schiff- base) to a chromophore retinal molecule (normally vitamin A or retinal). Once visual pigments absorb light, an isomerisation of the chromophore from -cis to –trans (or –trans to –cis if it is a photoisomerase) triggers a conformational change in the opsin, which activates the G-protein coupled receptor and initiates a downstream signalling known as phototransduction cascade.

The chromophore retinal absorbs light at a maximum spectral sensitivity (λmax) of 380 nm when the Schiff-base linkage to the opsin is not protonated, although visual pigments can be activated at different wavelengths since the protonation state of the Schiff-base linkage change their spectral absorption properties12.

Human rods and cones possess rod opsin and cone opsins respectively, which cause these cells to have different wavelength sensitivity: 498 nm (rods), 420 nm (S-cones), 530 nm (M-cones) and 560 nm (L-cones). Mice have similar rod opsins to humans, but they only possess S-cones

(λmax = 359 nm) and M-cones (λmax = 505 nm). However both can be co-expressed within the

13–16 same cone . Melanopsin (λmax = 479 nm) is found in ipRGCs, and surprisingly for an opsin found in vertebrates, it displays several hallmarks of invertebrate opsins17. The major

6 difference between vertebrate and invertebrate opsins is that after activation by light, invertebrate opsins do not need to release the retinal from the opsin pocket. Melanopsin is a tristable18 pigment that can bind both 11-cis-retinal or all-trans-retinal, acting as light sensor or photopigment regenerator respectively. Another important aspect of melanopsin is its activation profile, distributed by light across one active (blue light) and two silent (violet and red light) signalling states. Furthermore, pre-stimulation with red light enhances responses elicited with blue light18.

1.2.2. Phototransduction cascade

The process by which a photon triggers a biochemical cascade in photoreceptors is known as phototransduction. Rods and cones are in a continuous depolarised state during darkness, continually releasing glutamate at the ribbon synapse pre-synaptic to bipolar and horizontal cells. In response to light, rods and cones enter into a hyperpolarised state by a graded change in membrane potential, causing a reduction in glutamate release that is proportional to light intensity. Rod and cone photoresponses exhibit different speeds, photosensitivity and adaptation features in any given species, yet both are stable and reproducible within species.

Phototransduction in rods and cones starts when the visual pigment absorbs a photon. Then, the opsin molecule becomes active, following a sequence of intermediaries states19 it triggers the activation of the protein transducin. Transducin is a heterotrimeric G-protein with different but similar α-subunits in rods and cones20 (the subunits are encoded by different genes, but can be interchanged artificially21). Transducin activates the enzyme phosphodiesterase, which hydrolyses free cGMP in the cytoplasm. This reduction in cGMP results in the closure of cGMP-gated channels located in the membrane, which reduces cation influx and hyperpolarises the membrane at the outer segment of the photoreceptor causing a reduction in

7 glutamate release from the synaptic terminal. An important principle present at different stages of the phototransduction cascade is amplification, which is the process by which the activation of a small number of G protein coupled receptors triggers a robust downstream signalling event.

Following light exposure, for the photoreceptor to return to the initial dark-adapted state in a process known as dark adaptation, free levels of cGMP must be increased and all-trans-retinal needs to be released from the binding pocket and isomerised back to 11-cis-retinal in the RPE, so the photoreceptor can keep the responding to further light stimuli.

Rods and cones can operate in a dynamic range of light intensities because both can modulate their sensitivity to light responses in a process known as light adaptation, which is intimately related with variations at the phototransduction cascade level. While rods are classically known to have a limited ability to adapt to high light intensities, cones exhibit an almost limitless capacity to adapt to bright light conditions. However, some authors recently argued against the popular belief of rods being functional only at scotopic light levels22–28. Furthermore, a recent paper found that after the initial reduction of rod responses caused by bright light exposure, rods slowly recover, and the more light given the more robust the rod response. The mechanisms suggested for these observations were adjustments in gain and bleaching adaptation28, explained in more detail below.

In rods, long exposures to light decrease the probability of photon absorption because some of the rod opsins are bleached, as a result cGMP channels close, Ca2+ influx is reduced, and Ca2+ dissociates from Ca2+-binding proteins necessary for the functioning of the phototransduction cascade. The intracellular decrease in Ca2+ reduces three amplification steps of the phototransduction cascade and accelerates its termination by increasing the synthesis of cGMP, leading to a partial reopening of cGMP channels, the recovery of some circulating currents of cations, and therefore recovery of the operating range of the photoreceptor under increased

8 light conditions.Cones exhibit an additional resistance to saturation because their high rate of photopigment regeneration, low amplification of the signal, and fast termination of the phototransduction cascade29.

In contrast to rods and cones, light activation of melanopsin causes depolarisation of the ipRGCs as opposed to hyperpolarisation in classical photoreceptors. The elements of the phototransduction cascade of melanopsin differ from rods and cones and closer resemble invertebrate opsins, and determination of some of the elements are still active areas of research.

Light-activated melanopsin interacts with a G protein (Gq) that in turn activates phospholipase

C, which leads to the generation of inositol triphosphate (IP3) and diacylglycerol (DAG). IP3 increases intracellular Ca2+, while DAG activates protein kinase C, and is further metabolised into arachidonic acid and other polyunsaturated fatty acids that lead to the opening of non- selective cation channels30. It has been widely speculated that these channels are transient receptor potential (TRP) due to similarities with Drosophila phototransduction, immunoreactivity in ipRGCs, and blockage of the photoresponse with TRP antagonists31–33.

Very recently, Jiang et al.34 proposed the involvement of hyperpolarisation-activated cyclic nucleotide-gated (HCN) channels in specific ipRGC subtypes. Adaptation of ipRGCs is modulated by the background light level, neuromodulators and circadian rhythms. Similar to rods and cones, light adaptation follows Weber-Fechner’s law (constancy in object perception regardless background illumination) and is partially dependent on Ca2+ 35. When ipRGCs are stimulated by bright steady light, the voltage response transiently rises before relaxing and shortening the peak time36,37, and under flickering bright light the amplitude of the response is also attenuated37. Further, ipRGCs receive negative feedback from dopamine38, adenosine39 and melatonin40, which are neuromodulators associated with light adaptation and circadian rhythms.

1.2.3. Morphology, properties and functions

Rods and cones differ in shape, type of photopigment they contain, distribution across the retina, and pattern of synaptic connections. However, both are located in the photoreceptor layer, stratify in the outer plexiform layer, and exhibit homologous structures with similar functions such as the outer segment, inner segment and synaptic terminal.

The outer segment is a structure, conical in cones and cylindrical in rods, filled with a dense pile of discs made of cell membrane that contain densely packed transmembrane visual pigments (~25,000 molecules/μm2 41). Thus, the total number of pigments can be assessed using the outer segment volume, which is larger in rods compared to cones41. Additionally, rod discs are disconnected from the outer membrane, whereas in cones, the plasma membrane of the discs and the outer membrane are continuous. This fact accounts for the morphological difference in the surface area of rods, which is larger and more efficient than cones for photon detection. In rods, the noise (spontaneous chromophore isomerisation) present during the dark conditions is very low, and therefore single events can be detected above noise threshold due to the high stability of rod opsin and the large surface area of densely packed pigments. Rod opsin pigments exhibit high thermal stability and low spontaneous activation42 compared to cone pigments43, which improves the response accuracy at very low light intensities and allows a reliable single photon detection when dark-adapted44,45. Conversely, cones exhibit faster responses due to the fluidity in trafficking proteins and ions (such as Ca2+) provided by the high surface to volume ratio in their outer segment due to the connected discs. This fluency increases the termination of the phototransduction cascade and the traffic of regenerated photopigment, allowing cones to signal much faster than rods without undergoing saturation. For example, when 90% of the visual pigment is bleached, the current amplitude of cones is only half than the one measured in the dark46. These properties make rods perfect for high sensitivity and low acuity vision, whereas cones cover colour vision, high acuity and low sensitivity scenarios.

The outer segment is connected to the inner segment through a cilium, which is a bottled-neck point where all the proteins destined to the outer segment must pass through. The inner segment is filled with mitochondria that maintain the huge demand of energy required to maintain the membrane potential, especially in the dark.

Rods and cones make ribbon synapses onto bipolar and horizontal cells, which consists of a configuration known as a triad with two lateral processes from a horizontal cell and one central process from a bipolar cell that lies in an invagination at the photoreceptor terminal. This structure allows synaptic vesicles to be densely localized and released continuously under dark conditions47. At this point, many rods converge into a single rod bipolar cell, which makes the rod system more sensitive to light detection because signals can be pooled, but at the cost of a lower resolution. Conversely, the less convergent relationship of cones to cone bipolar cells maximises acuity by sacrificing sensitivity. The most extreme example with least convergence to maximise resolution occurs in the midget system in central fovea of human or monkey retinae, where connections are as low as one to one. Additionally, rods and cones can communicate via gap junctions at the synaptic endings48 and share downstream pathways.

At the ganglion cell layer, six populations of ipRGCs (M1-M6) can be distinguished in rodents and two in primates (M1 and M249–51), based on morphology, stratification, melanopsin expression level, and electrophysiological properties52–55. In contrast with classical photoreceptors, ipRGCs can stratify at different levels in both ON and/or OFF sublamina of the inner plexiform layer. While M1 cells solely stratify in the OFF sublamina; M2, M4, M5 stratify in the ON sublamina; and M3, M6 stratify in both ON and OFF sublaminae55,56. All populations express melanopsin in soma and dendrites (~3 molecules/μm2)57 and despite being only ~900 cells in the entire mouse retina (~1% of total retinal ganglion cells), ipRGCs produce a photoreceptive net across the retina due to their large dendritic fields58 with different kinetics

11 and sensitivities among the different ipRGCs populations56 (light response properties from the different ipRGCs subtypes are summarised in Table 1.1).

As previously mentioned, ipRGCs depolarise in response to both flash and steady light, in contrast with the hyperpolarising response of rods and cones59, with a transient inward current proportional to the intensity of the flash57. Under dim steady light, ipRGCs respond with a slow integration time (~8 s) that allows summation of photons arriving at different times57 (once the light intensity increases, the small signals sum creating larger and more sustained responses).

Particularly, the M1 subtype exhibits the fastest, largest and most sensitive responses among ipRGCs populations60, probably due to their high level of melanopsin expression61,62.

Surprisingly, M1 ipRGCs often enter into depolarization block, which significantly reduces their spiking rate as the stimulus gets above a threshold. However, this does not affect their overall efficiency at coding the stimulus intensity as many subsets of M1 cells exhibit different depolarisation block light thresholds meaning the population as a whole cover a very wide range of light intensities63,64.

In addition to intrinsic responses, ipRGCs can respond extrinsically to light stimulus via rod and cone inputs65. As previously mentioned, M1 stratify in the OFF sublamina and receive strong ectopic synapses from cone63,66,67 and rod63,68 ON bipolar cells . The role of OFF bipolar cells signalling into ipRGCs is minor and only evident after pharmacological manipulation of the retina63. Further, ipRGCs synapse with different amacrine cells50,69,70, and receive

71 inhibitory inputs mediated by GABAA but not by glycine receptors . Specifically, the M1 subtype establishes a reciprocal synapse with dopaminergic amacrine cells68,70,72 and it is suggested that some of these inhibitory inputs comes from dopaminergic amacrine cells since they co-release GABA and dopamine73–77, and in turn ipRGCs activate dopaminergic amacrine cells via glutamate release. While GABA inhibition seems to reduce the depolarisation block

12 at high irradiances and creates more transient responses at low light levels63, dopamine upregulates melanopsin expression78 and is suggested to improve the sensitivity of ipRGCs

(because the more melanopsin molecules the more chances of capturing photons) during daytime (dopamine levels are higher in the daytime), although the increase in excitability is modest79. Interestingly, exclusive and artificial acute activation of ipRGCs can be assessed using chemogenetic tools such as designer receptors exclusively activated by designer drugs

(DREADDs), which have been expressed exclusively in ipRGCs and activated with a designer drug that only targets that specific receptor during dark. Activation of ipRGCs using

DREADDs resulted in a decrease in both a and b waves of the scotopic electroretinogram

(ERG), suggesting that the modulation that ipRGCs exert onto the retina affects photoreceptors and, potentially, downstream pathways80.

Intrinsic light response Extrinsic light response

● ON-predominant, small and sustained M1 Large and high sensitivity81 ● Vey weak OFF with pharmacology82 M2 Small and low sensitivity81 ON-predominant, large and sustained81 M3 Small and intermediate sensitivity83 ON-predominant, large and sustained83 M4 Small and insensitive52 Small and sustained60 M5 Small and insensitive52 Small and sustained60 M6 Small and insensitive55 ON-predominant, small and sustained55 Table 1.1. Physiological properties of ipRGCs subtypes. Adapted from Schmidt et al. 201156

The axons from ipRGCs project to several brain areas such as the suprachiasmatic nuclei (SCN) of the hypothalamus for circadian rhythms, the olivary pretectal nuclei (OPN) for pupillary light reflex (PLR), the lateral geniculate nucleus for vision forming, the superior colliculus for contrast and motion, the ventral subparaventricular zone for masking behaviour and the ventrolateral preoptic nucleus for sleep52,61,84–89.

Despite different functions attributed to ipRGCs in the recent years, circadian entrainment and

PLR are the most well-known and are mediated by the SCN and OPN respectively. Several melanopsin functions have been discovered using Opn4-/- mutants, the transgenic mouse model for melanopsin knockout that replaces the opsin with a tau-lacZ sequence. Mice show no morphological changes or number or projections of ipRGCs, but loss of photosensitivity84. The main functions of melanopsin are discussed below:

Constriction of the pupil reduces light saturation, increases the depth of field and improves resolution. Although Opn4-/- mutants retain similar PLR responses as wild-types at low irradiances, they show incomplete (~80% 90) and brief91 constriction at high light intensities.

Further, animals lacking not only the photopigment melanopsin but all the ipRGCs65,92,93 completely lack PLR responses meaning ipRGCs form the principal conduits conveying this light information to the OPN.

Daily physiological and behavioural rhythms are a ubiquitous feature of mammals, and these are controlled by a molecular-based clock located in the SCN (central peacemaker) and in some organs94. However, the SCN needs a periodical reset based on environmental cues, and light is by far the most reliable. The process by which these rhythms repeat in a period of 24 hours under a light/dark cycle is known as photoentrainment, and ipRGCs are the main cells responsible for carrying the signal into the SCN95 (rods24 and cones96 also play a role in photoentrainment signalling through ipRGCs). Retinae themselves also possess a photoentrainable circadian oscillator and its photoentrainment, surprisingly, is not dependent on rods, cones or melanopsin, but neuropsin, a UV-sensitive opsin expressed in a subset of ganglion cells97.

While melanopsin is well known for its role in other non-image forming processes such as alertness98, melatonin suppression99, sleep modulation100 or photophobia101, increasing

14 evidence suggests that ipRGCs contributes to many aspects of image-forming vision such as environmental illuminance discrimination86,87,102,103, contrast89 and adaptation of visual responses104. In addition, melanopsin photoreception has been shown to influence circuits involved in colour discrimination (in primate49), spatial information (M2-M5)60 and motion detection (M1-M4)60. Furthermore, when rod and cone phototransduction is knocked out but ipRGCs are still functional, an elementary form of pattern vision is suggested based on an improved performance in water maze tests102 and c-fos activation in visual cortex52.

1.2.4. “Earliest photoreception”

The development of the retina starts during the embryonic period and comprises several phases until retinal maturation is achieved in adult organisms. Retinal cell types originate from common multipotent progenitors, known as retinal progenitor cells (RPC). The purpose of

RPCs is to expand the tissue during development, and stop their cycle at a certain point to start differentiation towards the mature form of a retinal neuron and establish synaptic connections.

Birthdates are different for every retinal cell, generally it is as follows: retinal ganglion cells

→ horizontal cells → cones → amacrine cells → rods → bipolar cells and Müller cells. The differentiation initiates in centre of the retina and spreads to the periphery105.

Consequently, the three types of photoreceptor are born at different dates. In mice, retinal ganglion cells and cones are born at embryonic day 10 (E10), whereas rods appear at E12106.

The outer segments of rods and cones both arise around postnatal day 4-6 (~P4-6) and reach adult length ~P19 – 25107. Similarly, dendrites from ipRGCs are clearly organized ~P6 and have adult-like stratification ~P19101. Melanopsin is the first opsin expressed in the retina

~E10.5, followed by S-opsin ~P1, rod opsin ~P5, and M-opsin ~P7108. Further, ipRGCs become photosensitive before birth109 in contrast with photoreception of rods and cones ~P10110,111.

The maturation of ipRGCs involve a drop in density ~70% from P0 to P14, while the dendritic arbour size, sensitivity and sustained responses increase112,113.

Proteins found at synaptic sites are a good indicator of the structural development of the retina.

In the outer plexiform layer glutamate transporters are expressed ~P2 in cones and ~P8 in rods114, and the proteins that form ribbon synapses anchor ~P10. Even though ribbon synapses are not mature at this stage, glutamatergic transmission is active and plays a role in synaptogenesis115. mGluR6 receptors from ON bipolar cells start their development ~P5-P8116, and ON bipolar cells release glutamate to propagate retinal waves from P10-P15117–119. In the inner plexiform layer, amacrine cells develop synapses with other amacrine cells and ganglion cells prior to bipolar cells120. Around eye-opening the density of ribbon synapses in the inner plexiform layer reaches half of those present in an adult retinae121 and photoreceptor signalling becomes essential for maintaining healthy synapses with bipolar cells ~P14122, causing retinal waves to cease123. Alterations from the photoreceptor flow of neurotransmitter into postsynaptic neurons (such as in rd/rd mice where classical photoreceptors degenerate) will cause sprouting of postsynaptic neurons into the outer nuclear layer124,125. Rd/rd mice are used as a model of human retinitis pigmentosa where rods and cones start degenerating during development until they are virtually gone ~P30126. Greferath et al.127 time-lined this degeneration in a double-mutant mouse rd-FTL, where a β-galactosidase reporter system was under the promoter of the c-fos gene. In photoreceptors c-fos is expressed before programmed cell death occurs128, which allows photoreceptor loss to be tracked. Until P8 no differences between wild-type and rd mice were observed, but rods started to degenerate ~P9-P18. Rods were virtually gone ~P18 (~2% of rods remained129) and cones started to degenerate ~P19, with a maximum peak between P22 - P26. In adult mice (>P60) no trace of rods could be found and only 5% of the original population of cones was still present in the periphery129, although they are thought to be non-functional.

1.3. Retinal pathways

Once classical photoreceptors respond to light triggering the phototransduction cascade and modulating the glutamate release at the post-synaptic site, the change in the gradient of neurotransmitter is sensed by second-order neurons in the visual pathway, bipolar cells, which in turn modulate their own glutamate release to feed information into third-order neurons, ganglion cells, that connect to the brain. Conversely to this flux of vertical information, horizontal and amacrine cells transmit information laterally between neurons in the same layer.

All these fluxes of information are explained in more detail in the following sections.

1.3.1. Cone pathways

Rods and cones synapse onto bipolar cells in the outer plexiform layer. Bipolar cells are classified into rod and cone bipolar cells, depending if they make contact with rods or cones predominately. In addition, bipolar cells can be categorized in subclasses depending on their response to light: depolarising (ON) or hyperpolarising (OFF), determined by the expression of different glutamate receptors on the bipolar cells. From the 13 types of bipolar cells that have been found so far, 12 belong to ON and OFF cone bipolar cells, whereas only one ON rod bipolar cell has been found130. Therefore, after the photoreceptor level, light signals diverge in two distinct pathways: ON or OFF, which are conserved across species131. The axon terminals of cone bipolar cells stratify in the ON or OFF sublamina of the inner plexiform layer innervating either ON or OFF retinal ganglion cells respectively (except M1 ipRGCs whose dendrites stratify within the OFF sublamina and receive en passant synapses from ON bipolar cells66,132).

All ON bipolar cells express mGluRs (mainly mGluR6 subtype133) on their presynaptic processes, which are essential for communication between photoreceptors and bipolar cells.

Signals onto mGluRs are called “sign-inverting” because ON bipolar cells depolarise in the presence of light contrary to photoreceptors. This modulation of polarity is caused by an associated ion channel (thought to be TRPM1134) that is coupled to a second messenger pathway triggered upon activation of the mGluR. mGluR6 responds to 2-amino-4- phosphonobutyric acid (APB or L-AP4) completely blocking ON bipolar cell responses, which mimics the increase in glutamate release from photoreceptors in the dark135.

Conversely, signals onto iGluRs are known as “sign-preserving” because OFF bipolar cells hyperpolarise in the presence of light as photoreceptors. In contrast with mGluRs, iGluRs are ligand-gated ion channels whose aperture is directly modulated by glutamate (without downstream signalling cascade). Two kinds of iGluRs are present in the dendrites of OFF bipolar cells (with distinct patterns of expression across species136–142): α-amino-3-hydroxy-5- methyl-4-isoxazolepropionic acid (AMPA) and kainate (named due to the ability to respond to the glutamate agonists AMPA and kainate respectively), with different resensitization rates that confer preference for transient and sustained components of the visual stimulus respectively143

(but see144). 6-Cyano-7-nitroquinoxaline-2,3-dione (CNQX) is a competitive antagonist of

AMPA and kainate receptors that completely blocks the responses of OFF-bipolar cells145.

Another type of iGluR are N-methyl-D-aspartate (NMDA) receptors (named due to the ability to respond to NMDA), which are assembled by different combinations of subunits (NR1,

NR2A through NR2D, and NR3A and NR3B in mice146–148), each conferring specific properties. NMDA receptors are non-selective cation channels that are especially permeable to Ca2+ and are unique in that at resting membrane potentials Mg2+ is bound to the receptor causing the blockage of the cation channel149,150. Only following membrane depolarisation, which displaces Mg2+ from the channels, allows it to open in response to glutamate151 (with slower kinetics than AMPA and kainate receptors). Additionally, both glycine and glutamate are necessary to activate NMDA receptors containing NR2 and NR1 subunits152–154.

Interestingly, NR1 and NR2 subunits have been found in photoreceptor terminals155,156, (cone) outer segments157 and bipolar cells158–163, although NMDA-mediated currents cannot be induced164–170 (except in isolated rod bipolar cells171). This suggests that NR1 subunits serve as glutamate autoreceptors, whereas NR2 subunits might provide a pathway for Ca2+ into the cone outer segments.

The outer retinal synapses between photoreceptors and bipolar cells are modulated by lateral synapses from horizontal cells (mice only have one subtype that contacts both rods and cones172,173, but different subtypes with input preference for rods and cones exist in fish and macaque174–176), which are laterally interconnecting neurons with huge receptive fields177 (~6 dendritic fields overlap at any given point of the retina178) made of gap junctions (Cx57 subtype179). Gap junctions (or electrical synapses) are clusters of intracellular channels that allow direct diffusion of ions and small molecules (up to 1KDa) between adjacent cells and are regulated by circadian rhythmicity and light-driven neuromodulators such as adenosine, dopamine and nitric oxide180–184. These neuromodulators trigger intracellular signalling pathways, and among the most observed regulatory mechanisms, phosphorylation controls tasks related to the lifecycle of the gap junctions such as assembly, trafficking, degradation and channel aperture185,186. For instance, the modulation that light exerts on the coupling of horizontal cells187–190 starts with the light-induced release of dopamine and nitric oxide

(explained in more detail in section “1.4. Dopamine in the retina”) that diffuses throughout the retina and causing an intracellular decrease of cAMP191,192 and cGMP193–195 concentration respectively, which results in a decrease in coupling196. However, the light-dependent modulation of coupling is more complex, exhibiting an n-shape with dark and light adapted retinae showing uncoupling between horizontal cells197–199, whereas mesopic (intermediate light intensities between photopic and scotopic) conditions exhibits maximum coupling200,201.

Horizontal cells provide a gain control system that improves the acuity of signals202–204 and contrast205,206 by measuring the environmental light illumination in a determined area and subtracting a proportional part by providing inhibitory input to rods and cones. Horizontal cells hyperpolarise in response to light causing γ-aminobutyric acid (GABA) release207,208 and activating GABA receptors located in photoreceptors209, bipolar cells137 and other horizontal cells210. GABA acts as an inhibitory neurotransmitter through Cl- mediated channels, and therefore activation of GABA receptors inhibits depolarisation of the neuron where these are expressed (different GABA receptor subtypes are suggested to mediate different signal

211,212 properties, GABAA transient and GABAC sustained ). The retina uses this lateral inhibition to create spatially opposing centre-surround receptive fields in order to increase the coding efficiency of the signal at an early stage of the visual process. An additional regulatory mechanism is caused by glutamate spillover from cones that receive negative feedback from horizontal cells, which triggers the activation of a glutamate-transporter associated Cl- current in neighbouring cones that increase their membrane potential213,214.

Axon terminals of cone bipolar cells exhibit ribbon synapses215–217 with the dendrites of ganglion cells within the inner plexiform layer, which is further divided in five sublaminae based on the stratification of OFF (1st and 2nd sublaminae) or ON (3rd to 5th sublaminae) neurons

(in complex retinae where bistratified ganglion cells218–220 and bipolar cells221–223 is common, the classification only applies to monostratified neurons). Retinal ganglion cells are positive for non-NMDA 224,225 and NMDA147,170,226,227 glutamate receptor subunits, whose combination mediates the final physiological responses of ganglion cells228–234. Transient responses235 and quantal release of glutamate236–238 is associated with non-NMDA receptors, whereas sustained responses145 and regulation of glutamate overflow238–241 is associated with NMDA receptors.

These distribution suggest that NMDA receptors are often localized extra-synaptically and those postsynaptic to bipolar cells are activated by glutamate spillover240.

Similar to the lateral regulation that horizontal cells exert, amacrine cells receive glutamatergic input from bipolar cells and regulate ganglion cell excitability either directly or indirectly through bipolar cells or additional amacrine cells using GABA and/or glycine as neurotransmitters242 (in fact, amacrine cells are commonly divided in wide/medium-field

GABAergic243,244 and narrow-field glycinergic245–247 amacrine cells, which present preference for ON or OFF pathways respectively248–250). Amacrine cells express subunits of both mGluRs and iGluRs with certain preference for rod or cone signalling. For example, NR1 and NR2

NMDA subunits, and mGluR7 are exclusive to cone bipolar cells155,156,159,251. On the inhibitory side, both GABA242,252,253 and glycine254–257 receptors are expressed in a variety of amacrine cells, and ON and OFF ganglion cells. Pharmacologically inhibiting GABA and glycine (with picrotoxin and strychnine respectively) modulates ganglion cell responses and provides information about the pattern of synaptic connections: GABAergic amacrine cells tend to do feedback (to bipolar cells) or feedforward inhibition (to ganglion cells) in reciprocal inhibitory circuits258–260 whereas glycinergic amacrine cells tend to do crossover inhibition261–264

(different neuronal populations inhibit each other’s activation). Conversely, excitatory amacrine cells such as vGluT3 has been shown to contact both ON and OFF bipolar cells265 and play a role in crossover and lateral excitation266, causing an increase in glutamate release in response to light267. Additionally, acetylcholine is released from starburst amacrine cells in response to light268,269 and has been suggested to modulate responses from other amacrine cells, bipolar cells and ganglion cells through nicotinic acetylcholine receptors270–273.

1.3.2. Rod pathways

Rod signals are highly dependent on the cone pathway at different levels to ultimately feed information into ganglion cells. Thus, three different rod pathways are known depending on environmental light demands (pink arrows, Figure 1.2).

The first and most sensitive pathway consists of rods → rod bipolar cells → AII amacrine cells

→ ON or OFF cone bipolar cell → ON or OFF ganglion cell. The ribbon synapse between rod spherules and rod bipolar cells is densely filled with glutamate vesicles that excite the postsynaptic glutamate receptors located on the dendrites of the bipolar cells. Then, rod ON bipolar cell makes a glutamatergic synapse with AII amacrine cells274 (rod bipolar cells respond to continuous light with maintained depolarisation, whereas the AII amacrine cells respond with a transient depolarisation275), which in turn communicate sign-conserving signals to ON cone bipolar cells (preferentially subtype CBb7)276 through gap junctions or sign-inverting signals to OFF cone bipolar cells through glycinergic synapses274. Additionally, A17 amacrine cells strongly modulate the kinetics of the synapse between rod bipolar cells → AII amacrine cells277,278 by releasing GABA onto the rod bipolar cell terminals279,280. Further, acetylcholine has been shown to trigger GABA released from A17 amacrine cells to rod bipolar cells281.

These connections are suggested to help setting the sensitivity level over a large area of rods and rod bipolar cells282. Gap junction permeability at this level affects the signalling distribution of rod bipolar cells to ON and OFF cone pathways, as signal transmission to ON bipolar signals will be eliminated by gap junction closure but OFF bipolar signal transmission will persist. For instance, dopamine and nitric oxide are light-induced neuromodulators that decrease electrical coupling between AII amacrines283,284 and AII amacrine to cone ON bipolar cells285,286 (explained in more detail in section “1.4. Dopamine in the retina). Artificially, meclofenamic acid (MFA) is a potent (non-selective) gap junction blocker that blocks AII →

AII and AII → cone bipolar electrical synapses287, in addition to others. AII amacrine cells are

22 the most abundant subtype of amacrine cells in the retina, they can be coupled with each other

(Cx36 subtype288) or to cone bipolar cells (Cx36 and Cx45 subtypes288–291, but see292) serving as an electrical bridge for rods to utilise one of the cone pathways293 AII-AII coupling seems to synchronise signals transmitted from rods using the primary pathway and maintains the sensitivity of the signal294. Similar to horizontal cells, bright light or dark adaptation results in a decrease of coupling whereas mesopic conditions increases conductance295,296. This adaptation occurs because detection of photons in low light conditions requires an uncoupling of AII amacrine cells to prevent a diffusion of the small signals from rods (which can be down to a single photon), whereas in mesopic conditions summing small and synchronised signals reduces the background noise. Under bright light conditions limiting lateral interactions enhances high acuity cone signals as rod input to cone ON bipolar cells is removed by closure of AII → cone ON bipolar cell gap junctions285,297. AII amacrine cells exhibit both NMDA and non-NMDA receptor-mediated responses to kainate, AMPA251 and NMDA298, that can be blocked with CNQX, but not L-AP4251. AII amacrine cells are also responsive to exogenous application of GABA, glycine and tetrodotoxin (TTX)251.

The second rod pathway is ~1 log unit less sensitive than the first299 and consists of rods coupling with cones through gap junctions at their spherules and pedicles respectively (via

Cx35/36300,301), and deletion of Cx36 results in a loss of signal transmission through primary and secondary rod pathways302. These gap junctions allow rods to have access to both ON and

OFF cone bipolar cells, and therefore ON and OFF ganglion cells. This pathway is used in high scotopic conditions and when the primary pathway is saturated303.

The third pathway is ~1 log unit less sensitive than the second299, and arose after finding direct connections between rods and OFF cone bipolar cells in different mammals304–311 although only one in five rods makes this synapse306. Rods involved in these connections exhibit high electrical coupling between them306, which suggests pooling of signals and therefore decreased sensitivity of the signal. The third rod pathway is consistent with reports of OFF ganglion cells responses after blockage of the first rod pathways using L-AP4 (blocking rod → rod bipolar signal) and strychnine (blocking AII → OFF cone bipolar signal). Further, depletion of cones showed that they are not necessary for function of the tertiary pathway304, which is used in backlight object detection during dusk or dawn306.

Figure 1.2. Possible routes conveying light signals to dopaminergic amacrine cells. Rod pathways are shown with pink arrows: Rod (dark grey) signals can enter cone ON bipolar cells (yellow) via three routes, rod → rod bipolar (white) → AII amacrine (dark green) → cone ON bipolar (primary rod pathway); rod → cone (light grey) → cone ON bipolar (secondary rod pathway); or rod → cone ON bipolar (tertiary pathway). The first two pathways require gap junction connectivity (red cylinders). Intrinsically photoresponsive retinal ganglion cells (ipRGCs; dark blue), may input directly or indirectly to dopaminergic amacrine cells (DA amacrine; bright green). PRL: photoreceptor layer, OPL: outer plexiform layer, INL: inner nuclear layer, IPL: inner plexiform layer, GCL: ganglion cell layer.

1.4. Dopamine in the retina

George Barger and James Ewens synthesised dopamine for the first time in 1910 at Wellcome

Laboratories in London, England312. During the following years research in biosynthesis and physiological actions of dopamine escalated313 until a remarkable finding occurred in 1957,

Arvid Carlsson demonstrated the role of dopamine as a neurotransmitter in the nervous system314. Today, we know that dopamine is involved in a broad range of functions as a brain neurotransmitter: mood, reward, cognition, locomotion, attention and sleep315. In the retina, adaptation to presenting light conditions relies considerably on modulation of retinal pathways by dopamine. In this section, the neuromodulatory roles of dopamine in the retina and the different inputs that trigger retinal dopamine release will be discussed.

Retinal monoamine presence was first observed in 1963 (using Falck-Hillarp’s method316) between the inner plexiform layer and the inner nuclear layer317. Most likely, researchers were looking at dopaminergic amacrine cells, which later would be found in that region in many different species318. Dopaminergic amacrine cells are the only neurons in the retina capable of synthetising and releasing dopamine, although they also co-release GABA319,320 and ATP321.

Tyrosine hydroxylase (TH) is the rate limiting enzyme to produce dopamine, noradrenaline and adrenaline, and has been extensively used to label dopaminergic amacrine cells. However, two types of cells express TH: type 1 and type 2 catecholaminergic amacrine cells, which differ in soma size (type 1 have larger somas than type 2), plane of arborisation in the inner plexiform layer (type 1 arborizes in stratum 1, whereas type 2 arborizes in stratum 3), levels of TH (type

1 contain ~3.5-fold more than type 2) and cell density (type 2 is more dense than type 1)322.

From these, type 1 is thought to be dopaminergic and type 2 noradrenergic323, which was assessed using different staining strategies in mouse models discussed below:

Human placental alkaline phosphatase (PLAP), an enzyme located on the outer surface of the cell membrane, was linked to a TH promoter to allow visualisation of these cells with indolyl phosphate324. The visualisation of PLAP is correlated with the amount of TH that the cell contains, which results in type 1 but not type 2 cell staining at short incubation times325. Further characterisation can be assessed with longer incubation times, revealing the aforementioned differences between both types in morphology, plane of arborisation and cell density.

Additionally, when TH antibody was applied in TH-PLAP retinae only the same type 1 cells were immunostained in both cases324, providing an extra manner of staining type 1 cells.

Isolated TH-PLAP+ cells were shown to release dopamine319, which provides further evidence to confirm their dopaminergic profile. However, a disadvantage TH-PLAP retina staining present is that cells can only be stained and visualised in dispersed cell culture. In order to visualise dopaminergic cells in living retinae, a transgenic mouse line was generated where red fluorescent protein (RFP) was driven by the TH promoter326. In TH-RFP retinae both types of catecholaminergic amacrine cells were stained and clearly exhibited previously described differences in morphology, plane of arborisation and cell density. This mouse model is probably the most used to characterise physiological responses from dopaminergic amacrine cells in the murine retina77,327–333. In contrast, transgenic models generated with a Cre recombinase and a reporter fluorescent protein system do not label type 1 catecholaminergic cells consistently, but GABAergic, glycinergic and few ganglion cells334. Some examples of these models are TH-BAC-Cre, TH-CRE and DAT-CRE animals, which are not suitable to study retinal dopaminergic amacrine cells. In addition, the Gensat retina project (founded by the National Institutes of Health) produced various BAC transgenic mice for general use within the neuroscience community335. For instance, in Drd2-eGFP mice the promoter for dopamine receptor 2 drives the expression of green fluorescent protein (GFP). However, one drawback of this model is that the expression of this specific subtype of dopamine receptor is not

27 exclusive of dopaminergic amacrine cells (all dopamine receptors subtypes are explained in further detail later in this section), and thus additional staining (for example TH) is needed to be certain of the cell identity. Further, to perform in vivo experiments, a two-photon microscope is needed in order to visualise the GFP+ cells and avoid excitation of the retina336.

In summary, dopaminergic amacrine cells (or type 1 catecholaminergic cells) are regularly distributed across the retina in a range of 10 to 200 cells per mm2 depending on species, and their dendrites form a dense network that virtually covers the entire retina324,326,337–339. The stratification of dopaminergic amacrine cells occurs mainly in the stratum 1 (OFF) of the inner plexiform layer, but axon-like processes can be emitted to the stratum 3 (ON) of the inner plexiform layer, ganglion cell layer or into the outer plexiform layer77,326,340–345. In fact, dopaminergic amacrine cells do not make morphologically defined synapses outside stratum

166,325, which are GABAergic since there are no dopaminergic synapses in the retina due to its propagation by volume transmission.

Dopaminergic amacrine cells respond to light inputs77,181,350–359,327,360–369,330,370,333,336,346–349 and circadian rhythms371–380 in several species due to the inputs received from rods, cones and ipRGCs (whether these inputs cause actual dopamine release or not is discussed in detail in this thesis). Three subtypes of dopaminergic amacrine cells can be identified depending on their electrophysiological responses to light: insensitive, transient elevation of spike frequency and sustained elevation of spike frequency77. Additionally, dopaminergic amacrine cells fire action potentials spontaneously75,324 that can be increased with kainate, and decreased with GABA or glycine324. Dopaminergic amacrine cells specifically express NR2A and NR2B NMDA subunits, the latter being sensitive to dopamine and causing NMDA receptor hypofunction381.

Since dopaminergic amacrine cells receive inputs from GABAergic and glycinergic amacrine cells (that in turn receive synapses from OFF bipolar cells330), authors suggested that

28 dopaminergic amacrine cells are inhibited during dark, and light removes that inhibition and triggers action potentials received from ON bipolar cells and ipRGCs330,336.

Apart from light influence, fluctuations in dopamine release at different times of day are strongly related to melatonin rhythms. Melatonin is a hormone synthesised in photoreceptors and controlled by a circadian rhythm modulated by dopamine and light. Contrary to dopamine rhythmicity, melatonin levels increase under darkness and decrease in the light382–384. Despite dopaminergic amacrine cells expressing the core circadian clock genes385–388, the ability to generate circadian rhythms of dopamine metabolism is dependent on the melatonin feedback loop and retinal dopamine is not able to follow a circadian rhythm in mice incapable of synthesising melatonin374,389. However, the reverse is not true and disruption of dopamine signalling does not stop the circadian rhythm of the enzyme responsible for melatonin production in photoreceptors390.

Dopamine, acting on dopamine receptors (D1-like; D2-like), mediates multiple trophic roles and physiological adaptation mechanisms across the entire retina. The metabolic pathway for dopamine formation initiates with L-tyrosine (an essential amino acid from the diet) gaining a hydroxyl group by the action of tyrosine hydroxylase, and becoming L-3,4- dihydroxyphenylalanine (L-DOPA), which is decarboxylated by DOPA-decarboxylase, finally becoming dopamine. Once synthesised, dopamine can be stored in vesicles, further metabolised or released extracellularly. Dopamine undergoes reuptake by dopaminergic amacrine cells themselves through dopamine transporters181,374,391 and 3,4-dihydroxyphenylacetic acid (DOPAC) is one of the major intermediaries of dopamine metabolism formed by monoamine oxidase followed reuptake392. The ratio of DOPAC to dopamine content is thought to be a good indicator of dopamine turnover extensively used in the literature181,368,400–403,370,393–

399, as it is generally only formed following dopamine release181.

As previously discussed, dopamine acts through dopamine receptors, which belong to the superclass of G-protein coupled family, and can be classified in two different families: D1-like

(D1 and D5) and D2-like (D2, D3 and D4)404 that enhance or inhibit internal cAMP production respectively by modulating the activity of adenylyl cyclase405. This suggests that dopamine has differential neuromodulatory effects depending on which receptors are expressed at its target.

Using immunocytohistochemistry, dopamine receptors have been identified in all types of retinal cells (Table 1.2), and the distribution of both families of dopamine receptors throughout different cell types on the entire retina provides a global fine-tuned modulation based on the exposure and concentration of dopamine181,406. In fact, despite the confined production and release of dopamine in the inner nuclear layer, dopamine spreads by volume transmission and travels tens of micrometers across the entire retina407. Thus, both families exhibit different sensitivity thresholds depending on the distance to the dopaminergic source181, D1-like being lower (~10 nM408) than D2-like (~1 nM409).

Localisation Dopamine receptor subtype RPE D2 (in mouse410 and human411) and D5 (in cow412 and human cell line413) Rods and cones D2 (in salamander and frog414) and D4 (in mouse415) Cone bipolar, D1 (in rat416, mouse, hamster and macaque417) and D2 in dopaminergic horizontal, amacrine amacrine cells418 to auto inhibit TH activity419 and ganglion cells Müller cells D2 (in rat420) Table 1.2. Dopamine receptor subtypes across the retina.

In photoreceptors, dopamine modulates light responses, melatonin synthesis, retinomotor movements, disc shedding and gap junction conductivity421–424. Since photoreceptors possess

D2-like receptors, dopamine leads to a decrease in cAMP (observed in chicken425, rat426, mouse415,427,428 and frog429) and a decreased capacity in further responses to light due to a downregulation of the phototransduction cascade429. Overall, dopamine modulates photoreceptor physiology to favour cone activity during daytime or high light intensities181,

30 and rod enhancement at night or dim light intensities430. Post-synaptic to photoreceptors, dopamine decreases coupling among horizontal cells431–433 by either a change in gap-junction permeability434 or number of channels435, and inhibits GABA release436. In bipolar cells, dopamine plays a role in gain control mechanism437, modulates the receptive field surround of

ON cone bipolar cells 438, enhances light responses of rod-driven bipolar cells (by controlling

- a sustained Cl current via GABAC receptor modulation, with horizontal cells as the suggested source of GABA430,439) and potentiates iGluRs of OFF cone bipolar cells440. In ganglion cells, dopamine modifies the spontaneous activity of ganglion cells, centre-surround balance of their receptive field441,442, and augments firing synchronisation between ganglion cells by increasing gap junction coupling (mouse443–445, but reduced in rat446 and cat447). In addition, dopamine modulates the photoresponsive activity of ipRGCs by attenuating melanopsin-based currents53.

1.4.1. Light inputs to dopaminergic amacrine cells

Dopaminergic amacrine cells respond to light via inputs from rods, cones and ipRGCs327,329,330,333,336, although their activation (assessed with electrophysiology and c-fos staining of TH+ cells) does not always correlate with actual dopamine release368,448–450. In this section, I review the different pathways that light information uses to travel from photoreceptors to dopaminergic amacrine cells that were assessed using various techniques, and upcoming experimental chapters will address which of these pathways trigger actual dopamine release using several mouse models under different light conditions.

Rods are thought to rely on cone pathways to reach dopaminergic amacrine cells in three pathways, the primary uses AII amacrine cells as a bridge between both photoreceptor circuitries: rod → rod bipolar → AII amacrine → cone ON bipolar cell → dopaminergic amacrine cells, the secondary uses gap junctions at photoreceptor level: rod → cone → cone

31 bipolar → dopaminergic amacrine, and the tertiary consists in a direct input from rod → cone bipolar cell → dopaminergic amacrine cell. Rod bipolar cells should not be directly involved in an excitatory input because they project to stratum 5 and dopaminergic amacrine cells stratify in stratum 1 (but see451). Rod bipolar cells however are suggested to be involved indirectly by inputting into glycinergic amacrine cells (different from AII) that inhibit dopaminergic amacrine cells upon illumination452.

Zhao et al. showed that under steady or flickering dim to intermediate light conditions, rods provide an excitatory electrophysiological input into dopaminergic amacrine cells mainly through pathways that involve gap junctions (primary and secondary), and that rods alone are responsible for stimulation of dopaminergic amacrine cells at low light intensities333 (with similar responses in both dorsal and ventral retina). Unfortunately, Zhao and colleagues did not test rods inputs at bright light intensities because for their preparations the RPE was removed and this would cause irreversible bleaching of rod opsins resulting in an experimental bias

(although months later, Tikidji-Hamburyan et al. recorded rod responses under bright light without the RPE attached to the retina28). Nonetheless, they suggest that rods could signal to some dopaminergic amacrine cells via the tertiary pathway at high light intensities. Indeed, the hypothesis of rods driving the majority of light-induced dopamine release was tested in Gnat1-

/- mouse model, that lack functional rods (although a small response is retained453–455). Those mice exhibited ~3-fold lower (although still significant) increase in dopamine turnover compared to wild-type mice in response to bright steady light, and activated ~5% of dopaminergic amacrine cells in contrast with ~100% in light-pulsed wild-type animals

(determined by c-fos staining)368. Additionally, Gnat1-/- animals also showed a decreased dopamine turnover overall during both day and night399. These results together suggest that rods are a main contributor to actual dopamine release even at bright light intensities.

Under intermediate and bright light conditions (both steady and flickering), cones provide an excitatory electrophysiological input into dopaminergic amacrine cells (with more pronounced responses in the dorsal retina) via cone ON bipolar cells → dopaminergic amacrine cells66

(which in fact is a reciprocal synapse with release of glutamate and GABA respectively that creates a balance of excitatory and inhibitory signals to potentially inhibit an excessive input345) or indirectly through ipRGCs330 (with more pronounced responses in the dorsal retina). Further, two inhibitory pathways are suggested to be cone-driven, an ON inhibition at light onset: cone

→ OFF bipolar cell → glycinergic amacrine cell → dopaminergic amacrine cell (mediated by

AMPA receptors) and an OFF inhibition at light offset: cone → OFF bipolar cell →

GABAergic/glycinergic amacrine cell → dopaminergic amacrine cell (mediated by kainate receptors). Indeed, Qiao et al. generated a “cone-only” mouse model (double knock-out Gnat1-

/- and Opn4-/-) and found that after light pulsing, cones evoke three different electrophysiological responses in dopaminergic amacrine cells: one ON excitatory current

(mediated by cone ON bipolar cells and ipRGCs), and an OFF and a delayed ON inhibitory currents (mediated by OFF bipolar cells; Figure 1.3). From those, ~70% of the tested dopaminergic amacrine cells exhibited exclusively ON responses, ~20% ON and OFF, and

~10% all three. However, these “cone-only” mice only caused ~1.2-fold increase in DOPAC: dopamine ratio after bright light exposure, in agreement with the results from Gnat1-/- mouse previously discussed330,368 and in contrast with ~3.5-fold increase in wild-type mice368. These results suggest that cones provide a balanced combination of excitatory and inhibitory inputs to dopaminergic amacrine cells, and that they contribute to dopamine release (at intermediate to bright light intensities) much less than rods. Interestingly, spiking from dopaminergic amacrine cells decline rapidly at the onset of a strongly depolarising synaptic input (caused by either steady or low rate flicker) and dopaminergic amacrine cells enter into depolarisation block, which is relieved when the depolarising input decays336.

Figure 1.3. Proposed cone pathways conveying cone signals to dopaminergic amacrine cells. Excitatory ON responses are mediated by ON bipolar cells directly (red) and ipRGCs (violet). OFF inhibition is mediated by type 2 and 3a OFF bipolar cells through GABAergic/glycinergic amacrine cells (cyan), whereas ON inhibition (d-ON response) is mediated by type 1 OFF bipolar cells via glycinergic amacrine cells (orange). a: sublamina a / OFF layer; b: sublamina b / ON layer; AC: amacrine cell; C: cone; DA: dopaminergic amacrine cell; GC: ganglion cell; GCL: ganglion cell layer; INL: inner nuclear layer; IPL: inner plexiform layer; ipRGC: intrinsically photoresponsive retinal ganglion cell; ONL: outer nuclear layer; OPL: outer plexiform layer. Image adapted from Qiao et al. Multiple cone pathways are involved in photic regulation of retinal dopamine. Scientific Reports (2016).

As previously mentioned, morphological evidence suggests that M1 ipRGCs and dopaminergic amacrine cells exhibit reciprocal synapses and these connections are glutamatergic from ipRGCs456 and act on AMPA/kainate receptors of dopaminergic amacrine cells457, whereas these in return inhibit ipRGCs458. The intrinsic electrophysiological input from ipRGCs to dopamine amacrine cells is independent from rods or cones signals327,328,330,333,339,458. Indeed, dopaminergic amacrine cells exhibited a light-induced robust and sustained response in rd1 animals327 mediated by AMPA and NMDA currents331. Signalling from ipRGCs provides sustained responses in dopaminergic amacrine cells of wild-type mouse in presence of L-AP4

(which blocks rod and cone ON pathways)327 and rd/rd animals (which lack rods or cones, but intact ipRGCs are preserved)329. However, when assessing actual dopamine release in Opn4-/- mice, those exhibit similar dopamine turnover levels as wild-type368. Further, rd/rd cl mice are not able to release dopamine in response to the same bright light stimulus as the one received by Opn4-/- and wild-type mice368. These results suggest that 1) dopamine release does not correlate with dopaminergic amacrine cells activation assessed via nuclear c-fos (although c- fos+ dopaminergic amacrine cells are reduced in rd/rd animals compared to wild-type) or depolarisation of the soma via electrophysiology, and 2) ipRGC signalling does not seem to be sufficient to elicit light-induced dopamine release despite the clear electrophysiological inputs.

In summary, responses from dopaminergic amacrine cells are the consequence of the integration of several intensity-dependent stimuli originating in the three types of retinal photoreceptors that result in a balance of excitatory and inhibitory inputs. Surprisingly however, actual light-induced dopamine release is mainly dependent on rod inputs with minor contribution from cones and/or ipRGCs, and occurs towards the photopic range. Table 1.3 summarises the current understanding of the pharmacological modulation that common drugs exert on retinal pathways that input into dopaminergic amacrine cells.

Drug Effects on dopaminergic amacrine cells AP5 Reduces ipRGC-based responses331 (NMDA antagonist) CNQX Blocks ipRGC, ON bipolar cell328 and OFF bipolar cell inputs330 and thus, (AMPA/kainate light-induced transient responses in wild-type324,328 and rd/rd animals327 antagonist) DACs themselves are inhibited by DA via D2R398,414,417,459. Attenuates rod inputs by uncoupling AII amacrine cells283 and photoreceptors398, and Dopamine attenuates DAC responses via reduction of melanopsin-based photocurrent38 and NMDA-gated currents381. Induces GABA release from amacrine cells that can be reversed with NMDA dopamine460,461 and increases AII gap junction connectivity462. Causes DAC depolarisation331 Hyperpolarises DAC324 and (muscimol, a GABA agonist) inhibits light- GABA induced DA release356,463. Co-released GABA from DACs in natural conditions, prevents excessive input from ON bipolar cells319,320.

Glycine Hyperpolarises DAC324 and reduces light-induced DA release464

L-AP4 Abolishes light-induced transient responses309,328 and DA release362,364 (mGluR6 agonist) MFA Blocks AII electrical synapses and prevents rod signals into DACs287, in (gap junction blocker) Cx36-/- mouse, DACs exhibit reduced responses to scotopic light333 Picrotoxin Stimulates TH activity in dark and has no effect on active TH. Increases

188,350,354,356,362 362350,465 (GABAA/C antagonist) DA turnover , the effects can be reversed with L-AP4 Abolishes hyperpolarisation triggered by glycine324 and suppresses the Strychnine OFF response of DACs330, increases DOPAC in dark (but see466) but not (Glycine R antagonist) additionally in light467 Blocks ipRGC direct and indirect (via ON bipolar cells) inputs into TTX DACs329, DAC themselves324,457 and reduces responses of ~50% DAC (Na-channel blocker) (the ones that receive input generated from ipRGCs)330 Table 1.3. Pharmacological modulation of dopamine release. AP5: (2R)-amino-5-phosphonovaleric acid; (2R)-amino-5-phosphonopentanoate, CNQX: 6-cyano-7-nitroquinoxaline-2,3-dione, D2R: Dopamine subtype 2 receptor, DA: dopamine, DAC: dopaminergic amacrine cell, GABA: γ- aminobutyric acid, ipRGCs: intrinsically photoresponsive retinal ganglion cells, L-AP4: L-2-amino-4- phosphonobutyric acid, MFA: meclofenamic acid, TH: tyrosine hydroxylase, TTX: tetrodotoxin

1.5. Aims

In the retina, dopamine is exclusively synthesised and released by dopaminergic amacrine cells and plays multiple roles in physiological functions such as light adaptation, circadian rhythms, development and cell survival. Dopaminergic amacrine cells are activated in response to light because of the inputs they receive from the three types of photoreceptors: rods, cones and ipRGCs. However, given the discrepancy in the literature between electrophysiological and/or nuclear activation of dopaminergic amacrine cells and actual release of the neurotransmitter dopamine, a systematic approach clarifying this disagreement seems necessary. Therefore, the work undertaken in this thesis aimed to:

1) Optimise an analytical protocol to reliably quantify dopamine and its primary

metabolite DOPAC in the murine retinae.

2) Determine the conditions that trigger light-induced dopamine release in the retina.

3) Determine the photoreceptive origin (rods, cones or ipRGCs) and pathways employed

of the inputs that trigger light-induced dopamine release in the retina.

CHAPTER 2:

Materials and Methods

2.1. Animals

All procedures were conducted in accordance with the Australian Code for the Care and Use of Animals for Scientific Purposes and were approved and monitored by the Western Sydney

University Animal Care and Ethic Committee under protocols A10396, A11900 and A12402.

Mice were bred at our animal facility (Campbelltown Campus, Western Sydney University,

Australia) and maintained under a 12 h light (~300 lux): 12 h dark cycle. Males and females ≥

60 days (unless specified otherwise) were used. Prior to experiments and with the purpose to dark-adapt retinae, mice were kept in complete darkness ≥ 18 h for CT6 (midday) or ≥ 6 h for

CT18 (midnight). Several mouse genotypes were used: wild-type C57BL/6J (Animal

Resources Centre, Canning Vale, Australia), Gnat2A518G (University of Melbourne,

Australia468; backcrossed to C57BL/6J), rd/rd (Animal Resources Centre, Canning Vale,

Australia), Opn4Cre/Cre (Johns Hopkins University52, USA), Opn1mwR (Dr. Jeremy Nathans469) and GFAP-IL6 (University of Sydney470, Australia). Mouse models are explained in more detail below:

C57BL/J6 animals are used as wild-type control in this project due to significant prior literature using them as a background strain for transgenic lines. However, these mice exhibit non detectable melatonin levels471, an important consideration when assessing retinal circadian rhythms because its absence affects both the regulation of retinal rhythmicity and inhibition of dopamine389.

Gnat2A518G mice are used as model for cone photoreceptor inactivation since they have a natural missense mutation (due to a single base pair substitution of adenine for guanine), that causes misfolding of the cone specific α-transducin subunit468. Indeed, ERGs from these animals show normal rod responses and no appreciable cone responses468. A reduction in the total number of cones occurs at 3 months of age compared to wild-type controls, accompanied by a

39 progressive remodelling at bipolar and horizontal cell synapses468. While more models for cone inactivation are available (such as Gnat2cpfl3, cpfl1 or CNGA3-/-), these suffer too much retinal degeneration or still exhibit ERG cone responses at young ages472–474. rd/rd is a mouse model used to study retinitis pigmentosa in humans. These animals possess a natural nonsense mutation in the rod phosphodiesterase475 that leads to rod and then cone degeneration within a month. Moreover, these animals suffer one of the most severe remodelling among retinal degenerate models476. After photoreceptor degeneration, the remaining cells in the retina (horizontal, amacrine, bipolar and retinal ganglion cells) form a tangle where new synapses are formed. Additionally, Müller cells and RPE migrate and disrupt the lamination of both plexiform layer477.

Opn4Cre/Cre is a transgenic mouse model of melanopsin inactivation created by expressing Cre recombinase in the place of melanopsin gene (OPN4)52. The ipRGCs of these animals do not show changes in number, morphology or projections478, but these cannot respond intrinsically to light. On the other hand, the activity of rods and cones remains unaltered.

Opn1mwR is a transgenic mouse model where the opsin of M-cones has been substituted by the one in human L-cones469, resulting in a ~45 nm red shift in spectral sensitivity. Other than that, the retinae of these animals seems as functional as wild-type animals with robust ERG responses, indistinguishable patterns of immunostained cones, and ganglion cell responses originated in “red cones”469.

GFAP-IL6 is a transgenic mouse model whose interleukin 6 (IL-6) is under the regulatory control of the glial fibrillary acidic protein (GFAP)479. This means that IL-6 is constitutively expressed in the central nervous system, causing severe neurologic symptoms and advanced neurodegeneration due to a lifelong inflammatory responses from microglia and astrocytes480.

2.2. Dopamine release experiments

2.2.1. Light sources

Dim red light (>650 nm) from a head torch or infrared light (>800 nm) from night-vision goggles were used for injections and dissections under dark conditions. A LED lamp (EDGE

SAD Lightbox, Litebook) was used for light pulsing, and different light intensities were achieved by adding 0.9 log unit neutral density (ND) film (Lee) to the lamp. For specific wavelength pulses, another LED lamp (MCWHLP1, Thor Labs) was coupled with a red (624 nm) or blue (434 nm) filter. All the light intensities used are summarised in Table 2.1, and the spectrograms presented in Figure 2.1. The measurements of the light source were conducted using different methodologies:

A photometer (L202 PMS, Macam) measured the light intensities displayed in lux units, and a spectrometer (USB2000+XR1-ES, Ocean Optics) measured the light spectrum. Effective photons were calculated using the relative photopigment absorption spectra by Govardovskii481 multiplied by the spectrum of the light source and the specific energy (in µW/cm2) of a single photon at each wavelength. Rod isomerisations were assessed by converting effective rod photons to corneal and retinal irradiance based on Lyubarsky calculations482, which included corrections that considered the wavelengths transmitted through the ocular tissues483, and the size of a dilated pupil with mydriatics (3.2 mm2). Corrections applied to the ex vivo experiments were conducted considering corneal and lens absence (since the retina is facing the light source directly), and the aperture of the pupil was considered to be the diameter of the eye, 8 mm2.

Effective photons are used in this thesis so that activation can be compared between photoreceptors. For publications, however, I have used Rh*/rod/s.

Light source + filter type W/m2 Lux Log rod photons cm-2 s-1 Log R*rod-1s-1 LED (max. intensity) 36.14 10369.21 15.99 7.34 LED + 1 0.9 ND Filter 4.54 1305.40 15.09 6.44 LED + 2 0.9 ND Filters 0.57 164.34 14.19 5.54 LED + 3 0.9 ND Filters 0.07 20.68 13.29 4.64 Head torch (>650 nm) 0.05 0.2 8.79 0.01 Infrared goggles (>800 nm) 0.6 <0.001 5.82 -2.47 Table 2.1. Intensity of light sources expressed in different units.

1 White light (LED) Red light (624 nm) 0,8 Blue liglightht (434(420 nm)

0,6

0,4

0,2 Normalised intensity Normalised

0 300 350 400 450 500 550 600 650 700 750 800

Wavelength (nm)

Figure 2.1 Light sources relative spectrograms. Normalised spectrograph of the raw LED light source (grey) or with the 434 nm filter (blue) or 624 nm filter (red).

1 L-cones M-cones S-cones 0,8 Rods Melanopsin 0,6

0,4

0,2 Normalised Sensitivity Normalised

0 300 350 400 450 500 550 600 650 700

Wavelength (nm)

Figure 2.2 Opsin nomograms of the murine retina. L-cones (which belong to human and are present mwR only in the genetically engineered Opn1 mice) in red (λmax= 556 nm), M-cones in green (λmax= 508 nm), S-cones in purple (λmax= 360 nm), rods in black (λmax= 498 nm) and melanopsin in blue (λmax= 480 nm).

2.2.2. Light pulsing, sample preparation and analysis

For in vivo experiments, mice were handled under “dark” conditions (either dim red or infrared), anaesthetised with an intraperitoneal injection of 70 mg/kg Ketamine + 7 mg/kg

Xylazine (Troy Laboratories), and their eyes dilated with Tropicamide 1% (Alcon) before light pulsing. Light pulses were conducted with the light source ~1.5 cm above the animal (meaning retinal illuminance was likely brighter in the ventral region). Unless specified otherwise, light pulses were 15 min in duration and maximum intensity (see Table 2.1). After light pulsing, mice were sacrificed by cervical dislocation, eyes were enucleated (under the same light conditions) and retinae extracted through a lateral incision along the ora serrata by adding pressure to both sides of the cut. Each retina was put into a 0.5 mL plastic tube and immediately frozen in liquid nitrogen (except postnatal day 0 and 7 animals where the whole eye was placed in the tube due to technical difficulties of extracting the retina). For dark-adapted mice, this process was conducted under dim red or under infrared light conditions (specified in each experiment). The tubes containing the retinae were stored at -80 C until analysed.

For ex vivo experiments, after adaptation to the dark, mice were sacrificed by cervical dislocation, their eyes enucleated, and the cornea and lens removed under dim red light conditions. The remaining eyecup was placed, with the retina facing up, into a 0.5 mL plastic tube with 40 µL of oxygen-bubbled, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

(HEPES)-buffered artificial cerebrospinal fluid (aCSF) at room temperature, and either light- pulsed or kept in the dark. The supernatant aCSF was then removed and frozen in liquid nitrogen and stored at -80 C until analysed.

Retinal homogenisation of in vivo experiments was completed with a pellet pestle cordless motor inside a 40 µL solution consisting of: 1% formic acid (Sigma-Aldrich), 0.5 mM ascorbic acid (Sigma-Aldrich) and 10 ng/mL the deuterated standards dopamine-d4 (CDN Isotopes) and

DOPAC-d5 (Cambridge Isotope Laboratories). For the ex vivo vials, media was supplied until

1% formic acid, 0.5 mM ascorbic acid, 10 ng/mL deuterated standards and drug (specified in each experiment) concentrations were reached. Then, for both in vivo and ex vivo preparations, samples were centrifuged at 14,000 r.p.m. for 10 min at 4 C and the supernatant placed in clear glass total recovery vials (Waters) for chromatographic analysis.

Mass spectrometry was performed using a Waters Xevo TQ-MS triple quadrupole mass spectrometer, fitted with an electrospray ionization source. The desolvation gas flow (nitrogen) was 800 L/h, desolvation temperature was 450 °C, cone gas 0 L/h, and collision gas (argon) flow of 0.15 mL/min, which gave a collision cell pressure of 2.6 × 10−3 mbar. In positive ion mode the capillary voltage was set at 1.2 kV and in negative mode 1.0 kV. Liquid chromatography was performed using a Waters Acquity UPLC, working at a flow rate of 0.20 mL/min. A Waters Acquity UPLC BEH C18 column of 1.7 μm particle size and dimensions

2.1 × 150 mm was used, operating at 40 °C. Solvent A consisted of 0.1% formic acid in ultrapure water and solvent B was 0.1% formic acid in methanol. A 20 min run was employed, commencing at 5% B for 1 min, increasing linearly to 100% B by 10 min, then returning immediately to 5% B at 15 min. The sample manager was kept at 4 °C and injections of 10 μL were made in full loop mode. Retention times for dopamine and dopamine-d4 were 2.1 and

2.04 min respectively and were analysed in positive ion mode. DOPAC and DOPAC-d5 were

5.09 and 5.01 min respectively and were analysed in negative ion mode.

Dopamine only (ex vivo) or dopamine and DOPAC (in vivo) for each retina were quantified using MassLynx software. Dopamine per retina (ex vivo) or DOPAC: dopamine ratio (in vivo) are expressed as mean ± S.E.M. in graphical data.

2.2.3. Viral injections (Methods conducted by Dr. Morven A. Cameron)

Intravitreal injections of AAV2-hSyn-DIO-hM3Dq-mCherry vector (2.3 x 1013 genomic particles/mL; The UNC Vector Core) were performed as previously reported80 and hyaluronan lyase and heparinase III (200 U each) was employed to maximise retinal penetration (total volume of 2.5 µL injected over 1 min). Control eyes underwent the same procedure, with injections including glycosydic enzymes but with the virus replaced by phosphate saline buffer

(PBS) as vehicle. Mice were allowed at least 6 weeks to recover before being used in in vivo studies.

For dopamine release assessment Opn4Cre/+ mice expressing hM3Dq in the right eye only, were dark-adapted overnight. Clozapine N-oxide (CNO, 5mg/kg) was administered intraperitoneally under dim red light and animals were then kept for 1 h in the dark. Both retinae were removed under dim red light and the left eye served as a control for the hM3Dq expressing eye. Pupil constriction in response to CNO was quantified 2 weeks before using the same methods and visualised using an infrared webcam (DCS-930L; D-Link) illuminated by a custom-made infrared LED array (> 950 nm). Animals with pupil constriction of less than 80% in response to CNO injection were excluded.

2.2.4. Statistical Analysis

All graphs and statistical analysis were performed with GraphPad Prism 7. Significance was defined as p-value smaller than 0.05. In vivo, both eyes were analysed separately and the average of the two was taken for each biological replicate reported. In contrast, for ex vivo experiments each retina was treated as an individual replicate as, once removed from the animal, retinae could be separated and subjected to different light conditions.

Student’s t-test was used to compare dopamine release between the two different “dark” conditions (dim red and infrared), between dark-adapted and light-pulsed mice at different development points, and between pharmacological treatments and controls. One-way ANOVA with Tukey’s post-test was used to test overall differences in the same light conditions across different genotypes. One-way ANOVA with Dunnett’s post-test was used to compare a range of increasing light intensities or exposures to dark-adapted controls.

To find the threshold of light-induced dopamine release across genotypes, irradiance response curves were fit to the data for each genotype using the following equation:

퐵표푡푡표푚 + (푇표푝 − 퐵표푡푡표푚) 푌 = 1 + 10(log 퐸퐶50−푋) 퐻푖푙푙 푆푙표푝푒

Where:

Y = Mean of dark-adapted value + 2 standard deviations

Bottom = Mean of lowest value

Top = Mean of highest value

EC50 = X where the Hill Slope is half

Hill Slope = Slope of the sigmoidal curve that fitted the data

Threshold was defined as the point at which the curve exceeds the mean of the “dark” condition

(dim red) + 2 standard deviations for the corresponding genotype.

To test differences across fitted sigmoidal curves, Extra sum-of squares F test was applied.

Significant differences are expressed with the p-value and the F value in the graphs.

2.3. Microglia characterisation

2.3.1. Immunohistochemistry

Mice were transcardially perfused with 4% paraformaldehyde (Merck Millipore) in 0.1 M PBS

(Sigma-Aldrich); eyes were enucleated and stored in 0.1 M PBS until used. Wholemount retinae were extracted and blocked in a 96-well plate with 5% donkey serum (Sigma-Aldrich) in a 0.1 M PBS and 3% Triton-X (Sigma-Aldrich) solution for 45 minutes at room temperature.

Retinae were placed in a solution containing 1:1000 rabbit-anti-Iba-1 primary antibody

(Wako), 1 % donkey serum, 0.1 M PBS and 3% Triton-X solution overnight at 4 C. Retinae were washed with 0.1 M PBS three times for 15 minutes, and transferred to a solution containing 1:200 tetramethylrhodamine (TRITC) secondary antibody (Jackson

ImmunoResearch), 2% donkey serum, 0.1 M PBS and 3% Triton for 90 minutes, washed with

0.1 M PBS three times for 15 minutes. Retinae were mounted in vectashield and maintained in a dark-proof box at 4 C until imaged.

2.3.2. Cell Counting and 3D Reconstruction

Mounted retinae were examined using a Zeiss ApoTome microscope. The number of Iba-1+ cells counted through the entire thickness of the tissue was divided by the area of the retina

(calculated by contouring the whole retina at 10X magnification) to obtain a density value.

For reconstruction, Z-stack images of 1 µm thickness/stack were taken in a Zeiss LSM 5 confocal microscope. Images were examined using Neurolucida 360 software and the processes were reconstructed using directional kernels interaction mode (Figure 2.3). Nuclei were drawn in two-dimensions.

2.3.3. Statistical Analysis

All graphs and statistical analysis were performed with GraphPad Prism 7. Significance was defined as p-value smaller than 0.05.

One-way ANOVA with Tukey’s post-test was used to test the following morphological differences of microglia between ages (4 or 24 months) and genotypes (wild-type or GFAP-

IL6): cell density, soma area, soma perimeter, convex area, convex perimeter, dendritic length, primary dendrites, branching points and shift in maximum complexity in relation with the soma. The latter was assessed by comparing the coordinates of the highest point of the curve, obtained from Sholl Analysis.

Sholl Analysis consists in drawing concentrical circles on the top of a reconstructed microglial cell with increasing radius of 3 µm starting in the centre of the soma (Figure 2.3 C) and plotting the size of the radius in the X axis and the different morphological variables (interactions, branching points, length, area, volume and endings) confined between that circle and the next in the Y axis.

Bivariate correlation of soma area, soma perimeter, convex perimeter, dendritic length, primary dendrites and branching points were plotted against convex area484. Correlation was tested using Pearson r test, and the slope of the linear regression compared among groups using extra sum of squares F test.

A B C

25 µm

Figure 2.3 3D reconstruction and analysis using Neurolucida 360 software. A) Z-stack confocal image of a single Iba1+ cell (red) at 40X magnification. B) Reconstruction of the dendritic arbour using directional kernels interaction mode (yellow). C) Sholl analysis consists of drawing concentrical circles (red) with a fixed increasing radius on top of the reconstructed dendritic arbour of the cell

CHAPTER 3:

Optimisation of a method to accurately quantify

dopamine turnover in the retina

3.1. INTRODUCTION

Several methods have been used to determine the activity of dopaminergic amacrine cells, however most of them were based on indirect measurements such as tyrosine hydroxylase or c-fos staining, which do not always correlate with actual dopamine release368. Thus, a quantitative method for measuring dopamine and its metabolites appeared: high performance liquid chromatography coupled with electrochemical detection (HPLC-ECD)485, which is still widely used. Since the main aim of this thesis was to comprehensively describe the release of dopamine in the mouse retina, a reliable and sensitive analytical method for measuring dopamine turnover needed to be developed. In contrast with HPLC, ultra-high performance liquid chromatography (UHPLC) operates under higher pressure (which allows better chromatographic separation) and mass spectrometry tandem (MS/MS) detects analytes of interest based on the molecular weight and fragmentation rather than the electrochemical properties of the molecule (which might be similar among different analytes). Overall,

UHPLC-MS/MS is faster, more sensitive and more selective.

The following paper comprehensively describes optimisation of each step during the sample preparation and analysis of dopamine and DOPAC in the mouse retina. The section “3.2.

Extended results” provides further results not included in the original paper relating to circadian changes in dopamine turnover, comparison between different dissection techniques and the impact of ethylenediaminetetraacetic acid (EDTA) addition to the sample before analysis.

WESTERN SYDNEY UNIVERSITY RESEARCHDIRECT NOTE: The following Supporting Information has been redacted from this version of the thesis document but is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.7b03216.

3.2. Extended RESULTS

Most of the information necessary to perform an understandable and reliable quantification of dopamine release is already explained in great detail the article. However, this extended section of the results provides even more fine-tuned characterisation of the elements that might impact dopamine quantification, such as the method of extracting the retina, addition of EDTA to the sample to prevent enzyme activation or influences of different circadian times.

First, I examined the best method for extracting retinal tissue, based on the maximum number of counts provided by the integration peak of both dopamine and DOPAC (Figure 3.1). This was done in two ways: 1) “popping” out the retina through a lateral cut along the ora serrata and squeezing the eye with the fingertips, and 2) dissecting the retina under the scope by first piercing the eyeball on the ora serrata, then cutting along the ora serrata with a pair of small surgical scissors, and finally removing the cornea and the lens, and detaching the RPE with a pair of surgical forceps. The “popping” method had increased counts for both dopamine and

DOPAC compared to retinae dissected under the scope (Figure 3.1). Since dopamine and

DOPAC oxidise rapidly under environmental oxygen without antioxidants (as previously shown in the article), it is likely that the method that involves the least time from extraction to snap freezing (“popping”) provides the best method to preserve these analytes.

Since EDTA has been previously used to stop enzyme activation in retinal samples (Table 3.2),

I examined the effect of adding EDTA in the homogenised sample containing each retina.

However, the use of EDTA is not entirely justified because the addition of the acid (either formic or perchloric) lowers the pH to a range where retinal enzymes are not active. Further, the addition of EDTA (or any salt) into samples analysed via UHPLC-MS/MS is not recommended because salts such as EDTA precipitate inside the conduits of the UHPLC instrument and remain there interacting with the mobile phase and further incoming samples

64 increasing ion suppression. Indeed, addition of EDTA to dopamine and DOPAC standards diluted in 0.1M perchloric acid and 0.01% of sodium metabisulfite did not increase the counts of integrated peak areas. In fact, dopamine counts were significantly lower compared to the solution of perchloric acid and sodium metabisulfite without EDTA (Figure 3.2).

Dopamine is released in a circadian manner due to the interaction with melatonin, with more dopamine being released during the day and less dopamine being released during the night.

Even though C57BL/6 mice are genetically incapable of synthetising melatonin, they do show significant circadian rhythmicity in retinal function368,389 meaning the light input to dopaminergic amacrine cells may undergo circadian modulation. Indeed, Cameron et al.368 reported more dopamine release at CT18 than CT6 in free-moving animals. While I showed in the article that anaesthesia did not alter the concentration of dopamine nor DOPAC, under dark or light conditions at CT6, I extended the analysis to examine dopamine release at subjective midnight (CT18). Surprisingly, dopamine turnover ratios (both dark-adapted and light-pulsed) were not statistically different when comparing CT6 to CT18 in either free-moving or anaesthetised mice (Figure 3.3). However, a trend of more dopamine released at CT18 than

CT6 in free-moving mice was abolished after application of anaesthesia, which provides a consistent reduction in the standard deviation of every measurement. This may be explained due to the animals’ behavioural circadian rhythms, presumably spending more time with their eyes open during their active phase (~CT12 - CT23) and not because of an actual increase in dopamine turnover in response to the same light stimulus. Therefore, it is likely that anaesthesia standardises the amount of light reaching the mice between circadian time points.

Figure 3.1 “Popping” the retina from the eye-cup provides increased dopamine and DOPAC counts instead of scope dissection. Under dim red light, A) dopamine absolute peak integration counts are significantly increased when “popping” is used over conventional dissection scope. B) DOPAC counts are also increased by using “popping” over scope dissection (t-test, **p < 0.01, mean ± SEM, n = 12 each).

Figure 3.2 Both sodium metabisulfite and EDTA cause a significant reduction in dopamine counts. A) Absolute peak integration counts of standard dopamine (100 ng/mL) dissolved in 0.1M perchloric acid show significantly reduced signal when 0.01% sodium metabisulfite (SMBS; black stripes) alone, or in combination with 0.5mM EDTA (black and grey stripes) are added. B) DOPAC standard (100 ng/mL) dissolved in perchloric acid do not exhibit any difference compared to the addition of SMBS or EDTA (one-way ANOVA with Tukey’s post-test, ***p < 0.0001, mean ± SEM, n ≥ 6 each group).

Figure 3.3 Anaesthesia reduces behavioural variation. Free-moving (A) and anaesthetised (B) mice exhibit significant light-induced dopamine release at both CT6 and CT18. No statistical differences are observed in light-pulses between CT6 and CT18 (t-test, p ≥ 0.22) in either condition. However, a trend of increased dopamine turnover is observed in free-moving mice at CT18, which is eliminated by anaesthesia (one-way ANOVA with Tukey’s post-test, ***p < 0.0001, mean ± SEM, n ≥ 6 each group).

3.3. Extended DISCUSSION

Measurement of retinal dopamine and DOPAC over the years has varied greatly from lab to lab. To illustrate the extent of the variety of analytical techniques, I took recent papers (~5 years old on average) that measured both DOPAC and dopamine and summarised the animal methods employed in Table 3.1. Most of the steps gathered in Table 3.1 are shown in this chapter to have an impact on dopamine turnover measurements, which makes comparisons between labs problematic. In addition, most of the papers report DOPAC: dopamine levels under extended environmental conditions, while others specifically compare dark-adapted vs light-pulsed retinae. For example, in DOPAC: dopamine column, basal values range from 0.02 to 1.2, with increases from 2 to 6-fold when comparing dark-adapted against light-pulsed animals. Also, it is difficult to reach a consensus on which light is better to use during the

“dark” dissections (dim red or infrared), as well as the collection method (retinal dissection under the scope, whole eye analysis or isolation of the retinae by “popping” it out of the eye).

Further, both light exposure times and intensities are very variable, (although a trend to use

ZT1 exists). Age does not seem to have an obvious correlation with increased or decreased dopamine turnover (but see486). Table 3.2. summarises the analytical methodology used in the same selection of papers from Table 3.1. Here, it is conspicuous the extent of the diversity in the solutions used to homogenise the sample, the flow rate, both mobile and stationary phases and internal standards.

From the conclusions obtained during our comprehensive method optimisation (including the extended results), it is recommended:

1. To reduce (when possible) strong acids such as perchloric acid, since it is more

hazardous to both the researcher and the sample compared to weaker acids such as

formic acid, which performs the same functions of lowering the pH to obtain both

protein precipitation and enzyme inactivation.

2. Not to add sodium metabisulfite or EDTA significantly since they reduced absolute

peak integration of dopamine likely due to an increased ion suppression while DOPAC

remained unchanged (Figure 3.1), meaning that DOPAC: dopamine ratio would be

significantly affected.

3. The flow rate determines the elution times of the metabolites meaning that if it were

too fast the analytes would co-elute with small molecules or if too slow analytes may

co-elute with lipids. This co-elution would likely impact the detection of the analytes.

Indeed, the peaks of the analytes are optimally distributed when the flow rate is kept

between 0.2 and 0.3 mL/min).

4. Acidified water and acidified methanol performed well compared to more complex

mobile phases.

5. The use of C18 columns since these exhibited stable performances across a variety of

sizes and models, although the cleaning of the column highly impacts on the quality of

the chromatogram, and is rarely mentioned on the literature. Overnight cleaning with a

low flow rate (~0.2 mL/min) of 100% acetonitrile significantly increased the half-life

of the column, as well as removed lipids and proteins.

In conclusion, I recommend the use of each of the steps followed in this article as a comprehensive troubleshooting guideline to improve or solve problems related with dopamine release quantification using UHPLC-MS/MS.

Light Light Euthanasia Dark light Collection DOPAC/DA Age Ref. intensity exposure n.s. infrared Dissection 0.02-0.08 42-56 170 lux From 1 to 12 h 389 n.s. dim red Dissection 0.06-0.13 70 100 lux 4 and 8 h 487 n.s. n.s. Whole eye 0.4 42 300 lux 4.5 h 402 Cervical infrared Whole eye 0.15-0.3 40-120 n.s. 5 and 11 h 398 dislocation “humanely dim red n.s. 0.1-0.2 40-120 100 lux 15-20 min 488 killed” n.s. dim red n.s. 1.2 18 200 lux 1 h 489 Cervical dim red n.s. 0.2 105 300 lux 2 and 14 h 490 dislocation Cervical dim red Dissection 0.2 35-50 200 lux 1 h 401 dislocation Cervical n.s. Dissection 0.1-0.3 84 n.s. n.s. 491 dislocation Cervical dim red "Popping" 0.1-0.6 70-120 1300 lux 90 min 368 dislocation Table 3.1. Light-pulsing methods across literature. Different features that affect dopamine turnover quantification are gathered in this table: euthanasia, light used during dissection in the dark, collection of the retinae, DOPAC: dopamine ratio, age, environmental illuminance in the dark: light cycle and time of exposure to light. (n.s.: not specified in the source)

Acid Antioxidant Additions Flow Mobile phase Column Standard 0.1M 20% 1.0 75mM PB, 25mM EDTA, 1.7mM C18, 3µm, DA, n.s. HClO4 ACN mL/min OA, 0.01% TEA, 7.5% ACN, pH 3 2.8 x 10 cm DOPAC 89.5% 0.1M TCA, 0.01M SA, C18, 0.1M 10mM 0.6 0.01M SA 0.1mM EDTA, 10.5% MeOH, pH 2.6µm, 4.6 IP TCA EDTA mL/min 3.8 x 10 cm 89.5% 0.1M TCA, 0.01M SA, C18, 0.2M 0.01M 0.1mM 0.6 0.1mM EDTA, 10.5% MeOH, pH 2.6µm, 4.6 n.s. HClO4 SMBS EDTA mL/min 3.8 x 10 cm 0.1M 50mM KP, 0.02% OSS, 15% C18, 5µm, DA, 0.5mM AA - - HClO4 MeOH, 0.1mM, EDTA, pH 3.1 4.6 x 15 cm DOPAC 89.5% 0.1M TCA, 0.01M SA, C18, 0.1M 10mM 0.6 0.01M SA 0.1mM EDTA, 10.5% MeOH, pH 2.6µm, 4.6 IP TCA EDTA mL/min 3.8 x 10 cm 95% PB, 0.05mM EDTA, 1.7mM C18, 0.1M 0.1mM 0.2 3,4- 10µM AA OSA, 90mM NP, 50mM CA, 5% 2.2µm, 2.1 HClO4 EDTA mL/min DHBA ACN x 10 cm 0.1M 0.3 50mM KP, 15 mg/ml EDTA, C18, 5µm, DA, - - HClO4 mL/min 0.26mM NOS, MeOH 8%, pH 4.5 4.6 x 25 cm DOPAC 95% PB, 0.05mM EDTA, 1.7mM C18, 0.1M 0.1mM 0.2 3,4- 10µM AA OSA, 90mM NP, 50mM CA, 5% 2.2µm, 2.1 HClO4 EDTA mL/min DHBA ACN x 10 cm 0.1M 0.01% 1.5 0.1M PA, 0.1mM EDTA 0.45mM C18, 5µm, 3,4- - HClO4 SMBS mL/min NOS, 6% ACN, pH 2.7 4.6 x 25 cm DHBA 0.2M 0.01% 1.5 0.1M NP, 0.1mM EDTA, 0.35mM C18, 5µm, 3,4- - HClO4 SMBS mL/min NOS, 5.5 % ACN, pH 2.7 4.6 x 25 cm DHBA Table 3.2. Analytical methods across the literature. Different methods to assess dopamine quantification: acid, antioxidant and other elements added during the homogenisation of the retinae (or whole eyes); flow rate; mobile and stationary phase (column); and internal standards. AA: ascorbic acid, ACN: acetonitrile, CA: citric acid, 3,4-DHBA: 3,4-dihydroxybenzilamine, EDTA:

ethylenediaminetetraacetic acid, HClO4: perchloric acid, IP: isoprotenerol, KP: potassium phosphate, MeOH: methanol, NOS: sodium octyle sulfate, NP: sodium phosphate, n.s.: not specified, OA: octanesulfonic acid, OSA: orthosilicic acid, OSS: octyl sodium sulphate, PA: phosphoric acid, PB: phosphate buffer, SA: sodium acetate, SMBS: sodium metabisulfite, TCA: trichloroacetic acid, TEA: triethylamine.

CHAPTER 4:

Light-induced dopamine release in wild-type mice

4.1. INTRODUCTION

Dopamine is the main light-driven neuromodulator of the retina and thus it is important to understand which environmental light conditions and pathways are involved to ultimately trigger its release. Approaches such as electron microscopy or immunohistochemical studies have contributed to the discovery of direct synaptic inputs into/from dopaminergic amacrine cells, while electrophysiological studies revealed the physiology of dopaminergic amacrine cells in response to light-driven inputs from all photoreceptors classes. However, the electrophysiological activation of dopaminergic amacrine cells does not always correlate with actual dopamine release368,449 and thus provides an incomplete picture of the role of dopamine in light adaptation. Quantitative studies helped to determine the effect of different drugs on the overall release of dopamine in the retina in several species, providing a valuable but still incomplete piece of the story. Further, as previously discussed in “Chapter 3. Optimisation of a method to accurately quantify dopamine turnover in the retina”, different laboratories use a huge diversity of methodologies and it is sometimes challenging to assume comparisons between results and yet, from all the extensive literature not a single study has fully characterised actual dopamine release in a single species with a consistent methodology both in vivo and ex vivo under a wide range of light conditions and/or pharmacological modulation.

Thus, the aim of this chapter is to fill the gap in the literature to better comprehend dopamine dynamics, release threshold and differences between in vivo and ex vivo preparations in wild- type mice. In this chapter I assessed retinal dopamine release in two complementary ways: in vivo with DOPAC: dopamine ratio from the entire homogenised retina, and ex vivo with dopamine released from the explanted retina into the media where the retina was submerged.

4.2. RESULTS

4.2.1. Dim light suppresses dopamine release

True dark conditions in retinal experiments are generally very rare, frequently “dark” refers to dim or near infrared illumination. Previous experiments examining dopamine release have used dim red light as the “dark” control for mouse experiments, taking advantage of their blue- shifted retinal sensitivity. However, while mouse rods are maximally sensitive to 498 nm wavelength light, they are ~4 log units more sensitive than cones under dark-adapted conditions meaning a significant rod signal is produced even under 8.79 log rod photons cm-2 s-1 illumination (equivalent to dim red light dissection). Despite the limited time spent on the dissection procedure (< 1 minute), surprisingly retinae dissected under dim red light (8.79 log rod photons cm-2 s-1) showed a significant decrease (t-test, p ≥ 0.02) in dopamine turnover compared to those dissected under infrared light (5.82 log rod photons cm-2 s-1) conditions at both CT6 (subjective midday) and CT18 (subjective midnight; Figure 4.1 A). Further, no differences between equivalent time points were observed at those circadian times (one-way

ANOVA with Tukey’s multiple comparison test, p ≥ 0.68). This suggests that rod input actually supresses dopamine release at dim light intensities. Dim red light would be subsequently used as the preferred dark-adapted control condition for further experiments (until specified otherwise) to compare with light-driven dopamine release as it offered the lowest basal level.

The ex vivo preparation exhibited a similar result with more dopamine released under infrared dissection conditions (Figure 4.1B), although it is not statistically significant (t-test, p = 0.059).

4.2.2. Light-induced dopamine release is high threshold

Dopamine is strongly released by bright light in a variety of species and yet the exact light threshold to elicit an increase in dopamine turnover remains mostly unknown. In fact, only one

75 article measured actual dopamine release along an irradiance-response curve in rabbit444. Since mice are such a pivotal model used in neuroscience and most of the literature contains experiments conducted with mice, I performed an irradiance-response curve to elucidate the threshold of dopamine release in wild-type mice in both in vivo and ex vivo approaches. In vivo, light-induced dopamine release was tested in mice exposed to different steady light conditions, and compared to dark-adapted mice dissected under dim red light (Figure 4.2). Experiments were conducted from 16 log rod photons cm-2 s-1 decreasing in 0.9 log units until no difference between the “dark” conditions were observed. Surprisingly, only the two highest irradiances

(~15 and 16 log rod photons cm-2 s-1; 15 min) caused a significant increase in light-induced dopamine release (one-way ANOVA with Dunnett’s multiple comparison test, p < 0.0001).

Further, mice were tested at different circadian points either CT6 or CT18, and no differences were observed between equivalent irradiance points (Figure 4.2), which can be fitted with the same curve (extra sum-of-squares F test, p = 0.83; F = 0.37). In the literature, flickering light has been reported to increase dopamine release in a number of species359,360,363,364,444,492,493 but

I show that dopamine release levels were the same (t-test, p = 0.18) in response to either steady or flickering light just below the release threshold (Figure 4.2).

Similarly, light-induced dopamine release in the supernatant of ex vivo eyecups (cornea and lens were removed but the retina remained attached to the RPE) exposed to different light intensities (from dark-adapted to 16 log rod photons cm-2 s-1, 15 min) at CT6 (Figure 4.3) exhibited a significant increase in light-induced dopamine release only at the two highest irradiances (one-way ANOVA with Dunnett’s multiple comparison test, ***p < 0.0001). The point of this experiment was to validate an alternative method to determine dopamine release with easier accessibility for further pharmacological manipulation (i.e. addition of inhibitory drugs in the media instead of intravitreal injections in vivo). The data illustrates that the two methods report the same, despite one measuring dopamine turnover based on the entire retinal

76 tissue in vivo (often used for dopamine quantification) and the other based on an ex vivo retinal preparation that releases dopamine into a small volume. Despite the light source and the distance of the retina was the same in both in vivo and ex vivo approaches, the lack of cornea and lens causes an increased aperture that causes a slight increase in the illuminance reaching the retina that was corrected (explained in section “2.2.1. Light sources”). Ex vivo results showed a very similar pattern to the in vivo methodology and both curves overlap (extra sum- of-squares F test, p = 0.57; F = 0.72) when plotted together (Figure 4.3). In fact, dopamine release thresholds (calculated by increasing two standard deviations to each irradiance point until it is above to the previous) are identical in vivo (CT6 = 14.59; CT18 =14.61) and ex vivo

(CT6= 14.59). The main conclusion from the irradiance response curve experiments is that light-induced dopamine release is very high threshold.Altogether, this data supports the use of either in vivo or ex vivo methodologies as a complementary assessment of dopamine release in the mouse retina. However, an increased inter-experiment variability in the baseline levels of the ex vivo preparation occurs, perhaps due to further manipulation of the retina and lower dopamine levels released into the supernatant compared to the amount of dopamine in the retinal tissue. Despite the variability between experiments, the intra-experiment variability

(assessed with the standard deviation) is similar to the in vivo results (~10% of the mean), and relative comparisons within the same experiments and dark-adapted controls provided reliable results.

Since many retinal experiments tend to remove the RPE from the retina, I tested the effect of both leaving the RPE attached or removing it from the retina (Figure 4.4). Removing the RPE prior to challenging the retina to bright light (16 log rod photons cm-2 s-1, 15 min) resulted in an impairment not only to the ability of the retina to produce significant light-induced dopamine release (comparing dark-adapted to light-induced RPE-detached retinae), but also lowered significantly the overall levels of dopamine released in the media by ~58.3%

(comparing RPE-attached to RPE-detached retinae at equivalent conditions, t-test, **p ≥

0.007). Thus, all experiments were performed keeping the RPE attached (as described in section “2.2.2. Light pulsing, sample preparation and analysis”).

The existing literature does not provide comprehensive details of the dynamics of dopamine release in mice. While changes in basal levels of dopamine turnover after several hours of dark- or light-adaptation have been measured, electrophysiological experiments record light-evoked responses of dopaminergic amacrine cells in response to very short light-pulses. Thus, little evidence provides information between seconds and hours of light exposure. Here, I measured dopamine release in response to bright light (16 log rod photons cm-2 s-1) at different exposure times from 30 seconds up to 1 hour in vivo or up to 15 minutes ex vivo (Figure 4.5). Further, after 1 hour of light-adaptation I switched off the lights to determine the time required for dopamine turnover to go back to overnight dark-adapted levels. Significant dopamine release occurred within 1 minute of light exposure from overnight dark-adapted mice (one-way

ANOVA with Dunnett’s multiple comparison test, ex vivo: *p = 0.03; in vivo: ***p < 0.0001), and remained constant through the following time points up to 15 or 60 minutes for ex vivo and in vivo preparations respectively. Importantly, while the ex vivo data can be fitted with a logarithmic function, the DOPAC: dopamine ratio showed a much more dynamic profile: it rapidly increased within 30 seconds and peaks at 1 minute, after which a reduction occurred between 1 to 15 minutes, and then stabilised at 15 minutes without further changes up to 1 hour. Conversely, once the lights were switched off from a light-adapted state, DOPAC: dopamine ratio quickly dropped within 5 minutes of darkness, and it continued decreasing for

60 minutes until overnight dark-adapted levels were reached (fitted with a logarithmic function;

Figure 4.4B). These data suggest that, in vivo and ex vivo approaches reflect different aspects of dopamine temporal dynamics: while the former seems to reflect a more instantaneous measurement, the latter reflects dopamine slowly accumulating in the media over time.

4.2.3. Dopamine release is modulated via excitatory and inhibitory inputs

Much is known about dopaminergic retinal responses to a wide variety of drugs in both in vivo and ex vivo approaches. However, drawbacks arise with some of these methods: 1) the RPE is removed in most of these experiments ex vivo (which impairs both dark-adapted light-induced dopamine levels, Figure 4.4), 2) when intravitreal injections are performed in vivo the final retinal concentration is at best an estimate, and 3) the half-life of some drugs are not exactly known. Here, I tested a range of drugs that modulate inputs into dopaminergic amacrine cells in wild-type mice, under both dark-adapted (with dim red dissection and extraction of the retinae) and light-pulsed conditions. Together, this summary of dopamine release in response to a battery of drugs provides further evidence to elucidate the pathways involved in light- induced dopamine release (Figure 4.6). I divided this summary of drug responses into different sections: inhibitory inputs, excitatory inputs, spiking and blocking of rod input into canonical cone pathways via gap junctions.

The activity of inhibitory pathways was tested by the addition of either strychnine (glycine receptor antagonist) or picrotoxin (GABAA/C receptor antagonist). Application of 10 µM strychnine significantly reduced dark levels of dopamine (t-test, **p-value = 0.009), but had no effect on light-pulsed retinae compared to dark and light controls respectively. Comparing between both strychnine conditions, retinae were still able to significantly release light-induced dopamine (t-test, #p-value = 0.02). Conversely, addition of 50 µM picrotoxin significantly raised dopamine levels in dark-adapted retinae (t-test, ***p-value < 0.0001) up to control light levels, but no further increase occurred by light-pulsing retinae while picrotoxin was present, and therefore dark-adapted retina resembled light-pulsed retinae when picrotoxin was present in the media (t-test, p-value = 0.75). This suggests that constitutive GABAergic inputs inhibit

79 dopaminergic amacrine cells in the dark and are relieved by the GABA antagonist causing dopamine release equivalent to a bright light pulse (as previously suggested75,465), while glycinergic inputs might regulate dopamine release at lower light levels and are overwhelmed at bright light intensities.

Signalling of excitatory pathways was assessed by addition of the following set of drugs individually: L-AP4 (mGluR type III agonist), CNQX (AMPA/kainate antagonist) and AP5

(NMDA antagonist). In accordance with previous experiments464, addition of 50 µM L-AP4 significantly reduced both dark and light levels (t-test, **p-value ≤ 0.008) of dopamine compared to their respective control groups and prevented light-induced dopamine release, suggesting that ON bipolar cells must be involved in the pathway governing light-induced dopamine release. Also, 10 µM CNQX significantly blocked light-induced dopamine release back to dark-adapted levels (t-test, **p-value = 0.002), but had no further reduction effect on dark-adapted retinae (t-test, p-value = 0.5). Surprisingly, 10 µM AP5 significantly decreased light-pulsed dopamine (t-test, ***p-value = 0.0002) back to dark-adapted levels, and further reduced dark-adapted dopamine release (t-test, *p-value = 0.02). A small but significant difference (t-test, #p-value = 0.04) could still be observed between AP5-treated dark-adapted and light-pulsed retina. These results suggest that light-inputs causing dopamine release must travel through ON bipolar cells and are mediated by both AMPA/kainate receptors at a certain point in the circuitry, and extrasynaptic NMDA receptors are activated by glutamate spillover from bipolar cells at bright intensities. A positive trend in dopamine turnover was observed in dark-adapted retinae treated with 50 µM NMDA, although removal of Mg2+ from the media to facilitate the activation of NMDA receptors did not further enhance the response (t-test, p =

0.77).

4.2.4. Gap junction coupling is required to drive the majority of light- induced dopamine release

Since rod signals use gap junctions to feed information into cone circuitry (through primary and secondary pathways), I used 200 µM MFA to non-selectively block gap junctions and study the effect of the remaining signalling pathways into dopaminergic amacrine cells. MFA- treated retinae significantly reduced dopamine release in both dark-adapted and light-treated conditions (t-test, **p-value ≤ 0.003). However, significant light-induced dopamine release remained when comparing between MFA-treated conditions (t-test, ##p-value ≤ 0.005) suggesting that this input could come from cone photoreceptors.

4.2.5. Dopamine release is enhanced with depolarisation

Next, the depolarisation-dependence of dopamine release was tested by either enhancing (with high potassium in the medium) or blocking (with TTX) overall excitability in the retina

(including dopaminergic amacrine cells themselves and ipRGCs). On the one hand, 10 µM

TTX caused a significant decrease in dopamine release in both dark-adapted and light-pulsed retinae (t-test, *p-value ≤ 0.01) compared to untreated controls. TTX-treated retinae were not able to drive significant light-induced dopamine release, although a trend was apparent (t-test, p-value = 0.22). On the other hand, when increasing KCl from 3 mM (basal level) to 15 mM

(considered to cause neuronal hyperexcitability494), dopamine release showed a positive trend in dark-adapted and a significant increase in light-pulsed retinae (t-test, *p-value = 0.04) compared to their respective controls. This suggests that dopamine release correlates with increased depolarisation by increasing extracellular potassium, although this is further enhanced by light.

4.2.6. Light-induced dopamine release occurs as early as P18

Taking advantage of the different development timelines of retinal photoreceptors, I explored light-induced dopamine release at different developmental time points to elucidate the relative contribution of each photoreceptor to dopamine release at different developmental ages. These results also provide valuable information for groups that study retinal development and myopia since low levels of dopamine are associated with this visual impairment495, and a better understanding of dopamine release might provide alternative approaches to treat the symptoms at early stages. Light-induced (16 log rod photons cm-2 s-1, 15 min) dopamine release was tested at different time points during development and compared to dark-adapted retinae of the correspondent post-natal days: P0, P7, P18, P22 and P26 (Figure 4.7). From these time points, the earliest age when light-induced dopamine release significantly differed to the dark-adapted control occurred at P18 (t-test, **p-value = 0.045). This difference was maintained at P22 (t- test, **p-value = 0.001), P26 (t-test, ***p-value = 0.0004) and as far as I know until P60, which is the age used for the rest of mice in this thesis (unless stated otherwise). Suggesting that circuit maturity (regarding dopaminergic amacrine cells inputs) is reached ~P18.

Figure 4.1 Dim light supresses dopamine turnover. A) A significant decrease in dopamine turnover occurs in dark-adapted mice dissected under dim red light (8.79 log rod photons cm-2 s-1, >650 nm) compared to those dissected under infrared light (5.82 log rod photons cm-2 s-1, >800 nm) at both CT6 and CT18 (t-test, *p < 0.05, ***p < 0.001, mean ± SEM, n ≥ 6 each). No statistical differences are observed between equivalent irradiance points at CT6 and CT18 (one-way ANOVA with Tukey’s multiple comparison test, p ≥ 0.68). B) Ex vivo preparation showed similar results at CT6 with increased dopamine release under infrared light (t-test, p = 0.059, mean ± SEM, n ≥ 3 each)

Figure 4.2 In vivo light-induced dopamine release occurs only at bright intensities. In an irradiance- response curve, significant light-induced dopamine release occurs at the two highest light intensities (>15 log rod photons cm-2 s-1, 15 min) compared to dark-adapted mice dissected under dim red light (one-way ANOVA with Dunnett’s multiple comparison test, ***p < 0.0001, mean ± SEM). No statistical differences are observed between equivalent irradiance points at CT6 (n ≥ 11 each) and CT18 (n ≥ 3 each; one-way ANOVA with Tukey’s multiple comparison test, p > 0.99; one curve can fit both data sets extra sum-of-squares F test, p = 0.83; F = 0.37). Flickering light (0.5 sec ON, 1.5 sec OFF) does not increase dopamine turnover near the threshold for significant dopamine release (~14 log rod photons cm-2 s-1, 15 min) compared to its equivalent irradiance point of steady light (t-test, p = 0.18, n = 4)

Figure 4.3 Ex vivo light-induced dopamine release occurs only at bright intensities and matches the in vivo approach. In an irradiance-response curve, significant light-induced dopamine release occurs at the two highest light intensities (>15 log rod photons cm-2 s-1, 15 min) compared to dark- adapted retinae dissected under dim red light, similarly to the in vivo irradiance-response curve (Figure 4.2; one-way ANOVA with Dunnett’s multiple comparison test, ***p < 0.0001, mean ± SEM, n ≥ 5 each; one curve (dotted line) can fit both datasets extra sum-of-squares F test, p = 0.57; F = 0.72).

Figure 4.4 Light-induced dopamine release is reduced in retinae without RPE. Significantly lower levels dopamine release in both dark and light (~16 log rod photons cm-2s-1, 15 min) conditions occur in retinae without the RPE attached compared to controls (t-test, p ≥ 0.007, mean ± SEM, n ≥ 3 each). Retinae without RPE attached are incapable of releasing light-induced dopamine (t-test, p = 0.06,).

Figure 4.5 Light-induced dopamine release occurs within 1 minute from a dark-adapted state. A) Significant light-induced dopamine release occurs after 30 s (in vivo) or 1 minute (ex vivo) of bright light exposure (~16 log rod photons cm-2s-1) compared to dark-adapted mice dissected under dim red light (one-way ANOVA with Dunnett’s multiple comparison test, *p < 0.05, ***p < 0.0001, mean ± SEM, n ≥ 5 each). This difference is maintained as long as 60 min (in vivo) or 15 min (ex vivo). B) When lights are turned off from a light-adapted retina in vivo, a significant decrease in dopamine release results after 10 min. The decrease continues for 60 min until overnight dark-adapted levels are reached (one-way ANOVA with Dunnett’s multiple comparison test, **p < 0.001, ***p < 0.0001).

Figure 4.6 Pharmacological modulation of retinal pathways. Bars filled with colour represent dark- adapted retinae, while white-filled, border-coloured bars represent light-pulsed retinae (~16 log rod photons cm-2s-1, 15 min). When two bars match in colour, the same drug treatment is applied in both. Stars indicate comparison between dark and light controls with dark or light treated retinae respectively. Hashtags indicate comparison between dark and light-pulsed retina with the same treatment. Strychnine (glycine antagonist) reduces dopamine in the dark while leaving light-induced dopamine intact.

Picrotoxin (GABAA/C antagonist) has no effect on dopamine release in the light while it causes a significant increase in the dark similar to light-pulsed controls. Overall, inhibition of excitatory inputs into dopaminergic amacrine cells with L-AP4 (mGluR agonist), CNQX (AMPA/kainate antagonist) and AP5 (NMDA antagonist) reduce dopamine levels under light conditions, L-AP4 and CNQX completely abolish light-inducible responses whereas retinae with the AP5 treatment exhibit a reduced but still significant response to light. NMDA show an increasing trend in dopamine release under dark conditions independently from addition of Mg2+. Blockage of gap junctions with MFA reduce both dark and light dopamine release with a small light-inducible response remaining possible Spiking reduction with TTX and depolarisation increase with high KCl concentration dramatically reduce or increase dopamine release respectively (only modestly in the dark with KCl). (t-test, *p < 0.05, **p < 0.001, ***p < 0.0001; t-test, #p < 0.05 ##p < 0.001, ###p < 0.0001 mean ± SEM, n ≥ 3 each condition). 88

Figure 4.7 Significant light-induced dopamine release occurs as early as P18. Significant light- induced (~16 log rod photons cm-2s-1, 15min) dopamine release compared to dark-adapted mice dissected under dim red light at equivalent development points starts at the age of 18 days and is maintained at least until P26. Mice under P18 do not exhibit light-inducible dopamine release in response to bright light. No differences are observed (one-way ANOVA with Tukey’s multiple comparison test, *p < 0.05, **p < 0.01, mean ± SEM, n = 5 each time point).

4.3. DISCUSSION

4.3.1. Dopamine is released under both “dark” and bright light

In this chapter I show that the irradiance profile of dopamine turnover in the retina is a U- shaped curve with more dopamine being released under dark and bright light, and less under intermediate light conditions. A significant increase in dopamine turnover under infrared light conditions compared to dim red light (Figure 4.1) might appear counterintuitive since dopamine is thought to be a light-adaptive neuromodulator. Further, this increase is very fast since the amount of either infrared or dim red light the mice are exposed to during the retinal extraction does not exceed 1 minute (after being dark-adapted for ≥ 6 hours). Here, I propose two hypotheses that might explain this phenomenon: First, rods exert a suppressive effect under dim red light. Although a direct measurement of global dopamine release has not been reported in the dark (except in fish and only during night-time363), indirect evidence based on the coupling of both horizontal cells and AII amacrine cells (which are suggested to be regulated by dopamine) showed increased coupling during dim and bright light, and decreased coupling at intermediate light conditions200,296. In addition, at the dim red light intensity (8.79 log rod photons cm-2 s-1) rods are active (~ 2 log units above its threshold) and thus are a likely candidate to be driving this suppressive effect under dim light. However, the application of L-

AP4 reduced dopamine release under dim red light suggesting that this suppression is not mediated by ON bipolar mechanisms, but rather rod synapses into OFF bipolar cells304.

The second hypothesis implies that this unnatural exposure to such low-light levels of light, such as experienced by the animal under infrared, might make the retina more excitable as a consequence of receiving inputs beyond the response capacity of the retinal system. This increased excitability could lead to a general increase in depolarisation of dopaminergic amacrine cells and subsequent dopamine release as shown in Figure 4.6. In the literature,

90 dopaminergic amacrine cells exhibited spontaneous activity (with at least four different profiles336) under dark conditions75,327,402,496, and these firing rates are correlated with release of dopamine vesicles (~40,000 molecules of dopamine per vesicle) via amperometry457. While it is true that mice live in very low light levels, moonlight and starlight produce a photon flux of ~9 log photons cm-2s-1 at wavelengths that range 410-700 nm. These light intensities are higher than the experimental conditions reproduced under IR light.

Moving to brighter light intensities, the literature often states that dopamine is released in response to bright light and yet only one article performed an irradiance response curve (in rabbit)497 that determined an actual dopamine release threshold (2 log units lower than the one

I present in Figures 4.2 & 4.3), with a similar pattern that showed an abrupt increase towards the brightest light intensities. However, other papers reported light induced dopamine release at lower485, similar187,364,492 or brighter368,498 light in different mammal and non-mammal species compared to my release threshold in mice. Furthermore, flickering light at a light intensity just below the abrupt increase of dopamine release did not show any difference compared to steady light (Figure 4.1), although dopaminergic amacrine cells have been shown to trigger strong bursts of spikes in rodent retinae in response to flickering light336. An increase in dopamine release was reported in both non-mammal360,363,492,493,499 and mammal species364,450, although for mammal species the flickering stimuli was brighter compared to the steady light.

While it is possible that the method I used to determine dopamine release is not sensitive enough to measure localised augments of dopamine, it is certainly able to statistically discern between two measurements whose dopamine release differs ~16 pg of dopamine/retina (Figure

4.6). In addition, a lot of the experiments reported in the literature were conducted in vitro retinae without the RPE attached and it is very rare to see light intensities as high as those used in this thesis because a significant portion of the rod photopigments would be irreversibly

91 bleached. It may be that these experiments miss the abrupt increase at light levels above 14 log rod photons cm-2 s-1. Indeed, I highlight the importance of keeping the RPE attached to the retina in ex vivo preparations in Figure 4.4, as without it retinae release significantly less dopamine likely due to a decrease in the regeneration of all-trans-retinal to 11-cis-retinal from rod opsin during bright light treatment. Certainly, while cones can additionally count on Müller cells to obtain 11-cis retinal500–502 and melanopsin can act as a photoisomerase with certain independence of the RPE503–505, rods strongly rely on the RPE506,507 suggesting that rods are necessary to drive light-induced dopamine release (as previously suggested by Cameron et al.368). Also, dopamine release occurred within a minute in response to bright light (Figure

4.5A) and while other studies reported the earliest light-induced dopamine release over longer timescales (within 10 to 15 minutes after light exposure187,485,492), actual dopamine vesicles are released within 3 seconds after dopaminergic amacrine cell depolarisation508.

Retinal dopamine was linked with eye growth a long time ago and hypothesised as a target for the regulation of myopia509. Since then, a number of studies showed that exposure to bright light (that cause light-induced dopamine release) reduced the risk of myopia, supporting the critical role of dopamine during development510–516. However, the contribution of each photoreceptor to light-induced dopamine release has not been treated in depth, although rods were considered to be involved ~P28 in mice399. In my experiments, overall light-induced dopamine release did not occur before P18 (Figure 4.7), even though all three photoreceptor types have been previously shown to be functional before that time. While it may be that the retinal circuitry is not mature enough at this age, dopaminergic amacrine cells can certainly be labelled as early as E19517, P7518, P10519 and P16520 in rat, and P6 in mouse, and those numbers increase until adult levels are reached ~P18 or ~P15 respectively338. Further, Arroyo et al.521 suggested that dopaminergic amacrine cells are functional very early (~P4) and can be elicited to release dopamine by retinal waves. The first functional photoreceptors in the retina are

92 ipRGCs, which have been shown to elicit light-evoked responses at P0 via electrophysiology, calcium imaging and c-fos activation of brain projections85,112,522 and even regulate light- induced retinal responses before birth521,523. However, no differences were observed between dopamine release in light-pulsed and dark-adapted retinae at that age (Figure 4.7), suggesting while melanopsin is active and signalling, its input alone is not enough to drive overall light- induced dopamine release. Later, S and M cone opsins are expressed at ~P1 and~P7 respectively, and rod opsin ~P5108, but no literature is available regarding actual light responses from these photoreceptors or downstream pathways at these ages (although excitatory post synaptic potentials can be recorded from ganglion cells at P7524 matching the time of arrival of axons from bipolar cells525). However, my experiments confirmed similar levels of dopamine turnover between light and dark conditions at P7. At eye opening (~P10-P12) light provokes electrophysiological responses in both ON and OFF retinal ganglion cells526,527, which correlates with the time that rods become effectively responsive to light85,110 (assessed by c-fos and electrophysiology). Unfortunately, no intermediate point of dopamine turnover between

P7 and P18 was tested, but one would expect at least a moderate increase in DOPAC: dopamine ratio in response to light ~P12-P15. In the literature, dopaminergic amacrine cells have been shown to respond to light ~P18127 (assessed with c-fos), correlating with the positive results in light-induced dopamine turnover between P18-P26 (Figure 4.7).

4.3.2. In vivo and ex vivo approaches provide complementary results

Treating retinae with in vivo or ex vivo approaches allows us to understand dopamine release from complementary perspectives. The in vivo approach of measuring DOPAC: dopamine ratio seems to provide an instantaneous picture of the moment when retinae were extracted and snap frozen. This is suggested by the fluctuation observed between 1 and 15 minutes after light exposure (Figure 4.5). For instance, at 5 minutes the animal has had more light exposure than

93 at 1 minute, but the DOPAC: dopamine ratio is smaller. These results may reflect the regulation of dopaminergic amacrine cells via different feedback mechanisms involving dopamine at different points of the visual pathways that input into dopaminergic amacrine cells, such as: 1) reducing rods activity under bright light amacrine cells528 and 2) reducing the amplitude and sensitivity of rod-driven depolarising bipolar cells via disinhibition of a GABAergic inhibitory input437 diminishes the excitatory input into dopaminergic amacrine cells; 3) increasing excitability of OFF bipolar cells440,529 reduces the inhibitory parallel inputs into dopaminergic amacrine cells330,336; 4) closing gap junctions284,445,530,531 reduces rod fed into cone pathways

(primary and secondary rod pathways); 5) auto-inhibition of dopaminergic amacrine cells via

D2 receptor decreases TH activity419.

Further, an important consideration should be made to researchers using the DOPAC: dopamine ratio as a measure of overall turnover in the retina since the dissection conditions of the retina (e.g. ~1 minute of bright dissection light) could significantly change this ratio over a short time-scale. Consequently, animals were light-pulsed for 15 minutes in further experiments since responses were stabilised by this time.

The ex vivo preparation is particularly useful for experimental designs that involve the use of drugs, since they can easily be added in the media without the troubleshooting of intravitreal injections. However, while the ex vivo measurements provide similar irradiance response graphs as the in vivo approach, the ex vivo time course shows a constant increase in dopamine release until a plateau is reached (Figure 4.5) without the oscillation observed in vivo. The differences in the reduction of temporal resolution can be explained because dopamine released in the supernatant (or pseudo-vitreous) is just accumulated over time taking longer to diffuse into a larger volume compared to the transmission within the retinal tissue. Also, dopamine molecules released from the ex vivo preparation might suffer from oxidation before the extraction (the artificial media was bubbling and/or atmospheric oxygen diffusion) and cause

94 a further reduction in the final quantification. Certainly, in “Chapter 3. Optimisation of a method to accurately quantify dopamine turnover in the retina”, I show that the standard counts of dopamine without an antioxidant such as ascorbic acid are significantly reduced compared to those with ascorbic acid. However, addition of ascorbic acid into the supernatant of the ex vivo preparation as an attempt to prevent dopamine oxidation was not possible because ascorbic acid affects both RPE532,533 and GABA receptors534 and might alter results of light-induced dopamine535. All these differences could account for the increased consistency in the in vivo baselines, compared to the variability in ex vivo baseline levels as previously discussed.

Additionally, the ex vivo retinae suffer more manipulation and small variations in the timing of the procedure might compromise the viability of the tissue (e.g. part of RPE detaches). To reduce that variability, viability assays in conjunction with protein normalisation536 can be performed in the future after the collection of the supernatant.

4.3.3. Dopamine release does not match electrophysiological inputs

Interestingly, the electrophysiological responses of dopaminergic amacrine cells upon application of some of the drugs from Figure 4.6 (strychnine, picrotoxin, L-AP4, CNQX, AP5,

TTX) did not correlate with dopamine quantification i.e. dopaminergic amacrine cells have electrophysiological light-induced responses after its application but not light-induced dopamine release77,327,330,333,336,458, suggesting a discrepancy between actual dopamine release and electrophysiological responses (as previously mentioned368). However, electrophysiological recordings are conducted at the soma of dopaminergic amacrine cells and actual dopamine release is suggested to occur at the processes451 that lack voltage-gated sodium channels537 suggesting they are electrically isolated from the soma. Even if the electrophysiological recordings were performed in the processes, in the striatum, the machinery necessary for dopamine release is only found ~20% of dopamine varicosities with the rest being

“silent”538. Therefore, it is possible that a similar scenario occurs in the retina and the recordings would still mismatch with dopamine release. However, stimulation of excitability with high potassium (15 mM KCl) increases dopamine release only under light stimulation compared to the control (3 mM KCl) similarly as reported in the literature, where a significant increase ~5-fold of dopamine release occurs with higher concentrations of KCl (between 25-

40 mM KCl)355,356,360,364,492,498,539. Depolarisation caused by high K+ presumably increases spiking, although there is the possibility that this depolarisation inactivates V-gated Na+- channels and thus reduces spiking. This last hypothesis can be easily tested in future experiments combining KCl with TTX to check if the increase in dopamine is maintained in the presence of both compounds. Conversely, TTX blocks sodium channels and therefore prevents firing of action potentials540. Indeed, application of TTX ex vivo significantly reduced dopamine levels in both dark and light, and that dopamine could not be released in response to light if TTX was present in the media. Sodium channels are mainly located in third-order neurons such as amacrine537,541,542 (including dopaminergic amacrine cells324,537) and retinal ganglion cells543,544 (including ipRGCs329,330); although in humans they appear to be present in rods545 and (controversially) in bipolar cells 546–549, which are suggested to be involved in the termination of a light response. Together these results (KCl and TTX) suggest that an increase in excitability is required for dopamine release at the specific discharge site. Alternatively, the increasing effect of KCl on dopamine release may be caused by a sustained depolarisation of dopaminergic amacrine cells rather than to an increase in spiking activity. Increasing extracellular K+ would shift the membrane potential to ~-50mV, potentially leading to Na+ channel inactivation but activating the voltage-dependent Ca2+ channels responsible for dopamine release.

Overall changes in light-induced dopamine release from dopaminergic amacrine cells in the presence of strychnine or picrotoxin did not have any effect compared to untreated light

96 controls (Figure 4.6). Glycine is released by half of the amacrine cells of the retina, and the majority of them are at the post-synaptic side of AII amacrine cells, which connect rod bipolar signals to cone OFF pathways550. In addition, glycine receptors are found in distinct subsets of amacrine cells including AII and dopaminergic amacrine cells452, and 3 µM strychnine effectively blocks different glycinergic receptors256. Relieving an inhibitory input hypothetically should increase excitability in dopaminergic amacrine cells, and indeed 1 µM strychnine has been shown to increase excitability of dopaminergic amacrine cells77. However, my results showed no additional increase in dopamine release, either in the light or the dark, in response to 10 µM strychnine. This agrees with other measurements of dopamine release in the literature466, further demonstrating the lack of correlation between electrophysiological response and dopamine release. However, Morgan et al.467 showed a dose-dependent increase of L-DOPA in dark-adapted rats (up to 50 nM strychnine), whereas my results show the opposite. These differences could be related to differences in both dosage (which has opposite effects at low and high doses on bipolar cells551) and analytical techniques (mice ex vivo eyecup against rat in vivo intravitreal injections). Additionally, 10 µM strychnine has been shown to block acetylcholine receptors552,553 which may affect the suggested link between ON bipolar terminals and dopaminergic amacrine cells through starburst amacrine cells322.

Picrotoxin has been shown to effectively block GABA-induced currents in GABAA and

554–559 GABAC receptors, both of which are ligand-gated chloride channels . GABA is released from horizontal cells, A17 and dopaminergic amacrine cells, and receptors can be found in all retinal cell types including dopaminergic amacrine cells themselves560. My experiments showed that addition of 50 µM picrotoxin raises dark-adapted dopamine levels to equal those of light-pulsed retinae and light and picrotoxin had no synergistic effect. Similar results are found in the literature with additions ~30-100 µM picrotoxin in fish356,492,561, amphibian562 and mammal350,465 retinae. Ultimately, the effects on upregulation of dopamine release in the dark

97 are most likely explained due to the relief of a tonic GABA inhibition of dopaminergic amacrine cells, as has been previously suggested356,465,562. The GABA source for this inhibition could come from the A17 amacrine cells that contact rod bipolar cells563,564. Alternatively, the presence of GABA receptors in bipolar cells565,566 and the increase in the amplitude of b wave in the electroretinogram after picrotoxin exposure567 suggests an increase light responsiveness of ON bipolar cells that may increase the excitatory input received by dopaminergic amacrine cells. As previously mentioned mice are nocturnal animals and the exposure to such high light intensities are certainly triggering volume transmission of dopamine in response to a bright light from a dark-adapted state. My hypothesis is that the release machinery of dopaminergic amacrine cells gets saturated with bright light and picrotoxin, and therefore we cannot see any synergistic effect. An alternative hypothesis would be that light and picrotoxin trigger pathways that share a common mechanism of dopamine release (supposing that more than one exists) and perhaps the addition of light, picrotoxin and more drugs that trigger DA release (such as high KCl) could answer if dopaminergic amacrine cells reached their saturation.

In contrast to the variability in responses from the relief of the inhibitory inputs to dopaminergic amacrine cells, the blockage of different excitatory inputs with L-AP4, CNQX and AP5 eliminated widespread dopamine release (Figure 4.6). L-AP4 is a potent and selective agonist of group III mGluR receptors (subtypes 4,6,7 and 8), from those, mGluR8 are found in photoreceptors, mGluR6-8 on the ribbon synapse between photoreceptors and bipolar cells227, and mGluR4,7,8 in amacrine and ganglion cells568 (although mGluR7 & mGluR8 are not the most prevalent GluR receptors, they have been found at the photoreceptor-bipolar cell synapse sites. Protein of mGluR7 can be found in rat and goldfish bipolar cells, and in mouse cones; whereas protein and mRNA of mGluR8 can be found in mouse and rat photoreceptors, and mRNA in mouse bipolar cells). L-AP4 is mainly used in the retina to eliminate ON bipolar cell responses by hyperpolarisation of these cells. Application of 50 µM L-AP4 in my experiments

98 was enough to prevent light-induced dopamine release and impact significantly on dopamine levels in the dark suggesting that basal dopamine release in the dark is mediated by a different circuitry than light-induced (possibly via rod →OFF bipolar cells304). In the literature, the same concentration of L-AP4 only abolished transient light electrophysiological responses in dopaminergic amacrine cells327,330 with the remaining sustained resistant component suggested to be driven by ipRGCs327. It may be that it is the transient, rather than sustained, electrophysiological response that ultimately elicits dopamine release.

CNQX is a competitive antagonist that blocks all excitatory synaptic activity of AMPA and kainate iGluR receptors, which are present in horizontal, OFF bipolar, amacrine (including dopaminergic amacrine345), and ganglion cells251,569–574. Thus, CNQX could be blocking key synapses such as ipRGCs → dopaminergic amacrine cells328,330. However, it is difficult to discern where the block is occurring, although dopaminergic amacrine cells express AMPA and kainate receptors77,324 that could explain the reduction observed in light-induced dopamine in the presence of 10 µM CNQX.

AP5 is a selective and competitive antagonist of NMDA receptors and, as previously mentioned, NMDA receptors are blocked by Mg2+ at resting membrane potentials below -65 mV. Therefore, removing this ion from the extracellular solution allows these ion channels to be activated by synaptic glutamate without coincident membrane depolarisation575,576. Indeed, my results suggest an involvement of NMDA receptors since blocking NMDA receptors with

10 µM AP5 reduced light-induced dopamine release, and enhancing NMDA receptor activity in a medium free of Mg2+ increased dopamine release. Further, different subunits that form

NMDA receptors largely impact on their affinity577,578 and blockade sensitivity of Mg2+ 579–581, with NR2A and NR2B more strongly blocked than NR2C and NR2D582 and preferentially expressed in both amacrine and ganglion cells583. Perhaps an uneven relief of NMDA blockage at different points of the retinal circuitry prevented the triggering signal into dopaminergic

99 amacrine cells to release dopamine. Even though NMDA receptors (and its subunits) are expressed ubiquitously throughout the retina583, more specifically, NMDA receptors have been shown to increase light responses from AII and A17 amacrine cells584, which are specific amacrine cells of rod pathways. Thus, it may be that at bright light intensities glutamate spillover from rod bipolar cells activates NMDA receptors at these synapses.

Finally, when gap junctions in the retina were non-specifically blocked with 200 µM

MFA287,585, light-induced dopamine release was significantly reduced but not eliminated suggesting that the remaining signal must be from cones, ipRGCs or an alternative gap-junction independent rod pathway (the role of cones and ipRGCs will be further discussed with transgenic animal models in “Chapter 5. Photoreceptive input driving light induced dopamine release”) or simply reflect an incomplete blockage of pathways. Certainly, the tonic light driven input that dopaminergic amacrine cells receive from ipRGCs is still present when MFA is applied544,586. Again, suggesting that spiking of the soma of dopaminergic amacrine cells does not equal widespread dopamine release. A more specific experiment performed by Zhang et al.333 knocking out specifically Cx36 (using a transgenic mouse model) to disrupt primary and secondary rod pathways (instead of all gap junctions with MFA), showed that light induced responses from dopaminergic amacrine cells were eliminated (in 5 out of 6 cells) at scotopic light levels, but responses recovered past the mesopic threshold. Therefore, they suggest that all three: primary (rods → rod bipolar cells → AII amacrine cells → cone bipolar cells), secondary (rods → cone bipolar cells) and tertiary (direct contact from rods to cone bipolar cells310,587) rod pathways involve rod signalling into dopaminergic amacrine cells. Since blockage of the electrical connection between AII and ON-bipolar cells287,585 still provided responses in dopaminergic amacrine cells, a remaining rod signal into dopaminergic amacrine cells may occur via chemical synapses through rod → ON rod bipolar → AII amacrine cells

→ OFF cone bipolar588,589 or ON rod bipolar → A17 amacrine cells451,590.

100

In conclusion, this chapter provides evidence that dopamine turnover is U-shaped with more dopamine being released only under very dim scotopic and very bright photopic light intensities. Temporal dynamics of dopamine are studied in vivo showing an “instantaneous” measurement of dopamine turnover after any given light or dark exposures, which exhibits much faster dopamine increase in response to light rather than dopamine decrease during dark adaptation. The ex vivo preparation is used as a tool to assess dopamine turnover under pharmacological modulation of pathways involved in dopamine release, providing results in agreement with the literature, ultimately indicating that dopamine release is a balance between excitatory and inhibitory inputs. Also, the circuitry involved in light-induced dopamine release reaches its maturity ~P18, despite evidence showing the activation of different photoreceptors at earlier stages during development. Importantly, different experiments from this chapter show that widespread dopamine release does not correlate with electrophysiological activation of dopaminergic amacrine cells, suggesting that the assumption made in few articles from the literature that somatic activation of dopaminergic amacrine cells equals dopamine release may be wrong. However, each analytical technique has its own limitations and much smaller amounts of localised dopamine (that I was unable to quantify) may be released in response to electrophysiological activation.

101

CHAPTER 5:

Photoreceptive input driving light-induced

dopamine release

102

5.1. INTRODUCTION

In the previous chapter I defined global light-induced dopamine release over a range of light intensities, exposure times and pharmacological modulation in wild-type mice. However, in order to determine the origin(s) of the input(s) that determine dopamine turnover, here, I used a range of mouse models with altered photoreception: Gnat2A518G mouse lacking cone phototransduction but without cone degeneration468; Opn4Cre/Cre mouse lacking melanopsin but without major ipRGCs reorganization52,478; rd/rd mouse whose rods and cones degenerate during development and suffer significant retinal remodelling, but have intact and functional ipRGCs; and Opn1mwR mouse that have a knock in (human) L-opsin substituting the (murine)

M-opsin and present a shifted sensitivity towards red469, which is used to spectrally segregate the activity of cones from the other photoreceptors.

The previous chapter showed that dopamine is released in response to bright light. Since both cones and ipRGCs are known to be functional at bright light intensities and present a well- established electrophysiological input into dopaminergic amacrine cells, I characterised the putative input of both ipRGCs and cones to light-induced dopamine release. First, to test the role of cones I used three different approaches: 1) irradiance-response curve in Gnat2A518G mouse compared to wild-type from the previous chapter, 2) Opn1mwR mouse to preferentially activate cones using a red-shifted light stimulus, and 3) light pulses at developmental ages in rd/rd mice where rods are lost but cones are still functional.

To determine the input of ipRGCs I relied on four different approaches: 1) irradiance-response curve in both Opn4Cre/Cre (no melanopsin) and rd/rd mouse (only ipRGCs active); 2) preferential activation of rods and cones but not melanopsin using distinct spectral stimuli in wild-type mice; 3) use of aged (>18 months) rd/rd mice to evaluate the impact of the reported increase in the number of dopaminergic amacrine cells activated by light at this age449,591

103

(presumably from ipRGCs); and 4) chemogenetic activation of ipRGCs (exclusively) using virally introduced Cre-dependent Designer Receptors Exclusively Activated by Designer

Drugs (DREADDs) in the Opn4Cre/+ mouse80.

Although rods are classically known to be active at low light intensities and become saturated under bright light, recent literature argues against this and suggests that rods might be functional at bright light intensities27,28. Therefore, to determine the input of rods in both dim and bright light conditions I performed three different experiments: 1) dissection of the retinae from all the aforementioned mutant mice under both dim red and infrared light to determine if the suppressive effect of dopamine turnover under dim red light was indeed caused by rods alone (Figure 4.1); 2) use of Gnat2A518G mouse light-pulsed with a red-shifted light to activate rods but not melanopsin above dopamine release threshold; and 3) use of MFA on light-pulsed

Gnat2A518G mouse to assess if an increase in dopamine turnover was due to primary, secondary or tertiary rod pathways, or had any additional photoreceptive inputs.

Further, I systematically provided complementary approaches to overcome the limitations of the experimental conditions and transgenic mice models themselves (discussed in detail throughout the chapter). Altogether these results suggest that rods alone are both sufficient and necessary to drive light-induced dopamine release at any given light intensity, and conversely, cones nor ipRGCs elicit significant changes in widespread light-induced dopamine release despite known electrophysiological input to dopaminergic amacrine cells.

104

5.2. RESULTS

5.2.1. Rod input supresses dopamine release under dim light

In the previous chapter I suggested that rods were responsible for suppression of dopamine release at dim light intensities, given the differences in dopamine turnover observed between dim red and infrared dissection in wild-type mouse (Figure 4.1). In order to clarify the photoreceptive origin that caused this reduction, I performed dissections under both dim red and infrared light in Gnat2A518G, Opn4Cre/Cre, Opn1mwR and rd/rd mice (Figure 5.2). Retinae from Gnat2A518Gand Opn4Cre/Cre dissected under dim red light exhibited a significant decrease in dopamine turnover compared to those dissected infrared light (t-test, **p-value ≤ 0.009).

Further, comparison of Gnat2A518G and Opn4Cre/Cre with wild-type mouse showed no differences at equivalent irradiance points (one-way ANOVA with Dunnett’s multiple comparison test against wild-type dim red p-value ≥ 0.21, and wild-type infrared p-value ≥ 0.34). No significant differences were observed between dim red and infrared light dissections in Opn1mwR mice (t- test, p-value = 0.32), although a trend of ~1.6-fold decrease under dim red light conditions was observed. However, no significant differences were observed when comparing equivalent irradiance points of Opn1mwR with wild-type mice (t-test, p ≥ 0.23). Given the red-shifted spectral activation of cones in this model, this result suggests that the increased activation of

L-cones under infrared light may drive a small reduction in dopamine release, reducing the difference between dim red and infrared conditions. Last, rd/rd mice did not exhibit any difference between dissection conditions (t-test, p-value = 0.17), suggesting that rods are necessary to modulate dopamine turnover at these light intensities and that ipRGCs alone do not elicit any measurable effect on the modulation of dopamine release by dim light.

105

5.2.2. Rods are both sufficient and necessary to drive light-induced dopamine release

In line with the rationale from the previous chapter, I tested dopamine release along an irradiance-response curve from dark-adapted (dim red) to 16 log photons cm-2 s-1, in both in vivo and ex vivo preparations. Here, I used three different genotypes (Gnat2A518G, Opn4Cre/Cre and rd/rd) to determine the role of each photoreceptor in light-induced dopamine release at given irradiance points. In agreement with wild-type in vivo data, only high irradiances (15 and

16 log rod photons cm-2 s-1; 15 min) caused a significant increase in light-induced dopamine release in Gnat2A518G and Opn4Cre/Cre (Figure 5.1A, one-way ANOVA with Dunnett’s multiple comparison test, ***p-value < 0.0001) suggesting that neither cones nor ipRGCs are necessary to drive light-induced dopamine release even at bright light intensities. The irradiance response curves from Gnat2A518G, Opn4Cre/Cre and wild-type mice could be fitted in the same curve (extra sum-of-squares F test, p = 0.89, F = 0.44) and exhibited very similar thresholds of dopamine release (Table 5.1), suggesting that neither cones nor ipRGCs are necessary to set the dopamine release threshold. Conversely, rd/rd mice did not exhibit light-induced dopamine release at any given light intensity (one-way ANOVA with Dunnett’s multiple comparison test, p-value >

0.99). However, the results obtained from rd/rd mice need to be interpreted with a degree of caution since those animals suffer severe retinal remodelling during development that might impair natural synaptic connections, and thus may not be the most reliable model to study isolated inputs from ipRGCs. Nonetheless, taken together these results suggest rods are both

1) necessary and sufficient to drive light-induced dopamine suppression and release, and 2) define the thresholds at which those occur.

I repeated these irradiance curves ex vivo and I observed very similar results as in vivo, with

Gnat2A518G and Opn4Cre/Cre mice releasing light-induced dopamine at the two brightest light

106 intensities compared to dim red light (one-way ANOVA with Dunnett’s multiple comparison test, ***p-value ≤ 0.0007), and rd/rd being irresponsive to light throughout the entire given range of light intensities (Figure 5.2B). However, as was apparent in the previous chapter, ex vivo baseline levels were variable, and comparing absolute values of either the two highest irradiances of Gnat2A518G mice or 15 log rod photons cm-2 s-1 of Opn4Cre/Cre mice with equivalent irradiance points in wild-types, showed a significant decrease in dopamine release

(one-way ANOVA with Tukey’s multiple comparison test, $$$p ≥ 0.0001). This may indicate that rods set the threshold for release, whereas cones and/or ipRGCs could play a role in amplitude modulation. However, this change in amplitude could also be a consequence of general morphological changes in the Gnat2A518G animals that prevents diffusion of dopamine out of the retina, or strain differences between the Opn4Cre/Cre animals (C57BL/6 and 129sv mix) compared to C57BL/6 wild-types. Despite this reduction in amplitude of responses, the absolute thresholds of dopamine release remain very close across genotypes (Table 5.1) and if the data is normalised one curve can fit all light-inducible genotypes ex vivo (wild-type,

Gnat2A518G and Opn4Cre/Cre; extra sum-of-squares F test, p = 0.71, F = 0.67). Further, if in vivo and ex vivo irradiance curves of each genotype are normalised together, one curve can fit both datasets in all genotypes (Table 5.1). These results strongly support the use of ex vivo as an equivalent measure for dopamine release, and support the same hypothesis raised by the in vivo results that rods are both sufficient and necessary to drive light-induced dopamine release.

107

Genotypes p-value F value In vivo threshold (log rod Ex vivo threshold (log rod photons cm-2 s-1) photons cm-2 s-1) Wild-type 0.9788 0.1097 14.5966 14.5988 Gnat2A518G 0.3888 1.051 14.5851 14.5624 Opn4Cre/Cre 0.6623 0.6058 14.3854 14.2424 rd/rd 0.9848 0.0153 N/A N/A Table 5.1. No difference between ex vivo and in vivo thresholds for dopamine release. F-test sum- of-squares was applied to normalised (R/Rmax) data from in vivo and ex vivo irradiance curves from each genotype, and one curve coukd fit each pair of data from each genotype. This suggests that both measurements are complementary when assessing light-induced dopamine release at equivalent irradiance levels at the same time exposures (15 minutes). Light-induced dopamine release threshold (defined as 2 standard deviations above dark release) was not different between genotypes or method.

While rd/rd mice might not be the best model to study circuitry due to the substantial remodelling they suffer, ipRGC inputs to dopaminergic amacrine cells are maintained in these animals327–329,592,593. Given that I did not observe any light-induced dopamine responses in rd/rd mice, I tested if the dopaminergic amacrine cells from rd/rd animals could indeed release dopamine. To assess this, I hypothesised two situations: 1) dopaminergic amacrine cells might be receiving additional glycinergic and/or GABAergic inhibition that cannot be relieved by ipRGC input, and 2) dopaminergic amacrine cells do not have the machinery necessary to synthesise and release dopamine as a consequence of degeneration and/or remodelling of the remaining retina. First, I inhibited inhibitory inputs into dopaminergic amacrine cells with either strychnine or picrotoxin as performed in the previous chapter. Application of 10 µM strychnine into the media did not affect dopamine levels of either dark-adapted or light-pulsed retinae compared to their respective untreated controls (t-test, p-value ≥ 0.17). However, addition of 100 µM picrotoxin caused a significant increase in light-induced dopamine release

(t-test, *p-value = 0.02), but not in dark-adapted controls (t-test, p-value = 0.1). Suggesting that indeed, dopaminergic amacrine cells in rd/rd animals receive GABAergic but not glycinergic inhibitory input but that this can only be relieved by a combination of light and picrotoxin to

108 cause dopamine release. These results are in agreement with an enhanced GABA and melanopsin signalling in rd animals reported in the literature592, and suggests that the relief of both stimulates dopaminergic amacrine cells via both disinhibition from GABA and excitation from ipRGCs.

Since depolarisation is correlated with dopamine release in the literature355,356,360,364,492,498

(Chapter 4, Figure 4.6), I stimulated the excitability of rd/rd retinae in a medium with high potassium (15mM; Figure 5.3). Not surprisingly, high potassium caused a significant increase in dopamine release in both dark-adapted and light-pulsed conditions compared to their respective non-treated controls (t-test, *p-value ≤ 0.03). No differences were observed between dark and light treated conditions (t-test, p-value ≥ 0.61), suggesting that the increase was based exclusively on increased excitability or sustained depolarisation of dopaminergic amacrine cells (as discussed in the previous chapter), and as opposed to wild-type mice (Chapter 4, Figure

4.6), was not enhanced by light. While the basal rate of dopamine didn’t increase significantly with high KCl, it certainly had a synergistic effect when combined with light. Taken together, these results indicate that the dopamine release machinery is still capable of functioning and an increase in excitability alone does cause dopamine release in the rd/rd retinae. Further, to better understand if the release of dopamine is based specifically on spiking, an experiment combining KCl with TTX would shine some insight on this hypothesis.

Using a Mg2+-free solution to disinhibit NMDA receptors and increase the excitatory inputs to dopaminergic amacrine cells did elicit significant dopamine release under both dark and light conditions (t-test, ***p-value ≤ 0.0001; Figure 5.3), suggesting that NMDA channels are necessary to trigger dopamine release. However, depletion of Mg2+ will cause an increase in the overall excitability of the retina in addition to unblocking NMDA channels.

109

However, these experiments were performed in young rd/rd mice (~P60), and paradoxically, aged rd/rd mice (~P720) are reported to have a wider activation pattern of dopaminergic amacrine cells in response to light driven by ipRGCs (~3% TH cells expressed c-fos at 3 months and ~36% at >14 months in response to light)593. Therefore, I examined light-induced dopamine release in aged animals to check if the increase in electrophysiological activation correlated with actual levels of dopamine release (Figure 5.4). Although aged animals exhibited an increased basal level of dopamine turnover under both dark and light conditions (t-test **p

≤ 0.0056), they remained irresponsive to a bright light pulse (16 log rod photon cm-2 s-1, 15 min) compared to dark-adapted retinae of an equivalent age. It is possible that increased innervation from ipRGCs in aged animals contributes to the increased basal release of dopamine rather than light-induced release.

Further examination of rd/rd mice was conducted at different developmental points to test if they had light-inducible dopamine at any developmental point before photoreceptor degeneration. Thus, light-induced dopamine release was tested at the highest irradiance (16 log photons cm-2 s-1, 15 min) at P0, P7, P18, P22 and P26 (Figure 5.5), and no light-induced dopamine release occurred at any developmental time point (t-test, p-value ≥ 0.25). This results further support the hypothesis that ipRGCs cannot drive light-induced dopamine alone despite its electrophysiological input (since ipRGCs are active before birth), and that cones alone

(present until P22) were not able to drive light-induced dopamine release at P18 (contrary to wild-types; Chapter 4, Figure 4.7).

To further assess the input and circuitry exploited by each photoreceptor to dopamine release in non-degenerate retinae, I used three additional and different approaches: 1) chemogenetic activation of exclusively ipRGCs in the dark (DREADDs), 2) spectral activation of the three photoreceptors using different wavelengths (624 nm; 434 nm) and light intensities, and 3) pharmacological modulation of Gnat2A518G mice.

110

DREADDs can have several applications depending on the construction of the vector594, and recently much work has been done on ectopic activation of ipRGCs using hM3D-Gq receptors80,595–597. Our vector of choice was pAAV-hSyn-DIO-hM3Dq-mCherry, so the floxed receptor hM3Dq (G-protein coupled receptor) could be expressed only in the ipRGCs of

Opn4Cre/+ mice. Depolarisation can be induced exclusively in ipRGCs with application of clozapine N-oxide (CNO). First, animals were unilaterally intravitreally injected with adeno- associated virus (AAV) containing floxed hM3Dq fused to mCherry. After 6 weeks of incubation, it was then possible to activate ipRGCs expressing hM3Dq with an intraperitoneal injection of CNO (5mg/kg). To verify expression in ipRGCs, pupillary constriction of mice was tested under infrared light (> 950 nm) in response to CNO injection. Figure 5.6 shows a representative image of a mouse pupil before (left) and after CNO (right) injection; this response lasted > 3 hours before the pupil started to re-dilate. Animals with < 80% constriction were excluded as constriction is correlated with expression levels80. As AAV injections were unilateral, the contralateral eye was used as a control. CNO was injected, pupil constriction verified, and eyes removed under dim red light after 1 hour. No significant melanopsin-induced dopamine release occurred in the virus-injected eye compared to the contralateral control

(paired t-test, p-value = 0.62) indicating that even sustained activation of ipRGCs in a non- degenerate retina cannot drive dopamine release.

Next, red and blue light pulses were tested on wild-type and Gnat2A518G mice to determine which photoreceptor is setting the threshold. Theoretical irradiance response curves were constructed for each photoreceptor based on their spectral sensitivity481 (Figure 5.7). These curves assume that the irradiance response to the white light stimulus given the threshold is set by each respective photoreceptor. As rods, M-cones, melanopsin and L-cones have similar sensitivity to white LED light (“Chapter 2. Materials and Methods”, Figure 2.2), effective photons cm-2 s-1 are very similar, however, as the white light does not contain UV light, if UV-

111 cones were driving this response they would be activated by a lower number of effective photons. The aim of these experiments was to use blue- and red-light stimuli to selectively activate each photoreceptor above its theoretical threshold. The blue light (solid blue line) activates all photopigments above threshold, whereas the red light (solid red line) activates only rods and cones above their dopamine release thresholds. Since rods are the prime suspect to drive dopamine release, high intensity red and blue light input intensities (red and blue lines) were matched for effective activation for rods (15 log effective photons cm-2 s-1). Both wild- type and Gnat2A518G mice were able to release light-induced dopamine in response to both red and blue light (Figure 5.8A; t-test, ***p < 0.0001), suggesting that the input of neither melanopsin nor cones was necessary to set the threshold. Further, comparing dopamine turnover levels in conditions where only rods were active above dopamine release threshold

(red light in Gnat2A518G) against activation of rods plus an additional photoreceptor, such as rods + melanopsin (blue light in Gnat2A518G) or rods + cones (red light in wild-type), showed no differences in dopamine turnover levels (one-way ANOVA with Dunnett’s multiple comparison test, p ≥ 0.87). These results suggest that rods alone set and define dopamine release. I also utilised a similar method in Opn1mwR mice to spectrally activate cones but neither rods nor ipRGCs above the dopamine release threshold (Figure 5.8B). The red pulse used here

(dotted line) is of a lower intensity that activates L-cones above threshold but not rods or ipRGCs (Figure 5.7). This stimulus was not able to cause light-induced dopamine release in

Opn1mwR mice (t-test, p = 0.27), although a positive trend was observed (Figure 5.8). To test that light-induced dopamine responses were normal in Opn1mwR mice, I light-pulsed them with white bright light causing significant dopamine release (t-test, **p = 0.002). This showed that the dopaminergic function of Opn1mwR mice was intact, but activating cone photoreceptors above their threshold was not enough to cause full amplitude light-induced dopamine release.

112

Pharmacological modulation in Gnat2A518G mice was the last approach used to test the contribution of the different photoreceptor to light-induced dopamine release by blocking primary and secondary rod pathways with MFA (Figure 5.9). Intravitreal injections of sham saline triggered normal light-inducible responses, but 1mM MFA injection (~100 µM final concentration in the eye) significantly reduced light-induced dopamine release compared to light sham control (t-test, #p < 0.05). In a similar manner, application of 200 µM MFA in the ex vivo preparation caused a reduction in light-induced dopamine compared to light control.

Therefore, both approaches (in vivo and ex vivo) suggest that rods can signal into dopaminergic amacrine cells through a gap junction independent pathway. However, these results may reflect an incomplete block, because while application of MFA ex vivo showed a significant increase in light-induced dopamine release comparing between MFA-treated retinae (t-test, **p < 0.01).

That light-induced difference between treated and untreated retinae was not observed in vivo

(t-test, p = 0.16) although a positive trend was apparent.

These results suggest that the majority of light-induced dopamine release driven from dopaminergic amacrine cells occurs via primary and secondary rod pathway signalling, and a minor amount occurs via a rod-driven gap-junction independent pathway.

113

Opn4Cre/Cre

Figure 5.1 Dopamine turnover is suppressed under dim light by rod input. Significant decrease in dopamine turnover occurs in dark-adapted mice dissected under dim red light (8.79 log rod photons cm- 2s-1, >650nm) compared to dissection under infrared light (5.82 log rod photons cm-2s-1, >800nm) at CT6 in A) Gnat2A518G and B) Opn4-/-, but not in C) Opn1mwR or D) rd/rd (t-test, *p < 0.05, ***p < 0.001, mean ± SEM, n ≥ 3 each). No statistical differences were observed between dim red dissections across genotypes (including wild-type from Figure 4.1), and under infrared conditions only CT6 wild-type statistically differed with rd/rd (one-way ANOVA with Tukey’s multiple comparison test, p = 0.01)

114

)

$$$ $$$

Figure 5.2 Light-induced dopamine release is defined by rod input. A) In an irradiance-response curve, significant light-induced dopamine release occurs at the two highest light intensities (>15 log rod photons cm-2 s-1) compared to dissection under dim red light in Gnat2A518G and Opn4Cre/Cre (one-way ANOVA with Dunnett’s multiple comparison test, ***p < 0.0001, mean ± SEM, n ≥ 5 each) similar to the curve from wild-type mice (Figure 4.2). One curve can fit all three datasets (extra sum-of-squares F test, p = 0.89; F = 0.44) and no significant differences are observed between equivalent irradiance points (one-way ANOVA with Tukey’s multiple comparison test, p ≥ 0.99). Conversely, rd/rd animals do not exhibit light-induced dopamine release at any given light intensity and no differences are observed between irradiance points (one-way ANOVA with Dunnett’s multiple comparison test, p ≥ 0.99 mean ± SEM, n ≥ 12 each). B) Similar to in vivo, light-induced dopamine release occurs at the two highest light intensities in Gnat2A518G and Opn4Cre/Cre, but not rd/rd retinae (one-way ANOVA with Dunnett’s multiple comparison test, ***p < 0.0001, mean ± SEM, n ≥ 5 each). Gnat2A518G retinae present a reduced amplitude of dopamine release at the two higher points of the irradiance curve compared to wild-type (one-way ANOVA with Tukey’s multiple comparison test, $$$p < 0.0001). Opn4Cre/Cre retinae present a reduced amplitude of dopamine release when comparing ~14 log rod photons cm-2 s-1 to wild-type equivalent irradiance point (one-way ANOVA with Tukey’s multiple comparison test, $p = 0.047). When normalised (not shown in the graph), ex vivo data can fit wild-type, Gnat2A518G and Opn4Cre/Cre datasets in a curve (extra sum-of-squares F test, p = 0.71; F = 0.67). Stars indicate comparison between dark and light conditions of each genotype, hashtags indicate comparison between rd/rd and the rest of cohorts at equivalent irradiance points, and dollar signs indicate comparison between wild-type to Gnat2A518G and Opn4Cre/Cre ex vivo.

115

Figure 5.3 Rd/rd mice are capable of releasing dopamine. Bars filled with colour represent dark- adapted retinae, while white-filled, border-coloured bars represent light-pulsed retinae (~16 log rod photons cm-2s-1, 15 min). Stars indicate comparison to respective control conditions (dotted lines). Regarding inhibitory inputs, no effects are observed after application of strychnine (glycine antagonist), but picrotoxin (GABAA/C antagonist) significantly increase dopamine of light-pulsed retinae. Both unblock of NMDA receptors in Mg2+-free media and increased spiking with high KCl concentration significantly increase dopamine release in both dark and light-pulsed conditions compared to untreated controls (t-test, *p < 0.05, ***p < 0.001, mean ± SEM, n = 2 each inhibitory condition and n = 6 each spiking dependent condition).

116

Figure 5.4 Aged rd/rd mice have higher basal dopamine turnover levels compared to young animals but cannot release dopamine in response to light. Similar to young rd/rd, aged rd/rd mice do not release dopamine in response to light. However aged rd/rd mice exhibit higher dopamine turnover levels in both dark-adapted and light-pulsed conditions (t-test, **p < 0.01, mean ± SEM, n = 5 each condition).

117

Figure 5.5 Rd/rd mice lack light-induced dopamine release at all ages during development. No significant differences are observed at any tested age from 0 to 26 days of light-induced (~16 log rod photons cm-2s-1, 15min) dopamine release compared to the respective points of dark-adapted mice dissected under standard dim red light. One-way ANOVA with Tukey’s multiple comparison test (p > 0.99, mean ± SEM, n = 4 each time point).

118

Figure 5.6 Chemogenetic activation of ipRGCs does not drive dopamine release. Opn4Cre/+ mice were infected with virally-introduced hM3Dq in the right eye 6 weeks prior to the experiment. Before the experiment, mice were injected with 5 mg/kg CNO to activate the infected ipRGCs. Pupil constriction was checked in both eyes (>80% of constriction as selection criteria in the treated eye) under infrared light (>950 nm) and dopamine release assessed. No differences in dopamine release turnover are found between treated and untreated retinae (paired t-test, p = 0.62, n = 5 each group), suggesting ipRGC activation is not sufficient to trigger dopamine release.

119

Figure 5.7 Theoretical chromatic contributions of photoreceptors to light-induced dopamine release. The relative activation of different photopigments (rod, melanopsin and UV, M and L cones) with the white light-irradiance response curve (obtained in wild-type mice) is presented assuming dopamine release is driven by each respective photoreceptor. Overlayed on each graph are the relative activation of light pulses delivered in Figure 5.8 for each photoreceptor: dotted red line: medium- intensity red light, red line: high-intensity red light, blue line: high-intensity blue light.

120

Med-red

Figure 5.8 Rods are both sufficient and necessary to drive light-induced dopamine release. A) Both high-intensity red light that activates rods and cones (above dopamine release threshold, see Figure 5.5) but not melanopsin, and blue light that activates all photoreceptors result in a significant increase in dopamine release (one-way ANOVA with Dunnett’s multiple comparison test, ***p < 0.0001 mean ± SEM, n = 5 each condition) compared to dark-adapted wild-type mice dissected under dim red light. No differences are observed at equivalent irradiance points between wild-type and Gnat2A518G mice (one-way ANOVA with Tukey’s multiple comparison test, p > 0.99), whose red light exposure is equivalent to rod-only activation. B) Medium-intensity red that activates exclusively cones in Opn1mwR (above dopamine threshold) is not enough to cause significant light-induced dopamine release (although a trend is observed) whereas a control bright white light exhibits a normal increase (one-way ANOVA with Dunnett’s multiple comparison test, **p < 0.001, mean ± SEM, n = 3 each condition).

121

Figure 5.9 Dopamine release is driven by gap-junction dependent and independent pathways. A) In Gnat2A518G mice significant light-induced (~16 log rod photons cm-2s-1, 15min) dopamine release compared to dark-adapted conditions is achieved in intravitreally injected mice with saline. Blocking gap junctions with 1 mM MFA significantly reduces, but does not eliminate, light-induced dopamine release (t-test, #p < 0.05, **p < 0.001, mean ± SEM, n = 3 each condition). B) Similarly, application of 200 µM MFA in ex vivo retinae of Gnat2A518G mice significantly reduces light-induced dopamine levels and has no effect on dark-adapted retinae. Significant light-induced dopamine release is possible in presence of MFA (t-test, # p < 0.05, **p < 0.001, ***p < 0.0001, mean ± SEM, n ≥ 4 each). Stars indicate comparison at equivalent light conditions between MFA-treated and control mice, whereas hashtags indicate comparison between dark and light treatments within the same condition (control or MFA).

122

5.3. DISCUSSION

5.3.1. Rods define light-induced dopamine release and set its threshold

As previously hypothesised in the literature368, data from this chapter suggests that rods are

(surprisingly) the main photoreceptor responsible for driving light-induced dopamine release and setting its absolute release threshold with no need of additional photoreceptors. Also, rods alone modulate dopamine turnover under dim light implying that they ultimately control dopamine release over a very wide range of light intensities (> 9 log units). More importantly, these data indicate that electrophysiological responses from dopaminergic amacrine cells do not always correlate with actual dopamine being released. Rods driving, and defining, such a high-threshold response may, at first, seem counterintuitive but the idea that rods participate in vision at high irradiances has gained acceptance in the recent years22,23,25,598,599. For instance, the most recent paper from Tikidji-Hamburyan et al.28 showed that while rod contrast sensitivity initially dropped in response to bright light, it progressively recovered with a speed proportional to the given light intensity (i.e. occurred faster under brighter backgrounds).

Bleaching adaptation and gain modulation were suggested as the mechanisms underlying this adaptation since rod responses became more robust with increased exposure times and light intensities. Thus, it is not hard to believe that rods could be signalling at bright intensities and triggering light-induced dopamine release over a wide range of light intensities as suggested throughout this thesis. However, the question as to why dopamine is not released at lower intensities where rods are robustly active is hard to reconcile. It may be that the answer lies in the inhibitory component of rod input that I identified under dim red light.

Dopamine turnover under dim red light was reduced compared to infrared light dissection in wild-type, Gnat2A518G and Opn4Cre/Cre animals, suggesting it is the input of rods alone that triggers this decrease. Conversely, no statistical differences were found between the two

123 dissection conditions in Opn1mwR mice which could intuitively lead to the thought of cones being involved. However, the lack of significance occurs because in this particular model the standard deviation is ~4-fold higher than any other conditions (± 0.18 vs ± 0.04). Further, a power calculation estimates that tripling the n should provide significant differences between conditions (inference for means, α = 0.05, ß = 0.8, n = 15). In addition, dopamine turnover remained unchanged in both dim red and infrared dissections when rods and cones were absent

(rd/rd), but dopamine turnover levels matched those from dim red light dissection in wild-type mice. The retina of rd/rd animals is known to exhibit spontaneous activity in dark conditions600,601, specifically in ON bipolar cells602, which may trigger an equivalent suppressive effect as the one driven by rods in wild-type.

At brighter light intensities, Gnat2A518G mice exhibited equivalent responses as wild-type during the whole irradiance curve, suggesting that cones do not play a quantifiable role in light- induced dopamine release even at bright irradiances. Conversely, the literature showed that light-induced activation of cones caused either electrophysiological or c-fos activation of dopaminergic amacrine cells330,333. However, as discussed in the previous chapter, depolarisation of the soma does not necessarily correlate with depolarisation of the processes where dopamine is released. In addition to indirect measurements, dopamine release has been measured and suggested to be driven by light-activated cones in Gnat1-/- mice in different experimental designs: 1) a bright light pulse (~15 log rod photons cm-2 s-1, 90 min) caused a small (~25%) but significant increase in dopamine turnover in Gnat1-/- mice, although significantly reduced compared to wild-type animals368; and 2) a long mesopic light pulse (with a non-specified light intensity corresponding to the one of the animal housing from the animal facility during 4.5 hours) caused ~30% increase in dopamine turnover in Gnat1-/- Opn4-/- mice.

Gnat1-/- animals have been reported to retain residual rod responses454,455 and those experiments light-pulsed animals for much longer than reported throughout this thesis, either of which could

124 account for the small increases in light-induced dopamine responses. However, Gnat2A518G mice too could be an incomplete model of photoreceptor activation where phototransduction cascade activates downstream of the mutated transducin454. To overcome this, I showed that preferential spectral activation of cones in Opn1mwR mice did not elicit any significant change in dopamine turnover, further supporting hypotheses derived from Gnat2A518G results. Only when red or blue light wavelengths activated rods above the threshold, dopamine release was observed regardless of the mice model used (Figure 5.8). Finally, the role of S-cones in light- induced dopamine release was not explored in depth with exposure to UV light. However, given that the bright red pulse (red solid line, Figure 5.7) can elicit significant dopamine release means UV-cones are not necessary to drive this release. Conversely, a very recent paper showed light-induced dopamine release in chicks in response to UV light603, suggesting that those could play a role. However, bird and mammalian retinae differ significantly in cone expression and circuitry, and the influence of UV-cones may be specific to birds.

Taking advantage of the different degeneration timeline of rods and cones in the developing rd/rd retinae, I assessed light-induced dopamine release at different developmental time points to elucidate the role of each after or before the disappearance of the other. For example,

Greferath et al.127 showed that dopaminergic amacrine cells were still activated (c-fos expression) by light in rd mice until P20, even when rods had dissapeared127. At P22 only half of dopaminergic amacrine cells were activated by light in parallel with degeneration of cones127. Therefore, one might think that if cones were driving light-induce dopamine release, an increase in dopamine would occur at that point in response to light (as in the wild-type developing retina; Figure 4.7). However, this was not the case, showing once more that dopaminergic amacrine cell activation does not equal dopamine release. At this stage (~P20), bipolar cells from rd/rd mice lose their response to glutamate604, severe morphological changes are conspicuous605–608 and rod bipolar cells show an increased sensitivity to inhibitory

125 neurotransmitters592,609, so it is perhaps not surprising that dopaminergic amacrine cells cannot release dopamine in response to these altered inputs.

These results differ from Cameron et al.368 who found that nuclear c-fos was mainly absent in adult Gnat1-/- animals in response to light, although these two animal models differ in their rod ablation, in rd/rd mice rods are absent whereas in Gnat1-/- animals rods are present but in a permanent dark-adapted state. If a balance of activation of ON and OFF pathways controls c-fos activation, one might assume that these two models would produce opposing results.

Furthermore, Greferath et al.127 examined long term c-fos expression in the rd-FTL mouse as opposed to transient c-fos activation after a 90 min light pulse.

Despite the controversy in the literature regarding c-fos activation of dopaminergic amacrine cells, my results show that rd/rd mice were not able to release any dopamine in response to bright light at any developmental point. These results suggest that while the circuitry was mature enough to drive light-induced release changes in wild-type mice, when rods disappeared cone signalling was not able to drive light-induced dopamine release in rd/rd mice.

The finding that inputs from ipRGCs into dopaminergic amacrine cells327,330,333,339,458,586 did not elicit light-induced dopamine release at any given intensity was further evidence that electrophysiological response (at the soma) does not necessarily equate to dopamine release

(Figure 5.2). In Opn4-/- animals, ipRGCs are still present and intact in the retina (although lacking melanopsin) and therefore may still be involved in extrinsic signalling due to rod and/or cone inputs49,71,79,82,333. Although this activation is not entirely understood, different papers suggest classic photoreceptor input via bipolar cells → amacrine cells → ON cone bipolar cells

→ (via ectopic synapse) ipRGCs65,66,82,100,132,610–612. Therefore, rods and/or cone could signal though ipRGCs and pair with conventional rod inputs to elicit dopamine release. Examining

126 dopamine release in animals lacking ipRGCs (instead of just melanopsin; Opn4DTA/DTA) would shed more light on this hypothesis.

Because rd/rd mice were able to release dopamine in high potassium and Mg2+-free media

(Figure 5.3), the rationale to perform an irradiance response curve was that high irradiances may induce depolarisation block in ipRGCs (M1)613 preventing signals from passing to dopaminergic amacrine cells. However, no light-induced dopamine release could be observed at any intensity in rd/rd animals. Surprisingly, when picrotoxin was applied in the dark dopamine release remained the same, but it significantly increased in response to light + picrotoxin, suggesting that the GABAergic disinhibition caused by the drug must be paired with an excitatory photoreceptive input. The most likely candidate driving this light response is ipRGCs, although an input from a small population of cones that do not degenerate at this stage could also be responsible129. It is possible that GABAergic inputs prevent spiking in the soma of dopaminergic amacrine cells from propagating to the areas, likely dopaminergic amacrine cell processes614, where dopamine is released.

Further, since rd/rd mice have retinal rearrangement during development, I tried an alternative method to test the contribution of ipRGCs in non-degenerate retinae by chemogenetically inducing activation of ipRGCs. I show that while ipRGCs were consistently active in the dark in response to CNO (driving the PLR), no significant release of dopamine occurred (Figure

5.6). These results mirror the lack of light-induced dopamine release in the rd/rd animals, but represent a more physiological state as the Opn4Cre/+ mice that do not show any circuit reorganization. This, again, demonstrates that spiking in the soma of dopaminergic amacrine cells does not correlate with widespread dopamine release.

The results from the ex vivo irradiances match the shapes of the in vivo counterparts in

Gnat2A518G, Opn4Cre/Cre, and rd/rd mice. However, differences in the baseline levels of these

127 measurements can be observed in Figure 5.2 B. Although it is true that in the literature rd/rd animals showed smaller concentrations of dopamine compared to wild-type449, the most likely explanation for the variability in ex vivo experiments is due to increased number of steps during the processing of the retinae and detrimental physiological conditions as happened in the previous chapter. While the reduction in amplitude of dopamine turnover in Gnat2A518G and

Opn4Cre/Cre mice at higher irradiances could indicate a role of cones in light-induced dopamine release (via ipRGCs) after rods set the absolute threshold, the conclusions obtained from the rest of the experiments performed in this chapter suggest just inter-experimental variability.

Further, if values are normalised, one curve can fit all values when comparing irradiance responses of the dopamine light-inducible genotypes (except rd/rd), meaning the threshold is the same regardless of the method used. As discussed in the previous chapter, in vivo and ex vivo methods have positive and negative attributes, but overall provided consistent and complementary results throughout a wide variety of experimental designs and animal models.

In summary, a large amount of evidence strongly suggests that rods drive light-induced global dopamine release over a wide range of light intensities. Cones and/or ipRGCs may be contributing to local dopamine release that cannot be measured with the current used technique, but widespread, high amplitude dopamine release is governed exclusively by rod photoreceptor input. Very recently, Munteanu et al.615 showed that basal dopamine levels were reduced during development when animals were dark reared. Under normal light/dark rearing conditions knocking out rods recapitulated the effect of dark rearing but removal of ipRGCs or cones had no effect. This model agrees with the results obtained in Gnat2A518G, Opn1mwR, and rd/rd mice that neither cone nor melanopsin photoreception has an effect on the retinal dopamine system.

128

Figure 5.10 Widespread light-induced dopamine release is driven by rods. Rods (blue bar) drive widespread dopamine release under both very low and very high light intensities using different known and suggested pathways discussed throughout Chapter 5. On the other hand, cones (green scalene triangle) and ipRGCs (purple circle) provide an electrophysiological input into dopaminergic amacrine cells but without eliciting widespread light-induced dopamine release.

129

CHAPTER 6:

Effect of inflammation on retinal microglia and

dopamine release

130

6.1. INTRODUCTION

Inflammation is part of the immune response of the body that protects from infection, injury or toxin exposure. Pain, heat and redness occur due to an increase in vascular permeability and activation of immune cells at the location of the injury. Once the aberration is eliminated, the affected tissue returns to its normal physiology. When this process is quick, effective and beneficial it is known as an acute inflammatory episode. However, when these processes persist after the initial insult, it is known as chronic inflammation, and within the nervous system it is called chronic neuroinflammation. The most common and studied cases of dysregulation and overactivation of resident immune cells in the brain are microglia and astrocytes, which are often associated with many neurodegenerative and neurological diseases616–622.

In the retina, chronic neuroinflammation and more specifically activated microglia, are key components associated with retinal degenerative diseases that cause visual impairments such as macular degeneration, diabetic retinopathy, retinal vein occlusion, retinitis pigmentosa and glaucoma623–626. Common observations of these, and many other chronic neuroinflammation- related pathologies, are overactivation of microglia, excessive release of pro-inflammatory mediators and high oxidative stress, which ultimately results in neuronal loss627,628.

Nonetheless, microglia are necessary for a healthy retina and some of the physiological functions of microglia involve several roles in development629,630, blood vessel formation631 and immunological surveillance during adulthood632; where microglia are highly mobile and can be found in different layers, tending to be close to blood vessels633–635.

I explored microglia in the retinae of GFAP-IL6 mice, a mouse model of chronic neuroinflammation where interleukin 6 (IL-6) is constitutively expressed by glial fibrillary acidic protein (GFAP) cells. GFAP is an intermediate filament protein expressed in different

131 cells of the central nervous system including astrocytes. IL-6 is a pro-inflammatory cytokine that plays a role in the induction of microglial activation636 and its expression is globally altered in the retina of wild-type mice in response to stress637. For instance, in glaucoma, it is overregulated and found in high concentrations637–640. Under physiological conditions, some authors suggest that the increase in IL-6 occurs to stimulate neuritogenesis in retinal ganglion cell axons640,641. While the GFAP-IL6 mouse model has been widely used to study chronic neuroinflammation in the brain, microglia in the retina has never been examined. Most of the literature that describes microglia in the retina covers developmental phases634,642, intravitreal injections to mimic or stop acute inflammatory responses643–645 and different animal models of retinal diseases646–650.

Here, I present a yet unexplored mouse model of inflammation (regarding the retina), in which

I evaluate the impact of increased IL-6 on retinal microglial number and morphology, and how this impacts light-induced dopamine release. I characterised retinal microglia populations in both wild-type and GFAP-IL6 mice in both young (4 months) and aged animals (24 months), with the purpose to further to test if light-induced dopamine release has any correlation with microglial numbers and morphology. For instance, prolonged exposure to light causes activation of retinal microglia that is suggested to lead to degeneration of photoreceptors651,652.

Further, microglia express D1 and D2-like dopamine receptors653,654 that are suggested to mediate their chemotaxis towards dopaminergic cells in Parkinson’s disease655. Microglia are a double-edged sword in that they are beneficial when regulated properly, but can quickly enter into a feedback loop of increasingly detrimental inflammation. A good example of this delicate state of balance is that sustained microglial depletion results in the degeneration of photoreceptor synapses656, while overactivation triggers a sustained release of inflammatory factors that contribute to neuronal dysfunction and to pathology657. However, activation of microglia during retinal detachment actually protects photoreceptors from degenerating658.

132

Most of the literature evaluating retinal health during neuroinflammation reports ERG amplitude and implicit time, however, since both of these parameters have been shown to be correlated with dopamine release659 an impact of inflammation on dopamine release itself could, in turn, affect ERG parameters. Certainly, while ERG responses are diminished at different points in diabetic retinopathy660, retinal vein occlusion661, retinitis pigmentosa662, hypoxia663, macular degeneration664 and glaucoma665; a correlation with changes in dopamine release remains to be examined.

Therefore, in this chapter I aimed to:

1) Assess if the density and activation state of retinal microglia increase with

inflammation and age, as it occurs in different regions of the brain.

2) Assess if an increase in density and activated state microglia cause a decrease in

basal or light-induced dopamine release due to the disruption of one or more key

elements involved

133

6.2. RESULTS

6.2.1. Microglial density increases with age

Since GFAP-IL6 mice lacked characterisation of their retinal microglia, I performed immunohistochemistry against Iba1+ (ionized calcium-binding adapter molecule 1, a common marker for microglia) at 4 and 24 months of age in wild-type and GFAP-IL6 mice, and counted the density of the stained cells. During youth, GFAP-IL6 mice showed a significant increase of ~31% in microglial density (Iba1+ cells/ mm2) compared to wild-type mice (***t-test, p- value = 0.0004), suggesting that the transgenic model triggers proliferation of microglia (Figure

6.1). Surprisingly, the differences between GFAP-IL6 and littermate controls is dramatically reduced in aged animals with no significant differences between the two groups (t-test, p-value

= 0.45). However, when comparing the microglial density of young and aged groups to their respective genotypes, both are significantly increased: ~48% for wild-type, (***t-test, p-value

< 0.0001) and ~28% for GFAP-IL6 (**t-test, p-value = 0.0022; Figure 6.1). This suggests that:

1) microglial proliferation does not reach its maximum during youth regardless of chronic exposure to IL-6; and 2) microglial responsiveness to IL-6 may be depleted in aged animals or microglial numbers have simply reached their maximum by this age.

6.2.2. Microglia tend towards an inflammatory morphological phenotype with inflammation and age

Activation of microglia is known to correlate with its morphology666–668, with resting microglia exhibiting extended dendritic arbours and small somas whereas reactive microglia have retracted dendrites and larger somas. Thus, I used different approaches to study morphological changes in retinal microglia to determine their state in both wild-type and GFAP-IL6 mice in young and aged animals. First, I stained microglia for Iba1 and reconstructed n ≥ 63 cells from

134 each experimental group. Then, 1) I determined the general morphological aspects such as soma perimeter and area, convex hull perimeter and area (the perimeter or the area that the cell occupies in a 2D space), primary dendrites, branching points and total dendritic length (Figure

6.2); 2) I performed Sholl analysis to visualise the radial distribution of the quantified characteristics, providing information about the spatial arrangement of the dendrites (Figure

6.3 & 6.5); and 3) I performed bivariate correlations with the different morphological aspects quantified in the first aim against convex hull area to find any correlation patterns associated with a specific age and/or genotype (Figure 6.6).

All the commonly tested microglial morphological characteristics across the literature (soma area and perimeter, convex area and perimeter, primary dendrites, branching points and dendritic length) were analysed in the retinae of wild-type and GFAP-IL6 mice and plotted in

Figure 6.2. A graphical scheme of the characteristics tested on a reconstructed microglial cell is drawn in Figure 6.2 H. From the tested parameters, an increase in the soma size, number of primary dendrites and branching points, and a decrease in the convex area, convex perimeter and dendritic length are suggestive of microglial activation667.

I observed an increased soma area and number of primary dendrites, and a reduced convex perimeter and area in young GFAP-IL6 animals compared to littermate controls, suggesting microglia activation. These results are in agreement with the increase in microglial density at

4 months of age between GFAP-IL6 and their littermate controls. Conversely, when aged

GFAP-IL6 animals are compared to aged wild-types, the differences observed between both genotypes during youth are abolished. Further, wild-types exhibited the smallest convex perimeter and area, actually suggesting a more reactive phenotype than the chronically inflamed animals. When young and aged animals from the same genotype are compared,

GFAP-IL6 mice only differed in an increase of primary dendrites with age. However, aged wild-type mice exhibited an increase in soma size and primary dendrites, and a decrease in

135 convex perimeter, area and dendritic length. These results indicate that wild-type mice develop increased inflammatory responses with age, however GFAP-IL6 mice do not exhibit this increase in neuroinflammatory responses possibly due to a continuous activated state triggered by chronic exposure to IL-6.

6.2.3. Aging, but not IL-6 upregulation, causes retraction of microglial dendrites

To study the spatial distribution of the previous quantifications, Sholl analysis data (specific morphological characteristic quantified at every radial checkpoint, plotted in X and Y axis respectively; “Chapter 2: Materials and Methods”, Figure 2.3) is shown in Figure 6.3. Sholl analysis consists of tracing concentric circles that start in the centre of the soma with a fixed increased in the radius (in this case 3 µM), and quantification of the amount of a determined morphological characteristic (i.e. number of branching points) comprised between two successive circles. The morphological characteristics tested were: intersections (of dendrites with the concentric circles that start in the centre of the soma), branching points, dendritic endings, dendritic length, dendritic area and dendritic volume. From the resulting plots of these characteristics I quantified “the shift parameter” (radius at Ymax; Figure 6.3). This analysis provides an idea of the overall dendritic complexity (i.e. retracted or expanded) and shift at the maximum radii of complexity (distance from the soma). Representative images of microglia from each cohort can be visualised in Figure 6.4, where young wild-type microglia display an extended dendritic arbour against retracted dendrites in GFAP-IL6 and aged animals.

Young GFAP-IL6 microglia do not show a shift in complexity compared to littermate controls, however in wild-types the complexity shifts towards the soma with age. Aged GFAP-IL6 microglia, however, do not show this shift in complexity with age. In other words, the dendrites

136 of aged microglia are more retracted than young microglia in wild- type but not in GFAP-IL6 mice. However, a couple of results such as the reduction in Ymax with age in both dendritic volume and branching points of aged GFAP-IL6 microglia does suggest a morphology towards an amoeboid shaped form (i.e. fewer branching points near the soma). These results show that a retraction of dendrites occurs in normal aging, but not due to upregulation of IL-6, suggesting

GFAP-IL6 microglia do not display this normal phenotypic change with age.

6.2.4. Bivariate correlations of morphological characteristics with convex hull area

The next analysis tests bivariate correlations of soma perimeter and area, convex perimeter, dendritic length, branching points, and primary dendrites against the convex hull area (Figure

6.6; values for R2 and correlation p-values appear on Table 6.1). Convex hull area is the total area the microglia and its dendrites occupies, therefore a change in the correlation between this and the morphological characteristics tested, means the cell’s proportions have changed, and a phenotypic change may have occurred. Bivariate analysis suggests that young wild-type and

GFAP-IL6 microglia covering a larger area do not have larger somas, suggesting microglia that cover larger areas simply extend their dendrites further. However, microglia from aged mice have a positive correlation between soma size and convex hull area suggesting that the larger microglia have proportionately larger somas, indicating a phenotypic change with age (Figure

6.6 A & B). These results suggest a phenotypically different population of microglia that forms with age, but not with overexpression of IL-6. As may be expected, convex perimeter, dendritic length and number of branching points are strongly correlated with convex area in all groups

(i.e. the more area the microglia occupy the more perimeter). However, the slopes of the linear regression of aged wild-type, and both young and aged GFAP-IL6 microglia between convex

137 perimeter, dendritic length and branching points with the convex area are significantly increased, i.e. a more circular than elliptical geometric distribution, compared to young wild- type (Figure 6.6 C, D & F). Activated microglia generally show a more circular distribution, so these data indicate GFAP-IL6 and aging cause a switch to a more activated microglial state.

Primary dendrites do not show any correlation or slope variation among groups (Figure 6.6 E).

WT 4m GFAP-IL6 4m WT 24m GFAP-IL6 24m R2 0.05 0 0.06 0.07 Soma perimeter Correlation 0.06 0.9 0.05 0.04 R2 0.03 0 0.05 0.09 Soma area Correlation 0.13 0.82 0.08 0.02 R2 0.03 0 0.07 0.01 Primary dendrites Correlation 0.19 0.83 0.05 0.55 R2 0.47 0.72 0.71 0.69 Dendritic length Correlation <0.0001 <0.0001 <0.0001 <0.0001 R2 0.67 0.93 0.79 0.94 Convex perimeter Correlation <0.0001 <0.0001 <0.0001 <0.0001 R2 0.07 0.42 0.47 0.35 Branching points Correlation 0.04 <0.0001 <0.0001 <0.0001 Table 6.1. R2 values and Pearson r test statistics for bivariate correlations (Figure 6.6). Level of significance is expressed with p-value: *p<0.05, ***p<0.0001. WT: wild-type.

6.2.5. Light-induced dopamine release is normal in GFAP-IL6 mice

After the characterisation of microglia density and reactive-like state in different experimental models, I assessed light-induced dopamine release in young wild-type and GFAP-IL6 (Figure

6.7). Surprisingly, light-induced dopamine release is not impacted by chronic inflammation as

GFAP-IL6 and wild-type have indistinguishable dark and light-induced dopamine release (one- way ANOVA with Tukey’s multiple comparison test, p ≤ 0.84). These results suggest that a significant increase in numbers, and chronic activation of microglia during 4 months does not impair dopaminergic amacrine cell function, or the pre-synaptic circuitry conveying the light signal.

138

Figure 6.1 GFAP-IL6 mice have a significantly higher density of microglia compared to wild-type mice at 4 but not at 24 months. A significant increase in density occurs in 4 months old GFAP-IL6 mice compared to 4 months old wild-type mice. Iba1+ density is significantly increased at 24 months in both genotypes compared to their respective young genotypes, but there is no statistical difference between genotypes in aged animals. One-way ANOVA with Tukey’s multiple comparison test (**p < 0.001, ***p < 0.0001, mean ± SEM, 4 months, n = 4; 24 months, n = 3). WT: wild-type.

139

Figure 6.2 Microglia from all GFAP-IL6 cohorts and aged wild-type mice present reactive-like morphology. Significant differences between 4 months old GFAP-IL6 and wild-type mice are observed in A & B) soma, C & D) convex perimeter and area, and E) primary dendrites. F) Branching points remain constant across ages and genotypes. Overall, young GFAP-IL6 mice exhibit increased soma sizes, retracted dendrites and more primary dendrites. At 24 months, convex perimeter, convex area, and G) dendritic length are significantly reduced in wild-type compared to GFAP-IL6, which also remain true when compared to 4 months old wild-type, in addition to an increased soma area, soma perimeter, and primary dendrites. When aged GFAP-IL6 microglia are compared to its young counterpart, only primary dendrites are increased. H) Visual schematic of the tested characteristics: branching points (blue) are bifurcations of dendrites, primary dendrites (red) contact with the soma, soma perimeter/area (green), dendritic length (black) is the sum of the length of all dendrites, and convex hull area and perimeter are the space that the cell occupies (highlighted in grey) One-way ANOVA with Tukey’s multiple comparison test (*p < 0.05, **p < 0.001, ***p < 0.0001, mean ± SEM, n ≥ 63 each). WT: wild-type.

140

Figure 6.3 Age causes dendritic retraction towards the soma. Sholl analysis consists of overlaying concentric circles of an increasing fixed radii on the reconstructed microglia (centre at soma), and tests each specific morphological characteristic between two radius checkpoints (every 3 µm). The number of morphological characteristics comprised within the distance between the two perimeters of the circles is plotted in a XY graph. Tested morphological characteristics are: intersections, branching points, endings, dendritic length, area and volume. Aged microglia show an overall trend towards the soma, with more severity in wild-type microglia. Representative images of these graphs are presented in Figure 6.4. WT: wild-type.

141

Figure 6.4 Representative image of microglial retraction with age and inflammation. Visually, young wild-type microglia (top left) display spread dendrites and small somas, whereas aged (top right) and GFAP-IL6 (young: bottom left; aged: bottom right) microglia exhibit retracted dendrites and large somas suggestive of a reactive phenotype.

142

Figure 6.5 Aged microglia retract dendrites. A significant shift to the soma occurs in all dendritic characteristics except the number of endings and branching points in aged wild-type microglia compared to young. Strong similarities in young and aged GFAP-IL6 microglia are observed, except for an increase in dendritic volume and a decrease in branching points in aged animals. One-way ANOVA with Tukey’s multiple comparison test (**p > 0.001, ***p < 0.0001, mean ± SEM, n ≥ 63 each group). WT: wild-type.

143

Figure 6.6 Bivariate morphological correlations to Convex Hull Area in GFAP-IL6 and wild-type mice. Strong correlations are found in dendritic length, convex perimeter and branching points in all groups. The size of aged somas and primary dendrites correlate with cell size, but not in the young groups (Pearson r correlation test, *p < 0.05, ***p < 0.0001). Linear regression of aged wild-type is reduced compared to the rest of groups in branching points, dendritic length and convex perimeter, no differences are found among any other groups regarding linear regression (Extra-sum-of-squares F test, ***p < 0.0001 n ≥ 63 each group). WT: wild-type.

144

Figure 6.7 Light-induced dopamine release is not altered in GFAP-IL6 mice. A significant increase in light-induced (~16 log rod photons cm-2s-1, 15 min) dopamine release occurred in GFAP-IL6 retinae; lacking any significant difference compared to wild-type retinae in either light or dark conditions. One- way ANOVA with Tukey’s multiple comparison test (***p < 0.0001, mean ± SEM, n ≥ 3). WT: wild- type.

145

6.3. DISCUSSION

6.3.1. Chronic neuroinflammation diminishes the hallmarks of the microglial reactive phenotype in aged GFAP-IL6 mice

Experiments conducted in this chapter are the first to systematically quantify microglial density and morphology in the retina of GFAP-IL6 mice at two different ages. The most relevant finding is the similarity of microglial morphology across time in GFAP-IL6 retinae in contrast with the overactivated state of wild-type microglia in aged animals, having more activated-like phenotype than the chronically inflamed mice at the same age. Previous reports have shown an age-dependent increase in microglia density in the retina, and reduction of their dendritic field and length632,656,669. These results suggest that chronic activation of microglia throughout life may limit this normal age-dependent microglial phenotypic change. Certainly, overactivation of microglia has been shown to induce apoptosis in up to 40% of the microglia population670, which causes release of even more pro-inflammatory cytokines671. Also, the motility of aged microglia has been shown to be significantly reduced compared to young microglia632, which raise the question: where does the microglia that replaces the apoptotic casualties migrate from?

Very recently, Huang et al.672 found that the origin of the repopulated retinal microglia has extra-retinal origins (optic nerve and ciliary body), therefore, different microglial subtypes might populate the retina of chronically inflamed mice, suggesting further characterisation needs to be done to elucidate potential diverse populations. For example, staining of the fractalkine receptor CXC3R1 could provide some insight on microglia dynamics, since its expression correlates with repopulation673,674. This characterisation of microglia in the retina could provide more specific targets for therapeutic interventions in order to limit retinal chronic inflammation.

146

Further complementary approaches could be tested using different biomarkers other than Iba1 staining, such as staining for translocator protein or peripheral benzodiazepine receptor (TSPO) to quantify the percentage of microglia that is actually reactive675. Also, it would be interesting to test microglial dynamics in living tissue in response to an external insult in both young and aged GFAP-IL6 mice, to provide information about physiological responses to harmful stimuli.

One would expect decreased responses compared to aged wild-type mice if microglia from

GFAP-IL6 indeed have reduced responses after constant overactivation during a lifetime. This would support previous suggestions based on the morphological analysis. Because all the work conducted in this chapter to characterise microglia is based on Iba1 staining, the use of other indicators such as classical microglia markers (CD11b, CD14, CD68), purinergic receptors

(P2X4, P2X7, P2Y6, P2Y12), fractalkine receptors (CX3CR1) and cytokines (TGF-ß, IL-1ß,

TNF-α)676 could determine a more refined profile of the characteristics of the potential changed microglial population(s) of GFAP-IL6 animals. For instance, the production of nitric oxide

(triggered by an external insult) increases the expression of CD11b, and nitric oxide scavengers reduce its expression together with the loss of dopaminergic cells in the brain677. Therefore, are chronically inflamed retinae more susceptible to retinal damage related with light-induced nitric oxide release? This can be answered by quantifying the expression of CD14 and/or CD68, which correlate with tissue damage–associated signals678 and lysosomal activity679. Purinergic receptors modulate microglial proliferation, activation, cell death and cytokine release680–687.

Specifically, an increased expression in P2X7 is common in different neuroinflammatory conditions and promotes the release of the pro-inflammatory cytokine IL-1ß, which has trophic effects on surrounding microglia688 promoting chemokine expression by Müller cells and

RPE689. In summary, having access to a battery of available microglial markers, much work can be done to better characterise and understand the phenotypic change in microglia associated with long-term inflammation.

147

During youth, GFAP-IL6 mice showed an increased number of microglia that phenotypically seems to be reactive compared to wild-type mice, which provides a first step in the validation of the mouse model as chronically inflamed in the retina. However, this difference was abolished with aging likely because “normal” ageing in wild-type mice triggers an increase in microglia proliferation that ends matching the density of GFAP-IL6 mice. Regardless of the differences between their activated-like phenotypes, the similarities in numbers raise the question: Is there a maximum (or optimum) number of microglia/mm2 an inflamed retina can handle? Lipopolysaccharides could be added in retinal tissue of already inflamed animals or aged wild-type (because GFAP-IL6 might have exhausted strong inflammatory responses as previously discussed) in an attempt to further increase microglial density. If this was the maximum number of microglia an adult retina can handle, future studies could have a good absolute reference (microglia/mm2) to evaluate their numbers in response to retinal therapies and perhaps establish a scale depending on the severity of the symptoms with its correlation with microglial density. Some studies approached similar questions using nearest neighbour distance calculations (in the brain), which showed that microglia get closer with increasing age without ovelapping690,691, but covering a greater percentage of the retina.

These results may suggest an IL-6-independent component regarding the increase of cell density, since GFAP-IL6 retinae do not reach its maximum density at 4 months of age.

However, there might be the possibility that young retinae just do not receive as much IL-6 as the aged animals. In addition, whether blood circulating IL-6 increases with age is still a matter of debate, although most publications report an increase in both rodents and primates692–698 rather than invariant699 or decreased levels700. An enzyme-linked immunosorbent assay

(ELISA) could provide further information in the experimental models used701 by quantifying retinal, rather than blood-circulating IL-6.

148

The results obtained using bivariate correlations mainly show differences between young wild- type microglia and other groups. In half of the measurements (convex hull perimeter, dendritic length and branching points) both the slope of the linear regression and the R2 are increased when the rest of groups are compared to young wild-type. This further suggests that microglia in GFAP-IL6 and aged retinae show activated characteristics (i.e. a more circular than elliptical shape of dendritic field, and fewer branching points per unit area). Young healthy microglia are likely capable of performing surveillance functions in healthy retinae whereas activated microglia anchor (or highly reduce their motility) and retract their dendrites in the rest of the cohorts, as reported in the literature632,667,702,703. To distinguish between different magnitudes of activation, Davis et al.704 established a correlation between soma size and roundness.

Together with the previous suggested experiments, this could be a good approach to further determine the activation profile of microglia in GFAP-IL6 mice.

Sholl Analysis offers a wide range of analysis depending on what you are specifically looking for705. Here, I wanted to test overall differences indicative of reactive-like states, and just by looking at the raw data (Figure 6.3) the most obvious is dendritic retraction. Therefore, I proposed a measurement I named “shift parameter” (radius at Ymax) to reveal changes at the maximum peak of complexity. Indeed, my measurements show an overall retraction of reactive microglial dendrites that relate with both age and inflammation. Plus, both dendritic area and dendritic volume of aged GFAP-IL6 microglia shift towards the soma compared to young

GFAP-IL6, suggesting thicker dendrites that resemble to those of a hypertrophic microglia with amoeboid-shape706. This might reflect a state of dystrophic microglia deteriorated due to independent age-related processes (caused by chronic neuroinflammation) rather than enhanced reactivity.

To conclude, most studies focus on the characterisation of microglia in inflamed or diseased conditions, whereas much fewer groups studied microglia in healthy non-aged conditions707–

149

712. Due to the initial nature of the experiments to test light-induced dopamine release, no further characterisation of microglia beyond individual morphology and total cell density was performed for this thesis. Ongoing experiments will further characterise GFAP-IL6 retinae by conducting measurements such as microglial spatial distribution713,714 and a complete characterisation of astrocytes715,716 and Müller cells717,718, which has been shown to be related with microglial pro-inflammatory activation. Also, functional vision tests such as ERG or the optokinetic drum test, might reveal more cues about the retinal physiological differences between genotypes. However, a big handicap is that most of the microglial studies occur in rodent retinae, and the direct translation of these findings to primates should be taken with care since the retinal complexity differs and few papers have already reported contradictory results.

For instance, Owen et al.675 showed that following pro-inflammatory activation, the pro- inflammatory marker TSPO is upregulated in rodent retinae while in humans it exhibits the opposite. Also, microglia density and populations are different in the macula and periphery of primate retinae719, and rodents lack any foveal specialisation.

6.3.2. Chronic neuroinflammation does not impair light-induced dopamine release

At a preliminary stage, a significantly increased population of microglia does not seem to impact light-induced dopamine release. This data may point to a lack of involvement of microglia with any of the key players that mediate light induced dopamine release, or a compensatory mechanism at some point(s) of the circuitry to maintain normal dopamine release since no differences are observed between wild-type and GFAP-IL6 mice. However, using the ex vivo approach and normalising dopamine to sample protein content could provide some insight in the absolute variation of dopamine. Interestingly, a link between microglia and

150 dopamine has been suggested in the literature: microglia have dopamine receptors and use them to migrate towards its source653,654, and the number of microglia correlates with the number of

TH+ cells in the brain720. Further, induction of damage that triggers an increase in microglia numbers inversely correlates with TH+ cells, suggesting that those play a role in phagocytosis721–723. Upstream of dopaminergic amacrine cells, some authors showed interaction of microglia with the RPE that might impact on rod integrity and therefore light- induced dopamine release724,725. Further, microglia can have a direct impact on rods by active phagocytosis, which leads to retinal degeneration726–728. Therefore, much more work is needed in order to unravel the interaction between microglia and the circuits that drive light-induced dopamine release. Future staining of retinal photoreceptors, second and third order cells, will reveal differences in total number or morphology, perhaps suggesting a compensatory mechanism (i.e. retinal remodelling as it occurs in rd/rd mice) or no alteration at all because neurons might have developed certain resistance to a persistent phagocytic environment729.

Because the nature of the retinal inflammation project was a collaboration with another group, experiments assessing light-induced dopamine release in aged mice have not been performed yet. In the literature it is reported that the number of rods in aged wild-type retinae are not different to young730 but the number of TH+ cells are reduced with age731. Therefore, it would be interesting to measure light-induced dopamine release in aged animals to see if dopaminergic amacrine cells have developed a compensatory mechanism that allows them to release similar amounts of dopamine as young animals. Further experiments involving the staining of key photoresponsive components (i.e. rods, cones and ipRGCs) and ERG measurements in GFAP-IL6 mice will provide more information regarding the retinal physiology related with chronic neuroinflammation using a novel mouse model approach.

151

CHAPTER 7:

General Discussion

152

Dopamine plays multiple roles in the healthy physiology of the retina and dopaminergic amacrine cells are the only cells in the retina that synthesise dopamine, which is released in response to light. Photoreceptors are the neurons that sense light changes and respond with biochemical signals that can reach dopaminergic amacrine cells. However, which of these light inputs are sufficient and/or necessary to drive dopamine release? To answer that question, I optimised an analytical technique routinely used for more than 30 years to be able to measure precise variations in retinal dopamine release in a wide variety of mouse models and experimental approaches. First, I performed a systematic dissection of each methodological step that involves treatment, collection or analysis of retinae, providing critical evidence for future experimental designs. These results emphasise the potential differences that can be observed in the final amount of dopamine based on common variations of the protocol such as use of anaesthesia, intensity, duration and wavelength of light-pulses, techniques for retinal dissection (classical dissection or “popping”), illumination during retinal dissection (dim red or infrared), solvents for retinal homogenisation, type of quantification (ECD, UV or MS/MS) or type of internal standards (deuterated or non-deuterated). Further, I characterised two complementary techniques to assess dopamine release in different experimental designs: in vivo and ex vivo, and discussed pros and cons of each. An important drawback of these methods however is that each measurement only provides a snapshot of the dopamine turnover/release after a given amount of time. Alternatively, the use of amperometry319,320,457 (with a carbon fibre electrodes) in explanted retinae could provide a continuous temporal measure of dopamine release across different light exposures and intensities in each subject.

Ultimately, all the results from this thesis suggest that dopamine release is controlled by a balance between excitatory and inhibitory inputs, and rods are the main mediator for its release

153 under a wide range of light intensities. The simplest explanation for the mechanism by which rods are signalling at such bright intensities seems to be related to bleaching adaptation, since recent publications showed that rods can be functional at these bright light intensities26–28.

Additional experiments can be performed to provide further support to this hypothesis such as light-pulsing rod-only mice, that can be obtained crossing models used for the thesis

(Gnat2A518G x Opn4Cre/Cre). In fact, preliminary results with this model suggest that rods alone can drive light-induced dopamine release at the brightest intensity tested. In addition, mice are nocturnal and have rod-dominated retinae, but the use of a diurnal retina (cone-dominated) would provide valuable information about how, and if, retinal circuitry undergoes adaptation in these animals (given that dopamine release is rod-mediated). Can cones mediate dopamine release in these animals or is it still mediated by the smaller population of rods? Are the intrinsic photoresponsive properties of rods more important than its downstream circuitry? A good model to find an answer to some of this questions would be the use of a diurnal mammal with a cone rich retina.

While there is no evidence in this thesis to support the contribution of cones or ipRGCs in the normal functioning light-induced dopamine release, their assistance in local modulation cannot be entirely ruled out. Certainly, extensive evidence is provided across the literature describing inputs from these photoreceptors into dopaminergic amacrine cells at distinct light intensities that cause somatic depolarisation of dopaminergic amacrine cells. Such small and localised variations in dopamine release could be resolved with a dopamine dye732, or dopamine sensors genetically encoded within the cell733,734. However, the sensitivity of those might need to be improved to distinguish dopamine concentrations below 1 µM.

The next obvious question is what is the impact of this large increase in dopamine release at bright light intensities on retinal function? This could be investigated by examining the modulation of gap junctions in response to bright-light stimulation (above dopamine

154 threshold). Choi et al.735 used a technique called cut-loading, that consists of measuring the diffusion of a dye after cutting the retina with a blade. The more diffusion of the dye, the more gap junction coupling and vice versa.

Further, it would be interesting to perform a systematic study of other light-induced neuromodulators such as nitric oxide or adenosine. Certainly, nitric oxide and dopamine release are suggested to have complicated interactions183,736,737 and its release results in similar overall outcomes such as gap junction uncoupling or inhibition of ocular growth. Both nitric oxide and adenosine have been shown to modulate the sensitivity of ipRGCs738–740, and perhaps pharmacological regulation of these could modify the inputs from ipRGCs to dopaminergic amacrine cells causing dopamine release. Also, adenosine (which is released in the dark741) is related to circadian rhythms39 and exert noticeable effects on the ERG (increase a and b waves)742, suggesting that it may very well having an effect on dopaminergic pathways398,743,744. Plus, its analysis would be very similar to dopamine since HPLC methods are already established745,746.

Finally, as a result of a collaborative investigation with another group that studies neuroinflammation in the brain, I had the chance to characterise microglia from retinae of chronically inflamed animals and relate that with dopamine release. Although chronic neuroinflammation does not seem to impair the regulation of retinal pathways involved in bright light-induced dopamine release, much work remains to be done to fully characterise the impact of microglia on the retina and (specifically for this thesis) the components that participate in inputs to dopaminergic amacrine cells that drive light-induced dopamine release.

Nonetheless, this relation might elucidate novel dopamine-microglia interactions, that have been shown to be dysregulated in Parkinson’s disease747,748. Because microglia is suggested to be the main contributor of the inflammation in Parkinson’s disease749,750 and has been shown to migrate towards dopamine and selectively attack dopaminergic amacrine cells655, perhaps

155 light variations can potentially modify microglial dynamics. Currently, different approaches to treat the abnormalities found in GFAP-IL6 animals (i.e. reducing the number of activated microglia) to try to reverse the genotype back to a healthy wild-type are being performed in our lab. For example, a promising therapy that involves feeding GFAP-IL6 mice with different doses of potent antioxidants such as curcumin seem to reduce retinal microglia dose- dependently, although further physiological characterisation remains to be completed.

To conclude, I systematically examined the light inputs driving dopamine release in the retina and reached the conclusion that rod photoreceptor input defines dopamine release in the retina through both excitatory and inhibitory pathways. However, interdisciplinary collaborations and much work remains to be done to fully unravel the secrecies of dopamine physiology.

156

References

157

1. Tashiro, T. et al. Early trace of life from 3.95 Ga sedimentary rocks in Labrador, Canada. Nature 549, 516–518 (2017). 2. Lamb, T. D. Evolution of phototransduction, vertebrate photoreceptors and retina. Prog. Retin. Eye Res. 36, 52–119 (2013). 3. Nilsson, D.-E. The evolution of eyes and visually guided behaviour. Philos. Trans. R. Soc. B Biol. Sci. 364, 2833–2847 (2009). 4. Reynolds, G. T. & Lutz, R. A. Sources of Light in the Deep Ocean. Rev. Geophys. 39, 123–136 (2001). 5. Panda-Jonas, S., Jonas, J. B., Jakobczyk, M. & Schneider, U. Retinal Photoreceptor Count, Retinal Surface Area, and Optic Disc Size in Normal Human Eyes. Ophthalmology 101, 519–523 (1994). 6. Ferguson, L. R., Dominguez, J. M., Balaiya, S., Grover, S. & Chalam, K. V. Retinal Thickness Normative Data in Wild-Type Mice Using Customized Miniature SD-OCT. PLoS One (2013), doi: 10.1371/journal.pone.0067265. 7. Hendrickson, A. E. Primate foveal development: A microcosm of current questions in neurobiology. Investigative Ophthalmology and Visual Science 35, 3129–3133 (1994). 8. Packer, O., Hendrickson, A. E. & Curcio, C. A. Photoreceptor topography of the retina in the adult pigtail macaque (Macaca nemestrina). J. Comp. Neurol. 288, 165–183 (1989). 9. Jeon, C. J., Strettoi, E. & Masland, R. H. The major cell populations of the mouse retina. J Neurosci 18, 8936–8946 (1998). 10. Strauss, O. The Retinal Pigment Epithelium in Visual Function. Physiol. Rev. 85, 845– 881 (2005). 11. Erskine, L. & Herreral, E. Connecting the retina to the brain. ASN Neuro (2015). doi: 10.1177/1759091414562107 12. Nathans, J. Determinants of Visual Pigment Absorbance: Identification of the Retinylidene Schiff’s Base Counterion in Bovine Rhodopsin. Biochemistry 29, 9746– 9752 (1990). 13. Calderone, J. B. & Jacobs, G. H. Regional variations in the relative sensitivity to UV light in the mouse retina. Vis. Neurosci. 12, 463–8 (1995). 14. Applebury, M. L. et al. The murine cone photoreceptor: a single cone type expresses both S and M opsins with retinal spatial patterning. Neuron 27, 513–523 (2000). 15. Haverkamp, S. The Primordial, Blue-Cone Color System of the Mouse Retina. J. Neurosci. 25, 5438–5445 (2005).

158

16. Ortiń-Martínez, A. et al. Number and distribution of mouse retinal cone photoreceptors: Differences between an albino (Swiss) and a pigmented (C57/BL6) strain. PLoS One (2014). doi: 10.1371/journal.pone.0102392 17. Provencio, I., Jiang, G., de Grip, W. J., Hayes, W. P. & Rollag, M. D. Melanopsin: An opsin in melanophores, brain, and eye. Proc. Natl. Acad. Sci. U. S. A. 95, 340–345 (1998). 18. Emanuel, A. J. & Do, M. T. H. Melanopsin tristability for sustained and broadband phototransduction. Neuron 85, 1043–1055 (2015). 19. Bartl, F. J., Ritter, E. & Hofmann, K. P. Signaling States of Rhodopsin. J. Biol. Chem. 276, 30161–30166 (2001). 20. Lerea, C. L., Somers, D. E., Hurley, J. B., Klock, I. B. & Bunt-Milam, A. H. Identification of specific transducin alpha subunits in retinal rod and cone photoreceptors. Science (80-. ). 234, 77–80 (1986). 21. Deng, W.-T. et al. Functional interchangeability of rod and cone transducin alpha- subunits. Proc. Natl. Acad. Sci. U. S. A. 106, 17681–6 (2009). 22. Demontis, G. C., Bisti, S. & Cervetto, L. Light sensitivity, adaptation and saturation in mammalian rods. Prog. Brain Res. 95, 15–24 (1993). 23. Yin, L., Smith, R. G., Sterling, P. & Brainard, D. H. Chromatic Properties of Horizontal and Ganglion Cell Responses Follow a Dual Gradient in Cone Opsin Expression. J. Neurosci. 26, 12351–12361 (2006). 24. Altimus, C. M. et al. Rod photoreceptors drive circadian photoentrainment across a wide range of light intensities. Nat. Neurosci. 13, 1107–1112 (2010). 25. Naarendorp, F. et al. Dark Light, Rod Saturation, and the Absolute and Incremental Sensitivity of Mouse Cone Vision. J. Neurosci. 30, 12495–12507 (2010). 26. Vinberg, F., Turunen, T. T., Heikkinen, H., Pitkänen, M. & Koskelainen, A. A novel Ca 2+ -feedback mechanism extends the operating range of mammalian rods to brighter light. J. Gen. Physiol. 146, 307–21 (2015). 27. Pahlberg, J. et al. Voltage-sensitive conductances increase the sensitivity of rod photoresponses following pigment bleaching. J. Physiol. 595, 3459–3469 (2017). 28. Tikidji-Hamburyan, A. et al. Rods progressively escape saturation to drive visual responses in daylight conditions. Nat. Commun. (2017). doi: 10.6084/m9.figshare.5492626 29. Schnapf, J. L., Nunn, B. J., Meister, M. & Baylor, D. A. Visual transduction in cones of the monkey Macaca fascicularis. J. Physiol. 427, 681–713 (1990).

159

30. Fain, G. L., Hardie, R. & Laughlin, S. B. Phototransduction and the Evolution of Photoreceptors. Curr. Biol. 20, 114–124 (2010). 31. Sekaran, S. et al. 2-Aminoethoxydiphenylborane Is an Acute Inhibitor of Directly Photosensitive Retinal Ganglion Cell Activity In Vitro and In Vivo. J. Neurosci. 27, 3981–6 (2007). 32. Hartwick, A. T. E. et al. Light-evoked calcium responses of isolated melanopsin- expressing retinal ganglion cells. J. Neurosci. 27, 13468–80 (2007). 33. Sonoda, T., Lee, S. K., Birnbaumer, L. & Schmidt, T. M. Melanopsin Phototransduction Is Repurposed by ipRGC Subtypes to Shape the Function of Distinct Visual Circuits. Neuron 99, 754–767 (2018). 34. Jiang, Z., Yue, W. W. S., Chen, L., Sheng, Y. & Yau, K.-W. Cyclic-Nucleotide- and HCN-Channel-Mediated Phototransduction in Intrinsically Photosensitive Retinal Ganglion Cells. Cell 175, 652–664 (2018). 35. Do, M. T. H. & Yau, K.-W. Adaptation to steady light by intrinsically photosensitive retinal ganglion cells. Proc. Natl. Acad. Sci. U. S. A. 2013, 1–6 (2013). 36. Warren, E. J., Allen, C. N., Brown, R. L. & Robinson, D. W. Intrinsic light responses of retinal ganglion cells projecting to the circadian system. Eur. J. Neurosci. 17, 1727– 1735 (2003). 37. Wong, K. Y., Dunn, F. A. & Berson, D. M. Photoreceptor adaptation in intrinsically photosensitive retinal ganglion cells. Neuron 48, 1001–10 (2005). 38. Van Hook, M. J., Wong, K. Y. & Berson, D. M. Dopaminergic modulation of ganglion- cell photoreceptors in rat. Eur. J. Neurosci. 35, 507–518 (2012). 39. Ribelayga, C. & Mangel, S. C. A circadian clock and light/dark adaptation differentially regulate adenosine in the mammalian retina. J. Neurosci. 25, 215–222 (2005). 40. Pack, W., Hill, D. D. & Wong, K. Y. Melatonin modulates M4-type ganglion-cell photoreceptors. Neuroscience 303, 178–88 (2015). 41. Fu, Y. & Yau, K. W. Phototransduction in mouse rods and cones. Pflugers Archiv European Journal of Physiology 454, 805–819 (2007). 42. Yau, K. W., Matthews, G. & Baylor, D. A. Thermal activation of the visual transduction mechanism in retinal rods. Nature 279, 806–807 (1979). 43. Kefalov, V., Fu, Y., Marsh-Amstrong, N. & Yau, K. W. Role of visual pigment properties in rod and cone phototransduction. Nature 425, 526–531 (2003). 44. Rieke, F. & Baylor, D. Single-photon detection by rod cells of the retina. Rev. Mod. Phys. 70, 1027–1036 (1998).

160

45. Tinsley, J. N. et al. Direct detection of a single photon by humans. Nat. Commun. (2016). doi: 10.1038/ncomms12172 46. Kenkre, J. S., Moran, N. A., Lamb, T. D. & Mahroo, O. A. R. Extremely rapid recovery of human cone circulating current at the extinction of bleaching exposures. J. Physiol. 567, 95–112 (2005). 47. Sterling, P. & Matthews, G. Structure and function of ribbon synapses. Trends in Neurosciences 28, 20–29 (2005). 48. Raviola, E. & Gilula, N. B. Gap Junctions Between Photoreceptor Cells in the Vertebrate Retina. Proc. Natl. Acad. Sci. 70, 1677–1681 (1973). 49. Dacey, D. M. et al. Melanopsin-expressing ganglion cells in primate retina signal colour and irradiance and project to the LGN. Nature 433, 749–54 (2005). 50. Jusuf, P. R., Lee, S. C. S., Hannibal, J. & Grunert, U. Characterization and synaptic connectivity of melanopsin-containing ganglion cells in the primate retina. Eur. J. Neurosci. 26, 2906–2921 (2007). 51. Neumann, S., Haverkamp, S. & Auferkorte, O. N. Intrinsically photosensitive ganglion cells of the primate retina express distinct combinations of inhibitory neurotransmitter receptors. Neuroscience 199, 24–31 (2011). 52. Ecker, J. L. et al. Melanopsin-expressing retinal ganglion-cell photoreceptors: Cellular diversity and role in pattern vision. Neuron 67, 49–60 (2010). 53. Hu, C., Hill, D. D. & Wong, K. Y. Intrinsic physiological properties of the five types of mouse ganglion-cell photoreceptors. J Neurophysiol 109, 1876–1889 (2013). 54. Reifler, A. N. et al. The rat retina has five types of ganglion-cell photoreceptors. Exp. Eye Res. 130, 17–28 (2015). 55. Quattrochi, L. E. et al. The M6 cell: A small-field bistratified photosensitive retinal ganglion cell. J. Comp. Neurol. (2018). doi: 10.1002/cne.24556 56. Schmidt, T. M., Chen, S. K. & Hattar, S. Intrinsically photosensitive retinal ganglion cells: Many subtypes, diverse functions. Trends in Neurosciences 34, 572–580 (2011). 57. Do, M. T. H. et al. Photon capture and signalling by melanopsin retinal ganglion cells. Nature 457, 281–287 (2009). 58. Provencio, I., Rollag, M. D. & Castrucci, A. M. Photoreceptive net in the mammalian retina. This mesh of cells may explain how some blind mice can still tell day from night. Nature (2002). doi: 10.1038/415493a 59. Berson, D. M., Dunn, F. a & Takao, M. Phototransduction by retinal ganglion cells that set the circadian clock. Science 295, 1070–1073 (2002).

161

60. Zhao, X., Stafford, B. K., Godin, A. L., King, W. M. & Wong, K. Y. Photoresponse diversity among the five types of intrinsically photosensitive retinal ganglion cells. J Physiol 592, 1619–1636 (2014). 61. Hattar, S. et al. Central projections of melanopsin-expressing retinal ganglion cells in the mouse. J. Comp. Neurol. 497, 326–349 (2006). 62. Baver, S. B., Pickard, G. E., Sollars, P. J. & Pickard, G. E. Two types of melanopsin retinal ganglion cell differentially innervate the hypothalamic suprachiasmatic nucleus and the olivary pretectal nucleus. Eur. J. Neurosci. 27, 1763–1770 (2008). 63. Wong, K. Y., Dunn, F. a, Graham, D. M. & Berson, D. M. Synaptic influences on rat ganglion-cell photoreceptors. J. Physiol. 582, 279–296 (2007). 64. Schmidt, T. M. & Kofuji, P. Functional and morphological differences among intrinsically photosensitive retinal ganglion cells. J. Neurosci. 29, 476–482 (2009). 65. Guler, A. D. et al. Melanopsin cells are the principal conduits for rod-cone input to non- image-forming vision. Nature 453, 102–105 (2008). 66. Dumitrescu, O. N., Pucci, F. G., Wong, K. Y. & Berson, D. M. Ectopic retinal ON bipolar cell synapses in the OFF inner plexiform layer: Contacts with dopaminergic amacrine cells and melanopsin ganglion cells. J. Comp. Neurol. 517, 226–244 (2009). 67. Hoshi, H., Liu, W.-L., Massey, S. C. & Mills, S. L. ON inputs to the OFF layer: bipolar cells that break the stratification rules of the retina. J. Neurosci. 29, 8875–83 (2009). 68. Østergaard, J., Hannibal, J. & Fahrenkrug, J. Synaptic contact between melanopsin- containing retinal ganglion cells and rod bipolar cells. Investig. Ophthalmol. Vis. Sci. 48, 3812–3820 (2007). 69. Belenky, M. A., Smeraski, C. A., Provencio, I., Sollars, P. J. & Pickard, G. E. Melanopsin retinal ganglion cells receive bipolar and amacrine cell synapses. J. Comp. Neurol. 460, 380–393 (2003). 70. Viney, T. J. et al. Local Retinal Circuits of Melanopsin-Containing Ganglion Cells Identified by Transsynaptic Viral Tracing. Curr. Biol. 17, 981–988 (2007). 71. Perez-Leon, J. A., Warren, E. J., Allen, C. N., Robinson, D. W. & Brown, R. L. Synaptic inputs to retinal ganglion cells that set the circadian clock. Eur. J. Neurosci. 24, 1117– 1123 (2006). 72. Vugler, A. A. et al. Dopamine neurones form a discrete plexus with melanopsin cells in normal and degenerating retina. Exp. Neurol. 205, 26–35 (2007). 73. Kosaka, T., Tauchi, M. & Dahl, J. L. Cholinergic neurons containing GABA-like and/or glutamic acid decarboxylase-like immunoreactivities in various brain regions of the rat.

162

Exp. Brain Res. 70, 605–617 (1988). 74. Wässle, H. & Chun, M. H. Dopaminergic and indoleamine-accumulating amacrine cells express GABA-like immunoreactivity in the cat retina. J Neurosci 8, 3383–94 (1988). 75. Feigenspan, a, Gustincich, S., Bean, B. P. & Raviola, E. Spontaneous activity of solitary dopaminergic cells of the retina. J. Neurosci. 18, 6776–6789 (1998). 76. Contini, M. & Raviola, E. GABAergic synapses made by a retinal dopaminergic neuron. Proc. Natl. Acad. Sci. U. S. A. 100, 1358–1363 (2003). 77. Zhang, D.-Q., Zhou, T.-R. & McMahon, D. G. Functional heterogeneity of retinal dopaminergic neurons underlying their multiple roles in vision. J. Neurosci. 27, 692– 699 (2007). 78. Sakamoto, K. et al. Dopamine regulates melanopsin mRNA expression in intrinsically photosensitive retinal ganglion cells. Eur. J. Neurosci. 22, 3129–3136 (2005). 79. Weng, S., Estevez, M. E. & Berson, D. M. Mouse Ganglion-Cell Photoreceptors Are Driven by the Most Sensitive Rod Pathway and by Both Types of Cones. PLoS One (2013). doi: 10.1371/journal.pone.0066480 80. Milosavljevic, N., Cehajic-Kapetanovic, J., Procyk, C. A. & Lucas, R. J. Chemogenetic Activation of Melanopsin Retinal Ganglion Cells Induces Signatures of Arousal and/or Anxiety in Mice. Curr. Biol. 26, 2358–2363 (2016). 81. Schmidt, T. M. & Kofuji, P. Functional and morphological differences among intrinsically photosensitive retinal ganglion cells. J. Neurosci. 29, 476–482 (2009). 82. Wong, K. Y., Dunn, F. A., Graham, D. M. & Berson, D. M. Synaptic influences on rat ganglion-cell photoreceptors. J. Physiol. 582, 279–296 (2007). 83. Schmidt, T. M. & Kofuji, P. Structure and function of bistratified intrinsically photosensitive retinal ganglion cells in the mouse. J. Comp. Neurol. 519, 1492–1504 (2011). 84. Hattar, S., Liao, H. W., Takao, M., Berson, D. M. & Yau, K. W. Melanopsin-containing retinal ganglion cells: architecture, projections, and intrinsic photosensitivity. Science 295, 1065–70 (2002). 85. Hannibal, J. & Fahrenkrug, J. Melanopsin containing retinal ganglion cells are light responsive from birth. Neuroreport 15, 2317–2320 (2004). 86. Barnard, A. R., Hattar, S., Hankins, M. W. & Lucas, R. J. Melanopsin regulates visual processing in the mouse retina. Curr. Biol. 16, 389–395 (2006). 87. Brown, T. M. et al. Melanopsin contributions to irradiance coding in the thalamo- cortical visual system. PLoS Biol. (2010). doi: 10.1371/journal.pbio.1000558

163

88. Hannibal, J. et al. Central projections of intrinsically photosensitive retinal ganglion cells in the macaque monkey. J. Comp. Neurol. 522, 2231–2248 (2014). 89. Schmidt, T. M. et al. A Role for Melanopsin in Alpha Retinal Ganglion Cells and Contrast Detection. Neuron 82, 781–788 (2014). 90. Semo, M. et al. Melanopsin retinal ganglion cells and the maintenance of circadian and pupillary responses to light in aged rodless/coneless (rd/rd cl) mice. Eur. J. Neurosci. 17, 1793–1801 (2003). 91. Zhu, Y. et al. Melanopsin-dependent persistence and photopotentiation of murine pupillary light responses. Investig. Ophthalmol. Vis. Sci. 48, 1268–1275 (2007). 92. Göz, D. et al. Targeted destruction of photosensitive retinal ganglion cells with a saporin conjugate alters the effects of light on mouse circadian rhythms. PLoS One (2008). doi: 10.1371/journal.pone.0003153 93. Hatori, M. et al. Inducible ablation of melanopsin-expressing retinal ganglion cells reveals their central role in non-image forming visual responses. PLoS One (2008). doi: 10.1371/journal.pone.0002451 94. Mohawk, J. A., Green, C. B. & Takahashi, J. S. Central and Peripheral Circadian Clocks in Mammals. Annu. Rev. Neurosci. 35, 445–462 (2012). 95. Panda, S. et al. Melanopsin (Opn4) requirement for normal light-induced circadian phase shifting. Science (80-. ). 298, 2213–2216 (2002). 96. Dkhissi-Benyahya, O., Gronfier, C., De Vanssay, W., Flamant, F. & Cooper, H. M. Modeling the Role of Mid-Wavelength Cones in Circadian Responses to Light. Neuron 53, 677–687 (2007). 97. Buhr, E. D. et al. Neuropsin (OPN5)-mediated photoentrainment of local circadian oscillators in mammalian retina and cornea. Proc. Natl. Acad. Sci. 112, 13093–8 (2015). 98. Zaidi, F. H. et al. Short-Wavelength Light Sensitivity of Circadian, Pupillary, and Visual Awareness in Humans Lacking an Outer Retina. Curr. Biol. 17, 2122–2128 (2007). 99. Panda, S. et al. Melanopsin is required for non-image-forming photic responses in blind mice. Science 301, 525–527 (2003). 100. Villa, K. L. et al. Rods-cones and melanopsin detect light and dark to modulate sleep independent of image formation. Proc. Natl. Acad. Sci. U. S. A. 105, 19998–20003 (2008). 101. Tu, D. C. et al. Physiologic diversity and development of intrinsically photosensitive retinal ganglion cells. Neuron 48, 987–999 (2005). 102. Brown, T. M. et al. Melanopsin-based brightness discrimination in mice and humans.

164

Curr. Biol. 22, 1134–1141 (2012). 103. Davis, K. E., Eleftheriou, C. G., Allen, A. E., Procyk, C. A. & Lucas, R. J. Melanopsin- derived visual responses under light adapted conditions in the mouse dLGN. PLoS One (2015). doi: 10.1371/journal.pone.0123424 104. Allen, A. E. et al. Melanopsin-driven light adaptation in mouse vision. Curr. Biol. 24, 2481–2490 (2014). 105. Reese, B. E. Development of the Retina and Optic Pathway. Vision Res. 51, 613–632 (2010). 106. Cepko, C. L., Austin, C. P., Yang, X., Alexiades, M. & Ezzeddine, D. Cell fate determination in the vertebrate retina. Proc. Natl. Acad. Sci. U. S. A. 93, 589–595 (1996). 107. LaVail, M. M. Kinetics of rod outer segment renewal in the developing mouse retina. J. Cell Biol. 58, 650–661 (1973). 108. Tarttelin, E. E. et al. Expression of opsin genes early in ocular development of humans and mice. Exp. Eye Res. 76, 393–396 (2003). 109. Rao, S. et al. A direct and melanopsin-dependent fetal light response regulates mouse eye development. Nature 494, 243–246 (2013). 110. Ratto, G. M., Robinson, D. W., Yan, B. & McNaughton, P. A. Development of the light response in neonatal mammalian rods. Nature 351, 654–657 (1991). 111. Bakall, B. et al. Expression and localization of bestrophin during normal mouse development. Invest Ophthalmol.Vis.Sci 44, 3622–3628 (2003). 112. Sekaran, S. et al. Melanopsin-dependent photoreception provides earliest light detection in the mammalian retina. Curr. Biol. 15, 1099–1107 (2005). 113. Schmidt, T. M., Taniguchi, K. & Kofuji, P. Intrinsic and extrinsic light responses in melanopsin-expressing ganglion cells during mouse development. J. Neurophysiol. 100, 371–384 (2008). 114. Sherry, D. M., Wang, M. M., Bates, J. & Frishman, L. J. Expression of vesicular glutamate transporter 1 in the mouse retina reveals temporal ordering in development of rod vs. Cone and ON vs. OFF circuits. J. Comp. Neurol. 465, 480–498 (2003). 115. Johnson, J. et al. Vesicular Glutamate Transporter 1 Is Required for Photoreceptor Synaptic Signaling But Not For Intrinsic Visual Functions. J. Neurosci. 27, 7245–7255 (2007). 116. Nomura, A. et al. Developmentally regulated postsynaptic localization of a metabotropic glutamate receptor in rat rod bipolar cells. Cell 77, 361–369 (1994). 117. Huberman, A. D., Feller, M. B. & Chapman, B. Mechanisms Underlying Development

165

of Visual Maps and Receptive Fields. Annu. Rev. Neurosci. 31, 479–509 (2008). 118. Ford, K. J. & Feller, M. B. Assembly and disassembly of a retinal cholinergic network. Vis. Neurosci. 29, 61–71 (2012). 119. Firl, A., Sack, G. S., Newman, Z. L., Tani, H. & Feller, M. B. Extrasynaptic glutamate and inhibitory neurotransmission modulate ganglion cell participation during glutamatergic retinal waves. J. Neurophysiol. 109, 1969–78 (2013). 120. Marquardt, T. & Gruss, P. Generating neuronal diversity in the retina: One for nearly all. Trends in Neurosciences 25, 32–38 (2002). 121. Fisher, L. J. Development of synaptic arrays in the inner plexiform layer of neonatal mouse retina. J. Comp. Neurol. 187, 359–372 (1979). 122. Hoon, M., Okawa, H., Della Santina, L. & Wong, R. O. L. Functional architecture of the retina: Development and disease. Prog. Retin. Eye Res. 42, 44–84 (2014). 123. Rosa, J. M. et al. Neuron-glia signaling in developing retina mediated by neurotransmitter spillover. Elife (2015). doi:10.7554/eLife.09590 124. Dick, O. et al. The presynaptic active zone protein bassoon is essential for photoreceptor ribbon synapse formation in the retina. Neuron 37, 775–786 (2003). 125. Raven, M. A. et al. Early afferent signaling in the outer plexiform layer regulates development of horizontal cell morphology. J. Comp. Neurol. 506, 745–758 (2008). 126. LaVail, M. M. & Sidman, R. L. C57BL/6J Mice With inherited Retinal Degeneration. Arch. Ophthalmol. 91, 394–400 (1974). 127. Greferath, U. et al. Mapping retinal degeneration and loss-of-function in Rd-FTL mice. Investig. Ophthalmol. Vis. Sci. 50, 5955–5964 (2009). 128. Smeyne, R. J. et al. Continuous c-fos expression precedes programmed cell death in vivo. Nature 363, 166–169 (1993). 129. Carter-Dawson, L. D., LaVail, M. M. & Sidman, R. L. Differential effect of the rd mutation on rods and cones in the mouse retina. Investig. Ophthalmol. Vis. Sci. 17, 489– 498 (1978). 130. Euler, T., Haverkamp, S., Schubert, T. & Baden, T. Retinal bipolar cells: elementary building blocks of vision. Nat. Rev. Neurosci. 15, 507–519 (2014). 131. Lamb, T. D. Evolution of vertebrate retinal photoreception. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 364, 2911–2924 (2009). 132. Hoshi, H., Liu, W. L., Massey, S. C. & Mills, S. L. ON inputs to the OFF layer: bipolar cells that break the stratification rules of the retina. J Neurosci 29, 8875–8883 (2009). 133. Masu, M. et al. Specific deficit of the ON response in visual transmission by targeted

166

disruption of the mGIuR6 gene. Cell 80, 757–765 (1995). 134. Morgans, C. W., Brown, R. L. & Duvoisin, R. M. TRPM1: The endpoint of the mGluR6 signal transduction cascade in retinal ON-bipolar cells. BioEssays 32, 609–614 (2010). 135. Slaughter, M. M. & Miller, R. F. 2-Amino-4-phosphonobutyric acid: A new pharmacological tool for retina research. Science (80-. ). 211, 182–5 (1981). 136. Maple, B. R., Gao, F. & Wu, S. M. Glutamate receptors differ in rod- and cone- dominated off-center bipolar cells. Neuroreport 10, 3605–3610 (1999). 137. Wässle, H., Haverkamp, S. & Grünert, U. The cone pedicle a complex synapse in the retina. Keio J. Med. 51, 19–20 (2002). 138. Cadetti, L., Tranchina, D. & Thoreson, W. B. A comparison of release kinetics and glutamate receptor properties in shaping rod-cone differences in EPSC kinetics in the salamander retina. J. Physiol. 569, 773–788 (2005). 139. Calkins, D. J. Localization of ionotropic glutamate receptors to invaginating dendrites at the cone synapse in primate retina. Vis. Neurosci. 22, 469–477 (2005). 140. Klooster, J., Yazulla, S. & Kamermans, M. Ultrastructural analysis of the glutamatergic system in the outer plexiform layer of zebrafish retina. J. Chem. Neuroanat. 37, 254– 265 (2009). 141. Lindstrom, S. H., Ryan, D. G., Shi, J. & Devries, S. H. Kainate receptor subunit diversity underlying response diversity in retinal Off bipolar cells. J Physiol 7, 1457–77 (2014). 142. Puthussery, T. et al. Kainate receptors mediate synaptic input to transient and sustained OFF visual pathways in primate retina. J. Neurosci. 34, 7611–21 (2014). 143. DeVries, S. H. Bipolar cells use kainate and AMPA receptors to filter visual information into separate channels. Neuron 28, 847–856 (2000). 144. Borghuis, B. G., Looger, L. L., Tomita, S. & Demb, J. B. Kainate receptors mediate signaling in both transient and sustained OFF bipolar cell pathways in mouse retina. J. Neurosci. 34, 6128–6139 (2014). 145. Hensley, S. H., Yang, X. L. & Wu, S. M. Identification of glutamate receptor subtypes mediating inputs to bipolar cells and ganglion cells in the tiger salamander retina. J Neurophysiol 69, 2099–2107 (1993). 146. Cull-Candy, S., Brickley, S. & Farrant, M. NMDA receptor subunits: diversity, development and disease. Curr. Opin. Neurobiol. 11, 327–335 (2001). 147. Sucher, N. J. et al. N-methyl-D-aspartate receptor subunit NR3A in the retina: developmental expression, cellular localization, and functional aspects. Invest. Ophthalmol. Vis. Sci. 44, 4451–4456 (2003).

167

148. Prybylowski, K. & Wenthold, R. J. N-Methyl-D-aspartate Receptors: Subunit Assembly and Trafficking to the Synapse. Journal of Biological Chemistry 279, 9673–9676 (2004). 149. Ruppersberg, J. P., Kitzing, E. V. & Schoepfer, R. The mechanism of magnesium block of NMDA receptors. Semin. Neurosci. 6, 87–96 (1994). 150. Amico-Ruvio, S. A., Murthy, S. E., Smith, T. P. & Popescu, G. K. Zinc effects on NMDA receptor gating kinetics. Biophys. J. 100, 1908–10 (2011). 151. Mayer, M. L., Westbrook, G. L. & Guthrie, P. B. Voltage-dependent block by Mg2+ of NMDA responses in spinal cord neurones. Nature 309, 261–263 (1984). 152. Johnson, J. W. & Ascher, P. Glycine potentiates the NMDA response in cultured mouse brain neurons. Nature 325, 529–531 (1987). 153. Kleckner, N. W. & Dingledine, R. Requirement for glycine in activation of NMDA receptors expressed in xenopus oocytes. Science 241, 835–837 (1988). 154. McBain, C. J. & Mayer, M. L. N-methyl-D-aspartic acid receptor structure and function. Physiol. Rev. 74, 723–760 (1994). 155. Fletcher, E. L., Hack, I., Brandstätter, J. H. & Wässle, H. Synaptic localization of NMDA receptor subunits in the rat retina. J. Comp. Neurol. 420, 98–112 (2000). 156. Kalloniatis, M., Sun, D., Foster, L., Haverkamp, S. & Wässle, H. Localization of NMDA receptor subunits and mapping NMDA drive within the mammalian retina. Vis. Neurosci. 21, 587–597 (2004). 157. Goebel, D. J., Aurelia, J. L., Tai, Q., Jojich, L. & Poosch, M. S. Immunocytochemical localization of the NMDA-R2A receptor subunit in the cat retina. Brain Res. 808, 141– 154 (1998). 158. Grünert, U., Lin, B. & Martin, P. R. Glutamate receptors at bipolar synapses in the inner plexiform layer of primate retina: Light microscopic analysis. J. Comp. Neurol. 466, 136–147 (2003). 159. Brandstatter, J. H., Hartveit, E., Sassoe-Pognetto, M. & Wassle, H. Expression of NMDA and high-affinity kainate receptor subunit mRNAs in the adult rat retina. Eur J Neurosci 6, 1100–1112 (1994). 160. Wenzel, A., Benke, D., Mohler, H. & Fritschy, J. M. N-methyl-D-aspartate receptors containing the NR2D subunit in the retina are selectively expressed in rod bipolar cells. Neuroscience 78, 1105–1112 (1997). 161. Muller, F., Greferath, U., Wassle, H., Wisden, W. & Seeburg, P. Glutamate receptor expression in the rat retina. Neurosci Lett 138, 179–182 (1992).

168

162. Hughes, T. E. Are there ionotropic glutamate receptors on the rod bipolar cell of the mouse retina? Vis. Neurosci. 14, 103–109 (1997). 163. Lo, W., Molloy, R. & E. Hughes, T. Ionotropic glutamate receptors in the retina: Moving from molecules to circuits. in Vision Research 38, 1399–1410 (1998). 164. Sarantis, M., Everett, K. & Attwell, D. A presynaptic action of glutamate at the cone output synapse. Nature 332, 451–453 (1988). 165. Eliasof, S. & Werblin, F. Characterization of the glutamate transporter in retinal cones of the tiger salamander. J. Neurosci. 13, 402–411 (1993). 166. Thoreson, W. B. & Miller, R. F. Membrane currents evoked by excitatory amino acid agonists in ON bipolar cells of the mudpuppy retina. J. Neurophysiol. 70, 1326–1338 (1993). 167. Euler, T., Schneider, H. & Wässle, H. Glutamate responses of bipolar cells in a slice preparation of the rat retina. J. Neurosci. 16, 2934–2944 (1996). 168. Sasaki, T. & Kaneko, A. L-glutamate-induced responses in OFF-type bipolar cells of the cat retina. Vision Res. 36, 787–795 (1996). 169. Hartveit, E. Functional organization of cone bipolar cells in the rat retina. J. Neurophysiol. 77, 1716–30 (1997). 170. Vandenbranden, C. a, Kamphuis, W., Nunes Cardozo, B. & Kamermans, M. Expression and localization of ionotropic glutamate receptor subunits in the goldfish retina an in situ hybridization and immunocytochemical study. J. Neurocytol. 29, 729–42 (2000). 171. Karschin, A. & Wässle, H. Voltage- and transmitter-gated currents in isolated rod bipolar cells of rat retina. J. Neurophysiol. 63, 860–876 (1990). 172. Gonzälez-soriano, J. Morphological types of horizontal cell in rodent retinae: A comparison of rat, mouse, gerbil, and guinea pig. Vis. Neurosci. 11, 501–517 (1994). 173. Trumpler, J. et al. Rod and cone contributions to horizontal cell light responses in the mouse retina. J Neurosci 28, 6818–6825 (2008). 174. Dacey, D. M., Lee, B. B., Stafford, D. K., Pokomy, J. & Smith, V. C. Horizontal cells of the primate retina: Cone specificity without spectral opponency. Science (80-. ). 271, 656–659 (1996). 175. Goodchild, A. K., Chan, T. L. & Grünert, U. Horizontal cell connections with short- wavelength-sensitive cones in macaque monkey retina. Vis. Neurosci. 13, 833–845 (1996). 176. Li, Y. N., Matsui, J. I. & Dowling, J. E. Specificity of the horizontal cell-photoreceptor connections in the zebrafish (Danio rerio) retina. J. Comp. Neurol. 516, 442–453 (2009).

169

177. Bloomfield, S. a & Miller, R. F. A physiological and morphological study of the horizontal cell types of the rabbit retina. J. Comp. Neurol. 208, 288–303 (1982). 178. Reese, B. E. Afferents and Homotypic Neighbors Regulate Horizontal Cell Morphology, Connectivity, and Retinal Coverage. J. Neurosci. 25, 2167–2175 (2005). 179. Hombach, S. et al. Functional expression of connexin57 in horizontal cells of the mouse retina. Eur. J. Neurosci. 19, 2633–40 (2004). 180. Djamgoz, M. B. et al. Light-adaptive role of nitric oxide in the outer retina of lower vertebrates: a brief review. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 355, 1199–1203 (2000). 181. Witkovsky, P. Dopamine and retinal function. Documenta Ophthalmologica 108, 17– 40 (2004). 182. Eldred, W. D. & Blute, T. A. Imaging of nitric oxide in the retina. in Vision Research 45, 3469–3486 (2005). 183. Sekaran, S., Cunningham, J., Neal, M. J., Hartell, N. A. & Djamgoz, M. B. A. Nitric oxide release is induced by dopamine during illumination of the carp retina: Serial neurochemical control of light adaptation. Eur. J. Neurosci. 21, 2199–2208 (2005). 184. Sato, M., Ohtsuka, T. & Stell, W. K. Endogenous nitric oxide enhances the light- response of cones during light-adaptation in the rat retina. Vision Res. 51, 131–137 (2011). 185. Lampe, P. D. & Lau, a F. Regulation of gap junctions by phosphorylation of connexins. Arch. Biochem. Biophys. 384, 205–15 (2000). 186. Lampe, P. D. & Lau, A. F. The effects of connexin phosphorylation on gap junctional communication. International Journal of Biochemistry and Cell Biology 36, 1171–1186 (2004). 187. Godley, B. F. & Wurtman, R. J. Release of endogenous dopamine from the superfused rabbit retina in vitro: effect of light stimulation. Brain Res. 452, 393–395 (1988). 188. Kirsch, M. & Wagner, H. J. Release pattern of endogenous dopamine in teleost retinae during light adaptation and pharmacological stimulation. Vision Res. 29, 147–154 (1989). 189. Zemel, E., Eyal, O., Lei, B. & Perlman, I. NADPH diaphorase activity in mammalian retinas is modulated by the state of visual adaptation. Vis Neurosci 13, 863–71. (1996). 190. Neal, M., Cunningham, J. & Matthews, K. Selective release of nitric oxide from retinal amacrine and bipolar cells. Investig. Ophthalmol. Vis. Sci. 39, 850–853 (1998). 191. Piccolino, M., Neyton, J. & Gerschenfelds2, H. M. Decrease of gap junction

170

permeability induced by dopamine and cyclic adenosine 3’:5’-monophosphate in horizontal cells of turtle retina. J. Neurosci. 4, 2477–2488 (1984). 192. Lasater, E. M. & Dowling, J. E. Dopamine decreases conductance of the electrical junctions between cultured retinal horizontal cells. Proc. Natl. Acad. Sci. U. S. A. 82, 3025–3029 (1985). 193. DeVries, S. H. & Schwartz, E. a. Hemi-gap-junction channels in solitary horizontal cells of the catfish retina. J. Physiol. 445, 201–230 (1992). 194. Lu, C. & McMahon, D. G. Modulation of hybrid bass retinal gap junctional channel gating by nitric oxide. J. Physiol. 499, 689–99 (1997). 195. Pottek, M., Schultz, K. & Weiler, R. Effects of nitric oxide on the horizontal cell network and dopamine release in the carp retina. Vision Res. 37, 1091–1102 (1997). 196. Hampson, E. C., Weiler, R. & Vaney, D. I. pH-gated dopaminergic modulation of horizontal cell gap junctions in mammalian retina. Proc. Biol. Sci. 255, 67–72 (1994). 197. Mangel, S. C. & Dowling, J. E. Responsiveness and receptive field size of carp horizontal cells are reduced by prolonged darkness and dopamine. Science (80-. ). 229, 1107–1109 (1985). 198. Mangel, S. C. & Dowling, J. E. The interplexiform-horizontal cell system of the fish retina: effects of dopamine, light stimulation and time in the dark. Proc. R. Soc. Lond. B. Biol. Sci. 231, 91–121 (1987). 199. Yang, X. L., Tornqvist, K. & Dowling, J. E. Modulation of cone horizontal cell activity in the teleost fish retina. II. Role of interplexiform cells and dopamine in regulating light responsiveness. J. Neurosci. 8, 2269–2278 (1988). 200. Xin, D. & Bloomfield, S. A. Dark- and light-induced changes in coupling between horizontal cells in mammalian retina. J. Comp. Neurol. 405, 75–87 (1999). 201. Baldridge, W. H., Weiler, R. & Dowling, J. E. Dark-suppression and light-sensitization of horizontal cell responses in the hybrid bass retina. Vis Neurosci 12, 611–620 (1995). 202. Rodieck, R. W. & Stone, J. Analysis of receptive fields of cat retinal ganglion cells. J. Neurophysiol. 28, 832–49 (1965). 203. Peichl, L. & Wässle, H. The structural correlate of the receptive field centre of alpha ganglion cells in the cat retina. J. Physiol. 341, 309–324 (1983). 204. Müller, J. F., Dacheux, R. F., Muller, J. F. & Dacheux, R. F. Alpha ganglion cells of the rabbit retina lose antagonistic surround responses under dark adaptation. Vis. Neurosci. 14, 395–401 (1997). 205. Balboa, R. M. & Grzywacz, N. M. The role of early retinal lateral inhibition: more than

171

maximizing luminance information. Vis. Neurosci. 17, 77–89 (2000). 206. Balboa, R. M. & Grzywacz, N. M. The minimal local-asperity hypothesis of early retinal lateral inhibition. Neural Comput. 12, 1485–1517 (2000). 207. Cueva, J. G. et al. Vesicular gamma-aminobutyric acid transporter expression in amacrine and horizontal cells. J. Comp. Neurol. 445, 227–37 (2002). 208. Hirano, A. A., Brandstätter, J. H., Morgans, C. W. & Brecha, N. C. SNAP25 expression in mammalian retinal horizontal cells. J. Comp. Neurol. 519, 972–988 (2011). 209. Pattnaik, B., Jellali, a, Sahel, J., Dreyfus, H. & Picaud, S. GABAC receptors are localized with microtubule-associated protein 1B in mammalian cone photoreceptors. J. Neurosci. 20, 6789–6796 (2000). 210. Liu, X., Hirano, A. A., Sun, X., Brecha, N. C. & Barnes, S. Calcium channels in rat horizontal cells regulate feedback inhibition of photoreceptors through an unconventional GABA- and pH-sensitive mechanism. J. Physiol. 591, 3309–3324 (2013). 211. Djamgoz, M. B. A. Diversity of GABA receptors in the vertebrate outer retina. Trends Neurosci. 18, 118–120 (1995). 212. Lukasiewicz, P. D. GABAC receptors in the vertebrate retina. Mol Neurobiol 12, 181– 194 (1996). 213. Szmajda, B. A. & DeVries, S. H. Glutamate Spillover between Mammalian Cone Photoreceptors. J. Neurosci. 31, 13431–41 (2011). 214. Vroman, R. & Kamermans, M. Feedback-induced glutamate spillover enhances negative feedback from horizontal cells to cones. J. Physiol. 593, 2927–40 (2015). 215. Heidelberger, R., Thoreson, W. B. & Witkovsky, P. Synaptic transmission at retinal ribbon synapses. Progress in Retinal and Eye Research 24, 682–720 (2005). 216. Schmitz, F. The making of synaptic ribbons: How they are built and what they do. Neuroscientist 15, 611–624 (2009). 217. Brandstätter, J. H. Differential expression of the presynaptic cytomatrix protein bassoon among ribbon synapses in the mammalian retina. Eur. J. Neurosci. 11, 3683–3693 (1999). 218. Dacey, D. M. Parallel pathways for spectral coding in primate retina. Annual Review of Neuroscience 23, (2000). 219. Dacey, D. M. & Lee, B. B. The ‘blue-on’ opponent pathway in primate retina originates from a distinct bistratified ganglion cell type. Nature 367, 731–735 (1994). 220. Masland, R. H. The fundamental plan of the retina. Nature Neuroscience 4, 877–886

172

(2001). 221. Wu, S. M., Gao, F. & Maple, B. R. Functional architecture of synapses in the inner retina: segregation of visual signals by stratification of bipolar cell axon terminals. J. Neurosci. 20, 4462–4470 (2000). 222. Wong, K. Y. Retinal Bipolar Cell Input Mechanisms in Giant Danio. III. ON-OFF Bipolar Cells and Their Color-Opponent Mechanisms. J. Neurophysiol. 94, 265–272 (2005). 223. Li, Y. N., Tsujimura, T., Kawamura, S. & Dowling, J. E. Bipolar cell-photoreceptor connectivity in the zebrafish (Danio rerio) retina. J. Comp. Neurol. 520, 3786–3802 (2012). 224. Brandstätter, J. H., Hartveit, E., Sassoè‐Pognetto, M. & Wässle, H. Expression of NMDA and High‐affinity Kainate Receptor Subunit mRNAs in the Adult Rat Retina. Eur. J. Neurosci. 6, 1100–1112 (1994). 225. Peng, Y. W., Blackstone, C. D., Huganir, R. L. & Yau, K. W. Distribution of glutamate receptor subtypes in the vertebrate retina. Neuroscience 66, 483–97 (1995). 226. Hartveit, E. et al. Localization and developmental expression of the NMDA receptor subunit NR2A in the mammalian retina. J. Comp. Neurol. 348, 570–582 (1994). 227. Brandstätter, J. H., Koulen, P. & Wässle, H. Diversity of glutamate receptors in the mammalian retina. Vision Res. 38, 1385–1397 (1998). 228. Slaughter, M. M. & Miller, R. F. The role of excitatory amino acid transmitters in the mudpuppy retina: An analysis with kainic acid and N-methyl aspartate. J. Neurosci. 3, 1701–1711 (1983). 229. Lukasiewicz, P. D. & McReynolds, J. S. Synaptic transmission at N‐methyl‐D‐aspartate receptors in the proximal retina of the mudpuppy. J. Physiol. 367, 99–115 (1985). 230. Coleman, P. A. & Miller, R. F. Do N-methyl-D-aspartate receptors mediate synaptic responses in the mudpuppy retina? J Neurosci 8, 4728–4733 (1988). 231. Mittman, S., Taylor, W. R. & Copenhagen, D. R. Concomitant activation of two types of glutamate receptor mediates excitation of salamander retinal ganglion cells. J. Physiol. 428, 175–197 (1990). 232. Diamond, J. S. & Copenhagen, D. R. The contribution of NMDA and Non-NMDA receptors to the light-evoked input-output characteristics of retinal ganglion cells. Neuron 11, 725–738 (1993). 233. Zhou, Z. J., Marshak, D. W. & Fain, G. L. Amino acid receptors of midget and parasol

173

ganglion cells in primate retina. Proc. Natl. Acad. Sci. U. S. A. 91, 4907–11 (1994). 234. Yu, W. & Miller, R. F. The mechanism by which NBQX enhances NMDA currents in retinal ganglion cells. Brain Res. 709, 184–196 (1996). 235. Cohen, E. D. & Miller, R. F. The role of NMDA and non-NMDA excitatory amino acid receptors in the functional organization of primate retinal ganglion cells. Vis.Neurosci. 11, 317–332 (1994). 236. Taschenberger, H., Engert, F. & Grantyn, R. Synaptic current kinetics in a solely AMPA-receptor-operated glutamatergic synapse formed by rat retinal ganglion neurons. J Neurophysiol 74, 1123–1136 (1995). 237. Taylor, W. R., Chen, E. & Copenhagen, D. R. Characterization of spontaneous excitatory synaptic currents in salamander retinal ganglion cells. J. Physiol. 486, 207– 221 (1995). 238. Matsui, K., Hosoi, N. & Tachibana, M. Excitatory synaptic transmission in the inner retina: paired recordings of bipolar cells and neurons of the ganglion cell layer. J Neurosci 18, 4500–4510 (1998). 239. Matsui, K., Hasegawa, J. & Tachibana, M. Modulation of excitatory synaptic transmission by GABA(C) receptor-mediated feedback in the mouse inner retina. J. Neurophysiol. 86, 2285–2298 (2001). 240. Vigh, J. Prolonged Reciprocal Signaling via NMDA and GABA Receptors at a Retinal Ribbon Synapse. J. Neurosci. 25, 11412–11423 (2005). 241. Chen, S. & Diamond, J. S. Synaptically released glutamate activates extrasynaptic NMDA receptors on cells in the ganglion cell layer of rat retina. J. Neurosci. 22, 2165– 2173 (2002). 242. Wässle, H., Koulen, P., Brandstätter, J. H., Fletcher, E. L. & Becker, C. M. Glycine and GABA receptors in the mammalian retina. in Vision Research 38, 1411–1430 (1998). 243. Pourcho, R. G. & Goebel, D. J. Neuronal subpopulations in cat retina which accumulate the GABA agonist, (3H)muscimol: a combined Golgi and autoradiographic study. J. Comp. Neurol. 219, 25–35 (1983). 244. Wassle, H. & Boycott, B. B. Functional architecture of the mammalian retina. Physiol. Rev. 71, 447–480 (1991). 245. Pourcho, R. G. & Goebel, D. J. A combined golgi and autoradiographic study of (3H)glycine‐accumulating amacrine cells in the cat retina. J. Comp. Neurol. 233, 473– 480 (1985). 246. Pourcho, R. G. & Goebel, D. J. A combined Golgi and autoradiographic study of 3H-

174

glycine-accumulating cone bipolar cells in the cat retina. J Neurosci 7, 1178–1188 (1987). 247. Pow, D. V., Wright, L. L. & Vaney, D. I. The immunocytochemical detection of amino- acid neurotransmitters in paraformaldehyde-fixed tissues. J. Neurosci. Methods 56, 115–123 (1995). 248. Eggers, E. D. & Lukasiewicz, P. D. Multiple pathways of inhibition shape bipolar cell responses in the retina. Visual Neuroscience 28, 95–108 (2011). 249. Eggers, E. D., McCall, M. A. & Lukasiewicz, P. D. Presynaptic inhibition differentially shapes transmission in distinct circuits in the mouse retina. J. Physiol. 582, 569–582 (2007). 250. Ivanova, E., Müller, U. & Wässle, H. Characterization of the glycinergic input to bipolar cells of the mouse retina. Eur. J. Neurosci. 23, 350–364 (2006). 251. Boos, R., Schneider, H. & Wassle, H. Voltage- and transmitter-gated currents of all- amacrine cells in a slice preparation of the rat retina. J Neurosci 13, 2874–2888 (1993). 252. Greferath, U., Grunert, U., Muller, F. & Wassle, H. Localization of GABAA receptors in the rabbit retina. Cell Tissue Res 276, 295–307 (1994). 253. Vitanova, L. et al. Immunocytochemical and electrophysiological characterization of GABA receptors in the frog and turtle retina. Vision Res. 41, 691–704 (2001). 254. Majumdar, S., Heinze, L., Haverkamp, S., Ivanova, E. & Wassle, H. Glycine receptors of A-type ganglion cells of the mouse retina. Vis. Neurosci. 24, 471–87 (2007). 255. Veruki, M. L., Gill, S. B. & Hartveit, E. Spontaneous IPSCs and glycine receptors with slow kinetics in wide-field amacrine cells in the mature rat retina. J. Physiol. 581, 203– 219 (2007). 256. Weiss, J. et al. Glycinergic input of small-field amacrine cells in the retinas of wildtype and glycine receptor deficient mice. Mol. Cell. Neurosci. 37, 40–55 (2008). 257. Majumdar, S., Weiss, J. & Wässle, H. Glycinergic input of widefield, displaced amacrine cells of the mouse retina. J. Physiol. 587, 3831–3849 (2009). 258. Caldwell, J. H. & Daw, N. W. Effects of picrotoxin and strychnine on rabbit retinal ganglion cells: changes in centre surround receptive fields. J. Physiol. 276, 299–310 (1978). 259. O’Brien, B. J., Richardson, R. C. & Berson, D. M. Inhibitory network properties shaping the light evoked responses of cat alpha retinal ganglion cells. Vis. Neurosci. 20, 351– 361 (2003). 260. Chen, X., Hsueh, H.-A., Greenberg, K. & Werblin, F. S. Three Forms of Spatial

175

Temporal Feedforward Inhibition Are Common to Different Ganglion Cell Types in Rabbit Retina. J. Neurophysiol. 103, 2618–32 (2010). 261. Wässle, H., Schäfer-Trenkler, I. & Voigt, T. Analysis of a glycinergic inhibitory pathway in the cat retina. J Neurosci 2, 594-604 (1986). 262. Hsueh, H.-A., Molnar, A. & Werblin, F. S. Amacrine-to-Amacrine Cell Inhibition in the Rabbit Retina. J. Neurophysiol. 100, 2077–88 (2008). 263. Manookin, M. B., Beaudoin, D. L., Ernst, Z. R., Flagel, L. J. & Demb, J. B. Disinhibition Combines with Excitation to Extend the Operating Range of the OFF Visual Pathway in Daylight. J. Neurosci. 28, 4136–50 (2008). 264. Werblin, F. S. Six different roles for crossover inhibition in the retina: Correcting the nonlinearities of synaptic transmission. Vis. Neurosci. 27, 1–8 (2010). 265. Haverkamp, S. & Wässle, H. Characterization of an Amacrine Cell Type of the Mammalian Retina Immunoreactive for Vesicular Glutamate Transporter 3. J. Comp. Neurol. 468, 251–63 (2004). 266. Lee, S. et al. An unconventional glutamatergic circuit in the retina formed by vGluT3 amacrine cells. Neuron 84, 708–15 (2014). 267. Grimes, W. N., Seal, R. P., Oesch, N., Edwards, R. H. & Diamond, J. S. Genetic targeting and physiological features of VGLUT3+ amacrine cells. Vis. Neurosci. 28, 381–392 (2011). 268. Masland, R. H. & Livingstone, C. J. Effect of stimulation with light on synthesis and release of acetylcholine by an isolated mammalian retina. J Neurophysiol 39, 1209–10 (1976). 269. Massey, S. C. & Neal, M. J. The light evoked release of acetylcholine from the rabbit retina iN vivo and its inhibition by γ‐aminobutyric acid. J. Neurochem. 32, 1327–9 (1979). 270. Masland, R. H. & Ames, A. Responses to acetylcholine of ganglion cells in an isolated mammalian retina. J. Neurophysiol. 39, 1220–35 (1976). 271. Yamada, E. S. et al. Synaptic connections of starburst amacrine cells and localization of acetylcholine receptors in primate retinas. J. Comp. Neurol. 461, 76–90 (2003). 272. Dmitrieva, N. A., Strang, C. E. & Keyser, K. T. Expression of alpha 7 nicotinic acetylcholine receptors by bipolar, amacrine, and ganglion cells of the rabbit retina. J. Histochem. Cytochem. 55, 461–476 (2007). 273. Sethuramanujam, S. et al. A Central Role for Mixed Acetylcholine/GABA Transmission in Direction Coding in the Retina. Neuron 90, 1243–1256 (2016).

176

274. Chun, M. H., Han, S. H., Chung, J. W. & Wassle, H. Electron microscopic analysis of the rod pathway of the rat retina. J Comp Neurol 332, 421–32 (1993). 275. Dacheux, R. F. & Raviola, E. The rod pathway in the rabbit retina: a depolarizing bipolar and amacrine cell. J Neurosci 6, 331–45 (1986). 276. Marc, R. E., Anderson, J. R., Jones, B. W., Sigulinsky, C. L. & Lauritzen, J. S. The AII amacrine cell connectome: a dense network hub. Front. Neural Circuits 8, 104 (2014). 277. Grimes, W. N., Zhang, J., Graydon, C. W., Kachar, B. & Diamond, J. S. Retinal Parallel Processors: More than 100 Independent Microcircuits Operate within a Single Interneuron. Neuron 65, 873–885 (2010). 278. Zhang, J., Li, W., Trexler, E. B. & Massey, S. C. Confocal analysis of reciprocal feedback at rod bipolar terminals in the rabbit retina. J Neurosci 22, 10871–10882 (2002). 279. Fletcher, E. L., Koulen, P. & Wässle, H. GABA(A) and GABA(C) receptors on mammalian rod bipolar cells. J. Comp. Neurol. 396, 351–365 (1998). 280. Schubert, T., Hoon, M., Euler, T., Lukasiewicz, P. D. & Wong, R. O. L. Developmental Regulation and Activity-Dependent Maintenance of GABAergic Presynaptic Inhibition onto Rod Bipolar Cell Axonal Terminals. Neuron 78, 124–137 (2013). 281. Elgueta, C., Vielma, A. H., Palacios, A. G. & Schmachtenberg, O. Acetylcholine induces GABA release onto rod bipolar cells through heteromeric nicotinic receptors expressed in A17 amacrine cells. Front. Cell. Neurosci. (2015). doi:10.3389/fncel.2015.00006 282. Nelson, R. & Kolb, H. A17: a broad-field amacrine cell in the rod system of the cat retina. J. Neurophysiol. 54, 592–614 (1985). 283. Hampson, E. C., Vaney, D. I. & Weiler, R. Dopaminergic modulation of gap junction permeability between amacrine cells in mammalian retina. J. Neurosci. 12, 4911–4922 (1992). 284. Kothmann, W. W., Massey, S. C. & O’Brien, J. Dopamine-stimulated dephosphorylation of connexin 36 mediates AII amacrine cell uncoupling. J. Neurosci. 29, 14903–14911 (2009). 285. Mills, S. L. & Massey, S. C. Differential properties of two gap junctional pathways made by AII amacrine cells. Nature 377, 734–737 (1995). 286. Xia, X. B. & Mills, S. L. Gap junctional regulatory mechanisms in the AII amacrine cell of the rabbit retina. Vis. Neurosci. 21, 791–805 (2004). 287. Veruki, M. L. & Hartveit, E. Meclofenamic acid blocks electrical synapses of retinal AII

177

amacrine and on-cone bipolar cells. J. Neurophysiol. 101, 2339–2347 (2009). 288. Feigenspan, A., Teubner, B., Willecke, K. & Weiler, R. Expression of neuronal connexin36 in AII amacrine cells of the mammalian retina. J Neurosci 21, 230–239 (2001). 289. Lin, B., Jakobs, T. C. & Masland, R. H. Different functional types of bipolar cells use different gap-junctional proteins. J. Neurosci. 25, 6696–6701 (2005). 290. Han, Y. & Massey, S. C. Electrical synapses in retinal ON cone bipolar cells: subtype- specific expression of connexins. Proc Natl Acad Sci U S A 102, 13313–13318 (2005). 291. Dedek, K. et al. Localization of heterotypic gap junctions composed of connexin45 and connexin36 in the rod pathway of the mouse retina. Eur. J. Neurosci. 24, 1675–1686 (2006). 292. Li, X. et al. Connexin45-containing neuronal gap junctions in rodent retina also contain connexin36 in both apposing hemiplaques, forming bihomotypic gap junctions, with scaffolding contributed by zonula occludens-1. J Neurosci 28, 9769–9789 (2008). 293. Strettoi, E., Dacheux, R. F. & Raviola, E. Synaptic connections of rod bipolar cells in the inner plexiform layer of the rabbit retina. J. Comp. Neurol. 295, 449–466 (1990). 294. Smith, R. G. & Vardi, N. Simulation of the AII amacrine cell of mammalian retina: functional consequences of electrical coupling and regenerative membrane properties. Vis. Neurosci. 12, 851–860 (1995). 295. Xin, D. & Bloomfield, S. A. Comparison of the responses of AII amacrine cells in the dark- and light-adapted rabbit retina. Vis Neurosci 16, 653–665 (1999). 296. Bloomfield, S. A. & Volgyi, B. Function and plasticity of homologous coupling between AII amacrine cells. in Vision Research 44, 3297–3306 (2004). 297. Bloomfield, S. a, Xin, D. & Osborne, T. Light-induced modulation of coupling between AII amacrine cells in the rabbit retina. Vis. Neurosci. 14, 565–576 (1997). 298. Hartveit, E. & Veruki, M. L. AII amacrine cells express functional NMDA receptors. Neuroreport 8, 1219–23 (1997). 299. Völgyi, B., Deans, M. R., Paul, D. L. & Bloomfield, S. A. Convergence and segregation of the multiple rod pathways in mammalian retina. J. Neurosci. 24, 11182–92 (2004). 300. Kothmann, W. W., Li, X., Burr, G. S. & O’Brien, J. Connexin 35/36 is phosphorylated at regulatory sites in the retina. Vis. Neurosci. 24, 363–75 (2007). 301. Asteriti, S., Gargini, C. & Cangiano, L. Mouse rods signal through gap junctions with cones. Elife 2014, (2014). 302. Güldenagel, M. et al. Visual transmission deficits in mice with targeted disruption of the

178

gap junction gene connexin36. J. Neurosci. 21, 6036–44 (2001). 303. Smith, R. G., Freed, M. A. & Sterling, P. Microcircuitry of the dark-adapted cat retina: functional architecture of the rod-cone network. J. Neurosci. 6, 3505–3517 (1986). 304. Soucy, E., Wang, Y., Nirenberg, S., Nathans, J. & Meister, M. A novel signaling pathway from rod photoreceptors to ganglion cells in mammalian retina. Neuron 21, 481–493 (1998). 305. Hack, I., Peichl, L. & Brandstatter, J. H. An alternative pathway for rod signals in the rodent retina: rod photoreceptors, cone bipolar cells, and the localization of glutamate receptors. Proc Natl Acad Sci U S A 96, 14130–14135 (1999). 306. Tsukamoto, Y., Morigiwa, K., Ueda, M. & Sterling, P. Microcircuits for night vision in mouse retina. J. Neurosci. 21, 8616–23 (2001). 307. Fyk-Kolodziej, B., Qin, P. & Pourcho, R. G. Identification of a cone bipolar cell in cat retina which has input from both rod and cone photoreceptors. J. Comp. Neurol. 464, 104–113 (2003). 308. Li, W., Keung, J. W. & Massey, S. C. Direct synaptic connections between rods and OFF cone bipolar cells in the rabbit retina. J. Comp. Neurol. 474, 1–12 (2004). 309. Protti, D. a, Flores-Herr, N., Li, W., Massey, S. C. & Wässle, H. Light signaling in scotopic conditions in the rabbit, mouse and rat retina: a physiological and anatomical study. J. Neurophysiol. 93, 3479–88 (2005). 310. Pang, J.-J. et al. Direct rod input to cone BCs and direct cone input to rod BCs challenge the traditional view of mammalian BC circuitry. Proc. Natl. Acad. Sci. 107, 395–400 (2010). 311. Pang, J. J., Gao, F., Paul, D. L. & Wu, S. M. Rod, M-cone and M/S-cone inputs to hyperpolarizing bipolar cells in the mouse retina. J. Physiol. 590, 845–54 (2012). 312. Barger, G. & Dale, H. H. Chemical structure and sympathomimetic action of amines. J. Physiol. 41, 19–59 (1910). 313. Fahn, S. The history of dopamine and levodopa in the treatment of Parkinson’s disease. Movement Disorders 23, 497-508 (2008). 314. Carlsson, A., Lindqvist, M. & Magnusson, T. 3,4-Dihydroxyphenylalanine and 5- hydroxytryptophan as reserpine antagonists. Nature (1957). doi: 10.1038/1801200a0 315. Björklund, A. & Dunnett, S. B. Dopamine neuron systems in the brain: an update. Trends in Neurosciences 30, 194–202 (2007). 316. Falck, B., Hillarp, N.-A., Thieme, G. & Torp, A. Fluorescence of catecholamines and related compounds condensed with formaldehyde. J. Histochem. Cytochem. 10, 348–

179

354 (1962). 317. Malmfors, T. Evidence of Adrenergic Neurons with Synaptic Terminals in the Retina of Rats Demonstrated with Fluorescence and Electron Microscopy. Acta Physiol. Scand. 58, 99–100 (1963). 318. Ballesta, J., Terenghi, G., Thibault, J. & Polak, J. M. Putative dopamine-containing cells in the retina of seven species demonstrated by tyrosine hydroxylase immunocytochemistry. Neuroscience 12, 1147–1156 (1984). 319. Hirasawa, H., Betensky, R. a. & Raviola, E. Corelease of Dopamine and GABA by a Retinal Dopaminergic Neuron. J. Neurosci. 32, 13281–13291 (2012). 320. Hirasawa, H., Contini, M. & Raviola, E. Extrasynaptic release of GABA and dopamine by retinal dopaminergic neurons. Philos. Trans. R. Soc. B Biol. Sci. 370, 1–7 (2015). 321. Ho, T. et al. Vesicular expression and release of ATP from dopaminergic neurons of the mouse retina and midbrain. Front. Cell. Neurosci. (2015). doi:10.3389/fncel.2015.00389 322. Mariani, A. P. & Hokoc, J. N. Two types of tyrosine hydroxylase-immunoreactive amacrine cell in the rhesus monkey retina. J. Comp. Neurol. 276, 81–91 (1988). 323. Ehinger, B. & Steinbusch, H. W. Quantitation and immunohistochemistry of catecholamines in the posterior segment of the eye. Histochemistry 82, 435–439 (1985). 324. Gustincich, S., Feigenspan, A., Wu, D. K., Koopman, L. J. & Raviola, E. Control of dopamine release in the retina: A transgenic approach to neural networks. Neuron 18, 723–736 (1997). 325. Contini, M. et al. Synaptic input of ON-bipolar cells onto the dopaminergic neurons of the mouse retina. J. Comp. Neurol. 518, 2035–2050 (2010). 326. Zhang, D.-Q., Stone, J. F., Zhou, T., Ohta, H. & McMahon, D. G. Characterization of genetically labeled catecholamine neurons in the mouse retina. Neuroreport 15, 1761– 1765 (2004). 327. Zhang, D.-Q. et al. Intraretinal signaling by ganglion cell photoreceptors to dopaminergic amacrine neurons. Proc. Natl. Acad. Sci. U. S. A. 105, 14181–14186 (2008). 328. Zhang, D. Q., Belenky, M. A., Sollars, P. J., Pickard, G. E. & McMahon, D. G. Melanopsin mediates retrograde visual signaling in the retina. PLoS One 7, (2012). 329. Prigge, C. L. et al. M1 ipRGCs Influence Visual Function through Retrograde Signaling in the Retina. J. Neurosci. 36, 7184–7197 (2016). 330. Qiao, S. N., Zhang, Z., Ribelayga, C. P., Zhong, Y. M. & Zhang, D. Q. Multiple cone

180

pathways are involved in photic regulation of retinal dopamine. Sci. Rep. (2016). doi: 10.1167/iovs.17-21835 331. Liu, L.-L., Spix, N. J. & Zhang, D.-Q. NMDA Receptors Contribute to Retrograde Synaptic Transmission from Ganglion Cell Photoreceptors to Dopaminergic Amacrine Cells. Front. Cell. Neurosci. (2017). doi:10.3389/fncel.2017.00279 332. Qiao, S. N., Zhou, W., Liu, L. L., Zhang, D. Q. & Zhong, Y. M. Orexin-a suppresses signal transmission to dopaminergic amacrine cells from outer and inner retinal photoreceptors. Investig. Ophthalmol. Vis. Sci. 58, 4712–4721 (2017). 333. Zhao, X., Wong, K. Y. & Zhang, D.-Q. Mapping physiological inputs from multiple photoreceptor systems to dopaminergic amacrine cells in the mouse retina. Sci. Rep. (2017). doi: 10.1038/s41598-017-08172-x 334. Vuong, H. E., Pérez de Sevilla Müller, L., Hardi, C. N., McMahon, D. G. & Brecha, N. C. Heterogeneous transgene expression in the retinas of the TH-RFP, TH-Cre, TH-BAC- Cre and DAT-Cre mouse lines. Neuroscience 29, 319–37 (2015). 335. Gong, S. et al. A gene expression atlas of the central nervous system based on bacterial artificial chromosomes. Nature 425, 917–925 (2003). 336. Newkirk, G. S., Hoon, M., Wong, R. O. & Detwiler, P. B. Inhibitory inputs tune the light response properties of dopaminergic amacrine cells in mouse retina. J. Neurophysiol. 110, 536–552 (2013). 337. Versaux-Botteri, C., Nguyen-Legros, J., Vigny, A. & Raoux, N. Morphology, density and distribution of tyrosine hydroxylase-like immunoreactive cells in the retina of mice. Brain Res. 301, 192–197 (1984). 338. Wulle, I. & Schnitzer, J. Distribution and morphology of tyrosine hydroxylase- immunoreactive neurons in the developing mouse retina. Dev. brain Res. 48, 59–72 (1989). 339. Keeley, P. W. & Reese, B. E. Morphology of dopaminergic amacrine cells in the mouse retina: Independence from homotypic interactions. J. Comp. Neurol. 518, 1220–1231 (2010). 340. Kolb, H., Nelson, R. & Mariani, A. Amacrine cells, bipolar cells and ganglion cells of the cat retina: A Golgi study. Vision Res. 21, 1081–1114 (1981). 341. Mariani, A. P. Amacrine cells of the rhesus monkey retina. J. Comp. Neurol. 301, 382– 400 (1990). 342. Hokoc, J. N. & Mariani, a P. Tyrosine hydroxylase immunoreactivity in the rhesus monkey retina reveals synapses from bipolar cells to dopaminergic amacrine cells. J.

181

Neurosci. 7, 2785–2793 (1987). 343. Nguyen-Legros, J. Functional neuroarchitecture of the retina: hypothesis on the dysfunction of retinal dopaminergic circuitry in Parkinson’s disease. Surg. Radiol. Anat. 10, 137–44 (1988). 344. Crooks, J. & Kolb, H. Localization of GABA, glycine, glutamate and tyrosine hydroxylase in the human retina. J. Comp. Neurol. 315, 287–302 (1992). 345. Contini, M. et al. Synaptic input of ON-bipolar cells onto the dopaminergic neurons of the mouse retina. J. Comp. Neurol. 518, 2035–50 (2010). 346. Nichols, C. W., Jacobowitz, D. & Hottenstein, M. The influence of light and dark on the catecholamine content of the retina and choroid. Invest. Ophthalmol. Vis. Sci. 6, 642– 646 (1967). 347. KRAMER, S. G. Dopamine: A Retinal Neurotransmitter :I. Retinal Uptake, Storage, and Light-Stimulated Release of H3-Dopamine in Vivo. Invest. Ophthalmol. Vis. Sci. 10, 438–452 (1971). 348. Iuvone, P. M., Galli, C. L., Garrison-Gund, C. K. & Neff, N. H. Light stimulates tyrosine hydroxylase activity and dopamine synthesis in retinal amacrine neurons. Science (80-. ). 202, 901–902 (1978). 349. Morgan, W. W. & Kamp, C. W. Dopaminergic amacrine neurons of rat retinas with photoreceptor degeneration continue to respond to light. Life Sci. 26, 1619–1626 (1980). 350. Kamp, C. W. & Morgan, W. W. GABA antagonists enhance dopamine turnover in the rat retina in vivo. Eur. J. Pharmacol. 69, 273–279 (1981). 351. Cohen, J., Hadjiconstantinou, M. & Neff, N. H. Activation of dopamine-containing amacrine cells of retina: light-induced increase of acidic dopamine metabolites. Brain Res. 260, 125–7 (1983). 352. Parkinson, D. & Rando, R. R. Effects of light on dopamine metabolism in the chick retina. J Neurochem 40, 39–46 (1983). 353. Parkinson, D. & Rando, R. R. Effect of light on dopamine turnover and metabolism in rabbit retina. Investig. Ophthalmol. Vis. Sci. 24, 384–388 (1983). 354. O’Connor, P. M., Zucker, C. L. & Dowling, J. E. Regulation of Dopamine Release from Interplexiform Cell Processes in the Outer Plexiform Layer of the Carp Retina. J. Neurochem. 49, 916–920 (1987). 355. Godley, B. F. & Wurtman, R. J. Release of endogenous dopamine from the superfused rabbit retina in vitro: effect of light stimulation. Brain Res. 452, 393–395 (1988). 356. Ishita, S., Negishi, K., Teranishi, T., Shimada, Y. & Kato, S. GABAergic Inhibition on

182

Dopamine Cells of the Fish Retina: A [3H]Dopamine Release Study with Isolated Cell Fractions. J. Neurochem. 50, 1–6 (1988). 357. Kolbinger, W., Kohler, K., Oetting, H. & Weiler, R. Endogenous dopamine and cyclic events in the fish retina, i: Hplc assay of total content, release, and metabolic turnover during different light/dark cycles. Vis. Neurosci. 5, 143–9 (1990). 358. Umino, O. & Dowling, J. E. Dopamine release from interplexiform cells in the retina: effects of GnRH, FMRFamide, bicuculline, and enkephalin on horizontal cell activity. J Neurosci 11, 3034–3046 (1991). 359. Umino, O., Lee, Y. & Dowling, J. E. Effects of light stimuli on the release of dopamine from interplexiform cells in the white perch retina. Vis Neurosci 7, 451–458 (1991). 360. Kolbinger, W. & Weiler, R. Modulation of endogenous dopamine release in the turtle retina: Effects of light, calcium, and neurotransmitters. Vis. Neurosci. 10, 1035–1041 (1993). 361. Witkovsky, P., Nicholson, C., Rice, M. E., Bohmaker, K. & Meller, E. Extracellular dopamine concentration in the retina of the clawed frog, Xenopus laevis. Proc. Natl. Acad. Sci. U. S. A. 90, 5667–5671 (1993). 362. Boatright, J. H., Gordon, J. R. & Michael Iuvone, P. Inhibition of endogenous dopamine release in amphibian retina by l-2-amino-4-phosphonobutyric acid (l-AP4) and trans-2- aminocyclopentane-1,3-dicarboxylate (ACPD). Brain Res. 649, 339–342 (1994). 363. Weiler, R., Baldridge, W. H., Mangel, S. C. & Dowling, J. E. Modulation of endogenous dopamine release in the fish retina by light and prolonged darkness. Vis. Neurosci. 14, 351–356 (1997). 364. Boelen, M. K., Boelen, M. G. & Marshak, D. W. Light-stimulated release of dopamine from the primate retina is blocked by 1-2-amino-4-phosphonobutyric acid (APB). Vis. Neurosci. 15, 97–103 (1998). 365. Witkovsky, P., Gabriel, R., Haycock, J. W. & Meller, E. Influence of light and neural circuitry on tyrosine hydroxylase phosphorylation in the rat retina. J Chem Neuroanat 19, 105–116 (2000). 366. Megaw, P., Morgan, I. & Boelen, M. Vitreal dihydroxyphenylacetic acid (DOPAC) as an index of retinal dopamine release. J. Neurochem. 76, 1636–1644 (2001). 367. Maaswinkel, H., Puppala, D., Mason, B., Legan, S. J. & Li, L. In vivo Microdialysis and HPLC Detection of Retinal Dopamine Release in Zebrafish . Invest. Ophthalmol. Vis. Sci. 45, 5285 (2004). 368. Cameron, M. A. et al. Light regulation of retinal dopamine that is independent of

183

melanopsin phototransduction. Eur. J. Neurosci. 29, 761–767 (2009). 369. Jackson, C. R. et al. Retinal Dopamine Mediates Multiple Dimensions of Light-Adapted Vision. J. Neurosci. 32, 9359–9368 (2012). 370. Pérez-Fernández, V., Harman, D. G., Morley, J. W. & Cameron, M. A. Optimized Method to Quantify Dopamine Turnover in the Mammalian Retina. Anal. Chem. 89, 12276–12283 (2017). 371. Iuvone, P. M. Evidence for a D2 dopamine receptor in frog retina that decreases cyclic AMP accumulation and serotonin N-acetyltransferase activity. Life Sci (1986). 372. Di Paolo, T., Harnois, C. & Daigle, M. Assay of dopamine and its metabolites in human and rat retina. Neurosci. Lett. 74, 250–4 (1987). 373. Cahill, G. M. & Besharse, J. C. Resetting the circadian clock in cultured Xenopus eyecups: regulation of retinal melatonin rhythms by light and D2 dopamine receptors. J. Neurosci. 11, 2959–71 (1991). 374. Nir, I., Haque, R. & Iuvone, P. M. Diurnal metabolism of dopamine in dystrophic retinas of homozygous and heterozygous retinal degeneration slow (rds) mice. Brain Res. 884, 13–22 (2000). 375. Doyle, S. E., McIvor, W. E. & Menaker, M. Circadian rhythmicity in dopamine content of mammalian retina: Role of the photoreceptors. J. Neurochem. 83, 211–219 (2002). 376. Ribelayga, C. & Mangel, S. C. Absence of Circadian Clock Regulation of Horizontal Cell Gap Junctional Coupling Reveals Two Dopamine Systems in the Goldfish Retina. J. Comp. Neurol. 467, 243–53 (2003). 377. Ruan, G. X., Allen, G. C., Yamazaki, S. & McMahon, D. G. An autonomous circadian clock in the inner mouse retina regulated by dopamine and GABA. PLoS Biol. 6, 2248– 2265 (2008). 378. Pozdeyev, N. et al. Dopamine modulates diurnal and circadian rhythms of protein phosphorylation in photoreceptor cells of mouse retina. Eur. J. Neurosci. 27, 2691–2700 (2008). 379. Tosini, G. & Michael Iuvone, P. Role of melatonin and dopamine in the regulation of retinal circadian rhythms. in The Retina and Circadian Rhythms 49–68 (2014). doi:10.1007/978-1-4614-9613-7_4 380. Korshunov, K. S., Blakemore, L. J. & Trombley, P. Q. Dopamine: A Modulator of Circadian Rhythms in the Central Nervous System. Front. Cell. Neurosci. (2017). doi:10.3389/fncel.2017.00091 381. Socodato, R. et al. Dopamine promotes NMDA receptor hypofunction in the retina

184

through D1 receptor-mediated Csk activation, Src inhibition and decrease of GluN2B phosphorylation. Sci. Rep. 7, (2017). 382. Pang, S. F., Yu, H. S., Suen, H. C. & Brown, G. M. Melatonin in the retina of rats: A diurnal rhythm. J. Endocrinol. 87, 89–93 (1980). 383. Cahill, G. M. & Besharse, J. C. Resetting the circadian clock in cultured Xenopus eyecups: regulation of retinal melatonin rhythms by light and D dopamine receptors. J. Neurosci. (1991). 384. Cahill, G. M. & Besharse, J. C. Light-sensitive melatonin synthesis by Xenopus photoreceptors after destruction of the inner retina. Vis. Neurosci. 8, 487–490 (1992). 385. Witkovsky, P. et al. Cellular location and circadian rhythm of expression of the biological clock gene Period 1 in the mouse retina. J. Neurosci. 23, 7670–7676 (2003). 386. Gustincich, S. et al. Gene discovery in genetically labeled single dopaminergic neurons of the retina. Proc. Natl. Acad. Sci. U. S. A. 101, 5069–5074 (2004). 387. Ruan, G.-X., Zhang, D.-Q., Zhou, T., Yamazaki, S. & McMahon, D. G. Circadian organization of the mammalian retina. Proc. Natl. Acad. Sci. U. S. A. 103, 9703–9708 (2006). 388. Dorenbos, R., Contini, M., Hirasawa, H., Gustincich, S. & Raviola, E. Expression of circadian clock genes in retinal dopaminergic cells. Vis. Neurosci. 24, 573–80 (2007). 389. Doyle, S. E., Grace, M. S., McIvor, W. & Menaker, M. Circadian rhythms of dopamine in mouse retina: the role of melatonin. Vis. Neurosci. 19, 593–601 (2002). 390. Sakamoto, K., Liu, C., Kasamatsu, M., Iuvone, P. M. & Tosini, G. Intraocular injection of kainic acid does not abolish the circadian rhythm of arylalkylamine N- acetyltransferase mRNA in rat photoreceptors. Mol. Vis. 12, 117–124 (2006). 391. Witkovsky, P., Nicholson, C., Rice, M. E., Bohmaker, K. & Meller, E. Extracellular dopamine concentration in the retina of the clawed frog, Xenopus laevis. Proc. Natl. Acad. Sci. U. S. A. 90, 5667–5671 (1993). 392. Nemeroff, C. B. The Biochemical Basis of Neuropharmacology, 8th ed. American Journal of Psychiatry 161, 379–379 (2004). 393. Cao, W. et al. Effects of experimentally induced ischemia on dopamine metabolism in rabbit retina. Invest. Ophthalmol. Vis. Sci. 34, 3140-6(1993). 394. Drumheller, A., Henni, H., Lafond, G., Brunette, J. R. & Jolicoeur, F. B. Differential effects of low versus high doses of apomorphine on retinal dopamine metabolism in light- and dark-adapted rabbits. Pharmacol. Biochem. Behav. 50, 83–90 (1995). 395. Pendrak, K. et al. Retinal dopamine in the recovery from experimental myopia. Curr.

185

Eye Res. 16, 152–7 (1997). 396. Pozdeyev, N. V. & Lavrikova, E. V. Diurnal changes of tyrosine, dopamine, and dopamine metabolites content in the retina of rats maintained at different lighting conditions. J. Mol. Neurosci. 15, 1–9 (2000). 397. Calamusa, M. et al. Specific alterations of tyrosine hydroxylase immunopositive cells in the retina of NT-4 knock out mice. Vision Res. 47, 1523–1536 (2007). 398. Li, H. et al. Adenosine and Dopamine Receptors Coregulate Photoreceptor Coupling via Gap Junction Phosphorylation in Mouse Retina. J. Neurosci. 33, 3135–3150 (2013). 399. Park, H. na et al. Visually-driven ocular growth in mice requires functional rod photoreceptors. Investig. Ophthalmol. Vis. Sci. 55, 6272–9 (2014). 400. Connaughton, V. P., Wetzell, B., Arneson, L. S., Delucia, V. & L. Riley, A. Elevated dopamine concentration in light-adapted zebrafish retinas is correlated with increased dopamine synthesis and metabolism. J. Neurochem. 135, 101–108 (2015). 401. Wu, X. H. et al. Unaltered retinal dopamine levels in a C57BL/6 mouse model of form- deprivation myopia. Investig. Ophthalmol. Vis. Sci. 56, 967–977 (2015). 402. Spix, N. J. et al. Vulnerability of dopaminergic amacrine cells to chronic ischemia in a mouse model of oxygen-induced retinopathy. Investig. Ophthalmol. Vis. Sci. 57, 3047– 3057 (2016). 403. Zhou, X., Pardue, M. T., Iuvone, P. M. & Qu, J. Dopamine signaling and myopia development: What are the key challenges. Progress in Retinal and Eye Research 61, 60–71 (2017). 404. Kebabian, J. W. & Calne, D. B. Multiple receptors for dopamine. Nature 277, 93–96 (1979). 405. Niznik, H. B. Dopamine receptors: molecular structure and function. Mol. Cell. Endocrinol. 54, 1–22 (1987). 406. Tran, V. T. & Dickman, M. Differential localization of dopamine D1 and D2 receptors in rat retina. Invest. Ophthalmol. Vis. Sci. 33, 1620–1626 (1992). 407. Bjelke, B. et al. Dopaminergic transmission in the rat retina: Evidence for volume transmission. J. Chem. Neuroanat. 12, 37–50 (1996). 408. Dearry, a et al. Molecular cloning and expression of the gene for a human D1 dopamine receptor. Nature 347, 72–6 (1990). 409. Dearry, A., Falardeau, P., Shores, C. & Caron, M. G. D2 dopamine receptors in the human retina: cloning of cDNA and localization of mRNA. Cell. Mol. Neurobiol. 11, 437–453 (1991).

186

410. Baba, K., Debruyne, J. P. & Tosini, G. Dopamine 2 Receptor Activation Entrains Circadian Clocks in Mouse Retinal Pigment Epithelium. Sci. Rep. (2017). doi: 10.1038/s41598-017-05394-x. 411. Dong, F. et al. [Expression of dopamine receptor D2 and adenosine receptor A2A in human retinal pigment epithelium]. Zhonghua. Yan Ke Za Zhi. 43, 1110–1113 (2007). 412. Nguyen-Legros, J., Versaux-Botteri, C. & Vernier, P. Dopamine receptor localization in the mammalian retina. Mol. Neurobiol. 19, 181–204 (1999). 413. Guha, S. et al. Stimulation of the D5 dopamine receptor acidifies the lysosomal pH of retinal pigmented epithelial cells and decreases accumulation of autofluorescent photoreceptor debris. J. Neurochem. 122, 823–833 (2012). 414. Muresan, Z. & Besharse, J. C. D2-like dopamine receptors in amphibian retina: localization with fluorescent ligands. J Comp Neurol 331, 149–160 (1993). 415. Cohen, a I., Todd, R. D., Harmon, S. & O’Malley, K. L. Photoreceptors of mouse retinas possess D4 receptors coupled to adenylate cyclase. Proc. Natl. Acad. Sci. U. S. A. 89, 12093–7 (1992). 416. Veruki, M. L. & Wässle, H. Immunohistochemical localization of dopamine D1 receptors in rat retina. Eur. J. Neurosci. 8, 2286–2297 (1996). 417. Nguyen-Legros, J., Simon, A., Caillé, I. & Bloch, B. Immunocytochemical localization of dopamine D1 receptors in the retina of mammals. Vis Neurosci 14, 545–51 (1997). 418. Cohen, J., Iuvone, P. M. & Neff, N. H. Neuroleptic drugs activate tyrosine hydroxylase of retinal amacrine cells. J Pharmacol Exp Ther 218, 390–394 (1981). 419. El Mestikawy, S. & Hamon, M. Is Dopamine‐Induced Inhibition of Adenylate Cyclase Involved in the Autoreceptor‐Mediated Negative Control of Tyrosine Hydroxylase in Striatal Dopaminergic Terminals? J. Neurochem. 47, 1425–1433 (1986). 420. Biedermann, B., Fröhlich, E., Grosche, J., Wagner, H. J. & Reichenbach, A. Mammalian M{ü}ller (glial) cells express functional D2 dopamine receptors. Neuroreport 6, 609– 612 (1995). 421. LaVail, M. M. Rod outer segment disk shedding in rat retina: Relationship to cyclic lighting. Science (80-. ). 194, 1071–1074 (1976). 422. Nguyen-Legros, J. & Hicks, D. Renewal of photoreceptor outer segments and their phagocytosis by the retinal pigment epithelium. Int. Rev. Cytol. 196, 245–313 (2000). 423. Ribelayga, C., Cao, Y. & Mangel, S. C. The Circadian Clock in the Retina Controls Rod-Cone Coupling. Neuron 59, 790–801 (2008). 424. Firsov, M. L. & Astakhova, L. A. The Role of Dopamine in Controlling Retinal

187

Photoreceptor Function in Vertebrates. Neurosci. Behav. Physiol. 46, 138–145 (2016). 425. Ivanova, T. N. & Iuvone, P. M. Circadian rhythm and photic control of cAMP level in chick retinal cell cultures: A mechanism for coupling the circadian oscillator to the melatonin-synthesizing enzyme, arylalkylamine N-acetyltransferase, in photoreceptor cells. Brain Res. 991, 96–103 (2003). 426. Nir, I. et al. Dysfunctional light-evoked regulation of cAMP in photoreceptors and abnormal retinal adaptation in mice lacking dopamine D4 receptors. J. Neurosci. 22, 2063–2073 (2002). 427. Cohen, A. I. & Blazynski, C. Dopamine and its agonists reduce a light-sensitive pool of cyclic AMP in mouse photoreceptors. Vis. Neurosci. 4, 43–52 (1990). 428. Jackson, C. R. et al. Essential roles of dopamine D4 receptors and the type 1 adenylyl cyclase in photic control of cyclic AMP in photoreceptor cells. J. Neurochem. 109, 148– 157 (2009). 429. Astakhova, L. A., Samoiliuk, E. V., Govardovskii, V. I. & Firsov, M. L. cAMP controls rod photoreceptor sensitivity via multiple targets in the phototransduction cascade. J. Gen. Physiol. 140, 421–433 (2012). 430. Herrmann, R. et al. Rod Vision Is Controlled by Dopamine-Dependent Sensitization of Rod Bipolar Cells by GABA. Neuron 72, 101–110 (2011). 431. Lasater, E. M. Retinal horizontal cell gap junctional conductance is modulated by dopamine through a cyclic AMP-dependent protein kinase. Proc. Natl. Acad. Sci. U. S. A. 84, 7319–23 (1987). 432. Ribelayga, C. & Mangel, S. C. Tracer coupling between fish rod horizontal cells: Modulation by light and dopamine but not the retinal circadian clock. Vis. Neurosci. 24, 333–44 (2007). 433. Pflug, R., Nelson, R., Huber, S. & Reitsamer, H. Modulation of horizontal cell function by dopaminergic ligands in mammalian retina. Vision Res. 48, 1383–1390 (2008). 434. McMahon, D. G., Knapp, a G. & Dowling, J. E. Horizontal cell gap junctions: single- channel conductance and modulation by dopamine. Proc. Natl. Acad. Sci. U. S. A. 86, 7639–7643 (1989). 435. Baldridge, W. H., Ball, a K. & Miller, R. G. Dopaminergic regulation of horizontal cell gap junction particle density in goldfish retina. J. Comp. Neurol. 265, 428–36 (1987). 436. Yazulla, S. & Kleinschmidt, J. Carrier-mediated release of GABA from retinal horizontal cells. Brain Res. 263, 63–75 (1983). 437. Smith, B. J., Côté, P. D. & Tremblay, F. D1 dopamine receptors modulate cone on

188

bipolar cell Nav channels to control daily rhythms in photopic vision. Chronobiol. Int. 32, 48–58 (2015). 438. Chaffiol, A., Ishii, M., Cao, Y. & Mangel, S. C. Dopamine Regulation of GABAAReceptors Contributes to Light/Dark Modulation of the ON-Cone Bipolar Cell Receptive Field Surround in the Retina. Curr. Biol. 27, 2600–2609.e4 (2017). 439. Travis, A. M., Heflin, S. J., Hirano, A. A., Brecha, N. C. & Arshavsky, V. Y. Dopamine- dependent sensitization of rod bipolar cells by GABA is conveyed through wide-field amacrine cells. J. Neurosci. 38, 723–732 (2017). 440. Maguire, G. & Werblin, F. S. Dopamine enhances a glutamate-gated ionic current in OFF bipolar cells of the tiger salamander retina. J. Neurosci. 14, 6094–6101 (1994). 441. Jensen, R. J. & Daw, N. W. Effects of dopamine antagonists on receptive fields of brisk cells and directionally selective cells in the rabbit retina. J. Neurosci. 4, 2972–85 (1984). 442. Jensen, R. J. & Daw, N. W. Effects of dopamine and its agonists and antagonists on the receptive field properties of ganglion cells in the rabbit retina. Neuroscience 17, 837– 855 (1986). 443. Hu, E. H., Pan, F., Völgyi, B. & Bloomfield, S. A. Light increases the gap junctional coupling of retinal ganglion cells. J. Physiol. 588, 4145–4163 (2010). 444. Mills, S. L. et al. Dopaminergic modulation of tracer coupling in a ganglion-amacrine cell network. Vis Neurosci 24, 593–608 (2007). 445. Bu, J. Y., Li, H., Gong, H. Q., Liang, P. J. & Zhang, P. M. Gap junction permeability modulated by dopamine exerts effects on spatial and temporal correlation of retinal ganglion cells’ firing activities. J. Comput. Neurosci. 36, 67–79 (2014). 446. Ogata, G., Stradleigh, T. W., Partida, G. J. & Ishida, A. T. Dopamine and full-field illumination activate D1 and D2-D5-type receptors in adult rat retinal ganglion cells. J. Comp. Neurol. 520, 4032–4049 (2012). 447. H Ikeda, T D Priest, J. R. and K. W. Silent dopaminergic synapse at feline retinal ganglion cells? Clin Vis. Sci 12, 25–38 (1986). 448. Frucht, Y., Vidauri, J. & Melamed, E. Light activation of dopaminergic neurons in rat retina is mediated through photoreceptors. Brain Res. 249, 153–156 (1982). 449. Nir, I. & Michael Iuvone, P. Alterations in light-evoked dopamine metabolism in dystrophic retinas of mutant rds mice. Brain Res. 649, 85–94 (1994). 450. Mills, S. L. et al. Dopaminergic modulation of tracer coupling in a ganglion-amacrine cell network. Vis. Neurosci. 24, 593–608 (2007). 451. Kolb, H., Cuenca, N., Wang, H. H. & Dekorver, L. The synaptic organization of the

189

dopaminergic amacrine cell in the cat retina. J. Neurocytol. 19, 343–366 (1990). 452. Newkirk, G. S., Hoon, M., Wong, R. O. & Detwiler, P. B. Inhibitory inputs tune the light response properties of dopaminergic amacrine cells in mouse retina. J. Neurophysiol. 110, 536–52 (2013). 453. Calvert, P. D. et al. Phototransduction in transgenic mice after targeted deletion of the rod transducin alpha -subunit. Proc. Natl. Acad. Sci. U. S. A. 97, 13913–13918 (2000). 454. Allen, A. E., Cameron, M. A., Brown, T. M., Vugler, A. A. & Lucas, R. J. Visual responses in mice lacking critical components of all known retinal phototransduction cascades. PLoS One (2010). doi: 10.1371/journal.pone.0015063 455. Hughes, S. et al. Characterisation of light responses in the retina of mice lacking principle components of rod, cone and melanopsin phototransduction signalling pathways. Sci. Rep. (2016). doi: 10.1038/srep28086 456. Wong, K. Y., Graham, D. M. & Berson, D. M. The retina-attached SCN slice preparation: an in vitro mammalian circadian visual system. J. Biol. Rhythms 22, 400– 10 (2007). 457. Puopolo, M., Hochstetler, S. E., Gustincich, S., Wightman, R. M. & Raviola, E. Extrasynaptic release of dopamine in a retinal neuron: Activity dependence and transmitter modulation. Neuron 30, 211–225 (2001). 458. Vuong, H. E., Hardi, C. N., Barnes, S. & Brecha, N. C. Parallel Inhibition of Dopamine Amacrine Cells and Intrinsically Photosensitive Retinal Ganglion Cells in a Non-Image- Forming Visual Circuit of the Mouse Retina. J. Neurosci. 35, 15955–15970 (2015). 459. Veruki, M. L. Dopaminergic neurons in the rat retina express dopamine D2/3 receptors. Eur. J. Neurosci. 9, 1096–1100 (1997). 460. Do Nascimento, J. L. M., Kubrusly, R. C. C., Reis, R. A. M., De Mello, M. C. F. & De Mello, F. G. Atypical effect of dopamine in modulating the functional inhibition of NMDA receptors of cultured retina cells. Eur. J. Pharmacol. 343, 103–10 (1998). 461. Calaza, K. C., De Mello, F. G. & Gardino, P. F. GABA release induced by aspartate- mediated activation of NMDA receptors is modulated by dopamine in a selective subpopulation of amacrine cells. J. Neurocytol. 30, 181–193 (2001). 462. Kothmann, W. W. et al. Nonsynaptic NMDA Receptors Mediate Activity-Dependent Plasticity of Gap Junctional Coupling in the AII Amacrine Cell Network. J. Neurosci. 32, 6747–59 (2012). 463. Morgan, W. W. A GABAergic Influence on the Light-Induced Increase in Dopamine Turnover in the Dark-Adapted Rat Retina In Vivo. J. Neurochem. 34, 1082–1086

190

(1980). 464. Boatright, J. H., Rubim, N. M. & Iuvone, P. M. Regulation of endogenous dopamine release in amphibian retina by gamma-aminobutyric acid and glycine. Vis Neurosci 11, 1003–1012 (1994). 465. Marshburn, P. B. & Iuvone, P. M. The role of GABA in the regulation of the dopamine tyrosine hydroxylase-containing neurons of the rat retina. Brain Res. 214, 335–347 (1981). 466. Boatright, J. H., Rubim, N. M. & Iuvone, P. M. Regulation of endogenous dopamine release in amphibian retina by gamma-aminobutyric acid and glycine. Vis Neurosci 11, 1003–1012 (1994). 467. Morgan, W. W. & Kamp, C. W. Effect of strychnine and of bicuculline on dopamine synthesis in retinas of dark-maintained rats. Brain Res. 278, 362–365 (1983). 468. Jobling, A. I., Vessey, K. A., Waugh, M., Mills, S. A. & Fletcher, E. L. A naturally occurring mouse model of achromatopsia: Characterization of the mutation in cone transducin and subsequent retinal phenotype. Investig. Ophthalmol. Vis. Sci. 54, 3350– 3359 (2013). 469. Smallwood, P. M. et al. Genetically engineered mice with an additional class of cone photoreceptors: implications for the evolution of color vision. Proc. Natl. Acad. Sci. U. S. A. 100, 11706–11711 (2003). 470. Campbell, I. L., Hofer, M. J. & Pagenstecher, A. Transgenic models for cytokine- induced neurological disease. Biochimica et Biophysica Acta - Molecular Basis of Disease 1802, 903–917 (2010). 471. Roseboom, P. H. et al. Natural melatonin ‘knockdown’ in C57BL/6J mice: Rare mechanism truncates serotonin N-acetyltransferase. Mol. Brain Res. 63, 189–197 (1998). 472. Biel, M. et al. Selective loss of cone function in mice lacking the cyclic nucleotide-gated channel CNG3. Proc. Natl. Acad. Sci. U. S. A. 96, 7553–7557 (1999). 473. Chang, B. et al. Cone photoreceptor function loss-3, a novel mouse model of achromatopsia due to a mutation in Gnat2. Investig. Ophthalmol. Vis. Sci. 47, 5017– 5021 (2006). 474. Chang, B. et al. A homologous genetic basis of the murine cpfl1 mutant and human achromatopsia linked to mutations in the PDE6C gene. Proc. Natl. Acad. Sci. 106, 19581–6 (2009). 475. Pittler, S. J. & Baehr, W. Identification of a nonsense mutation in the rod photoreceptor

191

cGMP phosphodiesterase beta-subunit gene of the rd mouse. Proc. Natl. Acad. Sci. U. S. A. 88, 8322–8326 (1991). 476. Jones, B. W. & Marc, R. E. Retinal remodeling during retinal degeneration. Experimental Eye Research 81, 123–137 (2005). 477. Marc, R. E. & Jones, B. W. Retinal remodeling in inherited photoreceptor degenerations. Mol. Neurobiol. 28, 139–147 (2003). 478. Lucas, R. J. et al. Diminished pupillary light reflex at high irradiances in melanopsin- knockout mice. Science (80-. ). 299, 245–247 (2003). 479. Campbell, I. L. et al. Neurologic disease induced in transgenic mice by cerebral overexpression of interleukin 6. Proc. Natl. Acad. Sci. 90, 10061–5 (1993). 480. Chiang, C. S., Stalder, A., Samimi, A. & Campbell, I. L. Reactive gliosis as a consequence of interleukin-6 expression in the brain: studies in transgenic mice. Dev. Neurosci. 16, 212–21 (1994). 481. Govardovskii, V. I., Fyhrquist, N., Reuter, T., Kuzmin, D. G. & Donner, K. In search of the visual pigment template. Vis. Neurosci. 17, 509–528 (2000). 482. Lyubarsky, A. L., Daniele, L. L. & Pugh, E. N. From candelas to photoisomerizations in the mouse eye by rhodopsin bleaching in situ and the light-rearing dependence of the major components of the mouse ERG. in Vision Research 44, 3235–3251 (2004). 483. Headington, K., Choi, S. S., Nickla, D. & Doble, N. Single cell imaging of the chick retina with adaptive optics. Curr. Eye Res. 36, 947–57 (2011). 484. Kongsui, R., Beynon, S. B., Johnson, S. J. & Walker, F. R. Quantitative assessment of microglial morphology and density reveals remarkable consistency in the distribution and morphology of cells within the healthy prefrontal cortex of the rat. J. Neuroinflammation (2014). doi:10.1186/s12974-014-0182-7 485. Proll, M. A., Kamp, C. W. & Morgan, W. W. Use of liquid chromatography with electrochemistry to measure effects of varying intensities of white light on DOPA accumulation in rat retinas. Life Sci. 30, 11–19 (1982). 486. Djamgoz, M. B. A., Hankins, M. W., Hirano, J. & Archer, S. N. Neurobiology of retinal dopamine in relation to degenerative states of the tissue. Vision Research 37, 3509–3529 (1997). 487. Jackson, C. R., Capozzi, M., Dai, H. & McMahon, D. G. Circadian Perinatal Photoperiod Has Enduring Effects on Retinal Dopamine and Visual Function. J. Neurosci. 34, 4627–4633 (2014). 488. Dai, H., Jackson, C. R., Davis, G. L., Blakely, R. D. & McMahon, D. G. Is dopamine

192

transporter-mediated dopaminergic signaling in the retina a noninvasive biomarker for attention-deficit/ hyperactivity disorder? A study in a novel dopamine transporter variant Val559 transgenic mouse model. J. Neurodev. Disord. (2017). doi: 10.1186/s11689- 017-9215-8. 489. Wu, X. H. et al. The role of retinal dopamine in C57BL/6 mouse refractive development as revealed by intravitreal administration of 6-hydroxydopamine. Investig. Ophthalmol. Vis. Sci. 57, 5393–5404 (2016). 490. Lahouaoui, H., Coutanson, C., Cooper, H. M., Bennis, M. & Dkhissi-Benyahya, O. Diabetic retinopathy alters light-induced clock gene expression and dopamine levels in the mouse retina. Mol. Vis. 22, 959–69 (2016). 491. Pardue, M. T. et al. High susceptibility to experimental myopia in a mouse model with a retinal on pathway defect. Investig. Ophthalmol. Vis. Sci. 49, 706–712 (2008). 492. Kirsch, M. & Wagner, H. J. Release pattern of endogenous dopamine in teleost retinae during light adaptation and pharmacological stimulation. Vision Res. 29, 147–154 (1989). 493. Dowling, J. E. Effects Of Light Stimuli On The Release Of Dopamine From Interplexiform Cells In The White Perch Retina. Vis. Neurosci. 7, 451–458 (1991). 494. O’Leary, T., van Rossum, M. C. W. & Wyllie, D. J. A. Homeostasis of intrinsic excitability in hippocampal neurones: Dynamics and mechanism of the response to chronic depolarization. J. Physiol. 588, 157–170 (2010). 495. Feldkaemper, M. & Schaeffel, F. An updated view on the role of dopamine in myopia. Exp. Eye Res. 114, 106–19 (2013). 496. Steffen, M. A. et al. Spontaneous activity of dopaminergic retinal neurons. Biophys. J. 85, 2158–2169 (2003). 497. Mills, S. L. et al. Dopaminergic modulation of tracer coupling in a ganglion-amacrine cell network. Vis. Neurosci. 24, 593–608 (2007). 498. Boatright, J. H., Hoel, M. J. & Michael Iuvone, P. Stimulation of endogenous dopamine release and metabolism in amphibian retina by light- and K+-evoked depolarization. Brain Res. 482, 164–168 (1989). 499. Umino, O. & Dowling, J. E. Dopamine release from interplexiform cells in the retina: effects of GnRH, FMRFamide, bicuculline, and enkephalin on horizontal cell activity. J Neurosci 11, 3034–3046 (1991). 500. Wang, J. S., Estevez, M. E., Cornwall, M. C. & Kefalov, V. J. Intra-retinal visual cycle required for rapid and complete cone dark adaptation. Nat. Neurosci. 12, 295–302

193

(2009). 501. Mata, N. L., Radu, R. A., Clemmons, R. S. & Travis, G. H. Isomerization and oxidation of vitamin A in cone-dominant retinas: A novel pathway for visual-pigment regeneration in daylight. Neuron 36, 69–80 (2002). 502. Ala-Laurila, P., Cornwall, M. C., Crouch, R. K. & Kono, M. The action of 11-cis-Retinol on cone opsins and intact cone photoreceptors. J. Biol. Chem. 284, 16492–16500 (2009). 503. Walker, M. T., Brown, R. L., Cronin, T. W. & Robinson, P. R. Photochemistry of retinal chromophore in mouse melanopsin. Proc. Natl. Acad. Sci. U. S. A. 105, 8861–8865 (2008). 504. Sexton, T., Buhr, E. & Van Gelder, R. N. Melanopsin and mechanisms of non-visual ocular photoreception. J. Biol. Chem. 287, 1649–56 (2012). 505. Zhao, X., Pack, W., Khan, N. W. & Wong, K. Y. Prolonged Inner Retinal Photoreception Depends on the Visual Retinoid Cycle. J. Neurosci. 36, 4209–17 (2016). 506. Saari, J. C. Biochemistry of visual pigment regeneration: The Friedenwald lecture. in Investigative Ophthalmology and Visual Science (2000). 507. McBee, J. K., Palczewski, K., Baehr, W. & Pepperberg, D. R. Confronting complexity: The interlink of phototransduction and retinoid metabolism in the vertebrate retina. Prog. Retin. Eye Res. 20, 469–529 (2001). 508. Hirasawa, H., Betensky, R. A. & Raviola, E. Corelease of Dopamine and GABA by a Retinal Dopaminergic Neuron. J. Neurosci. 32, 13281–13291 (2012). 509. Stone, R. A., Lin, T., Laties, A. M. & Iuvone, P. M. Retinal dopamine and form- deprivation myopia. Proc Natl Acad Sci U S A 86, 704–706 (1989). 510. Ashby, R., Ohlendorf, A. & Schaeffel, F. The effect of ambient illuminance on the development of deprivation myopia in chicks. Investig. Ophthalmol. Vis. Sci. 50, 5348– 54 (2009). 511. Cohen, Y., Belkin, M., Yehezkel, O., Solomon, A. S. & Polat, U. Dependency between light intensity and refractive development under light-dark cycles. Exp. Eye Res. 92, 40– 6 (2011). 512. Cohen, Y., Peleg, E., Belkin, M., Polat, U. & Solomon, A. S. Ambient illuminance, retinal dopamine release and refractive development in chicks. Exp. Eye Res. 103, 33– 40 (2012). 513. Smith, E. L., Hung, L. F. & Huang, J. Protective effects of high ambient lighting on the development of form-deprivation myopia in rhesus monkeys. Investig. Ophthalmol. Vis. Sci. 53, 421–8 (2012).

194

514. Wu, P. C., Tsai, C. L., Wu, H. L., Yang, Y. H. & Kuo, H. K. Outdoor activity during class recess reduces myopia onset and progression in school children. Ophthalmology 120, 1080–5 (2013). 515. Li, W. et al. The effect of spectral property and intensity of light on natural refractive development and compensation to negative lenses in guinea pigs. Investig. Ophthalmol. Vis. Sci. 55, 6324–6332 (2014). 516. Chen, S. et al. Bright light suppresses form-deprivation myopia development with activation of dopamine d1 receptor signaling in the ON pathway in retina. Investig. Ophthalmol. Vis. Sci. 58, 2306–2316 (2017). 517. Wu, D. K. & Cepko, C. L. Development of dopaminergic neurons is insensitive to optic nerve section in the neonatal rat retina. Brain Res Dev Brain Res 74, 253–260 (1993). 518. Martin-Martinelli, E., Simon, A., Vigny, A. & Nguyen-Legros, J. Postnatal development of tyrosine-hydroxylase-immunoreactive cells in the rat retina. Morphology and distribution. Dev. Neurosci. 11, 11–25 (1989). 519. Foster, G. A., Schultzberg, M., Goldstein, M. & Hökfelt, T. Differential ontogeny of three putative catecholamine cell types in the postnatal rat retina. Dev. Brain Res. 22, 187–196 (1985). 520. Kato, S., Nakamura, T. & Negishi, K. Postnatal development of dopaminergic cells in the rat retina. J. Comp. Neurol. 191, 227–236 (1980). 521. Arroyo, D. A., Kirkby, L. A. & Feller, M. B. Retinal Waves Modulate an Intraretinal Circuit of Intrinsically Photosensitive Retinal Ganglion Cells. J. Neurosci. 36, 6892– 6905 (2016). 522. Bertolesi, G. E., Hehr, C. L. & McFarlane, S. Wiring the retinal circuits activated by light during early development. Neural Dev. (2014). doi: 10.1186/1749-8104-9-3. 523. Kirkby, L. A. & Feller, M. B. Intrinsically photosensitive ganglion cells contribute to plasticity in retinal wave circuits. Proc. Natl. Acad. Sci. (2013). doi:10.1073/pnas.1222150110 524. Morgan, J. L., Schubert, T. & Wong, R. O. L. Developmental patterning of glutamatergic synapses onto retinal ganglion cells. Neural Dev. (2008). doi: 10.1186/1749-8104-3-8 525. Morgan, J. L., Dhingra, A., Vardi, N. & Wong, R. O. L. Axons and dendrites originate from neuroepithelial-like processes of retinal bipolar cells. Nat. Neurosci. 9, 85–92 (2006). 526. Tian, N. & Copenhagen, D. R. Visual stimulation is required for refinement on ON and

195

OFF pathways in postnatal retina. Neuron 39, 85–96 (2003). 527. Hilgen, G. et al. Pan-retinal characterisation of Light Responses from Ganglion Cells in the Developing Mouse Retina. Sci. Rep. 7, 42330 (2017). 528. Herrmann, R. et al. Rod Vision Is Controlled by Dopamine-Dependent Sensitization of Rod Bipolar Cells by GABA. Neuron 72, 101–110 (2011). 529. Yang, J., Pahng, J. & Wang, G. Y. Dopamine modulates the off pathway in light-adapted mouse retina. J. Neurosci. Res. 91, 138–150 (2013). 530. Bloomfield, S. A. & Völgyi, B. The diverse functional roles and regulation of neuronal gap junctions in the retina. Nat. Rev. Neurosci. 10, 495–506 (2009). 531. Zhang, A. J., Jacoby, R. & Wu, S. M. Light- and dopamine-regulated receptive field plasticity in primate horizontal cells. J. Comp. Neurol. 519, 2125–34 (2011). 532. Heckelen, a, Hermel, M., Kondring, B. & Schrage, N. F. Ascorbic acid reversibly inhibits proliferation of retinal pigment epithelial cells. Acta Ophthalmol. Scand. 82, 564–8 (2004). 533. Böhmer, J. A., Sellhaus, B. & Schrage, N. F. Effects of ascorbic acid on retinal pigment epithelial cells. Curr. Eye Res. 23, 206–14 (2001). 534. Calero, C. I. et al. Allosteric Modulation of Retinal GABA Receptors by Ascorbic Acid. J. Neurosci. 31, 9672–9682 (2011). 535. Neal, M. J., Cunningham, J. R. & Matthews, K. L. Release of endogenous ascorbic acid preserves extracellular dopamine in the mammalian retina. Investig. Ophthalmol. Vis. Sci. 40, 2983–2987 (1999). 536. Riss, T. L. et al. Cell Viability Assays. Assay Guid. Man. [Internet] 114, 785–796 (2013). 537. Witkovsky, P., Gábriel, R. & Križaj, D. Anatomical and neurochemical characterization of dopaminergic interplexiform processes in mouse and rat retinas. J. Comp. Neurol. 510, 158–174 (2008). 538. Liu, C., Kershberg, L., Wang, J., Schneeberger, S. & Kaeser, P. S. Dopamine Secretion Is Mediated by Sparse Active Zone-like Release Sites. Cell 172, 706–718.e15 (2018). 539. Bauer, B., Ehinger, B. & Åberg, L. [3H]-Dopamine release from the rabbit retina. Albr. von Graefes Arch. für Klin. und Exp. Ophthalmol. 215, 71–78 (1980). 540. Nakamura, Y., Nakajima, S. & Grundfest, H. The Action of Tetrodotoxin on Electrogenic Components of Squid Giant Axons. J. Gen. Physiol. 48, 985–996 (1965). 541. Bloomfield, S. A. & Xin, D. Surround inhibition of mammalian AII amacrine cells is generated in the proximal retina. J. Physiol. 523, 771–783 (2000).

196

542. Tian, M., Jarsky, T., Murphy, G. J., Rieke, F. & Singer, J. H. Voltage-Gated Na Channels in AII Amacrine Cells Accelerate Scotopic Light Responses Mediated by the Rod Bipolar Cell Pathway. J. Neurosci. 30, 4650–4659 (2010). 543. Taylor, W. R. TTX attenuates surround inhibition in rabbit retinal ganglion cells. Vis. Neurosci. 16, 285–290 (1999). 544. Reifler, A. N. et al. All spiking, sustained on displaced amacrine cells receive gap- junction input from melanopsin ganglion cells. Curr. Biol. 25, 2763–2773 (2015). 545. Kawai, F., Horiguchi, M., Suzuki, H. & Miyachi, E. I. Na+ action potentials in human photoreceptors. Neuron 30, 451–458 (2001). 546. CUI, J. & PAN, Z.-H. Two types of cone bipolar cells express voltage-gated Na+ channels in the rat retina. Vis. Neurosci. 25, 635 (2008). 547. Pan, Z. H. & Hu, H. J. Voltage-dependent Na(+) currents in mammalian retinal cone bipolar cells. J. Neurophysiol. 84, 2564–2571 (2000). 548. Zenisek, D., Henry, D., Studholme, K., Yazulla, S. & Matthews, G. Voltage-dependent sodium channels are expressed in nonspiking retinal bipolar neurons. J. Neurosci. 21, 4543–4550 (2001). 549. Ichinose, T., Shields, C. R. & Lukasiewicz, P. D. Sodium channels in transient retinal bipolar cells enhance visual responses in ganglion cells. J Neurosci 25, 1856–1865 (2005). 550. Grunert, U. & Wassle, H. Glycine receptors in the rod pathway of the macaque monkey retina. Vis. Neurosci. 13, 101–115 (1996). 551. Suzuki, S., Tachibana, M. & Kaneko, A. Effects of glycine and GABA on isolated bipolar cells of the mouse retina. J. Physiol. 421, 645–652 (1990). 552. Matsubayashi, H., Alkondon, M., Pereira, E. F., Swanson, K. L. & Albuquerque, E. X. Strychnine: a potent competitive antagonist of alpha-bungarotoxin-sensitive nicotinic acetylcholine receptors in rat hippocampal neurons. J. Pharmacol. Exp. Ther. 284, 904– 913 (1998). 553. Renna, J. M., Strang, C. E., Amthor, F. R. & Keyser, K. T. Strychnine, but not PMBA, inhibits neuronal nicotinic acetylcholine receptors expressed by rabbit retinal ganglion cells. Vis Neurosci 24, 503–511 (2007). 554. Woodward, R. M., Polenzani, L. & Miledi, R. Characterization of bicuculline/baclofen- insensitive gamma-aminobutyric acid receptors expressed in Xenopus oocytes. I. Effects of Cl- channel inhibitors. Mol Pharmacol 42, 165–173 (1992). 555. Newland, C. F. & Cull-Candy, S. G. On the mechanism of action of picrotoxin on GABA

197

receptor channels in dissociated sympathetic neurones of the rat. J. Physiol. 447, 191– 213 (1992). 556. Feigenspan, A., Wässle, H. & Bormann, J. Pharmacology of GABA receptor Cl- channels in rat retinal bipolar cells. Nature. 361, 159–62 (1993). 557. Yoon, K. W., Covey, D. F. & Rothman, S. M. Multiple mechanisms of picrotoxin block of GABA-induced currents in rat hippocampal neurons. J. Physiol. 464, 423–439 (1993). 558. Qian, H. & Dowling, J. E. Pharmacology of novel GABA receptors found on rod horizontal cells of the white perch retina. J Neurosci 14, 4299–4307 (1994). 559. Olsen, R. W. Picrotoxin-like channel blockers of GABAA receptors. Proc. Natl. Acad. Sci. U. S. A. 103, 6081–2 (2006). 560. Feigenspan, A., Gustincich, S. & Raviola, E. Pharmacology of GABA(A) receptors of retinal dopaminergic neurons. J Neurophysiol 84, 1697–1707 (2000). 561. O’Connor, P., Dorison, S. J., Watling, K. J. & Dowling, J. E. Factors affecting release of 3H-dopamine from perfused carp retina. J Neurosci 6, 1857–1865 (1986). 562. Boatright, J. H., Rubim, N. M. & Iuvone, P. M. Regulation of endogenous dopamine release in amphibian retina by melatonin: The role of gaba. Vis. Neurosci. 11, 1013– 1018 (1994). 563. Hartveit, E. Reciprocal synaptic interactions between rod bipolar cells and amacrine cells in the rat retina. J. Neurophysiol. 81, 2923–2936 (1999). 564. Eggers, E. D. & Lukasiewicz, P. D. Interneuron Circuits Tune Inhibition in Retinal Bipolar Cells. J. Neurophysiol. 103, 25–37 (2010). 565. Lukasiewicz, P. D. & Werblin, F. S. A novel GABA receptor modulates synaptic transmission from bipolar to ganglion and amacrine cells in the tiger salamander retina. J Neurosci (1994). doi; 10.1523/JNEUROSCI 566. Connaughton, V. P., Nelson, R. & Bender, A. M. Electrophysiological evidence of GABAA and GABAC receptors on zebrafish retinal bipolar cells. Vis. Neurosci. 25, 139–53 (2008). 567. Arnarsson, A. & Eysteinsson, T. The role of GABA in modulating the Xenopus electroretinogram. Vis. Neurosci. 14, 1143–1152 (1997). 568. Dhingra, A. & Vardi, N. mGlu receptors in the retina. Wiley Interdiscip. Rev. Membr. Transp. Signal. 1, 641–653 (2012). 569. Gilbertson, T. A., Scobey, R. & Wilson, M. Permeation of calcium ions through non- NMDA glutamate channels in retinal bipolar cells. Science (80-. ). 251, 1613–1615

198

(1991). 570. Zhou, Z. J., Fain, G. L. & Dowling, J. E. The excitatory and inhibitory amino acid receptors on horizontal cells isolated from the white perch retina. J. Neurophysiol. 70, 8–19 (1993). 571. Cohen, E. D. & Miller, R. F. The role of NMDA and non-NMDA excitatory amino acid receptors in the functional organization of primate retinal ganglion cells. Vis. Neurosci. 11, 317–32 (1994). 572. Yu, W. & Miller, R. F. NBQX, an improved non-NMDA antagonist studied in retinal ganglion cells. Brain Res. 692, 190–194 (1995). 573. Hughes, T. E. Are there ionotropic glutamate receptors on the rod bipolar cell of the mouse retina? Vis. Neurosci. 14, 103–109 (1997). 574. Vardi, N., Morigiwa, K., Wang, T. L., Shi, Y. J. & Sterling, P. Neurochemistry of the mammalian cone ‘synaptic complex’. Vision Res. 38, 1359–69 (1998). 575. Wang, L. Y. & MacDonald, J. F. Modulation by magnesium of the affinity of NMDA receptors for glycine in murine hippocampal neurones. J. Physiol. 486, 83–95 (1995). 576. Nikolaev, M. V., Magazanik, L. G. & Tikhonov, D. B. Influence of external magnesium ions on the NMDA receptor channel block by different types of organic cations. Neuropharmacology 62, 2078–2085 (2012). 577. Ishii, T. et al. Molecular characterization of the family of the N-methyl-D-aspartate receptor subunits. J. Biol. Chem. 268, 2836–2843 (1993). 578. Kutsuwada, T. et al. Molecular diversity of the NMDA receptor channel. Nature 358, 36–41 (1992). 579. Monyer, H., Burnashev, N., Laurie, D. J., Sakmann, B. & Seeburg, P. H. Developmental and regional expression in the rat brain and functional properties of four NMDA receptors. Neuron 12, 529–540 (1994). 580. Michaelis, E. K. Molecular biology of glutamate receptors in the central nervous system and their role in excitotoxicity, oxidative stress and aging. Prog. Neurobiol. 54, 369– 415 (1998). 581. Qian, A., Buller, A. L. & Johnson, J. W. NR2 subunit-dependence of NMDA receptor channel block by external Mg2+. J. Physiol. 562, 319–331 (2005). 582. Kuner, T. & Schoepfer, R. Multiple structural elements determine subunit specificity of Mg2+ block in NMDA receptor channels. J. Neurosci. (1996). 583. Shen, Y., Liu, X. L. & Yang, X. L. N-methyl-D-aspartate receptors in the retina. Mol. Neurobiol. 34, 163–179 (2006).

199

584. Zhou, Y., Tencerová, B., Hartveit, E. & Veruki, M. L. Functional NMDA receptors are expressed by both AII and A17 amacrine cells in the rod pathway of the mammalian retina. J. Neurophysiol. 115, 389–403 (2016). 585. Pan, F., Mills, S. L. & Massey, S. C. Screening of gap junction antagonists on dye coupling in the rabbit retina. Vis. Neurosci. 24, 609–618 (2007). 586. D.-Q. Zhang, P. J. Sollars, G. E. Pickard, D. G. M. Signaling by Ganglion-Cell photoreceptors to Dopaminergic Amacrine Cells Requires the Photopigment Melanopsin and AMPA-Type Glutamate Receptors. Invest. Ophthalmol. Vis. Sci. 51, 1206 (2010). 587. Tsukamoto, Y. et al. A Novel Connection between Rods and ON Cone Bipolar Cells Revealed by Ectopic Metabotropic Glutamate Receptor 7 (mGluR7) in mGluR6- Deficient Mouse Retinas. J. Neurosci. 27, 6261–6267 (2007). 588. Strettoi, E., Raviola, E. & Dacheux, R. F. Synaptic connections of the narrow-field, bistratified rod amacrine cell (AII) in the rabbit retina. J. Comp. Neurol. 325, 152–168 (1992). 589. Graydon, C. W. et al. Synaptic Transfer between Rod and Cone Pathways Mediated by AII Amacrine Cells in the Mouse Retina. Curr. Biol. 28, 2739–2751.e3 (2018). 590. Elgueta, C., Leroy, F., Vielma, A. H., Schmachtenberg, O. & Palacios, A. G. Electrical coupling between A17 cells enhances reciprocal inhibitory feedback to rod bipolar cells. Sci. Rep. (2018). doi: 10.1038/s41598-018-21119-0 591. Wang, S., Villegas-Pérez, M. P., Vidal-Sanz, M. & Lund, R. D. Progressive optic axon dystrophy and vascular changes in rd mice. Investig. Ophthalmol. Vis. Sci. 41, 537–545 (2000). 592. Atkinson, C. L., Feng, J. & Zhang, D.-Q. Functional integrity and modification of retinal dopaminergic neurons in the rd1 mutant mouse: roles of melanopsin and GABA. J. Neurophysiol. 109, 1589–99 (2013). 593. Semo, M., Coffey, P., Gias, C. & Vugler, A. Retrograde melanopsin signaling increases with age in retinal degenerate mice lacking rods and the majority of cones. Investig. Ophthalmol. Vis. Sci. 57, 115–25 (2016). 594. Roth, B. L. DREADDs for Neuroscientists. Neuron 89, 683–694 (2016). 595. Storchi, R. et al. Melanopsin-driven increases in maintained activity enhance thalamic visual response reliability across a simulated dawn. Proc. Natl. Acad. Sci. 112, E5734– E5743 (2015). 596. Li, S. et al. Promoting axon regeneration in the adult CNS by modulation of the

200

melanopsin/GPCR signaling. Proc. Natl. Acad. Sci. 113, 1937–1942 (2016). 597. Keenan, W. T., Fernandez, D. C., Shumway, L. J., Zhao, H. & Hattar, S. Eye-Drops for Activation of DREADDs. Front. Neural Circuits (2017). doi:10.3389/fncir.2017.00093 598. Altimus, C. M. et al. Rod photoreceptors drive circadian photoentrainment across a wide range of light intensities. Nat. Neurosci. 13, 1107–1112 (2010). 599. Invergo, B. M., Dell’Orco, D., Montanucci, L., Koch, K. W. & Bertranpetit, J. A comprehensive model of the phototransduction cascade in mouse rod cells. Mol. Biosyst. 10, 1481–1489 (2014). 600. Margolis, D. J., Newkirk, G., Euler, T. & Detwiler, P. B. Functional stability of retinal ganglion cells after degeneration-induced changes in synaptic input. J. Neurosci. 28, 6526–6536 (2008). 601. Menzler, J. & Zeck, G. Network Oscillations in Rod-Degenerated Mouse Retinas. J. Neurosci. 31, 2280–91 (2011). 602. Borowska, J., Trenholm, S. & Awatramani, G. B. An Intrinsic Neural Oscillator in the Degenerating Mouse Retina. J. Neurosci. 31, 5000–5012 (2011). 603. Wang, M., Schaeffel, F., Jiang, B. & Feldkaemper, M. Effects of Light of Different Spectral Composition on Refractive Development and Retinal Dopamine in Chicks. Invest. Ophthalmol. Vis. Sci. 59, 4413–4424 (2018). 604. Puthussery, T., Gayet-Primo, J., Pandey, S., Duvoisin, R. M. & Taylor, W. R. Differential loss and preservation of glutamate receptor function in bipolar cells in the rd10 mouse model of retinitis pigmentosa. Eur. J. Neurosci. 29, 1533–1542 (2009). 605. Barhoum, R. et al. Functional and structural modifications during retinal degeneration in the rd10 mouse. Neuroscience 155, 698–713 (2008). 606. Strettoi, E. & Pignatelli, V. Modifications of retinal neurons in a mouse model of retinitis pigmentosa. Proc Natl Acad Sci U S A 97, 11020–11025 (2000). 607. Strettoi, E., Pignatelli, V., Rossi, C., Porciatti, V. & Falsini, B. Remodeling of second- order neurons in the retina of rd/rd mutant mice. in Vision Research 43, 867–877 (2003). 608. Strettoi, E., Porciatti, V., Falsini, B., Pignatelli, V. & Rossi, C. Morphological and functional abnormalities in the inner retina of the rd/rd mouse. J Neurosci 22, 5492– 5504 (2002). 609. Varela, C., Igartua, I., De La Rosa, E. J. & De La Villa, P. Functional modifications in rod bipolar cells in a mouse model of retinitis pigmentosa. in Vision Research 43, 879– 885 (2003). 610. Lucas, R. J. et al. Diminished pupillary light reflex at high irradiances in melanopsin-

201

knockout mice. Science 299, 245–247 (2003). 611. Mrosovsky, N. & Hattar, S. Impaired masking responses to light in melanopsin- knockout mice. Chronobiol. Int. 20, 989–99 (2003). 612. Schmidt, T. M. & Kofuji, P. Differential cone pathway influence on intrinsically photosensitive retinal ganglion cell subtypes. J. Neurosci. 30, 16262–71 (2010). 613. Milner, E. S. & Do, M. T. H. A Population Representation of Absolute Light Intensity in the Mammalian Retina. Cell 171, 865–876 (2017). 614. Lee, J. W., Lim, M. Y., Park, Y. S., Park, S. J. & Kim, I. Reexamination of Dopaminergic Amacrine Cells in the Rabbit Retina : Confocal Analysis with Double- and Triple- labeling Immunohistochemistry. 26, 329–338 (2017). 615. Munteanu, T. et al. Light-dependent pathways for dopaminergic amacrine cell development and function. Elife 7, e39866 (2018). 616. Streit, W. J., Mrak, R. E. & Griffin, W. S. T. Microglia and neuroinflammation: A pathological perspective. J. Neuroinflammation (2004). doi:10.1186/1742-2094-1-14 617. Kim, Y. S. & Joh, T. H. Microglia, major player in the brain inflammation: Their roles in the pathogenesis of Parkinson’s disease. Experimental and Molecular Medicine 38, 333–347 (2006). 618. Mrak, R. E. Microglia in Alzheimer brain: A neuropathological perspective. Int. J. Alzheimers. Dis. (2012). doi:10.1155/2012/165021 619. Faden, A. I. & Loane, D. J. Chronic Neurodegeneration After Traumatic Brain Injury: Alzheimer Disease, Chronic Traumatic Encephalopathy, or Persistent Neuroinflammation? Neurotherapeutics 12, 143–150 (2015). 620. Kim, Y. K., Na, K. S., Myint, A. M. & Leonard, B. E. The role of pro-inflammatory cytokines in neuroinflammation, neurogenesis and the neuroendocrine system in major depression. Prog. Neuro-Psychopharmacology Biol. Psychiatry 64, 277–284 (2016). 621. Schain, M. & Kreisl, W. C. Neuroinflammation in Neurodegenerative Disorders—a Review. Current Neurology and Neuroscience Reports 17, (2017). 622. Notter, T. & Meyer, U. Microglia and schizophrenia: where next? Molecular Psychiatry 22, 788–789 (2017). 623. Chiu, K., Yeung, S., So, K. & Chang, R. C. Modulation of morphological changes of microglia and neuroprotection by monocyte chemoattractant protein-1 in experimental glaucoma. Nature 7, 61–68 (2010). 624. Fischer, F., Martin, G. & Agostini, H. T. Activation of retinal microglia rather than microglial cell density correlates with retinal neovascularization in the mouse model of

202

oxygen-induced retinopathy. J Neuroinflammation (2011). doi: 10.1186/1742-2094-8- 120 625. Karlstetter, M. et al. Progress in Retinal and Eye Research Retinal microglia : Just bystander or target for therapy ? Prog. Retin. Eye Res. 45, 30–57 (2015). 626. Madeira, M. H., Boia, R., Santos, P. F., Ambrósio, A. F. & Santiago, A. R. Contribution of microglia-mediated neuroinflammation to retinal degenerative diseases. Mediators of Inflammation (2015). doi: 10.1155/2015/673090 627. Block, M. L. & Hong, J. S. Microglia and inflammation-mediated neurodegeneration: Multiple triggers with a common mechanism. Progress in Neurobiology 76, 77–98 (2005). 628. Lull, M. E. & Block, M. L. Microglial Activation and Chronic Neurodegeneration. Neurotherapeutics 7, 354–365 (2010). 629. Huang, T., Cui, J., Li, L., Hitchcock, P. F. & Li, Y. The role of microglia in the neurogenesis of zebrafish retina. Biochem. Biophys. Res. Commun. 421, 214–220 (2012). 630. Ferrer-Martín, R. M. et al. Microglial cells in organotypic cultures of developing and adult mouse retina and their relationship with cell death. Exp. Eye Res. 121, 42–57 (2014). 631. Checchin, D., Sennlaub, F., Levavasseur, E., Leduc, M. & Chemtob, S. Potential role of microglia in retinal blood vessel formation. Invest. Ophthalmol. Vis. Sci. 47, 3595–3602 (2006). 632. Damani, M. R. et al. Age-related alterations in the dynamic behavior of microglia. Aging Cell 10, 263–276 (2011). 633. Diaz‐Araya, C. M., Provis, J. M., Penfold, P. L. & Billson, F. A. Development of microglial topography in human retina. J. Comp. Neurol. 363, 53–68 (1995). 634. Chen, L., Yang, P. & Kijlstra, A. Distribution, markers, and functions of retinal microglia. Ocul. Immunol. Inflamm. 10, 27–39 (2002). 635. Lee, J. E., Liang, K. J., Fariss, R. N. & Wong, W. T. Ex vivo dynamic imaging of retinal microglia using time-lapse confocal microscopy. Investig. Ophthalmol. Vis. Sci. 49, 4169–4176 (2008). 636. Sappington, R. M. & Calkins, D. J. Pressure-induced regulation of IL-6 in retinal glial cells: Involvement of the ubiquitin/proteasome pathway and NFκB. Investig. Ophthalmol. Vis. Sci. 47, 3860–3869 (2006). 637. Sims, S. M., Holmgren, L., Cathcart, H. M. & Sappington, R. M. Spatial regulation of

203

interleukin-6 signaling in response to neurodegenerative stressors in the retina. Am. J. Neurodegener. Dis. 1, 168–79 (2012). 638. Chen, K. H., Wu, C. C., Roy, S., Lee, S. M. & Liu, J. H. Increased interleukin-6 in aqueous humor of neovascular glaucoma. Investig. Ophthalmol. Vis. Sci. 40, 2627–2632 (1999). 639. Johnson, E. C. et al. Cell proliferation and interleukin-6-type cytokine signaling are implicated by gene expression responses in early optic nerve head injury in rat glaucoma. Investig. Ophthalmol. Vis. Sci. 52, 504–518 (2011). 640. Chidlow, G., Wood, J. P. M., Ebneter, A. & Casson, R. J. Interleukin-6 is an efficacious marker of axonal transport disruption during experimental glaucoma and stimulates neuritogenesis in cultured retinal ganglion cells. Neurobiol. Dis. 48, 568–581 (2012). 641. Sappington, R. M., Chan, M. & Calkins, D. J. Interleukin-6 protects retinal ganglion cells from pressure-induced death. Investig. Ophthalmol. Vis. Sci. 47, 2932–2942 (2006). 642. Santos, A. M. et al. Embryonic and postnatal development of microglial cells in the mouse retina. J. Comp. Neurol. 506, 224–39 (2008). 643. Kambhampati, S. P. et al. Systemic and intravitreal delivery of dendrimers to activated microglia/macrophage in ischemia/reperfusion mouse retina. Investig. Ophthalmol. Vis. Sci. 56, 4413–4424 (2015). 644. Seitz, R. & Tamm, E. R. Müller cells and microglia of the mouse eye react throughout the entire retina in response to the procedure of an intravitreal injection. Adv. Exp. Med. Biol. 801, 347–353 (2014). 645. Han, H. E., Kim, T. K., Son, H. J., Park, W. J. & Han, P. L. Activation of autophagy pathway suppresses the expression of iNOS, IL6 and cell death of LPS-stimulated microglia cells. Biomol. Ther. 21, 21–28 (2013). 646. Zhou, T. et al. Microglia Polarization with M1 / M2 Phenotype Changes in rd1 Mouse Model of Retinal Degeneration. 11, 1–11 (2017). 647. Noailles, A., Fernández-sánchez, L., Lax, P. & Cuenca, N. Microglia activation in a model of retinal degeneration and TUDCA neuroprotective effects. J Neuroinflammation (2014). doi: 10.1186/s12974-014-0186-3. 648. Zhang, C., Lam, T. T. & Tso, M. O. Heterogeneous populations of microglia/macrophages in the retina and their activation after retinal ischemia and reperfusion injury. Exp. Eye Res. 81, 700–709 (2005). 649. Johnson, L. V, Leitner, W. P., Staples, M. K. & Anderson, D. H. Complement activation

204

and inflammatory processes in Drusen formation and age related macular degeneration. Exp. Eye Res. 73, 887–896 (2001). 650. Zeng, X. X., Ng, Y. K. & Ling, E. A. Neuronal and microglial response in the retina of streptozotocin-induced diabetic rats. Vis. Neurosci. 17, 463–471 (2000). 651. Bejarano-Escobar, R., Blasco, M., Martin-Partido, G. & Francisco-Morcillo, J. Light- induced degeneration and microglial response in the retina of an epibenthonic pigmented teleost: age-dependent photoreceptor susceptibility to cell death. J. Exp. Biol. 215, 3799–3812 (2012). 652. Levine, E. S. et al. Rapid light-induced activation of retinal microglia in mice lacking Arrestin-1. Vision Res. 102, 71–9 (2014). 653. Färber, K., Pannasch, U. & Kettenmann, H. Dopamine and noradrenaline control distinct functions in rodent microglial cells. Mol. Cell. Neurosci. 29, 128–138 (2005). 654. Huck, J. H. J. et al. De novo expression of dopamine D2 receptors on microglia after stroke. J. Cereb. Blood Flow Metab. 35, 1804–11 (2015). 655. Mastroeni, D. et al. Microglial responses to dopamine in a cell culture model of Parkinson’s disease. Neurobiol. Aging 30, 1805–17 (2009). 656. Wang, X. et al. Requirement for Microglia for the Maintenance of Synaptic Function and Integrity in the Mature Retina. J. Neurosci. 36, 2827–2842 (2016). 657. Madeira, M. H., Boia, R., Santos, P. F., Ambrósio, A. F. & Santiago, A. R. Contribution of microglia-mediated neuroinflammation to retinal degenerative diseases. Mediators Inflamm. (2015). doi:10.1155/2015/673090 658. Okunuki, Y. et al. Microglia inhibit photoreceptor cell death and regulate immune cell infiltration in response to retinal detachment. Proc. Natl. Acad. Sci. 115, 6264–73 (2018). 659. Lavoie, J. et al. The electroretinogram as a biomarker of central dopamine and serotonin: Potential relevance to psychiatric disorders. Biol. Psychiatry 75, 479–486 (2014). 660. Reis, A. et al. Neuroretinal dysfunction with intact blood-retinal barrier and absent vasculopathy in type 1 diabetes. Diabetes 63, 3926–3937 (2014). 661. Matsui, Y., Katsumi, O., Sakaue, H. & Hirose, T. Electroretinogram b/a wave ratio improvement in central retinal vein obstruction. Br. J. Ophthalmol. 78, 191–8 (1994). 662. Seeliger, M., Kretschmann, U., Apfelstedt-Sylla, E., Rüther, K. & Zrenner, E. Multifocal electroretinography in retinitis pigmentosa. Am. J. Ophthalmol. 125, 214–226 (1998). 663. Klemp, K., Lund-Andersen, H., Sander, B. & Larsen, M. The effect of acute hypoxia and hyperoxia on the slow multifocal electroretinogram in healthy subjects. Investig.

205

Ophthalmol. Vis. Sci. 48, 3405–12 (2007). 664. Gerth, C., Delahunt, P. B., Alam, S., Morse, L. S. & Werner, J. S. Cone-mediated multifocal electroretinogram in age-related macular degeneration: Progression over a long-term follow-up. Arch. Ophthalmol. 124, 345–52 (2006). 665. Wilsey, L. J. & Fortune, B. Electroretinography in glaucoma diagnosis. Curr. Opin. Ophthalmol. 27, 118–24 (2016). 666. Wong, W. T. Microglial aging in the healthy CNS: phenotypes, drivers, and rejuvenation. Front. Cell. Neurosci. (2013). doi:10.3389/fncel.2013.00022 667. Spittau, B. Aging microglia-phenotypes, functions and implications for age-related neurodegenerative diseases. Front. Aging Neurosci. (2017). doi:10.3389/fnagi.2017.00194 668. Fernández-Arjona, M. del M., Grondona, J. M., Granados-Durán, P., Fernández- Llebrez, P. & López-Ávalos, M. D. Microglia Morphological Categorization in a Rat Model of Neuroinflammation by Hierarchical Cluster and Principal Components Analysis. Front. Cell. Neurosci. (2017). doi:10.3389/fncel.2017.00235 669. Preissler, J. et al. Altered microglial phagocytosis in GPR34-deficient mice. Glia 63, 206–215 (2015). 670. Liu, B. et al. Molecular consequences of activated microglia in the brain: Overactivation induces apoptosis. J. Neurochem. 77, 182–9 (2001). 671. Huang, Z. et al. Necroptosis in microglia contributes to neuroinflammation and retinal degeneration through TLR4 activation. Cell Death Differ. 25, 180–189 (2018). 672. Huang, Y. et al. Dual extra-retinal origins of microglia in the model of retinal microglia repopulation. Cell Discov. 4, (2018). 673. Liang, K. J. et al. Regulation of dynamic behavior of retinal microglia by CX3CR1 signaling. Investig. Ophthalmol. Vis. Sci. 50, 4444–51 (2009). 674. Zhang, Y. et al. Repopulating retinal microglia restore endogenous organization and function under CX3CL1-CX3CR1 regulation. Sci. Adv. (2018). doi: 10.1126/sciadv.aap8492 675. Owen, D. R. et al. Pro-inflammatory activation of primary microglia and macrophages increases 18 kDa translocator protein expression in rodents but not humans. J. Cereb. Blood Flow Metab. 37, 2679–2690 (2017). 676. Witcher, K. G., Eiferman, D. S. & Godbout, J. P. Priming the Inflammatory Pump of the CNS after Traumatic Brain Injury. Trends in Neurosciences 38, 609–620 (2015). 677. Roy, A., Fung, Y. K., Liu, X. & Pahan, K. Up-regulation of microglial CD11b

206

expression by nitric oxide. J. Biol. Chem. 281, 1497–80 (2006). 678. Janova, H. et al. CD14 is a key organizer of microglial responses to CNS infection and injury. Glia 64, 635–49 (2016). 679. Hendrickx, D. A. E., van Eden, C. G., Schuurman, K. G., Hamann, J. & Huitinga, I. Staining of HLA-DR, Iba1 and CD68 in human microglia reveals partially overlapping expression depending on cellular morphology and pathology. J. Neuroimmunol. 15, 12– 22 (2017). 680. Koizumi, S. et al. UDP acting at P2Y6receptors is a mediator of microglial phagocytosis. Nature 446, 1091–5 (2007). 681. Monif, M., Reid, C. A., Powell, K. L., Smart, M. L. & Williams, D. A. The P2X7 Receptor Drives Microglial Activation and Proliferation: A Trophic Role for P2X7R Pore. J. Neurosci. 29, 3781–91 (2009). 682. Monif, M., Burnstock, G. & Williams, D. A. Microglia: Proliferation and activation driven by the P2X7 receptor. Int. J. Biochem. Cell Biol. 42, 1753–6 (2010). 683. Vázquez-Villoldo, N. et al. P2X4 receptors control the fate and survival of activated microglia. Glia 62, 171–84 (2014). 684. Sipe, G. O. et al. Microglial P2Y12 is necessary for synaptic plasticity in mouse visual cortex. Nat. Commun. 7, 10905 (2016). 685. He, Y., Taylor, N., Fourgeaud, L. & Bhattacharya, A. The role of microglial P2X7: Modulation of cell death and cytokine release. J. Neuroinflammation (2017). doi:10.1186/s12974-017-0904-8 686. Mildner, A., Huang, H., Radke, J., Stenzel, W. & Priller, J. P2Y12receptor is expressed on human microglia under physiological conditions throughout development and is sensitive to neuroinflammatory diseases. Glia 65, 375–387 (2017). 687. Yang, X. et al. Microglia P2Y6 receptor is related to Parkinson’s disease through neuroinflammatory process. J. Neuroinflammation (2017). doi:10.1186/s12974-017- 0795-8 688. Monif, M. et al. Interleukin-1β has trophic effects in microglia and its release is mediated by P2X7R pore. J. Neuroinflammation (2016). doi:10.1186/s12974-016-0621-8 689. Natoli, R. et al. Microglia-derived IL-1β promotes chemokine expression by Müller cells and RPE in focal retinal degeneration. Mol. Neurodegener. 12, 31 (2017). 690. Hefendehl, J. K. et al. Homeostatic and injury-induced microglia behavior in the aging brain. Aging Cell 13, 60–69 (2014). 691. Tremblay, M.-È., Zettel, M. L., Ison, J. R., Allen, P. D. & Majewska, A. K. Effects of

207

aging and sensory loss on glial cells in mouse visual and auditory cortices. Glia 60, 541– 558 (2012). 692. Daynes, R. A. et al. Altered regulation of IL-6 production with normal aging. Possible linkage to the age-associated decline in dehydroepiandrosterone and its sulfated derivative. J. Immunol. 150, 5219–30 (1993). 693. Myśliwska, J., Bryl, E., Foerster, J. & Myśliwski, A. Increase of interleukin 6 and decrease of interleukin 2 production during the ageing process are influenced by the health status. Mech. Ageing Dev. 100, 313–328 (1998). 694. Dugan, L. L. et al. IL-6 mediated degeneration of forebrain GABAergic interneurons and cognitive impairment in aged mice through activation of neuronal NADPH oxidase. PLoS One (2009). doi: 10.1371/journal.pone.0005518. 695. Kiecolt-Glaser, J. K. et al. Chronic stress and age-related increases in the proinflammatory cytokine IL-6. Proc. Natl. Acad. Sci. 100, 9090–9095 (2003). 696. Tang, B. et al. Age-associated increase in interleukin 6 in MRL/lpr mice. Int. Immunol. 3, 273–278 (1991). 697. Foster, K. D., Conn, C. a & Kluger, M. J. Fever, tumor necrosis factor, and interleukin- 6 in young, mature, and aged Fischer 344 rats. Am. J. Physiol. 262, R211-5 (1992). 698. Kim, O. Y. et al. Effects of aging and menopause on serum interleukin-6 levels and peripheral blood mononuclear cell cytokine production in healthy nonobese women. Age (Omaha). 34, 415–425 (2012). 699. Beharka, A. A. et al. Interleukin-6 production does not increase with age. Journals Gerontol. - Ser. A Biol. Sci. Med. Sci. 56, B81–B88 (2001). 700. Kovacs, E. J. et al. Interleukin-6 contributes to age-related alteration of cytokine production by macrophages. Mediators Inflamm. (2010). doi: 10.1155/2010/475139 701. Grellner, W., Georg, T. & Wilske, J. Quantitative analysis of proinflammatory cytokines (IL-1β, IL-6, TNF-α) in human skin wounds. Forensic Sci. Int. 113, 254–64 (2000). 702. Caldeira, C. et al. Microglia change from a reactive to an age-like phenotype with the time in culture. Front. Cell. Neurosci. (2014). doi:10.3389/fncel.2014.00152 703. Norden, D. M. & Godbout, J. P. Review: Microglia of the aged brain: Primed to be activated and resistant to regulation. Neuropathol. Appl. Neurobiol. 39, 19–34 (2013). 704. Davis, B. M., Salinas-Navarro, M., Cordeiro, M. F., Moons, L. & Groef, L. De. Characterizing microglia activation: A spatial statistics approach to maximize information extraction. Sci. Rep. (2017). doi: 10.1038/s41598-017-01747-8 705. Karperien, A., Ahammer, H. & Jelinek, H. F. Quantitating the subtleties of microglial

208

morphology with fractal analysis. Front. Cell. Neurosci. (2013). doi:10.3389/fncel.2013.00003 706. Das Sarma, S., Chatterjee, K., Dinda, H., Chatterjee, D. & Das Sarma, J. Cytomorphological and Cytochemical Identification of Microglia. ISRN Immunol. (2013). doi: 10.1155/2013/205431 707. Raivich, G. Like cops on the beat: The active role of resting microglia. Trends in Neurosciences 28, 571–573 (2005). 708. Wake, H., Moorhouse, A. J., Jinno, S., Kohsaka, S. & Nabekura, J. Resting Microglia Directly Monitor the Functional State of Synapses In Vivo and Determine the Fate of Ischemic Terminals. J. Neurosci. 29, 3974–3980 (2009). 709. Tremblay, M. Ě., Lowery, R. L. & Majewska, A. K. Microglial interactions with synapses are modulated by visual experience. PLoS Biol. (2010). doi: 10.1371/journal.pbio.1000527 710. Davalos, D. et al. ATP mediates rapid microglial response to local brain injury in vivo. Nat. Neurosci. 8, 752–758 (2005). 711. Walker, F. R. et al. Dynamic structural remodelling of microglia in health and disease: A review of the models, the signals and the mechanisms. Brain, Behavior, and Immunity 37, 1–14 (2014). 712. Hristovska, I. & Pascual, O. Deciphering Resting Microglial Morphology and Process Motility from a Synaptic Prospect. Front. Integr. Neurosci. (2016). doi:10.3389/fnint.2015.00073 713. Zhang, Y., Peng, B., Wang, S., Liang, Y. & Yang, J. Image processing methods to elucidate spatial characteristics of retinal microglia after optic nerve transection. Sci Rep (2016). doi:10.1038/srep21816 714. Davis, B. M., Salinas-navarro, M., Cordeiro, M. F. & Moons, L. Characterizing microglia activation : a spatial statistics approach to maximize information extraction. Sci Rep (2017). doi:10.1038/s41598-017-01747-8 715. Von Bernhardi, R. & Ramirez, G. Microglia-astrocyte interaction in Alzheimer’s disease: Friends or foes for the nervous system? Biol. Res. 34, 123–128 (2001). 716. Shih, A. Y. Policing the Police: Astrocytes Modulate Microglial Activation. J. Neurosci. 26, 3887–3888 (2006). 717. Wang, M., Ma, W., Zhao, L., Fariss, R. N. & Wong, W. T. Adaptive Müller cell responses to microglial activation mediate neuroprotection and coordinate inflammation in the retina. J. Neuroinflammation (2011). doi:10.1186/1742-2094-8-173

209

718. Wang, M. & Wong, W. T. Microglia-Müller cell interactions in the retina. Adv. Exp. Med. Biol. 801, 333–338 (2014). 719. Singaravelu, J., Zhao, L., Fariss, R. N., Nork, T. M. & Wong, W. T. Microglia in the primate macula : specializations in microglial distribution and morphology with retinal position and with aging. Brain Struct. Funct. 222, 2759–2771 (2017). 720. Kapoor, K., Bhandare, A. M., Mohammed, S., Farnham, M. M. J. & Pilowsky, P. M. Microglial number is related to the number of tyrosine hydroxylase neurons in SHR and normotensive rats. Auton. Neurosci. Basic Clin. (2016). doi:10.1016/j.autneu.2016.05.005 721. Huh, Y. et al. Microglial activation and tyrosine hydroxylase immunoreactivity in the substantia nigral region following transient focal ischemia in rats. Neurosci. Lett. 349, 63–7 (2003). 722. Marinova-Mutafchieva, L. et al. Relationship between microglial activation and dopaminergic neuronal loss in the substantia nigra: A time course study in a 6- hydroxydopamine model of Parkinson’s disease. J. Neurochem. 110, 966–75 (2009). 723. Tanaka, S. et al. Activation of microglia induces symptoms of Parkinson’s disease in wild-type, but not in IL-1 knockout mice. J. Neuroinflammation (2013). doi:10.1186/1742-2094-10-143 724. Ma, W., Zhao, L., Fontainhas, A. M., Fariss, R. N. & Wong, W. T. Microglia in the mouse retina alter the structure and function of retinal pigmented epithelial cells: A potential cellular interaction relevant to AMD. PLoS One (2009). doi: 10.1371/journal.pone.0007945. 725. Ma, W., Zhao, L. & Wong, W. T. Microglia in the outer retina and their relevance to pathogenesis of age-related macular degeneration. in Advances in Experimental Medicine and Biology (2012). doi:10.1007/978-1-4614-0631-0_6 726. Zabel, M. K. et al. Microglial phagocytosis and activation underlying photoreceptor degeneration is regulated by CX3CL1-CX3CR1 signaling in a mouse model of retinitis pigmentosa. Glia 64, 1479–1491 (2016). 727. Zhao, L. et al. Microglial phagocytosis of living photoreceptors contributes to inherited retinal degeneration. EMBO Mol. Med. 7, 1179–1197 (2015). 728. Chen, Y. et al. Photoreceptor degeneration and retinal inflammation induced by very low-density lipoprotein receptor deficiency. Microvasc. Res. 78, 119–127 (2009). 729. Li, W. Eat-me signals: Keys to molecular phagocyte biology and ‘Appetite’ control. J. Cell. Physiol. 227, 1291–7 (2012).

210

730. Liets, L. C., Eliasieh, K., van der List, D. A. & Chalupa, L. M. Dendrites of rod bipolar cells sprout in normal aging retina. Proc. Natl. Acad. Sci. 103, 12156–12160 (2006). 731. Roufail, E. & Rees, S. Ageing has a differential effect on nitric oxide synthase- containing and catecholaminergic amacrine cells in the human and rat retina. J Comp Neurol 389, 329–47. (1997). 732. Rodriguez, P. C. et al. Fluorescent dopamine tracer resolves individual dopaminergic synapses and their activity in the brain. Proc. Natl. Acad. Sci. 110, 870–5 (2013). 733. Patriarchi, T. et al. Ultrafast neuronal imaging of dopamine dynamics with designed genetically encoded sensors. Science (80-. ). (2018). doi:10.1126/science.aat4422 734. Sun, F. et al. A Genetically Encoded Fluorescent Sensor Enables Rapid and Specific Detection of Dopamine in Flies, Fish, and Mice. Cell 174, 481–496 (2018). 735. Choi, H. J., Ribelayga, C. P. & Mangel, S. C. Cut-loading: A Useful Tool for Examining the Extent of Gap Junction Tracer Coupling Between Retinal Neurons. J. Vis. Exp. (2012). doi:10.3791/3180 736. M. M. Deshpande, W. D. E. Dopamine and Light Adaptation Modulate Nitric Oxide Production in Mouse Retina. Invest Ophtalmol , 49 5194(2008). 737. Carr, B. J. & Stell, W. K. Nitric Oxide (NO) Mediates the Inhibition of Form- Deprivation Myopia by Atropine in Chicks. Sci. Rep. (2016). doi:10.1038/s41598-016- 0002-7 738. Wang, G.-Y., van der List, D. A., Nemargut, J. P., Coombs, J. L. & Chalupa, L. M. The sensitivity of light-evoked responses of retinal ganglion cells is decreased in nitric oxide synthase gene knockout mice. J. Vis. 7, 7–13 (2007). 739. Nemargut, J. P. & Wang, G. Y. Inhibition of nitric oxide synthase desensitizes retinal ganglion cells to light by diminishing their excitatory synaptic currents under light adaptation. Vision Res. 49, 2936–47 (2009). 740. Sodhi, P. & Hartwick, A. T. Adenosine modulates light responses of rat retinal ganglion cell photoreceptors througha cAMP-mediated pathway. J Physiol 592, 4201–4220 (2014). 741. Ribelayga, C. & Mangel, S. C. Light/dark and circadian clock differentially control the extracellular level of adenosine in the mammalian retina. Invest. Ophthalmol. Vis. Sci. 45, 4644 (2004). 742. Jonsson, G. & Eysteinsson, T. Retinal A2A and A3 adenosine receptors modulate the components of the rat electroretinogram. Vis. Neurosci. (2017). doi:10.1017/S0952523816000171

211

743. Li, H., Chuang, A. Z. & O’brien, J. Regulation of photoreceptor gap junction phosphorylation by adenosine in zebrafish retina. Vis. Neurosci. 31, 237–43 (2014). 744. Sodhi, P. & Hartwick, A. T. E. Adenosine modulates light responses of rat retinal ganglion cell photoreceptors througha cAMP-mediated pathway. J. Physiol. 592, 4201– 4220 (2014). 745. Bryan, N. S. & Grisham, M. B. Methods to detect nitric oxide and its metabolites in biological samples. Free Radic. Biol. Med. 43, 645–57 (2007). 746. Van Dycke, A. et al. Quantitative analysis of adenosine using liquid chromatography/atmospheric pressure chemical ionization-tandem mass spectrometry (LC/APCI-MS/MS). J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 878, 1493–8 (2010). 747. Ouchi, Y. et al. Microglial activation and dopamine terminal loss in early Parkinson’s disease. Ann. Neurol. 57, 168–75 (2005). 748. Qian, L. & Flood, P. M. Microglial cells and Parkinson’s disease. Immunol. Res. 41, 155–64 (2008). 749. Moehle, M. S. & West, A. B. M1 and M2 immune activation in Parkinson’s Disease: Foe and ally? Neuroscience 302, 59–73 (2015). 750. Subramaniam, S. R. & Federoff, H. J. Targeting microglial activation states as a Therapeutic Avenue in Parkinson’s disease. Front. Aging Neurosci. (2017). doi:10.3389/fnagi.2017.00176

212