University of Pisa

BIOS School PhD School in BIOmolecular Sciences

Course in Molecular Biotechnology, 20th cycle

Effects of photoreceptor degeneration on

the architecture of retinal ganglion cells:

implications for Retinitis Pigmentosa therapy

Dr. Francesca Mazzoni

Tutors:

Dr. Enrica Strettoi Prof. Irma Nardi

CNR Neuroscience Institute, Pisa

Italian National Research Council

To my husband

“… once development is ended, the founts of growth and regeneration of

the axons and dendrites dried up irrevocably.

In adult centres the nerve paths are something fixed, ended, immutable.

Everything may die, nothing may be regenerated.

It is for the science of the future to change, if possible, this harsh decree.

Inspired with high ideals, it must work to impede or moderate gradual

decay of neurons, to overcome the almost invincible rigidity of their

connections, and to re-establish normal nerve paths, when disease has

severed centres that may intimately associated.”

Santiago Ramon y Cajal, 1913-14

Index Riassunto page 1

Abstract “ 3

Preface “ 5

1. Introduction “ 6 1.1. The retina “ 6

1.1.1. The mouse retina “ 7

1.1.2. Mouse Retinal Ganglion Cells “ 10

1.2. The Thy1-GFP transgenic mouse “ 17

1.3. Retinitis Pigmentosa (RP) “ 19

1.3.1. Genetics and molecular mechanisms of RP “ 21

1.3.1.1. RHO mutations “ 22

1.3.1.2. Pdeb mutations “ 24

1.4. Example of RP mutant mice: The rd10 mutant mouse “ 26

1.5. RP experimental therapeutic strategies: transplants and implants 27

1.5.1. Transplants “ 27

1.5.2. Implants “ 30

1.5.2.1. Subretinal prostheses “ 31

1.5.2.2. Epiretinal prostheses “ 32

1.5.3. Other therapeutic strategies to treat RP “ 34

1.5.3.1. Neuroprotective factors “ 34

1.5.3.2. Gene therapy “ 35

1.6. Aim of the project “ 36

1.7. Experimental design “ 38

2. Materials and Methods “ 40 2.1. Animals “ 40

2.1.1. Isolation of a new transgenic mouse “ 40

2.2. Analysis of second order neuron score in rd10 mutant mice “ 41

2.2.1. Immunocytochemistry (ICCH) “ 41

2.2.2. Rod bipolar and horizontal cell survival evaluation “ 42

Index

IV

2.3. Identification of mouse retinal ganglion cells (RGCs) page 43

2.3.1. Immunocytochemistry protocol “ 43

2.3.2. Classification parameters “ 44

2.4. RGC survival in adult mutant mice “ 45

2.4.1. Immunocytochemistry protocol for counting ” 46

2.5. Vascular density (VD) assessment “ 46

2.5.1. Immunocytochemistry for blood vessels “ 46

2.6. Anterograde axonal transport evaluation in adult mutant mice “ 47

2.6.1. Immunocytochemistry protocol for cholera toxin “ 47

3. Results “ 48 3.1. A new strain of transgenic mouse (rd10/Thy1-GFP) “ 48

3.2. Characterization of rd10 retinal phenotype “ 49

3.2.1. Photoreceptors “ 49

3.2.2. Rod bipolar cells “ 51

3.2.2.1. Survival of rod bipolar cell “ 53

3.2.3. Horizontal cells “ 54

3.2.3.1. Survival of horizontal cells “ 55

3.3. RGC morphology ” 56

3.3.1. Sampling ON and OFF RGC types “ 59

3.3.1.1. A types RGCs “ 60

3.3.1.2. Selection of B type RGCs “ 62

3.3.1.3. Selection of C type RGCs “ 65

3.3.2. Other morphological observations ” 67

3.3.3. RGC dendritic tree complexity assessment “ 68

3.4. RGCs survival “ 73

3.5. Vascular Density (VD) ” 73

3.6. RGCs axonal transport ” 74

4. Discussion “ 77 4.1. The rd10 mouse: a good model of human RP “ 77

4.2. Single RGC study in rd10/Thy1-GFP mice “ 79

5. Concluding remarks and future directions “ 88

Index

V

6. Bibliography “ 91 APPENDIX (Supplementary Materials) ” 103

(1-12)

Acknowledgments

1

Riassunto

La Retinite Pigmentosa (RP), una famiglia di malattie ereditarie che porta alla

progressiva morte dei fotorecettori, rappresenta una delle maggiori cause di

cecità nel mondo, senza una cura possibile.

Sebbene la causa primaria della malattia sia, tipicamente, un difetto in un gene

specifico dei bastoncelli, è noto che la degenerazione di queste cellule è seguita

dalla morte secondaria dei coni e da una cascata di eventi che portano ad una

degenerazione secondaria dei neuroni della retina interna, e particolarmente delle

cellule bipolari ed orizzontali.

Scopo di questo progetto è testare l’ipotesi che la perdita progressiva di

fotorecettori causi cambiamenti regressivi anche nelle cellule gangliari (RGCs),

l’unica stazione di uscita dell’ informazione visiva dalla retina verso i centri

encefalici superiori e piattaforma biologica di varie forme di trattamento della RP.

Questo studio dimostra che la morfologia, l’architettura cellulare e il grado di

sopravvivenza delle RGCs di un modello murino di RP sono molto ben conservati

a vari stadi della progressione della patologia. Specificatamente, abbiamo

generato un topo transgenico, il mutante rd10/Thy1-GFP, incrociando il topo GFP-

M, nel quale la proteina fluorescente GFP è espressa in una piccola popolazione

di RGCs di vari tipi, e il topo rd10, un recente modello di RP autosomica

recessiva. Il topo rd10 porta una mutazione missenso della subunità beta del

gene della fosfodiesterasi dei bastoncelli, che causa una tipica degenerazione dei

bastoncelli e dei coni con un picco a P24. L’espressione di GFP in un piccolo

numero di RGCs nella retina del mutante ha permesso un dettagliato studio della

fine struttura di questi neuroni a varie età.

Combinando tecniche di immunoistochimica, microscopia confocale e

morfometria analitica, abbiamo studiato le RGCs in un totale di 50 retine

appartenenti a 3 gruppi di età (3, 7 e 9 mesi), così da coprire la completa

degenerazione dei fotorecettori. Un numero di 572 RGCs è stato identificato e

raggruppato secondo la classificazione di Sun et al. (2002a). In particolare, le

RGCs sono state identificate in base a 4 parametri: il diametro dell’albero

dendritico, il diametro del corpo cellulare, la profondità media di stratificazione dei

dendriti nello strato plessiforme interno e la forma tipica della arborizzazione

Riassunto

2

dendritica. Per ogni punto temporale, 5 RGCs dello stesso tipo sono state

disegnate in tre dimensioni con un software per l’analisi di immagine

(NeuroLucida). Le tracce neuronali sono state valutate matematicamente per

ottenere alcuni parametri altamente indicativi della complessità dell’albero

dendritico e della sua fine architettura: la lunghezza dendritica totale, il numero

totale di nodi e l’area dell’albero dendritico. Otto differenti tipi di RGCs sono stati

analizzati e disegnati (per un totale di 164 cellule), compresi i tipi A (le cellule più

ampie), B (le più piccole) e C (intermedie), entrambe sia ON che OFF.

Abbiamo riscontrato una notevole conservazione della complessità strutturale di

tutti i tipi di cellule gangliari studiate fino a 9 mesi di età, periodo in cui le

alterazioni a carico dei neuroni di secondo ordine sono molto evidenti. Inoltre, a 9

mesi di età, la sopravvivenza delle cellule presenti nello strato delle gangliari

appare del 100%. Ancora a 9 mesi, iniezioni intraoculari di tossina del colera

hanno dimostrato la presenza di trasporto assonale anterogrado delle cellule

gangliari al nucleo genicolato laterale e al collicolo superiore. Ancora, alla stessa

età, abbiamo trovato una diminuzione del 30% nella densità dei vasi sanguigni

della retina che irrorano le cellule gangliari.

La notevole conservazione della fine struttura delle cellule gangliari nel mutante

rd10, la loro eccellente sopravvivenza, la mantenuta capacità di trasporto

assonale anterogrado, nonostante l’impoverimento dei vasi sanguigni e

l’alterazione e degenerazione dei neuroni di secondo ordine, aprono prospettive

terapeutiche per la RP che siano basate proprio sulla sopravvivenza delle cellule

gangliari. In particolare, la strategia di impiantare protesi epiretiniche che stimolino

direttamente questi neuroni per ripristinare la visione nei pazienti con RP potrebbe

essere applicata con successo in quei casi in cui le cellule gangliari risultassero

ancora ottimamente preservate, come nel caso della mutazione rd10.

3

Abstract Retinitis Pigmentosa (RP), a family of inherited diseases leading to progressive

photoreceptor death, is one of the major causes of blindness in the world, with no

cure yet.

Albeit the primary cause of the disease is, typically, a defect in a photoreceptor-

specific gene, it has been shown that the degeneration of rods and cones triggers

remodeling and secondary death of inner retinal neurons, and particularly of

bipolar and horizontal cells. Aim of this project is to test the hypothesis that, as an

effect of photoreceptor progressive loss, concomitant changes also occur in

retinal ganglion cells (RGCs), the last neurons of the retinal visual pathway and

the only exit of retinal information to higher brain centers.

We assessed the retention of morphology, overall architecture and survival rate of

RGCs in a mouse model of RP, at various stages of the disease progression.

Specifically, we generated a transgenic mouse, the rd10/Thy1-GFP mutant, by

crossing GFP-M mice, in which GFP is expressed in a small population of RGCs

of various types, and rd10 mice, a recent model of autosomal recessive RP. Rd10

mice carry a missense mutation of the beta-subunit of the rod-specific

phosphodiesterase gene, causing typical rod-cone degeneration with a peak at

postnatal day 24 (P24). The expression of GFP in a small number of RGCs in the

retina of the mutant allowed the detailed study of the fine structure of these

neurons at various ages.

By combining immunocytochemistry, confocal microscopy and analytic

morphometry, we studied RGCs in a total of 50 whole mounted retinas of 3 age

groups (3, 7 and 9 months of age, thus past the complete degeneration of

photoreceptors). A number of 572 RGCs were identified and grouped according to

the classification of Sun et al. (2002a). In particular, 4 parameters were taken into

account to identify RGCs: the diameter of the dendritic tree, the diameter of the

body, the mean stratification depth within the inner plexiform layer and the typical

shape of the dendritic arborization. Five RGCs of the same type, at each time

point, were drawn three dimensionally with a computer assisted image analyzer.

The neuronal tracings were mathematically evaluated to obtain some parameters

highly indicative of dendritic tree complexity and fine architecture: the total

Abstract

4

dendritic length, the total number of nodes and the dendritic tree area. Eight

different types of RGCs were analyzed and drawn (for a total of 164 cells),

including type A (the largest), B (the smallest) and C (medium sized), both ON

and OFF.

We found a remarkable preservation of the structural complexity of all the types of

RGCs studied, up to 9 months of age, even in the occurrence of major remodeling

among second order neurons. In addition, at 9 months of age, the survival of cells

in the GCL appeared comparable to the wt counterpart. Finally, still at 9 months,

injections of cholera toxin in the eyes demonstrated the presence of anterograde

axonal transport of RGCs to the lateral geniculate nucleus and to the superior

colliculus. Yet, at the same age, we detected a decrease of 30% in the density of

retinal blood vessels providing nourishment to ganglion cells.

The remarkable, long term preservation of RGC fine structure, survival and

capability of anterograde axonal transport in the retina of rd10 mutant mice,

despite blood vessel impoverishment and second order neurons remodeling and

degeneration, opens perspectives to therapy for RP based on ganglion cells

preservation. In particular, the strategy of implanting epiretinal prostheses, directly

stimulating ganglion cells, to restore vision in RP patients could be applied

successfully in those cases in which ganglion cells are still viable, such in the

case of the rd10 mutation.

5

Preface This work focuses on the retention of the morphology and architecture of retinal

ganglion cells (RGCs) in the rd10 mutant mouse, a recently characterized model

of human Retinitis Pigmentosa (RP). This is a family of inherited diseases leading

to progressive death of retinal photoreceptors and representing one of the major

causes of blindness in the world, with no cure yet. Perspective therapeutic

strategies under study, including cellular transplantation and implant of electronic

prostheses, rely upon the assumption that the inner layers of retina (located

“beyond” photoreceptors) do not undergo major adverse effects during the

progression of the pathology and retain their structure and function. Recent data

from this and other laboratories show that this is not the case: second order

neurons (bipolar and horizontal cells) undergo regressive remodeling and

eventually degenerate concomitantly and after photoreceptor death. RGCs are the

last station of the light information transfer from the eye to the brain. In view of the

adverse remodelling affecting bipolar and horizontal cells, it is legitimate to ask

whether RGCs undergo similar degenerative phenomena accompanying

photoreceptor death, especially when considering that RGCs represent the

biological substrate for several strategies to restore vision in RP, for instance by

means of epiretinal prostheses. Thus, this thesis investigates the effects of

photoreceptor degeneration upon the fine structure and overall architecture of

RGCs taking advantage of a mouse model of RP.

In the introductory chapters, the overall structure of the retina will be described

first (#1.1 and 1.1.1), particularly focusing on RGC morphology (#1.1.2), studied

by means of transgenic mouse technology (the Thy1-GFP mouse; #1.2).

Subsequently, the main characteristics of RP (#1.3 and #1.4) and its mouse

models will be described, later illustrating promising experimental strategies (#1.5,

1.5.2 and 1.5.3) and mostly epiretinal prostheses (#1.5.2.2).