Investigating damage, physical damage, or imitated via








drug-induced proliferation of Müller Glia proliferation in the retina








Calvin Yang

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now











BSCI 3961: 3 hours


Department of Biological Sciences

Vanderbilt University

December 7th, 2017


Research Advisor: Dr. James Patton










            This project focuses on retinal regeneration
in zebrafish via Müller glia. A lack of GABA signaling towards Müller glia from
horizontal cells implies photoreceptor cell death, resulting in dedifferentiation,
proliferation, and re-differentiation of Müller glia into photoreceptors. A
proliferative effect in the retinal can be achieved through light damage,
physical damage, or imitated via interference of GABA and the GABAA
receptor signaling by using certain drugs. Using drugs to imitate a
proliferative effect elucidates these signaling pathways and are helpful
towards experiments of utilizing exosomes in inducing retinal regeneration. We
use gabazine and picrotoxin, GABAA receptor antagonists to compare
their effectiveness at generating PCNA+ cells, which mark proliferation. The
results show that picrotoxin is effective at inducing proliferation and more
effective than gabazine. It also seems that cumulatively higher proliferation
levels occur in picrotoxin in combination with incidental physical damage from
performing eye injections.



















degenerative diseases such as retinitis pigmentosa (RP), diabetic retinopathy,
and macular degeneration (AMD) induce irreversible damage, poor vision, and
even blindness. These diseases arise genetically, environmentally, and from
other diseases that cause the death of photoreceptors, cells that detect light
and initiate the electrochemical signal for the visual system. In mammals,
death of photoreceptors is not naturally followed by any regenerative
processes. However, clinical research is ongoing in developing pluripotent stem
cell therapies to counter the effects of retinal degeneration, which would
require a mediated pathway that ensures proper differentiation and development
of stem cells into the diverse and complicated network of the retina.

zebrafish (Danio rerio) is a popular
model organism with remarkable regenerative abilities in tissues of the brain,
spinal cord, heart, fins, and eyes including the retina and the optic nerve. In
a laboratory setting, photoreceptor cell death can be induced by subjecting
dark-adjusted fish to prolonged and intense light exposure, permitting research
opportunities as photoreceptors naturally regenerate and fish regain their

A key cell member in
retinal regeneration is the Müller glia, which like all glial cells function as
support for neurons. When the retina is damaged, Müller glia dedifferentiate,
divide asymmetrically, and produce progenitor cells that are capable of
restoring all lost cell types (Bernardos et al., 2007; Fausett and Goldman,
2006; Nagashima et al., 2013; Rajaram et al., 2014; Ramachandran et al., 2012;
Thummel et al., 2008; Vihtelic and Hyde, 2000; Wan et al., 2012; Zhao et al.,
2014). Since retinas between fish and mammals are structurally and functionally
comparable, understanding how zebrafish induce retina regeneration will aid in
developing therapies for retina damage or diseases, in particular those that
target regeneration from mammalian Müller glia.


Figure 1: Progressive photoreceptor
regeneration from Müller glia after
light-induced lesion (Gramage et al, 2014)


Previous course work looked into ascl1a and its role
in Müller glia dedifferentiation by knockdown of its expression with a
morpholino to confirm its involvement in the GABAergic pathways. Glutamate from
photoreceptors target GABAergic horizontal cells. When photoreceptors die they
no longer stimulate horizontal cells to release GABA (Figure 2a). It was
hypothesized that Müller glia detect the decrease in GABA and initiate regeneration.
The choice drug to simulate the damaged retina’s decrease in GABA release was
gabazine, an allosteric inhibitor of channel opening of the GABAA
receptor (Ueno et al., 1997) (Figure 2b). The GABAA receptor is a
ligand-gated ion channel that allows the for the selective passage of Cl- ions.
Results show that interfering with GABA signaling cause spontaneous
proliferation in healthy zebrafish retinas and that higher levels GABA, even in
damaged retinas, suppresses proliferation.


Figure 2a: (Above left) Hypothetical model of GABA’s role in inhibiting
dedifferentiation of Müller glia, which fails in a damaged retina. Figure 2b:
(Above right): Experimental model of mimicking damage with Gabazine


            Achaete-scute homolog 1 or ASCL1, is a gene coding for a
basic helix-loop-helix transcription factor that is candidate for neural
development (Ball et al., 1993). Its zebrafish ortholog, ascl1a, is key in
lin-28-dependent let-7 miRNA
signaling pathway that results in Müller glia dedifferentiation (Ramachandran
et al., 2010). Ascl1a is needed for expression of the the lin-28 pluripotency
factor. Lin-28 inhibits the let-7
miRNA, which would otherwise negative feedback against ascl1a and lin-28 as
well as suppress genes associated with Müller glia dedifferentiation.

In prior experiments,
several conditions of Gabazine +/- and various Morpholino +/- PBS solutions
were injected into the eye of adult zebrafish. Scoring of PCNA positive cells,
which represented Müller glia, showed that the injection of certain ascl1a
morpholinos successfully counteracted the proliferative effected that Gabazine
induced, and implying that ascl1a knockdown could be involved in the GABAergic

A major project in the
Patton Lab is to induce Müller glia proliferation and development into other
cell types, particularly with the injection of cancer cell line derive
exosomes. Exosomes are a particular class of extracellular vesicles. They are
distinguished from microvesicles in that exosomes form in multivesicular
bodies, and are released extracellularly by exocytosis, whereas microvesicles
form directly from the plasma membrane. Both of these extracellular vesicles
may either fuse directly with the plasma membrane of a recipient cell, or are
internalized via endocytosis. Exosomes contain various proteins
and RNAs, mRNA and miRNA included, depending on the origin cell type. There is
much potential to utilize exosomes in medical therapy or as biomarkers.

It has been shown that KRAS
mutant colorectal cancer cell lines secrete exosomes containing miRNAs such as let-7a and miR-100 (McKenzie et al.,
2016). These miRNAs have been found specifically in Müller glia and are
involved in proliferative pathways. Therefore, eye injections of these exosomes
may be effective inducing a desired proliferative response.

A major problem with the
prior protocol of using PCNA labeling is that the labeling would only be
effective during DNA replication, therefore potentially only labeling a select
group of Müller glia. Furthermore, another issue with exosome injections, and
in general, is that we cannot distinguish which cells that are PCNA positive
are ones that have uptaken the exosomes. There is constantly a baseline count
of cells from innately occurring proliferation. And through the process of
performing eye injections, potential damage can induce more proliferation as
well, therefore convoluting the apparent effectiveness of exosome injections.

We have decided to
utilize Cre-Lox recombination as a tool to help us visual and screen out
proliferating Müller glia specifically from absorbing exosomes. Cre recombinase
performs a deletion event around loxP
sites on the same stand of DNA running in the same direction. In the process of
recombination, a contiguous plasmid with one loxP site and the stretch of DNA in between the two sites is
separated from the other loxP site
and the rest of the genome.

Our plan is to utilize a
transgenic fish line TgBeta-actin:loxp-mCherry-stop-loxp-GFP (

Ramachandran, et. al, 2010) (Figure 4) with
Cre expressing C6 rat glioma cell line as a primary vector for exosomes. A
recombination event would alter the fluorescence of the target cell from
mCherry to GFP, and this expressional effect is passed down to any progenitor
cell types. With this transgenic combination, we can identify proliferating Müller
glia specifically from being exosome recipients, as well as visualize the long
term destination and fates of these progenitor cells in the retina.

Figure 4: Layout
of transgenic zebrafish line recombination site (Ramachandran, et. al, 2010)


From exosomes, morpholinos, and drugs,
there are many biomolecules that can result in proliferating, Müller
glia derived cells. A new drug of interest for inducing regeneration is
picrotoxin, which is comparable to gabazine in that mimics “damage”, noncompetitively
inhibiting the opening of the GABAA receptor (Rho, et. Al, 1996). Therefore,
in this experiment we compare gabazine and picrotoxin in seeing which drug is quantitatively
better at inducing cell proliferation in the retina.






Figure 5: Outline of ultimate protocol of
exosome eye injection

the Cre positive exosomes, the progenitor cells cannot be traced and
visualization of Müller glia is only temporary in PCNA positive cells. Cre
injection into the transgenic line allows for long term GFP fluorescence of Müller
glia and progenitor cells.


and Methods

lines and Maintenance

The zebrafish line used
in this study was 6-month-old adult TgBeta-actin:loxp-mCherry-stop-loxp-GFP.
All fish were maintained in a 14:10 light:dark (L:D) cycle at 28°C.

Drug injections

            3 conditions were defined: PBS, gabazine(12.5nmol),
and picrotoxin (0.1nmol). Concentrations were determined from previous
literature. Briefly, zebrafish were anesthetized in 0.016% tricaine, an
incision was made in the sclera with a sapphire knife, and a blunt end 30 gauge
needle inserted. 0.5?L were injected into one eye of adult zebrafish.


After 72 hours, zebrafish were euthanized
in 0.08% tricaine and whole eyes were removed and fixed 4% paraformaldehyde
(all other staining) overnight. Eyes were then washed in PBS and cryoprotected
in 30% sucrose for 4 hours at room temperature. Eyes were then transferred to a
solution containing 2 parts OCT and 1 part 30% sucrose overnight followed by
transfer to 100% OCT for 2 hours and then embedded in OCT for cryosectioning.

Slides were warmed on slide warmer for 15
min and hydrated in 1XPBS in coplin jar for 30 min. Slides were incubated in
10mM sodium citrate, %0.05 Tween, pH 6.0 for 20 minutes at 95°C, and left to
cool in solution at room temperature for 20 minutes. Slides were washed with
500ul 1XPBS, blocked with blocking buffer (3% Donkey serum, 0.1% TritonX-100 in
1XPBS) overnight.

            Primary antibodies were added in ab
buffer (1%DS, 0.05% Tween in 1X PBS) overnight at 4°C. Antibodies used were
PCNA (Sigma, P8825; Abcam, ab2426) and Glutamine Synthetase (Millipore, mab302)
at 1:500 volume ratio to ab buffer. Slides were washed with 0.1%Tween in 1xPBS
(3 times 10 minutes each).

Cy3 secondary was used on PCNA primary;
AF488 secondary, Glutamine Synthase primary, both in 1:500 volume ratio to ab
buffer. TO-PRO®-3 Iodide was the choice of nuclear stain at a 1:2500 volume
ratio. Slides were left overnight at 40°C.

Slides were washed with 0.1%Tween in 1xPBS
(3 times 10 min each) and once with 1X PBS for 5min. Vectashield was added to
labeled slides.



a)                                                b)                                                 c)


e)                                                 f)


g)                                                  h)



Figure 6: Retinal
sections showing proliferating cells

PBS (a, b, c). Gabazine (e, f). Picrotoxin (g, h). Glutamine Synthase PCNA


            Bright red spots are PCNA positive proliferating
cells (Fig. 6) in a background of glutamine synthase. Some retinal sections and
regions are relative scarce in proliferating cells (Figure 6a). Others have
regions of extremely high densities and clusters of proliferating cells (Figure





Figure 7: One-way
ANOVA plots of PCNA+ cells

a) Counts of individual cells per section. b) Counts
of cells clusters per section.


4 sections per eye were scored for PCNA+
cells (Fig. 7). Individual sections were scored for single cells as well as
cell clusters. Any in contact group of PCNA+ cells made of at least 3 members qualifies
a cluster. Overall, there was a dichotomy in counts for both individual cells
and clusters. Some sections have sprawled and scarce PCNA+ with several or no
clusters. Other sections have one region of extremely dense PCNA+ cells,
meaning high cell counts and cluster counts. PBS and picrotoxin are statistically
significant in individual cell counts (Fig. 7a), and very significant in
cluster counts (Fig. 7b).






Overall, picrotoxin effectively induced
proliferations of PCNA+ cells in the retina. It significantly increases both individual
cell counts and clusters compared to PBS. Furthermore, it is more effective
than gabazine at inducing proliferation.

            However, as mentioned, many
sections, among all three conditions, had regions of high cell proliferation
and clusters, to the point of being practically impossible to count all cells. When
this occurs, there is only one concentration per section. However, adjacent
sections of an eye will have this high concentration area as well. Usually when
this is seen, it indicates that during the eye injection, the injection needle
had hit the back of the eye and damaged the retina. This direct physical damage
would be enough to induce the proliferative effect as mentioned. Some retinas definitely
had proliferation result from the needle hitting the retina, as this can be
identified by indentations into each sections.

            It would make sense to select
against sections damaged by retinas by not counting these highly concentrated
ones. But many sections that have a hotspot region show no indications of
physical damage. This would make it difficult to distinguish what would be usable
data. Some sections had high PCNA+ cell counts but the cells were well distributed
as well, which made it unreasonable at the point of analysis to remove all
replicates with high numbers. Others had counts that were intermediate, which
blurred the difference in dichotomy. The separation of a condition’s replicates
into subsets is clearest in the two drug injections but particularly in picrotoxin
(Fig. 7). The higher groupings for picrotoxin do seem collectively much higher
in value than for PBS. If we assume the worst case scenario, that all these
high “outliers” are the result of injection damage, then it is possible that the
aggregate of both physical damage and picrotoxin results in an even more profound
proliferative effect than just physical damage in a PBS injection.

            Looking forward, it would be ideal
to replicate this experiment keeping in better mind to avoid hitting the retina
during the injection. It could also be worth looking at the combined effects of
both gabazine and picrotoxin. Though the overall developmental effect is
similar, their interactions with the GABAA receptor differ at the molecular
level, and so their inhibitory effects could be in summation more effective
than merely increasing the concentration of individual drugs.

            Another drug of interest is muscimol,
a GABAA receptor agonist, which in theory would produce the opposite
of gabazine and picrotoxin. Muscimol could potentially yield lower PCNA+ cell
counts than a control PBS. It may also be effective against gabazine and
picrotoxin. It would be interesting to the future experiment singular and combinatorial
conditions with muscimol. This experiment was attempted this semester, but we
think due to a change in quality of histobond slides or matrix medium, the
sections are unable to stay on slides through antigen retrieval and staining.







thanks to Dr. Patton for letting me work in his lab for another semester. I’d
also like to thank Dominic Didiano (PhD) for his mentoring, and the rest of the
Patton Lab as well.