Stimulation of neural regeneration in themouse retinaMike O. Karla, Susan Hayesa, Branden R. Nelsona, Kristine Tana, Brian Buckinghamb, and Thomas A. Reha,1

aDepartment of Biological Structure, and bMedical Science Training Program, 357420 Health Science Center, University of Washington, School of Medicine,Seattle, WA 98195

Edited by Fred H. Gage, The Salk Institute for Biological Studies, San Diego, CA, and approved October 20, 2008 (received for review August 1, 2008)

Muller glia can serve as a source of new neurons after retinaldamage in both fish and birds. Investigations of regeneration inthe mammalian retina in vitro have provided some evidence thatMuller glia can proliferate after retinal damage and generate newrods; however, the evidence that this occurs in vivo is not conclu-sive. We have investigated whether Muller glia have the potentialto generate neurons in the mouse retina in vivo by eliminatingganglion and amacrine cells with intraocular NMDA injections andstimulating Muller glial to re-enter the mitotic cycle by treatmentwith specific growth factors. The proliferating Muller glia dedif-ferentiate and a subset of these cells differentiated into amacrinecells, as defined by the expression of amacrine cell-specific markersCalretinin, NeuN, Prox1, and GAD67-GFP. These results show forthe first time that the mammalian retina has the potential toregenerate inner retinal neurons in vivo.

amacrine � Gabaergic neuron � glia

I t is well established that the retina of cold-blooded vertebratesregenerates very well after damage (1–3). The avian retina is

also capable of limited regeneration of new neurons followingneurotoxic damage (4). In both systems, damage to retinalneurons causes Muller glial cells to re-enter the cell cycle, afterwhich they dedifferentiate into retinal progenitors, and ulti-mately differentiate into neurons. In fish, all types of neurons areregenerated. However, in chicks, only a limited number ofdifferent types of inner retinal neurons (amacrine, bipolar, andganglion cells) are produced; few, if any, photoreceptors areregenerated.

In the mammalian retina, by contrast, the proliferative re-sponse of Muller glia to injury is even more limited than inchicks. In response to injury in mouse or rat retina, the Mullerglia may become reactive and hypertrophy, but few re-enter themitotic cell cycle. Due to the lack of a spontaneous regenerativeresponse in the mammalian retina, several groups have at-tempted to stimulate regeneration with intraocular injections ofgrowth factors and/or transcription factors, in vitro or in vivo(5–10). Taken together, the studies of the mammalian retinaindicate that Muller glia have a very limited proliferative re-sponse to injury, but can be stimulated to re-enter the cell cycleafter photoreceptor or inner retinal neuron injury. These studiesalso reported that some of the progeny of Muller glial mitoticdivisions go on to differentiate characteristics of rod photore-ceptors. However, these studies failed to detect regeneratinginner retinal neurons after damage, unless they were transfectedwith genes that specifically promote amacrine fate (9). This ispuzzling, since in the chick, amacrine cells are the primaryneuronal cells that are regenerated after injury. In an attempt toresolve this issue, we have carried out a systematic analysis of theresponse to injury in the mouse retina, and the effects of growthfactor stimulation on Muller glial proliferation. The previousstudies have used rats because of the relative ease of intraocularinjections, but mice have the advantage that it is possible to usea variety of genetically manipulated animals to verify neuralregeneration. We find that as in the rat, very few Muller gliare-enter the cell cycle after N-methyl-D-aspartic acid (NMDA)

mediated damage to inner retinal neurons; however, the prolif-eration can be greatly stimulated by intraocular injection ofeither EGF, FGF1, or the combination of FGF1 and insulin.Although most progeny of the Muller glia maintain their ex-pression of Muller glial markers, a small number of bromode-oxyuridine (BrdU) labeled cells differentiate into amacrine cells,as indicated by their expression of NeuN, Calretinin, Pax6, andGAD67-GFP. Thus, our results show for the first time that themammalian retina has the potential to regenerate inner retinalneurons in vivo.

ResultsNeurotoxic Damage and Stimulation of Proliferation. To test whetherMuller glia can regenerate inner retinal neurons in the mouseretina, we elicited retinal damage with NMDA. Overstimulationof NMDA-activated ionotropic glutamate receptors causes celldeath of retinal ganglion cells (RGC) and amacrine cells (11–13).We confirmed in adult mouse retina that NMDA induceddamage leads to loss of Brn3-positive RGCs and GAD67-GFPGABAergic neurons, but not bipolar cells. (supporting infor-mation (SI) Fig. S1 A and B).

Previous studies have found only limited Muller glia prolifer-ation after various forms of retinal injury in the mammalianretina; however, a recent report shows that retinal damage causesincreased EGF receptor expression in Muller glia (5). Wetherefore reasoned that NMDA induced damage to the retinamight elicit a similar increase in the sensitivity of Muller glia toEGF. Previous studies in chick retina demonstrated that thecombination of FGF and insulin also stimulated proliferation ofMuller glia. Therefore, we made intraocular injections of eitherEGF or the combination of FGF and insulin in adult mice thathad received an injection of NMDA two days previously (Fig.1A). At this time, the mice also received BrdU to identify anycells that re-entered the cell cycle. We found that there were fewBrdU� cells 24 h following an injection of EGF in undamagedretina (Fig. 1B; 24 h after EGF), but there was a large, dosedependent increase in proliferation in animals that receivedEGF after prior NMDA-induced damage (Fig. 1 C and D).

We analyzed the distribution of the proliferating cells(BrdU�) in NMDA/EGF-treated retinas in flatmount prepara-tions at various time points after treatment. Systematic quanti-tative analysis of cell proliferation in mouse retina over time wasconducted by counting BrdU� nuclei on confocal images takenacross the entire depth of flatmounted retina (Fig. S1C). Injec-tion of either NMDA or EGF alone did not induce substantialproliferation in the retina (Fig. 1 B, E, and G). By contrast,

Author contributions: M.O.K., S.H., B.R.N., and T.A.R. designed research; M.O.K., S.H.,B.R.N., K.T., B.B., and T.A.R. performed research; M.O.K., S.H., B.R.N., K.T., B.B., and T.A.R.analyzed data; and M.O.K. and T.A.R. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

1To whom correspondence should be addresses. E-mail: tomreh@u.washington.edu.

This article contains supporting information online at www.pnas.org/cgi/content/full/0807453105/DCSupplemental.

© 2008 by The National Academy of Sciences of the USA

www.pnas.org�cgi�doi�10.1073�pnas.0807453105 PNAS � December 9, 2008 � vol. 105 � no. 49 � 19507–19512

NEU

ROSC

IEN

CE

Dow

nloa

ded

by g

uest

on

Janu

ary

13, 2

020

NMDA at day zero (d0) followed by a single injection of EGF48 h later (d2) resulted in a striking increase in proliferationacross the retina (day 3: Fig. 1 F–H). Even as early as 4 h after

the EGF injection there were up to 1728 BrdU� cells/mm2

(1168 � 203 BrdU� cells/mm2, n � 4, Fig. 1H). The number ofBrdU� cells declined over the next week (Fig. 1H, S2 B–D). By24 h (d3) after the EGF injection, 57% of the BrdU� cellsremained and at d 7 there was a further decline to 169 � 17BrdU� cells/mm2. However, these remaining BrdU� cells per-sist for at least 30 days (see below).

Several previous reports have found that in fish, birds, andmammals, Muller glia make up the majority of cells that re-enterthe mitotic cycle after retinal damage, although vascular cells,astrocytes and microglia also proliferate after damage (14–16).To confirm that Muller glia were among the cells that take upBrdU� after NMDA treatment in our experiments, we carriedout double-label immunofluorescence for BrdU and Sox2 orSox9, both nuclear markers for Muller glia (17–19). The Mullercell specific expression was confirmed by co-labeling sectionswith one of two well-characterized Muller glia markers: cellularretinaldehyde binding protein (CRALBP; Fig. 2 A and B) or aglial high affinity glutamate transporter, called solute carrierfamily 1 member 3 (GLAST; data not shown). Moreover, GFAP(glial fibrillary acidic protein) is up-regulated in many Muller gliaafter damage and therefore these cells can be labeled in micewith Cre-recombinase driven by GFAP crossed to ROSA-EYFPreporter mice (� GFAP-Cre::EYFP) (Fig. 2 C and D). Dam-aging adult mouse retina with NMDA and subsequently stimu-lating proliferation with various growth factors showed that asearly as 4 h after the mitogen, many of the BrdU� cells were alsoSox2� or Sox9�, residing in the presumptive Muller glia sub-layer of the INL (Fig. 2 E and G). Many of the GFAP-Cre::EYFPpositive Muller cells are PCNA positive after damage and growthfactor injections (Fig. 2F). We also observed GFAP-Cre::EYFPcells co-labeled for other mitotic markers: BrdU, Ki67, and PH3(Fig. 2 H–J).

To determine if mouse Muller glia dedifferentiate and toassess whether genes relevant to neurogenesis were re-expressedin these cells when they are stimulated to proliferate withinjections of EGF after NMDA damage, we carried out severaltypes of studies. Using immunohistochemistry we found thatMuller glia acquire characteristics of retinal progenitors afterNMDA/growth factor treatment, shown by their expression ofthe transcription factor Pax6. Pax6 is normally expressed byretinal progenitors in developing retina and in inner retinalneurons (amacrine/RGCs). Pax6 re-expression is a well-documented step in the dedifferentiation of Muller glia afterdamage of the fish and chick retina in vivo (20–22) as well asadult rat retina in vitro (10). Following NMDA induced retinaldamage and intraocular injections of either EGF or the combi-nation of FGF1 and insulin, the majority of GFAP-Cre::EYFPpositive Muller glia are also Pax6 positive (Fig. 2K).

We have recently reported in the chick retina that many genesnormally expressed only in developing retina are re-expressedfollowing NMDA induced damage (23). To assess whether thesame genes are up-regulated in mouse retina, we used a similarapproach. We harvested mouse retinas 4 h after EGF injection(d2), 24 h (d3), 48 h (d4) and seven days later (d9) (Fig. S3A).The left eyes were injected, while the right eyes served asuninjected controls; we used semiquantitative real-time PCR tomeasure changes in gene expression. Over the time coursestudied, Notch1 and one of its ligands, Delta1, were bothsignificantly increased, as well as other progenitor specific genes,FoxN4 and Pea3. Two additional progenitor-specific genes, Ascl1and Neurogenin2 were not up-regulated to levels detectable byRT-PCR (Fig. S3 B and C); however, the RT-PCR changes weobserve are likely to miss small changes in expression or evenlarge changes if they are in a small number of cells. Nevertheless,these experiments show that at least a part of the developmentalprogram relevant to neurogenesis is re-expressed in the mouse

A

B C D

E F

G

H

Fig. 1. Muller glia proliferation in adult mouse retina. (A) Treatmentscheme: NMDA induced damage at day 0 (D0) with subsequent EGF injectionat day 2 (D2). BrdU was applied at D2 to identify cell proliferation. (B–F) NMDAdamage followed by EGF treatment induces significant proliferation (C, D, F),while EGF (B) or NMDA neurotoxic damage (E) alone do not (E-F, BrdU � blackpseudosquares). RGCs are shown in retinal flatmounts of adult mice labeledfor neurofilament M (NFM, red) without (B) and after cell loss due to NMDAneurotoxic damage (C and D). (E) NMDA alone produces some proliferatingcells around the optic nerve head and occasional BrdU� cells appear at thevery periphery. NMDA damage followed by EGF (F) results in significantproliferation and correlates with the amount NMDA given (C and D). (G) Thereare only slightly more BrdU labeled cells in the central retina compared withthe peripheral retina after NMDA damage and EGF treatment (see also Fig.S2A). (H) The number of BrdU� cells in NMDA-damaged retinas decreaseswith time after treatment (see also Fig. S2 B–D). Images show single confocalplanes of 2 �m (B-D red channel) and z axis projections of 70 � 1 �m (B-D greenchannel). (Scale bars: 10 �m (B–D)).

19508 � www.pnas.org�cgi�doi�10.1073�pnas.0807453105 Karl et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

13, 2

020

retina following NMDA damage and stimulation of Muller glialproliferation.

Amacrine Cell Differentiation of Muller Glial Progeny. Pax6 is ex-pressed in both progenitors and postmitotic amacrine cells andganglion cells. We observed Pax6 expression in BrdU� cells up to12 days after injection of various factors (Fig. 3A), suggesting thatsome of the BrdU� cells differentiate into amacrine cells. Toinvestigate whether BrdU� cells were differentiating into amacrineor ganglion cells, we immunolabeled flatmounts or sections fromNMDA-damaged and growth factor-treated retinas with antibodiesagainst BrdU and one of two markers of amacrine and ganglioncells: NeuN or Calretinin. Fig. 3 shows examples of cells expressingboth BrdU and each of these markers. The double-labeled cellswere found in either the ganglion cell layer (GCL) or the innernuclear layer (INL) and had morphologies similar to the otherneurons in these layers. We found BrdU�/NeuN� cells five daysfollowing the growth factor injection in both the EGF and FGF1/insulin treatment groups (Fig. 3B). In either of those two treatmentsbetween one and three double-labeled NeuN�/BrdU� cells were

present in z-stacks across a 0.05 mm2 retinal area (Table 1). Usinga second marker for amacrine cells and RGC, Calretinin, we alsofound double-labeled cells five days after EGF or FGF1/insulintreatment. These experiments show that some of the cells thatincorporated BrdU after retinal damage and subsequent growthfactor treatment developed characteristics of inner retinal neurons.

To further confirm that new neurons are produced in theNMDA-treated retina, we used a GAD67-GFP transgenic re-porter mouse, which labels GABAergic cells in the retina. As inprevious experiments, we induced inner retinal neuronal damagewith NMDA; however, instead of giving single injections ofgrowth factors as above, we made multiple injections of FGF1/insulin to increase their potential for neural differentiation. Forthese experiments, we co-injected FGF1 and insulin once daily,for four consecutive days (4xFGF1�insulin) after NMDA dam-age, and analyzed the retinas at day (d) d 4, d 6, d 8, d 14 andd 30. Using confocal imaging and 3D visualization (Fig. S1), wefound clear examples of BrdU�/GFP� cells (6 � 0.3 cells permm2 overall mean across all retinas analyzed, d 8–30, n � 6) withmost cells in the in the INL (Fig. 3 D-J). We also observed cells

A B C D D’ D’’

E’

E E’’

F K

G H I J K’ K’’ K’’’

Fig. 2. Muller glia proliferation is inducible after retinal damage. Sox2 (A) and Sox9 (B) label Muller glia, as shown by colabeling with CRALBP. After NMDAdamage, GFAP-Cre::EYFP labels Muller glia, as shown by double-labeling with Sox9 (C) and CRALBP (D, D�,D�). (E) Confocal image of retina flatmounts showBrdU�/Sox2� double-labeled cells at depths corresponding to Muller glia as early as 4 h after induction of proliferation by various factors following NMDAdamage and EGF. The xz- (E�) and yz-views (E�) represent single optical sections through the indicated area (crosshair). (F) After NMDA damage followed by 4consecutive injections of FGF1/insulin, Muller glia up-regulate PCNA and treatment (F), BrdU (G and H), Ki67 (I) and phospho-histone3 (J). (K) GFAP-Cre::EYFPMuller glia express also express Pax6. Interestingly, at this time point (D6 after damage and treatment) some GFAP-Cre::EYFP Muller glia can be found in the ONL(C and D). (Scale bars: 10 �m.)

Karl et al. PNAS � December 9, 2008 � vol. 105 � no. 49 � 19509

NEU

ROSC

IEN

CE

Dow

nloa

ded

by g

uest

on

Janu

ary

13, 2

020

triple labeled for BrdU�/GFP�/Prox1� (Fig. 3H). Overall,3.6% of the BrdU� cells (0.04% of all Muller glia) differentiateinto GAD67-GFP cells. The location and morphology of theBrdU�/GFP� cells suggested that these new neurons in the INLare amacrine cells; however, both markers are also expressed inhorizontal cells, and so we cannot rule out that some of theBrdU� cells are new horizontal cells. We failed to find BrdU�/Brn3� cells, and so it is unlikely that any new ganglion cells aregenerated in our paradigm.

We also analyzed the retinas for the presence of new

ganglion cells and bipolar cells. For the bipolar cell analysis, weused a reporter mouse line that expresses EGFP under themetabotropic glutamate receptor 6 (Grm6-GFP) (24). Foranalysis of ganglion cells, we used an anti-Brn3 antibody. AfterNMDA damage, we did not find any evidence for BrdU�/Grm6-GFP bipolar cells (Fig. S4) or BrdU�/Brn3� ganglioncells in retinas treated with EGF or FGF1 and insulin. SinceOoto, et al. (9) reported that retinoic acid (RA) promotesbipolar generation of damaged adult rat retina in vivo, we alsotested whether RA treatment after NMDA damage would

A B C D

E F G HH

I J

Fig. 3. BrdU� cells express Pax6, NeuN and Calretinin (CalR) and GAD67-GFP. (A–C) After NMDA induced retinal damage and EGF or FGF1 treatment, we foundBrdU� neurons labeled with Pax6 (A), NeuN (B), and Calretinin (C) 8 to 14 days later. Images show single focal planes of 2 �m. (Scale bars: 5 �m in orthogonaland inset views; 10 �m all other views.) GAD67-GFP mice express GFP in immature neurons, a subset of mature RGCs, horizontal and all GABAergic amacrine cellsin adult mouse retina. NMDA/4xFGF1�insulin treatment of GAD67-GFP mouse retinas leads to BrdU� cells in both the ganglion cell layer (F) and the amacrinecell layer (D, E, G and H). (H) At D30 some BrdU� GAD67-GFP� cells are also labeled for Prox1. (I) Graphs showing the number of BrdU�/GAD67-GFP� cells asa function of days after NMDA damage. (J) Most of these cells are located in the inner nuclear layer, but some cells are found in the ganglion cell and horizontalcell layers. Images show z axis projections of 20 � 0.5 �m, i.e., 20 planes, 0.5 �m apart (D), 1 � 0.5 �m (d), 3 � 0.5 �m (E, e), 16 � 0.25 �m (F, f ), 2 � 0.25 �m(G, g), 2 � 1 �m (H), and 1 � 1 �m (h). (Scale bars: 5 �m in orthogonal and inset views; 10 �m all other views.)

19510 � www.pnas.org�cgi�doi�10.1073�pnas.0807453105 Karl et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

13, 2

020

promote new bipolar cell differentiation; however, we foundno BrdU�/Grm6-GFP cells.

DiscussionWe have found that Muller glia can be induced to proliferate invivo after NMDA induced neuronal degeneration, by subsequentinjections of EGF, FGF1, or a combination of FGF1 and insulin.The proliferation of the Muller glia occurs across the retina,regardless of eccentricity. Stimulation of proliferation afterNMDA damage also caused an up-regulation of progenitormarkers (Pax6, Notch, Dll1) similar to what has been reportedfor retinal regeneration in non-mammalian vertebrates—fishand chick (3, 23, 25, 26). Stimulation of proliferation in theNMDA-treated retina led to differentiation of some of theMuller glial progeny into cells that express many features ofretinal amacrine neurons, including Calretinin, NeuN, Pax6,Prox1, and GAD67-GFP. These results show for the first timethat the mammalian retina has the potential to regenerate innerretinal neurons in vivo.

Although we find that Muller glia make up the majority of thecells that incorporate BrdU in our damage/growth factor pro-tocol, three other cell populations are known to proliferate afterdamage in the retina: astrocytes, microglia, and endothelial cells(14–16). In our study, most of the proliferation of the astrocytesand endothelial cells takes place in the ganglion cell layer and isclearly separated from that of the Muller glia. Given previousresults in fish and chicks (see below) that Muller glia are thesource of new neurons after retinal damage and the data showingthat many of the BrdU� cells express Muller glial markers afterdamage in the mouse (this study), it is likely that the BrdU�inner retinal neurons we find are the progeny of Muller glia;however, we cannot rule out the possibility that endothelial cells,microglial cells, or astrocytes are the source of the new neurons.

It has long been recognized that poikilothermic vertebratesare capable of a remarkable degree of retinal regeneration (1–3).Teleost fish regenerate new retinal neurons following neurotoxindamage or surgical lesions through a blastema-like population ofprogenitor cells that arises adjacent to the site of injury. The fishretina contains at least two potential sources of progenitor cells:the mitotically active rod progenitors and Muller glia. For manyyears, it was thought that the rod progenitor was the primarysource of regenerated neurons following damage; however,recent studies using transgenic GFP reporter fish, have shownthat the primary source of regenerated neurons in the fish retinaare the Muller glia (21, 27–29). These cells are normally mitot-ically quiescent, but following damage to retinal neurons, theyre-enter the cell cycle, dedifferentiate, and express progenitorgenes, like Ascl-1 and Notch pathway components. The chickretina has a more limited regenerative capacity, although thereare many similarities with the fish (4, 23). Neurotoxic damage ofinner retinal neurons with NMDA in posthatch chicks leads toa robust proliferative response and as in the fish, the proliferatingMuller glia dedifferentiate and express retinal progenitor genes

(20). Some of the progeny spontaneously differentiate intoamacrine cells and bipolar cells.

While most mammalian Muller glia do not spontaneouslyre-enter the mitotic cycle after neurotoxic damage, as they do inthe bird and fish, Ooto, et al. (9) found a small number of Mullerglia incorporated BrdU in the rat retina after NMDA treatment.Other studies, have also reported that application of EGF, Shh,Wnt, or insulin and FGF stimulates Muller glia proliferation inadult rats following retinal damage in vivo or in vitro (5–10, 30).These previous studies have also reported that some of theprogeny of the Muller glia can also express markers of retinalphotoreceptors. BrdU� photoreceptors have been reportedafter NMDA damage (9), photoreceptor degeneration due toculture (10), in vivo following N-methyl-N-nitrosourea (MNU)(7, 8) or direct Muller glial stimulation with alpha-aminoadipate(6). Thus, outer retinal regeneration appears to be favored in therat, while in the chick, definitive evidence for photoreceptorregeneration has not been obtained. In this study, we have foundthat the mammalian retina can regenerate inner retinal neurons,most likely amacrine cells, similar to the avian retina (20).Interestingly, one of the most highly up-regulated genes in boththe mouse retina (Fig. S3) and the chick retina followingNMDA-induced damage is FoxN4 (23), a transcription factorthat biases retinal progenitor cells toward an amacrine cell fate.

Despite the similarities in amacrine cell regeneration, there isa striking difference between the fish and chick retinal responseto injury and that of the mouse or rat: Muller glia robustly andspontaneously re-enter the mitotic cell cycle following injury inthe former but not in the latter. The absence of a spontaneousproliferative response in mammalian Muller glia suggests thatsome mechanisms are in place to limit proliferation in these cellsin the adult. Close, et al. (31) investigated this question anddetermined that there are at least two molecular mechanisms inplace that maintain mammalian Muller glia in a mitoticallyquiescent state. Tgf�2, produced by retinal neurons, activates theTgfßR in progenitors, stimulating expression of the cyclin de-pendent kinase inhibitor, p27kip. In addition, the EGF receptorexpression in Muller glia gradually declines as the retina ma-tures, and the loss of a receptor for this important retinalmitogen also promotes cytostasis. Close, et al. (5) reported thatlight induced photoreceptor degeneration in albino rats causedan up-regulation in the EGF receptor in Muller glia, whichallowed them to be driven into the S-phase by intraocular EGFinjections. In the present study, we have confirmed and extendedthese results, showing that either retinal damage or EGF injec-tions alone fail to elicit a proliferative response, but bothtreatments together produce a robust response in mouse retina.In addition, we have found that EGF and FGF can stimulateMuller glial proliferation after damage in vivo. Similar resultshave been obtained in rat retina, either cultured as explants, orafter MNU treatment, which both lead to the damage of retinalphotoreceptors. In these reports, the combination of retinal

Table 1. Immunostain for BrdU� cells in the treated retina

Treatment

BrdU colabel and time point of analysis

Sox2 Pax6 NeuN CalR

D0 D2 D2 4 h D3 D8 D2 9 h D3 D8 D8 D8

Ctrl – – –NMDA Egf ��� ��� �� � �� �� �� �

Fgf1 ��� �� ��� �� �

Five confocal z-stacks across the entire retinal depths were taken (areas of 0.05 mm2) and scored to the following categories (N � 3):�, 1 cell / 0.05mm2; ��, � 3 cell / 0.05mm2; ���, � 3 cells / 0.05 mm2; blank, not quantified; Ctrl � control � solvent, EGF / FGF �Epidermal / Fibroblast growth factor.

Karl et al. PNAS � December 9, 2008 � vol. 105 � no. 49 � 19511

NEU

ROSC

IEN

CE

Dow

nloa

ded

by g

uest

on

Janu

ary

13, 2

020

damage, coupled with mitogen treatment, was required tostimulate Muller proliferation (7, 8, 10).

Taken together with earlier reports, it appears that at leastsome of the regenerative phenomena displayed in the fish andchick retina can be stimulated in the mouse retina. While manyof the progeny of the Muller cell divisions die within the firstweek after their production, the subset that survive are stable toat least 30 days and differentiate characteristics of amacrine cells(and possibly horizontal cells). Many reports of adult neurogen-esis/neuronal replacement have noted that between 50 and 80%of the newly generated neurons are eliminated (32–36) within afew weeks after they have been generated. It is not clear why thisoccurs, although it has been speculated that the new cells mustmake stable synaptic connections to survive. In zebrafish, wherefunctional retinal regeneration of all types of neurons occursafter damage, only approximately 30% of the original BrdU�Muller glia or their progeny remain after two weeks (27).

Therefore, it is possible that endogenous repair mechanismscan be stimulated in this system for replacement of other typesof retinal neurons. The process of dedifferentiation in the fish,avian, and mouse retina are remarkably similar; the Muller glialcells respond to damage (and mitogens) by re-entering the cellcycle, expressing key components of the neurogenic program ofretinal progenitors and producing new neurons. In the case of thefish, the new neurons are functionally integrated into the existingcircuitry, although this remains to be demonstrated in chicken ormammals. Muller glia exist in all vertebrate retinas, so thecellular source for regeneration is present in the human retina.Further understanding of the limits in its potential may lead tonew treatments for retinal degenerations.

Materials and MethodsMice were housed in the Department of Comparative Medicine at the Universityof Washington and maintained under a 12 h light/dark cycle with access to foodandwaterad libitum. TheGAD67-GFP (37)andGrm6-GFP (24,38) transgenicmiceused in this study were described previously. GFAP-Cre (39) mice were crossedonto B6.129 � 1-Gt(ROSA)26Sortm1(EYFP)Cos/J (Jackson Laboratory, #006148)reporter mice (� GFAP-Cre::EYFP). The use of animals in this study was in accor-dance with approved protocols and the guidelines established by the Universityof Washington, IACUC, and the National Institute of Health.

For in vivo studies, mice were anesthetized and graded glass micropipetteswith a fine tip aperture were used to inject the left eyes with 2 �l of 0.1 MNMDA, various growths factors, or control solutions. Any intraocular injectionafter the NMDA injection to the same animal included 1 mg/ml BrdU with afinal volume of 2 �l/eye. Factors were injected at the following final concen-trations unless indicated otherwise: 1 �g/�l EGF (R&D, recombinant mouse),100 ng/�l FGF1 (R&D, bovine), 1 �g/ul insulin (Sigma, bovine), and 0.3 �M/ulall-trans retinoic acid (Sigma). In addition, animals received BrdU (50 �g/g BW)by i.p. injection. Mice were killed at various time points after the injections andretinas were either processed for RNA extraction or immunohistochemistryand subsequent confocal microscopic analysis (Fig. S1 and Table S1).

For additional information, see SI Text.

ACKNOWLEDGMENTS. We thank Paige Etter and Christa Younkins for technicalassistance; Dr. Rachel A. Wong and all of the members of the Reh laboratory,especially Joe Brzezinski, for their constructive comments; Dr. Jack Saari (Univer-sity of Washington) for the generous gift of the anti-CRALBP antibody; and Z. J.Huang (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY) and N. Vardi(University of Pennslvannia, Philadelphia, PA) for providing GAD67-GFP andGrm6-GFP mice, respectively. This work was supported by a Postdoctoral Fellow-ship by the German Research Foundation (DFG, KA 2794/1-1) and ProRetinaTravelGrants toM.O.K.byNationalResearchServiceAwardF32EY016636toS.H.,byNationalResearchServiceAwardF32EY15631-01A1toB.R.N.andNationalEyeInstitute (R01 EY013475) grant to T.A.R.

1. Moshiri A, Close J, Reh TA (2004) Retinal stem cells and regeneration. Int J Dev Biol48:1003–1014.

2. Hitchcock P, Ochocinska M, Sieh A, Otteson D (2004) Persistent and injury-inducedneurogenesis in the vertebrate retina. Prog Retin Eye Res 23:183–194.

3. Raymond PA, Barthel LK, Bernardos RL, Perkowski JJ (2006) Molecular characterizationof retinal stem cells and their niches in adult zebrafish. BMC Dev Biol 6:36.

4. Fischer AJ (2005) Neural regeneration in the chick retina. Prog Retin Eye Res 24:161–182.

5. Close JL, Liu J, Gumuscu B, Reh TA (2006) Epidermal growth factor receptor expressionregulates proliferation in the postnatal rat retina. Glia 54:94–104.

6. Takeda M, et al. (2008) alpha-Aminoadipate induces progenitor cell properties ofMuller glia in adult mice. Invest Ophthalmol Visual Sci 49:1142–1150.

7. Wan J, et al. (2008) Preferential regeneration of photoreceptor from Muller glia afterretinal degeneration in adult rat. Vision Res 48:223–234.

8. Wan J, Zheng H, Xiao HL, She ZJ, Zhou GM (2007) Sonic hedgehog promotes stem-cellpotential of Muller glia in the mammalian retina. Biochem Biophys Res Commun363:347–354.

9. Ooto S, et al. (2004) Potential for neural regeneration after neurotoxic injury in theadult mammalian retina. Proc Natl Acad Sci USA 101:13654–13659.

10. Osakada F, et al. (2007) Wnt signaling promotes regeneration in the retina of adultmammals. J Neurosci 27(15):4210–4219.

11. Lucas DR, Newhouse JP (1957) The toxic effect of sodium L-glutamate on the innerlayers of the retina. Arch Ophthalmol 58:193–201.

12. Siliprandi R, et al. (1992) N-methyl-D-aspartate-induced neurotoxicity in the adult ratretina. Vis Neurosci 8:567–573.

13. Sucher NJ, Lipton SA, Dreyer EB (1997) Molecular basis of glutamate toxicity in retinalganglion cells. Vision Res 37:3483–3493.

14. Inman DM, Horner PJ (2007) Reactive nonproliferative gliosis predominates in a chronicmouse model of glaucoma. Glia 55:942–953.

15. Hanisch UK, Kettenmann H (2007) Microglia: Active sensor and versatile effector cellsin the normal and pathologic brain. Nat Neurosci 10:1387–1394.

16. Friedlander M (2007) Fibrosis and diseases of the eye. J Clin Invest 117:576–586.17. Taranova OV, et al. (2006) SOX2 is a dose-dependent regulator of retinal neural

progenitor competence. Genes Dev 20:1187–1202.18. Poche RA, Furuta Y, Chaboissier MC, Schedl A, Behringer RR (2008) Sox9 is expressed in

mouse multipotent retinal progenitor cells and functions in Muller glial cell develop-ment. J Comp Neurol 510:237–250.

19. Moshiri A, et al. (2008) Near complete loss of retinal ganglion cells in the math5/brn3bdouble knockout elicits severe reductions of other cell types during retinal develop-ment. Dev Biol 316:214–227.

20. Fischer AJ, Reh TA (2001) Muller glia are a potential source of neural regeneration inthe postnatal chicken retina. Nat Neurosci 4:247–252.

21. Bernardos RL, Barthel LK, Meyers JR, Raymond PA (2007) Late-stage neuronal progen-itors in the retina are radial Muller glia that function as retinal stem cells. J Neurosci27:7028–7040.

22. Hitchcock PF, Macdonald RE, VanDeRyt JT, Wilson SW (1996) Antibodies against Pax6immunostain amacrine and ganglion cells and neuronal progenitors, but not rodprecursors, in the normal and regenerating retina of the goldfish. J Neurobiol 29:399–413.

23. Hayes S, Nelson BR, Buckingham B, Reh TA (2007) Notch signaling regulates regener-ation in the avian retina. Dev Biol 312:300–311.

24. Dhingra A, et al. (2008) Probing neurochemical structure and function of retinal ONbipolar cells with a transgenic mouse. J Comp Neurol 510:484–496.

25. Kassen SC, et al. (2007) Time course analysis of gene expression during light-inducedphotoreceptor cell death and regeneration in albino zebrafish. Dev Neurobiol67:1009–1031.

26. Yurco P, Cameron DA (2007) Cellular correlates of proneural and Notch-delta geneexpression in the regenerating zebrafish retina. Vis Neurosci 24:437–443.

27. Fausett BV, Goldman D (2006) A role for alpha1 tubulin-expressing Muller glia inregeneration of the injured zebrafish retina. J Neurosci 26:6303–6313.

28. Thummel R, Kassen SC, Montgomery JE, Enright JM, Hyde DR (2008) Inhibition ofMuller glial cell division blocks regeneration of the light-damaged zebrafish retina.Dev Neurobiol 68:392–408.

29. Yurco P, Cameron DA (2005) Responses of Muller glia to retinal injury in adultzebrafish. Vision Res 45:991–1002.

30. Insua MF, et al. (2008) Trophic factors and neuronal interactions regulate the cell cycleand Pax6 expression in Muller stem cells. J Neurosci Res 86:1459–1471.

31. Close JL, Gumuscu B, Reh TA (2005) Retinal neurons regulate proliferation of postnatalprogenitors and Muller glia in the rat retina via TGF beta signaling. Development(Cambridge, UK) 132:3015–3026.

32. Cameron HA, McKay RD (2001) Adult neurogenesis produces a large pool of newgranule cells in the dentate gyrus. J Comp Neurol 435:406–417.

33. Dayer AG, Cleaver KM, Abouantoun T, Cameron HA (2005) New GABAergic interneu-rons in the adult neocortex and striatum are generated from different precursors. J CellBiol 168:415–427.

34. Lemasson M, Saghatelyan A, Olivo-Marin JC, Lledo PM (2005) Neonatal and adultneurogenesis provide two distinct populations of newborn neurons to the mouseolfactory bulb. J Neurosci 25:6816–6825.

35. Petreanu L, Alvarez-Buylla A (2002) Maturation and death of adult-born olfactory bulbgranule neurons: Role of olfaction. J Neurosci 22:6106–6113.,

36. Sun W, et al. (2004) Programmed cell death of adult-generated hippocampal neuronsis mediated by the proapoptotic gene Bax. J Neurosci 24:11205–11213.

37. Chattopadhyaya B, et al. (2004) Experience and activity-dependent maturation ofperisomatic GABAergic innervation in primary visual cortex during a postnatal criticalperiod. J Neurosci 24:9598–9611.

38. Morgan JL, Dhingra A, Vardi N, Wong RO (2006) Axons and dendrites originate fromneuroepithelial-like processes of retinal bipolar cells. Nat Neurosci 9:85–92.

39. Garcia AD, Doan NB, Imura T, Bush TG, Sofroniew MV (2004) GFAP-expressing pro-genitors are the principal source of constitutive neurogenesis in adult mouse forebrain.Nat Neurosci 7:1233–1241.

19512 � www.pnas.org�cgi�doi�10.1073�pnas.0807453105 Karl et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

13, 2

020