Assessment of intrinsic and extrinsic signaling pathway in excitotoxic retinal ganglion cell death

ABSTRACT Excitotoxicity leads to the activation of a cytotoxic cascade that causes neuronal death. In the retina, retinal ganglion cells (RGCs) die after an excitotoxic insult. Multiple

pathways have been proposed to contribute to RGC death after an excitotoxic insult, including TNF signaling, JNK activation, and ER stress. To test the importance of these pathways in RGC

death after excitotoxic injury, the excitotoxin N-methyl-D-aspartate (NMDA) was intravitreally injected into mice deficient in components of these pathways. Absence of _Tnf_ or its canonical

downstream mediator, _Bid_, did not confer short- or long-term protection to RGCs. Despite known activation in RGCs and a prominent role in mediating RGC death after other insults,

attenuating JNK signaling did not prevent RGC death after excitotoxic insult. Additionally, deficiency of the ER stress protein DDIT3 (CHOP), which has been shown to be involved in RGC

death, did not lessen NMDA induced RGC death. Furthermore, absence of both _Jun_ (JNK’s canonical target) and _Ddit3_, which together provide robust, long-term protection to RGC somas after

axonal insult, did not lessen RGC death. Collectively, these results indicate that the drivers of excitotoxic injury remain to be identified and/or multiple cell death pathways are activated

in response to injury. SIMILAR CONTENT BEING VIEWED BY OTHERS MOLECULAR MECHANISMS OF NMDA EXCITOTOXICITY IN THE RETINA Article Open access 27 October 2023 INTERCELLULAR COMMUNICATION ATLAS

REVEALS OPRM1 AS A NEUROPROTECTIVE FACTOR FOR RETINAL GANGLION CELLS Article Open access 11 March 2024 Σ2R/TMEM97 IN RETINAL GANGLION CELL DEGENERATION Article Open access 01 December 2022

INTRODUCTION Excessive stimulation of glutamate receptors has been shown to disrupt the intracellular environment and lead to activation of cytotoxic cascades culminating in neuronal death1.

In the retina, excitotoxic insult is thought to contribute to several diseases, including diabetic retinopathy and retinal ischemia2,3,4. Multiple cell types, including both neurons and

glia, are affected by excitotoxic insult. In the retina, excitotoxic insult is known to cause the death of retinal ganglion cells (RGCs) and at least some types of amacrine cells5,6,7,8,9.

Despite excitotoxicity’s importance in disease, the required cell death pathways controlling excitotoxic induced RGC death are not well-defined. Multiple intrinsic and extrinsic cell death

signaling pathways have been implicated in neuronal death after excitotoxic injury. Recently, TNFR1 activation has been shown to initiate the process of RGC death after excitotoxic insult7.

A key mediator downstream of TNFR1 activation, the pro-apoptotic protein BID, has also been shown to be important in mediating neuronal damage after glutamate toxicity10 including in RGCs11.

Interestingly, if BID is important for NMDA induced RGC death, it is not acting through BAX activation, since _Bax_ deficiency does not prevent RGC death after excitotoxic insult8,12.

Additionally, the JNK (c-Jun N-terminal kinases) pathway has been proposed as a regulator of neuronal death after excitotoxic insult. Canonical JNK signaling leads to activation of the

transcription factor JUN, a member of the activation protein-1 (AP-1) family of transcription factors13. In the eye, JNK and JUN activation in RGCs is present after induced ischemia, axonal

injury, increased intraocular pressure, and excitotoxicity14,15,16,17,18,19. Pan inhibitors against JNK, as well as _Jun_ antisense oligodeoxynucleotides, provided some protection to inner

retinal neurons after an excitotoxic insult20,21,22. Furthermore, ER stress has been proposed to mediate excitotoxic cell death23,24,25,26. In the retina, an increase in expression of the ER

stress gene _Ddit3_ (_Chop_) has been found in RGCs after intravitreal injection of the excitotoxin NMDA25. _Ddit3_ deficiency provided a modest, but significant level of protection of

ganglion cell layer neurons (which consists of RGCs and amacrine cells) after and intravitreal injection of NMDA23. It has also been observed that lack of _Ddit3_ significantly reduces RGC

death after axonal injury27. Thus, there appears to be numerous cell death pathways contributing to RGC death after an excitotoxic insult. Here, the importance of these pathways in

excitotoxic RGC death is critically tested using mice deficient in critical molecular components of these pathways. RESULTS INTRAVITREAL NMDA INJECTION KILLS RGCS Administration of

excitotoxins to the retina affects multiple cell types, including RGCs28. In order to establish a relevant concentration of NMDA to study NMDA-induced RGC death, the effect on RGC survival

of different concentrations of NMDA was examined. 2 μl of three concentrations of NMDA (2 mM, 20 mM, or 80 mM) were intravitreally injected into C57BL/6 J mice. For controls, the

contralateral eye was injected with 2 μl of PBS. The number of TUJ1+ RGCs remaining 7 days after injection was counted. All concentrations of NMDA (2 mM, 20 mM and 80 mM) resulted in a

significant decrease in the number of TUJ1+ RGCs compared to control eyes (_p_ < 0.001 for all concentrations compared to control; Fig. 1A,B). There was also significant decrease in cell

number between 2 mM and 20 mM injections of NMDA (_p_ < 0.001, for all comparisons); however, there was no significant decrease in cell number between 20 mM and 80 mM NMDA injection (_p_ 

= 0.819). The 20 mM NMDA concentration was chosen for future experiments based upon this concentration producing a significant amount of retinal neuron loss and relevance to previously

published studies7,29,30,31. To determine the window of RGC loss, TUJ1+ cells were counted 6hrs, 1d, 3d, 7d, 14d, and 28d after intraocular injection of 20 mM NMDA (Fig. 1C,D). There was a

significant loss of approximately 50% of TUJ1+ cells 6 hr and 1d after NMDA insult (_p_ < 0.001 comparing either 6 hour or 1 day after NMDA injection to control eyes). Between 1 day and 3

days after injection there was a further loss of RGCs (_p_ = 0.007, comparing 1d to 3d). However, after 3 days, there was not a further, significant decrease in TUJ1+ cell number _(p_ > 

0.05 comparing 3d to either 7d, 14d, or 28d). Thus, cell loss appears to be complete by 3 days after 20 mM NMDA insult. TNF AND BID ARE NOT REQUIRED FOR NMDA-INDUCED RGC DEATH TNF has been

suggested to be critical for inducing RGC death 6 hours after excitotoxic injury7. In order to determine if TNF is required for long-term NMDA induced RGC death, 2 µl of PBS (control,

contralateral eye) or 20 mM of NMDA was injected into the vitreous of _Tnf_ deficient (_Tnf_−/−) mice. The number of surviving TUJ1+ RGCs was counted at 6 hours and 7 days after intravitreal

NMDA injection, two time points that encompass when RGCs are dying and after the time window of cell death (see Fig. 1). _Tnf_ deficiency did not lessen RGC loss at either the 6 hour or 7 

day time point compared to wildtype mice (_p_ = 0.9694 comparing WT to _Tnf_−/− NMDA; Fig. 2). These data suggest that _Tnf_ is not required nor an important contributor to RGC death after

excitotoxic injury. Extrinsic death receptors, including TNFR1, may recruit the intrinsic pathway in order to amplify the cell death cascade. This is mediated through BID cleavage and

fragment translocation to the mitochondria where it activates BAX and/or BAK, two pro-apoptotic BCL2 proteins that are critical for apoptotic cell death32. To determine if BID is required

for NMDA induced RGC death, 2 μl of either 20 mM NMDA or PBS was injected into the vitreous of _Bid_ deficient (_Bid_−/−) and wildtype (_Bid_+/+) mice. The number of surviving RGCs (TUJ1+

cells) was counted at 6 hours and 7 days after insult. Significant RGC loss in both wildtype and _Bid_ deficient eyes was observed at both time points. Furthermore, _Bid_ deficiency did not

lessen RGC death (Fig. 3; _p_ ≥ 0.719 comparing WT to _Bid_−/− NMDA). These data suggest BID is not a critical mediator of short- and long-term excitotoxic RGC death. JUN IS ACTIVATED IN

MULTIPLE DIFFERENT RETINAL CELL TYPES AFTER EXCITOTOXIC INSULT The transcription factor JUN is activated by JNK signaling. JNK inhibitors have lessened RGC death after an NMDA

insult20,21,22. However, there was variability in RGC protection between inhibitors and it is probable that the inhibitors are not completely specific to JNKs, therefore it is important to

test the JNK pathway in excitotoxicity induced RGC death using genetics. In order to determine if JUN was activated (phosphorylated) in retinal cells after an excitotoxic insult, retinal

flat mounts were examined for pJUN at 6 hours after intravitreal injections of either NMDA or PBS. Similar to previous findings, pJUN was detected in TUJ1+ RGCs after NMDA insult (Fig. 4)22.

These results demonstrate that JUN dependent MAPK signaling is active in RGCs after an NMDA insult. MAPK SIGNALING DOES NOT APPEAR TO BE IMPORTANT FOR RGC DEATH AFTER NMDA INSULT DLK is a

MAP3K upstream of JNK and JUN that has been shown to be critical for excitotoxic-induced neuronal death. Deficiency of _Dlk_ significantly protects CA1 hippocampal neurons after excitotoxic

injury33. In order to assess the involvement of DLK in excitotoxic mediated RGC loss, DLK was conditionally deleted in the retina using a floxed allele of _Dlk_ and the retinal cre,

_Six3-cre_. Previously, we have shown that _Six3_-_cre_ successfully deletes _Dlk_ (and _Jun_, discussed below) in approximately 85% of RGCs34,35. Seven days post-NMDA injection, RGC loss

was comparable between wildtype and _Dlk_ deficient (_Dlk__fl/fl__; Six3-cre_+) mice (Fig. 5; _p_ ≥ 0.1843 comparing WT to _Dlk__fl/fl__; Six3-cre_+ NMDA injected eyes). These results

suggest that DLK is not critical for NMDA-induced RGC death. There are three JNKs in vertebrates, _Jnk1, Jnk2_, and _Jnk3_36. Deficiency of _Jnk3_ has been reported to confer protection to

hippocampal neurons in the brain after excitotoxic injury37. Inhibitors against JNK have produced mixed results, showing minimal or very significant protection for RGCs after

excitotoxicity20,22. After axonal injury, combined absence of _Jnk2_ and _Jnk3_ provides significant protection from RGC loss, showing that _Jnk2_ and _Jnk3_ can mediate pro-death signaling

in injured RGCs38. Therefore, in order to examine the role of JNK in NMDA mediated RGC death using a genetic approach, mice deficient for both _Jnk2_ and _Jnk3_ (_Jnk2_−/− _Jnk3_−/−) were

intravitreally injected with 2 μl of 20 mM NMDA. Compared to wildtype mice, the number of RGCs lost 7 days post-intravitreal injection of NMDA was similar in _Jnk2_−/− _Jnk3_−/− mice (Fig. 

6; _p_ ≤ 0.001 for comparison between PBS and NMDA for each genotype, _p_ ≥ 0.9769 comparing WT to _Jnk2_−/− _Jnk3_−/− NMDA). This result shows that the JNK involved in excitotoxic death in

other neurons (JNK3) and the JNKs that have been shown to be important in RGC death after an insult (JNK2 and JNK3) are not required for NMDA-mediated RGC death. This suggests that _Jnk1_

could have a role in the excitotoxic injury response, in the future it would be interesting to look at a conditional knockout of all three JNK isoforms. JUN is a canonical injury response

factor downstream of JNK and DLK that is known to mediate RGC death. JUN has been shown to be a mediator of RGC death after other injuries, for example inhibiting _Jun_ significantly lessens

RGC death after mechanical axonal injury or axotomy38,39. The role of JUN in excitotoxic-mediated RGC death is unclear, though protection of RGCs was found using a c-Jun antisense

oligodeoxynucleotide after excitotoxic insult21. In order to critically test the importance of JUN in NMDA induced RGC death, NMDA was injected into the vitreous of _Jun_ deficient mice.

Deletion of _Jun_ is embryonically lethal, therefore a conditional retinal deletion of JUN was achieved using a floxed allele of JUN and the early retinal deleter cre, _Six3-cre_

(_Jun__fl/fl__; Six3-cre_+; referred to as _Jun_ deficient mice)40,41,42,43. The number of surviving RGCs (TUJ1+ cells) were counted 7 days after intravitreal injection, a time point when

the majority of RGCs have died in wildtype retinas. There was a significant loss of RGCs in both the wildtype and _Jun_ deficient retinas after NMDA insult (Fig. 7; _p_ = 0.9846, comparing

WT to _Jun__fl/fl__; Six3-cre_+ NMDA). This result shows that the injury response factor JUN is not required for NMDA-mediated RGC death. THE ER STRESS PROTEIN DDIT3 (CHOP) IS NOT REQUIRED

FOR RGC EXCITOTOXIC DEATH ER stress has been suggested to mediate excitotoxic RGC death and an increase in expression of the ER stress protein _Ddit3_ has been found in RGCs after NMDA

injury25. _Ddit3_ has been shown to be involved in RGC death after several insults and modest protection of ganglion cell layer cells was seen in _Ddit3_ null mice after intraocular

injection of NMDA23. Thus, this ER stress pathway could play an important role in excitotoxic RGC death. Therefore, in order to determine the extent of involvement of DDIT3 in mediating RGC

death after excitotoxic injury, mice deficient in _Ddit3_ were injected with NMDA and RGC loss analyzed 7 days post-injection. Compared to wildtype, _Ddit3_ deficient mice had no significant

change in surviving TUJ1+ RGCs after NMDA insult (Fig. 8; _p_ = 0.8105 comparing WT to _Ddit3_−/− NMDA). This result suggests the ER stress pathway involving DDIT3 does not play a major

role in mediating RGC death after NMDA injury. _Jun_ or _Ddit3_ deficiency alone did not prevent RGC loss after excitotoxic injury, however it has been shown that deficiency in _Jun_ and

_Ddit3_ together provide increased protection to RGC somas after an axonal injury44. To test whether these two pathways together mediate RGC loss after NMDA insult, surviving RGCs were

counted in mice deficient in both _Jun_ and _Ddit3_ (_Jun__fl/fl__Ddit3_−/−_; Six3-cre_+) after an excitotoxic injury. Mice deficient in both _Jun_ and _Ddit3_ had a similar amount of RGC

loss as wildtype mice 7 days post-NMDA injection (Fig. 9; _p_ = 0.8582 between WT and _Jun__fl/fl__Ddit3_−/−_; Six3-cre_+ NMDA). KAINIC ACID INDUCED RGC DEATH IS NOT MEDIATED BY JUN OR DDIT3

There are differences in the severity and specificity of cell death based upon the type of excitotoxin45. Another excitotoxin that is used to induce excitotoxic cell death is kainic

acid45,46,47. Kainic acid, along with NMDA, is used for investigating neuroprotective strategies to limit RGC loss in glaucoma, retinal ischemia, and diabetic retinopathy48,49,50. _Jun_ and

_Ddit3_ have been shown to be important in KA-mediated death26,37,51, thus it is possible that in the retina these molecules are necessary for RGC death induced by KA, but not NMDA. In order

to find a concentration that produces consistent and reproducible RGC death, different concentrations of KA were intravitreally injected into wildtype mice (10 mM, 12 mM, 15 mM, or 20 mM;

Fig. 10). These KA concentrations are similar to what has been previously used to induce retinal injury50,52. All concentrations of KA produced significant RGC loss compared to PBS injection

(Fig. 10; _p_ < 0.001 for all concentrations compared to control), though the RGC loss achieved was not statistically significant between KA concentrations. Therefore, to assess the

involvement of _Jun_ and _Ddit3_ in KA-induced RGC death, 15 mM KA was intravitreally injected into mice deficient in both _Jun_ and _Ddit3_ (_Jun__fl/fl__Ddit3_−/−_; Six3-cre_+). Surviving

RGCs (TUJ1+ cells) were assessed 7 days post injection and there was no statistical difference between KA-injected wildtype and _Jun/Ddit3_ deficient mice (Fig. 11; _p_ = 0.9962 between WT

and _Jun__fl/fl__Ddit3_−/−_; Six3-cre_+ KA). This result is similar to what was described above with NMDA-mediated RGC injury. Overall, these data suggest that the JUN/JNK signaling pathway

does not play a role in either NMDA or KA-induced excitotoxic RGC death. DISCUSSION In this study, the importance of key components of the extrinsically activated mitochondrial cell death

pathway for RGC death after excitotoxic injury were tested. The absence of _Tnf_ or _Bid_ did not lessen RGC death after excitotoxic insult. The importance of intrinsic cell death signaling

pathways was also tested using mouse mutants of various components of the JNK-JUN signaling pathway (_Dlk_, _Jnk2_, _Jnk3_ and _Jun_) and ER stress pathway (_Ddit3_). Finally, we tested

whether combined deficiency in _Jun_ and _Ddit3_ provided additive protection to RGCs after excitotoxic insult. Collectively, our data show key extrinsic and intrinsic pathways that have

been implicated in RGC death do not appear to be necessary for RGC death after excitotoxic injury. TNF has been implicated in RGC death after excitotoxic insult7. Surprisingly, in our hands

_Tnf_ deficiency did not provide long-term protection to RGCs after excitotoxic injury nor did it prevent RGC loss at an early time point after insult. Additionally, absence of _Bid_ failed

to prevent RGC loss after NMDA insult. Together, these data show a key extrinsic pathway that has been implicated in RGC death does not appear to be necessary for mediating RGC death after

excitotoxic injury. Lebrun-Julien and colleagues previously showed a significant, but incomplete protection against RGC death in _Tnf_ deficient mice six hours after an intravitreal NMDA

injection7. These experiments are inconsistent with the work presented here showing no protection provided by _Tnf_ deficiency at six hours after intravitreal injection of NMDA7. It is

unclear why the experiments presented here, performed with similar concentrations of NMDA, similar procedures, and both using TNF deficient mice obtained different results compared to

Lebrun-Julien and colleagues previous work7. However, the longest time point examined by Lebrun-Julien _et al_. was only 6 hours after injury, and at this time point only a partial

protection was observed (about half of the cells that had died in the control were protected). It is important to note that Lebrun-Julien and colleagues7 found similar protection at this

time point using a TNF neutralizing drug, Etanercept. Therefore, it is possible that slight differences in the kinetics of cell loss between the different experiments explain the observed

differences. Nevertheless, the experiments described here show that TNF signaling is not required for RGC death after an excitotoxic insult. Multiple intrinsic pro-death signaling pathways

have been implicated in excitotoxic induced RGC death. JNK/JUN signaling has been shown to be involved in excitotoxic neuronal death in multiple systems. Manipulating upstream regulators of

JUN have been shown to reduce RGC death after excitotoxic injury, for example _Jnk3_ deficiency lessened neuronal loss after a kainic acid-induced excitotoxic injury37. Also, an upstream

kinase of JNK which is known to activate JNK in injured RGCs35,53,54, DLK, has been shown to be involved in excitotoxic insult induced neuronal death33,35. In RGCs, multiple lines of

evidence point to JNK/JUN signaling having a role in cell death after an excitotoxic insult. JNK and JUN are activated in RGCs after induced excitotoxicity (Fig. 4)18,22. Also, inhibiting

JNK signaling by an intravitreal injection of a JNK inhibitor decreased RGC loss after NMDA insult22. However, despite the potential importance of JNK signaling in excitotoxic injured

neurons, experiments herein which critically tested key components of JNK/JUN signaling showed that this pathway may not be a primary driver of excitotoxic RGC death. Specifically, we showed

that deficiency in _Jun_, _Jnk2/3_, and _Dlk_ did not prevent excitotoxicity-induced RGC death. It is unclear why the results presented here differ with previous studies showing protection

from excitotoxic injury after inhibiting the JNK/JUN signaling pathway20,21,22. Two previous studies used pharmaceutical approaches to inhibit JNK signaling20,22. For these studies it is

possible that there was inhibition of other kinases that contributed to the protection55. Munemasa and colleagues specifically tested the importance of JUN activation in NMDA induced RGC

death by intravitreally injecting _Jun_ anti-sense oligodeoxynucleotides21. It is possible that this experiment also resulted in off target effects. Furthermore, all of these experiments

counted total ganglion cell layer neurons and not just RGCs. The ganglion cell layer in the mouse consists of approximately equal numbers of RGCs and amacrine cells56 and both cell types die

after excitotoxic insult28. Thus, it is possible that there is protection of amacrine cells that explains the differences between the studies. It is also possible that the genetic deletion

of JNK signaling molecules during development alters the response of RGCs to excitotoxic insult. The lack of protection observed in this study by components of the JNK/JUN signaling pathway

is an intriguing result given the known importance of JNK/JUN signaling in RGCs after other insults, particularly after axonal injury38,57,58,59. Thus, these results suggest that in RGCs

different molecular pathways are activated depending on the insult. Furthermore, given the potential importance of excitotoxicity in diabetes and ischemic optic neuropathy and axonal injury

in glaucoma it suggests that RGCs die by different mechanisms in common human diseases. It is important to note that the experiments described here have not ruled out the potential

importance of a JNK1 contribution, particularly a JUN-independent JNK1 function, in mediating RGC death after an excitotoxic insult. It will be important in the future to test the role of

JNK1 and even the simultaneous inhibition of all three JNKs to fully define the importance of JNK signaling in excitotoxicity induced RGC death. After an excitotoxic insult, ER stress

components are upregulated and active in RGCs, including _Ddit3_25. Furthermore, _Ddit3_ deficiency has been shown to protect about 10 cells/mm in the ganglion cell layer more than in

C57BL/6 J after an excitotoxic injury23. _Ddit3_ deficiency did not provide protection to RGCs 7 days after insult, suggesting it is not required for RGC death after and excitotoxic insult.

Both JNK/JUN and ER stress signaling contribute to RGC death after axonal injury27,38. Recently, our group showed that these pathways appeared to be independently regulated in RGCs after

axonal injury44. In fact, absence of _Jun_ and _Ddit3_ provided robust, long-term protection to RGCs after an axonal insult44. Thus, it is possible that both of these pathways could

contribute to RGC death after an excitotoxic insult and that each is sufficient to kill RGCs. To critically test this hypothesis, RGCs were counted from mice deficient in both _Jun_ and

_Ddit3_ after an excitotoxic insult. Deficiency of both _Jun_ and _Ddit3_ failed to provide protection to RGCs after intravitreal NMDA injection 7 days after insult. Furthermore, RGC loss

after intravitreal injection of kainic acid was not lessened by _Jun/Ddit3_ deficiency. Together these data show that JUN and DDIT3 dependent pathways are not required to prevent RGC death

after an excitotoxic injury. This result was surprising given that both _Jun_ and _Ddit3_ are known RGC pro-death molecules and are expressed in RGCs after an excitotoxic injury. Studies

have reported that manipulating various pathways provide protection of RGCs7,21,22,23,60. However, most of these studies have shown only a minor or short-term protection, especially those

targeting parts of the pathway downstream of glutamate receptor activation. Furthermore, to date, no one method of intervention has shown complete protection from excitotoxic RGC death.

These studies coupled with the results described here indicate that multiple, independent pro-death pathways are active after an excitotoxic insult. This potentially makes preventing

excitotoxic RGC death very difficult and designing therapeutic targets challenging in chronic diseases where excitotoxicity has been implicated. Simply inhibiting NMDA and KA channels for

the duration of the chronic diseases are likely be detrimental to the system or become ineffective. Additionally, NMDA receptor antagonists failed to show efficacy in multiple clinical

trials for stroke and TBI, relegating development of potential therapeutics to downstream pathogenic signaling pathways61. Thus, multiple pro-death pathways may need to be targeted to

prevent vision loss in diseases were excitotoxicity contributes to RGC death. In the future, it will be important to define all of the pro-death pathways active in RGCs after excitotoxic

insult in order to identify  common upstream regulators of these pathways. It is important to note that this study used a typical procedure of giving a single intravitreal injection of a

fairly large amount of NMDA or KA. It is unlikely that this type of injury is what occurs during disease. To understand the signaling pathways that control excitotoxic RGC death in human

diseases it will be important to model the precise nature of the excitotoxic insult in terms of both concentration and length of insult. Therefore, to properly model disease relevant

exposures to excitotoxins, experiments should be performed where excitotoxins are chronically delivered to the retina at low or varying levels. MATERIALS AND METHODS MICE Mice carrying null

alleles of _Tnf (Tnf__tm1Gkl_), _Bid_ (_Bid__tm1Sjk_), _Jnk2_ (_Mapk9__tm1Flv_) and _Jnk3_ (_Mapk10__tm1Flv_), or _Ddit3_ (_Ddit3__tm2.1Dron_) were obtained from the Jackson Laboratory.

Additionally, floxed alleles of _Dlk_ (_Dlk__fl_)62; and _Jun_ (_Jun__fl_)40; were used. Floxed alleles were recombined using a retinal expressed cre, Tg(Six3-cre)69Frty mice (referred to as

Six3-cre)41 Note a “?” for an allele denotes that for controls some animals carried a WT allele while others carried either a null or floxed allele as appropriate (e.g.+/? means both +/+

and +/− genotypes were used). All colonies were maintained by intercrossing. All alleles were on a C57BL/6 J genetic background except for the _Jnk2_ and _Jnk3_ cross which was on a mixed

C57BL/6 J and DBA/2 J genetic background. Mice were housed in a 12-hour light dark cycle and were fed chow and water ad libitum. All mice used were adults (2–5 months of age). All

experiments were conducted in accordance with the Association for Research in Vision and Ophthalmology’s statement on the use of animals in ophthalmic research and were approved by the

University of Rochester’s University Committee on Animal Resources. INTRAOCULAR INJECTION Mice were anaesthetized with an intraperitoneal injection of 0.05 ml/10 g solution containing

ketamine (20 mg/mL) and xylazine (2 mg/mL). A small incision was made with a 30-gauge needle behind the limbus at the superiotemporal quadrant through the conjunctiva at a 45° angle. 2 mM,

20 mM, and 80 mM concentrations of N-Methyl-D-aspartate (NMDA; Sigma) and 10 mM, 12 mM, 15 mM, and 20 mM KA were made in sterile 0.1 M PBS. 2 μl of either NMDA or KA concentrations were

injected into the vitreous through the sclera just behind the limbus, using a 5 μL Hamilton syringe equipped with a 33-gauge removable needle. The contralateral eye was injected with 2 μL of

vehicle (0.1 M PBS) to serve as a control. To reduce an increase in intraocular pressure, the intravitreal injections were conducted over 2 minutes. The condition of the injected eyes was

monitored to ensure no problems arose due to the injection (e.g. cataract). IMMUNOHISTOCHEMISTRY AND CELL COUNTS RGC survival was assessed using the RGC marker TUJ1, whose expression is

sustained following injury63,64,65. Eyes were processed as previously described38,64. Following fixation in 4% paraformaldehyde (PFA), the anterior segment of each eye was removed and the

posterior eyecup was processed for whole mount immunostaining or cryosectioning. For whole mount immunostaining, retinas were blocked in 0.4% Triton X-100 in PBS containing 10% horse serum

for 3–4 hours. Retinas were then incubated in primary antibody against mouse anti-βIII tubulin (mouse anti-TUJ1, BioLegend, 1:1000) diluted in 0.3% Triton X-100 in PBS for 3 days at 4 °C.

Following washes in PBS, the retinas were incubated with Alexa Fluor conjugated secondary antibodies (Invitrogen, 1:1000) diluted in PBS for 1 day at 4 °C and then mounted ganglion cell

layer up onto slides. RGC density varies greatly with respect to retinal location. For each retina, images were obtained from eight 40× fields around the peripheral retina (two from each

quadrant), each field approximately 220 μm from the peripheral edge of the retina (one half of a 40× field in from the peripheral margin). The numbers of neurons immunolabeled with TUJ1 in

each image were quantified using the cell-counter tool in ImageJ (NIH). Counts from each field were summed. For immunohistochemistry on retinal sections, cryosections were blocked by

incubating in 10% horse serum in 0.1% Triton X-100 in PBS (PBST) for 2–3 h at room temperature. Sections were incubated with primary antibodies (mouse anti-TUJ1, BioLegend, 1:1000; rabbit

anti-pJUN, Cell Signaling, 1:250) diluted in PBST overnight at 4 °C. The following day the sections were washed and incubated with Alexa Fluor-conjugated secondary antibodies (Invitrogen,

1:1000) diluted in PBST for a minimum of 2 hours. Sections were then counterstained with 4′,6-diamidino-2-phenylindole (DAPI; Fisher Scientific). For terminal deoxynucleotidyl

transferase-mediated dUTP end labeling (TUNEL), mice were euthanized and cryosections were prepared as described above. TUNEL assays were performed according to the manufacturer’s

instructions (ApoTag, EMD Millipore). STATISTICAL ANALYSIS During quantification of results, the experimenter was masked to genotype and/or experimental group. Experiments with two groups

were analyzed for differences using a one-way ANOVA and three or more groups, either multiple time points or genotypes, underwent statistical analyses using a two-way ANOVA with significance

determined at _P_ values < 0.05, followed by the Tukey’s or Sidak’s multiple comparisons test for group comparisons. REFERENCES * Olney, J. W. Brain lesions, obesity, and other

disturbances in mice treated with monosodium glutamate. _Science_ 164, 719–721 (1969). Article  ADS  CAS  PubMed  Google Scholar  * Araszkiewicz, A. & Zozulinska-Ziolkiewicz, D. Retinal

neurodegeneration in the course of diabetes - pathogenesis and clinical perspective. _Curr. Neuropharmacol_. (2016). * Barber, A. J. A new view of diabetic retinopathy: a neurodegenerative

disease of the eye. _Prog. Neuropsychopharmacol. Biol. Psychiatry_ 27, 283–290 (2003). Article  CAS  PubMed  Google Scholar  * Ishikawa, M. Abnormalities in Glutamate Metabolism and

Excitotoxicity in the Retinal Diseases. _Scientifica_ 2013, e528940 (2013). Article  Google Scholar  * Fischer, A. J., Pickett Seltner, R. L., Poon, J. & Stell, W. K. Immunocytochemical

characterization of quisqualic acid- and N-methyl-D-aspartate-induced excitotoxicity in the retina of chicks. _J. Comp. Neurol._ 393, 1–15 (1998). Article  CAS  PubMed  Google Scholar  *

Lam, T. T., Abler, A. S., Kwong, J. M. K. & Tso, M. O. M. N-Methyl-D-Aspartate (NMDA)–Induced Apoptosis in Rat Retina. _Invest. Ophthalmol. Vis. Sci._ 40, 2391–2397 (1999). CAS  PubMed 

Google Scholar  * Lebrun-Julien, F. _et al_. Excitotoxic death of retinal neurons _in vivo_ occurs via a non-cell-autonomous mechanism. _J. Neurosci. Off. J. Soc. Neurosci._ 29, 5536–5545

(2009). Article  CAS  Google Scholar  * Libby, R. T. _et al_. Susceptibility to Neurodegeneration in a Glaucoma Is Modified by Bax Gene Dosage. _PLoS Genet_. 1 (2005). * Sucher, N. J.,

Lipton, S. A. & Dreyer, E. B. Molecular basis of glutamate toxicity in retinal ganglion cells. _Vision Res._ 37, 3483–3493 (1997). Article  CAS  PubMed  Google Scholar  * Tobaben, S. _et

al_. Bid-mediated mitochondrial damage is a key mechanism in glutamate-induced oxidative stress and AIF-dependent cell death in immortalized HT-22 hippocampal neurons. _Cell Death Differ._

18, 282–292 (2011). Article  CAS  PubMed  Google Scholar  * Uchibayashi, R., Tsuruma, K., Inokuchi, Y., Shimazawa, M. & Hara, H. Involvement of Bid and caspase-2 in endoplasmic reticulum

stress- and oxidative stress-induced retinal ganglion cell death. _J. Neurosci. Res._ 89, 1783–1794 (2011). Article  CAS  PubMed  Google Scholar  * Li, Y., Schlamp, C. L., Poulsen, K. P.

& Nickells, R. W. Bax -dependent and Independent Pathways of Retinal Ganglion Cell Death Induced by Different Damaging Stimuli. _Exp. Eye Res._ 71, 209–213 (2000). Article  CAS  PubMed 

Google Scholar  * Hibi, M., Lin, A., Smeal, T., Minden, A. & Karin, M. Identification of an oncoprotein- and UV-responsive protein kinase that binds and potentiates the c-Jun activation

domain. _Genes Dev._ 7, 2135–2148 (1993). Article  CAS  PubMed  Google Scholar  * Isenmann, S. & Bähr, M. Expression of c-Jun Protein in Degenerating Retinal Ganglion Cells after Optic

Nerve Lesion in the Rat. _Exp. Neurol._ 147, 28–36 (1997). Article  CAS  PubMed  Google Scholar  * Kim, B.-J. _et al_. _In vitro_ and _in vivo_ neuroprotective effects of cJun N-terminal

kinase inhibitors on retinal ganglion cells. _Mol. Neurodegener._ 11, 30 (2016). Article  PubMed  PubMed Central  Google Scholar  * Kwong, J. M. K. & Caprioli, J. Expression of

phosphorylated c-Jun N-terminal protein kinase (JNK) in experimental glaucoma in rats. _Exp. Eye Res._ 82, 576–582 (2006). Article  CAS  PubMed  Google Scholar  * Levkovitch-Verbin, H. _et

al_. The transcription factor c-jun is activated in retinal ganglion cells in experimental rat glaucoma. _Exp. Eye Res._ 80, 663–670 (2005). Article  CAS  PubMed  Google Scholar  * Munemasa,

Y. _et al_. P-JNK expression of the rat retina in NMDA-induced neurotoxicity. _ARVO Meet. Abstr._ 45, 724 (2004). Google Scholar  * Roth, S. _et al_. Mitogen-Activated Protein Kinases and

Retinal Ischemia. _Invest. Ophthalmol. Vis. Sci._ 44, 5383–5395 (2003). Article  PubMed  Google Scholar  * Munemasa, Y. _et al_. Contribution of mitogen-activated protein kinases to

NMDA-induced neurotoxicity in the rat retina. _Brain Res._ 1044, 227–240 (2005). Article  CAS  PubMed  Google Scholar  * Munemasa, Y. _et al_. Pro-apoptotic role of c-Jun in NMDA-induced

neurotoxicity in the rat retina. _J. Neurosci. Res._ 83, 907–918 (2006). Article  CAS  PubMed  Google Scholar  * Bessero, A.-C., Chiodini, F., Rungger-Brändle, E., Bonny, C. & Clarke, P.

G. H. Role of the c-Jun N-terminal kinase pathway in retinal excitotoxicity, and neuroprotection by its inhibition. _J. Neurochem._ 113, 1307–1318 (2010). CAS  PubMed  Google Scholar  *

Awai, M. _et al_. NMDA-induced retinal injury is mediated by an endoplasmic reticulum stress-related protein, CHOP/GADD153. _J. Neurochem._ 96, 43–52 (2006). Article  CAS  PubMed  Google

Scholar  * Concannon, C. G. _et al_. NMDA receptor-mediated excitotoxic neuronal apoptosis _in vitro_ and _in vivo_ occurs in an ER stress and PUMA independent manner. _J. Neurochem._ 105,

891–903 (2008). Article  CAS  PubMed  Google Scholar  * Shimazawa, M. _et al_. Involvement of ER stress in retinal cell death. _Mol. Vis._ 13, 578–587 (2007). CAS  PubMed  PubMed Central 

Google Scholar  * Sokka, A.-L. _et al_. Endoplasmic reticulum stress inhibition protects against excitotoxic neuronal injury in the rat brain. _J. Neurosci. Off. J. Soc. Neurosci._ 27,

901–908 (2007). Article  CAS  Google Scholar  * Hu, Y. _et al_. Differential effects of unfolded protein response pathways on axon injury-induced death of retinal ganglion cells. _Neuron_

73, 445–452 (2012). Article  CAS  PubMed  PubMed Central  Google Scholar  * Luo, X., Baba, A., Matsuda, T. & Romano, C. Susceptibilities to and Mechanisms of Excitotoxic Cell Death of

Adult Mouse Inner Retinal Neurons in Dissociated Culture. _Invest. Ophthalmol. Vis. Sci._ 45, 4576–4582 (2004). Article  PubMed  Google Scholar  * Bai, N. _et al_. NMDA receptor subunits

have different roles in NMDA-induced neurotoxicity in the retina. _Mol. Brain_ 6, 34 (2013). Article  CAS  PubMed  PubMed Central  Google Scholar  * De Groef, L. _et al_. Decreased TNF

Levels and Improved Retinal Ganglion Cell Survival in MMP-2 Null Mice Suggest a Role for MMP-2 as TNF Sheddase. _Mediators Inflamm._ 2015, e108617 (2015). Article  Google Scholar  * Zhao, J.

_et al_. (+)-Pentazocine Reduces NMDA-Induced Murine Retinal Ganglion Cell Death Through a σR1-Dependent Mechanism. _Investig. Opthalmology Vis. Sci._ 57, 453 (2016). Article  CAS  Google

Scholar  * Danial, N. N. & Korsmeyer, S. J. Cell Death. _Cell_ 116, 205–219 (2004). Article  CAS  PubMed  Google Scholar  * Pozniak, C. D. _et al_. Dual leucine zipper kinase is required

for excitotoxicity-induced neuronal degeneration. _J. Exp. Med._ 210, 2553–2567 (2013). Article  CAS  PubMed  PubMed Central  Google Scholar  * Fernandes, K. A., Harder, J. M., Kim, J.

& Libby, R. T. JUN regulates early transcriptional responses to axonal injury in retinal ganglion cells. _Exp. Eye Res._ 112, 106–117 (2013). Article  CAS  PubMed  PubMed Central  Google

Scholar  * Fernandes, K. A., Harder, J. M., John, S. W., Shrager, P. & Libby, R. T. DLK-dependent signaling is important for somal but not axonal degeneration of retinal ganglion cells

following axonal injury. _Neurobiol. Dis._ 69, 108–116 (2014). Article  CAS  PubMed  PubMed Central  Google Scholar  * Liu, J. & Lin, A. Role of JNK activation in apoptosis: A

double-edged sword. _Cell Res._ 15, 36–42 (2005). Article  ADS  PubMed  Google Scholar  * Yang, D. D. _et al_. Absence of excitotoxicity-induced apoptosis in the hippocampus of mice lacking

the Jnk3 gene. _Nature_ 389, 865–70 (1997). Article  ADS  CAS  PubMed  Google Scholar  * Fernandes, K. A. _et al_. JNK2 and JNK3 are major regulators of axonal injury-induced retinal

ganglion cell death. _Neurobiol. Dis._ 46, 393–401 (2012). Article  CAS  PubMed  PubMed Central  Google Scholar  * Lingor, P., Koeberle, P., Kügler, S. & Bähr, M. Down-regulation of

apoptosis mediators by RNAi inhibits axotomy-induced retinal ganglion cell death _in vivo_. _Brain_ 128, 550–558 (2005). Article  PubMed  Google Scholar  * Behrens, A. _et al_. Impaired

postnatal hepatocyte proliferation and liver regeneration in mice lacking c‐jun in the liver. _EMBO J._ 21, 1782–1790 (2002). Article  CAS  PubMed  PubMed Central  Google Scholar  * Furuta,

Y., Lagutin, O., Hogan, B. L. & Oliver, G. C. Retina- and ventral forebrain-specific Cre recombinase activity in transgenic mice. _Genes. N. Y. N 2000_ 26, 130–132 (2000). CAS  Google

Scholar  * Hilberg, F., Aguzzi, A., Howells, N. & Wagner, E. F. c-Jun is essential for normal mouse development and hepatogenesis. _Nature_ 365, 179–181 (1993). Article  ADS  CAS  PubMed

  Google Scholar  * Johnson, R. S., van Lingen, B., Papaioannou, V. E. & Spiegelman, B. M. A null mutation at the c-jun locus causes embryonic lethality and retarded cell growth in

culture. _Genes Dev._ 7, 1309–1317 (1993). Article  CAS  PubMed  Google Scholar  * Syc-Mazurek, S. B., Fernandes, K. A., Wilson, M. P., Shrager, P. & Libby, R. T. Together JUN and DDIT3

(CHOP) control retinal ganglion cell death after axonal injury. _Mol. Neurodegener._ 12, 71 (2017). Article  PubMed  PubMed Central  Google Scholar  * Jarrard, L. E. Use of excitotoxins to

lesion the hippocampus: Update. _Hippocampus_ 12, 405–414 (2002). Article  PubMed  Google Scholar  * Schwarcz, R. & Coyle, J. T. Kainic acid: neurotoxic effects after intraocular

injection. _Invest. Ophthalmol. Vis. Sci._ 16, 141–148 (1977). CAS  PubMed  Google Scholar  * Zheng, X.-Y., Zhang, H.-L., Luo, Q. & Zhu, J. Kainic Acid-Induced Neurodegenerative Model:

Potentials and Limitations. _BioMed Res. Int._ 2011, e457079 (2010). Google Scholar  * Cazevieille, C. & Osborne, N. N. Retinal neurones containing kainate receptors are influenced by

exogenous kainate and ischaemia while neurones lacking these receptors are not–melatonin counteracts the effects of ischaemia and kainate. _Brain Res._ 755, 91–100 (1997). Article  CAS 

PubMed  Google Scholar  * Johnson, T. V. & Tomarev, S. I. Rodent models of glaucoma. _Brain Res. Bull._ 81, 349–358 (2010). Article  PubMed  Google Scholar  * Mali, R. S., Cheng, M.

& Chintala, S. K. Plasminogen Activators Promote Excitotoxicity-Induced Retinal Damage. _FASEB J. Off. Publ. Fed. Am. Soc. Exp. Biol._ 19, 1280 (2005). CAS  Google Scholar  * Spigolon,

G., Veronesi, C., Bonny, C. & Vercelli, A. c-Jun N-terminal kinase signaling pathway in excitotoxic cell death following kainic acid-induced status epilepticus. _Eur. J. Neurosci._ 31,

1261–1272 (2010). Article  PubMed  Google Scholar  * Ali, S. A. M., Hosaka, Y. Z. & Uehara, M. Expression of Small Leucine-Rich Proteoglycans in the Developing Retina and Kainic

Acid-Induced Retinopathy in ICR Mice. _J. Vet. Med. Sci._ 73, 439–445 (2011). Article  CAS  PubMed  Google Scholar  * Watkins, T. A. _et al_. DLK initiates a transcriptional program that

couples apoptotic and regenerative responses to axonal injury. _Proc. Natl. Acad. Sci. USA_ 110, 4039–4044 (2013). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Welsbie, D. S.

_et al_. Functional genomic screening identifies dual leucine zipper kinase as a key mediator of retinal ganglion cell death. _Proc. Natl. Acad. Sci. USA_ 110, 4045–4050 (2013). Article 

ADS  CAS  PubMed  PubMed Central  Google Scholar  * Koch, P., Gehringer, M. & Laufer, S. A. Inhibitors of c-Jun N-terminal kinases: an update. _J. Med. Chem._ 58, 72–95 (2015). Article 

CAS  PubMed  Google Scholar  * Schlamp, C. L. _et al_. Evaluation of the percentage of ganglion cells in the ganglion cell layer of the rodent retina. _Mol. Vis._ 19, 1387–1396 (2013). CAS 

PubMed  PubMed Central  Google Scholar  * Crocker, S. J. _et al_. c-Jun mediates axotomy-induced dopamine neuron death _in vivo_. _Proc. Natl. Acad. Sci. USA_ 98, 13385–13390 (2001). Article

  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Pirianov, G. _et al_. Deletion of the c-Jun N-terminal kinase 3 gene protects neonatal mice against cerebral hypoxic-ischaemic injury.

_J. Cereb. Blood Flow Metab. Off. J. Int. Soc. Cereb. Blood Flow Metab._ 27, 1022–1032 (2007). Article  CAS  Google Scholar  * Ries, V. _et al_. JNK2 and JNK3 combined are essential for

apoptosis in dopamine neurons of the substantia nigra, but are not required for axon degeneration. _J. Neurochem._ 107, 1578–1588 (2008). Article  CAS  PubMed  PubMed Central  Google Scholar

  * Behrens, A., Sibilia, M. & Wagner, E. F. Amino-terminal phosphorylation of c-Jun regulates stress-induced apoptosis and cellular proliferation. _Nat. Genet._ 21, 326–329 (1999).

Article  CAS  PubMed  Google Scholar  * Ikonomidou, C. & Turski, L. Why did NMDA receptor antagonists fail clinical trials for stroke and traumatic brain injury? _Lancet Neurol._ 1,

383–386 (2002). Article  CAS  PubMed  Google Scholar  * Miller, B. R. _et al_. A dual leucine kinase-dependent axon self-destruction program promotes Wallerian degeneration. _Nat. Neurosci._

12, 387–389 (2009). Article  CAS  PubMed  PubMed Central  Google Scholar  * Cui, Q., Yip, H. K., Zhao, R. C. H., So, K.-F. & Harvey, A. R. Intraocular elevation of cyclic AMP

potentiates ciliary neurotrophic factor-induced regeneration of adult rat retinal ganglion cell axons. _Mol. Cell. Neurosci._ 22, 49–61 (2003). Article  CAS  PubMed  Google Scholar  *

Harder, J. M. & Libby, R. T. BBC3 (PUMA) regulates developmental apoptosis but not axonal injury induced death in the retina. _Mol. Neurodegener._ 6, 50 (2011). Article  PubMed  PubMed

Central  Google Scholar  * Park, K. K. _et al_. Promoting axon regeneration in the adult CNS by modulation of the PTEN/mTOR pathway. _Science_ 322, 963–966 (2008). Article  ADS  CAS  PubMed

  PubMed Central  Google Scholar  Download references ACKNOWLEDGEMENTS The authors would like to thank Drs. DiAntonio (Dlkfl), Furuta (Six3-cre) and Wagner (Junfl) for generously providing

mice. This work was supported by EY018606 (RTL) and Research to Prevent Blindness, an unrestricted grant to the Department of Ophthalmology at the University of Rochester Medical Center. The

funding agencies had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript. AUTHOR INFORMATION AUTHORS AND AFFILIATIONS *

Department of Ophthalmology, University of Rochester Medical Center, Rochester, NY, 14642, USA Berkeley K. Fahrenthold, Kimberly A. Fernandes & Richard T. Libby * Neuroscience Graduate

Program, University of Rochester Medical Center, Rochester, NY, 14642, USA Berkeley K. Fahrenthold * Department of Biomedical Genetics, University of Rochester Medical Center, Rochester, NY,

14642, USA Richard T. Libby * The Center for Visual Sciences, University of Rochester Medical Center, Rochester, NY, 14642, USA Richard T. Libby Authors * Berkeley K. Fahrenthold View

author publications You can also search for this author inPubMed Google Scholar * Kimberly A. Fernandes View author publications You can also search for this author inPubMed Google Scholar *

Richard T. Libby View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS B.K.F., K.A.F. and R.T.L. designed research; B.K.F. and K.A.F. performed

research; B.K.F., K.A.F. and R.T.L. analyzed data; B.K.F. and R.T.L. wrote the paper. All authors have reviewed the manuscript. CORRESPONDING AUTHOR Correspondence to Richard T. Libby.

ETHICS DECLARATIONS COMPETING INTERESTS The authors declare no competing interests. ADDITIONAL INFORMATION PUBLISHER'S NOTE: Springer Nature remains neutral with regard to

jurisdictional claims in published maps and institutional affiliations. RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons Attribution 4.0 International

License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source,

provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons

license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by

statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Fahrenthold, B.K., Fernandes, K.A. & Libby, R.T. Assessment of intrinsic and

extrinsic signaling pathway in excitotoxic retinal ganglion cell death. _Sci Rep_ 8, 4641 (2018). https://doi.org/10.1038/s41598-018-22848-y Download citation * Received: 20 December 2017 *

Accepted: 01 March 2018 * Published: 15 March 2018 * DOI: https://doi.org/10.1038/s41598-018-22848-y SHARE THIS ARTICLE Anyone you share the following link with will be able to read this

content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative