Glaucoma is a complex disease with the ultimate cause of blindness being the death of RGCs. Axonal injury in the lamina cribrosa, a specialized structure that RGC axons pass through as they exit the eye, is thought to be the critical insult for RGCs8,9,10,11
. Axonal insult induced RGC somal degeneration is BAX dependent2,3,7
, but the molecules that control BAX activation are undefined. BH3-only proteins are pro-death Bcl-2 family members that help control BAX activation12
. A likely candidate for activating BAX after axonal injury is the BH3-only protein BIM. Using Bim
knockout mice we sought to determine if BIM was critical for BAX activation (RGC death) after axonal injury, including in a mouse glaucoma model.
Complete deficiency or heterozygosity for a null allele of Bax
prevents RGC death even after extensive time after optic nerve crush injury and in glaucoma3,7
. These data suggest that the levels of BAX, and subsequently activated BAX, are critical in determining RGC somal death. BAX activation is dependent on the level of BH3-only proteins, particularly those, like BIM, BID and BBC3, that can directly activate BAX36,37
. A single member, BBC3, is required for the normal developmental death of RGCs, however, Bbc3
deficiency only provided a minor delay in RGC death after axonal injury1
. BIM has been shown to be involved in neuronal death during development and after injury13,24,28,38
. In fact in a retinal explant model, where RGCs are injured by axotomy and likely other insults, Bim
deficiency completely prevented RGC death for up to 4 days in culture13
. This result implicates BIM as an important factor controlling RGC death after injury. Due to limitations of the explant model it can only assess the earliest time points of cell loss after axonal injury, which is just beginning in vivo
at 3 days and occurs over at least 3 weeks1,22
. In vivo
deficiency significantly decreased RGC death after axonal injury at 3 days. However, by 5 days after injury there was substantial RGC death, though this death was also significantly reduced compared to Bim+/+
mice. The loss of BIM did have a corresponding minor effect on RGC survival at later time points, but did not provide significant and complete protection as observed in Bax
deficient mice after extensive time. Thus, it appears that the absence of BIM mainly affects the rate of death after axonal injury, but not the ultimate survival of RGCs.
The fact that single deficiency in Bim
and to a far lesser extent Bbc31
only delay death suggests that other factors contribute to RGC death after axonal injury. The expression of BID, the other BH3-only protein that is capable of directly activating BAX, is consistent with a role in RGC death after axonal injury39
. However, we have found that Bid
deficiency does not delay or prevent RGC death after axonal injury (unpublished observation). It is unclear if these molecules work in combination or if indirect or non-canonical BAX activation can occur after axonal injury. Interestingly, we recently showed that the transcription factor JUN was a key mediator of RGC death after axonal injury27
. JUN activation appears to be upstream of BIM (). Jun
deficiency provided a far more extensive protection of RGCs than Bim
, suggesting that other downstream targets of JUN activation are important mediators of BAX activation after axonal injury. Identifying additional targets of JUN will help to determine if BAX activation is completely dependent on pro-death Bcl-2 family members or whether alternative pathways are involved.
DBA/2J mice develop elevated IOP subsequent to an iris disease that is similar to pigment dispersion syndrome in humans29,40,41
. It is possible that cell death may play a role in the iris disease or in the viability of trabecular meshwork cells (the cells primarily responsible for regulating IOP in the iridocorneal angle) in response to an insult. The significant lessening of IOP elevation (glaucomatous insult) in D2.Bim-/-
mice was similar to that observed in D2.Bax-/-
mice. Lessening IOP elevation is not seen in many of the other genetic or therapeutic manipulations that effect RGC neurodegeneration in DBA/2J mice (e.g.42,43,44,45,46,47
). In D2.Bax-/-
mice there were no changes in the clinical presentation of the iris disease, suggesting that Bax
deficiency was affecting IOP regulation and not the iris disease. Loss of trabecular meshwork cells has been linked to IOP elevation in humans48,49,50,51
. In humans, pigment dispersion syndrome only leads to pathological ocular hypertension (pigmentary glaucoma) and RGC loss in a subset of patients41
. Why some patients are susceptible to IOP elevation is unknown. It is possible that this susceptibility results from the variability of the pigmentary insult causing death of trabecular meshwork cells. Our data suggests that a component of IOP elevation in pigment dispersion patients involves a BIM-BAX dependent cell death process. It will be important to test the role of BIM directly in trabecular meshwork cells and determine if a BIM dependent pathway can induce trabecular meshwork cell death. Manipulating this pathway may be a method of preventing IOP elevation in pigment dispersion syndrome and other ocular hypertensive diseases.
The lessening of IOP elevation in D2.Bim-/-
mice may not completely explain the large protection from optic nerve degeneration conferred by Bim
deficiency. The IOP profile of D2.Bim-/-
mice was similar to that observed in D2.Bax-/-
but the protection from glaucomatous optic nerve degeneration was far greater in D2.Bim-/-
deficiency caused several abnormalities with retinal development that could contribute to the protection. Normal developmental retinal vasculature remodeling is disrupted in Bim
leading to a significant increase in the amount of vasculature. Recently, work has implicated the vasculature as important unit in neurodegenerative disease, including glaucoma45,53
. It is possible that extra retinal vasculature directly or indirectly alters the way the retina responds to IOP elevation. Finally, optic nerve morphogenesis was disrupted in Bim
deficient mice. There was a clear lack of arrangement of glia in the area of the lamina cribrosa and abnormalities in the retina-optic nerve border and D2.Bim-/-
mice. Interestingly, optic nerve morphology has been suggested to be an endophenotype for glaucoma54,55
. This is perhaps not surprising since the lamina cribrosa is thought to be a key structure in many aspects of glaucoma and it is certainly plausible that alterations in its morphology change RGC susceptibility to IOP elevation. Thus, there are several roles for BIM in ocular development that may alter susceptibility to ocular hypertension.
Bim deficiency significantly reduced the number of eyes that had severe glaucoma, as judged by optic nerve degeneration. However, 15% of D2.Bim-/- eyes developed severe optic nerve degeneration. These degenerated nerves allowed us to test whether BIM was required for RGC death in glaucoma. Unlike in Bax deficient mutants, RGCs were lost in D2.Bim-/- eyes with severe glaucomatous optic nerve degeneration. Even though BIM is expressed in glaucomatous RGCs and plays a role in axonal injury induced neuronal death, BIM is not required for BAX activation and RGC death in glaucoma.
It appears that BIM plays several roles in ocular development and disease that could directly affect glaucoma pathophysiology. During retinal development BIM is critical for normal retinal vasculature development17
and optic nerve head morphogenesis, both of which have been implicated as endophenotypes in glaucoma. Bim
deficiency also lessened/delayed IOP elevation in DBA/2J mice, suggesting BIM might regulate trabecular meshwork cell death after insult. Finally, Bim
deficiency delayed RGC death after axonal injury, but did not prevent RGC death after a glaucomatous insult. In the future it will be important to uniquely manipulate BIM expression in each of the tissues where BIM has a role in ocular physiology and pathophysiology to gain a better understanding of its role in ocular hypertension and glaucomatous neurodegeneration.