ER stress and unfolded protein response in ocular health and disease

Abstract

The human eye is the organ that is able to react to light in order to provide sharp three-dimensional and colored images. Unfortunately, the health of the eye can be impacted by various stimuli that can lead to vision loss, such as environmental changes, genetic mutations, or aging. Endoplasmic reticulum (ER) stress and unfolded protein response (UPR) signaling have been detected in many diverse ocular diseases, and chemical and genetic approaches to modulate ER stress and specific UPR regulatory molecules have shown beneficial effects in animal models of eye disease. This review highlights specific eye diseases associated with ER stress and UPR activity, based on a recent symposia exploring this theme.

Introduction

The endoplasmic reticulum (ER) is found in every eukaryotic cell; it is a double-membrane bound organelle facilitating a wide variety of functions, such as protein folding, free calcium storage, and lipid/sterol synthesis. ER homeostasis can be compromised by many pathological events, including protein misfolding or accumulation of mutant protein, which can lead to ER stress [1]. The ER has its own regulatory mechanism, called the unfolded protein response (UPR), that is activated in an attempt to restore ER homeostasis and to prevent further damage to the cell [2]. In mammalian cells, the UPR is regulated by inositol-requiring enzyme 1 (IRE1), PKR-like ER kinase (PERK), and activating transcription factor 6 (ATF6) (Fig. 1). Under unstressed conditions, activation of IRE1, PERK, and ATF6 is prevented by the binding of the ER-resident protein, binding immunoglobulin protein (BiP) also known as 78-kDa glucose-regulated protein (GRP78) or heat shock 70-kDa protein 5 (HSPA5) gene, to their N-terminal domains. Once ER stress occurs BiP dissociates and activates IRE1, PERK, and ATF6 [3-6]. These three mediators activate a comprehensive transcriptional and translational signaling program; on one side, protein translation is slowed down, but on the other side proteins are being upregulated that belong to the folding machinery of the ER and the endoplasmic-reticulum-associated degradation (ERAD) to increase degradation events [2].

Fig. 1.

Unfolded protein response signaling pathways. Schematic Illustration of the three unfolded protein response signaling pathways, comprised of inositol-requiring enzyme 1 (IRE1), PKR-like ER kinase (PERK), and activating transcription factor 6 (ATF6). Under unstressed conditions, the activity of these three UPR inducers is prevented by BiP binding. Ones ER stress occurs BiP dissociates and activates IRE1, PERK and ATF6. IRE1’s luminal domain is coupled to the cytosolic kinase (K) and endoribonuclease (R) domains. ER stress leads to IRE1 oligomerization, transautophosphorylation, and the unconventional splicing of X-box binding protein-1 (usXBP1) mRNA to generate spliced XBP1 (sXBP1). The PERK protein bears a luminal domain that is linked to a kinase domain. PERK activation causes its dimerization and phosphorylation of the eukaryotic translation initiation factor 2 subunit alpha (eIF2α), leading to attenuation of global protein translation. ATF6 migrates as a monomer from the ER to the Golgi compartment upon activation. Two independent proteases (S1P and S2P) cleave ATF6 to generate a potent transcription factor that contains a bZip domain to bind to DNA to upregulate its downstream targets.

IRE1 is a key regulator of the UPR [2]. As an ER-resident transmembrane protein, IRE1 has a luminal domain, which is connected to its cytosolic kinase and endoribonuclease domain [Fig. 1, (K) and (R), respectively] [7-9]. In response to ER stress, IRE1 oligomerizes, resulting in the activation of its kinase domain leaving IRE1 autophosphorylated that ultimately leads to the activation of its RNase domain and the removal of an inhibitory intron from the mRNA of unspliced XBP-1 (usXBP1), generating the potent transcription factor spliced XBP1 (sXBP1) (Fig. 1) [10,11]. Transcriptional target genes of sXBP1 include several genes of the ERAD pathway [12,13].

The PERK branch of the UPR is responsible for a different response to ER stress [2]. In the presence of ER stress, PERK oligomerizes, which activates its cytosolic kinase domain leading to the inhibition of the ternary translation initiation complex (Fig. 1K) [14]. Inhibition of global protein translation is facilitated by PERK kinase activity by phosphorylating elongation initiation factor 2 alpha (eIF2α), which is an essential step to reduce the load of newly synthesized peptides in the ER and to alleviate ER stress to restore ER homeostasis [14,15]. Under prolonged ER stress, PERK signaling regulates a cell death program by upregulating proapoptotic transcriptional activators, activating transcription factor 4 (ATF4), and CCAAT/enhancer-binding protein homologous protein (CHOP) (Fig. 1) [16,17].

ATF6 is a N-linked glycosylated type 2 transmembrane protein able to exist as a monomer and oligomer, and also as heteromeric complexes with other ER proteins through its ability to form intra- and intermolecular disulfide bonds (Fig. 1) [18,19]. In response to ER stress, ATF6 is reduced to its monomeric confirmation and translocates from the ER to the Golgi compartment using COP II vesicles where site 1 (S1P) and site 2 (S2P) proteases cleave ATF6 in the transmembrane domain resulting in the release of ATF6’s cytosolic domain [20-22]. ATF6 transcriptional activity arises from its cleaved cytosolic domain, which contains a basic leucine zipper (bZIP)-domain that locates to the nucleus to upregulate ATF6 target genes, such as GRP78/BiP [11,13].

This review will provide an overview on what currently is known on the involvement of UPR signaling pathways in various eye diseases.

Chapter 1: Diabetic retinopathy – interference of UPR signaling events as potential treatment and prevention of DR disease pathologies

Diabetic eye disease, also known as diabetic retinopathy (DR) is a medical condition caused by damage to the retina so severe that it is described as the leading cause of blindness. DR is defined as a common vascular complication of diabetes with hallmarks of vascular dysfunction, neovascularization, increased permeability, inflammation, and vascular degeneration.

Yan et al. [23], have shown in one of their previous studies that P58^IPK/DNAJC3, an ER stress-related protein, is able to bind to PERK and therefore prevents the phosphorylation of the α-subunit of its downstream target eukaryotic translation initiation factor 2 (eIF2α). Several studies since have demonstrated that P58^IPK plays a protective role during ER stress [23,24]. Using human retinal capillary endothelial cells (HRCECs), Li and colleagues investigated the relationship between ER stress and P58^IPK [25]. It was demonstrated that when endogenous expression of P58^IPK was increased by transfections, expression of ATF4, and vascular endothelial growth factor (VEGF) was decreased in HRCECs when ER stress was induced. It is suggested that P58^IPK is able to down-regulate the activity of the ATF4 and VEGF pathways during ER stress. It was additionally shown that increased expression of P58^IPK resulted in decreased levels of CHOP in HRCECs, associated with the decline of ER stress-related apoptotic events. Furthermore, it is suggested that enhanced P58^IPK expression may reduce retinal blood vessel damage, since VEGF expression was also reduced in the presence of elevated P58^IPK expression. P58^IPK is therefore thought to be responsible for the inhibition of ER stress in endothelial cells of retinal blood vessels [25].

Vascular endothelial growth factor is a dimeric glycoprotein and a potent endothelial mitogen, and is thought to promote angiogenic growth of new blood vessels by regulating the proliferation, migration, and tubule formation of endothelial cells [26-28]. VEGF is a key player in pathology of retinal diseases, such as age-related macular degeneration (AMD), retinopathy of prematurity (ROP), sickle cell retinopathy (SCR), and retinal vascular occlusion (RVO) (reviewed in [29]). VEGF also plays a key role in DR pathologies [30,31], and its expression is regulated by UPR signaling [32].

One of the first studies that demonstrated a link between ER stress and the upregulation of intracellular VEGF expression as well as VEGF secretion was performed by Abcouwer and colleagues in 2002 [33]. Using the human retinal pigment epithelial cell line (ARPE-19) and a wide variety of ways to induce ER stress, such as amino acid deprivation and glucose deprivation (tunicamycin, brefeldin A, calcium ionophore A23187, thapsigargin), Abcouwer et. al. identified a significant increase in VEGF mRNA levels of 1.3- to 6-fold and 8- to 10-fold, respectively. The increased presence of VEGF mRNA levels was correlated with an increase in GRP78 mRNA levels under the same conditions [33]. In a different study, Li et. al. reported in an in vitro study that ER stress is increased in Akita mouse retinas, which is a genetic model of type 1 diabetes [34]. Using intravitreal injections of tunicamycin to induce ER stress in the retina, increased retinal VEGF expression was observed [34]. Additionally, it was reported that exposure of ER stress inducers, tunicamycin and thapsigargin, to vascular endothelial cells as well as Müller cells resulted in significant upregulation of VEGF expression [34-36]. Li and colleagues have further shown that ER stress induction leads to a significant induction of inflammatory molecules in diabetic retinas. Interestingly, when performing ER stress inhibitory experiments, amelioration of inflammation was observed in cultured human retinal endothelial cells that experience hypoxia – the findings that were confirmed using a diabetic animal model [34]. These results are also supported in additional animal models of type 1 and type 2 diabetes [37-40].

The activation of the transcription factor ATF4 through the PERK branch of the UPR on ER stress or high-glucose levels in retinal Müller cells (Fig. 2) has been shown to be essential for the induction of hypoxia-inducible factor 1-alpha (HIF-1α) and VEGF secretion [41]. Furthermore, genetic inactivation of ATF4 greatly impacted diabetic-induced retinal inflammation and vascular leakage attenuation [35], supporting a role for ATF4 in retinal inflammation and endothelial barrier dysfunction.

Fig. 2.

Schematic representation of the human eye and retina. (A) Anatomy of the human eye with (B) detailing the structure and cell components of the human retina.

A strong correlation between the distinct activation of UPR signaling events and the duration of diabetes has been reported. This fact is strongly supported by a research study performed by Yan and colleagues in 2012 [40]. A diabetic rat model was established by intraperitoneal injections of streptozotocin (STZ), which was used to analyze 89 ER stress-related factors by quantitative RT-PCR [40]. All 89 ER stress genes were expressed at both first and third month after development of diabetes, but significant changes were only detected in the expression of 13 genes and 12 genes in the first and third month after development of diabetes, respectively, when compared to nondiabetic animals [40]. Only two genes were identified that showed significant changes at both time point, HERP1 (homocysteine-responsive endoplasmic reticulum-resident ubiquitin-like domain member 1 protein) and ERdj4 (DNAJB9, DnaJ heat shock protein family (Hsp40) member B9) [40].

Research performed to date widely suggests the differential involvement of UPR signaling in retinal cells and tissues using multiple diabetes models [42-44]. It appears that all branches of the UPR are activated to some degree in diabetic retinas [34,35,37-39,41]. However, there are variations between the studies about the activation status of the individual UPR pathways, which could be due to the choice of retinal and diabetic animal model, severity of diabetes, diabetes development, time point of study, and methods of UPR pathway analysis.

Chapter 2: Affecting UPR signaling events as treatment of glaucoma

Glaucoma is a group of optic neuropathies and the second leading cause of irreversible blindness in the world affecting approximately 70 million people [45,46]. It is mainly characterized by the degeneration of retinal ganglion cell (RGC) axons resulting in vision loss (Fig. 2). Primary open-angle glaucoma (POAG) is the most common form of glaucoma and accounts approximately 70% of all cases [47]. POAG is usually accompanied by an elevated intraocular pressure (IOP), one of the major risk factors. Elevated IOP occurs when the outflow resistance of the aqueous humor, which is produced by the ciliary bodies and normally empties through the trabecular meshwork (TM), is increased and the liquid is no longer able to drain through the TM structures. The mechanisms that cause increased outflow resistance in the TM are not well understood yet.

A growing number of studies implicate ER stress in the TM as a cause for increase in IOP, and that reduction of ER stress can prevent ocular pathologies in glaucoma mouse models [48,49]. In human glaucomatous TM cells, elevated levels of ER stress-induced proteins, including GRP78, GRP94, and CHOP have been detected [50]. These results support the findings obtained in mouse models and point to chronic ER stress in human glaucomatous TM. Interestingly, little Xbp1 splicing or phosphorylated eIF2a were seen, which suggest dysregulated UPR signaling in glaucomatous TM cells compared to wild-type cells [50]. By contrast, genetic studies reveal that CHOP induction may play causal roles in glaucoma damage. Zode and colleagues demonstrated that deletion of CHOP protects against ER stress-mediated IOP elevation in glucocorticoid-induced ocular hypertension mouse models, and Hu and colleagues pointed out that the deletion of CHOP promotes RGC survival in mouse models of optic nerve crush and ocular hypertension [42].

Mutations in the myocilin gene (MYOC) are known to be the most common genetic cause of POAG and autosomal dominant juvenile-onset open-angle glaucoma [47,51-54]. Myocilin is expressed in many ocular tissues including TM cells, but its function is poorly understood. Recent studies on a transgenic glaucoma mouse model (Tg-MYOCY^437H) expressing the human mutated form of myocilin found increase in IOP in these mice already at an age of 3 months [48]. Zode and colleagues also showed that the topical administration of 4-phenylbutyrate (PBA) on the cornea significantly reduces the elevated IOP levels in those mice. Animals treated with PBA preserved pattern electroretinogram (PERG) amplitudes similar to control littermates and showed improved myocilin secretion as well as reduced ER stress in TM cells [48]. These results suggest that PBA, a chemical chaperone, enhances the folding of mutant myocilin and its secretion into the aqueous humor thereby preventing its accumulation in the TM and the evocation of ER stress. Thus, topical application of PBA eye drops could serve as a new therapy option for patients with myocilin-associated POAG.

It would provide great evidence if additional data could be provided to demonstrate that POAG is strongly associated with myocilin gene mutations, mice research studies could demonstrate a connection between POAG and mutation in UPR genes. Increasing ER stress due to potential dysfunction and/or upregulation of different UPR components seems to be another reason of IOP and for the development of POAG. Hence, to study the genetic background of this disease in more detail might reveal deeper understanding in the development of one of the world′s leading causes of blindness.

Chapter 3: Involvement of ER stress in retinal degeneration

Inherited retinal disorders (IRDs) form a genetically and clinically very heterogeneous group of diseases leading to severe visual impairments. One well-known IRD is retinitis pigmentosa (RP). RP refers to a very large number of genetically and phenotypically heterogeneous IRDs affecting more than one million people worldwide [55,56]. RP is primarily characterized by progressive rod photoreceptor degeneration, night blindness, and progressive constriction of the visual field, also known as tunnel vision. Secondarily, cone photoreceptors die resulting in a complete vision loss in later stages of the disease. Another well-studied IRD is Lebers congenital amaurosis (LCA). It is the most severe form of IRD, with congenital or early childhood blindness. Symptoms of LCA include nystagmus, impaired or absent papillary response. The protein that is linked to the development of LCA is tubby-like protein 1 (TULP1), which is a protein coding gene. Proteins of the TULP family share a conserved C-terminal domain of approximately 200 amino acid residues. TULP1 is also known as a photoreceptor specific protein that is involved in the transport of visual cascade proteins as rhodopsin, which is synthesized in the photoreceptor inner segment (IS) and transported through the connected cilium into the outer segment (OS) [57-60] (Fig. 2). TULP1 has been localized to the IS of the photoreceptor cell, closely to other parts of the biosynthetic machinery as the ER or the Golgi apparatus. Here, it functions as a chaperone or an adaptor protein linking vesicles with their proper motor proteins for effective cargo transport to the membrane [57-59]. Previous studies have shown that mutations in TULP1, among mutations in other genes, are the underlying cause of an early-onset form of autosomal recessive RP (arRP) and LCA [57-59]. Many missense, splice site, and stop mutations have been identified in TULP1 [61,62]. Lobo and colleagues demonstrated that missense mutations in TULP1 produce misfolded TULP1 protein accumulating in the ER and causing ER stress [63]. Immunohistochemical analyses show that wild-type TULP1 can be mainly found in the plasma membrane and cell processes of immortalized RPE cells, whereas mutant TULP1 predominantly accumulates in a spot-like pattern in the cytoplasm resembling that of the ER. Elevated levels of different ER markers, such as BiP, PERK, and XBP1 or XBP1s, could be detected in cells transfected with the mutant TULP1 in comparison to wild-type TULP1 [63]. Furthermore, overexpression of mutant TULP1 causes higher cell death rates as well as elevated CHOP levels compared to controls [63]. In addition, in vivo studies showed that mutant TULP1 accumulates in the ER of the mouse retina. These results suggest that various signaling components of the UPR become activated in retinal cells expressing mutant TULP1, promoting cell death, and may underlie the photoreceptor degeneration in TULP1-associated RP. Hence, mutated TULP1 is a prototype par excellence that misfolded proteins are retained in the ER of photoreceptor cells instead of being transported into the OS leading to ER stress and cell death. In this context, another well-known example is the misfolding of rhodopsin protein. Mice and rats carrying the P23H mutation are wildly used as RP animal models, expressing misfolded and nonfunctional rhodopsin, which similar to mutated TULP1 is retained in the ER and can no longer be transported into the OS.

Photoreceptor cells in the retina are constantly producing soluble and membrane proteins, for example, components of phototransduction machinery, that are destined to the cell surface or the OS disks (Fig. 2). Many of these proteins are synthesized and folded in the ER. If the retina is exposed to conditions that cause disturbance in ER homeostasis, it could induce ER stress in the photoreceptor cells and lead to retinal degeneration. These conditions include, but not limited to, expressing mutant proteins in the retina, exposing the eyes to bright light, and aging [64-69]. Studies performed over the past decade have implicated that ER stress is involved in retinal degenerative diseases [64,66-68,70]. Using the ERAI mouse GFP reporter line, it has been demonstrated that UPR is activated in mice photoreceptors expressing T17M or P23H mutant rhodopsin providing evidence for ER stress in these photoreceptors [64,65,67]. Since UPR signaling can restore ER homeostasis (Fig. 1), targeting UPR signaling may enhance retinal health or delay retinal degeneration.

In cell culture studies, overexpression of the BiP/Grp78 ER chaperone in cells expressing mutant P23H rhodopsin alleviates ER stress and reduces apoptosis [71]. Gorbatyuk et al. also demonstrated that subretinal gene delivery of BiP/Grp78 through adeno-associated virus type 5 (AAV5) to transgenic rats expressing P23H rhodopsin reduced the expression of a proapoptotic gene, Chop, and partially restored retinal function in the P23H transgenic rats [71]. Recently, it has been shown that overexpressing ATF4 accelerates retinal degeneration in mice expressing T17M mutant rhodopsin [72]. However, knocking down ATF4 in mice expressing T17M mutant rhodopsin prevented the T17M mice from losing their retinal functions and retinal integrity [72]. Interestingly, when Chop was ablated from the mice expressing either T17M or P23H rhodopsin, the retinal function was not rescued and the mice still underwent severe retinal degeneration, suggesting that Chop was dispensable in retinal degeneration of these mouse models [64,73-75]. Genetic and chemical targeting of other UPR components and signaling pathways may reveal new ways to alleviate ER stress in retinal degenerative diseases.

Chapter 4: ATF6α, UPR mediator, and inherited retinal dystrophies gene

As described in chapter 3, IRDs constitute a large group of genetically inherited retinal degenerative disorders and can be classified as to whether they cause degeneration of rod and/or cone photoreceptor cells (Fig. 2). To date, approximately 250 inherited retinal dystrophy genes have been identified (Retinal Information Network http://www.sph.uth.tmc.edu/RetNet/). Achromatopsia is characterized by cone photoreceptor dysfunction and eventual degeneration of cone photoreceptors, whereas cone-rod dystrophy (CRD) is characterized by progressive loss of cone function; followed by loss of rod photoreceptor function (Fig. 2).

Using next-generation whole-exome sequencing, we recently discovered autosomal recessive mutations presented in the activating transcription factor 6 α (ATF6α) gene in patients with achromatopsia and CRD [76-78]. ATF6α is a glycosylated transmembrane protein found in the ER known for its function in the ATF6-linked UPR pathway [19] (Fig. 1/3). In response to ER stress, ATF6α is transported from the ER to the Golgi apparatus where site 1 and site 2 proteases cleave ATF6α to release the cytosolic domain of ATF6α, which is a basic leucine zipper (bZIP) transcription factor [19]. The cleaved cytosolic domain of ATF6α then locates to the nucleus where it binds to DNA and transcriptionally upregulates downstream target genes, such as ER protein folding chaperones and enzymes [18,19,79,80] (Fig. 1/3).

The identified ATF6α mutations span the entire coding region of ATF6α gene. The changes include missense, nonsense, splice-site, and small frame-shifting deletions, insertions and duplications mutations. These mutations were classified into three different functional classes based on the effect and the location of the mutation [77] (Fig. 3). Class 1 ATF6α mutations code for mutant ATF6α that exhibit impaired trafficking from the ER to the Golgi apparatus and, therefore, demonstrate reduced proteolysis to release the cytosolic domain under ER stress conditions [77]. Class 2 ATF6α mutations code for truncated ATF6α containing fully functional bZIP and transcription activation domains. These truncated ATF6α mutants are able to constitutively activate downstream targets [77]. Class 3 ATF6α mutations code for ATF6α mutants that have absent or defective bZIP domain, resulting in the loss of transcriptional activity [77].

Fig. 3.

Classification of ATF6 mutations. Full length ATF6 can be present as a monomer, dimer, or oligomer using disulfide bond formations, once UPR signaling is activated BiP dissociates, and reduced monomeric ATF6 traffics from the ER to the Golgi compartment (see Fig. 1). Class 1 ATF6 mutations are trafficking mutations that show an impairment of translocation from the ER to the Golgi apparatus. Class 2 mutations of ATF6 present a fully intact cytosolic domain of ATF6 and show constitutive transcriptional activator function. Class 3 ATF6 mutations demonstrate a nonfunctional bZip domain and fail to bind and upregulate ATF6 specific targets.

The identification of ATF6α as a novel retinal dystrophy gene reveals that ATF6α has a crucial role in human retinal (foveal) development and cone photoreceptor function (Fig. 2). Intriguingly, despite ATF6α ubiquitous expression, mutations in ATF6α solely manifest in a cone photoreceptor phenotype in human

Chapter 5: Functional role of ER stress during cellular senescence and pathological angiogenesis

Senescence is a cellular state that cells adopt upon exposure to DNA damage, telomere shortening, and many other cellular stresses. Increased numbers of senescence cells are a feature of the normal aging process that results in aggravation of tissue function [81-83]. Cellular senescence is also a common feature in complex physiological processes, such as embryogenesis and tissue repair [84-87]. Biochemical and/or inflammatory stressors can compromise the function of cells in our nervous system, such as breakdown of the vascular system or a part thereof as witnessed in vascular retinopathies (DR and ROP) [88]. The retinal neurons directly exposed to the degenerating vasculature in ischemic retinopathies are retinal ganglion cells (RGC) (Fig. 2). RGCs require a stable metabolic supply for proper function. One of the central questions in ischemic retinopathies is how can retinal ganglion cells escape ischemia-induced cell death. A recent research study from Oubaha et al. [89] have provided insight into why retinal neurons survive in ischemic retinopathies despite compromised metabolic supply and explored the function of cellular senescence. When a cell undergoes senescence, it acquires a permanent state of cell cycle arrest, in which cells remain viable [90]. Oubaha et al. [89] found that retinal ischemia can trigger cellular senescence to protect retinal cells, which enables retinal cells to survive periods of limited metabolic supply and to escape hypoxia-associated cell death in a mouse model of ischemic retinopathy. Interestingly, using transcriptome analysis combined with loss-of-functions genetics, Oubaha’s results indicated that those cells require the activity of the endoribonuclease domain of IRElα to promote senescence and adopt a senescence-associated secretory Phenotype (SASP) [89]. The presence of SASP-associated cytokines, such as plasminogen activator inhibitor-1 (PAI-1), interleukin-6 (IL-6), interleukin-8 (IL-8), and VEGF were found to be present in the vitreous humor of patients suffering from proliferative DR. Furthermore, they identified a clinical relevance of SASP inhibition using metformin, a widely used biguanide antidiabetic drug known to reduce SASP without interfering with the growth arrest program [91]. They demonstrated that intravitreal injection of metformin into mice abolished SASP, decreased ER stress, and reduced retinal neovascularization and pathological retinal angiogenesis [89].

Additional research performed by Binet and colleagues demonstrates that failure of the reparative angiogenesis system in the ischemic retina is temporally and spatially linked to ER stress [92]. PERK and IRE1α activation (Fig. 1) were found in hypoxic/ischemic retinal ganglion cells and linked to the activation of signaling cascades of the angiostatic response in those regions [92]. Authors further identified that the endoribonuclease domain of IRE1α degrades netrin-1, which normally play an essential role in myeloid-dependent revascularization. Inhibition of IRE1α or treatment with netrin-1 showed enhanced retinal vascular regeneration leading to a relief of the hypoxic stimulus and prevention of destructive neovascularization [92]. The degradation of netrin-1 through IRE1α activation by persisting neuronal ER stress could underlie the vascular regeneration and direct regulation of reparative angiogenesis seen in ischemic retinas.

Chapter 6: UPR signaling in retinal ganglion cells

RGCs play an essential role in the processing of visual information, which they receive from the photoreceptors that passes on the information to bipolar and retina amacrine cells before it reaches the RGCs (Fig. 2). The axons of RGCs form the optic nerve; they are the only type of retinal neuron that is able to directly pass visual information to the brain, which makes them extremely vulnerable to axon damage, resulting in vision loss due to optic neuropathy [93]. Traumatic optic nerve injury followed by RGC loss is a common feature of head injuries caused by falls or traffic accidents. It has been shown in rodents that the majority of RGCs undergo cell death at approximately 2 weeks after intraorbital optic nerve injury [94-96].RGC loss as hallmark of optic neuropathy, other than optic nerve trauma, is also seen in glaucoma [97-102]. Apoptosis has been described as one of the leading causes of RGC loss in these diseases [97-101,103]. In the search for therapeutic targets, it becomes apparent that targeting apoptotic effectors may not be a valuable approach since RGC death most likely occurs at the end stage of the disease, which suggest that identifying upstream effectors (before RGCs are terminally committed to apoptosis) is more fruitful. In 2012, Hu and colleagues employed an optic nerve crush model known to result in loss of the majority of RGC [104]. They identified that distinct pathways of the ER stress response are activated in axotomized RGCs in this optic nerve crush model [104]. Furthermore, they found that optic nerve injury causes transcriptional upregulation of CHOP. CHOP knockout mice showed dramatically increased RGC survival in their optic nerve crush model [104]. Also, overexpression of spliced XBP1 using adeno-associated virus in CHOP KO mice further enhanced RGC survival after optic nerve injury [104]. These studies suggest that UPR signaling pathways may be attractive therapeutic targets to promote RGC survival in optic neuropathies.

Conclusions and future directions

Endoplasmic reticulum stress and the activation of the UPR signaling response are a well-known feature in many neurodegenerative diseases; under various physiological conditions, ER stress can induce apoptotic cell death leading to neuronal cell death in the brain and retina [69,105-112]. Targeting UPR signaling pathways at various stages is thought to create an opportunity to support cell survival events, prevent neuronal cell loss, and hence create treatment for neuronal diseases.

Patients suffering from diabetes have a profound chance to develop DR, which is the leading cause of blindness caused by the loss of neural and vascular retinal cells. Yan and colleagues have successful demonstrated that inhibition of PERK, prevents the phosphorylation of its downstream target eIF2α resulting in the inhibition of apoptotic cell death in endothelial cells of retinal blood vessels [23,25]. It is further thought that the downregulation of VEGF would take preventive measures in the development and advancement of DR pathologies [25]. These findings are promising but as the authors stated further clarification is needed to precisely determine how P58^IPK correlates with the VEGF signaling pathway under ER stress conditions. Additional in vivo models are required to support their research findings. It would also be valuable to gain insights on how P58^IPK is regulated during DR pathologies and how it could be addressed as direct target for potential treatments in the prevention of neuronal cell death.

One of the biggest challenges in the development of DR treatments is that patient that suffer from diabetes go through different disease stages, which makes it more challenging to identify a protein that can be directly target for disease treatments or prevention. This complex problem was addressed in a study from Yan et al. in 2012 [40]. Researchers studied the correlation between the activation of certain ER stress genes and the duration of diabetes. A total of 89 genes were analyzed at two time points of 1- and 3-month old diabetic mice. Of all genes analyzed, they identified that various ER stress genes are upregulated at the two different time points. However, only two of them were consistently elevated at both time point, named HERP1 and ERDj4 [40]. The very important message from this valuable study is that there is clearly a change in the expression profile of ER stress genes and UPR signaling events throughout disease development and progression. To develop effective targets to treat DR, researchers have to gain continuous insights on gene expression profiles throughout disease progression to be able to develop a therapy that can directly target an ER stress event that has the biggest potential as DR treatment.

Animal models are invaluable tools to study retinal diseases, but in some instances, other research models are required to gain a deeper understanding on how eye developmental is linked to retinal disease. Mutations in ATF6α have been linked to the development of achromatospia [76-78]. Recently, a new role for ATF6α signaling in early stem cell differentiation was identified [113]. It was shown that the activation of ATF6α during stem cell differentiation resulted in the acceleration of differentiation events toward the mesodermal lineage generating functional endothelial cells that were able to undergo in vitro angiogenesis to form blood vessels; ATF6α additionally stimulated the loss of pluripotency [113]. In the same study, ATF6α was found to support the growth and maturation of the ER without activating a comprehensive ER stress responds. It is now questioned if impairment or lack of functional ATF6α could cause the development of immature ER in cone photoreceptor cells, which could partly contribute to disease pathologies of achromatopsia. The maturation and survival of cone photoreceptor cells require the production and proper folding of cone OS proteins in the ER, lack of fully developed ER, and fully functional ATF6α may lead to the production of misfolded and nonfunctional cone proteins resulting in the death of cone photoreceptors or the under development of cone photoreceptor cells. Therefore, investigating the development of ER in cone photoreceptor cells carrying ATF6α mutation will provide more insight in how impaired ATF6α function impacts cone photoreceptor maturation.

In summary, the activation of UPR signaling events during the progression of retinal degeneration is well-known; however, it remains unclear on how exactly the ER stress response or UPR-related genes contribute to either photoreceptor cell death or the development of cone and rod cells that results in the development of retinal diseases. UPR signaling is a highly regulated cellular program that is initiated to provide a protective role, but can turn toward cell death via apoptosis in a supportive effort to protect connective tissue. The timing of IRE1, PERK, and ATF6 activation and deactivation is very sensitive and depends drastically on the severity and the extent of ER stress and its causes. It is therefore essential to gain insights on how UPR signaling mediates the fate of accumulated misfolded proteins in photoreceptor; and additionally, how the UPR regulates the switch from a prosurvival to an apoptotic response. Gaining this knowledge will be a valuable tool to address potential targets for disease treatment and/or prevention in more precise fashion.

Abbreviations

AAV5

adeno-associated virus type 5

AMD

age-related macular degeneration

ARPE-19

human retinal pigment epithelial cell line

arRP

autosomal recessive RP

ATF4

activating transcription factor 4

ATF6

activating transcription factor 6

BiP

binding immunoglobulin protein

bZIP

basic leucine zipper

CHOP

CCAAT/enhancer-binding protein homologous protein

DR

diabetic retinopathy

eIF2α

elongation initiation factor 2 alpha

ERAD

endoplasmic-reticulum-associated degradation

ERdj4

DNAJB9, DnaJ heat shock protein family (Hsp40) member B9

ER

endoplasmic reticulum

GRP78

78-kDa glucose-regulated protein

HERP1

homocysteine-responsive endoplasmic reticulum-resident ubiquitin-like domain member 1 protein

HIF-1α

hypoxia-inducible factor 1-alpha

HRCECs

human retinal capillary endothelial cells

HSPA5

heat shock 70-kDa protein 5

IL-6

interleukin-6

IL-8

interleukin-8

IOP

intraocular pressure

IRDs

inherited retinal disorders

IRE1

inositol-requiring enzyme 1

IS

inner segment

LCA

Lebers congenital amaurosis

MYOC

myocilin gene

OS

outer segment

PAI-1

plasminogen activator inhibitor 1

PBA

4-phenylbutyrate

PERG

pattern electroretinogram

PERK

PKR-like ER kinase

POAG

primary open-angle glaucoma

RGC

retinal ganglion cell

ROP

retinopathy of prematurity

RP

retinitis pigmentosa

RVO

retinal vascular occlusion

S1P

site 1 proteases

S2P

site 2 proteases

SASP

senescence-associated secretory phenotype

SCR

sickle cell retinopathy

STZ

streptozotocin

sXBP1

spliced XBP1

TM

trabecular meshwork

TULP1

tubby-like protein 1

UPR

unfolded protein response

usXBP1

X-box binding protein-1

VEGF

vascular endothelial growth factor