Novel Neuroprotective Function of Apical-Basal Polarity Gene Crumbs in Amyloid Beta 42 (Aβ42) Mediated Neurodegeneration

Alzheimer's disease (AD, OMIM: 104300), a progressive neurodegenerative disorder with no cure to date, is caused by the generation of amyloid-beta-42 (Aβ42) aggregates that trigger neuronal cell death by unknown mechanism(s). We have developed a transgenic Drosophila eye model where misexpression of human Aβ42 results in AD-like neuropathology in the neural retina. We have identified an apical-basal polarity gene crumbs (crb) as a genetic modifier of Aβ42-mediated-neuropathology. Misexpression of Aβ42 caused upregulation of Crb expression, whereas downregulation of Crb either by RNAi or null allele approach rescued the Aβ42-mediated-neurodegeneration. Co-expression of full length Crb with Aβ42 increased severity of Aβ42-mediated-neurodegeneration, due to three fold induction of cell death in comparison to the wild type. Higher Crb levels affect axonal targeting from the retina to the brain. The structure function analysis identified intracellular domain of Crb to be required for Aβ42-mediated-neurodegeneration. We demonstrate a novel neuroprotective role of Crb in Aβ42-mediated-neurodegeneration.


Background
Alzheimer's disease (AD) is a progressive neurodegenerative disorder with no effective cure to date. AD is characterized by the progressive loss of neurons in the hippocampus and cortex causing decline in cognitive and behavioral functions eventually leading to the death of the patient [1,2]. AD neuropathology is associated with two types of abnormal protein deposition in the human brain viz.: (1) neurofibrillary tangles (NFTs) containing hyperphosphorylated forms of a microtubule associated protein Tau, and (2) the accumulation of the amyloid-beta (Ab42) peptide [1][2][3][4][5][6][7][8]. Ab42 is generated by improper (b-and c-) cleavage of the transmembrane receptor amyloid precursor protein (APP), as well as by mutations linked to familial AD that affect APP processing [9]. The abnormal cleavage of APP causes the protein to be 42 amino acids long (Ab42), whereas, the normal length of the protein is 40 amino acids long (Ab40) [1,2,7,8]. The amyloid hypothesis suggests that Ab42 forms protofibrils and fibrils. Accumulation of Ab42 impairs basic cellular processes due to oxidative stress, misregulation of intracellular calcium, ER stress [10], and aberrant signaling through interaction with several receptors [3,5,6,8], which results in the death of neurons [7]. Therefore, it is important to understand the mechanism underlying Ab42 mediated cell death and neurotoxicity.
Since the genetic machinery and basic cell biological pathways are conserved from insects to humans, several animal models have been employed to model AD. Despite the immense amount of information available from modeling AD in animal models such as the mouse [2,7] and the fruit fly [7,[11][12][13][14][15], the exact mechanism(s) mediating Ab42-dependent cell death are yet to be determined. The fruit fly has been a model organism for human diseases for many years since nearly 70% of human disease genes are conserved in flies [16]. We have used a Drosophila melanogaster eye model to express the human Ab42 peptide [3].
The Drosophila eye model has been extensively employed to investigate patterning, growth, and cell biological processes [7,[14][15][16]. The adult Drosophila compound eye develops from an epithelial bi-layer structure housed inside the larva called the eye-antennal imaginal disc, which gives rise to an eye, antenna and head cuticle of the adult fly [17]. A synchronous differentiation event in the developing third instar larval eye imaginal disc differentiates retinal precursor cells to photoreceptor neurons [18]. The eye imaginal disc metamorphose to a pupal retina which develops into the adult eye comprising of about 800 units called ommatidia [18]. Each ommatidium contains eight photoreceptors, pigment cells and several support cells. In the pupal retina, the extra undifferentiated cells are eliminated by programmed cell death (PCD) [19]. PCD is not observed during earlier stages of larval eye development, however, abnormal extracellular signaling due to inappropriate levels of morphogens may trigger cell death in the developing larval eye imaginal disc [20]. We have found that Ab42 dependent cell death is mediated, in part, through activation of the JNK signaling pathway [3]. However, blocking the JNK signaling pathway does not completely rescue the Ab42dependent cell death [3]. Therefore, there may be other genetic components that remain to be identified.
Using the Gal4/UAS system [21], we have developed an AD model with transgenic flies [3] where high levels of Ab42 are misexpressed in the differentiating photoreceptor neurons of the fly retina using a Glass Multiple Repeat driver [22] (GMR-Gal4.UAS-Ab42, hereafter GMR.Ab42). These GMR.Ab42 transgenic flies exhibit progressive neurodegenerative pathology in the developing retina, which is similar to that observed in AD [3]. Moreover, the misexpression of Ab42 in the differentiating retina (GMR.Ab42) exhibits a stronger neurodegenerative phenotype at 29uC [3]. The expression of the cell fate marker like disc large (dlg, a membrane specific marker) in the developing eye imaginal disc was studied. In comparison to the wild type adult eye ( Figure 1A) and the larval eye imaginal disc ( Figure 1B), misexpression of Ab42 (GMR.Ab42) in the Drosophila eye imaginal disc resulted in a reduced eye size with disorganized photoreceptors on the posterior margin as evident from the expression of pan neural marker, Elav (DSHB), in the photoreceptor neurons ( Figure 1G), and a highly reduced adult eye which did not show any wild type ommatidium within the compound eye ( Figure 1F) [3].
Our earlier studies showed that in the GMR.Ab42 retina, the ommatidia delaminated from the retinal layers possibly due to loss of polarity and/or cell adhesion [3]. We tested various components of the apical-basal polarity gene pathway in a forward gain of function genetic screen by individually co-expressing the apical basal polarity genes with Ab42 (GMR.Ab42+apical basal polarity genes) in the differentiating photoreceptor neurons. From this screen, we identified a transmembrane protein Crumbs (Crb), as a strong genetic modifier of the Ab42 mediated neurodegenerative phenotype. Crb is highly conserved and has three homologs CRB1, CRB2 and CRB3 in humans. An apical basal polarity gene crb encodes Crb protein, which is localized to the apical domain of the epithelial cells, where it is involved in setting up the apico-basal axis of the cell [23]. Furthermore, Crb is required for organizing apico-basal polarity specification, adherens junctions (AJ) and remodeling in epithelial cells [23,24]. Crb works by forming a complex with Stardust (Sdt/Pals1) [25]. Sdt, in turn, binds to the intracellular domain of Crb and recruits Pals associated tight junction protein (Patj) [26] and Lin7 [27]. To date, Crb has not been reported to play any role in Ab42 mediated neurodegeneration.
We have employed Gal4/UAS system for targeted misexpression studies [21]. All Gal4/UAS crosses were maintained at 18uC, 25uC and 29uC, unless specified, to sample different induction levels. The adult fly cultures were maintained at 25uC, while the egg laying (progeny) were transferred to 29uC. Misexpression of Ab42 in the differentiating retina (GMRGal4.UAS-Ab42, GMR.Ab42) exhibits a stronger neurodegenerative phenotype at 29uC [3]. All the targeted misexpression experiments were conducted using the Glass Multiple Repeat driver line (GMR-Gal4), which directs expression of transgenes in the differentiating retinal precursor cells of the developing eye imaginal disc and pupal retina [22].

Detection of cell death
Cell death was detected using TUNEL assays from Roche Diagnostics [3,31]. TUNEL assays were used to identify the cells undergoing cell death where the cleavage of double and single stranded DNA is labeled by a Fluorescein. The fluorescently labeled nucleotides are added to 39 OH ends in a templateindependent manner by Terminal Deoxynucleotidyl transferase (TdT). The fluorescent label tagged fragmented DNA within a dying cell can be detected by fluorescence microscopy. Eye antennal discs after secondary antibody staining [32] were blocked in 10% normal donkey serum in phosphate buffered saline with 0.2% Triton X-100 (PBT) and labeled for TUNEL assays using a cell death detection kit from Roche Diagnostics.
The TUNEL positive cells were counted from five sets of imaginal discs and were used for statistical analysis using Microsoft Excel 2010. The P-values were calculated using one-tailed t-test and the error bars represent Standard Deviation from Mean [3].

Adult eye imaging
Adult eye images were taken on the Axioimager.Z1 Zeiss Apotome. Adult flies were mounted onto a needle and the image was completed by using extended depth of focus function of the Axiovision software version 4.6.3 by compiling the individual stacks from the Z-sectioning approach. The final images and figures were prepared using Adobe Photoshop CS4 software.

Results
We tested Crb protein levels using Crb antibody (Cq4, DSHB) [23] in the GMR.Ab42 eye imaginal disc using a modified protocol [30]. The Crb protein is localized to the apical domain of the epithelial cells. We observed higher levels of Crb protein in the GMR.Ab42 background ( Figure 1F) as compared to the wild type eye imaginal disc ( Figure 1C). Misexpression of Ab42 peptide with full length Crb [28] using GMR-Gal4 driver (GMR.Ab42+Crb (FL), as evident from Crb antibody staining ( Figure 1I), resulted in increased neurodegeneration as shown by highly disorganized morphology due to clumping of photoreceptor neurons (Red channel, marked by Elav) of neighboring ommatidia of the eye imaginal disc ( Figure 1H). Large gaps were observed among the photoreceptors of the ommatidia where the cells begin to die or clump together. The adults failed to form due to early pupal lethality ( Figure 1G). These animals died in the early pupal stages; as a result we could not observe any pupal retina like structures (data not shown). Downregulating Crb levels by using a heterozygous combination of FRT82B crb 11A22 allele [33] ( Figure 1L) or crb RNAi (Figure 1 O) resulted in the rescue of the GMR.Ab42 mediated neurodegeneration as seen in the eye imaginal disc ( Figure 1K, N) as well as in the adult eye ( Figure 1J, M). We found significant rescue although complete restoration to the wild type eye was not seen. These results suggested that higher levels of crbs are associated with the retina undergoing neurodegeneration due to misexpression of Ab42. Furthermore, Ab42 mediated neurodegeneration can be rescued by downregulating crb function.
We employed TUNEL staining to discern the mechanism of neurodegeneration due to misexpression of Crb in the developing retina. The TUNEL staining marks the nuclei of the dying cells, where the cleavage of double and single stranded DNA is labeled by Fluorescein [31]. Here we utilized TUNEL staining to quantitate the effects of Crb protein levels on neurodegeneration in the GMR.Ab42 background (Figure 2A-F). The TUNEL positive cells were counted from five sets of imaginal discs and were used for statistical analysis using Microsoft Excel 2010. The P-values were calculated using one-tailed t-test and the error bars represent Standard Deviation from the Mean [3]. It is known that a few cells undergo cell death in the wild-type eye imaginal disc ( Figure 2A) which does not affect the final morphology of the adult compound eye ( Figure 1A). The number of TUNEL positive nuclei of the dying cells in the GMR.Ab42 flies ( Figure 2B) was almost three times as high when compared to the wild-type eye imaginal disc (p = 1.943610 26 ; Figure 2F). We investigated the levels of Crb with reference to the induction of cell death and found that when Crb levels were increased in a GMR.Ab42 background (GMR.Ab42+Crb FL), the TUNEL positive cell number increased ( Figure 2C) and was almost seven times higher than the wild type eye imaginal disc (p = 9.536610 28 ; Figure 2F) and nearly two times higher than the GMR.Ab42 eye imaginal disc ( Figure 2F). Reducing levels of crb by using crb 11A22 allele [33] ( Figure 2D) or crb RNAi ( Figure 2E) reduced cell death as evident from reduction in the number of TUNEL positive cells to almost two fold with respect to the GMR.Ab42 eye imaginal disc (for crb 11A22 p = 8.386610 25 , for crb RNAi p = 8.030610 25 ; Figure 2F).
Next, we investigated the effects of modulating levels of Crb on retinal axon targeting from the retina to the brain using chaoptin (24B10, a marker for photoreceptor cells and their axons [34], DSHB) staining. Disruption of axonal transport mechanisms that leads to axonal vesicle stalling has been shown to contribute to the neurodegenerative phenotypes in the AD fly model [35]. During Drosophila visual system development, stereotypical targeting of the axons from the retinal neurons to the special layers of the optic ganglion, medulla and lamina of the brain occurs. The axons of the eight photoreceptor neurons from each ommatidium [36] fasciculate together and project as a single bundle towards the optic lobes of the brain. The Drosophila photoreceptors (R cells) seek specific targets to connect in distinct layers of the optic lobes of the brain, viz., R1-R6 axons project to the lamina; R7 and R8 axons project to the separate layers of the medulla [37]. In comparison to the wild-type eye disc where retinal neurons innervate different layers (medulla and lamina) in the brain ( Figure 3A), the GMR.Ab42 eye disc shows complete loss of axonal targeting ( Figure 3B). Additional upregulation of full length Crb levels in GMR.Ab42 (GMR.Ab42+Crb FL) strongly affected the retinal axon targeting from the retina to the brain ( Figure 3C) as compared to the wild type ( Figure 3A) and the GMR.Ab42 alone ( Figure 3B). The axonal targeting was restored when crb levels were reduced in the GMR.Ab42 background by using either FRT82B crb 11A22 allele ( Figure 3D) or crb RNAi ( Figure 3E). These results further validated our hypothesis that higher levels of Crb enhanced the neurodegenerative phenotype of Ab42 aggregate accumulation.
In order to discern how different domains of Crb protein ( Figure 4A) are involved in preventing GMR.Ab42 mediated neurodegeneration, we used the structure function analysis approach. The full length Crb, a type I transmembrane protein, has 28 EGF domains and four Laminin-AG like repeats in its large extracellular domain (ECD), a transmembrane domain (TM), and a short intracellular domain (ICD) ( Figure 4A). The Crb protein's TM domain consists of 37 amino acids spanning the region of the membrane [38]. The ICD contains two motifs, juxtamembrane FERM-binding motif (FBM or JM) domain and C-terminal PDZ (Postsynaptic density/Discs large/ZO-1) binding motif (PBM) domain ( Figure 4A). Through its PBM domain, Crb forms a complex with PDZ domain proteins, Stardust and PatJ [25]. It is important to note that the ICD of Crb protein interacts with a variety of conserved proteins including apical basal polarity genes Par6 and aPKC [39,40]. Prior structure-function studies using the different Crb domains, for example, in the gastrulating embryo, showed that the ubiquitous expression of a membranebound cytoplasmic ICD, suppressed the crb mutant phenotype to the same extent as full length crb [23,28]. Thus, the different domains of Crb carry out different downstream signaling interactions of the protein, so it is important to investigate which domains are involved in the rescue or enhancement of the neurodegeneration caused by Ab42.
We employed targeted misexpression [21] of Ab42 and various domains of Crb protein using the GMR-Gal4 driver [22] for a structure function analysis. The rationale of these studies was to determine which domain of Crb protein is required for its function in Ab42 mediated neurodegeneration ( Figure 1D, E). As discussed previously, in comparison to the wild type eye ( Figure 4B), GMR.Ab42 exhibited strong reduction in size due to neurodegeneration as seen in the adult eye ( Figure 4C), whereas GMR.Crb [41] resulted in an increase of the adult eye size with minimal necrosis on the margin ( Figure 4D) [42]. Targeted misexpression of Crb ICD (the Crb ICD construct used has been Figure 1. Levels of apical basal polarity gene crb modulates Ab42 mediated neurodegeneration. Wild type (A) adult compound eye, a highly organized structure comprising of 750-800 ommatidia [18], which develops from (B, C) eye-imaginal disc. Third-instar eye imaginal disc stained with membrane specific marker, Disc large (Dlg; white), a pan neural marker Elav (red, marks photoreceptors), and (C) Crb protein expression. The Crb expression is localized on the apical surface of epithelial cells and accumulates at the apical membrane's outer margin [51]. (D-F) Misexpression of Ab42 using GMR-Gal4 driver [22] in the differentiating photoreceptor neurons results in the induction of neurodegeneration as seen in (D) the highly reduced adult eye with a glazed surface and (E, F) eye imaginal disc. Note that in GMR.Ab42 eye imaginal discs (E) pan neural marker, Elav, exhibits clumping of the photoreceptor neurons and holes in the developing retina, and (F) strong enrichment of Crb expression in the GMR domain. (G-I) Misexpression of Crb full length [41] in GMR.Ab42 background (GMR.Ab42+Crb FL) strongly enhances the neurodegeneration phenotype which results in (G) pupal lethality (adults failed to form due to early pupal lethality and as a result lacked the adult eye structure) and (H, I) severe neurodegeneration observed in the eye imaginal disc as evident from (H)  Role of Crumbs in Ab42 Mediated Neurodegeneration PLOS ONE | www.plosone.org referred to as Crb intra [28]; Figure 4A) in a GMR.Ab42 background (GMR.Ab42+Crb intra ) resulted in strong enhancement of the neurodegenerative phenotype of GMR.Ab42 alone as seen in the eye imaginal disc ( Figure 4G) as well as in the adult eye ( Figure 4F). The GMR.Ab42+Crb intra adult eye showed strong neurodegeneration as evident from the dark necrotic patch in place of the adult eye ( Figure 4F). However, the control GMR.Crb intra also showed some neurodegeneration ( Figure 4E), which was not as strong as GMR.Ab42+Crb intra ( Figure 4F). Since both the control ( Figure 4E) as well as GMR.Ab42+  Crb intra (Figure 4F, G) showed a neurodegenerative phenotype, it raised the possibility of an additive effect. Further experimentation using the truncated constructs of Crb intra domains disproved this additive effect hypothesis. Targeted misexpression of GMR.Ab42 with Crb intra DPBM [28] or Crb intra DJM [28] in developing retina resulted in the rescue of the GMR.Ab42 neurodegenerative phenotype as seen in the eye imaginal disc ( Figure 4J, M) as well as the adult eye ( Figure 4I, L). The controls GMR.Crb intra DPBM ( Figure 4H) and GMR.Crb intra DJM ( Figure 4K) exhibit a slightly reduced adult eye. The Crb intra construct lacking both the JM and PBM domains (GMR. Crb intra DJM DPBM ( Figure 4A)) resulted in a near normal adult eye ( Figure 4N). Targeted misexpression of GMR.Ab42 with Crb intra DJM DPBM resulted in the rescue of the GMR.Ab42 neurodegenerative phenotype as seen in the eye imaginal disc (Fig. 4P), and the adult eye (Fig. 4O). All these results clearly demonstrated that like the full length Crb (Crb FL), the entire ICD (Crb intra ) is also responsible for the enhancement of the neurodegenerative phenotype of GMR.Ab42. It suggests that Crb ICD is sufficient enough to carry out the Crb FL function in Ab42 mediated neurodegeneration. When we remove either one or both of the JM and PBM domains from the ICD of Crb, the GMR.Ab42 phenotype is rescued and the ommatidia are restored to near wild-type. This data strongly indicates that both the JM and PBM domains in Crb are essential to suppress the Ab42 effects. There might be a correlative interaction between the JM and PBM domains of Crb in the Ab42 mediated neurodegeneration. However, when we have an intact ICD or full length Crb, there is a severe enhancement of the GMR.Ab42 phenotype. Also, in the loss-of-function crb flies (GMR.Ab42+crb 11A22 and GMR.Ab42+crb RNAi) where we see reduced Crb level expression ( Figure 1L, O) as compared to the wild-type ( Figure 1C), there is a rescue of Ab42 mediated neurodegeneration further validating our hypothesis that Crb levels can modify the neurodegenerative phenotype of Ab42 accumulation. Thus, Crb levels can serve as an excellent biomarker for AD.
To further verify the structure function analysis results, TUNEL assays were performed on all of the constructs. The rationale was to examine if the reduced eye phenotype seen in GMR.Ab42+Crb intra was due to cell death or, on the other hand, if the restored eye as shown by removing either or both of the JM and PBM domains of ICD motifs ( Figure 4A) is due to reduced number of TUNEL cells. As mentioned earlier, TUNEL marks the nuclei of dying cells, therefore a reduced number of TUNEL positive cells nuclei corresponds to less cells dying, which will lead to a rescue of GMR.Ab42 neurodegenerative phenotype in the adult eye. We found that the severely reduced adult eye phenotype of GMR.Ab42+Crb intra is in fact due to an increase in the number of TUNEL positive cells as compared to the wild-type and the GMR.Ab42 eye disc ( Figure 5A, B, I). The GMR.Ab42+Crb intra exhibits strong neurodegenerative phenotype as evident from disorganized photoreceptor neurons (marked by Elav, green) in the ommatidia. Furthermore, the number of TUNEL positive cells nuclei are increased ( Figure 5A, B; red). The TUNEL staining explains the reason for a highly reduced adult eye in GMR.Ab42+Crb intra ( Figure 2F). Additionally, when any either JM or PBM or both JM and PBM domains of the ICD motifs were removed in the GMR.Ab42 background, the severity of neurodegenerative phenotypes was significantly reduced. In GMR.Ab42+Crb intra DJM ( Figure 5C, D), GMR.Ab42+ Crb intra DPBM ( Figure 5E, F), or GMR.Ab42+Crb intra DJM DPBM ( Figure 5G, H), the number of TUNEL positive dying cells nuclei were significantly less than GMR.Ab42 and GMR.Ab42+ Crb intra ( Figure 5I). All of these results further validate the data shown in Figure 4 and conforms to the adult eye phenotypes of each of these structures.
For all the ICD motifs of Crb, the TUNEL positive cells were counted from five sets of imaginal discs and were used for statistical analysis using Microsoft Excel 2010. The P-values were calculated using one-tailed t-test and the error bars represent Standard Deviation from the Mean [3]. All the p-values showed the TUNEL count to be significantly different from GMR.Ab42 and the wild-type ( Figure 5I). By studying the domains of Crb with reference to the cell death, we found that misexpression of intact Crb ICD domain in GMR.Ab42 background (GMR.Ab42 +Crb intra ), resulted in the increased number of TUNEL positive cell ( Figure 5I) and was almost six times higher than the wild type eye imaginal disc (p = 1.5559610 27 ) and nearly two times higher than the GMR.Ab42 eye imaginal disc (p = 8.  Crb intra DJM DPBM p = 8.9278610 26 ; Figure 5I). This TUNEL assay further validated our hypothesis that the reduced eye phenotype seen in GMR.Ab42+Crb intra (with its fully intact ICD) is primarily due to induction of cell death and the restored eye phenotypes seen when any one or both of the ICD motifs of Crb is/are removed, does indeed have reduced number of dying cells as evident from TUNEL staining.
To further test our hypothesis, we looked at the axonal targeting from the retina to the brain using 24B10 (Chaoptin) in these constructs ( Figure 4A). As mentioned earlier, 24B10 shows an organized and orderly axon branching from the retina to the brain in the wild-type background ( Figure 3A). However, when we observed the 24B10 staining in the GMR.Ab42+Crb intra eye there is extreme disorganization marked by the clumping of axons, as well as Elav (red) positive photoreceptors which results in impairing of axonal targeting from retina to the brain ( Figure 6A, B). This data further confirms our TUNEL data using GMR.Ab42+Crb intra . Additionally, when we analyzed other constructs of Crb by removing either or both of the JM or PBM domains from the ICD motif, there is a rescue of the adult eye ( Figure 4A, H-P), a reduction in the number of TUNEL positive ( Figure 5C-I), and restoration of the organization of axons from the retina to the brain (Fig. 6 C-H) in all three constructs (GMR.Ab42+Crb intra DJM ( Figure 6C, D), GMR.Ab42+Crb intra DPBM ( Figure 6E, F), GMR.Ab42+Crb intra DJM DPBM (Figure 6G, H). When the JM motif (GMR.Ab42+Crb intra DJM ( Figure 6C, D) or the PBM motif (GMR.Ab42+Crb intra DPBM ( Figure 6E, F) was removed, there is restoration of the axonal targeting as evident from the 24B10 staining and marking the axonal projections innervate the two layers of the brain. Furthermore, when we remove both of the ICD motifs (GMR.Ab42+Crb intra DJMDPBM ( Figure 6G, H), the axonal connection to the brain is restored to near wild type axonal targeting. These data further validates that the ICD domain of Crb is sufficient enough for Crb function in Ab42 mediated neurodegeneration.

Discussion
Our studies strongly suggest that transmembrane protein Crb is involved in Ab42 mediated neurodegeneration. During wing development, N upregulates crb transcription at the dorso-ventral (DV) boundary, and the ability of Crb to inhibit the activity of the c-secretase complex has been proposed to help refine the N activity domain [43]. Crb functions as a negative regulator of the N signaling pathway [42]. Notch is involved in the development and organization of the dorso-ventral boundary through cell proliferation of the developing eye. Because N and Amyloid Precurssor Protein (APP) are cleaved by similar secretases [44] and Crb regulates N, the Crb effects on Ab42 could be caused through N regulation. However, in the GMR.Ab42 model used in our studies, the Ab42 protein is already cleaved from of APP and does not require cleavage by band c-secretase. Therefore, our data using the transgenic model suggests that Crb also acts downstream of c-secretase mediated cleavage of APP. Furthermore, higher levels of Crb can enhance human Ab42 mediated neurodegeneration [3]. Thus, Crb role in modulating Ab42 mediated neurodegeneration is downstream of N signaling pathway.
In addition, Crb is an upstream regulator of the organ size growth control pathway viz., Hippo signaling pathway. Recently, it was shown that Crb interacts with its juxtamembrane FERMbinding motif (JM) with the FERM domain of Expanded (Ex) to regulate growth by affecting the Hippo pathway activity [45][46][47]. Our structure function analysis studies exhibited that ICD of Crb is sufficient for its role in Ab42 mediated neurodegeneration suggesting that Crb may act independent of its interaction with Hippo pathway member Ex in Ab42 mediated neurodegeneration.
Since Crb ICD is involved in its interaction with apical basal polarity gene localization, there is a strong possibility that higher level of Crb in a GMR.Ab42 background might affect the apical basal polarity of the retinal photoreceptor neurons which result in neurodegeneration. Mutations in Crb homolog 1 (CRB1) has been shown to cause autosomal recessive retinitis pigmentosa (arRP) and autosomal Leber congenital amaurosis (arLCA) [36]. During Drosophila eye development, Crb is required in photoreceptors for stalk elongation [42,48], and in preventing light-dependent retinal degeneration [49]. Mutations in the human Crb homolog (CRB1) result in abnormalities like thick retina and lamination problems [50]. Furthermore, mutant Crb protein is thought to be responsible for retinal degenerations [50]. However, in GMR.Ab42 background higher levels of Crb protein were responsible for neurodegeneration. Therefore, it is a strong possibility that higher Crb levels may impair apical basal polarity leading to the Ab42 neurodegeneration. Thus, regulating Crb levels can help prevent the onset of neurodegeneration and Crb may serve as one of the biomarker as well as the key therapeutic targets for the AD. Figure 6. Misexpression of Crb intracellular domain (ICD) can impair axonal targeting. The eye-antennal disc is stained with Elav (red), which marks the photoreceptors, and 24B10 (chaoptin; green), which stains the axons from the retina to the brain [34]. (A, B) Misexpression of intact instar cellular domain ICD GMR.Ab42+Crb intra results in clumping of photoreceptors (Elav; A), disorganization of axonal targeting from the retina to the brain as evident from 24B10 staining. (C, D)When the JM motif using GMR.Ab42+Crb intra DJM , photoreceptor organization as well as the axonal targeting is restored to the wild type. Similarly, removing the (E, F) PBM motif of the ICD using GMR.Ab42+Crb intra DPBM , or both PBM and JM domain in GMR.Ab42+Crb intra DJMDPBM , result in restoration of axonal targeting and photoreceptors. Thus, ICD domain of Crb is required for its role in neurodegeneration. doi:10.1371/journal.pone.0078717.g006