Bax-inhibitor-1 knockdown phenotypes are suppressed by Buffy and exacerbate degeneration in a Drosophila model of Parkinson disease

Introduction

Bax inhibitor-1 (BI-1) belongs to a diverse group of proteins, known as transmembrane Bax inhibitor-1 motif-containing (TMBIM) family (Henke et al., 2011; Li et al., 2014; Reimers et al., 2008; Rojas-Rivera & Hetz, 2015), that have been determined to be regulators of cell death. A different nomenclature categorises these proteins into the LFG family, adopted from the family member Lifeguard (Hu, Smith & Goldberger, 2009), which consists of at least six highly conserved members in a wide range of organisms (Chae et al., 2003; Henke et al., 2011; Huckelhoven, 2004). These regulators of cell death, accomplish this role by the regulation of the death receptor, modulation of the endoplasmic reticulum (ER) calcium homeostasis, ER stress signalling pathways, autophagy, reactive oxygen species (ROS) production, cytosolic acidification, and other cellular activities (Li et al., 2014; Rojas-Rivera & Hetz, 2015). The founding member of this group is BI-1, or TMBIM6 also known as testis enhanced gene transcript, and has been demonstrated to inhibit the effect of Bax-induced cell death (Walter et al., 1995; Xu & Reed, 1998). Members of this protein family possess a BI-1-like domain with six to seven transmembrane (TM)-spanning regions that are strongly associated with the ER membranes (Carrara et al., 2012; Chae et al., 2004; Xu & Reed, 1998). BI-1 is highly conserved across diverse species with eukaryotic homologues of BI-1 able to block Bax-induced cell death when expressed in yeast (Chae et al., 2003), thus implying it regulates an evolutionarily conserved cytoprotective pathway.

This protein though not structurally related to the B cell lymphoma 2 family of proteins, forms a complex with the pro-survival members Bcl-2 and Bcl-X_L but not with Bax or Bak (Lisbona et al., 2009; Xu & Reed, 1998). Therefore, it is likely the anti-apoptotic activity of BI-1/TMBIM6 is mediated by interaction with pro-survival members of the Bcl-2 family and acts downstream of Bcl-X_L (Xu et al., 2008). BI-1 deficient cells, that include neurons, are more sensitive to apoptosis induced by ER stress and has been linked to the modulation of ER calcium homeostasis (Chae et al., 2004; Dohm et al., 2006). This implicates BI-1 in a variety of human diseases that include numerous cancers, obesity, liver diseases, autoimmune response, and diabetes (Kiviluoto et al., 2012; Li et al., 2014; Lisak et al., 2016; Robinson et al., 2011; Rojas-Rivera & Hetz, 2015). Neuroprotective roles include, protection from oxygen–glucose deprivation, promotion of neuronal proliferation and differentiation, and stress-induced protection (Dohm et al., 2006; Hunsberger et al., 2011; Jeon et al., 2012; Krajewska et al., 2011). It regulates ROS production by modulation of unfolded protein response (UPR) induction in the ER (Lee et al., 2007), suppression of mitochondria-mediated ROS production (Kim et al., 2012), reduction of cytochrome P450 2E1 activity, and regulation of the ER membrane lipid peroxidation (Kim et al., 2009). BI-1 undoubtedly has significant cytoprotective roles and their abrogation lead to cellular homeostatic dysfunction and disease.

Drosophila melanogaster appear to possess most of the TMBIM protein family homologues with TMBIM6/BI-1 represented by BI-1/CG7188 (Attrill et al., 2015; Hu, Smith & Goldberger, 2009; Rojas-Rivera & Hetz, 2015). Drosophila has been used as a model organism in the study of gene expression and in human disease models, albeit with very promising results (Staveley, 2015). Several studies have used Drosophila to elucidate the importance of this protein in cellular homeostasis; including functional conservation of this protein in evolutionarily diverse organisms (Chae et al., 2003), BI-1 as a negative regulator of the ER stress sensor IRE1α and its role in the UPR (Lisbona et al., 2009), and its modulation of autophagy (Castillo et al., 2011). Expression in the Ddc-Gal4-expressing neurons is the focus of our studies as they are very sensitive to subtle differences in gene products and can be used to study ROS, ER stress, apoptosis, autophagy, and many other cellular processes. This is mainly because they degenerate in an age-dependent manner and this degeneration manifests as deficiency in locomotor function (Botella et al., 2009; Feany & Bender, 2000; Park, Schulz & Lee, 2007; Staveley, 2015). The key elements of the Drosophila model of Parkinson disease (PD) that utilizes the expression of a human α-synuclein transgene to induce the PD-like symptoms (Feany & Bender, 2000); is its ability to recapitulate some features of human PD that include, age-dependent loss of dopaminergic (DA) neurons that manifest in age-dependent loss in locomotor function (Auluck et al., 2002; Botella et al., 2009; Buttner et al., 2014; Feany & Bender, 2000; Kong et al., 2015; Staveley, 2015; Wang et al., 2015; Zhu et al., 2016). The spatio-temporal UAS/GAL4 expression system (Brand & Perrimon, 1993), and the availability of a plethora of promoters or enhancers of which TH-Gal4, elav-Gal4, and Ddc-Gal4 are employed to model PD in flies.

The Bcl-2 family member homologues in Drosophila are limited to the single anti-apoptotic Buffy and the pro-apoptotic Debcl (Colussi et al., 2000). In previous studies, the overexpression of Buffy has been shown to confer survival advantages specifically in response to external stimuli and in conditions of cellular stress (M’Angale & Staveley, 2016a; Monserrate, Chen & Brachmann, 2012; Sevrioukov et al., 2007; Tanner et al., 2011). This point to an important role for this protein in aspects of cell death. We investigated the outcome of the knockdown of BI-1 in Drosophila neurons, and further determined whether there is an interaction with the anti-apoptotic Bcl-2 protein Buffy. We employed two different RNAi lines to determine the specificity of the effects of knockdown of this gene and compared them to a control line. We further co-expressed BI-1 in DA neurons along with α-synuclein to investigate whether it possesses neuroprotective functions by assessing the phenotypes that would result from knockdown of BI-1 and expression of α-synuclein. Lastly, in supportive experiments we attempted to establish a role for BI-1 in the Drosophila developing eye.

Materials and Methods

Results

Drosophila BI-1 is closely related to the human homologue

The 245 amino acids Drosophila BI-1 isoform A has a 42% identity and 68% similarity to the 295 amino acids human isoform B. The Drosophila homologue has a BI-1 domain between amino acids 21–223 and the human version at 74–286 (Fig. 1) as determined by the NCBI CDD (Marchler-Bauer et al., 2015). An alignment of the protein sequences using Clustal Omega (Goujon et al., 2010; Sievers et al., 2011) shows high conservation of the BI-1-like domain in the organisms analysed (Fig. 1A). Six TM domains in both Drosophila and human BI-1 were identified using both ELM (Dinkel et al., 2014) and TMpred (Artimo et al., 2012). An analysis of membrane-spanning domains by Phyre2 (Kelley et al., 2015) reveals seven TM domains (Fig. 1B) in both sequences that are highly identical in the cytoplasmic to intracellular orientation. An inhibitor of apoptosis binding motif at amino acids 1–5, an ER retention motif at position 221–224, and binding motifs for Atg8 at position 212–224 and calmodulin at amino acids 226–242 were identified by ELM. The presence of NES was detected in both Drosophila and human BI-1 using NetNES (la Cour et al., 2004) and only in Drosophila using the ELM. The 3D modelling of these proteins using Phyre2 (Fig. 1C) shows a close similarity in the structure and the orientation of the TM domains with the image coloured by rainbow from the N → C terminus.

Knockdown of BI-1 in DA neurons decreases lifespan and severely impairs locomotor function

The expression of both BI-1-RNAi lines in the Ddc-Gal4-expressing neurons results in decreased lifespan and impaired locomotor function. The median lifespan for these flies was 54 days for BI-1-RNAi (1) and 46 days for BI-1-RNAi (2) when compared to 70 days for the controls that express the lacZ transgene as determined by the Log-rank (Mantel–Cox) test (Fig. 2A). When BI-1 is suppressed in these neurons, the flies develop an early onset impairment of locomotor ability as determined by the nonlinear fitting of the climbing curves (Fig. 2B). The 95% CI for the slope were 0.033–0.050 and 0.0175–0.0355 for the two RNAi lines respectively when compared to 0.052–0.070 for the lacZ control flies. These results appear to suggest a role for BI-1 in the protection of neurons in Drosophila.

Buffy suppresses the BI-1-RNAi-induced phenotypes

The directed overexpression of the pro-survival Bcl-2 homologue Buffy results in increased lifespan and improved climbing ability (M’Angale & Staveley, 2016a). When Buffy is co-expressed with both BI-1-RNAi lines in the Ddc-Gal4-expressing neurons, the results indicate an increased median lifespan of 70 days and 72 days respectively when compared to 74 days for Buffy co-expressed with lacZ control flies and 70 days for the lacZ flies, as determined by Log-rank test (Fig. 3A). The climbing ability of the BI-1-RNAi flies was not significantly different from the Buffy co-expressed with lacZ controls as determined by comparison of the BI-1-RNAi climbing curves (Fig. 3B) with the control curve. The 95% CI for the slope of BI-1-RNAi (1) was 0.0340–0.057 and that of BI-1-RNAi (2) was 0.040–0.061 when compared to 0.035–0.050 and 0.052–0.070 for the controls. Taken together these results suggest a pro-survival role for BI-1; as the phenotypes induced by its knockdown are significantly counteracted by the pro-survival Bcl-2 homologue Buffy.

Knockdown of BI-1 with the expression of α-synuclein slightly alters phenotypes

The expression of α-synuclein in dopaminergic neurons results in impaired locomotor function that is attributed to cellular toxicity. The co-expression of BI-1-RNAi along with α-synuclein in the Ddc-Gal4-expressing neurons, slightly exacerbated the reduced survival and the loss in climbing ability observed with the expression of α-synuclein. The median lifespan was 52 days and 54 days for flies that express BI-1-RNAi along with α-synuclein compared to 58 days for controls that co-express α-synuclein along with the lacZ transgene and 70 days for flies that express the benign lacZ transgene (Fig. 4A) as determined by Log-rank test with P < 0.001. A comparison of the climbing curves by nonlinear fitting at 95% CI revealed they were significantly different (Fig. 4B), with a CI of 0.038–0.049 for BI-1-RNAi (1) and 0.025–0.033 for BI-1-RNAi (2) co-expressed along with α-synuclein and compared to 0.052–0.069 for the α-synuclein co-expressed with lacZ control flies. This implies that the knockdown of BI-1 in the Ddc-Gal4-expressing neurons abrogates its cytoprotective function and enhances the α-synuclein-induced phenotypes.

Knockdown of BI-1 in the eye decreases ommatidia number and increases degeneration, phenotypes that are rescued upon Buffy overexpression

The directed knockdown of BI-1 in the Drosophila developing eye using the GMR-Gal4 transgene resulted in eyes with decreased number of ommatidia and a higher disruption of the ommatidial array in both the RNAi lines that were tested (Figs. 5A–5C and 5J) as determined by a one-way analysis of variance with a P value < 0.0001. Co-expression of both BI-1-RNAi lines with Buffy restored the mean number of ommatidia to control levels as determined by a one-way analysis of variance with P = 0.2439 and 0.2342. The percentage disruption was significantly different from the control flies (Figs. 5D–5F and 5K). Taken together, these results suggest that BI-1 may play a pro-survival role in the development of the Drosophila eye and that Buffy suppresses the developmental eye defects that result from the knockdown of BI-1. The knockdown of BI-1 along with α-synuclein expression resulted in a significant decrease in the number of ommatidia or increase in percentage disruption of the eye as determined by a one-way analysis of variance with a P value < 0.0001 in both instances (Figs. 5G–5I and 5L). The number of ommatidia and percentage of disruption was worse than with either α-synuclein expression or BI-1 knockdown. This indicates that the knockdown of BI-1 enhances the α-synuclein-induced eye defects.

Discussion

The knockdown of BI-1 via stable inducible RNAi in the Ddc-Gal4-expressing neurons of Drosophila results in decreased survival and impaired climbing ability over time. Although there is no known homologue of Bax in Drosophila, the only pro-apoptotic Bcl-2 homologue is Debcl (Brachmann et al., 2000; Colussi et al., 2000; Igaki et al., 2000; Zhang et al., 2000), and has been demonstrated to possess pro-apoptotic functions. The Drosophila BI-1 is able to block Bax-induced cell death in yeast (Chae et al., 2003), and reduction of BI-1 function induces cell death (Xu & Reed, 1998). These results suggest neuronal dysfunction may result from degeneration or death when the function of BI-1 is reduced in the Ddc-Gal4-expressing neurons. The BI-1-induced cell death could occur through interaction with pro-survival Bcl-2 proteins at the ER membrane (Xu & Reed, 1998) and especially Bcl-2 and Bcl-X_L in humans and possibly Buffy in Drosophila. BI-1 seems to be involved in cellular functions that are protective to ER stress-induced apoptosis (Chae et al., 2004). It seems to do this by the regulation of calcium ions (Lisak et al., 2016; Xu et al., 2008) and ROS (Kim et al., 2009; Lee, Kim & Chae, 2012). BI-1 regulates ER stress by controlling ER-generated ROS accumulation and stress linked to the UPR. Therefore, the knockdown of this important ER stress regulator in the DA neurons would result in neuronal degeneration and death. The only pro-survival Bcl-2 homologue in Drosophila is Buffy (Quinn et al., 2003) and the overexpression of Buffy is known to confer survival advantages to cells under normal conditions and under conditions of stress (Clavier et al., 2014; M’Angale & Staveley, 2016a, 2016c, 2017; Monserrate, Chen & Brachmann, 2012; Quinn et al., 2003; Sevrioukov et al., 2007). The overexpression of Buffy along with the knockdown of BI-1 resulted in the suppression of the BI-1-induced phenotypes. This Buffy action may be specific to its interaction with BI-1 or to its general pro-survival signalling pathways. The rescue of the BI-1-induced phenotypes in both the Ddc-Gal4-expressing neurons and in the developing eye may indicate a pro-survival role for BI-1 in Drosophila, as the pro-survival action of Buffy can abrogate its phenotypes.

The expression of human α-synuclein in DA neurons of Drosophila results in impaired climbing ability (Feany & Bender, 2000), similar to what is observed in reduced BI-1 function. The expression of α-synuclein along with the reduction of BI-1 activity significantly altered the impaired locomotor ability observed. The age-dependent reduction of climbing ability could be a result of BI-1-induced apoptosis coupled with neurotoxicity that result from α-synuclein accumulation and the subsequent dysfunction of cellular mechanisms. Although we observed the enhancement of the α-synuclein-induced phenotypes by the knock down of BI-1, marked by the reduction in longevity and a precocious loss in climbing ability, the BI-1-induced phenotypes were hardly altered by the expression of α-synuclein. This observation may exclude the involvement of α-synuclein in the BI-1 pathway as its expression does not enhance the BI-1 phenotypes, and the observed phenotypes may be because of the knockdown of BI-1. All the same, it appears that the presence of either of the mechanisms, vis-a-vis BI-1-induced apoptosis or α-synuclein aggregation neurotoxicity, confers a great disadvantage to Ddc-Gal4-expressing neurons.

The suppression of BI-1 in the Drosophila eye under the direction of the GMR-Gal4 transgene results in a lower ommatidia number when compared to the control. BI-1 is an apoptosis suppressor gene and the down-regulation of its protein product results in programmed cell death (Li et al., 2014). The reduction in the ommatidia number observed is mainly due to the fusion of ommatidia and the resulting ommatidia disarray. The knockdown of BI-1 in the Drosophila eye seems to exacerbate the Gal4-induced apoptosis that manifests as roughened-eye phenotype (Kramer & Staveley, 2003). The co-expression of the Bcl-2 pro-cell survival homologue Buffy with BI-1-RNAi results in the suppression of the phenotype, with the number of ommatidia and the roughened eye restored to control levels. Buffy seems to ameliorate this phenotype and it is possibly via a general action on survival signals or an interaction with BI-1.

The expression of α-synuclein in the Drosophila eye results in reduced ommatidia and a highly disrupted ommatidial array (Feany & Bender, 2000). This α-synuclein-induced developmental eye defects model is a viable system to show the effects of altered gene expression and its role in neuroprotection. The co-expression of α-synuclein with BI-1-RNAi in the Drosophila eye resulted in decreased ommatidia number and a highly disrupted ommatidial array when compared to the control that expresses α-synuclein. The number of ommatidia decreased further when α-synuclein was co-expressed with BI-1-RNAi. Additionally, the degree of disruption of the ommatidial array was also increased. Though it did not appear to be additive in nature, it seems that the combination of the expression of the neurotoxic α-synuclein and the knockdown of the activity of the anti-apoptotic BI-1 results in a worsening of the roughened eye phenotype. The accumulation of α-synuclein has been implicated in breakdown of cellular homeostasis that include apoptosis, ROS production, and autophagy (Chinta et al., 2010). The knockdown of BI-1 disrupts regulation of similar mechanisms as those implicated in α-synuclein-induced neurotoxicity that include apoptosis, autophagy, and ROS production (Li et al., 2014). It therefore, follows that the combined action of α-synuclein expression and BI-1 knockdown worsened the phenotypes that result from either α-synuclein expression or BI-1 knockdown.

Conclusion

The knockdown of BI-1 in the Ddc-Gal4-expressing neurons of Drosophila results in reduction in lifespan and an age-dependent loss in climbing ability, phenotypes that are strongly associated with the degeneration and loss of dopaminergic neurons. The co-expression of the pro-survival Buffy with BI-1-RNAi results in the rescue of the phenotypes observed, it is possible that Buffy and BI-1 participate in cellular pathways that promote anti-apoptosis. Finally, BI-1 appears to be neuroprotective as its knockdown along with α-synuclein expression result in enhanced phenotypes.

Supplemental Information