All Drosophila stocks were maintained on the standard food media at 25°C, and the genetic crosses were performed at standard lab conditions at a set temperature of 25°C. Fly stocks used in the present study are Oregon-R⁺, UAS-Hsc70-4RNAi (BL-28709 and BL-34836), UAS-Hsc70-4WT (BL-5846), obtained from Bloomington Drosophila Stock Center; w¹¹¹⁸; UAS-127Q.HA; +/+, and w¹¹¹⁸; UAS-20Q.HA; +/+ (Kazemi-Esfarjani and Benzer, 2000), w¹¹¹⁸;GMR-GAL4; +/+ (Freeman, 1996), w¹¹¹⁸;GMR-GAL4:UAS-127Q/CyOGFP; +/+ (Yadav and Tapadia, 2013), w¹¹¹⁸; UAS-httex1PQ93/CyO; +/+ (Steffan et al., 2001).

To examine the external morphology of adult eyes, flies of desired genotypes were etherized and their eyes were photographed using a Sony Digital Camera (DSC-75) attached to Nikon PDSL-42 stereo binocular microscope.

For nail polish imprint (Arya and Lakhotia, 2006), flies of the desired genotype were etherized followed by decapitation of their head with a needle. Decapitated heads were dipped in the transparent nail polish and then allowed to dry at room temperature. Then, the dry nail polish was peeled off with the help of a needle to create an exact replica of the external surface of the Drosophila eye, which was subsequently observed under the DIC optics in a Nikon Ellipse 800 microscope.

This assay was performed using Y-maze. The flies of the desired genotype were subjected to a phototaxis assay. A group of 10 flies were put in the Y-maze. One side of the arm was dark and was covered with black paper while another side of the arm was illuminated by external light, and the stem of the Y-maze was also covered with black paper (Singh et al., 2014). Flies were transferred into the stem of the Y-maze. After 60 s, the number of flies present in each arm was counted. These experiments were performed in triplicate for each genotype having 100 flies. Flies of light and dark arms were calculated into mean valves of percentage and presented as ± standard error mean.

Eye imaginal discs of the desired genotypes were dissected out in 1X PBS followed by 20 min of 4% PFA fixation. Then, tissues were washed in 0.1% PBST three times (1× PBS, 0.1% Triton-X) for 10 min and then kept in blocking solution (1× PBS, 0.1% Triton-X, 0.1% BSA, 10% FCS, and 0.02% Thiomersal) for 2 h at RT followed by the overnight incubation of the primary antibody at 4°C. Post-incubation, the tissues were subjected to three times washing in 0.1% PBST for 10 min followed by blocking for 1 h and then incubated in a secondary antibody for 2 h at RT followed by washing three times in 0.1% PBST, counterstained with DAPI for 20 min, rinsed in 0.1% PBST, and finally mounted in glycerol containing an antifading agent, DABCO (Sigma, D8001).

Primary antibodies which were used in the experiments are rabbit anti-HA (1:1500, Sigma-H6908), rat anti-ELAV (1:100, DSHB-7E8A10), mouse anti-Futsch (1:100, DSHB-22C10), mouse anti-Dlg1 (1:20, DSHB-4F3), rabbit p-JNK (1:100, Promega-pTPpY), mouse anti-Relish (1:50, DSHB-21F3), rat anti-HSP70 (1:500, Sigma-SAB5200204), mouse anti-Hsc70 (1:100, Abnova-MAB6130), mouse anti-beta-tubulin (1:500, DSHB-E7), acridine orange (10 μg/ml, Sigma-A6014), and DAPI (1 mg/ml, Sigma-D9542). Imaging was done using Zeiss LSM 510 Meta confocal microscope. Images were processed with Adobe Photoshop7.

Protein samples were prepared by dissecting out eye discs from wandering third instar larvae from the desired genotypes. Eye discs were dissected in 1× PBS, after which they were transferred in RIPA buffer (100 mM Tris-Cl pH 6.8, 100 mM NaCl, NP-40, EDTA, NaF, NaO₂V₄, 20% glycerol, protease inhibitor cocktail, and 2 mM PMSF). Then, the tissues were homogenized in ice and then subjected to centrifugation at 12,000 rpm for 20 min to collect the protein-containing supernatants. Then, protein quantification was done using Bradford’s protein quantification technique. Finally, 25 μg of protein were taken from the supernatant and added in a 1:1 ratio with sample buffer (100 mM Tris-Cl pH 6.8, 4% SDS, 0.2% bromophenol blue, 20% glycerol, 100 mM DTT, and 2 mM PMSF). Then, samples were denatured in a boiling water bath for 5 min and were centrifuged at 12,000 rpm for 10 min. After this, the supernatant was collected and subjected to SDS-PAGE. Protein samples were then transferred onto PVDF membrane using wet electrotransfer apparatus at 100 V for 1.5 h at 4°C. After the completion of the transfer, PVDF membranes were incubated in blocking solution (4% BSA in 0.1% TBST) for 2 h followed by primary antibody incubation for overnight at 4°C. Membranes were then washed in 0.1% TBST for 15 min in a repetition of three times. After this, membranes were incubated in HRP-tagged secondary antibody for 2 h at room temperature followed by three times washing with 0.1% TBST for 15 min each. Signals were detected by using ECL via ChemiDoc (VILBER).

Eye imaginal discs were dissected in 1× PBS and then were collected in RIPA buffer (50 mM Tris-Cl pH 7.4, 150 mM NaCl, 1% NP-40, 0.1% SDS, 1 mM EDTA pH 8, 1 mM PMSF) accompanied with 1 μg/ml w/v protease inhibitor cocktail. Then, homogenized in ice and after centrifugation at 5,000 g at 4°C for 5 min, supernatants were quantified. Then, 1 mg of total proteins were taken and incubated with Protein G-Dynabeads (Invitrogen-10004D) at 4°C overnight. Then, beads were separated from the supernatant using a magnet stand and were subjected to washing in chilled RIPA buffer three times at 4°C for 5 min each. Then, beads were separated from the washing solution completely and subjected to elution. Protein elution was done in SDS sample buffer (100 mM Tris-Cl pH 6.8, 4% SDS, 0.2% bromophenol blue, 20% glycerol, 100 mM DTT, and 2 mM PMSF) and was denatured by boiling it for 5–10 min. Finally, beads were separated and the supernatant was subjected to Western blotting.

Head of 5-day-old flies were chopped and washed in 1× phosphate buffer (PB) and then fixed in Karnovsky’s fixative for 2 h at RT, after which head was post-fixed in 1% Osmium tetroxide in 0.1M PB, pH 7.4 for 1 h followed by three washes with 0.1 PB, pH 7.4, 15 min each at RT. The sample was then dehydrated in a graded series of ethanol (30, 50, 70, 90, 95, and 100%) and then air-dried. The dried samples were then mounted on the carbon-taped SEM stub and sputter-coated with the platinum. The SEM images were taken using a Carl Zeiss, EVO-18, scanning electron microscope.

Total RNA was isolated from the eye imaginal discs of healthy wandering third instar larvae of the desired genotypes in order to check the expression pattern of several immune genes, such as relish, attacin, cecropin, diptericin, drosocin, drosomycin, and different isoforms of hsc70 and hsf. In order to achieve so, wandering third instar larvae of the desired genotypes were dissected to isolate total RNA using TRIzol reagents by following the recommended protocol provided by Ambion, India. The RNA pellets were resuspended in 15 μl of DEPC-MQ, and after the pellets were dissolved, their quantitative estimation was done using spectrophotometric analysis. Then, 1 μg of each RNA sample was incubated with 1U of RNase-free DNaseI (Roche, Sigma-4716728001) for 30 min at 37°C to remove any residue DNA. Then, first-strand cDNA was synthesized from these incubated samples, by following the standard cDNA preparation protocol. The prepared cDNA was subjected to real-time PCR using forward and reverse primer pairs of target genes. Real-time PCR was done by using 5 μl of qPCR master mix (SYBER Green, Puregene, Genetix), 2 picoM/μl of each primer per reaction in 10 μl of the final volume in the ABI QuantStudio Real-Time PCR machine. To analyze the change in the expression, the relative fold change mRNA expression of different genes was calculated using ^ΔΔCt value and then represented as ± standard error mean. Data normalization was done using rp49 as an internal control. Sequences of primers used in this study are mentioned in Supplementary information 1.

All statistical analyses were done using Prism software using either unpaired Student’s t-test to determine the significance between two independent variables or one-way ANOVA along with a suitable post-test in order to determine the significance in the variance of means between the different genotypes. Data are presented as ± standard error mean as values taken from three independent experiments. P values ≤ 0.05 (*) were considered statistically significant.

Huntington phenotype in Drosophila was obtained using the UAS-GAL4 system (Brand and Perrimon, 1993) for spatiotemporal control of any transgene. Here, we used an eye-specific GAL-4 driver, GMR-GAL4, to express 127 glutamate residues under the UAS promoter in the rhabdomeres, the neuronal cells of the Drosophila compound eye. Alteration in the transcription of various hsp70 gene families has been reported in neurodegenerative diseases (Shieh and Bonini, 2011; Malik et al., 2013), but a lack of information about the expression of alternatively spliced transcript variants encoding different hsc70 isoforms prompted us to check their expression. Hsc70 has five isoforms, namely, hsc70-1, hsc70-2, hsc70-3, hsc70-4 and hsc70-5 in Drosophila. Expression of all five isoforms was assessed by RT-PCR using isoform-specific primers in three independent experiments, followed by Student’s t-test to check the significance level of the transcript’s fold change, following which hsc70-1, hsc70-3, and hsc70-4 were found upregulated in GMR-GAL4:UAS-127Q background (Figure 1A), where hsc70-4 gene expression was increased with the highest value of significance. No significant change was observed for hsc70-2 and hsc70-5 isoforms in polyQ-expressing progenies.

Heat shock factor 1 (Hsf1) is activated by physiological stresses and activates heat shock genes required to mitigate stress. Studies have shown that activation of HSF1 in mammalian cells requires interaction with Hsc70 (Ahn et al., 2005), so we checked the expression of hsf in Drosophila, which was found to be downregulated. This observation suggests that under the polyQ condition, hsf is probably not involved in the regulation of hsc70 expression. As the aggregate proteins impair the HSF-mediated heat shock response (HSR) induction in diseased conditions (San Gil et al., 2020), and a recent study has shown that hsc70 is involved in the progression of the epithelial tumor in a hsf-independent manner (Singh et al., 2022). Therefore, these pieces of information suggest that the overexpression of these hsc70 isoforms might be independent of hsf in the polyQ-expressed condition.

Since the transcriptional activity of hsc70-4 was altered in the polyQ background, we looked at the protein levels of Hsc70. In the Western blot we detected two protein bands around 70kDa, the lower one was at ∼68 kDa, which is inducible Hsp70. It showed a typical pattern of elevated levels of Hsp70 in polyQ conditions in comparison to the wild-type. It also remained elevated in hsc70-4 overexpressed polyQ background, but downregulating hsc70-4 reduced its expression (in Supplementary information 4C). In the case of cognate Hsc70 (the upper band at ∼71 kDa), a significant upregulation of Hsc70 in the polyQ background was observed, which was reduced after expressing hsc70-4 RNAi in the polyQ condition. While the level of Hsc70 elevates after overexpressing hsc70-4 in the polyQ condition, the upregulation was not significant in comparison to the polyQ condition alone (Figure 1B). This observation was validated after calculating the fold changes in the Hsc70 from three independent experiments, and statistical significance was determined by using one-way ANOVA followed by Tukey’s multiple comparisons test (Figure 1C).

To find if constitutively expressed heat shock protein augments the polyQ degenerative phenotype, we downregulated it using UAS-Hsc70-4RNAi in GMR-GAL4:UAS-127Q background, and the eye phenotype of adult flies was observed. The compound eyes of Drosophila have approximately 800 repeating unit eyes or ommatidia that are organized into a hexagonal array and contain an invariant number of photoreceptor neurons and accessory cells that can be unambiguously identified by their position within the ommatidium. The typical array of hexagonal lattice in the adult eyes exhibited loss of ommatidial clusters and regular spacing because of multiple fusions after polyQ misexpression (Figures 1D, e–h), as compared to the wild-type eyes (Figures 1D, a–d). A significant improvement in the eye pigmentation and ommatidial arrangement of GMR-GAL4:UAS-127Q > UAS-Hsc70-4RNAi adult flies was observed, with a remarkable restoration of pigmentation loss and ommatidial disruption, even at the age of 20 days (Figures 1D, i–l), in comparison to GMR-GAL4:UAS-127Q flies (Figures 1D, e–h), where, along with the pigmentation loss and ommatidial disruption, there was a gradual reduction in eye size from 5-day-old to 20-day-old flies. In GMR-GAL4:UAS-127Q > UAS-Hsc70-4WT, no restoration in the eye pigmentation or ommatidial pattern of the degenerated eyes was observed (Figures 1D, m–p), further suggesting that only downregulation of hsc70-4 will significantly rescue the degenerated eye phenotype in polyQ background.

Before validating the results obtained with hsc70-4 RNAi (Figure 1D), we compared different internal controls to ensure that the results are specific for 127Q residues. For this, we compared the eye phenotype of wild-type with no PolyQ aggregates (Figures 2A, a), GMR-GAL4>UAS:20Q.HA with only 20 Polyglutamine residues (Figures 2A, b), and a single copy of GMR-GAL4/+ (Figures 2A, c), and did not observe any significant difference among them, suggesting that the degenerated eye phenotype in GMR-GAL4:UAS-127Q.HA is only due to the expanded repeats of 127 glutamine. Next, we used two different RNAi lines of hsc70-4 from BDSC (Bl-28709 and BL-34836) to downregulate hsc70-4 and one overexpression line (BL-5846) in order to upregulate hsc70-4 in polyQ background and compare the eye phenotype in the 5-day-old flies. The degenerating eye phenotype was categorized into three groups based on the severity of the morphological disruption in the eye, which were mild (Figures 2B, c,f), moderate (Figures 2B, a,e), and severe (Figures 2B, a,d) (Sanokawa-Akakura et al., 2010). After overexpressing hsc70-4 in GMR-GAL4:UAS-127Q condition, the mean percentage of flies with severe eye phenotype increased significantly. However, the percentage of flies with either mild or moderate eye phenotypes remained unchanged (Figure 2D). These observations were made by observing the eye phenotype of 5-day-old flies from their respective genotype, and the mean percentage of those adult flies was recorded from three independent experiments (where n ≥ 100) followed by Student’s t-test in between each phenotypic category, to check for their significance level.

Likewise, two hsc70-4 RNAi lines were used separately to downregulate it in GMR-GAL4:UAS-127Q background. Afterward, the transcript expression of hsc70-4 was checked in the same and compared to the polyQ, which was found to be significantly depleted, thus, validating the knockdown effect of these two RNAi lines (in Supplementary information 3). The resulting progenies showed the rescue in the degenerated eye phenotype, which was categorized further into three different groups—mild, moderate, and high—based on the strength of their phenotypic rescue. The mild rescued phenotype (Figures 2C, c,f) showed a slight improvement in the ommatidial arrangement and mild restoration of the pigmentation loss. Moderate rescued phenotype (Figures 2C, b, e) was demarcated as the recovery of eye pigmentation throughout the eye along with the recovered ommatidia arrangement, and strong rescued phenotype (Figures 2C, a,d) showed a remarkable improvement in the ommatidial arrangement and the recovery of the eye pigmentation.

After the phenotypic confirmation, we wanted to observe the eye surface topography in detail for the confirmation of the previously speculated rescued phenotype. So, we did the SEM imaging of the eye of the same genotypes. We observed severely disrupted eye morphology in the polyQ background, showing small asymmetrical ommatidia along with very few irregularly arranged mechanosensory bristles present (Figures 2E, b, f), in comparison to the wild-type eye (Figures 2E, a, e). The ommatidial disruption elevated in polyQ flies upon overexpressing hsc70-4 (Figures 2E, d, h), as the ommatidia were scarcely present and were smaller in size, with even lesser mechanosensory bristles. However, after downregulating hsc70-4 in GMR-GAL4:UAS-127Q background, the shape and arrangement of ommatidia were restored, as there was a clear boundary between the ommatidium and an increase in the abundance of mechanosensory bristles, which were in much more orderly fashion (Figures 2E, c, g). Thus, all these observations suggest that PolyQ aggregates impact neurodegeneration, which is probably associated with the dysregulation of hsc70-4, as downregulating it rescues the polyQ-mediated degenerated eye phenotype.

The cell-type differentiation of eye cells begins midway through the third instars stages, with the appearance of the ‘morphogenetic furrow,’ and posterior to it, progressive induction of cell fate occurs, with the development of photoreceptor neurons/rhabdomeres (first neuron R8, then R2 and R5, R3/R4, R1/R6, and finally R7 within each ommatidium). Aggregate accumulation in polyQ background led to the disruption in these photoreceptor neurons, leading to the structural and functional loss in the aggregates-infested eye of the Drosophila.

Having observed the phenotypic rescue after downregulating hsc70-4, we looked for the aggregates deposition in the eye imaginal discs of developing late third instar larvae and observed that the number and size of PolyQ aggregate decreased drastically in hsc70-4 RNAi-driven GMR-GAL4:UAS-127Q progenies (Figures 3A, b, e) in comparison to undriven GMR-GAL4:UAS-127Q (Figures 3A, a,d). In contrast, no change in the aggregates level was observed in the case of GMR-GAL4:UAS-127Q > UAS-Hsc70-4WT genotype (Figures 3A, c, f) [this observation was supported by the quantitative analysis (Figure 3B) using the mean value of intensity/pixel of PolyQ aggregates measured from the immunofluorescence images of the respective genotype which was mentioned in Figure 3A]. Hence, the downregulation of hsc70-4 in GMR-GAL4:UAS-127Q background led to a significant reduction in aggregates deposition in the developing eye disc.

To trace the effect of reduction in the PolyQ aggregates on degenerative eye phenotype, we checked the pattern of rhabdomeres, which can be easily visualized by ELAV staining, a pan-neuronal marker. Disruption in the rhabdomere pattern occurred due to polyglutamine toxicity in GMR-GAL4:UAS-127Q-driven conditions (Figures 3C, b). After downregulating hsc70-4 in polyQ-overexpressing eye discs, the rhabdomere arrangement was restored (Figures 2C, c) and resembled wild-type (Figures 3C, a). The pattern of rhabdomeres after overexpressing hsc70-4 remained disrupted (Figures 3C, d). We also checked the ommatidial arrangement using Dlg, a membrane marker, to validate previous observation (in Supplementary information 5A), which further validates the severe rhabdomere disruption after overexpressing hsc70-4. As the number of rhabdomeres reduces from seven in the wild-types to an average of four per ommatidium in polyQ condition, which furthers reduces to 3-rhabdomere/ommatidium in GMR-GAL4:UAS-127Q > UAS-Hsc70-4WT and downregulating, the hsc70-4 recovers this loss to an average of > 5-rhabdomeres per ommatidial unit in the eye disc (in Supplementary information 5B). Therefore, downregulating hsc70-4 restores the loss of rhabdomeres, thus leading to the recovery of the ommatidial disruption caused by polyQ background.

Axons of photoreceptor cells innervate into the optic lobe of the brain, and the proper connection of these axons is a hallmark of neuronal integrity (Andrews et al., 2010). Futsch is one of the microtubular-associated proteins present in the axons and can be used as an axonal marker to visualize the axonal connection. In the polyQ expressing eye disc of the third instar larvae, the axonal connections were disrupted severely, confirming that the entire morphology gets disrupted (Figures 3C, f) compared to the wild-type (Figures 3C, e). The axonal connections get restored in the eye disc of GMR-GAL4:UAS-127Q > UAS-Hsc70-4RNAi genotype (Figures 3C, g), but it remained disrupted in the case of GMR-GAL4:UAS-127Q > UAS-Hsc70-4WT genetic background (Figures 3C, h).

After visualizing the structural phenotypes, we checked the phototaxis response to investigate its functional implications. The phototaxis response was severely hampered in the polyQ-expressing eyes of 10-day-old flies. Less than 20% of the total adult population responded to the light, which was slightly improved after the overexpression of hsc70-4. However, the total percentage of flies responding to light remains less than half its population. On the other hand, downregulating hsc70-4 restored the phototaxis response in more than 60% of the adult population (Figure 3D). Therefore, as the rhabdomere pattern and the neuronal cytoskeletal arrangement were rescued after downregulation of hsc70-4, it led to the restoration of normal cellular processes that includes maintenance of cell shape and morphology, cytokinesis, adhesion, migration, phagocytosis, and endocytosis (Pelissier et al., 2003; Riggs et al., 2003; Myers and Casanova, 2008). Thus, the structural rescue in polyQ-expressing condition that occurs after reducing the transcript expression of hsc70-4 also leads to functional rescue.

Previous studies have reported the elevation of immune response in neurodegenerative conditions, which aggravates the physiological complications of the disease (Cao et al., 2013; Heneka et al., 2014), and it has been shown in Drosophila that polyQ enhances immune response (Dubey and Tapadia, 2018). This dysregulation happens due to the accumulation of Relish, trapped by PolyQ aggregates (Takano and Gusella, 2002), which ultimately causes the hyperactivation of the immune response and elevates the expression of AMPs. Studies in human and mouse models have shown the involvement of HSPA8 in the activation of NF-κB, thus, regulating the level of the innate immune response (Sheppard et al., 2014; Kumada et al., 2019). As hsc70-4 is the Drosophila ortholog of human HSPA8, we checked whether hsc70-4 regulates the hyperactivated immune response in polyQ-mediated neurodegeneration. The transcript levels of relish and AMPs in Oregon-R⁺ and GMR-GAL4:UAS-127Q were checked by qPCR and compared with the GMR-GAL4:UAS-127Q > UAS-Hsc70-4RNAi genotype. In comparison to Oregon-R⁺, the expression of relish and the different AMPs were significantly upregulated in GMR-GAL4:UAS-127Q. Maximum upregulation was observed in attacin and drosomycin, but diptericin, drosocin, and cecropin were upregulated too. Downregulating hsc70-4 in polyQ background resulted in a significant reduction in gene expression of relish and all the AMPs (Figure 4A).

The C-terminal of Relish protein harbors 49kDa inhibitory ankyrin repeats, which are cleaved off to release the transcriptionally active nuclear localizing Rel homology domain. So, when we compared the expression of Relish in GMR-GAL4:UAS-127Q progenies (Figures 4B, h) to the Oregon-R⁺ (Figures 4B, c), we found that it had significantly increased. After downregulating hsc70-4 in polyQ background, the expression of Relish reduced (Figures 4B, m), suggesting a significant drop in the Relish activation and accumulation. We further looked at the expression of HSP70 to see if the enhanced inflammatory response in the polyQ background upsurges the stress-inducible HSP70, after which a significant surge was observed in the level of HSP70 in the GMR-GAL4:UAS-127Q background (Figures 4B, g) and was reduced upon downregulating hsc70-4 in the same genetic background (Figures 4B, l). The quantitative analysis from the Western blot also validates the similar expression pattern of Hsp70 (in Supplementary information 4C). Furthermore, colocalization between these two proteins was detected after co-staining Relish and HSP70 in polyQ conditions (Figures 4B, i,j), which has also been reported earlier (Ran et al., 2004; Somensi et al., 2017). However, reduction in the expression of HSP70 and Relish upon downregulating hsc70-4 also depleted the colocalization signal (Figures 4B, n,o), which might be happening because the elevation of Relish in polyQ condition caused the accumulation of HSP70 and reduction in the inflammation depletes the HSP70 accumulation. Thus, these results suggest that hsc70-4 downregulation in the polyQ-expressing background may reduce hyperactivated innate immune response along with the reduction in the level of cellular stress.

In continuation to the earlier observation, the question arises as to how alteration in the expression of hsc70-4 in polyQ-overexpressed background affects the activation of Relish. Therefore, to find the answer, we looked into the possibility of the physical interaction of Hsc70 with Relish in polyQ-expressed conditions. Studies mention the role of Hsc70 in the clearance of aggregated proteins (Wyttenbach, 2004; Shieh and Bonini, 2011), physical interaction of Hsc70 with the polyglutamine stretch (Monsellier et al., 2015; Kim et al., 2016; Imler et al., 2019; Johnson et al., 2020), and polyglutamine aggregate interaction with NF-κB (Relish/Dorsal-Dif family) via HEAT-like motif and co-translocated into the nucleus (Takano and Gusella, 2002, Khoshnan et al., 2004). These studies suggest the possible role of polyglutamine aggregates in dysregulating both Hsc70 and NF-κB by physically interacting with both of them.

Therefore, we hypothesized that PolyQ-mediated dysregulation of NF-κB might be caused because of the hyperactivation of NF-κB by Hsc70. To confirm this, we needed to find out whether these three proteins interacted with each other in polyQ-overexpressed conditions. Thus, we performed a colocalization experiment in GMR-GAL4:UAS-127Q.HA background for PolyQ, Hsc70, and Relish, after which results showed colocalization between PolyQ and Hsc70 (Figures 5A, a–d), PolyQ and Relish (Figures 5A, e–h), and Relish and Hsc70 (Figures 5A, i–l) (yellow color in panel c indicates colocalization). Quantitative analysis of colocalization was done using Zen software along with the colocalization scattered plot (Figures 5A, d,h,l). Moreover, the overlap coefficient was also checked, which was ≥ 0.5, with a positive correlation of ≥ 0.1 in all three conditions (Figures 5A, c,g,k). Additionally, upon analyzing the surface-rendered image of the 3D projection of the eye disc stained for PolyQ, Hsc70, and Relish, the coappearance of Hsc70 and Relish was found within the PolyQ aggregate (data in the Supplementary information 6). Thus, these findings project the possibility of physical interaction of Relish and Hsc70 while PolyQ aggregates sequester both proteins.

For further confirmation of these interactions, we performed the co-immunoprecipitation experiment. To achieve this, we used anti-HA and anti-Relish antibodies to pull down PolyQ aggregates and Relish, respectively, from the protein sample extracted from the eye disc of third instars of GMR-GAL4:UAS-127Q.HA genetic background. Then, the interacting partners were detected by Western blotting of the eluted sample, denoted as the IP sample, and compared with the crude protein sample from GMR-GAL4:UAS-127Q.HA genetic background, which is represented as INPUT. After the immunoprecipitation of PolyQ aggregates using an anti-HA antibody, a band of ∼70 kDa in the blot of IP1 detected the Hsc70 labeled as 2 (Figures 5B, i). Also, three different isoforms of Relish were observed in the Western blot of the IP1, two sharp bands of whole-Relish protein labeled as 1 (∼110 kDa), C-Relish labeled as 1# (∼49 kDa), and a very faint band of N-Relish labeled as 1* (∼68 kDa). Thus, it was confirmed that Hsc70 and Relish were coprecipitated with PolyQ aggregates in the IP1. Similarly, after the immunoprecipitation of Relish in polyQ background, a band of Hsc70 around ∼70 kDa in the blot of IP2 was detected (Figure 5B, ii), which is inferred as the physical interaction of Hsc70 with Relish. PolyQ aggregate oligomers were detected in the Western blot of the IP2 sample (Figures 5B, iii), where several bands with a smeary appearance were observed due to the presence of fragmented PolyQ oligomer during the processing along with the three clear bands labeled as * which might be the PolyQ trapped with the IgG present in elute sample of IP2 (Figures 5B, iii). Therefore, the co-immunoprecipitation experiments confirmed the possibility of co-interaction of PolyQ aggregates, Hsc70, and Relish in polyQ-expressing background, which highlights one of the underlying causes of Hsc70 dysregulation in GMR-GAL4:UAS-127Q.HA expressing background.

c-Jun N-terminal kinase (JNK) is one of the central signaling cascades of the MAPK signaling pathway and functions as a regulator of cellular processes like proliferation, development, and apoptosis (Liu and Lin, 2005). So, we checked its expression in the polyQ-overexpressing eye discs. The level of activated JNK, i.e., p-JNK in the polyQ-overexpressing eye disc was considerably elevated in GMR-GAL4:UAS-127Q (Figures 6b,e), in comparison to Oregon-R⁺ (Figures 6a,d), which is similar to an earlier report (Dubey and Tapadia, 2018), which reduced significantly in GMR-GAL4:UAS-127Q > UAS-Hsc70-4RNAi (Figures 6c,f). This suggests that the elevated p-JNK level in the polyQ-expressing condition restores normal upon hsc70-4 downregulation. Quantitative analysis of the immunofluorescence images (data in the Supplementary information 7) further confirms the rescue in the p-JNK level after hsc70-4 downregulation in the polyQ condition. More than a 15-fold reduction in the p-JNK expression was recorded after downregulating hsc70-4 in polyQ background compared to the polyQ condition alone, whereas no significant fold change was seen when compared with the wild-type.

Cellular death ensues an increase in p-JNK, as it enhances the caspase activity (Borsello and Forloni, 2007; Perrin et al., 2009), so we looked for the real-time status of cellular death by using acridine orange (AO) live staining in eye discs. The eye disc of GMR-GAL4:UAS-127Q showed a drastic elevation in the AO puncta (Figure 6h) compared to Oregon-R⁺ (Figure 6g), suggesting a significant increase in the number of apoptotic cells in the eye disc, which again confirms that elevated JNK expression leads to increased apoptosis. However, the level reduced significantly in hsc70-4 downregulated condition almost to near normal (Figure 6i), which speculates a considerable rescue in the cellular apoptosis because of the back to the near-normal restoration of the p-JNK level after downregulation of hsc70-4 in polyQ condition.

Since the above studies were done using transgenic Drosophila having 127 glutamate residues to simulate the polyglutamine-associated disease condition like Huntington’s disease, so we validated it further by expressing mutant exon I of the huntingtin gene having 93 CAG repeats (UAS-httex1PQ93) in the eye discs of Drosophila in GMR-specified manner. As wild-type flies show normal eye pigmentation with a proper ommatidial arrangement (in Supplementary information 8A), these flies show age-wise progressive eye degeneration showing the loss of eye pigmentation along with the disrupted ommatidial arrangement (in Supplementary information 8B). The adults show almost complete loss of phototaxis response after the 10th day (Mallik and Lakhotia, 2009). Downregulating hsc70-4 in GMR-GAL4:UAS-httex1PQ93 background (in Supplementary information 8C) showed almost complete rescue in the loss of pigmentation and more compact arrangement of ommatidia. On the contrary, overexpressing the hsc70-4 in GMR-GAL4:UAS-httex1PQ93 background (in Supplementary information 8D) has an almost detrimental effect on the degenerated eye phenotype with higher pigmentation loss and more disrupted ommatidial arrangement.

These observations were further confirmed by the phototaxis assay performed on the 10-day-old flies (in Supplementary information 9). The population of GMR-GAL4:UAS-httex1PQ93 flies showed more than a 50% reduction in the phototaxis response, and overexpressing hsc70-4 did not show any significant changes. But approximately 70% of GMR-GAL4:UAS-httex1PQ93 > UAS-Hsc70-4RNAi flies were found to show a positive phototaxis response, which validates the rescue in the eye phenotype. Thus, mutant huntingtin flies validate the rescue effect of the hsc70-4 downregulation on the diseased eye phenotype.