The modulation of the exterior polyQ-induced rough eye phenotype served as a readout of neurotoxicity and its modulation by modifier candidates. Additionally, integrity of eye structure and the connection between aggregation and neurodegeneration can be analysed in greater detail by histological and immunohistochemical approaches on the one hand and biochemical filtration methods on the other hand. We utilised both for certain candidate genes in order to gain insight into the modes of action of the discovered modifiers.
5.3.1 Evaluation of tissue integrity of SCA3tr-Q78-shRNA-coexpressing flies
A suitable approach to study morphology and thereby integrity of the compound eye is frontal sectioning of fly heads and the subsequent histological staining of the tissue.
The morphological changes of the eye surface of polyQ flies originate from the degeneration of the underlying tissue of the compound eye. Namely, photoreceptor and adjacent cell architecture are severely deteriorated, leading to the overt rough eye phenotype and eventually collapse of the eye.
Selected suppressor candidates were used in order to assess preservation of deeper eye tissue in contrast to retinal damage in polyQ flies. Flies co-expressing polyQ protein and enhancer RNAi could not be analysed due to severe degeneration of eye structure. As expected, silencing of genes leading to improvement of REP was also able to alleviate the detrimental effects of polyQ expression in the retina, albeit to different degrees (Figure 13). For example, the two RNAi lines for knockdown of Hsc70-4 (Transformant IDs 50222/26465) both ameliorated cell demise in the retina of SCA3 flies in line with their suppression effect on REP in the RNAi screen. Nevertheless, the effect of tissue preservation was significantly different for these shRNAs, with line 50222 showing almost wild-type retinal extend and structure whereas line 26465 exhibited good external phenotype mitigation yet retinal organisation presented diffuse and width was decreased.
5.3.2 Filter retardation analysis of RNAi influence on polyQ aggregates
Due to the proposed toxicity of certain polyQ aggregate species and the possible influence of modifier gene knockdown thereon, it was intended to biochemically study whether there is an impact of silencing of screened modifier genes on the levels of SDS-insoluble polyQ aggregates as previously described [276].
In order to address this question, expanded polyQ protein was co-expressed together with the candidate shRNA in the eye. Where possible, offspring were collected and SDS-treated head lysates were subjected to filter retardation assay (FRA) analysis. Filtration of the protein lysates through a nitrocellulose membrane would lead to trapping of aggregates exceeding a certain size (0.2 µm), allowing for assessment of polyQ aggregate load. Of course, lethal enhancers could not be analysed due to the absence of viable progeny. The hypothesis was that REP-suppressing candidates would decrease toxic aggregate levels, whereas enhancer shRNA expression would result in higher aggregate load. Nevertheless only a minor number of suppressor candidates was observed to effectively ameliorating aggregate number (6/34, 17.6 %) with shRNA against CG3808 being the most potent aggregation suppressor. On the contrary, a considerable large group did not change Figure 13. Influence of selected shRNAs on tissue integrity of SCA3tr-Q78 fly head sections.
Control sections feature intact retinal tissue in GMR-GAL4 and serious degeneration of eye tissue in SCA3tr-Q78 fly sections. Following introduction of shRNA lines suppressing Ataxin-3-induced REP, retinal thickness is improved to different degrees and retinal tissue architecture is restored towards GMR control situation.
All scale bars apply to 50 µm respectively.
aggregate levels significantly or even enhanced (2/34, 5.8 %) the cellular polyQ aggregate burden after normalisation to polyQ control. Additionally, the majority of enhancer RNAi candidates exhibited a trend towards decreasing aggregate levels with only a few gene knockdowns resulting in higher aggregate load. However, absolute number of significant changes is smaller in enhancers compared to suppressors. Concluding from these results, aggregate levels in this experimental setting do not appear to correlate to exterior REP and vice versa (Figure 14).
dve-s EloA ppk14 bwa CG17048 RpII15 LvpH
aux trbd 5PtaseI Hsc70-4(26465) 13F
G CG5
687 CG3808 CG9153 CG14619 CG15618 CG16890 Brd8 MRG
15 50
DNApol- Hsc70-1 CG17919 Hsc70-4 (50222) Cad88C Dab Hrb27C Doa LanB1 Droj2 CG4266 roq Hop
CG6758 chm CCS timeout CG9601 DAAM CG33128 Pkc CG34372 Dnz1 Hsf
AR-2 CG4288 CG15339 CG31048 CG15534 l(3)neo38 Marf CG12935
-2 -1 0 1
fold change (log) normalized to control *** ** **
** * *
**
* *
n 2 n 2
Suppressors Enhancers
Figure 14. Analysis of SDS-insoluble SCA3tr-Q78 aggregate load with shRNA modifiers.
(A) Exemplary filter retardation analysis for visualisation of aggregate load. GMR-GAL4 control is negative, SCA3tr-Q78 lysates exhibit heavy aggregation which is mitigated by suppressor shRNA. (B) Densitometric measurement of filter retardation analysis compared to SCA3tr-Q78 for suppressors and enhancers of polyQ-induced toxicity. n ≥ 3 if not indicated otherwise. Significant changes are: * p < 0.05; ** p < 0.01: *** p < 0.001.
A
B
5.3.3 RNAi effects on polyQ inclusions in situ
Expression of expanded polyQ protein leads to formation of protein aggregates in the compound eye as shown before and verified biochemically by the filter retardation experiments. Aggregation of toxic gene products is a hallmark of polyQ disease and considered to be at least in part causative for neurotoxicity and degeneration. On a microscopic level, inclusion bodies in retinal cells are detectable, presumably consisting of diverse polyQ aggregate species and various other proteins recruited to the agglomerate. In the eyes of the offspring of polyQ flies and flies concomitantly expressing modifier shRNA we intended to address the question whether improvement or worsening of the REP corresponds to the aggregate load in situ.
In the frontal head sections representative for select candidates we were however not able to show a robust connection between decrease of inclusions in the eye tissue and change of the REP. Three of the analysed RNAi lines featured a reduction of SDS-insoluble aggregates in filter retardation assays (Brd8, CG17919, CG33128, Figure 15A-C) with the latter being an enhancer of the REP. The two suppressor lines still featured inclusions, however to a seemingly decreased amount. The CG33128 shRNA (Figure 15C) led to an enhanced number of inclusions and concomitantly had the worst tissue integrity. The suppressor lines at least presented with improved retinal morphology compared to polyQ alone. Two lines with increased SDS-insoluble aggregate load in filter retardation analysis, CG17048 and Hsc70-4 (26465) (Figure 15D, E), had clearly delimited small inclusions in moderate numbers in parallel with an overall well-preserved tissue integrity. Thus, there was no clear trend such that modified eye structure and therefore altered neurotoxicity have their origin in a differential number of immunohistochemically detectable aggregates.