SRCP1 conveys resistance to polyglutamine aggregation

SUMMARY

The polyglutamine (polyQ) diseases are a group of nine neurodegenerative diseases caused by the expansion of a polyQ tract that results in protein aggregation. Unlike other model organisms Dictyostelium discoideum is a proteostatic outlier, naturally encoding long polyQ tracts yet resistant to polyQ aggregation. Here we identify serine-rich chaperone protein 1 (SRCP1) as a molecular chaperone that is necessary and sufficient to suppress polyQ aggregation. SRCP1 inhibits aggregation of polyQ-expanded proteins allowing for their degradation via the proteasome where SRCP1 is also degraded. SRCP1’s C-terminal domain is essential for its activity in cells, and peptides that mimic this domain suppress polyQ aggregation in vitro. Together our results identify a novel type of molecular chaperone and reveal how nature has dealt with the problem of polyQ aggregation.

eTOC

Santarriaga et al. identify a molecular chaperone that is both necessary for Dictyostelium’s unusual resistance to polyglutamine aggregation and is sufficient to impart resistance to polyglutamine aggregation in other organisms. These findings describe how nature has dealt with the problem of polyglutamine aggregation.

INTRODUCTION

The polyglutamine (PolyQ) diseases are a group of nine inherited neurodegenerative diseases caused by the expansion of a polyQ repeat in the coding region of specific proteins (Williams and Paulson, 2008). PolyQ-expanded proteins misfold and lead to the formation of protein aggregates, ultimately resulting in the loss of specific types of neurons (Paulson et al., 2000). PolyQ aggregation is thought to be a key early event in polyQ toxicity and suppression of polyQ aggregation is one potential way to treat these diseases.

PolyQ aggregation has been studied in a wide variety of organisms ranging from yeast to primates (Bates and Davies, 1997; Kazemi-Esfarjani and Benzer, 2000; Meriin et al., 2002; Santarriaga et al., 2015; Satyal et al., 2000; Scherzinger et al., 1997; Tomioka et al., 2017). In each case, expression of a polyQ-expanded protein results in the formation of protein aggregates, with the exception of one organism, Dictyostelium discoideum (Malinovska et al., 2015; Santarriaga et al., 2015). Dictyostelium discoideum has a unique genome among sequenced organisms in that it encodes large numbers of homopolymeric amino acid tracts (Eichinger et al., 2005). Among the most common homopolymeric amino acid repeats are polyQ repeats, with 1,498 proteins containing 2,528 polyQ tracts of 10 or more glutamines (Eichinger et al., 2005). Endogenous polyQ tracts in Dictyostelium reach well beyond the disease threshold (~37Q), reaching repeat lengths of 80 glutamines, yet these proteins remain soluble (Santarriaga et al., 2015). Moreover, unlike other model organisms, overexpression of a polyQ-expanded huntingtin exon-1 construct (^GFPHtt^ex1Q103) does not result in protein aggregation (Malinovska et al., 2015; Santarriaga et al., 2015). Together, this suggests that Dictyostelium encodes novel proteins or pathways to suppress polyQ aggregation.

Here, we have analyzed known protein quality control pathways and show that Hsp70, autophagy, and the ubiquitin-proteasome system are not responsible for suppressing polyQ aggregation in Dictyostelium. Using a forward genetic screen, we identified a single Dictyostelium discoideum specific gene that is necessary for suppressing polyQ aggregation. This gene encodes serine-rich chaperone protein 1 (SRCP1), a small 9.1kDa protein that suppresses polyQ aggregation. In the presence of SRCP1 aggregation-prone polyQ proteins are degraded via the proteasome where SRCP1 is also degraded. Upon conditions where polyQ degradation is impaired, SRCP1 also suppresses polyQ aggregation, consistent with a chaperone function for SRCP1.

SRCP1 does not contain any identifiable chaperone domains, but rather utilizes a C-terminal domain that resembles amyloid (pseudo-amyloid) to suppress aggregation of polyQ-expanded proteins. Together, our findings provide insight into how Dictyostelium discoideum resists polyQ aggregation and identifies a new type of molecular chaperone that conveys resistance to polyQ aggregation.

RESULTS

SRCP1 is necessary for Dictyostelium discoideum to evade polyQ aggregation

Among known protein quality control pathways, molecular chaperones, autophagy, and the ubiquitin-proteasome pathway assist in combating polyQ aggregation (Koyuncu et al., 2017; Nath and Lieberman, 2017). To determine if these pathways suppress polyQ aggregation in Dictyostelium, we stably expressed ^GFPHtt^ex1Q103 in Dictyostelium and inhibited select protein quality control pathways. Inhibition of known protein quality control components, including Hsp70, autophagy, and the ubiquitin-proteasome system did not lead to an accumulation of ^GFPHtt^ex1Q103 puncta (Figure S1 1A-J. Related to Figure 1), suggesting that other protein quality control pathways are responsible for Dictyostelium’s unusual resistance to polyQ aggregation.

Figure 1: Identification of a novel protein that suppresses polyQ aggregation in Dictyostelium.

(A-C) A REMI screen identifies clones with ^GFPHtt^ex1Q103 aggregates. A REMI screen was performed in Dictyostelium expressing ^GFPHtt^ex1Q103, and clonal isolates were plated in 96-well plates prior to analysis by high-content imaging. Shown are representative negative (A) and positive (B, C) hits from the REMI screen.

(D) Quantification of the number of cells with ^GFPHtt^ex1Q103 puncta from (A-C) (n=4, **** p<0.0001). Error bars indicate SD.

(E-G) Deletion of SRCP1 results in ^GFPHtt^ex1Q103 aggregation in Dictyostelium. SRCP1 knockout cells were generated by homologous recombination and selected with blasticidin. ^GFPHtt^ex1Q103 was electroporated into wild-type (E) or SRCP1 knockout cells (F, G), selected with G-418, and imaged by fluorescent microscopy.

(H) Quantification of cells with ^GFPHtt^ex1Q103 puncta from SRCP1 +/+ and SRCP1 −/− cell lines (n=4, * p<0.05). Error bars indicate SD.

(I) Deletion of SRCP1 results in the accumulation of aggregated ^GFPHtt^ex1Q103. Wildtype or SRCP1 knockout cells expressing ^GFPHtt^ex1Q103 were lysed in NETN buffer and quantified by BCA protein assay. Samples were prepared with SDS and 40 μg of protein were used for filter trap assay. A representative image is shown (n=3).

(J) Quantification of ^GFPHtt^ex1Q103 present in the filter trap assay (I). The amount of ^GFPHtt^ex1Q103 present in (I) was quantified using ImageJ (n=3, ** p<0.01). Error bars indicate SD.

(K-L) SRCP1 expression rescues ^GFPHtt^ex1Q103 aggregation in SRCP1 knockout cells. Wild-type and SRCP1 knockout cells were transfected with either RFP alone (K) or ^RFPSRCP1 (L) and selected with hygromycin B. Cells were then subsequently transfected with ^GFPHtt^ex1Q103, selected with G-418, and imaged by fluorescent microscopy.

(M) Quantification of cells with ^GFPHtt^ex1Q103 puncta from SRCP1 knockouts expressing either RFP or ^RFPSRCP1 from (K-L) (n=4, * p<0.05). Error bars indicate SD.

(N-O) Deletion of SRCP1 results in the accumulation of insoluble species. Lysates from wildtype and SRCP1 knockout cells were quantified by BCA protein assay and subjected to differential centrifugation to isolate soluble and insoluble fractions. Samples were then run on SDS-PAGE and analyzed by western blot for either ubiquitin (N) or polyglutamine-expanded proteins (O).

We next utilized a forward genetic screen to identify genes responsible for suppressing polyQ aggregation in Dictyostelium. We performed a restriction enzyme-mediated integration (REMI) screen in Dictyostelium stably expressing ^GFPHtt^ex1Q103 and coupled it with high-content imaging to identify clonal isolates where a minimum of 5% of the cells contained ^GFPHtt^ex1Q103 puncta (versus the ~1% of cells that normally have aggregates). From this screen we identified a single uncharacterized Dictyostelium discoideum specific gene responsible for suppressing ^GFPHtt^ex1Q103 aggregation (Figure 1A-D). The protein encoded by this gene is a member of a large gene family of undefined function that encodes proteins with a serine-rich domain; therefore, we named it serine-rich chaperone protein 1 (SRCP1). To confirm that SRCP1 is responsible for suppressing ^GFPHtt^ex1Q103 aggregation, we generated SRCP1 knockout Dictyostelium strains. Knocking out SRCP1 led to no obvious growth phenotypes (data not shown), while expression of ^GFPHtt^ex1Q103 in SRCP1−/− cells resulted in the formation of numerous ^GFPHtt^ex1Q103 puncta (Figure 1E-H) that are insoluble via filter trap assay consistent with a role for SRCP1 in suppressing polyQ aggregation (Figure 1I, J). We next wanted to confirm that our results were due removal of SRCP1, and not an indirect effect. To accomplish this, we transformed SRCP1−/− cells with ^GFPHtt^ex1Q103 and either RFP or ^RFPSRCP1. Consistent with SRCP1 suppressing ^GFPHtt^ex1Q103 aggregation, expression of ^RFPSRCP1, but not RFP alone, prevented ^GFPHtt^ex1Q103 aggregation (Figure 1K-M). Together these data are consistent with SRCP1 being responsible for Dictyostelium discoideum’s unusual resistance to polyQ aggregation.

We next wanted to determine SRCP1’s endogenous function in Dictyostelium. Because Dictyostelium contain an abnormally high number of homopolymeric amino acid tracts, we hypothesized that SRCP1 played a role in maintaining proteostasis in Dictyostelium’s repeat-rich proteome. To test this hypothesis, we collected lysates from two independent SRCP1−/− cell lines and performed differential centrifugation to separate soluble and insoluble fractions. Consistent with SRCP1 playing an important role in maintaining proteostasis in Dictyostelium, we observed an accumulation of ubiquitinated species in the insoluble fraction of the SRCP1−/− cells consistent with the presence of polyubiquitinated protein aggregates (Figure 1N). In addition to polyubiquitinated proteins, the insoluble fraction was also enriched for endogenous polyQ proteins in the absence of SRCP1, suggesting that SRCP1 functions, at least in part, to maintain Dictyostelium’s polyQ-rich proteome (Figure 1O). Together these data suggest that SRCP1 plays an important role in maintaining proteostasis in Dictyostelium.

SRCP1 is sufficient to suppress polyQ aggregation

Because reducing polyQ aggregation in patients is one potential therapeutic avenue for the polyQ diseases, we next assessed the ability of SRCP1 to suppress ^GFPHtt^ex1Q74 aggregation in human cells. To test this, we co-transfected HEK293 cells with ^GFPHtt^ex1Q74 in the presence or absence of ^RFPSRCP1 and assessed the ability of ^RFPSRCP1 to prevent ^GFPHtt^ex1Q74 aggregation. Expression of ^GFPHtt^ex1Q74 resulted in the formation of ^GFPHtt^ex1Q74 puncta, while co-expression of ^RFPSRCP1 led to a dramatic decrease in the number and size of ^GFPHtt^ex1Q74 puncta (Figure 2A-C). Importantly, expression of ^RFPSRCP1 alone did not display any toxicity (Figure S2. Related to Figure 2). PolyQ aggregates migrate more slowly in SDS-PAGE gels and form high molecular weight aggregates (Scherzinger et al., 1999a). To determine if SRCP1 suppressed the formation of polyQ aggregates, we next analyzed the amount of ^GFPHtt^ex1Q74 aggregates in the presence or absence of ^RFPSRCP1. Expression of ^GFPHtt^ex1Q74 resulted in the presence of both soluble ^GFPHtt^ex1Q74 and ^GFPHtt^ex1Q74 aggregates (Figure 2D). Co-expression of ^RFPSRCP1 with ^GFPHtt^ex1Q74 greatly reduced the amount of ^GFPHtt^ex1Q74 aggregates as measured by SDS-PAGE, filter trap assay, and confocal microscopy, while having no effect on the levels of monomeric, soluble ^GFPHtt^ex1Q74 (Figure 2D-I). Consistent with SRCP1 having no effect on soluble polyQ protein, SRCP1 did not alter the levels of ^GFPHtt^ex1Q23 suggesting that SRCP1 selectively affects polyQ-expanded proteins that have a propensity to form aggregates (aggregation-prone) (Figure 2J, K). Because the polyQ diseases are neurodegenerative diseases, we next tested if SRCP1 prevented polyQ aggregation in human neurons. To this end, induced pluripotent stem cell (iPSC) derived neurons were transfected with either ^GFPHtt^ex1Q74 alone, or ^GFPHtt^ex1Q74 and ^RFPSRCP1. Similar to HEK293 cells, co-expression of ^RFPSRCP1 led to a dramatic decrease in ^GFPHtt^ex1Q74 puncta (Figure 2L, M). In addition to human neurons, we also assessed SRCP1’s ability to suppress polyQ aggregation in an intact animal. We injected zebrafish embryos with RNA encoding either ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1, and counted the number of ^GFPHtt^ex1Q74 puncta in spinal cord neurons. Similar to human cells, co-expression of ^RFPSRCP1 in zebrafish led to a significant reduction of ^GFPHtt^ex1Q74 puncta, consistent with SRCP1 suppressing ^GFPHtt^ex1Q74 aggregation (Figure 2N, O). Together, these data demonstrate that SRCP1 is sufficient to suppress polyQ aggregation in human cells and in zebrafish neurons.

Figure 2: SRCP1 reduces levels of aggregated, but not soluble ^GFPHtt^ex1Q74.

(A) Expression of SRCP1 in HEK293 cells results in a decrease in ^GFPHtt^ex1Q74 puncta. HEK293 cells were transfected with either ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1, and imaged by fluorescent microscopy. Co-expression of ^RFPSRCP1 decreased ^GFPHtt^ex1Q74 puncta. Shown is a representative image (n=3).

(B) Quantification of the number of ^GFPHtt^ex1Q74 puncta in (A). Images obtained for (A) were analyzed using ImageJ to determine the number of puncta in a field of confluent HEK293 cells (n=3, ** p<0.01). Results indicate that co-expression of SRCP1 significantly decreases the number of puncta. Error bars indicate SD.

(C) Quantification of the size of ^GFPHtt^ex1Q74 puncta in (A). Images obtained for (A) were analyzed using ImageJ to determine the size of puncta in cells expressing ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1 (n=3, ** p<0.01). Results indicate that co-expression of SRCP1 significantly decreases ^GFPHtt^ex1Q74 puncta size. Error bars indicate SD.

(D) SRCP1 suppresses the accumulation of aggregated ^GFPHtt^ex1Q74. HEK293 cells were transfected with either ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1. Cells were collected 48 hours post-transfection and analyzed by western blot. Two replicates of the experiment are shown (n=6).

(E) Co-expression of ^RFPSRCP1 decreases aggregated ^GFPHtt^ex1Q74. The amount of ^GFPHtt^ex1Q74 aggregates in (D) was quantified and standardized against a loading control (actin) (n=4, * p<0.05). Error bars indicate SD.

(F) Co-expression of ^RFPSRCP1 does not alter levels of soluble ^GFPHtt^ex1Q74. Quantification of soluble ^GFPHtt^ex1Q74 in (D). The amount of soluble ^GFPHtt^ex1Q74 was quantified and standardized against a loading control (actin) (n=4). Error bars indicate SD.

(G) SRCP1 reduces the levels of aggregated ^GFPHtt^ex1Q74 in filter trap assays. HEK293 cells expressing either ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1 were lysed in NETN buffer and quantified by BCA protein assay. Samples were prepared with SDS and 40 μg of protein were used for filter trap assay. A representative image is shown (n=3).

(H) Quantification of ^GFPHtt^ex1Q74 present in the filter trap assay (G). The amount of ^GFPHtt^ex1Q74 present in (G) was quantified using ImageJ (n=3, ** p<0.01). Error bars indicate SD.

(I) Soluble, but not punctate ^GFPHtt^ex1Q74 is present in cells co-expressing ^GFPHtt^ex1Q74 and ^RFPSRCP1. Representative confocal images of HEK293 expressing ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1 are shown. Cells expressing ^GFPHtt^ex1Q74 and ^RFPSRCP1 have similar levels of diffuse ^GFPHtt^ex1Q74, but reduced levels of aggregated ^GFPHtt^ex1Q74. Arrowheads indicate puncta. A representative image is shown (n=3).

(J) SRCP1 does not alter levels of ^GFPHtt^ex1Q23. HEK293 cells were transfected with either ^GFPHtt^ex1Q23 or ^GFPHtt^ex1Q74 in the presence and absence of ^RFPSRCP1. Cells were collected 48 hours post-transfection and analyzed by western blot. A representative image is shown (n=3).

(K) Quantification of protein levels of ^GFPHtt^ex1Q23 in (J). The amount of ^GFPHtt^ex1Q23 was quantified and standardized against a loading control (actin) (n=3). Error bars indicate SD.

(L) SRCP1 suppresses ^GFPHtt^ex1Q74 in iPSC-derived neurons. iPSC derived neurons were stained with Tuj1 (white) and astrocytes with GFAP (red). Nuclei are labeled with Hoechst (blue). Arrowheads indicate aggregated ^GFPHtt^ex1Q74; asterisks indicate diffuse ^GFPHtt^ex1Q74. Representative images are shown.

(M) Quantification of ^GFPHtt^ex1Q74 puncta present in (L) (n=3, * <p=0.05). Error bars indicate SD.

(N) SRCP1 suppresses ^GFPHtt^ex1Q74 aggregation in zebrafish spinal cord neurons. Zebrafish embryos were injected with RNA for ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1 and imaged 24 hours later for the presence of ^GFPHtt^ex1Q74 aggregates. Representative images are shown.

(O) Quantification of ^GFPHtt^ex1Q74 puncta present in (N). The number of ^GFPHtt^ex1Q74 aggregates were blindly scored for GFP puncta (n=10, **** p<0.0001). Error bars indicate SD.

SRCP1 prevents polyQ aggregation and allows for proteasomal degradation of aggregation-prone polyQ

In human cells co-expression of ^RFPSRCP1 and ^GFPHtt^ex1Q74 led to a dramatic decrease in ^GFPHtt^ex1Q74 aggregates but did not cause a corresponding increase in soluble ^GFPHtt^ex1Q74 levels (Figure 2D, F). This suggests that SRCP1 was either increasing the clearance of ^GFPHtt^ex1Q74 aggregates or promoting degradation of ^GFPHtt^ex1Q74 that misfolds prior to aggregation. One well established route for the clearance of both soluble and aggregated polyQ is autophagy (Koyuncu et al., 2017). To determine if expression of SRCP1 resulted in autophagic degradation of polyQ protein, we transfected HEK293 cells with ^GFPHtt^ex1Q74 in the presence or absence of ^RFPSRCP1 and treated with the autophagy inhibitor 3-MA 24-hours post-transfection. Consistent with previous publications, treatment with 3-MA led to a significant increase in soluble and aggregated ^GFPHtt^ex1Q74 (Qin et al., 2003; Ravikumar et al., 2002). However, under conditions of autophagy inhibition, co-expression of ^RFPSRCP1 did not lead to a significant increase in soluble ^GFPHtt^ex1Q74 levels, suggesting that SRCP1 does not promote ^GFPHtt^ex1Q74 clearance via autophagy (Figure 3A-C).

Figure 3: SRCP1 expression results in proteasomal degradation of aggregation-prone ^GFPHtt^ex1Q74.

(A) Autophagy inhibition does not significantly increase soluble ^GFPHtt^ex1Q74 in the presence of ^RFPSRCP1. HEK293 cells were transfected with either ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1 and treated with either DMSO or 5 mM 3-MA 24 hours post-transfection. Samples were collected 24 hours later, and levels of ^RFPSRCP1 and ^GFPHtt^ex1Q74 were determined by western blot. Shown is a representative blot (n=3).

(B) Quantification of soluble levels of ^GFPHtt^ex1Q74 in (A). The amount of soluble ^GFPHtt^ex1Q74 present in (A) was quantified using ImageJ (n=3, ** p<0.01). Error bars indicate SD.

(C) Autophagy inhibition does not increase punctate ^GFPHtt^ex1Q74 in HEK293 cells co-expressing ^GFPHtt^ex1Q74 and ^RFPSRCP1. HEK293 cells were transfected with either ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1 and treated with either vehicle (DMSO) or 5 mM 3-MA 24 hours post-transfection. Cells were imaged 24 hours after treatment by fluorescent microscopy.

(D) Proteasome inhibition stabilizes soluble ^GFPHtt^ex1Q74 in the presence of ^RFPSRCP1. HEK293 cells were transfected with either ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1 in the presence or absence of MG132. Samples were collected 18 hours later, and levels of ^RFPSRCP1 and ^GFPHtt^ex1Q74 were determined by western blot. Shown is a representative blot (n=3). Levels of ^GFPHtt^ex1Q74 increased in both the insoluble and soluble fraction in the absence of ^RFPSRCP1. In cells co-transfected with ^GFPHtt^ex1Q74 and ^RFPSRCP1 levels of soluble, but not aggregated ^GFPHtt^ex1Q74 increased, consistent with a chaperoning role for SRCP1.

(E) Quantification of soluble levels of ^GFPHtt^ex1Q74 in (D). The amount of soluble ^GFPHtt^ex1Q74 present in (D) was quantified using ImageJ (n=3, ** p<0.01). Error bars indicate SD.

(F) Proteasome inhibition leads to an increase in diffuse, but not punctate ^GFPHtt^ex1Q74 in HEK293 cells co-expressing ^GFPHtt^ex1Q74 and ^RFPSRCP1. HEK293 cells were transfected with either ^GFPHtt^ex1Q74, or ^GFPHtt^ex1Q74 and ^RFPSRCP1 for 24 hours prior to the addition of 10 mM MG132 or vehicle (DMSO). Images were taken 18 hours after addition of MG132 or DMSO by fluorescent microscopy.

Because inhibiting autophagy did not increase levels of soluble ^GFPHtt^ex1Q74 in the presence of SRCP1, we next turned our attention to the proteasome, the other major route for ^GFPHtt^ex1Q74 degradation (Michalik and Van Broeckhoven, 2004). To determine if the presence of SRCP1 resulted in proteasomal degradation of ^GFPHtt^ex1Q74, HEK293 cells were transfected with ^GFPHtt^ex1Q74 in the presence or absence of ^RFPSRCP1 and treated with proteasome inhibitor 24 hours post-transfection. Proteasome inhibition led to increased levels of both soluble and aggregated ^GFPHtt^ex1Q74 (Michalik and Van Broeckhoven, 2004; Miller et al., 2005; Waelter et al., 2001) (Figure 3D-F). However, co-expression of ^RFPSRCP1 in the presence of proteasome inhibition led to an even further increase in the amount of soluble ^GFPHtt^ex1Q74 when compared to proteasome inhibition alone. This indicates that in the presence of SRCP1 aggregation-prone, polyQ-expanded protein was being degraded by the proteasome (Figure 3D-F). Furthermore, in the presence of proteasome inhibition, SRCP1 prevented polyQ aggregation, consistent with SRCP1 functioning as a molecular chaperone (Figure 3D-F). Together, these data demonstrate that the expression of SRCP1 results in the degradation of aggregation-prone ^GFPHtt^ex1Q74 via the proteasome, but not the lysosome, and that SRCP1 acts as a molecular chaperone preventing ^GFPHtt^ex1Q74 aggregation upon conditions where ^GFPHtt^ex1Q74 degradation is impaired.

PolyQ accelerates the degradation of SRCP1

In the experiments where ^RFPSRCP1 and ^GFPHtt^ex1Q74 were co-transfected, we observed that ^RFPSRCP1 levels were difficult to detect. However, in experiments where ^RFPSRCP1 was transfected alone, we could easily detect ^RFPSRCP1. This led us to hypothesize that ^GFPHtt^ex1Q74 accelerated the turnover of ^RFPSRCP1. We next analyzed levels of ^RFPSRCP1 in human cells expressing ^GFPHtt^ex1Q74, ^RFPSRCP1, or ^GFPHtt^ex1Q74 and ^RFPSRCP1, and found that the presence of ^GFPHtt^ex1Q74 led to a dramatic decrease in ^RFPSRCP1 levels (Figure 4A, B). To determine if ^RFPSRCP1 is degraded by the proteasome, we transfected human cells with ^RFPSRCP1 in the presence or absence of proteasome inhibition and assessed levels of ^RFPSRCP1. Proteasome inhibition led to a marked stabilization of ^RFPSRCP1, consistent with ^RFPSRCP1 being degraded by the proteasome (Figure 4C, D). Together, these data demonstrate that ^RFPSRCP1 is degraded by the proteasome in a manner that is stimulated by the presence of ^GFPHtt^ex1Q74.

Figure 4: ^GFPHtt^ex1Q74 accelerates proteasomal degradation of SRCP1.

(A) ^RFPSRCP1 turnover is accelerated by ^GFPHtt^ex1Q74. HEK293 cells were transfected with ^RFPSRCP1, or ^GFPHtt^ex1Q74 and ^RfPSRCP1. Samples were analyzed by western blot. Shown is a representative blot (n=3).

(B) Quantification of levels of ^RfPSRCP1 in the presence or absence of ^GFPHtt^ex1Q74 from (A). Levels of ^RfPSRCP1 were quantified using ImageJ (n=3, *** p<0.001). Error bars indicate SD.

(C) SRCP1 is degraded by the proteasome. HEK293 cells were transfected with ^RFPSRCP1, or ^GFPHtt^ex1Q74 and ^RFPSRCP1 for 24 hours prior to the addition of 10 μΜ MG132 or vehicle (DMSO). Samples were collected 18 hours later, and levels of ^RFPSRCP1 were determined by western blot. Shown is a representative blot (n=3).

(D) Levels of ^RFPSRCP1 in the presence and absence of ^GFPHtt^ex1Q74 and MG132 from (C) were quantified using ImageJ (n=3, * p<0.05, **p<0.01). Error bars indicate SD.

SRCP1’s serine-rich domain is dispensable for SRCP1 function

SRCP1’s ability to suppress polyQ aggregation in the presence of proteasome inhibition is consistent with SRCP1 being a molecular chaperone. SRCP1 does not contain a canonical chaperone domain; however, it does contain a serine-rich N-terminus. Because DNAJB6 utilizes a serine-rich domain to suppress polyQ aggregation (Kakkar et al., 2016), we hypothesized that SRCP1’s serine-rich N-terminus may be important for suppressing polyQ aggregation. To test this, we generated constructs where all serine and threonine residues in SRCP1’s N-terminus were mutated to alanine (Figure 5A). However, mutation of SRCP1’s serine and threonine residues (^RFPSRCP1^ST1) did not disrupt SRCP1’s ability to suppress polyQ aggregation (Figure 5B-D). Together, these data demonstrate that SRCP1’s serine-rich domain is dispensable for its ability to suppress polyQ aggregation.

Figure 5: SRCP1’s serine-rich domain is dispensable for SRCP1 function.

(A) SRCP1 contains a serine-rich N-terminal region. Schematic depicting the N-terminal serine-rich region of SRCP1. The SRCP1^St1 construct has all N-terminal serine and threonine residues mutated to alanine.

(B) SRCP1’s serine-rich region does not suppress ^GFPHtt^ex1Q74 aggregation. HEK293 cells were transfected with ^GFPHtt^ex1Q74, ^GFPHtt^ex1Q74 and ^RfPSRCP1, or ^GFPHtt^ex1Q74 and ^RfPSRCP1^ST1. Samples were collected 48 hours after transfection and analyzed for levels of ^GFPHtt^ex1Q74 by western blot. No difference between ^RfPSRCP1 and ^RfPSRCP1^ST1 was detected. Shown are three replicates (n=9).

(C) The amount of ^GFPHtt^ex1Q74 aggregates in (B) was quantified and standardized against a loading control (actin) (n=9, * p<0.05). Error bars indicate SD.

(D) Serine and threonine residues in SRCP1’s serine-rich N-terminal region do not suppress ^GFPHtt^ex1Q74 puncta formation. HEK293 cells were transfected with ^GFPHtt^ex1Q74, ^GFPHtt^ex1Q74 and ^RfPSRCP1, or ^GFPHtt^ex1Q74 and ^RfPSRCP1^ST1. Samples were imaged 48 hours after transfection by fluorescent microscopy.

SRCP1 utilizes a C-terminal pseudo-amyloid domain to suppress polyQ aggregation

Because mutating ~40% of the residues in SRCP1’s N-terminal serine-rich domain did not cause any measurable defect in SRCP1 activity, we next turned our attention to SRCP1’s C-terminal region to identify a region that may be important for SRCP1’s chaperone function. One striking feature of SRCP1’s C-terminal region is a highly hydrophobic, glycine-rich region that resembles amyloid. Peptides that resemble amyloid can form mixed amyloid with amyloid-forming proteins and influence amyloid formation (Cheng et al., 2012; Sato et al., 2006). To determine if SRCP1 encodes any predicted amyloidogenic domains, we utilized in silico approaches, including Tango, FISH Amyloid, FoldAmyloid, PASTA 2.0, and AmylPred2 and identified two potential amyloid-forming regions in SRCP1’s C-terminal region, which we termed pseudo-amyloid domains (Figure 6A, B; Figure S3. Related to Figure 6). We next mutated regions within SRCP1’s pseudo-amyloid domains (^RFPSRCP1^61−70A and ^RFPSRCP1^71−80A) and determined their ability to suppress ^GFPHtt^ex1Q74 aggregation. Consistent with SRCP1’s first pseudo-amyloid domain being important for suppressing polyQ aggregation, the ^RFPSRCP1^61−70A mutant lost the ability to suppress ^GFPHtt^ex1Q74 aggregation, while the ^RFPSRCP1^71−80A mutant retained full activity (Figure 6C-E). This is consistent with amino acids 61–70, but not 71–80 of SRCP1 containing critical residues for SRCP1 function (Figure 6C-E). To gain more detailed insight into residues important for SRCP1 function, we next mutated individual amino acid residues in SRCP1 from 61–70 to alanine and determined their ability to suppress ^GFPHtt^ex1Q74 aggregation via fluorescence microscopy. Two individual point mutations, ^RFPSRCP1^V65A and ^RFPSRCP1^I69A, resulted in a decrease in SRCP1 function, consistent with an important role for suppressing polyQ aggregation (data not shown, Figure 6F).

Figure 6: SRCP1’s pseudo-amyloid domain prevents polyQ aggregation.

(A) SRCP1’s C-terminal region contains two predicted to form amyloid. Schematic depicting the sequence of SRCP1’s C-terminal region. Multiple in silico programs predict an aggregation-prone, amyloid-forming region in SRCP1. Amino acids that are predicted to form amyloid are indicated by asterisks.

(B) Sequence of SRCP1’s two pseudo-amyloid domains that are mutated in (C-E) or used as peptides (N, O).

(C) Amino acids 61–70 are essential for suppressing ^GFPHtt^ex1Q74 aggregation. HEK293 cells were transfected with ^GFPHtt^ex1Q74, ^GFPHtt^ex1Q74 and ^RFPSRCP1, ^GFPHtt^ex1Q74 and ^RFPSRCP1^61−70A, or ^GFPHtt^ex1Q74 and ^RFPSRCP1^71−80A Cells were imaged 48 hours post-transfection by fluorescent microscopy.

(D) Amino acids 61–70 of SRCP1 are essential for suppressing polyQ aggregation. HEK293 cells were transfected with ^GFPHtt^ex1Q74, ^GFPHtt^ex1Q74 and ^RFPSRCP1, ^GFPHtt^ex1Q74 and ^RFPSRCP1^61−70A, or ^GFPHtt^ex1Q74 and ^RFPSRCP1^71−80A. Samples were collected 48 hours post-transfection and analyzed by western blot (n=7).

(E) Levels of aggregated ^GFPHtt^ex1Q74 from (D) were quantified using ImageJ (n=4, * p<0.05, ** p<0.01, *** p<0.001). Error bars indicate SD.

(F) Amino acids V65 and I69 are essential for suppressing ^GFPHtt^ex1Q74 aggregation. HEK293 cells were transfected with ^GFPHtt^ex1Q74, ^GFPHtt^ex1Q74 and ^RFPSRCP1, ^GFPHtt^ex1Q74 and ^RFPSRCP1^V65A, or ^GFPHtt^ex1Q74 and ^RFPSRCP1^I69A. Cells were imaged 48 hours post-transfection by fluorescent microscopy.

(G) A peptide derived from SRCP1’s pseudo-amyloid domain suppresses polyQ aggregation. In vitro Htt^Q46 aggregation assays were performed in the absence or presence of increasing ratios of SRCP1 peptide to Htt^Q46. A representative image is shown (n=5).

(H) SRCP1 decreases Htt^Q46 fibrils. In vitro Htt^Q46 aggregation assays were performed with Htt^Q46 and SRCP1 61–80 peptide (3:1 peptide to Htt^Q46) for 5 hours and imaged by EM.

(I) SRCP1 decreases aggregated Htt^Q46. In vitro Htt^Q46 aggregation assays were performed with Htt^Q46 and SRCP1 61–80 peptide (3:1 peptide to Htt^Q46) for 5 hours. Samples were then prepared with SDS, subjected to filter trap assay, and analyzed via western blot for polyglutamine.

(J) SRCP1 decreases larger Htt^Q46 species. In vitro Htt^Q46 aggregation assays were performed with Htt^Q46 and SRCP1 61–80 peptide (3:1 peptide to Htt^Q46) for 5 hours. Samples were analyzed by dynamic light scattering.

(K) SRCP1 peptide delays but does not prevent Htt^Q46 amyloid fiber formation. In vitro Htt^Q46 aggregation assays were performed with Htt^Q46 and SRCP1 61–80 peptide (3:1 peptide to Htt^Q46) for 72 hours and imaged by EM.

(L) SRCP1 peptide does not prevent Htt^Q46 aggregation over 72 hours. In vitro Htt^Q46 aggregation assays were performed with Htt^Q46 and SRCP1 61–80 peptide for 72 hours (3:1 peptide to Htt^Q46). Samples were then prepared with SDS, subjected to filter trap assay, and analyzed via western blot for polyglutamine.

(M) A SRCP1 peptide delays but does not prevent the formation of larger Htt^Q46 species. In vitro Htt^Q46 aggregation assays were performed with Htt^Q46 and SRCP1⁶¹⁻⁸⁰ peptide for 72 hours (3:1 peptide to Htt^Q46). Samples were analyzed by dynamic light scattering.

(N) A peptide of SRCP1’s amino acids 61–70 suppresses polyQ aggregation. In vitro Htt^Q46 aggregation assays were performed with Htt^Q46 and either SRCP1 61–80 peptide, SRCP1 61–70 peptide, or SRCP1 71–80 peptide (3:1 peptide to Htt^Q46). Htt^Q46 alone was used as a positive control. A representative image is shown (n=3).

(O) Amino acids V65 and I69 are essential for SRCP1 61–70 peptide to suppress polyQ aggregation. In vitro Htt^Q46 aggregation assays were performed with Htt^Q46 and either SRCP1 61–70 peptide or SRCP1 61–70 peptide with amino acids V65 and I69 mutated to alanine (VI-A) (3:1 peptide to Htt^Q46). Htt^Q46 alone was used as a positive control. A representative image is shown (n=3).

We next wanted to determine if SRCP1 could directly suppress polyQ aggregation in vitro. Because we were unable to generate soluble recombinant SRCP1 protein, we generated a 20-amino acid peptide that mimics SRCP1’s C-terminal pseudo-amyloid region. This peptide remained soluble in vitro (data not shown) and was sufficient to suppress aggregation of Htt^Q46 in vitro as measured by Thioflavin-T fluorescence (Figure 6G). We next wanted to confirm that our SRCP1 derived peptide was inhibiting Htt^Q46 aggregation and not disrupting the binding of Thioflavin-T to Htt^Q46 aggregates. To accomplish this, we analyzed Htt^Q46 aggregate formation at a 5 hour time point by electron microscopy (EM) and found that the addition of the SRCP1 peptide decreased Htt^Q46 fiber formation and led to the appearance of spherical structures similar to previously described soluble oligomers (Tsigelny et al., 2008) (Figure 6H). The presence of these spherical structures and decrease in fibers also correlated with a decrease in Htt^Q46 aggregates as measured by filter trap assay and an increase in smaller species via dynamic light scattering (DLS) (Figure 6I, J). This, coupled with the observation that the 1.5:1 and 3:1 molar ratios of peptide to Htt^Q46 resulted in a delay, but not prevention of Htt^Q46 aggregation (Figure 6G), led us to hypothesize that the SRCP1 peptide delayed, but did not prevent, Htt^Q46 aggregation. To test this hypothesis, we analyzed Htt^Q46 aggregation in the presence or absence of the SRCP1 peptide after a 72-hour incubation. Consistent with the SRCP1 peptide delaying, but not preventing aggregation, Htt^Q46 aggregates were detected by EM and filter trap analysis (Figure 6K, L) and the smaller species observed via DLS at 5 hours were no longer detectable at 72 hours (Figure 6M). To further analyze SRCP1’s pseudoamyloid domain in vitro, we tested peptides of the two predicted individual domains (amino acids 61–70 or 71–80) within the SRCP1 peptide. Consistent with our cell data, a peptide that encodes amino acids 61–70 was sufficient to suppress Htt^Q46 aggregation in vitro whereas a peptide that consists of amino acids 71–80 did not (Figure 6N). Similarly, a peptide consisting of amino acids 61–70 with V65 and I69 mutated to alanine resulted in a loss in activity in accordance with the fluorescent microscopy data in cells (Figure 6F, O). Together, these data support a role for SRCP1’s C-terminal pseudo-amyloid domain in contributing to SRCP1’s ability to suppress polyQ aggregation.

SRCP1 Reverses Neurite Shortening in Huntington’s Disease (HD) iPSC-derived Neurons

Because SRCP1 suppressed polyQ aggregation, we next wanted to determine if SRCP1 reversed disease phenotypes associated with huntingtin aggregation. One such phenotype is the degeneration of neurons, which is believed to contribute to the early pathology of Huntington’s disease (DiFiglia et al., 1997; Li et al., 2001). To this end, we utilized two independent HD iPSC-derived neurons with either 60 or 180 glutamines (HD iPSC Consortium, 2012). Neurons derived from HD iPSCs exhibit shortened neurites (Chae et al., 2012; Kaye and Finkbeiner, 2013; The Hd iPsc Consortium, 2012), so we tested if SRCP1 could reverse this phenotype. We expressed ^RFPSRCP1 in HD iPSC-derived neurons and quantified the length of the neurites in the presence or absence of ^RFPSRCP1. Overexpression of SRCP1 significantly increased neurite length and reversed the phenotype seen in the untreated HD iPSC-derived neurons (Figure 7A, B). This is consistent with SRCP1 preventing aggregation-related phenotypes and suggests that SRCP1 may prevent toxic events associated with human HD.

Figure 7: SRCP1 rescues defects in neurite outgrowth in HD iPSC-derived neurons.

(A) SRCP1 increases neurite outgrowth in HD iPSC-derived neurons. HD iPSC-derived neurons were transfected with ^RFPSRCP1 or the transfection reagent alone (control). HD iPSC-derived neurons were stained with Tuj1 (green). Nuclei labeled with Hoechst. Representative images depict neurite length variability within the two treatment conditions.

(B) Quantification of (A) shows a significant increase in neurite length μm) in HD iPSC-derived neurons treated with ^RFPSRCP1 compared to the transfection reagent alone (control) (n=3, **** <p=0.0001). Error bars indicate SD.

DISCUSSION

The model organism, Dictyostelium discoideum, is a proteostatic outlier that naturally encodes for a large number of homopolymeric amino acid tracts. Among the repeat sequences encoded in the Dictyostelium genome, polyQ is among the most abundant, and the length of these polyQ repeats can reach well within the disease range in humans (Eichinger et al., 2005). Here we have identified SRCP1 a chaperone that suppresses polyQ aggregation in Dictyostelium (Figure 1). We show that SRCP1 is sufficient to suppress polyQ aggregation in multiple systems, including Dictyostelium, HEK293 cells, iPSC-derived human neurons, and zebrafish spinal cord neurons (Figure 2). SRCP1 accomplishes this by suppressing polyQ aggregation ultimately resulting in degradation of aggregation-prone polyQ proteins by the proteasome (Figure 2–4). In addition, SRCP1 is also degraded by the proteasome and its degradation is enhanced by the presence of polyQ-expanded protein (Figure 3, 4). We show that SRCP1’s serine-rich N-terminus is dispensable for SRCP1 activity in cells (Figure 5). However, we did identify a C-terminal pseudo-amyloid domain that is necessary for SRCP1’s ability to suppress polyQ aggregation (Figure 6C-F). Finally, we show that expression of SRCP1 in Huntington’s Disease (HD) iPSC-derived neurons rescues defects in neurite outgrowth (Figure 7A, B). Together, our findings identify a new type of chaperone that effectively prevents the accumulation of polyQ aggregates and provides insight into how nature has dealt with the problem of polyQ aggregation.

SRCP1 is a Dictyostelium discoideum specific protein

While no previous reports have analyzed SRCP1 function and SRCP1 does not have any readily recognizable domains, our data indicate it functions as a molecular chaperone. SRCP1 is a member of a large class of Dictyostelium discoideum genes that encode small proteins (~6–11kDa) with serine-rich regions that have been implicated in Dictyostelium’s developmental process (Vicente et al., 2008). This suggests that Dictyostelium’s developmental process may have led to proteins and pathways that allow Dictyostelium to suppress aggregation of its repeat-rich proteome. Unlike most of these genes, SRCP1 levels are not sharply developmentally regulated and instead, SRCP1 is expressed throughout the Dictyostelium life cycle (Scaglione, unpublished results). In the future, more work is needed to understand the function of other members of this gene family to determine if they play similar roles to SRCP1 in suppressing protein aggregation. Because Dictyostelium encode for amino acid repeats for every amino acid except tryptophan, it will be important to expand studies beyond polyQ.

SRCP1 is a molecular chaperone

SRCP1 does not contain any readily identifiable chaperone domains. Instead we identified a C-terminal domain in SRCP1 that resembles amyloid. This pseudo-amyloid domain is necessary for SRCP1’s chaperone activity in cells (Figure 6C-F), and peptides that mimic its C-terminus alter the rate of amyloid formation in vitro (Figure 6G-O). Previous work identified that peptides that efficiently integrated into amyloid fibers (Cheng et al., 2012; Eskici and Gur, 2013; Sato et al., 2006) served as potent inhibitors of amyloid formation. Supportive of this, our in vitro data suggests SRCP1 inhibits aggregation via a direct interaction, indicating that SRCP1’s pseudo-amyloid domain may bind ^GFPHtt^ex1Q74 once it has converted to an aggregation-prone amyloid fold. Alternatively, SRCP1 may recognize aggregated species and function as a disaggregase dismantling amyloid once it has formed. Additional work will be necessary to determine what species SRCP1 recognizes and at what stage of aggregation SRCP1 functions.

In addition to SRCP1’s chaperone activity, the presence of SRCP1 results in the degradation of ^GFPHtt^ex1Q74 via the proteasome where SRCP1 is also degraded. The mechanism by which SRCP1 and polyQ are targeted to the proteasome is unclear. Proteasomal substrates are typically targeted to the proteasome via ubiquitination (Kwon and Ciechanover, 2017). However, it is unclear if SRCP1 targets ^GFPHtt^ex1Q74 for ubiquitination or is ubiquitinated itself. Alternatively, some proteins, like the neurodegenerative disease protein tau, are degraded in the absence of ubiquitination (Blair et al., 2013; David et al., 2002; Grune et al., 2010). This raises the possibility that SRCP1 may stimulate ^GFPHtt^ex1Q74 degradation independent of ubiquitin signaling. In addition to a potential direct role in targeting polyQ-expanded proteins for degradation, SRCP1 may also function in a more indirect manner. SRCP1 could also function to maintain polyQ-expanded protein in a soluble state where it can be recognized by the cell’s degradation machinery. This would be more similar to proteins like small heat shock proteins that prevent protein aggregation or protein disaggregases like Hsp104 that block prion seed formation (Park et al., 2014; Shorter and Lindquist, 2006; Zhu and Reiser, 2018). In the future the identification of members of the proteostatic network that function with or downstream of SRCP1 will be important.

What polyQ species does SRCP1 recognize?

While there remains some debate about whether large aggregates observed in Huntington’s disease are toxic or protective, it is increasingly clear that smaller oligomeric huntingtin species are toxic (Hoffner and Djian, 2014). In addition to smaller oligomeric species causing toxicity, monomeric polyQ has also been shown to cause toxicity in cells (Nagai et al., 2007). Interestingly, the toxicity of this monomeric polyQ protein is conformation dependent as only polyQ that has adopted a β-sheet conformation induces toxicity in cells (Nagai et al., 2007). This is interesting because SRCP1’s pseudo-amyloid domain is necessary for its ability to prevent polyQ aggregates in cells (Figure 6C-F) and peptides that mimic this domain can delay the formation of polyQ aggregates in vitro (Figure 6G-O). As a conformational change to a β-sheet-rich structure is an integral component of amyloid formation, we hypothesize that SRCP1’s pseudo-amyloid domain forms mixed amyloid nand selectively identifies monomeric polyQ that has adopted this β-sheet conformation. Consistent with this we see that SRCP1 has no effect on the levels of ^GFPHtt^ex1Q23 or soluble ^GFPHtt^ex1Q74 suggesting SRCP1 can selectively identify monomeric, aggregation-prone polyQ protein. (Figure 2D-F, J,K). Additionally, we observe that overexpression of SRCP1 results in proteasomal degradation of aggregation-prone ^GFPHtt^ex1Q74 (Figure 3D-F). This supports the notion that SRCP1 is selectively influencing the monomeric version of ^GFPHtt^ex1Q74 as the proteasome is inefficient at degrading aggregated polyQ protein (Holmberg et al., 2004). SRCP1 could be influencing monomeric levels of ^GFPHtt^ex1Q74 through multiple mechanisms, potentially stabilizing monomers prior to oligomerization or perhaps recognizing oligomers and promoting the formation of monomers. In the future, more mechanistic details are needed to determine how SRCP1 influences the oligomeric status of polyQ.

SRCP1 provides insight into therapies

SRCP1’s ability to selectively allow for the degradation of aggregation-prone, but not soluble ^GFPHtt^ex1Q74 (Figure 2, 3) has important therapeutic implications. For example, huntingtin has essential roles in development (Duyao et al., 1995; Nasir et al., 1995; Zeitlin et al., 1995), and ataxin-3, the polyQ protein that causes Spinocerebellar ataxia type 3, plays an important role in autophagy (Ashkenazi et al., 2017). These results argue that depletion of soluble polyQ proteins may be problematic. Instead, development of molecules that selectively identify and clear aggregation-prone, polyQ-expanded proteins may be a better therapeutic avenue. In the future, delineation of the molecular details that mediate the SRCP1/polyQ interaction may lead to the development of novel therapeutics to suppress polyQ aggregation.

STAR * Methods

Key Resources Table

Contact For Reagent and Resource Sharing

For all reagents and resource requests, please email the Lead Contact, Dr. Scaglione (mscaglione@mcw.edu).

Experimental Model and Subject Details

Mammalian Cell Culture

Human embryonic kidney293 (HEK293) cells were grown at 37°C and 5.2% CO ₂. HEK293 cells were maintained in Dulbecco’s Modified Eagle’s Medium (Gibco by Life Technologies) supplemented with 10% fetal bovine serum (Atlanta biologicals) and 1% Penicillin-Streptomycin (Gibco by Life technologies).

Dictyostelium discoideum Cell Culture

Dictyostelium discoideum AX4 cells were maintained in shaking cultures at 22°C in HL5 (17.8 g peptone, 7.2 g yeast extract, 0.54 g Na₂HPO₄, 0.4 g KH₂PO₄, 130 μl B12/Folic acid, 20 ml of 50% w/v glucose, ampicillin 100 μg/ml, pH 6.5) media. Cells were maintained at a density no greater than 6×10⁶ cells/ml. For growth on bacteria, Dictyostelium cells between a density of 1×10⁶ - 6×10⁶ cells/ml, were diluted 1:100, 1:1000, and 1:10,000. Approximately, 500 μl of dilutions were spread on bacterial plates. To prepare bacterial plates, K. aerogenes (Dictybase) was grown at room temperature for two days and then plated on freshly made SM plates (35 ml per 100 mm Petri dish; 1 Liter: 10 g glucose, 10 g proteose peptone, 1 g yeast extract, 1 g MgSO₄*7H₂O (or 0.5 g MgSO₄), 1.9 g KH₂PO₄, 0.6 g K₂HPO₄, 20 g Agar) and incubated overnight at room temperature.

Bacterial Cell Culture

To obtain plasmid DNA, overnight bacterial cultures were grown overnight at 37°C with a selection anti biotic at 50 μg/ml. One Shot TOP10 chemically competent E.coli (Invitrogen by ThermoFisher Scientific) cells were used for any cloning procedures. For all other plasmids XL-10 E.coli cells were used. All E. coli cells were stored at −80°C until necessary.

iPSC Culture

Induced pluripotent stem cells (iPSCs) used were karyotypically normal and mycoplasma negative. iPSCs were cultured in T25 ultra-low attachment culture flasks (Corning) as non-adherent neural progenitor cell aggregates in Stemline medium (Sigma-Aldrich) supplemented with 100 ng/ml epidermal growth factor (Miltenyi), 100 ng/ml fibroblast growth factor (Stem Cell Technologies), 5 μg/ml heparin (Sigma-Aldrich), and 0.5% N2 (Life Technologies) in humidified incubators at 37°C and 5.0% CO₂. Subsequently, neural progenitor cells were differentiated into neurons and astrocytes as previously described (Ebert et al., 2013). Briefly, cells were dissociated with TrypLE (Life Technologies) and seeded onto Matrigel (Corning) coated glass coverslips at 2.6×10⁴ cells/cm². For differentiation, GFAP+ astrocytes and Tuj1+ neurons were grown in Neurobasal medium (Life Technologies) supplemented with 2% B27 (Life Technologies) and Antibiotic-Antimycotic (Life Technologies) for 1–2 weeks. Following differentiation, iPSC-derived astrocytes exhibit functional calcium responses to ATP (McGivern et al., 2013) and potassium currents (Ebert et al., 2013), and iPSC-derived neurons exhibit NR2B NMDA receptor expression (Schwab et al., 2017) and appropriate electrophysiological properties (2012; The Hd iPsc Consortium, 2012)

Zebrafish Colony Management

Zebrafish were housed in a closed circulating system using water purified by reverse osmosis, and subjected to 10% daily flush. Conductivity was maintained at 800 μS. Particulates were removed by drum filtration. The light:dark cycle was 14L:10D. Fish were fed three times per day with hatched artemia. For experiments, embryos derived from group crosses of the ZDR strain were used.

Method Details

Expression Constructs

For Dictyostelium expression huntingtin exon-1 with 103 glutamines was cloned into pTX-GFP (Dictybase) using KpnI and XbaI (^GFPHtt^ex1Q103). For restriction-enzyme mediated integration, the kanamycin resistance cassette was swapped in place of the ampicillin resistance cassette using BamHI and SpeI sites in the pDM323 (Dictybase) vector resulting in a kanamycin resistant pDM323 (pDM323 KAN) plasmid. ^GFPHtt^ex1Q103 with 103 glutamines was cloned into the generated pDM323 KAN plasmid. The pBSR3 plasmid (Dictybase) was used for the integration step. SRCP1 (DDB_G0293362), SRCP1 with amino acids 61–70 or 71–80 mutated to alanine, and SRCP1 with serine and threonine mutated to alanine were synthesized for mammalian expression into the PS100049 vector by Blue Heron. Individual point mutants of SRCP1 61–70 were generated using QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent). Plasmids for mammalian expression encoding pEGFP huntingtin exon-1 with 23 (^GFPHtt^ex1Q23) or 74 (^GFPHtt^ex1Q74) glutamines were obtained from Addgene.

Dictyostelium Cell Transformation

Transformations were performed via electroporation as described previously (Knecht and Pang, 1995). Briefly, 5×10⁶ cells were spun down at 500 g for 5 minutes at 4°C. Cells were then washed three times with ice-cold H-50 buffer (20 mM HEPES, 50 mM KCl, 10 mM NaCl, 1 mM MgSO₄, 5 mM NaHCO₃, 1 mM NaH₂PO₄), resuspended in 100 μl of ice-cold H-50 buffer, and combined with 10 pg of DNA. Cell and DNA mixture was added to a pre-cooled 1 mm cuvette and electroporated at 0.85 kV/25 μF/time constant 0.6 msec twice with about 5 seconds between pulses. Cells were then incubated on ice for 5 minutes and added to 10 ml of HL5 media on 10 cm plates. Appropriate selection drugs were added the following day and included in media going forward. Selection drugs include G-418 (Gibco by Life technologies) at 10 μg/ml, blasticidin (GoldBio) at 10 μg/ml (except where indicated), and hygromycin B (ThermoFischer Scientific) at 50 μg/ml.

Restriction Enzyme-Mediated Integration (Kuspa, 2006)

10 μg of the pBSR3 construct was linearized with BamHI, electroporated into AX4 cells along with 50 units of DpnII, and selected for one week with blasticidin at 4 μg/ml. Following selection, pBSR3 cells were electroporated with pDM323(Kan) ^GFPHtt^ex1Q103 construct and selected for one week with blasticidin at 4 μg/ml and G-418 at 10 μg/ml. To obtain individual colonies, cells were then plated on SM bacterial plates in serial dilutions. This was performed until the number of clones desired was obtained. Upon the appearance of Dictyostelium colonies on the bacterial plates, colonies were picked and grown in HL5 media in 96-well plates. Once cells had reached confluency, 96-well plates were washed with ice-cold starvation buffer (0.1 M MES, 0.2 mM CaCl₂, 2 mM MgSO₄, pH 6.8) twice and screened for aggregates using the ArrayScan High-Content Imaging System (ThermoScientific). Cells with greater than 5% aggregates were considered “hits” and grown up in 10 cm plates.

For gene identification, genomic DNA was first isolated from Dictyostelium “hit” clones. Genomic DNA was then digested with either ClaI, HindIII, or Bglll for an hour at 37°C, purified and ligated for 30 min utes at room temperature. Ligated DNA was then transformed into One Shot TOP10 chemically competent E.coli. Bacterial colonies were screened by restriction digest for insert and then sent for sequencing.

Dictyostelium Knockout Generation

Dictystelium knockout vectors were generated following the StarGate® Acceptor Vector pKOSG-IBA-Dicty1 system (iba-lifesciences) (Wiegand et al., 2011). The knockout vector was then linearized, electroporated into AX4 cells, and selected for one week with blasticidin at 10 μg/ml. To isolate clones, the electroporated cells were plated on SM bacterial plates. Individual colonies were then picked and grown up to confluency in 10 cm dishes. To screen knockouts, genomic DNA was obtained and utilized in PCR to confirm blasticidin insertion as well as flanking regions of SRCP1. To confirm SRCP1 knockouts, clones were electroporated with ^GFPHtt^ex1Q103 and selected with G-418 for one week at 10 μg/ml. Cells were then imaged by fluorescent microscopy and analyzed for ^GFPHtt^ex1Q103 aggregates.

Differential Centrifugation in Dictyostelium

For differential centrifugation, 1 × 10⁷ cells were washed with 1x PBS and lysed with NETN ((0.5% NonidetP-40, 150 mM NaCl, 50 mM Tris, and protease inhibitor (GoldBio)). Samples were centrifuged at 15,000 rpm for 30 minutes at 4°C. The supernatant (soluble fractio n) was removed and subjected to BCA protein assay. Remaining pellet (insoluble fraction) was washed three times with NETN, resuspended in 1x Laemmli buffer (4x stock: 40% Glycerol, 240 mM Tris/HCl pH 6.8, 8% SDS, 0.04% bromophenol blue, 5% beta-mercaptoethanol) and analyzed by western blot.

Genomic DNA Isolation in Dictyostelium (Pilcher et al., 2007)

Approximately, 3 × 10⁷ cells were pelleted by centrifugation for 4 minutes at 500 g at room temperature and washed twice with 1x PBS. Cells were resuspended in 1 ml of nuclei buffer (20 mM Tris-HCl pH 7.4, 5 mM MgOAc, 1 mM EDTA pH 8.0, 5% (w/v) sucrose, 1 mM EGTA) and lysed by the addition of 200 μl of 20% Triton X-100 and incubation on ice for 20 minutes. To obtain genomic DNA, lysates were centrifuged at 12,000 g for 5 min at 4°C. The supernatant was removed and pellet was resuspended in 300 μl of proteinase K Buffer (100 mM Tris-HCl pH 7.4, 5 mM EDTA pH 8.0, 0.1 mg ml⁻¹ proteinase K, 1% (v/v) SDS) and incubated at 65°C for 30 min. Nucleic acids were extracted by adding an equal volume (300 μl of phenol:chloroform (1:1). Samples were inverted to mix and centrifuged at 12,000 g for 20 minutes at 4°C. The aqueous (upper) layer was transferred into a fresh tube and mixed with one volume (300 μl) of chloroform by inversion. Samples were centrifuged at 12,000 g for 10 minutes at 4°C. The aqueous (upper) layer was transferred into a fresh tube and precipitated overnight at −20°C with 2.5 volumes (7 50 μl of ice-cold 100% ethanol. DNA was pelleted by centrifugation at 12,000 g for 30 minutes at 4°C. S upernatant was removed and pellet was resuspended in 100 μl of TE (10 mM Tris-HCl pH 7.4, 1 mM EDTA) containing 10 μg ml⁻¹ RNase A and incubated at room temperature for 15 minutes. DNA was precipitated again overnight at −20°C by the addition 1/10 volume (10 μl of 3 M NaOAc and 2.5 volumes (250 μl of ice-cold 100% ethanol. DNA was pelleted by centrifugation at 12,000 g for 30 minutes at 4°C. The supernatent was removed and pellet was washed with two volumes (250 μl of ice-cold 70% ethanol. DNA was pelleted by centrifugation at 12,000 g for 2 minutes at 4°C. Super natant was removed and pellet was dried at room temperature for 10 minutes. DNA pellet was resuspended in 50 μl of TE pH 7.4 and stored at 4°C.

Chemical Inhibition in Dictyostelium

For proteasome inhibition, 5×10⁶ cells were incubated with either DMSO or 100 μΜ MG132 (Sigma-Aldrich) for 18 hours. Following treatment, cells were imaged and/or harvested by spinning at 500 g at room temperature for 5 minutes. Cell pellets were washed three times with 1x PBS and lysed with ice-cold NETN (with protease inhibitor). Lysates were then sonicated twice for 10–15 seconds. To verify inhibition, samples were run on SDS-PAGE and analyzed by western blot for ubiquitin.

For autophagy inhibition, 1×10⁷ cells were incubated either with vehicle or 150 μM NH₄Cl (Sigma-Aldrich) for 8 hours. Following treatments, cells were imaged by fluorescent microscopy and plated on coverslips at a density of 2×10⁶ cells per ml and allowing cells to adhere. Cells were then washed three times with 1x PBS and fixed with 100% ice-cold methanol at −20°C for 10 m inutes. Methanol was aspirated and cells were washed twice with 1x PBT (0.1% Triton X, 0.5% BSA in 1x PBS), incubated in blocking buffer (1% Triton X, 2% BSA in 1x PBT) for 30 minutes at room temperature, and put in primary Rb anti ATG8a (courtesy of Jason King, University of Sheffield) (1:5000) overnight at 4°C. Following primary cells were washed three times with 1x PBT and incubated in secondary goat anti rabbit (Jackson ImmunoResearch Laboratories; 711–166-152) for 2 hours at room temperature. Following secondary, cells were washed an additional three times with 1x PBT and mounted on slides with ProLong Gold Antifade reagent (Invitrogen by ThermoFisher Scientific). Slides were imaged with a Nikon eclipse 90i confocal microscope.

For Hsp70 inhibition, 1×10⁷ cells were incubated either with DMSO or 80 μM VER155008 (ApexBio) for 24 hours. Following treatments, cells were imaged and/or harvested by spinning at 500 g at room temperature for 5 minutes. Cell pellets were washed three times with 1x PBS and lysed with ice-cold NETN (with protease inhibitor). Lysates were then sonicated twice for 10–15 seconds. To verify inhibition, samples were run on SDS-PAGE and analyzed by western blot for ATG8.

Cell Viability Assay

HEK293 cells were transfected as described with RFP or ^RFPSRCP1. Following transfections, cells were diluted to 500,000 cells/mL with 50,000 cells plated per well in duplicate from each transfection reaction in a 96-well plate. Serial dilutions were then performed down each column of the plate down to 390 cells per well. Cells were cultured for an additional 24 hours, then lysed using the CellTiter-Glo Kit (Promega).

Luminescence was quantified between 565–700nm wavelengths on a Spark microplate reader (Tecan) to determine cell viability.

Mammalian Cell Transfections

Transfections were performed with Lipofectamine 2000 (Invitrogen by ThermoFisher Scientific) and adapted from the manufacturer’s instructions. Briefly, cells were plated on either 12- well or 6- well plates and transfected at 50–70% confluency. For 12- well transfections, 0.833 μg of DNA per well was mixed with 50 μl of OptiMEM (Gibco by Life Technologies) media and incubated for 5 minutes at room temperature. Lipofectamine 2000 in a 2:1, μl of lipofectamine: μg DNA ratio was mixed with 50 μl of OptiMEM media and incubated for 5 minutes at room temperature. DNA and Lipofectamine were mixed and incubated for 15 minutes at room temperature. Fresh media was added to cells prior to the addition of DNA and lipofectamine. For 6- well transfections, the same protocol is followed except that 1.25 μg of DNA per well and 100 μl of OptiMEM for dilution was used. Media was changed 24 hours post-transfection and cells were harvested 48 hours post-transfection. Prior to harvesting, HEK293 cells were washed three times with ice-cold 1x PBS. Samples were then lysed with either 150 μl of ice-cold NETN (with protease inhibitor) or 1x Laemmli Buffer and sonicated twice for 5–8 seconds.

iPSC transfections were performed with mixed cultures of astrocytes and neurons adhered to matrigel-coated coverslips, seeded at 2.6×10⁴ cells/cm². iPSC transfections were performed using Lipofectamine 2000 according to the manufacturer’s instructions and recommended DNA concentrations for a 24-well transfection.

Chemical Inhibition in HEK293 cells

For proteasome inhibition, 30 hours post-transfection cells would be replenished with DMEM containing either DMSO or 10 μΜ MG132 (Sigma-Aldrich) for 18 hours. Prior to harvesting, HEK293 cells were washed three times with ice-cold 1x PBS and imaged by fluorescent microscopy with a 20x objective using the Evos FL Auto Imaging System. Cells were then lysed with either 150 μl of ice-cold NETN (with protease inhibitor) or 1x Laemmli Buffer and sonicated twice for 5–8 seconds.

For autophagy inhibition, 24 hours post-transfection cells would be replenished with DMEM and treated with either DMSO or 5 mM 3-Methyladenine (Sigma-Aldrich) for an additional 24 hours. Prior to harvesting, HEK293 cells were washed three times with ice-cold 1x PBS and imaged by fluorescent microscopy with a 20x objective using the Evos FL Auto Imaging System. Cells were then lysed with either 150 μl of ice-cold NETN (with protease inhibitor) or 1x Laemmli Buffer and sonicated twice for 5–8 seconds.

Microscopy

For HEK293s’, slides were prepared by placing coverslips in 6- well plates and incubating with poly-L-lysine hydrobromide (Sigma-Aldrich) for an hour at room temperature. Coverslips were rinsed with 1x PBS three times prior to plating HEK293 cells. HEK293 cells were then transfected and fixed by incubation in 4% paraformaldehyde for 20 minutes at room temperature. Coverslips were mounted with ProLong Gold Antifade reagent and imaged with a 20x objective using the Evos FL Auto Imaging System or with a 40x objective using the Leica TCS SP5 Confocal Microscope System. For confocal, Z-stack images were obtained at 0.5- m intervals at 1024 × 1024 pixel resolution and merged using Fiji.

Plated iPSC derived cultures of astrocytes and neurons were fixed with 4% paraformaldehyde in PBS for 20 minutes at room temperature, 3–5 days after transfection. Cells were permeabilized and blocked prior to antibody labeling. Nuclei were labeled with Hoechst nuclear dye. Primary antibodies used were rabbit anti-Tuj1 (Covance MRB-435P), chicken anti-Tuj1 (Gene Tex, GTX85469) and mouse anti-GFAP (Cell Signaling, 3670). Images were taken using an upright Nikon fluorescent microscope with a 40X objective. The number of neural cells with GFP positive puncta was quantified using NIS Elements object quantification tool (Nikon) and neurite length was measured using NIS Elements length measurement tool (Nikon).

For Dictyostelium cells, 1×10⁷ cells were washed three times with ice-cold starvation buffer and then imaged with a 20x objective using the Evos FL Auto Imaging System.

Zebrafish Analysis

To test the effect of ^RFPSRCP1 on ^GFPHtt^ex1Q74 aggregation in zebrafish, each plasmid was used for in vitro synthesis of mRNA (mMessage mMachine T7 transcription kit, ThermoFisher Scientific). Newly fertilized embryos were injected with either 200 pg mRNA each of ^RFPSRCP1 and ^GFPHtt^ex1Q74, or RFP alone and ^GFPHtt^ex1Q74. Embryos were developed until 54 hours post fertilization, at which time live specimens were anesthetized in Tricane, embedded in 1% low-melt agarose, and subjected to confocal microscopy. Spinal cord images were taken at 100 μm x 100 μm and a scan depth of 50 μm. Single image planes from 488 nm and 568 nm excitation were collected. Images were then processed for threshold analysis using ImageJ and the total number puncta were scored.

SDS-PAGE and Western Blot

Following sonication, lysates were subjected to BCA protein assay (ThermoFisher Scientific) and prepared for loading by the addition of 4x laemmli buffer and boiling for four minutes. Samples then loaded on SDS-polyacrylamide gels, ran at 175 V, and transferred onto an Immuno-blot PVDF membrane (Biorad) overnight at 30 V. Membranes were blocked in 5% milk in TBS with 0.1% Tween (TBST) and incubated in primary antibody overnight at 4°C. Following primary, membranes were washed for 10 minutes at room temperature three times with TBST and incubated in secondary for an hour at room temperature. Membranes were then washed for 10 minutes at room temperature three times with TBST and then incubated in buffer for enhanced chemiluminescence (50 mM Na₂HPO₄; 50 mM Na₂CO₃; 10 mM NaBO3·4H2O; 250 mM luminol; 90 mM coumaric acid).

Anti-GFP (Invitrogen by ThermoFisher Scientific; A11122), anti-RFP (Invitrogen by ThermoFisher Scientific; MA5–15257), anti-polyglutamine-expansion (EMD Millipore; MAB1574), and anti-ubiquitin (BD Pharmingen; 550944) were used at 1:1000. Peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch Laboratories; 111–035-045; 115–035-174) were used at 1:5,000 dilution. Anti-β-Actin was used as a loading control at 1:1000 (Invitrogen by ThermoFisher Scientific, PA121167).

Filter Trap Assay

For HEK293 lysates, BCA protein assay was used to determine protein concentration. Forty micrograms of protein were diluted up 90 μl with NETN (with protease inhibitor) 10 μl of 10% SDS was added to protein sample and diluted to 1 ml with 1% SDS in 1x PBS, and filtered through a 0.2 mm cellulose acetate membrane filter (Sterlitech) using a DHM-48 filter trap hybridization manifold. Membrane was then washed with 1 ml of 1% SDS in 1x PBS and analyzed by western blotting (Scherzinger et al., 1999b; Wanker et al., 1999).

Huntingtin Exon-1 46Q Purification

Huntingtin exon-1 with 46 glutamines (Htt^Q46) in pET-32 was obtained from Addgene. Htt^Q46 was grown in BL21 cells at 37°C to an optical density of 0.6 and indu ced with IPTG at 1 mM overnight at 16°C. After indu ction cells were spun down at 7,000 rpm for 5 minutes and resuspended in resuspension buffer (15 mM Tris-HCl buffer, pH 8.0). For lysis, lysozyme was added and cells were incubated at 4°C for 45 minutes. To obtain solub le fraction, lysates were spun down at 12,000 rpm for 10 minutes, after which supernatant was added to 3 ml of Ni-beads (GoldBio) per 100 ml of lysate and tumbled for 4 hours at 4°C. Beads were then washed three t imes with resuspension buffer and then washed three more times with wash buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 1 mM PMSF, 1 mM EDTA). Protein was then eluted off beads by tumbling overnight at 4°C in 25 mL o f wash buffer with 250 mM imidazole.

Thioflavin Aggregation Assays

Aggregation assays were performed as previously described in Kakkar, V. et al. Briefly, 15 μM Htt^46Q was mixed with 90 pM SRCP1 peptide (GenScript) (1:6, 1:3, 1:1.5) and 10 pM Thioflavin-T (Sigma-Aldrich). All samples were prepared on ice in buffer containing 20 mM Tris-HCl pH 8.0, 50 mM NaCl, 2 mM CaCl₂. Enterokinase (New England BioLabs) was added at 1.6 units/sample to initiate aggregation. All samples were prepared on ice and 50 μL aliquots were transferred to a flat, black 384-well plate and allowed to aggregate at 37°C for 16–18 hours. Fluorescence was measured wit h the excitation at 440 nm and emission at 480 nm, every 5 minutes using a Tecan Plate reader.

Filter Trap Analysis of Recombinant Protein

For filter trap assays, Htt^46Q was used at a concentration of 15 μM alone or with 90 μM of SRCP1 peptide. Peptide alone was used at 50 μM. All samples were prepared on ice in buffer containing 20 mM Tris-HCl pH 8.0, 50 mM NaCl, 2 mM CaCl₂. Enterokinase at 1.6 units/sample was added to initiate the reactions and samples were allowed to aggregate at 37°C for 5 and 72 hours. Approximately 40 μg of protein sample were then used for filter trap assay.

Dynamic Light Scattering (DLS)

For dynamic light scattering (DLS) measurements, Htt^46Q was used at a concentration of 15 μM alone or with 90 μM of SRCP1 peptide. Peptide alone was used at 50 μM. All samples were prepared on ice in buffer containing 20 mM Tris-HCl pH 8.0, 50 mM NaCl, 2 mM CaCl₂. Enterokinase at 1.6 units/sample was added to initiate the reactions and samples were allowed to aggregate at 37°C for 5 and 72 hours. Following aggregation, 10 μL of sample was loaded into a Hellma Analytics QC High Precision Cell Quartz Suprasil Cuvette (cuvette model #ZMV1002, light path of 1.25 × 1.25 mm). DLS measurements were collected at 25 °C by using a Malvern Zetasizer μV with a 50 mW laser at 830 nm, using a detector angle of 90°. The laser power was set to 30% power and integration time was set to 100 seconds. Each measurement consisted of 50–60 collections and are representative of n = 2, where each n consisted of at least 2 technical replicates. DLS spectra were analyzed and visualized in R (Team, 2016) by using the tidyverse (Wickham, 2017), scales (Wickham, 2016), and directlabels (Hocking, 2017) packages.

Electron Microscopy (EM)

For EM, samples were prepared on ice in buffer containing 20 mM Tris-HCl pH 8.0, 50 mM NaCl, 2 mM CaCl2. Htt^46Q was used at a concentration of 30 μM alone or with 174 μM of SRCP1 peptide. Peptide alone was used at 50 μM. Enterokinase at 2.2 units/sample was added to initiate the reactions and samples were allowed to aggregate at 37°C for 5 and 72 hours.

Following aggregation assays, freshly ionized 400mesh Formvar / carbon coated copper grids were floated onto 10μl droplets of sample and stain for 2 minutes to allow adsorption of the sample to the formvar /carbon film. After adsorption, the sample was then wicked away from the edge of grid surface and the grid was immediately floated on the droplet of negative stain (2% aqueous uranyl acetate) for 1 minute. The stain was wicked away from the edge of the grid and the grid was then allowed to air dry. Samples were examined in a Hitachi H600 TEM.

Quantification and Statistical Analysis

For the high-content imaging, HCS program was used to develop custom algorithms to detect puncta and number of cells. Using the two algorithms, the HCS program quantified the percent of cells with aggregates. For general microscopy analysis, ImageJ was used to quantify the number of cells per puncta and the puncta size where indicated. ImageJ was also used for the quantification of western blots. Briefly, band of interest would be measured for intensity and normalized to the intensity of the loading control.

Values were then entered in Graph Pad where they were analyzed by either an ANOVA for multiple comparisons and Students t-tests for 1:1 comparisons followed by Tukey’s post-hoc analysis. Differences were considered statically significant at a p-value less than 0.05.

For control iPSCs, four biological replicates from two independent differentiations were analyzed and for HD iPSCs, four biological replicates were analyzed from four separate differentiations using two independent HD lines.

Highlights.

• SRCP1 is necessary for Dictyostelium’s resistance to polyglutamine aggregation

• SRCP1 is sufficient to convey resistance to polyglutamine aggregation in other organisms

• SRCP1 reduces the level of SDS-insoluble but not soluble polyQ-expanded protein

• Residues in SRCP1’s C-terminus are necessary for its function