Prion-like propagation of protein aggregates has recently emerged as being implicated in many neurodegenerative diseases. The goal of this protocol is to describe, how to use the nematode C. elegans as a model system to monitor protein spreading and to investigate prion-like phenomena.
Prions are unconventional self-propagating proteinaceous particles, devoid of any coding nucleic acid. These proteinaceous seeds serve as templates for the conversion and replication of their benign cellular isoform. Accumulating evidence suggests that many protein aggregates can act as self-propagating templates and corrupt the folding of cognate proteins. Although aggregates can be functional under certain circumstances, this process often leads to the disruption of the cellular protein homeostasis (proteostasis), eventually leading to devastating diseases such as Alzheimer’s disease (AD), Parkinson’s disease (PD), Amyotrophic lateral sclerosis (ALS), or transmissible spongiform encephalopathies (TSEs). The exact mechanisms of prion propagation and cell-to-cell spreading of protein aggregates are still subjects of intense investigation. To further this knowledge, recently a new metazoan model in Caenorhabditis elegans, for expression of the prion domain of the cytosolic yeast prion protein Sup35 has been established. This prion model offers several advantages, as it allows direct monitoring of the fluorescently tagged prion domain in living animals and ease of genetic approaches. Described here are methods to study prion-like behavior of protein aggregates and to identify modifiers of prion-induced toxicity using C. elegans.
Many neurodegenerative diseases, including Alzheimer’s disease (AD), Parkinson’s disease (PD), Amyotrophic lateral sclerosis (ALS), and transmissible spongiform encephalopathies (TSEs), are associated with aggregation-prone proteins and are hence collectively known as protein misfolding disorders (PMDs). TSEs or prion diseases constitute a unique class of PMDs in that they can be infectious in both humans and animals1. At the molecular level, prions replicate by recruiting and converting monomeric α-helix-rich host-encoded cellular PrP (PrPC) into the pathological β-sheet-rich PrPSc conformation2,3. Self-propagating protein aggregates have been also identified in fungi, which share important characteristics with mammalian prions4,5. Additionally, mammalian prions are capable of moving from cell-to-cell and infect naïve cells6,7.
While PMDs other than TSEs are not infectious, they share a common pathogenic principle with prion diseases8,9. Although the proteins linked to each of the PMDs are not related in structure or function, they all form aggregates via a crystallization-like process called nucleated or seeded polymerization; moreover proteinaceous seeds grow by recruiting their soluble isoforms2,10,11. The efficiency to self-propagate varies in vivo, depending on the intrinsic properties of the protein, which together with additional cellular factors such as molecular chaperones ultimately determine rates of aggregate nucleation, seeding, fragmentation and spreading12-15. Hence, there must exist a fine balance among these factors that allows efficient propagation of protein aggregation. This might also explain why only some amyloidogenic aggregates harbor the characteristics of a prion, and thus not all PMDs are infectious. Prions seem to represent ‘top-performers’ of a wide spectrum of self-replicating proteinaceous aggregates, which makes them a powerful tool to study PMDs8,13.
Intriguingly, the toxicity associated with disease-related aggregates often has a non cell autonomous component16,17. This means that they affect neighboring cells that do not express the corresponding gene, in contrast to a strictly cell autonomous effect, which implies that only the cells expressing the gene exhibit the specific phenotype. This was compellingly demonstrated by tissue-specific expression or knock down of the respective proteins in numerous models of neurodegenerative diseases18-26. Various mechanisms have been suggested as a basis for this non-cell autonomous toxicity in PMDs, including diminished nutrient supply, imbalance in neuronal signaling, glutamate excitotoxicity, and neuroinflammation16,27,28. In addition, a prion-like movement of disease-linked aggregates between cells might contribute to this aspect29,30. Increasing evidence suggests that protein inclusions other than prions can transmit from cell-to-cell, which may explain the characteristic spreading of pathology observed in many PMDs30-36. However, it has yet to be determined whether there is a clear causal link between intercellular movement of disease proteins and the toxic effect on neighboring cells. Therefore, a better understanding of the cellular pathways that underlie cell-to-cell transmission and non cell autonomous toxicity is necessary and essential for the development of novel therapeutics. However, many aspects of prion-like spreading and cellular factors that influence cell-to-cell transmission of misfolded proteins in metazoans are not well understood, in particular at the organismal level.
The nematode Caenorhabditis elegans has several advantages that provide the potential to discover new facets of prion-like spreading in metazoans17. It is transparent, allowing for in vivo tracking of fluorescently tagged proteins in the living organism. Furthermore, many cellular and physiological processes affected by disease are conserved from worms to human, and C. elegans is also amenable to a wide variety of genetic manipulations and molecular and biochemical analyses37-39. Exactly 959 somatic cells make up the adult hermaphrodite with a simple body plan that still has several distinct tissue types, including muscle, neurons and intestine.
To establish a new prion model in C. elegans, we chose to exogenously express the well characterized glutamine/asparagine (Q/N)-rich prion domain NM of the cytosolic yeast prion protein Sup35, since there are no known endogenous prion proteins in worms4,40. Yeast prions have been invaluable in elucidating basic mechanisms of prion replication41-44. Furthermore, NM is the first cytosolic prion-like protein that has been shown to recapitulate the full life cycle of a prion in mammalian cell culture45,46. Likewise, when expressed in C. elegans, the Sup35 prion domain adopted remarkably well to the different requirements for propagation in metazoan cells compared to yeast cells and exhibited key features of prion biology40. NM aggregation was associated with a profound toxic phenotype, including the disruption of mitochondrial integrity and appearance of various autophagy related vesicles on the cellular level, as well as embryonic and larval arrest, developmental delay, and a widespread disturbance of the protein folding environment on the organismal level. Strikingly, the prion domain exhibits cell autonomous and non cell autonomous toxicity, affecting neighboring tissues in which the transgene was not expressed. Furthermore, the vesicular transport of the prion domain within and between cells is monitored real time in vivo40.
Here we describe how to examine prion-like dissemination in C. elegans. We will explain how to monitor the intra- and intercellular transport of vesicles containing the prion domain using time-lapse fluorescence microscopy. We will emphasize the use of tissue-specific folding sensors and ubiquitously expressed stress reporters to evaluate cell autonomous and non cell autonomous effects on cellular fitness. Finally, we will describe the procedure of a recently performed genome wide RNA interference (RNAi) screen to identify new modifiers of prion-induced toxicity. In combination, these methods can help to tease apart genetic pathways involved in the intercellular movement of proteins and their non cell autonomous toxicity.
1. Monitoring Transcellular Spreading of Prion-like Proteins By In Vivo Time-lapse Imaging
NOTE: Grow C. elegans wild-type (WT) (N2) and transgenic lines according to standard methods and carefully control the cultivation temperature47.
2. Using Folding Sensors and Stress Reporters to Investigate Cell Autonomous and Non cell Autonomous Affects on Proteostasis and Toxicity
3. Genome-wide Screen for Suppression of Prion-induced Toxicity in C. elegans
Figure 4. Schematic representation of the RNAi screening protocol. See protocol section 3 for a detailed description of the individual steps.
4. Confirmation of Preliminary Screen Hits
Monitoring intercellular spreading of prion-like proteins by in vivo time-lapse imaging
Transgenic C. elegans lines expressing the prion domain are particularly well suited for the analysis of certain aspects of prion-like proteins, e.g., cell-to-cell transmission and non cell autonomous toxicity. The transparency of the animals enables tracking of fluorescently tagged proteins from within the living organism at every stage of life. Taking advantage of this, movement of prion-like proteins across cells and tissues using a fluorescence microscope are visualized. A 63X or 100X/1.4NA oil objective is necessary to resolve the vesicular structures. Importantly, the prion domain has to be tagged with mRFP, in order to remain fluorescently visible within these presumably acidic vesicles40. Yellow fluorescent protein (YFP), which is often used as a tag, is not suitable because it is quenched in a low pH environment40,55. Strain AM815 expresses RFP-tagged R2E2 (a highly aggregation-prone and toxic form of the prion domain NM) that accumulates in tubular vesicles, while the RFP tag alone remains soluble (Figure 1A and B). Using in vivo time-lapse imaging, it could be observed that these tubular vesicles are transported within and between cells40 (Figure 1C and D, Video 1 and 2) [place links to Video 1 and 2 here].
Using folding sensors and stress reporters to investigate cell autonomous and non-cell autonomous effects on proteostasis and toxicity
Prions are able to cross seed related proteins and disrupt the cellular protein folding environment. This can be revealed through the coexpression of appropriate folding sensors. Folding sensors are non-essential metastable proteins that depend on a functional proteostasis network to fold correctly. Some examples are the well-known firefly luciferase and proteins harboring temperature-sensitive (ts) mutations56-66. A disruption of proteostasis leads to misfolding of the reporter, which can be detected by visual examination of aggregates or by exposure of the respective ts mutant phenotype at permissive temperatures. Stretches of polyglutamine (polyQ) with a length at the threshold of aggregation (around 40 residues) are also often used because they exhibit an age-dependent aggregation67-69. Earlier onset of aggregation reveals a compromised folding environment. Here, we employed a prion mutant RΔ2-5 that remains soluble when expressed in C. elegans body wall muscle (BWM) cells and intestine40 (Figure 2A and C). Upon simultaneous expression with R2E2 in BWM cells, RΔ2-5 readily aggregates (Figure 2B), revealing a cell autonomous effect of R2E2 on the protein folding environment. The widespread effect on proteostasis and induction of misfolding across tissues can be evaluated by expression of the folding sensor in a distinct tissue that does not express the highly aggregation-prone form of the prion domain. R2E2 was expressed under a BWM cell specific promoter (myo-3p), whereas RΔ2-5 was expressed under an intestine-specific promoter (vha-6p). Aggregation of intestinal RΔ2-5 demonstrates that R2E2 affects protein folding in a non cell autonomous manner40 (Figure 2D). Instead of using RΔ2-5, where aggregates can only be detected at high resolution, the use of polyQ44 expressed in the intestine (Q44i) allows visualization of aggregates at low resolution under a stereomicroscope (Figure 2E and F). Furthermore, this enables a quantification of animals with aggregates, as well as a quantification of the number of aggregates.
To investigate non cell autonomous toxicity of the prion domain, we previously examined neighboring tissues that do not express R2E2 by electron microscopy and found that expression of the prion domain in muscle cells leads to a non cell autonomous disruption of mitochondrial integrity in adjacent tissues, such as the intestine40. Mitochondrial dysfunction results in increased reactive oxygen species (ROS) production and oxidative stress. A noninvasive way to visualize induction of the oxidative stress response is to use an oxidative stress inducible reporter70. The strain CF1553 expresses green fluorescent protein (GFP) under the oxidative stress inducible promoter sod-370. When crossed to animals expressing the prion domain R2E2 in BWM cells, the reporter was significantly upregulated in BWM cells as well as in numerous non-BWM tissues (Figure 3). This strain can be used to examine if certain candidate genes influence this non cell autonomous induction of the oxidative stress response. Table 1 provides a list of C. elegans strains expressing folding sensors and stress reporters that can be used analogously.
Genome-wide screen for suppression of prion-induced toxicity in C. elegans
The complex toxicity phenotype of the prion domain expressed in C. elegans is particularly interesting because of the lack of discernible toxicity observed in most in vitro prion models71. This model might reveal new pathways that can influence the toxicity associated with prions in metazoans. To address this, we have screened for genes that, when RNAi-depleted, would improve the motility defect of R2E2 transgenic animals. Control treated WT worms can be used as a positive control, although most candidate genes will not provide a full movement phenotype rescue, due to the variable RNAi penetrance and the high toxicity of the transgene. The animals were treated for 4 days with RNAi, from the first larval state L1 to day 2 of adulthood. By that time, the thrashing of untreated R2E2 animals is significantly slower than the thrashing of WT animals, which is important because we are screening for an improvement of motility. F1 progeny will be present (~L1s), but is easy to distinguish from the P0 generation. Only the P0 generation is scored. This initial screen is visual and non quantitative and therefore, a low threshold was used to identify preliminary hits, meaning that every bacterial RNAi clone that seemed to increase the thrashing of prion-expressing animals was retested in a quantitative crawling assay. Figure 4 outlines the main steps of the screen setup. The high resolution of the Falcon 4M60 camera enabled a visual screening of a quarter of the plate at a time on a computer screen, which allowed the screen to be completed more efficiently. While helpful, this is not necessary. The visual examination of worm thrashing can also be performed by using a stereoscope at low magnification (10X).
Confirmation of preliminary screen hits
High throughput screens are performed as an approach that allows analyzing genome wide effects to obtain a primary list of candidate genes that can be refined by complementary methods. Validation of the screen results is therefore essential to eliminate possible false positives, which might result from performing a screen with only one experimental sample per RNAi. R2E2 transgenic animals were fed with the RNAi clones of primary hits and measured their speed on NGM plates on the second day of adulthood, as a quantifiable readout of motility and decreased toxicity. We used the wrMTrck plugin for ImageJ, developed in our laboratory, which accurately identifies C. elegans in videos and follows their movement. The wrMTrck plugin and scripts for automated analysis are open-source and publicly available at http://www.phage.dk/plugins/wrmtrck.html, along with detailed instructions on how to use it. After downloading, copy the whole wrMTrck folder to the Plugins folder of ImageJ. This protocol describes the wrMTrck_Batch method, preferred for analyzing a large number of videos at a time, but instructions for a manual, movie by movie, analysis are also available with the plugin download. WrMTrck_Batch converts an original video (Figure 5A) into a binary format with background subtraction and each worm is assigned a track. Each of the processed videos can later be viewed with these tracks superimposed, in order to manually eliminate any artifacts that were falsely identified as worms (e. g., track 1 in Figure 5B). One effective way to eliminate background artifacts (indicated with an arrow on Figure 5B) is to carefully choose the minimum and maximum size of the objects to be identified (Figure 5C). From the different outputs of the wrMTrck plugin, we use the body lengths per second (BLPS) parameter which measures the average speed of each animal, represented by the length of each track divided by the body length of each object, thus normalizing for potential differences in size of the nematodes (Figure 5D). Using this method, we identified 35 suppressors of prion-induced toxicity (sopt) that increased the motility phenotype in R2E2 expressing C. elegans to at least 133% and up to 256% when compared to empty vector-treated worms (Figure 6). These sopt genes represent the final confirmed hits that resulted from our high throughput screening efforts.
Figure 1. The prion domain spreads between cells and tissues in C. elegans by vesicular transport. (A) Collapsed confocal z-stacks of nematodes expressing RFP in body wall muscle (BWM) cells (RFPm). (B) Collapsed confocal z-stacks of nematodes expressing the highly toxic and aggregation prone NM variant, R2E2, in BWM cells (R2E2m). Arrows indicate tubular vesicles, arrowhead indicates aggregate. (C+D) Time-lapse series of animals expressing RFPm (C) and R2E2m (D). Arrows indicate areas of vesicular movement. RFPm exhibits mostly a diffuse staining. Even when RFPm is found in vesicles, these barely move. R2E2m localizes to numerous vesicles that are transported within and between cells. Scale bars: 10 µm. The accompanying Videos 1 and 2 [place links to Video 1 and 2 here] correspond to the Figures 1C and D, respectively.
Figure 2. Folding sensors reveal the widespread effect of the prion domain on proteostasis. (A+B) Coexpression of R2E2m promotes RΔ2-5m aggregation. Confocal images of (A) RΔ2-5m control and (B) RΔ2-5m coexpressed with R2E2m in BWM cells. (C+D) Muscle expressed R2E2m promotes intestinal RΔ2-5i aggregation. (C) Confocal image of control animal expressing RΔ2-5i in intestinal cells. (D) Confocal image of animals expressing R2E2m in BWM cells and RΔ2-5i in the intestine. Scale bars: 10 µm. (E+F) R2E2m induces non cell autonomous aggregation of Q44i. (E) Representative fluorescent images of Q44i and Q44i;R2E2m animals 4 days after transferring synchronized L1 larvae onto fresh OP50-seeded NGM plates. The expression of R2E2 in BWM cells leads to an earlier onset of Q44i aggregation in the intestine. (F) Quantification of animals with aggregates (in %) on indicated days after synchronization. Error bars represent S.D. This figure has been modified from40.
Figure 3. The sod-3p::GFP transcriptional stress reporter reveals that the prion domain causes cell autonomous and non cell autonomous induction of oxidative stress. (A+B) Representative fluorescent images (using the same exposure time) of sod-3p::GFP expressing control animals (A) and sod-3p::GFP;R2E2m animals (B) on day 2 of adulthood. (C) The expression of R2E2 in BWM cells leads to an induction of the reporter not only in BWM cells (I), but also in the intestine (II), nerve cord (III), pharynx and head neurons (IV).
Figure 5. Analysis of worm movement with the wrMTrck plugin for ImageJ. (A) Example of an input movie generated by using an appropriate magnification (~10X) to monitor several animals at a time. (B) “*_labels.tif” files are an output of wrMTrck_Batch and allow for manual curation of analysis results. Arrow indicates an object that was correctly excluded from the analysis due to an appropriate choice of minimum and maximum size parameters. (C) Input window of wrMTrck_Batch, showing parameters used for this protocol. These parameters need to be adjusted according to the individual microscope and movie settings. (D) Output results from movement analysis of (B), with the BLPS column highlighted in yellow. Please click here to view a larger version of this figure.
Figure 6. R2E2 worms treated with RNAi clones that are supressors of prion-induced toxicity (sopt) show improved motility. Motility is shown as relative BLPS when compared to the negative control (empty vector). Shown are all statistically significant (student t-test) RNAi clones identified in the HTP screen. represents the hit shown as an example in Figure 5. Results are from 3 movies, with 17 <n <53 tracks per condition. Error bars represent SEM.
Table 1. C. elegans strains expressing folding sensors and stress reporters.
Strain name | genotype | Comments | Strain source |
Folding sensors | |||
transgenes | |||
AM140 | unc-54p::polyQ35::yfp | muscle-specific expression of polyQ35, age dependent aggregation | CGC or Morimoto lab |
AM141 | unc-54p::polyQ40::yfp | muscle-specific expression of polyQ40, age dependent aggregation | CGC or Morimoto lab |
AM801 | unc-54p::RΔ2-5::yfp | muscle-specific expression of nucleation incompetent prion domain | Morimoto lab |
OG412 | vha-6p::polyQ44::yfp | intestine-specific expression of polyQ44, age dependent aggregation | CGC |
AM809 | vha-6p::RΔ2-5::gfp; myo-2p::mcherry | intestine-specific expression of nucleation incompetent prion domain | Morimoto lab |
AM47 | F25B3.3p::polyQ40::cfp | neuron-specific expression of polyQ40, cell type specific aggregation | Morimoto lab |
AM982 | unc54p::luciferase::yfp | muscle-specific expression of WT firefly luciferase | Morimoto lab |
myo-3p::luciferase::gfp; rol-6 | muscle-specific expression of WT firefly luciferase | Behl lab | |
sng-1p::luciferase::gfp; rol-6 | muscle-specific expression of WT firefly luciferase | Behl lab | |
FUH55 | unc54p::FLUC::egfp; rol-6 | muscle-specific expression of WT firefly luciferase | Hartl lab |
FUH134 | unc54p::FLUCSM::egfp; rol-6 | muscle-specific expression of R188Q mutant firefly luciferase | Hartl lab |
FUH135 | unc54p::FLUCDM::egfp; rol-6 | muscle-specific expression of R188Q+R261Q double mutant firefly luciferase | Hartl lab |
FUH48 | F25B3.3p::FLUC::egfp; rol-6 | neuron-specific expression of WT firefly luciferase | Hartl lab |
FUH136 | F25B3.3p::FLUCSM::egfp; rol-6 | neuron-specific expression of R188Q mutant firefly luciferase | Hartl lab |
FUH137 | F25B3.3p::FLUCDM::egfp; rol-6 | neuron-specific expression of R188Q+R261Q double mutant firefly luciferase | Hartl lab |
ts mutants | |||
CB1402 | unc-15(e1402) I | Paramyosin(ts), temperature sensitive Unc-paralyzed, Let | CGC |
CB1157 | unc-54(e1157) I | Myosin(ts), temperature sensitive Unc | CGC |
CB1301 | unc-54(e1301) I | Myosin(ts), temperature sensitive Unc | CGC |
CB286 | unc-45(e286) III | Unc-45(ts), temperature sensitive, slow moving Unc, Egl | CGC |
HE250 | unc-52(e669su250) II | Perlecan(ts), temperature sensitive Unc, stiff paralysis | CGC |
SD551 | let-60(ga89) IV | Ras(ts), temperature sensitive Muv, Ste, Let, Lva, Osm | CGC |
CX51 | dyn-1(ky51) X | Dynamin(ts), temperature sensitive Unc | CGC |
CW152 | gas-1(fc21) X | Gas-1(ts), temperature sensitive EtOH sensitivity | CGC |
ZZ26 | unc-63(x26) I | Acetylcholine receptor(ts), temperature sensitive levamisole resistance | CGC |
stress reporters | |||
CL2070 | hsp-16.2p::gfp;rol-6 | thermal stress; UPRcyto | CGC |
TJ375 | hsp-16.2p::gfp | thermal stress; UPRcyto | CGC |
TJ3000, TJ3001 | hsp-16.2p::gfp; Cbr-unc-119(+) | thermal stress; UPRcyto | CGC |
AM446 | hsp70p::gfp;rol-6 | thermal stress; UPRcyto | Morimoto lab |
AM722 | hsp70p::mcherry; myo-2p::cfp | thermal stress; UPRcyto | Morimoto lab |
AM799 | hsp90p::gfp | thermal stress; UPRcyto | Morimoto lab |
OG497 | hsf-1p::hsf-1::gfp; Cbr-unc-119(+) | thermal stress; UPRcyto | CGC |
CF1553 | sod-3p::gfp;rol-6 | oxidative stress | CGC |
KN259 | sod-3p::gfp;rol-6 | oxidative stress | CGC |
CL2166 | gst-4p::gfp::nls | oxidative stress | CGC |
LD1171 | gcs-1p::gfp;rol-6 | oxidative stress | CGC |
LD1 | skn-1p::skn-1::gfp;rol-6 | oxidative stress | CGC |
IG274 | nlp-29p::gfp | osmotic stress | CGC |
BC20309 | mtl-1p::gfp | metal stress | Baillie lab |
BC20342 | mtl-2p::gfp | metal stress | Baillie lab |
BC20314 | elt-2p::gfp | metal stress | Baillie lab |
SJ4005 | hsp-4p::gfp | UPRER | CGC |
SJ4100 | hsp-6p::gfp | UPRmito | CGC |
SJ4058 | hsp-60p::gfp | UPRmito | CGC |
CGC: Caenorhabditis Genetics Center; ts = temperature sensitive; UPR = unfolded protein response; cyto = cytosolic; mito = mitochondrial; ER = endoplasmic reticulum.
The methods described here help to illustrate spreading and the complex cell autonomous and non cell autonomous toxicity of prion-like proteins. We recently discovered that an aggregation-prone cytosolic prion domain is taken up into membrane-bound vesicles in an autophagy related process. A specific subset of these vesicles transports the prion domain within and between cells and tissues40. The key to monitor their movement in the living animal is that the protein has to be tagged with mRFP, because only mRFP-tagged proteins were visible in these presumably acidic vesicles. Using the same approach, we are currently investigating the potential prion-like behavior of certain disease-related proteins, such as superoxide dismutase 1 (SOD1) (linked to ALS), alpha-synuclein (linked to PD) or TAR DNA-binding protein 43 (TDP-43) (linked to ALS). Although this intercellular transport within vesicles seems to be used by several other proteins, there might be additional pathways that allow spreading.
The use of well characterized folding sensors is a powerful tool to monitor effects on the cellular protein folding environment in C. elegans63-66. Here we have shown how to use the age dependent accumulation of fluorescently tagged aggregation prone proteins expressed under tissue-specific promoters to visually monitor the folding capacity of distinct tissues simultaneously. Other possibilities to study the organismal proteostasis network include the use of endogenous ts mutant proteins. These metastable proteins will function normally at the permissive temperature (15 °C), but misfold and become nonfunctional at the restrictive temperature (25 °C), exhibiting their characteristic mutant phenotype. In the presence of an aggregation prone protein or during aging, the ts mutant phenotype gets exposed even at permissive conditions63-66. Some of these ts mutants are expressed in only a subset of tissues or exhibit tissue-specific phenotypes, making these particularly useful to test for cell autonomous and non cell autonomous proteotoxic effects. For example, a ts mutant of LET-60, a member of the GTP-binding RAS protooncogene family, Ras(ts), is ubiquitously expressed, but shows tissue-specific mutant phenotypes63,66. By using this ts mutant, the effect of polyQ expansions on the cellular proteostasis was shown to be cell autonomous63. Expression of polyQ in the genetic background of ras(ts) revealed only the mutant phenotype specific to the tissue in which the protein was expressed. The exposure of multiple mutant phenotypes would indicate that a protein has non-cell autonomous proteotoxic effects.
Additionally, to test for effects on cellular stress pathways, there is a large selection of in vivo fluorescent stress reporters in C. elegans72. These are usually transcriptional reporters that express green fluorescent protein (GFP) under a stress-inducible promoter, such as hsp-16.2p or sod-3p, which report on thermal or oxidative stress, respectively72. By using appropriate reporters in combination with prion-like proteins, one can monitor a potential induction of stress responses in tissues other than the one expressing the respective transgene. One caveat, however, is that these stress reporters are often not sensitive enough to reveal subtle upregulation of a given pathway (unpublished observations).
On our quest to examine whether genes that influence the non-cell autonomous toxicity also affect protein spreading, we started by screening for genes that, when knocked down, would ameliorate movement defects. Using visual evaluation of swimming rates as a read-out, we were able to comfortably assay ~25 plates per session and perform two rounds of this protocol per week, resulting in the completion of the initial screen within 6 weeks. The confirmation of the hits, by quantitative analysis of worm movement on solid plates, in triplicates, using wrMTrck was performed in an additional 3 weeks. One caveat, however, is that by using crawling in plates instead of swimming in liquid media as a read-out one might lose certain candidate genes, as swimming and crawling are not identical motions and can be influenced by distinct pathways73. Thus, if some candidate genes cannot be confirmed on solid plates, they should be re-tested in liquid media, where swimming can be quantified by counting the thrashing frequency within a given time.
The screening protocol described here uses automated liquid handling workstations, which greatly reduces variability. The disadvantage is that an excess of fluids for the reservoir of the robots is needed, which might not always be feasible. This screening setup can be easily adapted for other screens in C. elegans that use thrashing as a read-out.This screen was not designed to directly identify genes that influence the spreading or non cell autonomous toxicity of the prion domain, as genome wide screens should be simple in design, so as to be performed in a timely manner. Motility defects can originate not only from muscular, but also from neuronal damage and from a number of other gene defects53. Thus, by using this readout, we might find not only modifiers of muscle-specific toxicity. We are currently testing the candidate genes for their effects on prion domain spreading and non cell autonomous toxicity using the methods described above. These experiments will eventually reveal whether transcellular spreading of protein aggregates is directly linked to the toxicity observed in other tissues or if cell-to-cell transmission and non cell autonomous toxicity can be uncoupled.
Taken together, the described methods can be used to examine the potential prion-like behavior of proteins at the cellular and organismal level using C. elegans.
The authors have nothing to disclose.
We thank Cindy Voisine and Yoko Shibata for helpful discussion and critical comments on the manuscript. We acknowledge the High Throughput Analysis Laboratory (HTAL) and the Biological Imaging Facility (BIF) at Northwestern University for their assistance. This work was funded by grants from the National Institutes of Health (NIGMS, NIA, NINDS), the Ellison Medical Foundation, and the Daniel F. and Ada L. Rice Foundation (to R.I.M.). C.I.N.-K. was supported by the Deutsche Forschungsgemeinschaft (KR 3726/1-1).
Reagent | |||
Nanosphere size standards 100 nm | ThermoScientific | 3100A | |
Levamisole | Sigma | L-9756 | |
IPTG | Sigma | 15502-10G | |
Ahringer RNAi library | Source BioScience LifeSciences | http://www.lifesciences.sourcebioscience .com/clone-products/non-mammalian/c-elegans/c-elegans-rnai-library/ |
|
Equipment | |||
Sorvall Legend XTR Refrigerated Centrifuge, 120VAC | ThermoScientific | 75004521 | http://www.coleparmer.com/Product/Thermo_Scientific_Sorvall_Legend_ XTR_Refrigerated_Centrifuge_120 VAC/EW-17707-60 |
96 pin replicator | Scionomix | http://www.scinomix.com/all-products/96-pin-replicator/ | |
HiGro high-capacity, incubating shaker | Digilab | http://www.digilabglobal.com/higro | |
Multidrop Combi Reagent Dispenser | Titertrek | http://groups.molbiosci.northwestern.edu/hta/titertek.htm | |
Biomek FX AP96 Automated Workstation | Beckman Coulter | http://groups.molbiosci.northwestern.edu/hta/biomek_multi.htm | |
Innova44 shaker | New Brunswick | http://www.eppendorf.com/int///index.php?sitemap=2.3&pb=d78efbc05310ec 04&action=products&contentid=1& catalognode=83389 |
|
M205 FA | Leica | http://www.leica-microsystems.com/de/produkte/stereomikroskope-makroskope/fluoreszenz/details/product/leica-m205-fa/ | |
ORCA-R2 C10600-10BDigital CCD camera | Hamamatsu | http://www.hamamatsu.com/jp/en/community/life_science_camera/product/search/C10600-10B/index.html | |
Spinning Disc AF Confocal Microscope | Leica | http://www.leica-microsystems.com/products/light-microscopes/life-science-research/fluorescence-microscopes/details/product/leica-sd-af/ | |
Falcon 4M60 camera | Teledyne Dalsa | http://www.teledynedalsa.com/imaging/products/cameras/area-scan/falcon/PT-41-04M60/ | |
Software | |||
MetaMorph Microscopy Automation & Image Analysis Software | Molecular Devices | http://www.moleculardevices.com/products/software/meta-imaging-series/metamorph.html | |
Hamamatsu SimplePCI Image Analysis Software | Meyer Instruments | http://meyerinst.com/imaging-software/hamamatsu/index.htm | |
ImageJ | NIH | http://rsbweb.nih.gov/ij/download.html | |
wrMTrck plugin for ImageJ | http://www.phage.dk/plugins/wrmtrck.html | ||
C. elegans strains | |||
N2 (WT) | Caenorhabditis Genetics Center (CGC) | http://www.cgc.cbs.umn.edu/strain.php?id=10570 | |
AM815 rmIs323[myo-3p::sup35(r2e2)::rfp] | Morimoto lab | available from our laboratory | |
See table 1 for a source for folding sensor and stress reporter strains |