Proteostatic decline is a hallmark of aging, facilitating the onset of neurodegenerative diseases. We outline a protocol to quantifiably measure proteostasis in two different Caenorhabditis elegans tissues through heterologous expression of polyglutamine repeats fused to a fluorescent reporter. This model allows rapid in vivo genetic analysis of proteostasis.
The ability to maintain proper function and folding of the proteome (protein homeostasis) declines during normal aging, facilitating the onset of a growing number of age-associated diseases. For instance, proteins with polyglutamine expansions are prone to aggregation, as exemplified with the huntingtin protein and concomitant onset of Huntington’s disease. The age-associated deterioration of the proteome has been widely studied through the use of transgenic Caenorhabditis elegans expressing polyQ repeats fused to a yellow fluorescent protein (YFP). This polyQ::YFP transgenic animal model facilitates the direct quantification of the age-associated decline of the proteome through imaging the progressive formation of fluorescent foci (i.e., protein aggregates) and subsequent onset of locomotion defects that develop as a result of the collapse of the proteome. Further, the expression of the polyQ::YFP transgene can be driven by tissue-specific promoters, allowing the assessment of proteostasis across tissues in the context of an intact multicellular organism. This model is highly amenable to genetic analysis, thus providing an approach to quantify aging that is complementary to lifespan assays. We describe how to accurately measure polyQ::YFP foci formation within either neurons or body wall muscle during aging, and the subsequent onset of behavioral defects. Next, we highlight how these approaches can be adapted for higher throughput, and potential future applications using other emerging strategies for C. elegans genetic analysis.
Protein homeostasis (proteostasis) is defined as the cellular ability to maintain proper function and folding of the proteome. The inherent challenge to proteostasis is ensuring all proteins are properly folded and maintained in a native conformation, which is further amplified by the varied nature of protein size, amino acid composition, structural conformation, stability, turnover, expression, sub-cellular compartmentalization, and modifications1. Proteostasis is maintained through the coordinated action of a large proteostatic network, consisting of approximately 2000 unique proteins, which regulate proper synthesis, folding, trafficking, and degradation within the proteome2,3. The workhorse components of the proteostatic network are nine major families of molecular chaperones4. Every tissue and cell type preferentially utilizes specific subsets of molecular chaperones, presumably in alignment with the differing demands of distinct proteomes5.
One hallmark of normal organismal aging is the progressive decline and collapse of cellular proteostasis, which is thought to be an underlying basis for the onset and progression of a growing number of age-associated diseases. For instance, Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, and Amyotrophic Lateral Sclerosis (ALS) share a common characteristic: in each case manifestation of neurodegeneration is driven by genetic alterations that predisposes a mutant protein to aggregation (amyloid-β/Tau, α-synuclein, HTT, FUS/TBD-43/SOD-1, respectively)6,7,8,9,10. During aging, the integrity and inducibility of the proteostatic network declines, which results in the accumulation of proteotoxic aggregates that result in cellular dysfunction and neurodegeneration. Of note, protein conformational diseases are not unique to neurons, and occur across multiple tissues, as highlighted by type II diabetes, multiple myeloma, and cystic fibrosis11,12,13,14. Therefore, elucidating mechanisms capable of preserving proteostasis will facilitate the development of targeted interventions for the treatment of disease and to promote healthy aging.
The small soil nematode Caenorhabditis elegans (C. elegans) has been instrumental in discovering genes and elucidating pathways that alter proteostasis. Many components of the proteostatic network and the signal transduction pathways that regulate proteostasis are evolutionarily conserved. Furthermore, C. elegans has reduced complexity and redundancy relative to vertebrate systems, making it more amenable to genetic analysis and gene discovery. Additional advantages of C. elegans that have led to it being widely used as a model system to study proteostasis include: powerful genetic and functional genomics, a short life cycle (3 days) and lifespan (3 weeks), a compact and well-annotated genome, the availability of a wide assortment of genetic mutants, and the ease of visualizing tissue-specific changes in cell biology using fluorescent reporters.
The progressive decay of proteostasis during aging can be easily quantified in C. elegans. The Morimoto laboratory first demonstrated that a polyglutamine expansion fused to yellow fluorescent protein (polyQ::YFP) could be used to quantify proteostatic decline in C. elegans during aging15,16,17,18. YFP fusions to 35 glutamine repeats or more result in an age-associated formation of fluorescent foci along with signs of cellular pathology. Of note, this range of glutamine expansion mirrors the length of the polyglutamine tract of the Huntingtin protein at which Huntington’s Disease pathology begins to be observed in humans (typically >35 CAG repeats)19. Strains with expression of polyQ::YFP within muscle, intestinal, or neuronal cells have been utilized to confirm that the age-associated decline of proteostasis occurs across different cell and tissue types. Muscle-specific polyQ::YFP expression (i.e., unc-54p::Q35::YFP) has been the most widely used tissue-specific reporter, as accumulating fluorescent foci are easy to quantify over the first few days of adulthood using a simple fluorescent dissecting microscope (Figure 1A-1B). Additionally, animals become paralyzed during mid-life, as the proteome within the muscle collapses due to the proteotoxic effect of the reporter (Figure 1C). Similarly, the age-associated decline in neuronal proteostasis can be followed (rgef-1p::Q40::YFP) by directly quantifying foci/aggregate formation and age-associated declines in coordinated body-bends after placing animals into liquid (Figure 2).
Here, we present a detailed protocol on how to measure the age-dependent progression of protein aggregate accumulation and the associated proteotoxicity induced by the expression of polyglutamine repeats within neuronal and muscle tissue in C. elegans. We provide examples of typical results generated using these strains and methods. Further, we show how we have utilized these methods to study transcriptional regulation of the proteostatic network. We discuss additional ways these reporters can be easily integrated with other existing reagents or adapted for larger screens.
1. Preparation of reagents
2. Synchronization of C. elegans
NOTE: Choose whether to synchronize C. elegans by either alkaline hypochlorite treatment of gravid adults or egg lay.
3. Progeny production
NOTE: Steps must be taken to either prevent progeny production or to separate the synchronized starting population from their progeny. Preventing progeny production can be achieved chemically with addition of 5-Fluoro-2'-deoxyuridine (FUdR) to plates, which is described here. Some studies have reported that FUdR can alter proteostasis24,25. Alternative approaches to prevent progeny production are discussed below.
4. Measuring the decline in proteostasis in muscle tissue by using polyglutamine-expressing animals
NOTE: Two methods can be used to identify proteostasis decline in muscle cells: imaging the formation of protein aggregates during aging (4.1) and measuring the proteotoxicity these aggregates cause with age through the onset of paralysis (4.2).
5. Measuring the decline in proteostasis in neuronal tissue by using polyglutamine expressing animals.
NOTE: Two complementary methods are used to assay proteostasis decline in neurons (1) by quantifying the formation of protein aggregates (fluorescent foci) during aging and (2) by measuring age-associated decline in the neuronal proteome via a movement-based assay.
In C. elegans, the polyglutamine repeat model has been instrumental for the identification of genes that regulate the proteostatic network. For instance, we previously showed that the homeodomain interacting protein kinase (hpk-1), a transcriptional cofactor, influences proteostasis during aging by regulating expression of autophagy and molecular chaperones31. We found that loss of hpk-1, either by RNAi silencing or in hpk-1(pk1393) null mutant animals, increases the number of Q35::YFP aggregates that accumulate during aging. Day 2 adult control animals display 18.0 ± 2.7 aggregates while the hpk-1(pk1393) null mutant and hpk-1 RNAi-treated Q35::YFP animals averaged 28 ± 5.3 and 26.0 ± 5.1 aggregates, respectively (Figure 3A-D). Similarly, by day 8 of adulthood, 77–78% of hpk-1(RNAi) and hpk-1(pk1393) animals were paralyzed while 50% of control Q35::YFP animals were paralyzed (Figure 3E). In addition, HPK-1 is sufficient to regulate protein aggregate formation as ubiquitous overexpression of hpk-1 reduces the number of Q35::YFP foci in muscle tissue and protects aging animals from Q35::YFP-associated paralysis during aging (Figure 3E). Collectively, these results demonstrate that HPK-1 regulates proteostasis and highlights how the polyQ::YFP model can be utilized for reverse genetic analysis of changes in proteostasis during aging.
Figure 1: Expression of polyQ::YFP within C. elegans muscle results in progressive foci accumulation and paralysis during aging. (A) C. elegans unc-54p::Q35::YFP expression at days 1 and 4 of adulthood (upper and lower panel, respectively). Arrows indicate representative foci. (B) Quantification of fluorescent foci over the first 4 days of adulthood. Foci are resistant to FRAP15,32,33, consistent with an insoluble protein aggregate. Error bars represent standard error of the mean (SEM) (C) unc-54p::Q35::YFP animals become paralyzed during aging. Error bars represent standard error of proportion. Raw data for (B-C) is provided in Supplemental Table 1. The scale bar represents 100 μm in all panels. Please click here to view a larger version of this figure.
Figure 2: Expression of polyQ::YFP within C. elegans neurons results in progressive foci accumulation and disruption of normal body bends. (A, top) DIC image of the C. elegans anterior. The pharynx is a bi-lobed structure in the head of the animal, which is surrounded by the nerve ring, an interconnected cluster of 180 neurons. Red brackets indicate region to score for foci within head neurons. (A, middle) rgefp-1::Q40::YFP fluorescence at day 2 adulthood. Note that YFP expression is largely diffuse, with the exception of an occasional aggregate (arrow). (A, bottom) rgefp-1::Q40::YFP fluorescence at day 10 adulthood. Foci/aggregates are indicated (red arrow). (B) Quantification of fluorescent foci over the first 10 days of adulthood. Foci are resistant to FRAP15,32,33, consistent with an insoluble protein aggregate. Error bars represent standard error of the mean (C) Typical frequency of body bends of wild type and rgef-1p::Q40::YFP animals maintained at 20°C feeding on empty vector RNAi (L4440) at days 2 adulthood. Increased glutamine expansion correlates with movement defects15. Error bars represent standard error of the mean. Raw data for (B-C) is provided in Supplemental Table 1. The scale bar represents 20 μm in all panels. Please click here to view a larger version of this figure.
Figure 3: HPK-1 promotes proteostasis. (A-C) hpk-1 activity affects the accumulation of Q35::YFP foci in muscle cells. Shown are representative images of Punc-54::polyQ::YFP animals treated with (A) control RNAi or (B) hpk-1 RNAi, and (C) transgenic animals overexpressing hpk-1 (Psur-5::HPK-1::CFP). (D) Time course of polyQ::YFP foci accumulation in conjunction with: treatment with control RNAi (black circles), hpk-1 RNAi (white circles), hpk-1(pk1393) (white squares), or hpk-1 overexpression (open triangles). Data points display the mean ± standard deviation (S.D.) of at least 15 animals per biological replicate; at least 5 independent experiments were performed. (E) Time course of paralysis of Punc-54::polyQ::YFP animals in conjunction with: treatment with control RNAi (black circles), hpk-1 RNAi (white circles), hpk-1(pk1393) (white squares), or hpk-1 overexpression (open triangles). Plotted data display the results for a single representative trial. This figure is reprinted from reference31 with permission via a Creative Commons Attribution (CC BY) license. The scale bar represents 100 μm in all panels. Please click here to view a larger version of this figure.
Supplemental Table 1: Raw data of results. Table includes foci, paralysis, and body bend data from Figure 2 and Figure 3. Please click here to download this table.
Supplemental File 1: Common reagents for routine C. elegans work. Common reagents include recipes for preparing various types of plates and buffers. Please click here to download this file.
Aging is characterized by a gradual decline in proteostasis. Proteostasis is maintained by a complex system, the proteostatic network, for the coordinated, dynamic, stress-responsive control of protein folding, degradation, and translation. Why proteostasis fails in the course of aging is poorly understood, but a decaying epigenome, declining inducibility of stress responses, and loss of compensatory crosstalk all coincide with this breakdown. In C. elegans, the transcriptional inducibility of multiple forms of stress response rapidly decline within a few hours after the onset of reproduction due to the formation of repressive chromatin marks at stress loci2,34,35. Proteostasis collapse is a massive clinical problem as it underlies the development of protein misfolding diseases. Thus, having a method suitable to genetic analysis to quantify cellular proteostasis in vivo is essential to gain deeper mechanistic insight into how organisms maintain proper folding and function of the proteome.
Transgenic C. elegans with tissue-specific expression of a polyglutamine fluorescent reporter are an effective and robust method to study age-dependent decline of proteostasis. The two most prominent benefits of the polyglutamine model are: (1) tissue-specific expression combined with powerful genetics and functional genomics allows discovery and characterization of proteostasis regulators in the context of an intact multicellular organism, and (2) visualization of foci formation permits direct quantification of age-associated proteostatic decline in vivo. While in many cases a correlation exists between aggregate accumulation and increased proteotoxicity, in some instances, these two phenotypes are negatively correlated. For example, decreased insulin signaling increases lifespan and stress resistance: daf-2 mutant animals (a hypomorphic loss of function in the insulin/insulin-like growth factor 1 receptor), show an increased aggregate load while improving proteostasis capacity by the upregulation of a protective mechanism that prevents the formation of toxic aggregate species33,36,37,38,39. Thus, in some cases aggregate formation is protective by sequestering harmful protein species from the rest of the proteome. Since readouts of cellular proteotoxicity can be assessed in conjunction with aggregate formation, one can test for inverse correlations to identify genetic interactions involved in protective sequestration mechanisms. Lastly, polyglutamine fluorescent reporters can be combined with either tissue specific knockdown (e.g., classically through tissue specific RNAi or RNA hairpin expression), and more recently through tissue-specific protein degradation, such as the Tir1-auxin system40– or with tissue-specific overexpression of a gene of interest using tissue-specific reporters, thereby gaining insight into regulation of proteostasis within and across cells and tissues.
Methods to measure changes in proteostasis during aging require chronologically matched animals. Thus, it is necessary to either prevent production of offspring or to separate starting animals. In the protocol, we outline how to use FUdR to prevent progeny production, but some studies have reported that FUdR can alter proteostasis24,25. Alternatively, feminized genetic backgrounds can be used (e.g., fer-15(b26);fem-1(hc17)41). Lastly, adult animals can be moved to fresh RNAi plates away from progeny. This simplifies background considerations at the expense of throughput. Periodically moving animals to fresh food has the additional advantages of renewing exposure to dsRNA and preventing possible starvation.
Because fluorescent foci are quantified through visual observation, one must be precise and consistent in the identification of foci. C. elegans polyQ::YFP foci have been experimentally verified to be in vivo protein aggregates via fluorescence recovery after photobleaching (FRAP)15,32,33. Additional biochemical approaches can be used to assess protein aggregation42. It is important to note that by day 3 adulthood, early foci become larger and begin to accumulate what appear as satellite foci (analogous to moons around a planet). It is essential to decide before beginning whether to count these satellite foci as separate aggregates, or to be more conservative and consider the collective area as a single aggregate, and then to remain consistent when scoring throughout all experiments. We prefer the latter; it extends the dynamic range during which foci formation can be assessed, but even with these more conservative estimates, by day 4 – 5 there is an apparent plateau effect. In actuality proteostasis continues to decline, but foci are so numerous it is no longer possible to quantify the number accurately by eye. To improve reproducibility across biological trials, between individuals, and between laboratories, it is critical to define what is being scored as foci.
While we detail methods for simple reverse genetic approaches to test a limited set of conditions for changes in proteostasis using the unc-54p::Q35::YFP reporter, we have developed methods with higher throughput where changes in proteostasis can be quantified for up to 100 conditions (e.g., RNAi clones) simultaneously. This method involves the use of replica sets, which we have used extensively to quantify changes in lifespan after genetic perturbation. Briefly, independent samples derived from a large isogenic population are scored at each timepoint, rather than a single sample over time. This approach is easily adapted for assessing age-associated proteotoxic changes, such as paralysis (using the poly-Q::YFP with the muscle promoter) or body bends (via neuronal promoter). Replica set scoring in this manner entails adding liquid to the wells of a multi-well plate, which stimulates C. elegans to move. An analogous approach can be readily applied to foci formation within muscle cells. Previously, we identified 159 genes necessary for a normal (wild-type) lifespan or the increased lifespan of decreased insulin/IGF1 receptor mutant animals, and quantified changes in healthspan and lifespan. Of these, 103 gene inactivations result in a progeric phenotype, with animals showing one or more signs of premature aging43. In a secondary screen using the unc-54p::Q35::YFP reporter and a replica set scoring approach, we were able to identify approximately 50 progeric gene inactivations that produced accelerated foci formation (A.V.S. unpublished results), which subsequently facilitated focused mechanistic studies where we identified a critical role for both the Myc network of transcription factors and the transcriptional co-factor HPK-1 in maintaining proteostasis31,44. A detailed description of the replica set method can be found in23.
The methods described here for statistical analysis of polyQ foci across conditions focus on comparing conditions within each time point, however there may be cases where the trend across time is of more interest than a comparison at a single point. If it is possible to track individual animals over time, such as if animals are singled in wells of multi-well plates or microfluidic devices, a relatively simple repeated-measures ANOVA would be appropriate. However, most often animals are kept in bulk on petri plates, making such individual tracking infeasible, and thus different methods are needed. The polyQ aggregate count data is expected to be autocorrelated between time points (e.g., foci counts should only increase, making observations for a given condition not independent between time points), necessitating a statistical analysis approach which can appropriately handle correlated error, such as ARIMA.
Both foci formation and movement defects are well-defined and quantifiable molecular hallmarks of aging that are tractable to genetic and/or functional genomic analysis. For example, a forward genetic screen using C. elegans expressing Q35::YFP within muscle cells identified enhanced aggregation after loss of unc-30, a neuron-specific transcription factor that regulates the synthesis of the inhibitory neurotransmitter gamma-aminobutyric acid (GABA)32. The development of feeding-based RNAi in C. elegans also led to a period of gene discovery in proteostasis: a genome-wide RNAi screen of C. elegans expressing Q35::YFP identified approximately 340 genetic modifiers that either enhance or inhibit the proteostatic network and thus increase or decrease polyQ aggregation39,42.
While straightforward functional genomic screens revealed a core proteostatic network, these strains remain an invaluable phenotypic resource. Tissue-specific expression of polyglutamine fluorescent reporters in Caenorhabditis elegans is an ideal system in which established genetic and functional genomic tools can be applied via enhancer, suppressor, or synthetic screens to identify novel genetic interactions45,46. Cellular proteostasis, and the myriad physiological systems that operate either to preserve or to challenge overall protein homeostasis, is emerging as a complicated cell biological and endocrine picture. Proteostasis is not represented by a single biological pathway, protein complex, or organelle. Instead, proteostasis is the product of the mass action of multiple intricate and interdependent biological pathways, machines and systems, and is challenged by myriad environmental stressors. Future studies utilizing the polyQ::YFP model will be essential to unravel the complexity of how cells maintain proper function and folding of the proteome.
The authors have nothing to disclose.
We would like to thank past and present members of the Samuelson laboratory for their assistance in the refinement of this method and/or discussion that aided the development of this manuscript. Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under Award Numbers RF1AG062593 and R21AG064519. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
24 Well Culture Plates | Greiner Bio-One | #662102 | |
2 mL 96-well plates | Greiner Bio-One | #780286 | |
600 µL 96-well plates | Greiner Bio-One | #786261 | |
96-pin plate replicator | Nunc | 250520 | |
Air-permeable plate seal | VWR | 60941-086 | |
bacteriological agar | Affymetrix/USB | 10906 | |
bacto-peptone | VWR | 90000-368 | |
C. elegans RNAi clone library in HT115 bacteria- Ahringer | Source Bioscience | C. elegans RNAi Collection (Ahringer) | See also Kamath et. al, Nature 2003. |
C. elegans RNAi clone library in HT115 bacteria- Vidal | Source Bioscience | C. elegans ORF-RNAi Resource (Vidal) | See also Rual et. al, Genome Research 2004. This library is also available from Dharmacon. |
FuDR (5-Fluoro-2'-deoxyuridine) | Alfa Aesar | L16497 | |
Glass microscope cover slips | VWR | 48404-455 | |
Glass microscope slides | VWR | 160004-422 | |
IPTG (isopropyl beta-D-1-thigalactopyranoside) | Gold Bio | 12481C100 | |
Retangular non-treated single-well plate, 128x86mm | Thermo-Fisher | 242811 | |
Sodium Azide, CAS #26628-22-8 | Sigma-Aldrich | S2002 | |
Zeiss Axio Imager M2m microscope with AxioVision v4.8.2.0 software | Zeiss | unknown | |
Zeiss StemiSV11 M2 Bio Quad microscope | Zeiss | unknown |