Iron promotes protein insolubility and aging in C. elegans

Iron and other metals accumulate with age in C. elegans

Aging results in losses of various homeostatic mechanisms leading to alterations in the proteomic landscape [31, 32]. We hypothesized that the loss of proteostasis with age may be in part the result of alterations in the levels of metals previously associated with protein aggregation. In order to address this hypothesis, we first tested whether endogenous metal levels change with age. We established a metallomic profile during aging in C. elegans by measuring the elemental composition of age-synchronous mass cultures of young hermaphrodites versus older adults. Worms gradually exhibit aging characteristics such as decreased locomotion beginning around day 5 of adulthood at 25°C. The majority of the population lacked spontaneous movement by day 11 of adulthood and over 50 percent of animals were dead at this point. We sampled live animals at day 1, 5, 9 and 11 representing young, young-mid, mid- and late-life animals, avoiding transgenerational contamination by utilizing the temperature conditional sterile strain TJ1060 [spe-9(hc88) I; fer-15(b26)II]. 300,000 hermaphrodite worms were maintained for each of the age-synchronous cohorts that were subsequently sampled, snap frozen and prepared for analysis via an inductively coupled plasma-atomic emissions spectrophotometer (ICP-AES; Varian Inc). The abundance of many elements remained at steady levels between day 1 to day 11 of adulthood. However, we found dramatic alterations in levels of many elements known to be associated with human disease. Calcium, iron, copper and manganese levels were found to markedly increase from day one of adulthood in an age-dependent manner while potassium and phosphorus levels decreased (Fig. 1). These changes suggest that aging poses a vulnerability to the metal homeostatic network.

Increase in dietary iron accelerates protein aggregation in C. elegans

We noted that iron exhibited a high magnitude of accumulation with aging and we were aware of its role in neurodegenerative-associated protein aggregation. Consequently, we decided to investigate whether an increase in dietary iron could promote aging in C. elegans. In agreement with previous reports, we found that supplementation with 15 mM ferric ammonium citrate (FAC) resulted in a modest but robust reduction in mean and maximum lifespan (Fig. 2A). Iron has been shown to affect protein aggregation associated with various neurodegenerative diseases and aging in C. elegans has been strongly associated with a loss in proteostasis. In order to investigate whether the effects of dietary iron enrichment on lifespan could be via effects on altered proteostasis, we took advantage of a series of well-characterized C. elegans protein aggregation models. We first tested a temperature-conditional strain, CL4176 [smg-1 (cc546ts); dvIs27(myo-3::Aβ_3–42 let 39UTR(pAF29))]. The hall-mark protein associated with Alzheimer's disease (AD) inclusions, Aβ, accumulates in the body wall muscle when this strain is grown at 25°C, leading to paralysis of the animals. We exposed animals to a diet containing 15 mM FAC simultaneously shifting them to 25°C while scoring for the frequency of paralysis each subsequent day. We found that exposure to a diet of 15 mM FAC increased the rate of paralysis in the Aβ transgenic animal (Fig. 2B). Polyglutamine (PolyQ) aggregation is a key pathological hallmark of Huntington's disease (HD) and other related neurodegenerative disorders. We examined PolyQ aggregation in Q40::YFP transgenic animals AM141 rmIs133[P(unc-54) Q40::YFP)] upon transient exposure to 15 mM FAC from the young adult stage. We found that iron had a modest but significant effect on the numbers of polyglutamine inclusions in this second protein aggregation model (Fig. 2C, Fig. 2D).

More importantly, an iron-rich environment diet was found to adversely affect functional physiology in this strain as demonstrated by reduction in movement (Fig. 2E). We then tested a strain expressing a metastable mutant perlecan protein; HE250 [unc-52(e669su250)II]. Perlecan is the core protein of the mammalian basement membrane heparan sulfate proteoglycan and UNC-52 mutants are often used as indicators of the protein homeostatic network capacity [33]. As with the previous models, worms shifted to the restrictive temperature of 25°C become paralyzed, in this case likely due to a disruption in extracellular matrix fibers. Iron did not, however, exacerbate protein misfolding phenotype in this paradigm (data not shown). This raised the question as to whether specific protein classes are more susceptible to aggregation upon iron supplementation, and if the cellular location plays a significant role in protein vulnerability.

Various major disease-specific proteins have been found to become more prone to aggregate or change conformational structure in response to iron supplementation. However, a systematic evaluation of protein aggregation in response to elevated environmental iron exposure to date is lacking. In order to determine proteomic changes in response to iron in a non-biased manner, we undertook a biochemical approach. We and others have shown that a fraction of the C. elegans proteome becomes increasingly insoluble with age [28, 29]. The insoluble fraction of the C. elegans proteome can be separated with strong detergent buffers such as SDS, where the remaining fraction (the ‘insolublome’) can be re-solubilized by formic acid and visualized on SDS-PAGE gels. We used this approach to identify proteins vulnerable to loss of solubility upon dietary iron supplementation. Consistent with previous findings, we observed that the SDS insoluble protein fraction as assessed by SDS PAGE gel significantly increased with age. Furthermore, 7 day old worms exposed to 15 mM FAC from day 1 of adulthood had a striking increase in insoluble proteins compared with 7 day old controls (Fig. 3A). When comparing age-induced aggregation with aggregation promoted by iron supplementation, we were struck by the apparent similarity in the banding pattern on SDS-PAGE gels. Based on this, we conjectured that elevated dietary iron may result in an acceleration of aging-associated protein insolubility. To further investigate, we proteolytically digested the SDS-insoluble fractions (in an in-solution digestion protocol) and utilized mass spectrometry to first identify and then quantitate proteins within the insoluble fraction in the absence and presence of iron supplementation. Reflecting the SDS PAGE result, we identified 681 proteins for control worms and 1068 proteins for iron-treated worms, with only 30 unique proteins exclusively observed in controls, but 417 unique proteins in iron-treated counterparts (Fig. 3B, Fig. 3C, Table S1 and S2), suggesting that iron potentiates the tendency of numerous proteins to aggregate. Once these sets of aggregation-prone proteins were identified, we looked for common functional or structural features. Using the DAVID v.6.7 software (The Database for Annotation, Visualization and Integrated Discovery) [34], we found a strong similarity in the functional classes of proteins that were identified in iron-supplemented versus control cohorts. Insoluble proteins from iron-supplemented worms were found to be enriched in several functional categories (Kegg pathways) including the ribosome, TCA cycle, oxidative phosphorylation and the proteasome (Table 1A). Interestingly, we observed a strong similarity in the functional classes of aggregation-prone proteins in aged [29] and iron-treated animals, suggesting that iron supplementation and aging have very similar effects on protein insolubility.

We next asked whether a particular cellular compartment or tissue was over-represented in the iron-induced insolublome. We hypothesized that iron, based on its known functions, would induce more specific and localized protein damage compared with the general aging phenotype. However, when we examined expression patterns, we found that iron-induced aggre-gation appears widespread, affecting different tissues and different intracellular compartments, similarly to the aging phenotype (Table S2A). We also used a quantitative label-free mass spectrometry approach (Skyline MS1 Filtering) [35] to determine robust and statistically significant fold changes between control and iron treated worms. Multiple biological, process and technical replicates were used to generate a candidate list (Table 1B). The latter list of 66 changing proteins upon iron treatment was further confirmed in an independent MRM-like quantitative approach called SWATH MS2 [36]. (For all quantitative details see Table S3 and S4). We found that many proteins that become insoluble with age also aggregate in response to iron supplementation, such as several ribosomal proteins, vitellogenins and heat shock protein 60. We also found an interesting seemingly age-independent aggregation by iron supplementation for a smaller set of proteins. For example, iron-sulfur clusters are found in a variety of metalloproteins, but are best known for their role in oxidation-reduction reactions as part of the mitochondrial electron transport chain. We found iron-sulfur proteins to be enriched in the iron-insoluble fraction compared with controls. From the candidate list of robustly increased proteins in iron treated worms we also noted several transthyretin (TTR) proteins, which are known to be involved in human neurodegenerative disease.

CaEDTA reduces iron levels in C. elegans

Based on our data, iron appears to accelerate at least one important feature of normal aging, the accumulation of insoluble protein. We reasoned that if we reduced the impact of metals on insoluble protein formation, we might extend healthspan and possibly lifespan. In order to investigate the role of metal accumulation, we needed an efficient method of reducing the endogenous metal load. Metals are known to play important functions in growth, development and reproduction. After testing various metal chelators, we found that exposure to CaEDTA from the egg stage delayed development and that exposure from the 4^th larval stage delayed reproduction (Fig. 4A), suggesting that CaEDTA is readily absorbed and bioavailable in C. elegans. In order to test the ability of CaEDTA to reduce metal levels in adult worms, we exposed hermaphrodite worms to CaEDTA continuously from day 1 of adulthood and measured their metal profile by ICPaes 4 days later. We found that exposure to 2.5 mM CaEDTA significantly reduced levels of iron and, to a lesser extent, zinc (Fig. 4B).

CaEDTA is protective in models of protein aggregation

Since iron exacerbated protein aggregation phenotypes, we speculated that lowering metal levels would ameliorate this effect. We examined the Aβ expressing strain CL4176 for effects of metal chelation via CaEDTA. We s hifted animals to 25°C while simultaneously exposing them to CaEDTA at day 1 of adulthood and subsequently performed daily scoring for paralysis. By day 5 of adulthood, ov er 85% percent of animals in the control group were paralyzed whereas CaEDTA supplementation from time of temperature upshift lead to a nearly complete prevention of the aggregation-induced paralysis (Fig. 4C). When we investigated effects in the perlecan model, we found that exposure to CaEDTA from the time of temperature upshift dramatically suppressed the paralysis phenotype associated with protein misfolding (Fig. 4D). Lastly, we examined formation and toxicity of polyglutamine inclusions in the AM141 strain following CaEDTA treatment. CaEDTA treatment was found to reduce the number of aggregates in this model, as well as delaying onset of muscle dysfunction (Fig. 4E, Fig. 4F). Thus, transient exposure to the metal chelator CaEDTA suppresses pathologies associated with protein mis-folding in several protein aggregation disease models.

CaEDTA extends healthspan and lifespan in C. elegans

We next investigated the effects of CaEDTA on functional aging phenotypes. As a measure of healthspan, we quantified the number of body bends throughout adulthood. We found that CaEDTA slowed spontaneous age-related declines in movement (Fig. 5A, Fig. 5B). We next addressed whether lower metal levels leads to an increase in stress resistance. We exposed day 1 adult animals to CaEDTA and measured survival after an upshift to 35°C. Transient CaEDTA exposure was found to increase the animals' survival upon heat stress (Fig. 5C). Dietary restriction is known to suppress protein aggregation, delay reproduction and increase the lifespan of C. elegans. To test whether CaEDTA acts as a dietary restriction mimetic, we measured pharyngeal pumping after treatment. We did not find any significant change in pumping rate upon CaEDTA exposure (Suppl. Fig. 2A), indicating that CaEDTA does not alter food intake. Finally, we asked if reducing the metal load would slow aging by exposing synchronized animals to CaEDTA from day 1 of adulthood and assessing survival. We found that CaEDTA significantly increased the survival of wild-type animals (Fig. 5D).

Caenorhabditis elegans strains

The TJ1060 strain [spe-9(hc88)I; fer-15(b26)II] strain was a gift from Thomas E. Johnson (Univ. of Colorado, Boulder). The following worm strains were obtained from the Caenorhabditis Genetics Center (CGC, University of Minnesota) and cultured using standard conditions: Bristol N2 (wild-type), CB4037 [glp-1(e2141)], CF1038 [daf-16(mu86)]. For paralysis and motility experiments, the following strains were used: CL4176 [smg-1 (cc546ts); dvIs27(myo-3::Aβ_3–42 let 39UTR(pAF29))]; pRF4], HE250 [unc-52(e669su250)II], AM141 [rmIs133(P(unc-54) Q40::YFP)].

Elemental analysis

Synchronous populations of 300,000 worms were grown in mass cultures on enriched peptone plates with NA22 as the food source [56]. 300 μL of worm pellet per sample was collected on day 1, 5, 9 and 11, respectively, with isotonic buffer (HEPES, choline chloride Omnitrace metal free water), dead animals were omitted. Worms exposed to 2.5 mM CaEDTA were collected at day 5 of adulthood, after continuous exposure from day 1 of adulthood. The worms were pelleted by centrifugation and washed in isotonic buffer three times over a total of 30 minutes to clear gut content. The worm pellets were dried for 48 hours at 60°C. Samples were digested in 250 μL of 70% HNO₃. Samples were diluted to a final volume of 3.125 ml at 5% HNO₃ for ICPaes analysis with a Vista Pro inductively coupled plasma-atomic emissions spectrophotometer (ICP-AES; Varian Inc).

Adult lifespan assays

Compound preparation: Solutions of CaEDTA and FAC were prepared at 75 mM and 450 mM concentrations respectively. Stocks were sterile filtered and 100 μL of solution spotted onto 3mL NGM/OP50 E. coli plates at a final concentration of 2.5 and 15 mM respectively. Control plates were spotted with 100 μL water (CaEDTA) and 9% ammonium citrate (FAC). Plates were dried at room temperature for 24 hours and stored and 4°C for no longer than one week.

60 or more synchronous hermaphrodite worms were grown at 20°C on standard NGM/OP50 plates and transferred to CaEDTA or FAC on day 1 of adulthood. The worms were transferred every day for one week and every third day subsequently while scoring for survival. Animals that crawled off the plates or died due to internal hatching were censored. All experiments were done in the absence of FUdR.

Aggregation-associated paralysis assays

CL4176: Synchronous populations were grown at 15°C until the animals were transferred onto drugs (L4 stage FAC exposure, day 1 adult CaEDTA exposure) and simultaneously shifted to 25°C to allow for Aβ expression. The number of paralyzed animals was scored every day post temperature upshift.

HE250: Synchronous populations were grown at 15°C until the young adult stage at which point the animals were transferred onto compounds and simultaneously shifted to the permissive temperature of 25°C. Paralysis was scored 48 hours post temperature upshift.

Microscopy and quantification

Animals were paralyzed in 500 uM levamisole and mounted on 2 mM agar pads with glass coverslips. YFP expression was analyzed using an Olympus BX51 upright microscope. Quantification of inclusions was done using ImageJ^TM.

Motility assays

Movement was measured by placing individual day 2 adult animals in 10uL S-basal and allowed to recover for 20 seconds. The number of body bends was then counted for 30 seconds.

Insoluble protein extraction

TJ1060 [spe-9(hc88)I; fer-15(b26)II] temperature sensitive mutants were grown in synchronous mass cultures and exposed to compound on day one of adulthood. Samples of approximately 300 μL were collected and the worms were washed several times with S-basal. Samples were snap frozen, thawed on ice and resuspended in aqueous lysis buffer (20 mM Tris base, 100 mM NaCl, 1 mM MgCl₂, pH=7.4) with protease inhibitor cocktail COMPLETE™ (Roche Diagnostics, Mannheim, Germany). The worms were sonicated on ice and then briefly centrifuged at 3000xg to remove remaining carcasses. All samples were normalized for total protein concentration as assessed by BCA assay. Samples were centrifuged at 16,000xg in 4°C and washed in lysis buffer. This step was repeated three times. The pellet was resuspended and washed in lysis buffer containing 1% SDS and centrifuged at 16,000xg. This step was repeated three times to retain the SDS insoluble fraction. The remaining pellet was then treated with 100 μL of 70% formic acid for one hour to dissolve the SDS insoluble fraction. The acidic fractions were dried for 2 hours in a Speed-Vac at 45°C.

Mass spectrometry and protein quantification

The formic acid re-suspended insolublome pellets obtained from 7 day old worms exposed to 15 mM FAC and 7 day old controls, respectively, were denatured in a final solution of 6 M urea, and 100 mM Tris and further processed for proteolytic digestion. The protein mixture was reduced with 20 mM DTT (37°C for 1 h), and subsequently alkylated with 40 mM iodoacetamide (30 min at RT in the dark). Samples were diluted 10-fold with 100 mM Tris pH 8.0 and incubated overnight at 37°C with sequencing grade trypsin (Promega, Madison WI) added at a 1:50 enzyme:substrate ratio (wt/wt). Samples were then acidified with formic acid and desalted using HLB Oasis SPE cartridges (Waters, Milford MA). Proteolytic peptides were eluted, concentrated to near dryness by vacuum centrifugation, re-suspended and further desalted (C-18 zip-tips) for mass spectrometric analysis. Iron treated samples and control samples of 7 day old worms were each prepared in 3 separate biological replicates, and each in two independent process (sample preparation/workflow) replicates. Technical duplicates (MS injection duplicates) were acquired to assess technical variability. All samples were analyzed by reverse-phase HPLC-ESI-MS/MS using an Eksigent UltraPlus nano-LC 2D HPLC system (Dublin, CA) connected to a quadrupole time-of-flight TripleTOF 5600 mass spectrometer (AB SCIEX). Typically, mass resolution for MS1 scans and corresponding precursor ions was ~35,000 while resolution for MS2 scans and resulting fragment ions was ~15,000 (‘high sensitivity’ product ion scan mode). Specifically, samples were acquired by reverse-phase HPLC-ESI-MS/MS using an Eksigent Ultra Plus nano-LC 2D HPLC system (Dublin, CA) which was directly connected to a quadrupole time-of-flight (QqTOF) TripleTOF 5600 mass spectrometer (AB SCIEX, Concord, CAN). The auto sampler was operated in μl-pickup injection mode filling a 3 μl loop with 3 μl analyte per injection. Briefly, after injection, peptide mixtures were transferred onto the analytical C18-nanocapillary HPLC column (C18 Acclaim PepMap100, 75 μm I.D. × 15 cm, 3 μm particle size, 100 Å pore size, Dionex, Sunnyvale, CA) and eluted at a flow rate of 300 nL/min using the following gradient: at 5% solvent B in A (from 0-13 min), 5-35% solvent B in A (from 13-58 min), 35-80% solvent B in A (from 58-63 min), at 80% solvent B in A (from 63-66 min), with a total runtime of 90 min including mobile phase equilibration. Solvents were prepared as follows, mobile phase A: 2% acetonitrile/98% of 0.1% formic acid (v/v) in water, and mobile phase B: 98% acetonitrile/2% of 0.1% formic acid (v/v) in water. For selected samples, additional data sets were recorded in data-independent mode (DIA) using SWATH MS2 acquisitions for quantitative analysis. In the SWATH-MS2 acquisition, a Q1 window of 25 m/z was selected to cover the mass range of m/z 400-1000 in 24 segments (24 × 100 msec), yielding a cycle time of 3.25 sec, which includes one 250 msec MS1 scan. Data acquisition was performed in data dependent mode (DDA) on the TripleTOF 5600 to obtain MS/MS spectra for the 30 most abundant precursor ions (approx. 50 msec per MS/MS) following each survey MS1 scan (250 msec). Each sample was analyzed in 3 biological and 3 technical injection/MS replicates. For protein identification all data were searched using Protein Pilot v. 4.5 beta [57] using a false discovery rate (FDR) of 1%. For MS instrumentation details and bioinformatics database search engine specifics see Procedure S1. MS1 chromatogram based quantification was performed in Skyline 2.5 an open source software project (http://proteome.gs.washington.edu/software/skyline) as described recently in detail by Schilling et al. [35]. A candidate protein list showing a robust increase in iron-treated versus control worms was confirmed quantifying additional, MRM-like data-independent SWATH MS2 acquisitions. Mass spectrometric raw data can be accessed at ftp://ftp.buckinstitute.org:225/Iron_Insoluble_Proteins/. The processed MS data is provided in Table S1-S4 to show comprehensive lists of peptides that were identified with all their mass spectrometric information, as well as detailed MS1 Filtering and SWATH quantification results. Annotated MS/MS spectral libraries can be accessed at https://daily.panoramaweb.org/labkey/project/Gibson/Lithgow_Iron/begin.view?

Thermotolerance assays

Briefly, animals were grown at 20°C and transferred to drug plates on day one of adulthood. After 10-14 hours of exposure, the temperature was raised to 35°C and animals were scored for survival beginning after 5 hours and every 1.5 hour thereafter until all animals were dead. Data analyses for lifespan and thermotolerance assays were carried out using Prism™ (GraphPadSoftware Inc., SanDiego,USA).