Fluorodeoxyuridine Improves Caenorhabditis elegans Proteostasis Independent of Reproduction Onset

Protein homeostasis (proteostasis) networks are dynamic throughout the lifespan of an organism. During Caenorhabditis elegans adulthood, the maintenance of metastable proteins and the activation of stress responses are inversely associated with germline stem cell proliferation. Here, we employed the thymidylate synthase inhibitor 5-fluoro-2′-deoxyuridine (FUdR) to chemically inhibit reproduction, thus allowing for examination of the interplay between reproduction and somatic proteostasis. We found that treatment with FUdR modulates proteostasis decline both before and after reproduction onset, such that effective induction of the heat shock response was maintained during adulthood and that metastable temperature-sensitive mutant phenotypes were rescued under restrictive conditions. However, FUdR treatment also improved the folding capacity of germline- and gonadogenesis-defective mutants, suggesting that proteostasis modulation by FUdR is independent of germline stem cell proliferation or inhibition of reproduction. Our data, therefore, indicate that FUdR converges on alternative regulatory signals that modulate C. elegans proteostasis capacity during development and adulthood.


Introduction
The long-term health of all cells is dependent on the coordinated activity of protein quality control networks. Despite these protective networks, aging is associated with a general decline in folding capacity and increased susceptibility to protein misfolding diseases. In Caenorhabditis elegans, protein homeostasis (proteostasis) is modulated early in adulthood, such that maintenance of metastable proteins and stress-resistance decline sharply following the onset of reproduction [1][2][3][4].
To uncover the roles of GSCs in modulating somatic functions, both laser ablation and mutations were previously employed to specifically remove GSCs or inhibit GSC proliferation, respectively [5,19]. However, as these techniques are both laborintensive when used for analyzing different genetic backgrounds, a simpler approach was sought to inhibit reproduction and then examine the subsequent impact of such treatment on proteostasis. Given that the thymidylate synthase inhibitor 5-fluoro-29-deoxyuridine (FUdR) [20] prevents DNA replication and reproduction in C. elegans [21] and was shown to prevent over-proliferation of GSCs in adulthood [22], we asked whether FUdR could also affect the induction of a protective heat shock response and protect metastable proteins in adulthood.
FUdR has been customarily used to inhibit reproduction in experiments, such as lifespan assays in which minimal animal handling is desired [21,23,24]. However, recent studies have shown that FUdR significantly increases the lifespan of mutant animals, such as tub-1 and gas-1 [25,26], and affects C. elegans metabolism [27]. This suggests that FUdR can modulate different signaling pathways that play a role in determining lifespan, possibly in association with the effect of this compound on fertility, even though the lifespan of wild type individuals is not modulated by treatment with this reagent [24]. Here, we show that FUdR treatment not only modulates heat shock response activation and protein folding capacity after reproduction onset but also improves the functions of these protective pathways during development, leading to a rescue of metastable temperature-sensitive mutant phenotypes under restrictive conditions. However, modulation of these proteostasis pathways by FUdR before and after reproduction onset can be dissociated from GSC proliferation or reproduction since FUdR improved these proteostasis functions in germline-and gonadogenesis-defective mutants, suggesting that other signaling pathways can modulate proteostasis during the lifespan of C. elegans.

FUdR Negates the Sharp Decline in Thermo-resistance that Occurs Early in Adulthood
Thermo-resistance declines sharply with the onset of C. elegans reproduction [1,14]. The survival rate of wild type nematodes challenged with a prolonged heat shock (HS) on the second day of adulthood was significantly lower than the survival rates of animals at the fourth larval (L4) stage or the first day of adulthood (11.0863.2% compared to 76.762.6% and 6666.8%, respectively, p,0.005). To test whether FUdR can modify this decline in thermo-resistance after reproduction onset, HS survival rates of wild type animals that had been moved to Nematode Growth Media (NGM) plates supplemented with 100 mg/ml FUdR at the L3 stage (34 h at 25uC) were monitored. This treatment was designed to inhibit reproduction while avoiding the induction of developmental phenotypes associated with earlier exposure to FUdR [21,23,24]. When naïve wild type animals raised on FUdR were challenged with prolonged HS on the second or third day of adulthood, survival rates remained high, namely 75.865.7% and 71.963.3%, respectively ( Fig 1A). Even by day 9 of adulthood, 7165.7% of the FUdR-treated animals were able to survive this stress (data not shown). Such a shift in survival rates was not observed when animals were raised on non-proliferating bacteria, such as ampicillin-sensitive bacteria grown on ampicillin-supplemented plates [28] (Fig. S1). This suggests that FUdR-dependent changes in HS survival are not due to inhibition of bacteria growth but rather result from FUdR-mediated effects on C. elegans stress response activation.
To determine whether the increased stress survival observed in FUdR-treated animals was associated with changes in the ability to activate the HS response, we compared the induction of HS genes using a HS transcriptional reporter in which green fluorescent protein (GFP) expression is regulated by the hsp-16.2 promoter in a stress-dependent manner and then followed GFP expression patterns in wild type animals raised on regular or FUdRsupplemented plates. To complement this approach, mRNA levels of different HS genes were monitored using real-time PCR. When challenged on the first day of adulthood, GFP fluorescence was detected in various somatic tissues but was predominantly observed in intestinal cells of animals raised in the presence or absence of FUdR. FUdR-treated animals maintained their ability to induce hsp-16.2-dependent GFP expression, whereas GFP fluorescence was not detected in non-treated animals ( Fig. 1B and Fig. S2A). In agreement, the ability of wild type animals raised on regular plates to induce mRNA levels of hsp-70 and hsp-16.11 declined strongly on the second day of adulthood, while FUdRtreated animals maintained their ability to induce hsp-70 and hsp-16.11 (Fig. 1C). No significant changes were detected in the basal expression levels of HS genes in FUdR-treated animals (Fig. S2B). Likewise, induction of other stress responses, such as oxidative stress response genes (gcs-1, gst-4, and sod-3), UPR ER (hsp-4) and UPR mt (hsp-6), was not detected in FUdR-treated animals, suggesting that FUdR-mediated effects on HS response activation are not due to hormesis. These data indicate that treatment with FUdR can regulate the HS response during adulthood but does not elicit the main protein damage responses.

FUdR Treatment Improved Protein Folding before and after the Onset of Reproduction
FUdR-mediated modulation of thermo-resistance before reproduction onset (L4 and day 1 of adulthood) prompted us to probe the capacity of cellular protein folding in FUdR-treated animals, employing metastable proteins as folding sensors [1]. While proteostasis perturbations induce the misfolding of metastable proteins under permissive conditions, small molecules or genetic modulation that improve proteostasis can rescue metastable protein folding under restrictive conditions [1,[29][30][31]. We, therefore, examined whether treatment with FUdR could rescue the phenotypes of metastable proteins under conditions that induce misfolding.
We first examined the effects of FUdR on metastable temperature-sensitive (ts) mutant proteins expressed during larval development. We employed two metastable missense mutations that result in myosin mislocalization in the sarcomere under restrictive temperatures, namely a mutation in the myosin chaperone unc-45(e286) (unc-45(ts)) and a mutation in myosin B, unc-54(e1301) (unc-54(ts)). These temperature-sensitive mutations lead to embryonic lethality and movement defects, the extent of which depends on the developmental stage at which the animals are shifted to the restrictive temperature [32]. To examine how FUdR treatment affected proteostasis during development, embryos were allowed to hatch at 15uC (24 h) and were then shifted to 25uC (L1-L2 larvae) to avoid embryo lethality. These animals were subsequently moved to plates containing FUdR after 24 h at 25uC (L3 larvae) or maintained on regular NGM plates. Motility was examined on the first and second days of adulthood. As expected, 75.566.5% and 88.365.7% of the unc-45(ts) and unc-54(ts) animals, respectively, that were grown under restrictive conditions (25uC) in the absence of FUdR were paralyzed by the first day of adulthood. In contrast, treatment with FUdR resulted in a complete rescue of paralysis of both unc-45(ts) and unc-54(ts) animals, with 1.461.4% and 2.161.4% of the population being paralyzed, respectively. At the same time, the mobility of the wild type animals was unaffected (data not shown). Motility was also maintained on the second day of adulthood, with 2.362.3% and 2.261.5% of these populations being paralyzed, respectively ( Fig. 2A and 2B).
The motility decline of unc-54(ts) mutant animals is associated with a severe disruption of myofilament organization [1]. We, therefore, performed immunostaining with UNC-54-specific antibodies  to monitor myosin localization in animals raised under restrictive conditions in the presence or absence of FUdR. As expected, the myofilaments of unc-54(ts) animals shifted to 25uC were disrupted and UNC-54 was completely mislocalized, compared to wild type or unc-54(ts) animals raised at 15uC. In contrast, myofilaments of FUdR-treated animals were partially maintained, and showed reduced number of myo-filaments (Fig. 2C). These observations support our finding that FUdR treatment modulated proteostasis during development, thereby increasing both stress resistance and folding capacity.
We then examined the effects of FUdR treatment on protein folding after the onset of reproduction by monitoring a temperature-sensitive mutation in dynamin-1 dyn-1(ky51), a protein that functions in endocytosis and in neuronal synaptic vesicle recycling. This mutation leads to a rapid and reversible arrest of endocytosis when animals are shifted to 25uC [1,33]. 83.763.9% of dyn-1(ts) mutant animals that were raised at 20uC coiled around themselves when shifted to 28uC on day 2 adulthood. In contrast, animals treated with FUdR were mostly unaffected, with only 19.364% of dyn-1(ts) mutant animals showing coiling in response to this shift (Fig. 3A).
Similar behavior was observed for animals presenting a temperature-sensitive mutation in perlecan, unc-52(e669su250). This mutation leads to detachment of myo-filamants from the basal membrane in body wall muscle of adult C. elegans and results in paralysis under restrictive conditions (growth at 25uC) [1]. This mutation is in a splicing variant that is only expressed in adulthood. When animals expressing unc-52(ts) are raised at 25uC, they become paralyzed starting from day 1 of adulthood, allowing us to differentiate the effects of FUdR before and after the onset of reproduction. The motility of mutant animals expressing unc-52(ts) declined quickly, such that these animals were completely paralyzed (97.761.6%) and showed reduced number of myo-filaments by day 2 of adulthood when maintained under restrictive conditions. In contrast, only 10.364.4% of animals grown on FUdR were paralyzed, with myo-filament structure being maintained in such individuals, similar to wild type animals ( Fig. 3B and 3C). Thus, FUdR treatment can rescue the folding of metastable proteins before and after the onset of reproduction. FUdR-mediated Effects on Thermo-resistance and Protein Folding are Independent of the GSC Pathway HS survival rates of FUdR-treated animals on day 2 of adulthood were similar to those observed previously for GSC proliferation mutant glp-1(e2141ts) (glp-1) sterile animals [14]. However, unlike the glp-1 animals, the survival rates of FUdR-treated animals were also improved at L4 and day 1 of adulthood, where 94.463.3% and 91.462.1% of the animals survived, respectively, as compared to untreated animals (p,0.05) (Fig. 1A). Given that glp-1 animals did not show any effects on thermoresistance before the onset of reproduction [14] whereas FUdRtreated animals did, we examined the effects of FUdR treatment on the HS survival of glp-1 animals before the onset of reproduction. If FUdR functions by inhibiting GSC proliferation, then FUdR treatment would not be expected to affect glp-1 mutant animals. In contrast, were FUdR to function in a parallel pathway to that used for GSC inhibition, then glp-1 animals treated with FUdR should display an additive effect on thermo-tolerance upon treatment with this reagent. Treatment of glp-1 animals with FUdR resulted in a significant increase in HS survival on day 1 of adulthood, with 91.164% of the glp-1 animals treated with FUdR surviving, as compared to 70.963.5% of the non-treated animals (p,0.01) (Fig. 4A). Increased thermo-resistance was also observed when mes-1(bn7) (mes-1) individuals, corresponding to a different germline proliferation mutant, were treated with FUdR (96.162%, as compared to 68.165.6% for untreated animals, p,0.005) (Fig. 4A). This increase in survival was similar to that observed following FUdR treatment of wild type animals (Fig. 4A), suggesting that FUdR-mediated effects on proteostasis before the onset of reproduction occur independently of GSC inhibition. In agreement with this proposal, it was noted that when unc-54(ts) animals were crossed with glp-1 mutant animals, 91.862% of the unc-54(ts);glp-1 mutant progeny were paralyzed on the first day of adulthood, with FUdR treatment completely rescuing the paralysis phenotype (2.7.61.2% on day 2 of adulthood, p,0.005) of these animals (Fig. 4B).
We then examined the effects of FUdR treatment on glp-1 animals after the onset of reproduction. FUdR-treated glp-1 animals showed a small increase in survival rates (77.864%, as compared to 6666.5% for non-treated animals) when examined on day 2 of adulthood ( Fig. 5A). Prolonging the HS to 9 hours increased this difference, with 57.264% survival being noted following treatment, as compared to 35.966.4% (p,0.005) for glp-1 non-treated animals (Fig. S3). Likewise, mes-1 animals treated with FUdR showed increased thermo-resistance in the face of  prolonged HS (87.562%, compared to 62.967.9% for untreated animals, p,0.05), suggesting that the thermo-resistance improvement observed in FUdR-treated animals occurs independently of GSC inhibition (Fig. 5A).

FUdR Treatment can Improve Thermo-resistance and Protein Folding of Gonadogenesis-defective Mutants
FUdR inhibits reproduction [23]. Animals exposed to FUdR using our treatment protocol (i.e. a shift to FUdR-containing surroundings after 34 h at 25uC) showed some germline stem cell proliferation but no progeny production (Fig. S4), suggesting that other stages of reproduction might be affected by FUdR. Given that removal of the gonad also affects some aspects of somatic function [34], we reasoned that FUdR might affect proteostasis by inhibiting reproduction in a GSC-independent manner. We, therefore, examined the HS survival of gonadogenesis-defective mutants gon-2(q388) (gon-2) and asked whether proteostasis in animals that lack a reproductive system was modulated by FUdR.
Similar to wild type and glp-1 animals, the survival rates of gon-2 animals treated with FUdR were elevated on day 1 of adulthood (84.967.5%, as compared to 6567% for untreated animals, p,0.005). Likewise, the survival rates of gon-2 animals that declined on day 2 of adulthood (9.761.8%, a value similar to what is seen with wild type) were rescued by FUdR treatment (62.667.7%) (Fig. 5A). FUdR treatment also rescued the paralysis of unc-52(ts);gon-2 animals from 97.561.3% to 10.364.8% (p,0.05). In agreement, myo-filament organization in FUdRtreated animals was partially preserved even under restrictive conditions ( Fig. 5F and 5G). Thus, FUdR treatment improved protein-folding capacity in adulthood in the absence of a reproductive system, suggesting that FUdR-mediated effects on proteostasis are not necessarily related to its effects on reproduction. Indeed, a significant improvement in thermo-resistance was observed even when FUdR concentrations were reduced 10-fold to a concentration that does not inhibit reproduction (Fig. S5) [21,25]. We, therefore, propose that other signaling pathways, independent of reproduction, exist and that these can modulate proteostasis capacity during adulthood, as well as during development.

Discussion
Here, we showed that FUdR is a potent modulator of proteostasis in C. elegans. Treatment with FUdR increased HS response activation both before and after the onset of C. elegans reproduction. This was not associated with an induction of protein damage responses, such as the HSR or UPR, and resulted in a maintenance of the folding and function of temperature-sensitive mutant proteins under restrictive conditions. Moreover, we showed that FUdR-mediated actions during development and adulthood occur independently from the effects of this compound on reproduction. While GSC proliferation mutants modulated proteostasis only during adulthood [14,16,18], FUdR treatment improved proteostasis of wild type as well as germline-and gonadogenesis-defective mutant animals both before and after the onset of reproduction. These data suggest that FUdR affects proteostasis via an alternative pathway unrelated to GSC or gonadal signaling. In agreement, FUdR was found to increase the lifespan of glp-1 animals [25]. Given that nutrient availability and environmental challenges were also shown to reset stress response activation and modulate the capacity of the cellular maintenance machinery [34][35][36][37][38], we propose that a switch in quality control function can result from a combination of exogenous and endogenous signals, possibly acting at specific times point during development [39] or adulthood [1,4,40], that dictate the ability of the organism to maintain its proteome.
How then does FUdR modulate proteostasis? FUdR inhibits thymidylate synthase, [20], resulting in an imbalance of deoxyribonucleoside triphosphate pools and causing accumulation of DNA double strand breaks [41]. Recent publications have shown that induction of the DNA damage response in C. elegans germline stem cells transpires in direct response to DNA damage by inactivating checkpoint proteins, blocking programmed cell death or upon impairing nucleotide excision repair and that such damage repair leads to improved HS survival [35,36,42,43]. Signals to the soma are suggested to function via the induction of an innate immune response [35], while somatic signals acting via KRI-1 are thought to regulate germline stem cell apoptosis [42]. This in turn suggests that DNA damage responses in the germline can be linked to improved somatic proteostasis. However, we find that germline stem cells are not required for FUdR to confer somatic resistance. Indeed, FUdR treatment was able to enhance stress survival and proteostasis even in gon-2 animals that lack the entire reproductive system to a similar extent as seen in intact animals. Thus, FUdR is not likely to affect HS activation by inducing DNA damage responses in the germline. Given that ionizing radiation caused DNA damage to somatic cells of wild type and glp-1 animals but did not activate the DNA damage response and induce programmed cell death [44], we cannot exclude the possibility that FUdR-dependent DNA damage in somatic cells modulates thermo-resistance and protein folding in the soma. Figure 5. FUdR treatment modulates proteostasis of germline-and gonadogenesis-defective mutants upon transition to adulthood. (A) Age-synchronized glp-1(e2141), mes-1(bn7) or gon-2(q388) animals raised in the absence (black) or presence (gray) of FUdR were exposed to a 37uC HS for 6 h on the second day of adulthood and survival was assayed. Data represent means 6 SEM of .5 independent experiments. P values compare age-matched treated and untreated animals. (*) P,0.05, (**) P,0.01 and (n.s.) not significant. (B) The motility of agesynchronized unc-52;glp-1 animals raised in the absence (black) or presence (gray) of FUdR was examined on the second day of adulthood and the percent of paralyzed animals was scored. Data represent means 6 SEM of .4 independent experiments. P values compare age-matched treated and untreated animals. (*) P,0.05, (**) P,0.01 and (n.s.) not significant. (C) Confocal images of age-synchronized unc-52;glp-1 animals raised in the absence or presence of FUdR, and stained with anti-MYO-3 antibodies (green) and Phalloidin (red). Scale bar is 10 mm. (D) The motility of agesynchronized unc-52;mes-1 was examined as in B. (E) Confocal images of age-synchronized unc-52;mes-1 as in C. (F) The motility of age-synchronized unc-52;gon-2 was examined as in B. (G) Confocal images of age-synchronized unc-52;gon-2 as in C. Scale bar is 10 mm. doi:10.1371/journal.pone.0085964.g005 Treatment with amyloid-binding compounds, such as Thioflavin T (ThT), also enhanced cellular proteostasis during C. elegans development and adulthood. Indeed, ThT treatment could rescue the misfolding of unc-52(ts) and unc-54(ts) mutant animals grown under restrictive conditions, both before and after the onset of reproduction, similarly to what was observed here upon FUdR treatment [29]. ThT was suggested to induce stress-response signaling directly by acting as a ''stress response mimetic''. Likewise, NG-094, a hydroxylamine derivative, was shown to potentiate HSF-1 function and improve proteostasis [45]. Thus, FUdR may modulate different modes of cellular stress signaling and improve organismal stress survival and proteostasis capacity. Indeed, a metabolomics analysis of FUdR-treated animals showed that FUdR treatment had a significant effect on the profile of metabolites [27]. Given that changes in lipid composition can modulate proteostasis [16,46], it is possible that FUdR modulates different quality control systems by modulating cellular metabolism. Additional studies are required to determine which pathway or pathways modulate proteostasis in a FUdR-dependent manner. However, our data clearly demonstrate that FUdR, even at low concentrations, changes the ability of C. elegans to respond to the environment and, therefore, suggest the use of caution when employing FUdR to inhibit reproduction and maintain a synchronized population [25][26][27].

Nematodes and Growth Conditions
For a list of strains used in this work and name abbreviations, see Table S1. Nematodes were grown on NGM plates seeded with Escherichia coli OP50-1 strains at 15uC. Unless otherwise stated, embryos laid at 15uC were set on new plates at 25uC. After 34 h, the animals were transferred to control or 100 mg/ml FUdRcontaining plates, unless stated otherwise. The first day of adulthood (day 1) was set at 50 h after temperature shift, before the onset of egg laying. Animals were moved every one or two days during the reproductive period to avoid progeny contamination.

Statistical Analysis
Experiments were repeated at least 3 times and data are presented as means 6 SEM. P values were calculated by using the Wilcoxon Mann-Whitney rank sum test to compare two independent populations. Data were compared to age-matched untreated animals. For Figure 1C, data were compared to those obtained with same-treated animals on day 1 of adulthood. (*) denotes P,0.05 and (**) denotes P,0.01.

HS Treatment
A total of 15-40 age-synchronized animals grown at 25uC were used for each assay. Animals were moved to new plates, which were then sealed and placed in a 37uC bath for 90 min. Animals were frozen or fixed immediately following stress.

Thermo-resistance Assay
Animals were picked at the indicated age and transferred to a 24-well plate containing HS buffer (100 mM Tris-HCl, pH 7.4, 17 mM NaCl and 1% cholesterol supplemented with bacteria). Animals were then subjected to a 37uC HS for 6 h (or 9 h in the experiment depicted in Fig. S3). HS buffer was supplemented with SYTOX orange (Invitrogene) and animal survival was scored by monitoring dye uptake using a Leica M165 FC fluorescent stereoscope with a TXR filter. Fluorescent animals were scored as dead. .45 animals per experimental condition were scored.

RNA Levels
Total RNA was extracted from wild type animals raised in the absence or presence of FUdR that were untreated or subjected to HS (see HS treatment). RNA was extracted using the TRIzol reagent (Invitrogene). For cDNA synthesis, mRNA was reverse transcribed using the iScriptTM cDNA Synthesis Kit (Bio-Rad). Quantitative PCR was performed on a C1000 Thermal Cycler (Bio-Rad) with SsoFas EvaGreen Supermix (Bio-Rad).

GFP Stress Reporters
Animals expressing GFP under control of the hsp-16.2 promoter (phsp-16.2::GFP) [47] were subjected to HS (see HS treatment) and GFP fluorescence in the gut was monitored 18-24 h later using a Leica M165 FC fluorescent stereoscope with a GFP2 filter and capture using Leica DFC360FX fluorescent camera. Likewise, GFP fluorescence in the gut of untreated wild type animals raised in the absence or presence of FUdR was examined.

Paralysis Assay
A total of 15-40 age-synchronized animals were used for each assay. Animals were moved to clean plates containing FUdR or not every day and paralysis was scored by monitoring animal movement after 10 min. Those animals that did not move were scored as paralyzed. .45 animals were scored per experimental condition.
Coiling Assay dyn-1(ts) animals were raised at 20uC and 15-40 L3 animals were set on plates in the presence or absence of FUdR. The animals were then shifted to plates equilibrated to 28uC for 5 min and the percentage of animals that were coiled was scored. .50 animals were scored per experimental condition.

Immunostaining, Phalloidin and DAPi Staining
Immunofluorescence studies were performed as previously described [1]. Animals were stained with 28-2 anti-UNC-54 (a gift from Prof. Barral) or 5-6 anti-MYO-3 antibodies (Hybridoma Bank) [48]. Secondary Dylight 488-conjugated goat anti-mouse IgG antibodies (Jackson Immuno Research) were used to reveal primary antibody binding. Animals were also stained with Rhodamine-Phalloidin to mark actin. Animals were imaged using a Leica DM550 confocal microscope through a 6361.0 numerical aperture objective with a 488-nm or a 532-nm line for excitation. For DAPI staining (#40043, Biotium), animals were fixed as previously described [1] and were then incubated overnight at 4uC with 20 mg/ml DAPI in x1PBS. Animals were imaged using an Olympus Fluoview FV1000 confocal microscope through a 6061.0 numerical aperture objective set with a 405 nm line for excitation. Figure S1 HS Survival rates of wild type animals fed on non-proliferating ampicillin-treated bacteria. Age-synchronized wild type animals raised on FUdR-or ampicillinsupplemented plates were exposed to a 37uC HS for 6 h and survival was assayed on day 2 of adulthood. Data represent means 6 SEM of .4 independent experiments. P values compare agematched treated and untreated animals. (*) P,0.05. (TIF) Figure S2 FUdR treatment does not induce stress responses associated with damaged proteins. (A) Images of age-synchronized wild type animals expressing GFP under control of the hsp-16.2 promoter (phsp-16.2::GFP) raised in the absence or presence of FUdR on the first or second day of adulthood. (B) Quantification of mRNA levels from age-synchronized wild type animals raised in the absence (black) or presence (gray) of FUdR on the first day of adulthood. The data presented are normalized to those obtained with non-treated animals. Data represent means 6 SEM of .3 independent biological samples. P values compare age-matched treated and untreated animals. No significant difference in expression levels was observed for genes, but for hsp-4, which decreased upon FUdR treatment. (*) P,0.05. (TIF) Figure S3 FUdR can further improve HS Survival of glp-1 animals. Age-synchronized wild type or glp-1 animals raised in the absence (black) or presence (gray) of FUdR were exposed to a 37uC HS for 9 h on day 2 of adulthood and survival was assayed. Data represent means 6 SEM of .4 independent experiments. P values compare age-matched treated and untreated animals. (*) P,0.05. (TIF) Figure S4 FUdR affects GSC proliferation. Confocal images of age-synchronized day 2 adults, wild type animals raised in the absence or presence of FUdR and stained with DAPI.

Supporting Information
(TIF) Figure S5 The effects of FUdR concentrations on the ability to mount a protective HS response. (A-B) Age-synchronized wild type animals raised on plates containing different concentrations of FUdR (as indicated) were exposed to a 37uC HS for 6 h on the first (A) or second (B) day of adulthood and survival was assayed. Data represent means 6 SEM of .4 independent experiments. P values compare age-matched treated and untreated animals. (*) P,0.05 and (**) P,0.01.

(TIF)
Table S1 Strains used in this work. (PDF)