|Home | About | Journals | Submit | Contact Us | Français|
Gene expression in Drosophila melanogaster changes significantly throughout life and some of these changes can be delayed by lowering ambient temperature and also by dietary restriction. These two interventions are known to slow the rate of aging as well as the accumulation of damage. It is unknown, however, whether gene expression changes that occur during development and early adult life make an animal more vulnerable to death. Here we develop a method capable of measuring the rate of programmed genetic changes during young adult life in Drosophila melanogaster and show that these changes can be delayed or accelerated in a manner that is predictive of longevity. We show that temperature shifts and dietary restriction, which slow the rate of aging in Drosophila melanogaster, extend the window of neuronal susceptibility to GRIM over-expression in a way that scales to lifespan. We propose that this susceptibility can be used to test compounds and genetic manipulations that alter the onset of senescence by changing the programmed timing of gene expression that correlates and may be causal to aging.
The idea of using the temporal pattern of specific gene expression changes to measure the rate of aging has been suggested and has in several cases correlated extremely well with percent lifespan and not chronological age (Helfand et al. 1995; Rogina et al. 1997; Rogina and Helfand 1995; Rogina and Helfand 1997). It is currently unknown whether these and other gene expression changes play some role in the causality of aging or if the rate of aging would be slowed if Drosophila could maintain their youthful levels of gene expression indefinitely.
It is intriguing to postulate that the negative aspects of aging begin only once Drosophila and other organisms undergo their full developmental program and that delaying particular genetic changes could prevent the onset of senescence and the accumulation of damage associated with it. In support of this theory, the accumulation of protein oxidation does not increase linearly with age, and shows an accumulation only later in life (Levine and Stadtman 2001). Lipid oxidation also shows a similar pattern, and the onset of this damage can be delayed by a reduction in ambient temperature and caloric intake, manipulations known to extend lifespan and delay particular genetic changes (Zheng et al. 2005b).
The number of genes in D. melanogaster that change with age has been reported to be on the order of 10% with microarray studies (Kim et al. 2005; Landis et al. 2004; Pletcher et al. 2005; Pletcher et al. 2002). Over 80% of enhancer trap lines, however, show age-dependent changes in not only temporal expression but also in the tissues they’re expressed (Seroude et al. 2002), suggesting that there are many tissue specific genetic differences that previous whole body and head and thorax microarray studies are incapable of identifying.
The use of a single gene as a biomarker of aging is problematic in that a genetic or environmental manipulation may basally alter the expression of a single gene and thus the identification of multiple genes that change expression in a way that scales to percent lifespan would be much more desirable. Another alternative would be to identify a physiological window that changes with age and also scales to lifespan.
D. Melanogaster reaches adulthood very quickly upon eclosion and can effectively mate and reproduce within 24hrs. The study of many enhancer trap lines indicates that there are many programmed genetic changes that occur during the first week of life (Seroude et al. 2002). One of these changes is the loss of endogenous caspase activity in brains (Zheng et al. 2005a). We explored the nature of the genetic changes involved in the loss of apoptosis in the brain and found that the ability to drive caspase activity in neurons is not initially lost. In this study we find that neurons retain the ability to undergo apoptosis into early adult life with an adequate pro-apoptotic stimulus but that this ability declines with age in a predictable manner with interventions known to extend lifespan. Expanding upon a previous study that found an age-dependent change in neuronal susceptibility to GRIM over-expression (Poon, 2007), we develop an assay that allows the measurement of the rate of particular genetic changes that appear to coincide and be predictive of the rate of aging. This assay will be useful for screening mutants and drugs that may extend lifespan in part by delaying the rate of gene expression changes.
Standard cornmeal food (1x food): For every 1L of food made, 8.9g of agar, 124.4g sucrose, 31.1g inactive yeast, and 53.6g of cornmeal was added, mixed and autoclaved. Food was cooled to 65°C and 2.7g tegosept, dissolved in 11.8mL of ethanol was added along with 5.3mL of propionic acid. For RU486 containing food 20ml of RU486 dissolved in etoh at either 200μM or 500μM concentration was added at temperature below 65C. In food without RU486, 20mL of etoh was added. 0.5x food was prepared similarly to 1x food except that only 50% of the yeast 15.6g and 50% of the sucrose 62.2g was added per Liter. 2x food was similar to 1x food except that twice as much yeast 62.2g and twice as much sugar 248.8g was added per Liter. 9% yeast food was prepared similarly to 1x food except that three times the yeast 93.3g was added per Liter. 18% yeast food was the same as 1x food except that 186.6g yeast was added per Liter.
The ELAV-GeneSwitch line was from Haig Keshishian (Yale University, New Haven, CT). UAS-p35 and UAS-GFP were from the Bloomington Stock center lines 5072 and 1521, respectively. UAS-DIAP1 was from Bruce Hay (California Institue of Technology, Pasadena, CA). UAS-DRONC and UAS-GRIM were from John Abrams (UT Southwestern Medical Center, Dallas, TX). All stocks were grown up and used as received.
All flies were kept in a temperature-controlled incubator at 60% humidity with 12 h on/off light cycle at 25°C in vials containing standard cornmeal medium with lightly sprinkled live yeast on top. For all data shown in the GRIM susceptibility experiments, UAS-GRIM virgins were used and crossed to GeneSwitch ELAV males.
Newly eclosed adults were collected within 24 hrs, and 25 males and 25 females were placed into individual vials with the indicated food and temperature. Multiple independent vials were set up per experiment ranging from 4 to 10. Every 1–2 days flies were passed into new vials, and the number of dead flies was counted. The data from table 2 and figure 3a was collected from an earlier study and done separately from the other experiments. It was determined later that passing vials every day often extends the median lifespan of mated flies and in all other experiments vials were passed almost daily, while in the first study flies were passed every 2 days. We also began our experiments using 200μM RU486 but decided that 500μM RU486 was a more effective dose for inducing Grim expression and death.
Total RNA was isolated by homogenizing 40–55 fly heads at the indicated date and temperature and purified using the RNeasy kit (Qiagen, Valencia, CA). 0.5–1μg of RNA was first DNase treated with DNA-free kit (Ambion, Austin, TX) and then cDNA was made using superscript SSII or SSIII (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol. QPCR was performed on an ABI 7500 Real-Time PCR machine using the ABI SYBR-Green PCR master mix (Applied Biosystems, Foster City, CA) following the manufacturer’s instructions. Each PCR reaction was done in triplicate to control for pipetting errors using 20μM of pairs of the following primers:
To obtain an estimate of relative expression levels we used the comparative cycle threshold (Ct) method with Tubulin as the endogenous reference gene.
2–3 female heads or 4 male heads were homogenized on ice in 80–100μl of assay buffer. Assay buffer was composed of the following: 10mM tris-HCI (pH 8.4), 100 mM NaCl, 1 mM MgCl2. Protease inhibitors pepstatin A, leupeptin, antipain, aprotinin and chymostatin were added immediately before use to a final concentration each of 2.5μg/mL. Samples were spun for 5 min at ~15,000g @ 4°C and 30μl of the supernatant was added to a costar white, opaque 96 well half area plate (Corning Incorporated, Corning, NY). 30μl of Glo reagent from the Caspase-Glo 3/7 Assay kit (Promega, Madison, WI) was added to each sample and mixed and incubated for 45 min to 1hr before luminescence was read with a BioTek Synergy HT plate reader (BioTek Instruments, Winooski, VT). The remaining supernatant was used to quantify protein/μl using a Bio-rad protein assay (Bio-Rad Laboratories, Hercules, CA) with Bovine Gamma Globulin (Pierce, Rockford, IL) as a standard measuring 595 absorbance with a Fisherbrand flat bottom clear 96 Well Plate.
Single fly heads, bodies, or whole flies were homogenized in 80μl of assay buffer and spun for 5 min at 15000g @ 4C. 50μl of supernatant was used to add to a costar 96 well half area black opaque plate (Corning Incorporated, Corning, NY). Fluorescence was measured using a BioTek Synergy HT plate reader (BioTek Instruments, Winooski, VT) with (485 excitation/530 emission) filters. Protein/μl was quantified the same way as the caspase assay. For homozygous GFP; ES flies we detected ~30 fold increase in background GFP readings over buffer alone from eye pigment from the miniwhite gene. This background actually increases as the pigment changes color during the first week of life. To control for this we subtracted age matched controls not fed RU486. In contrast, bodies have ~0 background compared to buffer alone.
Statistical analyses, including log rank tests, anova, and unpaired t-tests, were performed by using the Prism suite of biostatistical software (GraphPad, San Diego, CA). Error bars on graphs represent standard error mean while QPCR error bars were calculated as 95% confidence intervals using the software analysis package of the ABI 7500 Real-Time PCR machine.
Flies that allowed the inducible targeting of the pro-apoptotic gene, GRIM, selectively to the nervous system were created by mating female UAS-GRIM virgins to males containing the GeneSwitch ELAV driver (ES) (Osterwalder et al. 2001). This driver is active primarily in neurons where the modified yeast GAL4 transcription protein is expressed and can bind to the UAS promoter and activate transgene gene expression in the presence of the ligand RU486. When 500μM RU486 (to activate GRIM gene expression) was mixed into our standard cornmeal food and fed to flies during the first week after eclosion there was a significant increase in death compared to flies not fed RU486 (Figure 1a,b, Table 1). Within 48 hours of exposure to RU486, females did not lay any eggs and many displayed a bloated abdomen phenotype, which was a phenotype occasionally found in males. By delaying the initial feeding of RU486 to day 10, female flies no longer died significantly more than controls (figure 1a) although egg laying was still prevented and the bloating of the abdomen often remained. The change in susceptibility to death correlated with a loss of caspase induction in the heads of flies fed 500μM RU486 for 24 hours compared to age-matched controls (Figure 1c): d1 = 3.0+/−0.2 p<.0001, d4 = 6.7 +/− 1.5 p=.005, d7 = 1.1 +/−0.2, d10 = 0.8+/−0.1, d13 = 0.7+/−0.1, n = 5 each. The increase in caspase activity between day 1 and 4 may reflect an initial increase in GRIM susceptibility or possibly a diminished food intake on the first day of eclosion.
The change in susceptibility to GRIM was gradual and showed individual variability in that the earlier a population of flies was fed RU486 for the first time, the more flies showed a high level of susceptibility. For example, there is a clear difference in the number of flies that die in the first 15–20 days post GRIM induction between day 4 and day 7 (Figure 1a). More flies are resistant at day 7, which suggests that susceptibility is determined by temporal genetic variation.
In contrast, males were still susceptible to GRIM induced death at day 10 (Figure 1b) and this correlated with the continued ability of GRIM induction to induce caspase activity in heads (Figure 1d). Males did, however, show a decrease in susceptibility with age as measured by the reduction in fold increase of head caspase activity (Figure 1d): d4 = 4.7 +/− 0.8 p < .003 compared to uninduced control, d13 = 2.3 +/− 0.3 p = .005 compared to uninduced control, n = 5 each and p < .03 that d13 is less induced than d4. Males also had a delayed onset of death at later ages. For example, it took only 6 days for males to start dying faster than uninduced controls when fed 500μM RU486 for the first time at day 6, compared to 12 days when induced for the first time on day 12 (data not shown). That the males are more susceptible than females may be due to a different expression pattern than females as GRIM-induced death rates are highly dependent on dose. It should be considered that RU486 and/or this particular over-expression system could enhance sensitivity to Grim and a different system is needed to confirm this that this is not the case.
It was previously reported that there is a dramatic decrease in endogenous caspase activity in the brain upon eclosion (Zheng et al. 2005a). We also found that this decrease was extremely rapid and correlated with the still developing morphology of the adult fly. The endogenous caspase activity (in arbitrary units) of w1118 flies dropped significantly in only 10 hours; males 0–2 hrs old = 201.7+/−48.2, 10–22 hrs old = 29.3+/− 9.1 p < .004, and females 0–2 hrs old = 153.3 +/− 33.1, 10–22 hrs old = 18.3 +/−2.2 p < .002 n=5 each (Figure 1e).
The measured caspase activity in the brain of UAS-GRIM; ES female flies (not fed RU486) remained constant from 24 hours after emergence to at least day 13 (data not shown). We did however, detect at least a 3-fold increase in caspase activity compared to buffer alone with only two female heads and at least 10 -fold increase with 4 male heads, suggesting that there is a low but basal level of caspase activity due to the leakiness of the genetic system as endogenous caspase activity is not present in the adult fly brain (Zheng et al. 2005a).
To determine if the change in GRIM susceptibility was intrinsic to the flies or if instead a decrease in the expression of the GAL4 driver was causing an apparent change in susceptibility we used GFP as a reporter to measure ELAV-GAL4 driver expression over time. To do this homozygous UAS-GFP; ES flies were fed 500μM RU486 continuously from the first day of eclosion and sampled throughout the first 38 days of life. Arbitrary units of GFP/μg protein was measured using a plate reader in both the head (Figure 2a, top): d6 = 63.3+/−14.5, d12 = 87.3+/−5.8, d18 = 155.2+/−14.0, d24 = 139.7+/−6.4, d31 = 168.6+/−14.8, d38 = 141.0+/−37.5 (n = 4 each) d6 is significantly less p < 0.04 compared to all days except d12, and the body (Figure 2a, bottom): d6 = 4.0 +/−0.3, d12 = 7.4+/− 0.4, d18 = 9.8 +/−0.4, d24 = 9.4+/− 0.5, d31 = 10.8+/−1.1, d38 = 10.6+/−1.9. (n = 4 each) d6 is significantly less p < 0.005 compared to all days except for d12. Age-matched flies not fed RU486 were used to subtract out GFP auto-fluorescence from tissues and especially the eye pigment from the miniwhite gene. Plate reader background readings (in the absence of RU486) gave a 30-fold increase in GFP/μg protein in heads compared to bodies. This background actually increased during the first 10 days of life as the pigments darkened. In contrast, bodies did not have a significant increase in signal compared to buffer at any age.
GFP fluorescence increased greater than 2-fold from 6 days of feeding to 18 days in both the head and body and then reached a plateau. This suggested that driver expression was increasing during the first 1–2 weeks of life or that RU486 and/or GFP was accumulating. To determine this, mated female flies were fed RU486 for the first time at day 5 or day 10 for 1 week (Figure 2b). GFP expression was ~2x higher when initiated at day 10, suggesting that GAL4 expression was increasing in the brain with age: GFP/μg protein at day 5 = 72.7 +/−9.1 (n = 6), day 10 = 152.9+/− 40.8 (n = 5) p < 0.07.
To show that GFP did not remain indefinitely in the flies once expressed, female flies were removed from RU486 after 12 days of continuous induction and GFP expression was measured over time (Figure 2c). Fluorescence was significantly diminished after 6 days and was virtually gone by 12 days: d12 RU486 = 7.4+/−0.4, 6d etoh = 4.8+/−0.3, 12d etoh = 0.4+/−0.3, 26d etoh = 0 +/−0.1. (n = 4 each) Each bar is different p ≤0.0002 from every other bar except 12d etoh and 26d etoh. This result is consistent with western blot studies measuring the rate of decline of epitope tagged proteins (data not shown).
Our studies were now aimed at determining whether reduced temperature and a change in caloric intake could alter the window of neuronal GRIM susceptibility in a predictable manner that correlated with predicted lifespan. To do this, it was necessary to determine whether these two interventions themselves altered GRIM expression. It was previously shown that flies had a greater death rate with 500μM RU486 dissolved in standard cornmeal food compared to 200μM (Poon, 2007), which is consistent with greater GRIM gene expression as estimated using UAS-GFP flies as a surrogate reporter: whole body GFP/μg fluorescence (in arbitrary units) of homozygous mated female GFP; ES flies maintained on 200μM RU486 1x cornmeal food for 10 days post eclosion = 12.0 +/− 1.0 (n = 8) vs. 500μM = 23.6 +/− 2.2 (n = 8) p = .0003.
We next tested the death rates of flies fed the same concentration of RU486 on different caloric foods. The rate of death was much greater on the lower calorie food compared to the higher (Figure 3a, Table 2) p < .0001 that median lifespan of flies maintained on 0.5x with RU486 is less than 1x and 2x with RU486. To determine if this was related to a difference in intrinsic susceptibility due to increased protection from greater caloric intake or was instead related to GRIM dose (either from a difference in RU486 intake or a change in GAL4 driver expression), 200μM RU486 was fed continuously for 10 days to UAS-GFP; ES female flies on different caloric food media and GFP was measured. Flies raised on the food with the lowest caloric content had the greatest GFP expression: 0.5x food = 17.3+/−1.1, 1x food = 12.0+/−1.0, 2x food = 7.6+/−0.5 n = 8 each, p < .01 that all means are different from each other (Figure 3b), which is consistent with compensatory feeding behavior on a low yeast diet (Carvalho et al. 2005; Min et al. 2007; Min and Tatar 2006).
Temperature also strongly influenced the inducible GAL4/UAS gene expression system. UAS-GRIM; ES flies fed 500μM RU486 at 18ºC from day 1 of life did not die faster than controls not fed RU486. This is consistent with the inefficiency of the genetic expression system at low temperature (Figure 3c). 10-day GFP/μg expression of homozygous UAS-GFP; ES of female flies at 18ºC is less than 50% that of age-matched flies maintained on the same food conditions at 25ºC: 18ºC = 10.4+/−0.8, at 25ºC = 23.6+/−2.2 n = 8 each, p<.0001.
To address the differences in gene expression caused by changes in ambient temperature and caloric intake, flies were initially maintained using different conditions (food and temperature) for a number of days and then switched to a uniform food for GRIM induction (Figure 4a). This food was our standard cornmeal food and 500μM RU486.
When flies were maintained on several different initial conditions for 10 days and then switched for 10 days to the standard RU486 food, GFP expression of UAS-GFP; ES flies was similar (Figure 4b) n = 5 each. To test whether GRIM expression was similar to that of GFP, and also was not initially different immediately after switching conditions, UAS-GRIM; ES flies were maintained for either 3 or 12 days at 18 or 25ºC on standard cornmeal food without RU486 and then both were switched to 25ºC on standard cornmeal food with 500μM RU486 for 24 hrs. RNA from the heads of female flies was extracted and analyzed for fold GRIM induction and found to be increased at day 12 (consistent with the results from Figure 2b) but the prior temperature maintenance had no bearing on current GRIM induction with RU486 (Figure 4c). These results indicate that this experimental protocol was adequate for similar GRIM induction despite an initial difference in environmental conditions.
The window of susceptibility for females on standard cornmeal food at 25ºC ended by day 10, with very minimal susceptibility to death at day 9 (Table 1). We hypothesized that this change in susceptibility is due to intrinsic genetic expression changes that decrease susceptibility over time and that lowering ambient temperature would slow down the rate of temporal gene expression, thus extending the chronological time the flies were susceptible (Figure 4d).
Lifespan of UAS-GRIM; ES flies in the absence of RU486 was increased dramatically at 18ºC compared to 25ºC and also reduced significantly at 29ºC (Table 3). To test if there were age dependent changes in susceptibility of flies maintained at different temperatures, flies were switched to 500μM RU486 at 25ºC at day 4 (Figure 5a), day 7 (Figure 5b), or day 10 (Figure 5c). At day 4, flies maintained at 29ºC had less susceptibility to GRIM than flies maintained at 18 or 25ºC, which were similar. By day 7, flies previously maintained at 29ºC were no longer susceptible compared to uninduced controls while flies maintained at 25ºC were now clearly less susceptible as a population compared to flies maintained at 18ºC. By day 10, only flies maintained at 18ºC for that time remained susceptible. By day 20, flies maintained at 18ºC also showed no susceptibility and actually outlived the 25ºC control flies (Figure 5d, Table 3) as expected if aging was slowed during the first 20 days of life. Flies maintained for 16 days at 18ºC did not live significantly shorter than the 25ºC control curve but did live less than the 20-day switch flies, suggesting that the susceptibility window at 18ºC ends completely between day 17 and 19 on standard cornmeal food.
The susceptibility to death by GRIM induction at different temperatures correlated with the ability of the 24hr feeding of 500μM RU486 to induce caspase activity in female heads compared to age-matched uninduced controls (Figure 5e): d4, 29ºC = 1.4+/− 0.1 p < 0.01, 25ºC = 6.7 +/− 1.5 p < 0.003, 18ºC = 5.2 +/− 0.8 p = .0004; d7, 29ºC = 0.8+/− 0.2, 25ºC = 1.1+/− 0.2, 18ºC = 5.8+/−2.8 p < 0.07; d10, 25ºC = 0.8+/− 0.1, 18ºC = 2.0 +/− 0.4 p < 0.04; d13, 25ºC = 0.7 +/− 0.1, 18ºC = 2.3 +/− 0.7 p < 0.06, n = 5 each. We confirmed that susceptibility to GRIM and the induction of caspase activity in female heads was not due to an increase in total GRIM expression (Figure 5f). Normalizing GRIM mRNA expression levels to day 3 induction of flies previously maintained at 18ºC did not demonstrate any correlation between total GRIM levels and the ability to induce caspase activity or death. The values of relative GRIM mRNA expression after 24 hrs of 500μM RU486 feeding were 18ºC d3 = 1 (+/−0.1), 25ºC d3 = 1.4+/−0.2, 29ºC d3 = 0.9 +/− 0.1, 18ºC d12 = 2.4+/− 0.2, 18ºC d20 = 2.5 +/− 0.2, 25ºC d12 = 2.3 +/− 0.1, 29ºC d6 = 1.3 +/− 0.1.
We next tested whether increasing the percentage yeast in the food medium would shorten the window of GRIM susceptibility due to an increase in caloric intake. All flies were raised on the same sugar and cornmeal diet but were supplemented with different percentages of yeast (3%, 9%, and 18%). When flies were switched after 4 days to standard cornmeal food (3% yeast) and fed 500μM RU486, flies that were previously maintained on 18% yeast showed a reduced susceptibility to GRIM, with no difference between 3% or 9% (Figure 6a).
If the switch to RU486 food was made on day 7, only flies maintained on the lowest caloric food had a reduced median lifespan (Figure 6b). These data suggest that caloric restriction is able to delay some of the same genetic changes that a reduction in temperature does but to a lesser extent.
The single control curve of 3% cornmeal food appeared adequate and did not require the additional control curves of 9% yeast switched to 3% yeast and 18% yeast switched to 3% yeast without RU486 because the time window of only 4 and 7 days was short enough to minimally effect lifespan. Evidence for this is found by examining the data in Tables 1 and and44 and comparing the median lifespans of flies maintained on 3%, 9% and 18% food before being switched to 3% food with RU486 after a time when the flies are no longer susceptible to GRIM over-expression. Median lifespans are all similar to the median lifespan of the 3% ethanol control, ~45 days when switched within 10 days.
We have shown that the ability to induce death by apoptosis in the adult nervous system with the pro-apoptotic gene GRIM diminishes with age and that this period of change correlates to percent lifespan such that the time window of inducing death can be dramatically extended by lowering ambient temperature and slightly extended by lowering the percentage yeast in the diet. An interesting finding is observed by looking at mortality curves of flies subjected to GRIM over-expression. Under a particular temperature or food condition, each successive day that GRIM induction is delayed results in fewer of the population being affected (figure 1a, ,5d)5d) suggesting that stochastic genetic changes are occurring and there is variability in the regulation of these changes that can be manipulated.
There are seven known caspases in Drosophila; these include the initiator caspases DRONC, Strica, and Dredd and the executioner caspases Dcp-1, Decay, Drice, and Damm. Dark is also a player in apoptosis and is necessary for DRONC activation (Hay and GUO 2006). It is the relative levels of DIAP1 and active caspases that regulate apoptosis in Drosophila. DIAP1 is essential for inhibiting cell death and inhibits many caspases either by driving their degradation (Chai et al. 2003; Olson et al. 2003) or rendering them inactive. GRIM, along with the proteins Hid and REAPER are upstream of these factors and bind to DIAP1 and hinder its ability to disrupt caspase activity (Hay and Guo 2006).
We did not detect an increase in DIAP1 mRNA or any decrease in the mRNA of 8 pro-apoptotic genes tested during the first 12 days of life (data not shown). This does not exclude the possibility that protein levels are unchanged during this time, as DIAP1 protein levels are highly regulated.
To determine if death could still be induced in the adult fly downstream of the GRIM-DIAP1 interaction we over-expressed DRONC in the nervous system and observed a significant although not huge increase in death compared to ethanol control at days 6 and 12: control median lifespan = 38 days (n = 48), d6 = 30 days (p < .002, n = 96), d12 = 29 days (p < .0001, n =97). Previous studies indicate that 200μM RU486 is enough to induce death at day 6 but not day 12 (Poon, 2007). That death was not more rapid or severe is consistent with endogenous DIAP1 levels being high enough to partially inhibit excess DRONC.
In support of increasing neuronal DIAP1 levels mediating the switch in GRIM susceptibility with age, the simultaneous neuronal over-expression of DIAP1 and GRIM inhibited death and caspase activity (figure 7a,b) as well as the bloating and inhibition of egg laying phenotypes observed in females when GRIM alone is expressed. The baculovirus p35 protein, which inhibits apoptosis, (Lannan et al. 2007) was also able to prevent death and other GRIM-induced phenotypes when over-expressed simultaneously with GRIM (Figure 7a,b), proving that GRIM is killing flies due to the activation of caspases.
Using our standard cornmeal food, mated female GRIM; ES flies maintained at 25ºC no longer show an increase in death rate between days 9 or 10 when fed 500μM RU486 compared to uninduced controls. This was extremely consistent and repeatable in 3 independent experiments. The reduction of ambient temperature to 18ºC extends this window over 50%, while changing the caloric composition of the food altered the window by up to 25%. Using UAS-GRIM; ES flies in combination with a large-scale drug screen may identify compounds capable of extending or accelerating the time period of GRIM susceptibility and give key insights into processes that mediate the rate of aging at the genetic level. These experiments could be done by using standard cornmeal food with particular drugs mixed in for five to ten days before switching the food media to RU486 and comparing survivorship curves to a no drug control population.
GRIM susceptibility is only one specific genetic change or group of genetic changes that happens to scale to lifespan. It is unclear whether this specific change can be used as a marker for interventions that delay the timing of gene expression in general. That this assay works in a predictable manner for both dietary restriction and temperature shifts is strong evidence that the nature of the changes associated with neuronal GRIM susceptibility can predict of the rate of aging.
The UAS-GRIM; GeneSwitch ELAV system could also be used in transposon-based mutagenesis screens to attempt to identify novel genes involved in regulating the timing of gene expression. One method of performing the screen would be to create a fly line with a piggyBac transposable element with homozygous GeneSwitch ELAV and a piggyBac transposase homozygous for GeneSwitch ELAV. These flies could then be crossed together and the resulting offspring crossed to a homozygous UAS-GRIM line so all resulting offspring would express UAS-GRIM and GeneSwitch ELAV. Flies with the transposase would be discarded via the use of a specific selectable marker and inverse PCR would be done on single female flies that were still susceptible to GRIM induced death after day 10 on standard cornmeal food at 25ºC.
We thank Peter Poon for his initial observation of a change in susceptibility to GRIM with the GeneSwitch ELAV driver, Will Lightfoot for preparing food and Stephen Helfand for providing the funding for this research, which was supported by the National Institute of Aging Grants RO1AG024353 and RO1AG025277.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.